The map does not zoom out past Finland. Show European statistics?
Switch to European statistics?
The map zooms out to Europe and the areas become countries. The figures come from Eurostat and national statistics for 2021–2025 and are not directly comparable with the Finnish municipal map.
Immigration as a cartogram
Immigration in charts
The charts run from demographic change to economic effects, crime, migration and the asylum system, and societal effects. Under each chart's title is a short finding from its figures. The table and the sources are under the chart, and the share button gives a link, an image or the data as a CSV file.
Source: Finnish Immigration Service (Migri) · tilastot.migri.fi · The two protection rows come from different Migri tables than the other three: international protection is positive asylum decisions and temporary protection is a permit type of its own. Both are permits granted rather than applications, but a decision can fall in a different year from the application. About a thousand permits a year are granted on other grounds, which is why the lines do not add up to the all-grounds figure.
Source: Statistics Finland · tyokay/115t and vaerak/11rs · not an employment rate, see Sources and notes
Origin groups
The charts on benefit dependency and on taxes and incomes divide people of foreign background into five groups by background country. The groups come straight from the source, kuntaluvut.fi.
Large refugee countries
Iraq, Somalia, Afghanistan, Syria
Middle East
Middle East and North Africa without the large refugee countries, Central Asia included
Eastern Europe
Eastern European countries
Western countries
EU/EEA, North America, Australia and similar
Other countries
All remaining foreign countries
The chart of the share outside the labour force groups people by citizenship, and its groups are narrower. The large refugee countries are the same four, but Eastern Europe there means Russia, Ukraine, Poland and Estonia, and Western countries Sweden, Germany and the United States. The Middle East there is the Middle East without the large refugee countries.
In the other charts' menus, the quick choice Large refugee countries picks the same four: Iraq, Somalia, Afghanistan and Syria.
Sources and notes
Knowledge base
This section collects statistics, research and source material on immigration and integration, and on their effects on crime, the economy and society more broadly. The material is intended for policy and social analysis, not for drawing conclusions about individuals or population groups. Migrants are a large and highly diverse group of people with very different backgrounds and reasons for moving. The aim is not to characterize the group as a whole, but to highlight the partially negative effects associated with certain types of immigration to the country.
Comparison sitesComparable sites and statistical sources by country. Finland is roughly 10–20 years behind
mafrance.app is a French site documenting immigration-related statistics and incidents on an interactive map, conceptually similar to this site. It illustrates how much further the same development has progressed in France. The site holds thousands of immigration-related cases, and the number of reception centres and mosques, for instance, is far higher than in Finland. Finland is clearly behind France in this development, but the statistics and cases point in the same direction.
migrationskartan.se is a Swedish site mapping immigration-related statistics in Sweden. Sweden has taken in substantially more immigrants than Finland relative to population size. The site's map marks, among other things, the locations of so-called "no-go zones" and mosques.
la.stnight.in is a site recording recent shootings, stabbings and bombings in Sweden. It originated from the US president's remark "look at what happened last night in Sweden," for which no clear "incident from yesterday" could be found at the time. The site therefore holds the kind of incidents Trump might be imagined to have meant, even though they did not happen on the day before his remark.
demografie-europa.eu is a portal providing interactive maps of demographic change and the age structure of different population groups in Germany and Sweden. The maps visualise how immigration is reshaping the demographic composition of these countries and regions.
An interactive map from the 2021 census (ONS) showing the share of the White British population (English, Welsh, Scottish, Northern Irish) by area. It shows the decline of the native population and its minority position in large cities such as London, Leicester and Birmingham.
datosinmigracion.es is a Spanish platform visualising the foreign-born population across Spain as an interactive map. Data covers 1998–2025 and can be filtered by country of origin, province, gender, and age group. Spain is one of Europe's fastest-growing immigration countries.
Nederland in Beeld compiles official statistics on the societal impact of immigration in the Netherlands. The site presents daily-updated figures on prison populations, welfare use, crime suspects, housing allocation, and costs — sourced from CBS StatLine, IND, COA, and court databases. The Netherlands is one of Europe's best-documented examples of the consequences of mass immigration.
Statistics Norway's (SSB) interactive map of the share of people with an immigrant background by municipality. The site illustrates the regional distribution well.
Statistics Denmark's official statistics on immigrants and their descendants. Covers population trends, countries of origin, and residential distribution.
Statistics Denmark's interactive map of the share of people with an immigrant background by municipality. Illustrates the geographic concentration and uneven distribution of demographic change across Denmark.
Statistik Austria's official atlas maps the distribution of the population by citizenship across Austria. Based on official population register data, it shows how the foreign-national population is concentrated in Vienna and other urban centres.
messerinzidenz.de is a German site documenting knife attacks and blade crime across Germany on a case-by-case map. It collects incidents from news sources and visualises their geographic distribution. Knife crime in Germany has risen significantly alongside increased immigration.
Official interactive map from Germany's Federal Statistical Office (Destatis) showing the share of people with a migration background by district. Germany admitted over a million asylum seekers in 2015–2016 and has by far the largest immigrant population in the EU. The regional distribution shown here illustrates the long-term demographic consequences of large-scale migration.
demografie-europa.eu is a portal providing interactive maps of demographic change and the age structure of different population groups in Germany and Sweden. The maps visualise how immigration is reshaping the demographic composition of these countries and regions.
Demographic change
The population structure of Finland and Europe is changing faster than public debate often suggests, driven by immigration and birth-rate differences. The statistics below show the scale of the change in Finland and how it compares with the rest of Europe. The debate also often overlooks that part of the growth in the foreign-background population in fact arises in Finland and does not come from immigration alone.
Sections in this group and their figures
Births and fertilityThe total fertility rate in 2024 was 1.39 for foreign-background women and 1.23 for Finnish-background women, both far below the replacement level of 2.1. In some groups (Syria, Pakistan, Iraq, Somalia) it is above it.
The dependency ratio and ageingA hundred people of working age are not a hundred workers. Of people of Finnish background aged 20–64, 79.2 per cent were in work in 2023, against 29.3 to 47.3 per cent for the largest refugee-origin backgrounds and over 80 per cent for those who came on a work permit.
Muslim population growthScenarios for Finland: the 2.7% Muslim share (2016) would reach 4.2% by 2050 with zero migration, 11.4% in the medium scenario and 15.0% in the high-migration scenario.
Key figures
Between 2000 and 2025 the number of domestic-language speakers fell by about 75,400 while foreign-language speakers grew by about 547,000. All of Finland's 21st-century population growth has come from foreign-language speakers.
The number of speakers of Finland's domestic languages (Finnish, Swedish, Sámi) has fallen every year since 2014.
In 2015–2025 the number of domestic-language speakers fell in 264 of Finland's 308 municipalities.
Demographic change is sometimes branded a conspiracy theory. What can be called a conspiracy theory is the claim of a deliberate and planned replacement of the population with people of foreign background. The two are often conflated, however, so that the discussion of demographic change itself can be set aside.
Whatever the label, the change is there in the statistics. The number of speakers of the domestic languages (Finnish, Swedish, Sámi) has fallen every year since 2014, and over 2000–2025 it shrank by about 75,400 while the number of foreign-language speakers grew by about 547,000. Finland's 21st-century population growth has therefore come entirely from people of foreign background. The same shows in the municipalities: in 2015–2025 the number of domestic-language speakers fell in 264 of 308 municipalities, and in 22 of the 66 municipalities whose population grew, the net growth came entirely from foreign-language speakers.
There were about 37,600 people of foreign background in Finland in 1990 and about 660,800 in 2025. Most of the growth is still foreign-born, but the Finland-born second generation is growing fast and passed 100,000 in 2025.
Stacked bars: the lower part is those born abroad, the upper part those born in Finland, and the bar's full height is the foreign-background population.
People with a foreign background in Finland, by place of birth · 1990–2025
Foreign background, born abroadForeign background, born in Finland
Foreign background does not mean immigrant. Background country passes from the parents, so the group includes children born in Finland, and how large that part is says which group is turning from an immigrant population into a settled one. Of people of Somali background 40.5 % were born in Finland, of Ukrainian background 5.2 %. Two things sit behind the difference. Somali migration began in the early 1990s, while the Ukrainian-background population only reached its present size after 2022, so the Finland-born part has had different lengths of time to build up. The other is the group's fertility, which adds to that part on top of time. The total fertility rate was 2.21 for Somali background and 1.22 for Ukrainian background in 2024, and through the 2000s the Somali-background figure ran between 3.8 and 5.8.
Share of the background group born in Finland · by background country, 2025
The share reflects both when the migration happened and how many children the group has. A group that arrived long ago has had time to have children born in Finland, one that arrived recently has not, and the higher the group's fertility rate the faster its share grows. Ukrainian migration began in 2022, which is why it sits at the bottom of the list, and the Somali background tops it on both counts: the migration began in the early 1990s and the fertility rate ran between 3.8 and 5.8 through the 2000s.
Included are background countries with at least 5,000 people. In smaller groups the share swings from year to year.
The former Soviet Union and the former Yugoslavia are classes of the background-country classification rather than current states: people born before the break-up are recorded under them, which is why they appear separately.
Sweden's high reading is a different phenomenon from the rest: many children of Swedish-born parents live in Finland because of return migration.
The chart breaks population change down by language group. The annual change in domestic-language speakers turned negative in 2014 and has not returned to positive since.
Population change by language group · Finland 1991–2025
Change from previous year, people
Total population changeDomestic languages (Finnish, Swedish, Sámi)Foreign-language speakers
Details
Domestic languages = total population minus foreign-language speakers (covers Finnish, Swedish, Sámi and unknown).
The 2022–2023 spike partly reflects refugees from Ukraine entering the population register.
Mother tongue is a register entry: people who change their registered language move between groups.
Show the figures
Year
Total population change (Change from previous year, people)
Domestic languages (Finnish, Swedish, Sámi) (Change from previous year, people)
Foreign-language speakers (Change from previous year, people)
Population projection if current demographic trends continue · 2000–2100
% of population
Native backgroundForeign background
Details
2000–2025 is Statistics Finland's realised population statistics. From 2025 onward (dashed), the chart projects current fertility, mortality and net migration forward unchanged — it shows where today's trends would lead if they continued. It is not an official Statistics Finland forecast.
The calculation combines five Statistics Finland datasets: age structure by background, total fertility rate by background, the whole population's fertility age shape, the national survivorship curve, and net migration.
The migrant age distribution is based on Finnish Immigration Service (Migri) statistics on first residence permits, 2021–2025: about 18% under 14, 4% aged 14–17, 51% aged 18–34, 27% aged 35–64, and under 1% over 65.
Fertility is distributed across ages using the whole population's age shape (not background-specific), scaled to each group's own total fertility rate (native-background 1.23 and foreign-background 1.39 in 2024).
Mortality uses a single national life-expectancy curve, since Statistics Finland does not publish mortality broken down by background.
Net migration is held constant at the 2019–2023 average (about 30,900 people/year).
Statistics Finland classifies a child as Finnish-background as soon as at least one parent was born in Finland. The model does not apply this reclassification and keeps every child in its mother's background group instead, so the real trend would likely move slower than this chart shows. Data on how common mixed-background couples are does exist, but the model is not built to use it.
Because every current rate is held fixed for the whole horizon, the chart shows what continuing today's trends would imply, not a prediction of what will actually happen.
Finland's foreign-language population by language group, if current trends continue · 2000–2040
People
Other Asian languagesMiddle Eastern and North African languagesRussian and other former Soviet languages (excl. Baltic)Western European languagesOther African languagesEastern European languagesBaltic languages
Details
2000–2025 is Statistics Finland's realised population statistics (vaerak/11rl). From 2025 onward (dashed), the chart projects current fertility, mortality and migration forward unchanged — it shows where today's trends would lead if they continued. It is not an official Statistics Finland forecast.
The language grouping is the same one used in the Uusimaa projection by the cities of Helsinki, Espoo and Vantaa and the Regional Council of Uusimaa (below), so the two charts are directly comparable. The eighth group (other or unknown mother tongue, about 3% of foreign-language speakers) is included in the calculation but left off the chart.
Each group's total fertility rate is borrowed from the Uusimaa projection (corrected for the child's registered mother tongue), because fertility by language group is not published at national level: Baltic 1.08, Russian and ex-Soviet 0.89, Western Europe 1.12, Eastern Europe 1.63, Middle East and North Africa 2.12, other Africa 2.69 and other Asia 0.98. The report's declining paths for the Middle Eastern and African groups are carried through to 2040 (1.87 and 2.42).
Mortality uses a single national life-expectancy curve, since no mortality statistics are published by language.
Net migration is not taken from a migration statistic: for each language group the model computes the natural change implied by its own age structure and subtracts it from the observed change in population — the residual is net migration, so it also absorbs changes of registered mother tongue.
Net migration is held constant at the 2015–2025 average (about 25,800 people a year in total). The eleven-year average deliberately damps both the Ukraine-war spike and the 2022–2024 labour-migration surge: the Russian and ex-Soviet group's realised net migration ranged from 2,100 to 24,600 over those years.
The migrant age distribution is based on Finnish Immigration Service (Migri) statistics on first residence permits.
Age structure is published in 5-year bands. The split into single years of age uses the shape of the national foreign-background population.
Because every current rate is held fixed for the whole horizon, the chart shows what continuing today's trends would imply, not a prediction of what will actually happen.
Show the figures
Year
Other Asian languages (People)
Middle Eastern and North African languages (People)
Russian and other former Soviet languages (excl. Baltic) (People)
The cities of Espoo, Helsinki and Vantaa together with the Regional Council of Uusimaa published a forecast of Uusimaa's foreign-language population for 2022–2040 in August 2023. The forecast projects the total foreign-language population growing from about 53,000 (2000) to over 554,000 (2040). Realised statistics run through 2021. Everything from 2022 onward is the report's own projection, not measured data. The vertical line marks the report's publication year, 2023. Actual population growth in 2022–2025 has in fact already exceeded the report's projections, as immigration to Finland has been at record highs in those years.
Uusimaa's foreign-language population by language group · 2000–2040
People
Other Asian languagesMiddle Eastern and North African languagesRussian and other former Soviet languages (excl. Baltic)Western European languagesOther African languagesEastern European languagesBaltic languages
Details
Up to 2023 the figures are realised statistics. The years after that are a joint projection by the cities of Espoo, Helsinki and Vantaa with the Uusimaa Regional Council — not a Statistics Finland population forecast.
The language group is the registered mother tongue, not background country or citizenship. A bilingual person is recorded under one language only.
Show the figures
Year
Other Asian languages (People)
Middle Eastern and North African languages (People)
Russian and other former Soviet languages (excl. Baltic) (People)
The foreign-background population is markedly younger than the Finnish-background population. In the 35–39 age band, for instance, 12.2% of the foreign-background population falls in that band, against just 6.0% of the Finnish-background population. The foreign-background share would therefore keep growing to some extent even entirely without further immigration. The Finnish-background population, by contrast, is concentrated in older age bands and no longer grows on its own. Finland's population growth today rests almost entirely on immigration, and the difference in age structure only accelerates demographic change.
Age structure of the Finnish- and foreign-background population, share of each group by age band · 2025
% of group
Finnish backgroundForeign background
Details
Each group's shares sum to 100% – the curve shows the shape of the age structure, not group size.
Age bands are five years wide. The last band is 95+.
The Finland-born foreign-background generation is the only foreign-background group that loses people to migration every year. Its net migration has been negative without interruption for as long as the statistic exists, and the number leaving has risen from a few dozen a year in the 1990s to close to a thousand.
The scale is worth keeping in view. In the same year the foreign-born ran a net gain of tens of thousands, and the native population also loses people to migration in most years. The groups' age structures are far too different for these counts to yield a relative propensity to move: the table has no age dimension, and the second generation's median age is 9, so much of its migration is children moving with their parents.
Migration of the Finland-born foreign-background population · Finland 1990–2025
People per year
Immigration to FinlandEmigration from FinlandNet migration
Details
The chart covers only the Finland-born foreign-background population. The foreign-born run net migration of a different order, about 33,000 in 2025, which would not fit on the same axis.
The figures are counts, not rates against the population. The table has no age dimension, so propensity to move cannot be age-adjusted, and the groups' age structures differ sharply.
Emigration is recorded when a person ceases to be registered as permanently resident. Some return later, and the return then shows as immigration.
Statistics Finland population projection 2021 (medium scenario): the foreign-language population share is projected to grow from its current ~11–12% to approximately 18–20% by 2040. Since actual net migration in 2022–2025 significantly exceeded the medium-scenario assumptions, the real share may be higher.
Helsingin kaupunki, kaupunkitietopalvelut (Pasi Saukkonen)2024-08-15
Summary
According to senior researcher Pasi Saukkonen, at the end of 2023 Helsinki had 134,084 residents of foreign background, about one fifth of the area's population. Of these, 109,300 were themselves born abroad and 24,784 born in Finland, so the growth is still overwhelmingly first-generation immigrants. The city grew by 10,472 people in 2023, of which 10,219 came from net migration. Of Helsinki's net migration, 81% was international and 19% domestic. The largest background countries were Russia and the former Soviet Union, Estonia, Somalia, Iraq, China and Sweden, but the fastest growth in 2020–2023 came from the Philippines (+1,457), Bangladesh (+958) and Sri Lanka (+765). The Asian-origin share of the foreign-born population rose from 28% (2020) to 32% (2023), and students' share of residence permits granted from 14% (2020) to 25% (2023).
Statistics Finland's latest population projection (24 October 2024) estimates that net migration continuing at its current level would raise Finland's population to 6.5 million by 2070. For 2060 it gives 6.34 million, where the projections made between 2009 and 2021 put the same figure between 5.3 and 6.2 million. It does not break the population down by language, so the 2021 projection remains the most recent published estimate of the foreign-language share. The newer figure makes the same point from the other direction: in the projection, population growth rests on immigration rather than on births.
Iltalehti, 21 May 2025. An analysis published by the Kalevi Sorsa Foundation, written by Aapo Hiilamo, a doctor of demography. It is a stochastic population projection built on the model developed by University of Washington researchers for the UN's projections, which forecasts a probability distribution rather than a single figure. The seven million of the headline is the upper end of that distribution, not its central value: the median for 2050 is about 6.2 million, very close to Statistics Finland's own point projection. Hiilamo's central claim is that growth in the population and in the number of employed rests on migration alone, because the forecast ranges for births and deaths are far narrower than the one for migration, and that successful immigration could add more than a million people by 2050 against a scenario where net migration settles lower. The analysis cites the Ministry of Finance's estimate that a 7,500-person rise in annual net migration would cut the sustainability gap by nearly a billion euros, and notes that Statistics Finland raised its net migration assumption by 25,000 to 40,000. The same article carries the opposing figure: Sitra's spring 2025 stochastic projection puts about 24,000 a year as the best match to what has actually happened, and says that even at that level the problems of ageing are not removed but postponed by a few decades. Both are projections, not measurements.
In 2025, 14,168 foreign citizens permanently resident in Finland acquired Finnish citizenship — 2,656 more than in 2024 (+23%) and the highest number during Finland's independence. Most went to Iraqi citizens (1,960, or 14% of all citizenships granted), followed by Russian (1,494), Syrian (1,282) and Afghan (1,075) citizens. 92% of them kept their previous citizenship. Naturalisation was tightened on 1 October 2024: the required period of residence is now generally 8 years instead of the previous 5, or 5 years if the language requirement is met. The change applies only to applications submitted on or after that date, so it is barely reflected in the 2025 figures yet.
Helsingin Sanomat published a calculation suggesting the number of descendants of today's Finnish-background population could fall from about 5 million to under 1 million by 2113 at the current fertility rate (1.25). It was a thought experiment by Heikki Bergholm — not a scientific study — based on 2022 birth data (~38,600 births to the native-background population) and an unchanged 81-year life expectancy.
The population stood at 5,652,881 at the end of 2025 and grew by 16,910, almost half of the previous year's increase. There were 660,800 people of foreign background, 11.7% of the population. The second generation passed 100,000 for the first time and has doubled in 11 years. The share was highest in Uusimaa (20.1%) and lowest in South Ostrobothnia (4.9%), and the highest share among municipalities was in Vantaa (30.5%). There were 646,392 foreign-language speakers, 11.4% of the population, the largest groups Russian (102,618), Estonian (48,495) and Ukrainian (46,696).
There were 660,800 people with a foreign background at the end of 2025, or 11.7% of the population. According to calculations by Professor Matti Viren, the current growth rate would produce 11.8 million by 2060, more than the estimated 4.5 million with Finnish background.
The number of people speaking a domestic language (Finnish, Swedish, Sámi) fell below five million for the first time in 34 years: 4,999,717 people, 88% of the population. The population declined because more people died than were born, though net migration remained positive. In 2025, 18% of babies were born to foreign-language mothers. Emigration recorded from Uusimaa rose by 2,913, but the increase comes from the Finland–Estonia population-register treaty rather than a new wave of departures.
Europe
Replacementclock.org shows the European-descended share of the population in 45 countries from 1950 to 2100. The site's figures are recorded statistics up to 2025 and forecasts after that.
The foreign-born share of the population has risen rapidly in most of Western Europe over the past 15 years. In Germany and Sweden, for instance, it has grown to well over one and a half times its level in little more than a decade.
The UK is not in the Eurostat series. Its figures come from ONS population estimates (June 2024 and 2010) and are not fully methodologically comparable to the other countries.
Births and fertilityPopulation structure
Key figures
The total fertility rate in 2024 was 1.39 for foreign-background women and 1.23 for Finnish-background women, both far below the replacement level of 2.1. In some groups (Syria, Pakistan, Iraq, Somalia) it is above it.
Alongside migration, fertility is a significant factor in demographic change. Fertility has fallen steeply since the early 1990s among both Finnish-background and foreign-background women, and the gap between the groups has narrowed in the process. That combined foreign-background figure hides wide variation by background country. Women from Syria, Pakistan, Iraq, Bangladesh, Somalia and Afghanistan still exceed 2.0 children per woman, while those from China, the Philippines and Vietnam fall below the Finnish-background rate.
The difference shows up a generation later. A small first generation still produces a large second one where fertility is high.
Total fertility rate = 5 × the sum of age-specific fertility rates (births in an age band / women in that band), over ages 15–49.
The denominator is the mean of the populations for y−1 and y, because population is recorded on 31 December each year. Year-end alone would understate fertility for fast-growing groups.
The age bands do not align exactly: the lowest births band is 0–19 and the highest 45–, so children of mothers under 15 or over 49 sit in the numerator but not the denominator. The effect is small.
The method was checked against Statistics Finland's own figure: it produces 1.23 for Finnish background, matching table 13zb, and 1.41 for foreign background against their 1.39.
Year-to-year variation is large in small groups, where both births and women are few.
The figure is a single-year cross-section, not the number of children the group's women will ultimately have.
The share of newborns with a foreign-background mother has grown many times over since 1990. Over the same period total births have fallen steeply, and markedly fewer children are now born to Finnish-background mothers than at the start of the 1990s. The foreign-background share of births is therefore clearly larger than the foreign-background share of the population.
Share of live births to foreign-background mothers · Finland 1990–2024
%
Foreign-background share of mothers
Definition
The share means the number of children born to mothers of other than Finnish background as a proportion of all live births.
According to ETLA's review, the total fertility rate was down to 1.25 in 2024, the lowest in measurement history, against a replacement level of 2.07. Replacement was last exceeded in 1968. The population stood at 5,656,900 at the start of 2026 (0.34% growth on the year), and 623,949 people — 11% of permanent residents — had a foreign background at the start of 2025. Life expectancy in 2024 was 79.6 years for men and 84.8 for women, there were 1.3 million single-person households (28% of the population), and 22,000 marriages were solemnised against 33,000 in 2008. ETLA stresses that migration decides the direction of population change — and that migration is the most volatile part of any such projection.
A piece in HS's Visio section on the Belgian demographer Ron Lesthaeghe, who estimates that the number of people of Finnish background — those with a Finnish mother or father — will fall below half of the present figure by 2100. He describes a "double storm" facing Finland: a care crisis followed by a severe labour shortage. Lesthaeghe endorses the calculation by Heikki Bergholm published in HS on 25 March 2024 on how the number of people of Finnish origin and their descendants would develop at a total fertility rate of 1.25.
International
VisQuill draws a population pyramid, the age distribution of mortality and the age profile of fertility for every country from the UN World Population Prospects 2024, covering 1953–2023, and countries can be compared side by side. Its fertility measure is PASFR, which says at what age children are born rather than how many there are. It is a visualisation, not a new source. Finland's figures in this section come from Statistics Finland and are not directly comparable with the UN estimates.
The dependency ratio and ageingPopulation structureEconomy
Key figures
A hundred people of working age are not a hundred workers. Of people of Finnish background aged 20–64, 79.2 per cent were in work in 2023, against 29.3 to 47.3 per cent for the largest refugee-origin backgrounds and over 80 per cent for those who came on a work permit.
A working-age person out of work counts as a supporter in the demographic dependency ratio although economically they are a dependant. In 2025 registered unemployed jobseekers averaged 25.1 per cent of the foreign-citizen labour force and 10.9 per cent of the Finnish-citizen one.
Immigration is offered as the answer to ageing. The dependency ratio, however, counts heads: it says how many people of working age there are per hundred others, but pensions are paid out of the taxes of people in work. The question is therefore not how many arrive, but how many of them find work and what the rest cost.
Immigration does fix the dependency ratio as a number, because it adds people of working age. The site's own calculation puts the demographic ratio for 2060 at 68.7 people under 15 or over 64 per hundred aged 15 to 64, against 83.8 with no migration. It does not fix the problem itself, because a hundred people of working age are not a hundred workers. Of people of Finnish background aged 20–64, 79.2 per cent were in work in 2023, of the largest refugee-origin backgrounds 29.3 to 47.3 per cent, and of those who came on a work permit more than 80 per cent. When the same calculation divides the dependants by the people in work rather than by the people of working age, at the arrivals' present employment rate of 56 per cent migration moves the figure the other way from the head count. In 2030 there would be 106.5 dependants per hundred workers with migration and 105.8 without it, and in 2060 the figures are 133.6 and 132.0. Migration would start to ease this figure only once the arrivals' employment rate passes about 62 per cent in 2030 and 64 per cent in 2040. The refugee-origin shares fall below that threshold and the work-permit ones clear it comfortably. What strengthens the funding base is therefore migration of a particular kind rather than migration as such, and the reason for entry decides it more precisely than the country of origin.
A working-age person out of work counts as a supporter in the demographic ratio although economically they are a dependant. The figure can therefore improve while the funding base weakens. In 2025 registered unemployed jobseekers were 25.1 per cent of the foreign-citizen labour force on average and 10.9 per cent of the Finnish-citizen one. Foreign-language speakers are about 11 per cent of the population, 29.4 per cent of basic social assistance recipients and 41 per cent of the euros paid. The differences between groups are large, and they are set out in the section on benefit use.
The usual answer to this is that the second generation closes the gap. The site's own figures say it narrows but does not close. Of foreign-background people born in Finland, 54.4 per cent complete a vocational qualification against 62.4 per cent of the native population. A tertiary degree is held by 31.7 per cent of second-generation 25 to 34 year olds and by 41.3 per cent of the native population. In PISA 2025 the reading gap to second-generation pupils was 66 points. How far these gaps reduce to socio-economic background, and how far they do not, is set out in the education section.
The improvement in the head count can be read from the authority's own figures too, not only from the site's calculation. Statistics Finland publishes the same projection twice, once with its own net-migration assumption of 40,025 people a year and once as a self-sufficiency projection with no net migration. Without migration the old-age dependency ratio would be 49.4 in 2045 and with it 40.1. The level the self-sufficiency projection reaches in 2045 the main projection reaches in 2065, so migration buys about two decades. In 2070 the figure is 53.3, higher than today, because the arrivals age too. The time bought is real, but it is a different measure from the one pensions are paid out of.
The section's calculation of what holding today's ratio would take is not the authority's figures but the site's own arithmetic on the same assumptions. It can be run directly in the dependency calculator, where moving net migration and the employment rate shows which of the two moves the figure, and by how much.
The line is measured up to 2025 and projected after it. The two branches differ in net migration alone: the upper one is the self-sufficiency projection with none, the lower one the actual projection with 40,025 net arrivals a year. The gap in 2045 is 9.2 points.
Old-age dependency ratio 1990–2045: the projection and the same projection without net migration
People 65 and over per 100 of working age
Without net migration (self-sufficiency projection)Projection, net migration 40,025/yr
Details
The self-sufficiency projection is not a scenario in which migration stops. It is a calculation that isolates migration's effect by holding the fertility and mortality assumptions unchanged.
A projection is not a forecast. It carries current fertility, mortality and migration assumptions forward, and every new round has revised the fertility assumption downwards.
The line ends in 2045 because the self-sufficiency variant does not run further. The main projection's longer horizon is in the next chart.
The old-age ratio counts people 65 and over against people of working age. It says nothing about who is in work or when people retire.
Show the figures
Year
Without net migration (self-sufficiency projection) (People 65 and over per 100 of working age)
Projection, net migration 40,025/yr (People 65 and over per 100 of working age)
Every bar is from the same projection, with 40,025 net arrivals a year throughout. The old-age ratio barely moves to 2040 and then climbs steeply, because the people arriving now and working now are themselves pensioners in the 2060s. In 2070 it is 53.3, higher than today's 38.5.
The same projection to 2070: migration postpones the deterioration, it does not stop it
People 65 and over per 100 of working age
Details
The total dependency ratio, which also counts children, stays flatter over the same period (61.6 in 2025 and 71.9 in 2070). That is because there are fewer children, which is not good news for the finances: a child costs less than a pensioner and later becomes a payer.
The timing of the upturn depends on the fertility assumption. If fertility stays below the assumption, the curve rises earlier.
The retirement age rises with life expectancy, which moves part of the 65-and-over group to the working side. This chart uses a fixed age of 65, as Statistics Finland's own indicator does.
The arithmetic is direct. In 2070 the projection has 2,016,441 people aged 65 or over. For them to face as many working-age people as today, there would have to be 5,234,026 of working age — the projection has 3,781,749. The gap is 1,452,277 people, about 32,273 more every year for 45 years on top of the assumed 40,025. Even that would not hold permanently, because these arrivals age too, so the next decade would need more again.
What holding today's ratio would take
40 025The projection's net migration assumption, people per year
20Years of postponement from that assumption
1 452 277Shortfall of working-age people in 2070 to hold today's ratio
32 273Extra annual migration that would take, on top of the assumption
Details
The calculation is deliberately simple: it asks only how many working-age people would be needed to hold the old-age ratio at its 2025 level of 38.5. It does not model the children of the extra arrivals, who are later of working age, nor the arrivals moving into the 65-and-over group themselves.
The figure is an order of magnitude, not a forecast. The same arithmetic is the core of the UN's 2000 replacement migration report, and its conclusion was the same: stabilising the age structure would take migration volumes no country has proposed.
The ratio can be held down by other means than migration: retirement age, employment rate, length of working lives and productivity. This chart does not claim migration is the only lever, it measures what migration alone would have to do.
The dependency ratio counts heads, but pensions are paid out of the taxes of people in work. Of people of Finnish background aged 20–64, 79.2 percent were in work. For the largest refugee-origin backgrounds the share is 29.3–47.3 percent, so a hundred more people of working age from that group means about 29–47 more workers. The same gap runs the other way: people who came on a work permit are above 80 percent.
A hundred people of working age are not a hundred workers · employment rate, ages 20–64, 2023
% in work, ages 20–64
Details
The figure is a cross-section. Employment rises with time in the country, and for protection-based arrivals who have been here ten years it is about 60 percent, higher than shown here.
A background country is not a reason for entry. People from the same country have come for protection, for work and for family, and the reason for entry explains employment better than the country does.
The groups are small and young and contain many recent arrivals, so year-to-year variation is large.
Pick a scenario from the menu. Each line is the dependency ratio's path under one net-migration assumption, and the gap between the lines is what migration does.
Total dependency ratio 2025–2060: Statistics Finland's assumption, 40,025/yr
Scenario
The base is Statistics Finland's self-sufficiency projection, with 40,025 net arrivals a year added on top. They are aged with the authority's own mortality trend and have children here at the measured foreign-background fertility. The authority's own projection at the same migration is drawn as a comparison line, which is how close the model lands.
dependants per 100 of working age
The site's modelStatistics Finland's projection, 40,025/yrWithout net migration
Details
The total dependency ratio counts both under-15s and over-64s per hundred of working age. The section's other charts draw the old-age ratio, which counts only the latter, so the figures cannot be compared with those.
The dashes start in 2046. Up to there the base is Statistics Finland's self-sufficiency projection, after it the site's continuation on the authority's own mortality trend and birth scale.
The model is not a forecast but a calculation of where the given assumptions lead. Against Statistics Finland's own projection the model's path is 0.7 points higher in 2030, 3.5 points higher in 2045 and 3.0 points higher in 2060.
Show the figures
Year
The site's model (dependants per 100 of working age)
Statistics Finland's projection, 40,025/yr (dependants per 100 of working age)
Without net migration (dependants per 100 of working age)
Muslim population growthPopulation structureValues and attitudes
Key figures
Scenarios for Finland: the 2.7% Muslim share (2016) would reach 4.2% by 2050 with zero migration, 11.4% in the medium scenario and 15.0% in the high-migration scenario.
An estimated 6.6–7.0 million Muslims lived in Germany in 2025, or 8.0–8.5% of the population, up 1.6–1.8 percentage points from 2019.
Spain's Muslim population was 2.4–2.5 million, about 5% of the population, at the start of 2025, and about 11% of babies born in 2024 had at least one Muslim-background parent.
Immigrants arriving from Muslim-majority countries are among the largest and fastest-growing immigrant groups in Europe and in Finland. Several of those countries are sources of refugee-based migration in particular, which raises the Muslim population of the receiving countries.
Fertility among people from Muslim-majority countries is mostly above two, and well below that among Finnish-background women. The direction is steeply downward, though. Over ten years it has fallen in every one of them except Bangladesh, where it rose, and Iran, where it held steady, and in some it has dropped by nearly half. The figures by background country.
The Pew Research Center has estimated Finland's Muslim share and set out three alternative paths to 2050. They differ only in the volume of migration, which is zero in one, medium in another and high in the third.
A population share does not by itself say anything about consequences. The site classifies by background country rather than religion, because people from a given country are often adherents of its majority religion. Pew builds its own estimates on the same assumption. Consequences are therefore examined by background country: employment and benefit use on the economic side, crime and the views of Europe's Muslims on the others. Background country is a rough proxy for religion, and that is worth keeping in mind when reading the figures.
Finland
Registered Islamic communities in Finland20254 findings
Membership of registered religious communities is official register data (Statistics Finland, population structure, table 11rx: membership of religious communities). The total of 120,000–130,000 is an estimate rather than a register figure: it is based on a census by country of origin, because Muslims who belong to no religious community are not recorded
This is a measurement, not a forecast. It reports what happened between 2010 and 2020, so it complements the 2017 forecast below.
Aside from Christians and unaffiliated people, no other religious group saw its share in any European country change by 5 percentage points or more between 2010 and 2020.
In Sweden the Muslim share rose to 8%, a rise of 4 percentage points. In Germany it rose by about one point to 7%. Albania rose by about 4 points, but Muslims were already a majority there in 2010.
Europe's largest religious shift was not about Muslims but about Christians and the unaffiliated: the Christian share fell to 67% (down 8 points) and the absolute number by 9% to 505 million, while the religiously unaffiliated rose to 25% (up 7 points) and grew 37% to 190 million.
Pew Research Center — Europe's growing Muslim population201730 European countries, scenarios to 20506 findings
Pew estimated Europe's Muslim population at 25.8 million in 2016, or 4.9% of the population.
Even with no further migration, the Muslim share would still rise to 7.4% by 2050, driven only by age structure and fertility.
In the medium scenario the share would be 11.2%, and in the high-migration scenario it would be 14.0% in 2050.
Among migrants who arrived in Europe between 2010 and 2016, an estimated 53% were Muslim. Among those granted or expected to be granted refugee status, 78% were Muslim.
The median age of Muslims was 30.4 years versus 43.8 for non-Muslims. The total fertility rate was 2.6 children for Muslim women and 1.6 for non-Muslim women.
Finland scenario: Pew estimated the Muslim share at 2.7% in 2016. For 2050 the zero-migration scenario is 4.2%, the medium scenario 11.4% and the high scenario 15.0%.
The Muslim population in Europe: projections and country-level examples
A collection of estimates and news reports on the growth of the Muslim share of the population in various European countries. The figures are estimates and projections as given by their sources — some official statistical calculations, some academic, some news or opinion media — and the results depend heavily on assumptions about fertility and migration.
SWI swissinfo.ch, drawing on Pew data: Switzerland's Muslim population was 6.1% (500,000) in 2016. With zero migration the share would rise to 8.2% (660,000) by 2050, to 10.3% (1,140,000) in the medium scenario, and to 12.9% (1,500,000) in the high scenario. Muslims are on average 13 years younger than other Europeans. Pew assumed a fertility rate of 2.1 for Muslims versus 1.5 for others.
A peer-reviewed article uses wavelet analysis to project a Muslim majority in 13 European countries between 2085 (Cyprus) and 2215 (Lithuania). Majority status would be reached in different countries in: Sweden 2125, France 2135, UK 2180. It assumes higher fertility (~1 more child per woman) and 2017-level migration (~706,000 asylum seekers/yr, 86% Muslim). The authors offer no uncertainty caveat about the time horizon — they call the mid-point scenario the “most likely” one and go on to describe concrete social consequences out to 2220. The method is signal processing rather than a demographic cohort model, and the rule used to grow the Muslim population is problematic: each five-year figure adds a share of the 0–4 age group to a share of the total population, although those children are already inside the total. The raised share is then fed into the next period, so the deviation compounds about 40 times by 2220. The model has no mortality, no age structure for the Muslim population and no fertility convergence. Compare this with the Pew and IIASA cohort-component projections.
The scholar of religion Teemu Pauha estimates that Finland had about 140,000 Muslims at the start of last year, growing by roughly 10,000 a year. The article covers the Muslim community in Vaasa, which has grown partly through international students — the University of Vaasa's international student numbers have more than doubled in a year — and prays in three rented spaces. The community would be interested in buying the Vetokannas church and converting it into a mosque. According to Edla Mäkelä, chief architect of the Church Administration, no church in Finland has previously been converted into a mosque. That church is not for sale.
Jerusalem Post: Registered Muslim membership in Norway grew from 72,000 (2006) to an estimated 197,000+ (2025), about +174%. Church of Norway membership, meanwhile, fell from 85% (1994) to 62% (2024).
The European Conservative reports Dutch demographer Jan Latten's estimate: Muslims would grow from about one million today (~6%) to 3–4 million by 2050 (~20% of a 20-million population). The estimate assumes migration continues at the current rate.
CEU CEFAS, Observatorio Demográfico (Informe 23)2026-03
Summary
A report by the demographic observatory of San Pablo CEU University estimates Spain's Muslim population at 2.4–2.5 million at the start of 2025, about 5% of the population: roughly 3.6% first-generation immigrants and about 1.4% people born in Spain with at least one Muslim-background parent. About 11% of babies born in Spain in 2024 had at least one Muslim-background parent. The report attributes the gap to Muslim women's fertility being two to three times that of Spanish women. About 85% of first-generation Muslim immigrants come from Africa, and around 65% of the whole group is of Moroccan origin. Here too, this publication defines Muslim identity by country of origin and parental background rather than religious conviction.
Bundesamt für Migration und Flüchtlinge (BAMF), Forschungsbericht 552026-06-18
Summary
A calculation by Germany's Federal Office for Migration and Refugees (BAMF) puts the number of Muslims living in Germany at 6.6–7.0 million in 2025, or 8.0–8.5% of the population. The share is up 1.6–1.8 percentage points on the 2019 estimate, mainly because of migration from crisis regions in the Middle East. The largest background group is Turkish (39.4%), followed by Arab (29.4%). The Turkish share has been falling since 2008. The figure is an extrapolation combining first results from the 2025 microcensus with the 2020 Muslim Life in Germany survey. Muslims are counted as people with a background in Muslim-majority countries — both immigrants and their German-born descendants — rather than by religious self-identification, which is why the result is given as a range.
Integration
The immigration debate often holds that investing in integration in particular is what reduces the problems associated with immigration. Finland spends a great deal on integration, yet many of its measures — especially for certain immigrant groups — show weak results. Integration can be measured from many angles: the effectiveness of integration measures, school outcomes, residential segregation, partner choice and the sharing of civic obligations. Whatever the measure, the differences between background groups are large and narrow slowly. Some groups need practically no integration measures at all, while others absorb considerable spending without particularly successful outcomes.
Sections in this group and their figures
EducationAt Kirsti school in Espoo 92% of pupils (2026) speak a foreign first language and 55 different mother tongues are spoken, the most common being Albanian.
Language proficiencyIn the 2022 population survey 63 percent of immigrants reported at least moderate Finnish or Swedish. Of those who completed integration training in 2022–2023, 32 percent reached the B1.1 target and 60 percent stopped at A2 or below.
Residential segregationFinland has four municipalities where Finnish is not even the second most spoken language (2022). In Swedish-speaking Korsnäs, the second-largest language group is Bosnian.
Child welfare placementsAccording to a register study of 1991–2015, the Finland-born second generation of foreign background had been placed outside the home almost three times as often as the majority population.
Partner choice and mixed-background couplesA shrinking share of foreign-background people's couples include a Finnish-background partner: down from 82% (1992) to 47% (2023).
At Kirsti school in Espoo 92% of pupils (2026) speak a foreign first language and 55 different mother tongues are spoken, the most common being Albanian.
Espoo already has eight schools where more than half the pupils speak a foreign first language, and across the city the share is about 32% (2026).
The foreign-background share of Finland's comprehensive-school pupils rose from 8.1% to 13.1% in just five years (2020–2025).
Education is a basic measure of integration, because it predicts later employment and income. Learning outcomes also say how well school evens out differences in starting points.
Pupils of immigrant background achieve worse learning outcomes in Finland than the native-background population. In PISA 2025 the native group was 66 points ahead of second-generation pupils in reading and 113 points ahead of the first generation. The gap narrows in the second generation but does not close. About half of it is associated with socio-economic background, and once every background factor the report controls for is held constant the reading gaps are 32 and 63 points. Finland's gaps are among the largest in the OECD. To the extent that the gap appears to narrow, part of the reason is a decline in native-background results. Between 2012 and 2025 native pupils' mathematics score fell by 45 points, the second generation's by 22 and the first generation's by 17, the last of which was not statistically significant.
The figure for adult education level is a lower bound for the first generation, because degrees earned abroad are not recorded in the Finnish register and about half of them therefore have an unknown level. The reliable comparison is between the native population and the Finland-schooled second generation. The start of the chain is under language proficiency and its other end in who one goes to school with: residential segregation.
Educational attainment is also lower among people of immigrant background than among native Finns. Among foreign-background people schooled in Finland, fewer hold a tertiary degree than among the native population, and more are left with only basic education (13.7% vs. 7.2% for natives).
Share with a tertiary degree among 25–34-year-olds by origin · Finland (2025)
% of 25–34-year-olds in the group
Details
Native = Finnish background. 2nd gen = foreign background born in Finland. 1st gen = foreign background born abroad.
Degrees earned abroad are not entered in Finland's qualification register, so about 53% of the 1st generation have an “unknown” education level — their 23% figure is a floor, not a real level. The reliable comparison is the native population vs. the Finnish-schooled 2nd generation.
Finnish-born people of foreign background aged 25–34 are still few (~8,400) and skew toward the earliest immigrant cohorts.
Learning outcomes describe what someone knows on a given day. Completing a qualification describes how many finish what they began, which is a different question: it follows each group from its own starting line and needs no age control.
The answer does not point one way. In higher education the Finland-born foreign-background generation is essentially at the native level, in general upper secondary a little behind, and in vocational education the weakest of the three groups, below even the foreign-born. At university the foreign-born complete more often than anyone else, mostly because that group contains international students who came to Finland for the degree itself.
Share holding the qualification they started, 5.5 years after starting, by education sector · Finland 2024
% of those who started
Finnish background, born in FinlandForeign background, born in FinlandForeign background, born abroad
Details
The measure is the share of those who started in that sector and hold that same sector's qualification 5.5 years later. Someone who switched sector and graduated there does not count as having completed.
No age control is needed, because the measure follows each starting cohort from its own starting line.
At university and university of applied sciences, the foreign-born include international degree students who came to Finland for the degree. That raises the group's completion share and does not describe foreign-background people who grew up in Finland.
The second generation's starting numbers are small, about 850 in general upper secondary and about 200 at university in 2024, so a single year's movement can be noise.
PISA 2025 assessed 15-year-olds with science as the main domain for the third time. The gap between native and immigrant-background pupils is large in all three domains, and in Finland it is among the largest in the OECD. Only Mongolia, Brazil, the Philippines and Lebanon had wider second-generation gaps in science.
The report gives every gap three ways: as measured, with the pupil's socio-economic background held constant, and with home language and grade level held constant as well. The chart shows the first and the last of those. Adjustment roughly halves the gap but does not remove it. Even the smallest remainder, 25 points in mathematics in the second generation, is more than a school year.
Finland's overall result has fallen, but the fall is not the immigrant-background pupils'. Between 2012 and 2025 native pupils lost 45 points in mathematics and 35 in science, while the first generation's corresponding changes were 17 and 9 points and neither was statistically significant. Between 2022 and 2025 the only statistically significant decline was the native group's.
A bar to the left is points below the native population. The upper bar is the measured gap, the lower one the gap after background factors are held constant, and the difference between them is the part that background explains.
Immigrant-background pupils' gap to the native group, before and after background factors are held constant · Finland, PISA 2025
PISA score points relative to native-background pupils
As measuredBackground held constant
Reading, first generation
−113−63
Science, first generation
−100−48
Science, second generation
−73−37
Mathematics, first generation
−69−35
Reading, second generation
−66−32
Mathematics, second generation
−46−25
About 20 PISA points corresponds to one school year of learning on average.
Native = pupils with at least one parent born in Finland. Second generation = born in Finland with both parents born abroad. First generation = born abroad themselves.
Held constant here means three things together: the pupil's socio-economic background (PISA's ESCS index), the language spoken at home and the grade level. The report notes that the last two are very weakly related to attainment once socio-economic background is accounted for, so most of the narrowing comes from that.
Holding socio-economic background alone constant gives the intermediate step: 48 and 89 points in reading, 29 and 47 in mathematics, 54 and 74 in science.
Mean scores were 487, 421 and 374 in reading, 478, 432 and 409 in mathematics and 517, 444 and 417 in science. Finland's national averages were 474, 469 and 504.
There were 6,758 participants from the native group, 243 from the second generation and 294 from the first. The groups are small, so even large point differences are not always statistically significant. Country of birth was missing for 373 pupils, 5 per cent, which is 3 percentage points more than in 2015.
Holding a factor constant is not a causal claim. It says how much of the gap remains when pupils of similar background are compared, not what causes the gap.
The share of foreign-background pupils in comprehensive-school grades 1–9 has risen from 8.1% to 13.1% in five years (2020–2025). Regional differences are large: the share in Uusimaa is several times that of the regions with the least immigration.
Foreign-background pupils' share in comprehensive school by region · Finland 2020 → 2025
Foreign background = both parents (or the only known parent) born abroad. Finland-born second-generation pupils are included, so the share exceeds the foreign-language share.
Kainuu's 2020 marker is missing (value suppressed in the source table).
Early school leaving is markedly more common among foreign-born youth than among the native population. The gap is moderate by EU standards, but combined with weaker learning outcomes (PISA) and lower tertiary completion it foreshadows more persistent divergence in education and labour-market pathways.
Early leavers from education and training (18–24) by origin · Finland (2024)
9.2%Native-born
13%Foreign-born
3.8 pp higher for the foreign-born than for the native population, 9.6% nationwide.
Details
An early leaver = an 18–24-year-old with at most lower-secondary education who is not in further education or training.
Finland's gap is small by EU standards, and this compares the foreign-born (1st generation) with the native population — no separate 2nd-generation figure is published for this indicator.
On the NEET measure (youth not in employment, education or training) the origin gap is typically larger: across OECD countries the foreign-born average is about 19% vs about 14% for the native-born, but a Finland-specific NEET-by-origin figure is not available here.
A school's pupil composition comes up whenever school choice, catchment areas and neighbourhood segregation are discussed. Concern about immigrant-background pupils concentrating in particular schools is commonly answered by saying that composition makes no difference to a school's results. That claim — “foreign-language pupils don't drag down a school's results” — can be checked, as long as one is clear about what the word results covers: Finland's PISA 2022 data splits schools into six groups by the share of pupils with an immigrant background.
The claim splits into two questions that get different answers. A school's own mean does fall, and substantially, because its immigrant-background pupils are part of its result. Computed from the report's figures, the mathematics mean in a top-group school sits about 28 points, or roughly a year and a half of schooling, below a bottom-group school. That follows from the composition of the pupil body rather than showing that the school teaches worse. Native pupils' own mean scores, by contrast, show no fall as the share rises, so on that count the data does not support the concern. The finding carries less weight than it sounds, though. All the report says is that no similar trend was observable among native pupils. It gives no statistical test for that comparison, although it reports significance in nearly every other comparison it makes, and it publishes no standard errors or confidence intervals anywhere. The top group is also small, so ”no observed difference” is not the same as ”no difference”.
The comparison also sets one native population against another rather than following the same group across environments, because schools with a high immigrant share are in the suburbs of large cities and those with the lowest share are rural. The same chapter carries a mechanism that could hold the native mean up: in schools where at least a tenth of pupils have an immigrant background, pupils are streamed by ability into separate classes or teaching groups considerably more often than elsewhere. Some native families also choose a different school or move away, with PISA measuring 15-year-olds only after that sorting has happened: departure shows up as absence from the school, not as a lower score in it.
Immigrant-background pupils' own results, however, fall sharply as the share rises, by something on the order of two years of schooling. Holding socio-economic status constant narrows the gap only slightly, so most of the association survives. The researchers consider weak command of the language of instruction the most likely explanation. Pupils are not allocated to schools at random, so this is an association rather than an effect of the school.
The figures are from PISA 2022, because no equivalent school-level analysis of the 2025 data has been published yet. The most recent native-immigrant gaps are at the top of this section.
The bar is the change in points when a school's share of immigrant-background pupils rises from under 5 per cent to at least 40. The upper bar is unadjusted, the lower one adjusted for socio-economic background.
Change in mean score between schools with under 5% and at least 40% immigrant-background pupils · Finland, PISA 2022
PISA score points
UnadjustedAdjusted for socio-economic status
Native (reading)
+8+1
Native (mathematics)
+7+1
Immigrant background (mathematics)
−40−32
Immigrant background (reading)
−44−38
About 20 PISA points corresponds to one school year of learning on average.
Native pupils' mean shows no fall as the immigrant-background share rises — it is 7 to 8 points higher in the highest-share schools. The difference is small and it has not been tested. Immigrant-background pupils' own mean collapses, and adjusting for socio-economic status does not remove the association: 32 of the 40-point mathematics gap and 38 of the 44-point reading gap remain.
The authors themselves read this more cautiously: the Finnish data supports the international finding only in part, socio-economic status rather than the immigrant-background share as such may explain the association, and they put the gap that survives adjustment down to things like limited command of the language of instruction.
Values were read off the report's charts. Data: PISA 2022 Finland, 241 schools and 10,239 pupils, of whom 1,866 (18%) had an immigrant background (oversample).
The median immigrant-background share across sampled schools was 6%, and in only about 10% of schools was at least one pupil in three of immigrant background — the top group is therefore small and its figures noisier.
“Immigrant background” here means a pupil whose parents were both born abroad (1st and 2nd generation combined). It is not the same as speaking a foreign first language, though the groups overlap heavily.
Adjustment means the pupil's socio-economic status (PISA's ESCS index: parents' education and occupational status plus household assets) is computationally raised to the native level. The adjustment covers the pupil's own background and nothing else. The report does not adjust for a school's location, for the type of municipality, or for the mean socio-economic status of the school's pupils, and the Finnish words for rural area, built-up area and municipality group appear nowhere in it. In large cities, then, a school's result may be lifted by things that have not been separated from the immigrant-background share, school choice and weighted-curriculum classes among them.
This is an observed association, not a causal effect — pupils are not allocated to schools at random.
The fall in a school's own mean is this site's calculation from the report's figures, not a result the report states. It comes from weighting the group's native and immigrant-background means by their shares of the pupils: in the top group at least 40% of pupils have an immigrant background and their mean sits 89 points below the native one. The calculation assumes a share of exactly 40%, the floor of the group, so the real difference is larger. This follows from the composition of the pupil body, not from an effect of the school on any individual pupil.
The oversample does not distort the school grouping. A school's immigrant-background share is computed across its full PISA target population rather than across the pupils drawn for the test, and the national figures correct the over-representation with sampling weights. Two other sampling points remain: even weighted, the immigrant-background share in the data (7%) is lower than the foreign-background share among 15-year-olds in population statistics (9%), and the pupils who had to be invited were as a rule those with more than a year in Finnish- or Swedish-language instruction, so the most recent arrivals may be missing. Both missing groups are likely to be below average, so the measured gap is more likely an under- than an over-estimate.
The report gives no statistical test for the native comparison, although it reports significance in nearly every other comparison it makes, and it publishes no standard errors or confidence intervals anywhere. A reader therefore cannot judge how precise the flat native result is.
The same chapter reports that in schools where at least 10% of pupils have an immigrant background, pupils are streamed by ability into separate classes or teaching groups considerably more often than elsewhere: 42% of pupils against 25%, a statistically significant difference. Streaming is a possible mechanism that could hold the native mean up even where the school's overall level falls. The report does not connect the two findings — that link is this site's observation.
The flat native result is partly geographic. The report puts native pupils' socio-economic status lowest in the rural schools with the fewest immigrant-background pupils and second lowest in the large-city suburban schools where they are most numerous, so the comparison sets one native population against another. Adjustment takes the native advantage from 7 to 8 points down to one point, but the report itself stresses that native socio-economic status did not vary greatly between the school groups. Some native families also choose a different school (see this section's articles).
The gap widens steadily with the school's share of immigrant-background pupils, from 42 points in the lowest group to 89 in the highest. Adjusting for socio-economic status narrows it by 17–31 points — most in the top group, where the socio-economic difference was largest — but the gap still widens with the share after adjustment (25 → 58 points).
Native–immigrant mathematics gap by the school's share of immigrant-background pupils · Finland, PISA 2022
PISA score points (native − immigrant background)
UnadjustedAdjusted for socio-economic status
Details
Differences were computed from means read off the report's charts. The horizontal axis is the school's share of immigrant-background pupils.
Same caveats as the previous chart: the top group is small, this is an association rather than a causal effect, and the adjustment covers only socio-economic status as measured by the ESCS index.
Own-mother-tongue instruction supplements basic education, and it is voluntary both for the municipality and for the pupil. The state grant covers at most two weekly hours per notional group of four pupils. In autumn 2025 there were 24 827 participants, 63 languages and 85 providers.
Pupils in own-mother-tongue instruction · 2018–2025
Pupils and students
AutumnSpring
Details
The figures come from grant applications rather than from a school register. They therefore say what teaching a grant was applied for.
Spring and autumn are two terms of the same calendar year, not the same group counted twice. Autumn is larger in almost every year because new pupils start at the beginning of the school year.
The cost is not published separately. The 2024 general grant for foreign-language pupils was EUR 22 475 000, and it covers Finnish or Swedish as a second language, other support for teaching, and own-mother-tongue instruction in one sum.
Computed with the grant's own coefficients, own-mother-tongue instruction comes to something on the order of EUR 12 million a year: 24 827 pupils in notional groups of four, two weekly hours, 38 school weeks and EUR 25 per teaching hour. This is the site's own calculation rather than an official figure, and it assumes the groups are full.
The teaching is voluntary in both directions, it runs to at most two hours a week, and its cost sits in the same EUR 22.5 million budget line as second-language teaching. A national evaluation exists, but it did not measure learning outcomes.
Kansallinen koulutuksen arviointikeskus (Karvi)2023
Summary
The main recruitment problems are that applicants lack teacher training, teaching experience or sufficient Finnish. A third of providers can and a third cannot give adequate support with current resources. Classes are often held in the evening, and only about half of the mother-tongue teachers worked with a preparatory-education teacher or a special-needs teacher. Teachers say quality would improve if classes were grouped by age and the syllabus were returned to the national core curriculum. The evaluation has effectiveness in its title, but the report defines it as the teaching's capacity and opportunities to support the intended changes, and the material is a survey. Learning outcomes were not measured, so whether the teaching affects language skills or school performance is not known in Finland.
The 2024 general grant was EUR 22 475 000 for foreign-language pupils and EUR 75 000 for Sami and Romani pupils. The same sum covers three things: Finnish or Swedish as a second language, other support for teaching, and own-mother-tongue instruction. The coefficients are public even though the split is not: the computed cost of a teaching hour is EUR 25, the grant covers at most 86 % of accepted computed costs, and own-mother-tongue instruction is funded for at most two weekly hours per notional group of four pupils. Two pupils form a group where four cannot be found.
In autumn 2025 the teaching reached 24 827 pupils and students, in 63 languages, from 85 providers. In autumn 2018 the figures were 20 732, 58 and 88, so participation has grown by about a fifth while the number of providers has stood still. The largest languages in autumn 2025 were Russian 5 931, Arabic 3 019, Somali 2 108, Ukrainian 1 746, English 1 361 and Albanian 1 206. Ukrainian grew from 105 pupils to 1 746 in three years.
First-generation immigrants in Finland fall about 105 points behind the native population on the PIAAC literacy scale — the largest gap of any PIAAC participating country. The gap narrows to about 51 points when the immigrant speaks Finnish or Swedish at home and has lived in Finland for at least 5 years. 83% of TE-services experts say poor literacy has led to a jobseeker being rejected.
Finland · PISA 2022: 58% of first-generation immigrant pupils fell below the baseline level in mathematics (native-background pupils: 22%). In reading, 61% were below the baseline. The gap is also visible in the second generation (43% / 39%).
Euroopan komissio, Education and Training Monitor (Suomen maaraportti)2025
Summary
According to the Commission's country report, underachievement among foreign-born pupils in Finland stands at 57%, one of the highest rates in the EU, and the gap remains significant even after accounting for socio-economic background and the language spoken at home. Early leaving from education is 13.0% among the foreign-born against 9.2% among the native-born. The share of disadvantaged pupils reaching a good level of basic skills has fallen by 8 percentage points since 2015: Finland remains more equitable than the EU average, but the trend is downwards.
A study by the University of Jyväskylä and the Ministry of Education and Culture using PISA 2022 data (241 schools, over 10,200 pupils, about 1,900 of them of immigrant background): 61% of first-generation immigrant pupils had weak reading skills and 58% weak mathematics skills. In the second generation the figures were 39% and 43%, against roughly 21% and 25% for all pupils. The researchers note that the gap between immigrant-background and native pupils has narrowed above all because the proficiency of majority-population pupils has declined.
An Yle report from two Helsinki schools where more than two thirds of pupils have an immigrant background and around 30 languages are spoken. Pupils call the language of the schoolyard "mamusuomi" (immigrant Finnish). Behind it is the same research finding as above: 61% of first-generation immigrant pupils read too poorly to take full part in society. The recommendation cited is that the immigrant-background share of a class should stay at 20–30% at most. At the Itäkeskus school, over 70% of pupils go on to upper secondary education. Education Minister Anders Adlercreutz promised more weekly hours of mother tongue and literature in primary school, a reform of learning support, and €50 million in equity funding for hiring staff. This is a good example of how immigration adds to public costs in education too.
A first-grade Finnish lesson at Itäkeskus comprehensive school in Helsinki cannot be called a mother-tongue lesson, because not one pupil has Finnish as a first language. The split groups of two first-grade classes were present, 17 pupils, all of them S2 pupils (Finnish as a second language). Ten different languages are spoken in one of the school's classes, and S2 pupils are numerous at other year levels too. Some 7-year-olds read more fluently than many native Finnish speakers, while others decode mechanically without fully understanding the text. Teachers say preparing materials at several levels adds work, and interpreters are needed often because no common language can be found. A particular challenge is that for some children Finnish is present only at school: the mother tongue is spoken at home, and the child may be the best Finnish speaker in the family.
At Kirsti school in Espoo, 92% of the roughly 450 pupils speak a foreign first language, and 55 different native languages are spoken (Albanian the most common, then English). Across Espoo about 32% of pupils speak something other than Finnish, Swedish or Sámi as their first language, and in eight Espoo schools the foreign-language share exceeds 50%. The principal calls the school's multiculturalism a genuine asset, but also notes that once the share of foreign-language children passed 40–50%, Finnish-speaking parents began considering moving their child to another school.
A follow-up on the Kirsti school in Espoo (about 450 pupils, 92% of them foreign-language speakers): Espoo now has eight schools where foreign-language pupils are more than half. Politicians from the Greens, the National Coalition and the Social Democrats describe the trend as segregation that has gone "in a really difficult direction". The measures floated include mixing rental and owner-occupied housing, redrawing school catchment areas and expanding specialised teaching programmes. They disagree about where responsibility lies and which measures would work.
Helsinki is redrawing the catchment areas of comprehensive schools in its eastern districts: children from Puu-Myllypuro will move to Itäkeskus comprehensive school from 2027, and later so will children from the Karhunkaataja area. The official justification is evening out pupil numbers, but the change is also driven by concern over residential segregation. Itäkeskus has very few children who speak Finnish as their first language, and officials are now saying out loud that without the transfer, the risk is that its pupils will not learn Finnish. The school boards disagreed: Itäkeskus called the change excellent, Myllypuro opposed it.
Last year 65% of pupils at Itäkeskus comprehensive school studied Finnish as a second language. Among first-graders the share was 84%. Helsinki has decided to redraw its catchment areas so that more pupils transfer to Itäkeskus comprehensive school, particularly from areas with a Finnish-background population. Parents in Herttoniemi have appealed the decision to the administrative court. Their main concern is the longer journey to school. One parent interviewed also raises the number of second-language pupils, while saying they have heard a great deal of good about the school. According to Ville Raatikainen, the city's director of basic education, the sole reason is evening out pupil numbers, because Myllypuro is full and the Itäkeskus school is being extended, while Kaisa Eskola (Left Alliance), who chairs the committee division, also cites an aim to reduce segregation. Doctoral researcher Oskar Rönnberg's work puts the school argument in proportion: families with children are indeed keen to move away from more disadvantaged and ethnically diverse areas, but schools are not the reason. By far the most important reason was a perceived lack of safety in the area, problems of social order and visible substance use. Turku does not treat redrawing catchment areas as its primary tool. It invests instead in the schools of weaker areas, with smaller teaching groups, more learning materials, more assistants and special-needs teachers, and with specialisation classes.
Teachers interviewed by MTV describe everyday life with Finnish-as-a-second-language pupils: some pupils speak no Finnish at all when they start, and differentiating the teaching takes time away from the other pupils. In Helsinki, several schools have a majority of foreign-language pupils, while at Sammu school in Huittinen S2 pupils are about 7% and the teacher says the language catches on fast in a Finnish-speaking environment. The state funds language support at €50 million a year, rising to €75 million from 2027. The piece is based on teacher interviews and a survey answered by 31 of some 1,700 class teachers.
Language proficiencySkillsEconomy
Key figures
In the 2022 population survey 63 percent of immigrants reported at least moderate Finnish or Swedish. Of those who completed integration training in 2022–2023, 32 percent reached the B1.1 target and 60 percent stopped at A2 or below.
One of the most essential parts of integration is knowing the language. Without it, neither an understanding of the surrounding culture nor ties to it form as readily. There are two figures for language proficiency, and they measure different groups. A population survey covers everyone who has moved to Finland. The training statistics cover only those who finished a course, and only some are eligible for one. What the language is enough for at school is under education, and what the teaching costs under integration policy.
Finland
The first figure covers everyone who has moved to Finland, including those who never attended a course. The two that follow cover only those who completed integration training, that is the training's best possible result. The figures are not comparable with one another.
Two figures for language proficiency, two different groups
63%Report at least moderate Finnish or Swedish (population survey)
32%Reached the B1.1 target, of those completing integration training
60%Stopped at A2 or below, of those completing
Details
The first figure is a self-assessment in a survey; the two that follow are a measured level. They do not measure the same thing or the same group.
The training figures cover only those who finished: people who dropped out or never started are not in them.
Language proficiency across the whole immigrant population
THL's MoniSuomi 2022 is the only language figure that also covers people who have attended no course at all: 63% reported speaking Finnish or Swedish at least at intermediate level, leaving 37% below it. The spread between background countries runs from 89% to 40%. These are self-assessments rather than test results, “intermediate” corresponds to no level of the Common European Framework, and self-assessment usually overstates.
From the 2014 survey on the work and well-being of people of foreign origin: three in four rated their Finnish or Swedish at least intermediate. The figure cannot be set beside MoniSuomi's 63% as a time series. The populations differ — UTH covered people of foreign origin aged 15–64, MoniSuomi foreign-born people aged 20–74 — and the question is not the same. The gap mainly shows that two self-assessment surveys do not make a trend.
A THL population survey of 7,838 foreign-born respondents, collected between September 2022 and March 2023. The question was put to everyone regardless of whether they had attended any course, which makes it the only figure covering the whole population. Of all immigrants 63% reported speaking Finnish or Swedish at least at intermediate level: 68% of women and 57% of men aged 20–74. The spread between background countries is steep: 89% of women from Estonia, 80% of women from Russia or the Soviet Union against 60% of the men, while the figure is 44% for people from South-East Asia and 40% for those from the rest of Asia, the Americas and Oceania. The regional pattern runs against expectation: 70–71% in Kymenlaakso, Päijät-Häme and Central Uusimaa, 60% in Helsinki and 59% in Western Uusimaa.
Breaks the MoniSuomi distribution down further than the combined share: about a quarter rate themselves excellent, about a third beginners, and about one in ten report speaking neither national language at all. Proficiency was weakest among those who came for work or study, who may not need the language day to day and some of whom are in the country temporarily. The page notes that survey data is needed because register data cannot measure language skills.
What integration training aims at, and what it reaches
The linguistic target of integration training is B1.1, and about a third of those who complete the training reach it. The figure covers only those who finished: people who dropped out or never started are not in it, so it is the training's best possible result rather than an average for immigrants. There is no sign of improvement — in 2013–2016 just under 35% of those finishing reached B1.
Owal Group, työ- ja elinkeinoministeriön tilauksesta2025
Summary
The linguistic target of integration training is B1.1, a functional basic proficiency. According to the review, a ministry survey of the ELY centres in summer 2024 found that in the training the centres purchased in 2022 and 2023, 32% of everyone who completed it reached at least B1.1 and about 60% stopped at A2 or below. Authors Risto Karinen, Taru Aihinen, Jenna Siltala and Olli Oosi. The cover is dated May 2024 although it cites a survey run that summer. The document was published in the government project portal in May 2025.
Who is eligible for the training
An integration plan and the training that follows are limited to registered unemployed jobseekers in their first three years. The conditions leave out employed work-based migrants and their spouses, people at home with children, students, and anyone whose three-year window has closed. The scale says the same: the places available are a fraction of the foreign-language population.
The conditions for a plan and the training that follows from it: the person is a registered unemployed jobseeker, no more than three years have passed since the first residence permit or the registration of residence, and they are able to take part full-time in services promoting integration and employment.
The established form of integration training — labour market training lasting about a year — has 13,000–14,000 students annually, and the full-time training at folk high schools about 7,000. There were 610,148 foreign-language speakers in Finland at the end of 2024. Statistics Finland counted 57,000 students in liberal adult education aimed at immigrants in 2024, but those courses are lighter than the targeted integration training, and attendance says nothing about the level reached.
What interpretation costs
The absence of language skills shows up as an interpretation bill whose size no one knows.
Suomen kääntäjien ja tulkkien liitto (SKTL)2023-05-02
Summary
The profession's own background report on the lack of regulation: the law sets no competence requirement for an interpreter, so a single assignment may be handled by anyone. Only the title of court interpreter is protected, and the register the National Agency for Education has kept since 2016 exists, but no authority is obliged to use an interpreter from it. The report comes from an interest organisation and its proposed remedy is an interpreting act, but its description of the current absence of regulation can be checked against the law.
The wellbeing services counties spent EUR 35.3 million on interpretation and translation in 2024, about four million more than in 2023. The largest state buyers are the court service at 4.5 million, the Immigration Service at 3.8 million (2.9 million in 2023) and the Prime Minister's Office at 2.6 million. Municipal spending was unchanged. Part of the growth is remote interpretation, which has increased use. No one holds a total, because the purchases are not recorded in a uniform way. In Finland the right to service in one's own language does not depend on time in the country. In Denmark someone who has lived there more than three years pays for health-care interpretation themselves.
The language requirement for citizenship
The language level citizenship asks for is the same B1.1 that integration training aims at and that 32% of those completing it reach. Citizenship can also be obtained in Swedish, and some switch language for exactly that reason: Swedish is perceived as easier than Finnish because it is close to English and the other European languages. The route is small but growing — in 2022, 530 people took the intermediate test in Swedish against 6,869 in Finnish. The requirement can therefore also be met in a language that is the mother tongue of 285,360 people in Finland, about five per cent of the population, a third of them in Ostrobothnia.
Natalia Tolmacheva, a PhD in economics, studied Finnish for seven years without reaching the required level, switched to Swedish and passed the test at the first attempt after ten months of study. She speaks mainly English in Finland. In 2022, 530 people took the intermediate National Certificate in Swedish against 6,869 in Finnish, and demand for the Swedish test has grown by about a hundred takers a year. Lidiia Salo, who prepares candidates for the test, says Swedish is perceived as considerably easier than Finnish. Saara Kalajoki of the National Agency for Education offers another explanation: the choice usually follows from moving to a Swedish-speaking region or a Swedish-speaking spouse. One interviewee, Kia Lindberg of the Helsinki Adult Education Centre, raises the integration concern.
Citizenship requires satisfactory spoken and written Finnish or Swedish, normally demonstrated through the National Certificate of Language Proficiency. The bill adds a citizenship test on knowledge of society from the beginning of 2027. The test would not be required of an applicant who has demonstrated proficiency matching the Integration Act's target level, B1.1, in at least two skill areas.
There were 285,360 Swedish speakers at the end of 2024, about five per cent of the population, and the number is projected to fall to 281,789 by 2030 and 260,973 by 2050. The distribution is uneven: about a third live in Ostrobothnia, where Swedish speakers are a narrow majority, and although the capital region has the largest numbers — 36,945 in Helsinki, 20,246 in Espoo — the share of each city's population is small. The compiler is an advocacy body for Swedish speakers, but the figures are Statistics Finland's.
Professional language skill, and whether anyone tests it
The general language certificate that suffices for citizenship does not measure the language of the job, and professional language skill is not measured in Finland at all: the law requires a health or social care professional to have sufficient language skills without defining what sufficient means, and no test exists. In SuPer's member survey in September 2024, 53% had seen dangerous situations caused by a foreign-language colleague's lack of Finnish and 83% found it a burden on themselves. The figures come from a union survey and measure a perception rather than a recorded incident, and SuPer is campaigning for a compulsory professional language test — but the absence of such a test can be checked against the law and does not rest on the survey.
“In Finland nurses' professional language skills are not tested at all. No such test even exists.” The law requires a health or social care professional to have sufficient language skills without defining what sufficient means, and only general proficiency is tested. Both nursing unions, SuPer and Tehy, regard this as a clear patient-safety risk. In SuPer's survey one per cent of respondents judged the professional language skills of those recruited from abroad to be sufficient. The unions blame employers' unwillingness to pay for language teaching.
A member survey run 2–22 September 2024 among members in employment, 3,255 responses. On the question about dangerous situations (n = 2,157), 14% answered yes, often and 39% yes, sometimes — 53% had seen dangerous situations caused by a foreign-language colleague's lack of Finnish. 23% had not and 24% could not say. Of the respondents 83% agreed partly or fully that a colleague's weak language skills are a burden on them. The open answers return to medication, record-keeping, understanding care instructions, and tasks shifting to the Finnish-speaking staff. The survey is the union's own, the respondents are self-selected and the question measures a perception rather than a recorded incident. SuPer is campaigning in the same document for a compulsory B2-level test as a condition of graduation.
Finland has four municipalities where Finnish is not even the second most spoken language (2022). In Swedish-speaking Korsnäs, the second-largest language group is Bosnian.
In the Havukoski district of Vantaa, over 55% of residents speak a foreign first language (2026).
Finland already has nearly 20,000 Albanian speakers (2026). A community that could be called “Little Kosovo” has grown up in Suvela, Espoo.
One measure of integration is whether people end up living among the local population or concentrated together with their own community. The more the same immigrant groups concentrate in the same areas, the more segregation forms. A higher foreign-background share can also affect whether the native population wants to move to an area, because it often changes people's impression of the place.
The foreign-background population is heavily concentrated in the Helsinki Metropolitan Area, and at neighbourhood level the gap is sharper than at municipal level. The largest shares are in eastern and north-eastern Helsinki, but the most segregated areas are in Vantaa and Espoo, where the foreign-background share already reaches around half in places.
Socioeconomic segregation is rising in all of Finland's largest cities. Ethnic segregation is measured with a dissimilarity index, which says what share of foreign-language speakers would have to move to another neighbourhood for them to be spread evenly across the city. The index rose most in Helsinki, but fell across the large-city average. It measures only how evenly a group is spread, not how large it is. Part of the shift in the area distribution comes from the group growing alone. The foreign-background share of the whole country rose from 5.9% to 11.1%, so the distribution moves right even if the areal split stayed the same. The foreign-language share has meanwhile risen steeply, and their concentration in the most disadvantaged neighbourhoods has continued.
Neighbourhood-level figures are available from Suomen Perusta's lähiöluvut.fi: population share, ARA subsidised rental housing, income and housing support, employment, income and education level by origin group across Helsinki, Espoo and Vantaa. Some figures in this section concern people of foreign background and others foreign-language speakers. These are different concepts, and each entry states which one it uses. How the area shows at school is under education, and where the make-up of the population comes from under demographic change.
Where people of foreign background live in Finland · by the postal area's share
% of all people with a foreign background
20142024
Details
2,984 postal areas with data for both years are included. The source's area 99999 (“Tuntematon”) is left out: it is not a place but where the source files people whose postal code is unknown — 78,206 of them in 2024, 22,955 of foreign background. At 2.5 times the largest real postal area it skewed the measure while it was in.
The national foreign-background share rose from 5.9% to 11.1%, so the distribution shifts right on the group's growth alone.
Dissimilarity index by postal area · Finland 2014 and 2024
Dissimilarity index
20142024
Details
The index is the share of foreign-background residents who would have to move to another postal area for the group to be spread evenly within the area.
Cities with at least 8 postal areas are included. The national figure is computed across all postal areas, so it is not on the same scale as a single city's.
Vantaa: 25% of residents are foreign-background — the highest in the Helsinki Metropolitan Area. Espoo: 22%. Helsinki: about 19–20% (over 134,000 people). Across Finland, foreign-background residents are about 9–10% of the population.
The number of crowded households has grown especially among foreign-language speakers, and their number doubled over ten years. Nationally there were 227,000 crowded households holding 937,000 people, and about 56,000 households in the Helsinki region, 10.5 per cent of its residents. Statistics Finland counts a household as crowded when it has more than one person per room, the kitchen not counted as a room.
An Yle report marking Varissuo's 40th anniversary in Turku. The suburb has about 9,000 residents, 12,000 at its peak, and roughly 40% of them now speak a first language other than Finnish or Swedish. More than 40 nationalities live there and nearly 60 languages are spoken. The Varissuo school has under 200 pupils with 19 different mother tongues and 20 nationalities. Varissuo is Turku's second-largest suburb after Runosmäki.
There are four municipalities in Finland where Finnish is not even the second most spoken language. In Swedish-speaking Korsnäs (about 2,000 residents) the second-largest language group is Bosnian. Finland's foreign-language population grew by more than 25,000 in the previous year, the largest increase in 40 years.
An Yle report from Hervanta in Tampere: 48% of preschoolers at the Etelä-Hervanta school speak a language other than Finnish or Swedish, and 52% across all Hervanta preschools, against 9% for Tampere preschools as a whole. For comparison, the figure is 27% in Helsinki's municipal preschools and 20% in Turku. Besides Finnish and Swedish, Hervanta's preschoolers have 25 other mother tongues, the most common being Somali, Arabic, Dari, Albanian and Russian. The district's foreign-language population doubled from about 3,300 (2010) to 6,800 (2022), and roughly one in three Hervanta residents now speaks something other than Finnish or Swedish as a first language.
Peer-reviewed study covering Finland's 20 largest cities with register data from 2000–2018: the income-based segregation index rose from 0.32 to 0.39 even though income inequality (Gini) barely changed, and socioeconomic segregation increased in all 20 cities and at every scale of analysis. Ethnic segregation developed differently: it rose most in Helsinki, but the 20-city average slightly FELL (0.29 to 0.28) and fell in 9 cities out of 20. The share of foreign-language speakers nevertheless rose steeply — Helsinki 5% to 14%, Espoo 4% to 16%, Vantaa 4% to 17%. The authors reconcile this themselves: a dissimilarity index can fall while the foreign-language share grows sharply, and foreign-language speakers remain overrepresented in the most socioeconomically disadvantaged areas — in the mid-rise suburbs built between the 1960s and 1990s their share is about twice the city average.
Lapland had nearly 7,000 foreign nationals in 2023, or 3.9% of its population, against 7.3% nationally. The number has doubled in a little over a decade, but growth is slower than in the rest of the country: around 2,000 people were added between 2022 and 2024. The Rovaniemi area has roughly 3,000 foreign nationals. These figures count foreign citizens, not people of foreign background, which is a considerably wider group.
HS on a report by Hypo bank and the regional research company MDI: foreign-language residents who have moved to Finland live in rented housing and on the outskirts of cities more often than the native population. Their housing aspirations, however, are much the same: over 80% of those renting hope to move into owner-occupied housing, but doubt it will happen because of their employment situation and the uncertainty of residence permits. Hypo's managing director Antti Tolonen says immigration means cities need more large, family-sized flats. The report is published by a bank and a research company, which is worth keeping in mind when reading its housing-policy conclusions.
Net migration cost Vantaa 2.6 million euros in tax revenue in 2023, because those moving in paid less tax than those moving out. Cross-country skier Vilma Nissinen recently called Hakunila and Vantaa as a whole "Finland's Mogadishu", a characterisation the piece sets against the area's sports facilities and hiking trails. The planned tram line is expected to bring housing for 60,000 people and 30,000 jobs.
An Iltalehti report from a daycare centre in an area where more than half of residents are of foreign background. The centre has only three Finnish-speaking children left. Some Finnish parents avoid daycare centres like it, though the matter is rarely discussed in public debate.
A Kosovo-background community has formed in Suvela, Espoo. Families fleeing the Yugoslav wars settled near their compatriots from the 1990s on. About 1,500 Albanian speakers live in the Suvela, Espoo Centre and Tuomarila areas, out of some 4,000 in Espoo as a whole. Finland has 19,695 Albanian speakers in total, 11,718 of them in the Helsinki metropolitan area.
The mall was Finland's largest when it opened in 1965. Today it is known as the country's most multicultural marketplace, with shops selling ethnic foods and spices, a small mosque, and community events such as Somalia's national day and the Puhos Loves People festival. About 20% of Helsinki residents speak a first language other than Finnish, Swedish or Sámi. Itäkeskus and several of the residential areas around it have a relatively high foreign-background share. The downsides mentioned are the drug trade and a sense of insecurity.
Child welfare placementsSafetySocial cohesion
Key figures
According to a register study of 1991–2015, the Finland-born second generation of foreign background had been placed outside the home almost three times as often as the majority population.
A child is placed outside the home when the support given at home is not enough. The reason may be the child's own symptoms or the conditions at home.
According to a register study covering 1991–2015, children of foreign background are clearly overrepresented among those placed. In the Finland-born second generation, placements were nearly three times as common as in the majority population. Among children of refugee background the share is somewhat higher than among others of foreign background. The difference also shows in the kind of placement. Foreign-background placements more often begin as emergency placements, and the first placement is less often a foster family and more often an institution such as a children's home.
Placed outside the home at least once in 1991–2015, by origin · Finland
% of the group's children and young people
Details
Data: persons born 1976–1999 (40,972 with a foreign background, matched against a random sample of 35,772 from the majority population). Share of each group with at least one out-of-home placement in 1991–2015.
1st gen = born abroad, 2nd gen = born in Finland with both parents born abroad. Refugee background = parents' birth country a refugee origin country.
The differences were highly statistically significant.
THL's annual statistics do not break placements down by background, so monitoring the phenomenon depends on individual studies.
Janus – sosiaalipolitiikan ja sosiaalityön tutkimuksen aikakauslehti2023
Summary
3.6% of the majority population had at least one placement, against 6.0–7.5% of the first generation and 9.6–10.4% of the Finland-born second generation. The register study links THL's child welfare register with Statistics Finland population data (born 1976–1999, 40,972 with foreign background, 35,772 majority-population controls). About a third of foreign-background children's first placements were emergency placements (20% for the majority), and nearly one in four emergency-placed foreign-background children was emergency-placed twice. Family placements accounted for 5.8–9.4% of foreign-background placements and about 20% for the majority, so children's home placements dominated. Of placed second-generation children, 31.1–36.7% were under school age (14.7% in the majority population). The authors set out two explanations side by side: the group's above-average welfare risks and structural features of the service system. They read the frequency of emergency placements as a sign that early support and non-institutional services had not reached these families in time, and attribute the prevalence of institutional placements partly to a shortage of suitable foster families. The register data describes service use and official responses, not the need for services or their accessibility.
Partner choice and mixed-background couplesSocial cohesion
Key figures
A shrinking share of foreign-background people's couples include a Finnish-background partner: down from 82% (1992) to 47% (2023).
In 2022, couples in which both partners are of foreign background outnumbered couples with one Finnish-background partner for the first time.
Partner choice varies sharply by background country: 91% of Thai-background couples include a Finnish-background partner, against 8.0% for Somali and 7.9% for Syrian background (2023).
Partner choice is one measure of integration. The differences reflect the reason for migration (marriage migration vs. refuge) and community size, but also how strongly a group orients itself toward the surrounding society.
In some groups a partner is chosen almost exclusively from within the community, leaving everyday life separate from the native population and letting parallel communities settle in across generations.
The share paired with a Finnish-background partner varies enormously by background country: a clear majority of couples with a Thai, Swedish or other Western background include a Finnish-background partner, whereas among refugee-origin groups (Syria, Somalia, Afghanistan, Iraq) the share stays below a quarter.
Share of couples with a Finnish-background partner, by background country · Finland (2023)
All family types are included (married, registered and cohabiting couples). For each background country: the share of couples with at least one partner of that background whose other partner is of Finnish background.
Calculation: (X–Finnish-background couples) / (all couples with at least one X-background partner).
People of Russian / former-Soviet background are split across several classes. The largest, “former Soviet Union”, is shown here.
One row per country, two directions. To the left, how many per thousand women aged 20–64 of that background are in a couple with a Finnish-background man; to the right, the same for the group's men. Both are divided by the group's own male and female population, because a raw count of couples would also reflect the sex ratio: protection-based migration is male-heavy. The top of the list leans right, the bottom leans left.
Partnered with a Finnish-background person: women left, men right · by background country · Finland (2025)
couples per 1,000 adults aged 20–64 of that sex
women with a Finnish-background partnermen with a Finnish-background partner
Iraq
−15+96
Afghanistan
−10+46
Syria
−11+40
Turkey
−54+198
Morocco
−84+297
Pakistan
−12+41
Nigeria
−45+123
Democratic Republic of the Congo
−21+54
Somalia
−10+23
Nepal
−16+30
Former Yugoslavia
−54+95
India
−32+52
Iran
−54+86
Bangladesh
−12+18
Germany
−446+646
Sri Lanka
−18+22
Spain
−344+432
United States
−517+618
Ethiopia
−61+65
Sweden
−601+633
Romania
−126+112
Kenya
−178+139
Hungary
−282+215
Brazil
−467+355
Bulgaria
−157+113
Unknown
−263+189
Myanmar
−55+36
Estonia
−182+95
Poland
−332+159
Latvia
−171+68
Former Soviet Union
−218+79
Vietnam
−169+58
Ukraine
−66+22
Russia
−187+52
China
−213+42
Philippines
−199+32
Thailand
−644+102
At group level: from Muslim-majority countries 24 per thousand for women and 83 for men; from other countries 211 for women and 112 for men. On the men's side the gap is small and runs the other way. On the women's side it is almost ninefold.
The left-hand figures are positive rates, not negative quantities. The side distinguishes the sexes, not the sign.
Families on 31 December, not marriages contracted. The table's dimensions are man or older partner and woman or younger partner, so same-sex couples are included as a small minority.
Countries are included when they have at least 1000 adults aged 20–64 of each sex in Finland and at least 100 mixed couples. The threshold is the same for every country, so none was picked for its result.
The denominator is the group's whole population aged 20–64, not those in a relationship, because the number in a relationship by background country and sex is not published. The figure is a rate, not a share of couples.
Without that adjustment the figures mislead. Among mixed couples with a Muslim-majority background, 4811 are male-led and 954 female-led, but the same group has about 58,000 men and 39,000 women here, so part of the difference is headcount.
Background country is not religion. Statistics Finland does not record religion and this chart follows that. Background country is a rough proxy for a country's majority religion, and the roughness shows here: Indonesia is Muslim-majority and is not at the top of the list.
A shrinking share of foreign-background people's couples are with a Finnish-background partner: down from 82% in 1992 to 47% in 2023. As communities grow, partners are increasingly found within one's own background group rather than among the native population. In 2022 the number of couples where both partners have a foreign background overtook mixed couples.
Of marriages formed by Finnish-background women in 2024, 6.4% were with a foreign-background man, and 6.6% of Finnish-background men's marriages were with a foreign-background woman. In 1990–2010 the shares were at the 4–5 percent level. Women's foreign-background spouses most often came from the former Soviet Union, Iraq and the United Kingdom, men's from the former Soviet Union, Estonia and Thailand. Of marriages formed by resident women 7.8%, and of those formed by resident men 11.5%, were between two foreign-background spouses.
Integration policy: costs and effectivenessEconomySocial cohesion
Key figures
Under 35% of those completing integration training in 2013–2016 reached the target language level B1.
Of those finishing integration training in 2016, more than four in five stayed below B1.1, the level vocational training requires, and nearly 60% stayed below A2.2.
A municipality receives a calculated compensation from the state for every refugee it takes in: €6,845 a year for a child under school age and €2,300 a year for anyone older.
Finland spends hundreds of millions of euros a year on immigrant integration, mostly as state compensations to municipalities and wellbeing services counties, plus integration training and integration plans. The key question is how much of that spending helps.
The evidence suggests that individual low-cost measures may work, but integration outcomes, employment and language skills, remain weak in international comparison. Language skills have been measured in Finland too. According to the National Audit Office, only about a third of those who completed integration training in 2013–2016 reached the B1 target level, so two in three fell short of it. The evidence base is thin: there is one credible causal study of effectiveness, it covers people who arrived in the 1990s, its effect is measured for about a third of them, it contains no cost-benefit calculation, and the reform it evaluates did not move employment — only earnings and benefit use. Against an annual spend in the hundreds of millions, that is not much.
Data on unemployment, and on the high welfare use that follows from it, can also be found on this site: employment rate by background country and sex and subsidy dependency by origin. Employment is generally low among people of foreign background and particularly so among groups who arrived from refugee-producing countries, whose share of welfare use is correspondingly high. Differences in employment between the sexes can also say something about how far integration has come in different groups. Women who moved from Muslim-majority countries such as Pakistan, Bangladesh, Afghanistan and Turkey have markedly lower employment rates than men.
Compensation paid to municipalities and wellbeing services counties is the largest single item of integration spending, and the sum rose by well over a hundred million in a year.
Integration compensation in the state budget
279Compensation for promoting integration, 2023, EUR m
384The same in 2024, EUR m
6 845Per-child compensation to the municipality, under school age, EUR/yr (2024)
Details
The figures are budget compensation items, not the total cost of integration: training, interpretation and social security sit under other headings.
Integration spending: cost and effectiveness20247 findings
Compiled from the state budget (budjetti.vm.fi), the Centre of Expertise in Immigrant Integration / Ministry of Employment (TEM), VATT and the OECD (Settling In 2023, Economic Survey Finland 2025)
Public funding for integration consists mainly of state compensations paid to municipalities and wellbeing services counties, plus integration training and integration plans. State-budget compensations for promoting integration were about €279 million (2023) and about €384 million (2024). Compensation specifically for the reception and integration of people receiving international protection came to about €137 million paid to municipalities in 2020.
The calculated per-person compensation to a municipality was €6,845/year for a child under school age (0–6) and €2,300/year for a person aged 7 or over (2024 level), paid for 3 years — 4 years for quota refugees.
Effectiveness: according to VATT, receiving an individual integration plan raised an immigrant's cumulative gross income by more than €20,000 over ten years and cut the transfers they received by roughly €8,000 — cumulative earnings rose 47% without raising total costs (Sarvimäki & Hämäläinen 2016, VATT). The figure is a local average treatment effect for the unemployed immigrants whose treatment the reform actually changed, not an average across all immigrants.
One low-cost measure may therefore work, but the study contains no cost-benefit calculation and the reform did not move employment, only earnings and benefit use. Integration outcomes — employment and language skills — remain weak in international comparison: on OECD measures the employment gap between the native- and foreign-background population is among the widest in the Nordics and the whole OECD, and fewer than half of immigrants are proficient in a domestic language (EU average 62%) (OECD Settling In 2023, Economic Survey Finland 2025).
What the design does and does not carry. The threshold could not be manipulated: it was published a year after the cut-off date, and there is no jump in the number of arrivals at that point. The sample, though, is drawn from a 15 % register sample and then restricted to men aged 25–60 who were unemployed or on social assistance during their first three years. The compliers — those whose treatment the rule changed — are about a third of immigrants, and arrivals from the former Yugoslavia are over-represented among them by a factor of 4.1. The authors themselves describe the balance test on background characteristics as unstable: the estimates are mostly insignificant but sensitive to the specification, and the sign changes. They also note that some immigrants may have avoided the plan by finding work before the programme began, which would bias the estimate upwards.
The reform did not increase the number of training days. It changed what they contained, shifting them towards language training. It is the same money spent differently.
Set against the international literature the figure sits at the top of the range rather than at its middle. A European meta-analysis (93 estimates from 33 studies) found that only private-sector wage subsidies improved immigrants' labour market outcomes. Training and job-search assistance did not, on average (Butschek & Walter 2014). Sweden's equivalent 2010 reform — an individual plan and a coach — showed up in neither employment nor income over the first two years, and in France language training raised labour force participation by 15–27 percentage points but not the probability of employment (Lochmann et al. 2019). Of the six forms of support in Denmark's own integration programme, only wage subsidies significantly shortened the time from benefits to work (Clausen et al. 2009). Lock-in is a standard finding in this literature too: taking part in training lowers employment while the course runs, and the return appears only years later. A recent survey puts the range of estimates from zero to 54 %, and notes that displacement — whether the jobs simply move from one group to another — is essentially unmeasured (Kiviholma 2026). Denmark's 1999 language-training reform, by contrast, produced permanent earnings gains and similar effects on the children (Foged et al. 2024).
Intergenerational spillovers of integration policy20245 findings
Pesola & Sarvimäki, Review of Economics and Statistics 2024 (IZA DP 15310). Same 1999 reform discontinuity, register data
The 1999 integration-plan reform increased adult immigrants' language training and improved their labour-market position — and the benefit reached their children too.
The group the reform reached arrived in the late 1990s: about half were born in the former Soviet Union, a quarter are Ingrian Finns and just under a fifth refugees. The result therefore does not speak directly to what follows from integrating those who arrived after 2015.
A parent's participation raised the children's school grades by 0.5 standard deviations and kept them in education more than a year longer.
Two decades after arrival, children of the immigrants covered by the policy earned 42% more than children whose parents narrowly missed it.
The estimate is imprecise. The effect is €7,373 in annual earnings at age 31 against an expected €17,390 for the comparison group, with a standard error of €3,773: the 95% confidence interval runs from close to zero to roughly double. Across bandwidths and estimators the effect ranges between €2,567 and €9,006, the sample is about 1,300 children, and the tests of the mechanisms — the parents' colleagues, the children's schoolmates — are statistically underpowered.
Integration training — performance audit (National Audit Office)20183 findings
National Audit Office of Finland, audit report 15/2018 (follow-up 2022)
Of those who completed integration training in 2013–2016, just under 35% reached language level B1 — roughly two in three fell short of the target.
The audit's core finding: the Ministry of Economic Affairs and Employment did not have enough information on how integration training affects immigrants' language skills or employment.
The only effectiveness metric the ministry tracked — jobseeker status three months after training — was judged inadequate: the unemployment rate alone says nothing about the training's effectiveness.
Working Together — the labour-market integration of immigrants in Finland (OECD)20183 findings
OECD (2018), Working Together: Skills and Labour Market Integration of Immigrants and their Children in Finland, OECD Publishing
The integration-training language courses delivered weak results: in 2016 more than four in five participants failed to reach B1.1 (the level needed for vocational training) and close to 60% failed to reach even A2.2. 28% ended up at beginner level or below.
In 2015, close to two in five integration-training participants fell into unemployment or left the labour force at the end of their training.
Foreign-born women's employment rate trailed comparable native-born women by 51.5 percentage points in year one (28.8 pp for men) and still by 40 pp five years after arrival — narrowing to about 20 pp only after 15 years.
EU-harmonised integration-outcome indicators show the same picture across several domains: the foreign-born trail the native population in education, housing and material well-being. Self-reported health is the exception, and even that is largely explained by their younger age structure. "EU-harmonised" means every member state collects the figures using the same definitions and the same surveys — Eurostat's income and living conditions survey EU-SILC and the Labour Force Survey — so that countries can be compared with each other. Overcrowding and severe material deprivation, for instance, are defined identically across the EU rather than nationally.
NEET youth (15–29) by country of birth · Finland 2004–2025
Immigration is often defended through labour supply and public-finance benefits. Labour migration can be economically positive, but register data and fiscal models show that the positive effects are not seen among immigrant groups of every migration reason or background country. Humanitarian and family migration in particular show up in the data as higher benefit use and less tax paid.
Sections in this group and their figures
Employment by background countryIn 2024 the employment rate of the foreign-born of foreign background was 56.0 percent, against 74.5 percent for the Finnish background together with the Finland-born of foreign background. For the African background it was 49.9 percent.
Employment by grounds for arrivalThose who arrive on a work permit are over 80% employed in their first year. Among people granted international protection the rate starts near 10% and is about 60% after ten years (register data 2011–2021).
Income levelIn 2024 the median income of working-age people was €33,615 for those of Finnish background and €24,060 for those of foreign background. The gap is 28%, exactly what it was in 1995.
Welfare useForeign-language speakers are about 11% of the population but 29.4% of basic social assistance recipients (2024). In 2017 the share was about 19%.
Remittances abroadRemittances paid from Finland grew from about USD 16 million (1990) to USD 1.57 billion (2024), almost a hundredfold.
Cost estimatesThis site's own calculation by the Danish method: the non-Western background population, 529,000 people in 2025, is an estimated annual cost of €5.3…9.2 billion to public finances at average cost, or €10,000…17,300 per person. The range is wide because service use is modelled rather than measured, and the scope includes people who came for work: the protection background alone is €1.4…2.0 billion.
Key figures
What decides net transfers is the reason for migrating: the gap between work-based and protection-based migration was about €19,500 per person a year (register data 2023).
Excluding collectively consumed public services, the net effect on public finances was −€7,900 a year for people born in Somalia, +€3,400 for the native-born population and +€5,100 for people born in Germany (2011 data).
Over a lifetime the net effect on public finances of a person who moved from Somalia was −€951,000 without children and −€1,343,000 with them. The spread is wide, ±€801,000 (2019).
Labour shortage is the usual justification for work-based immigration, but the domestic labour reserve is large. In June 2026 Finland had 297,000 unemployed. A further 357,000 or so people were studying at universities and polytechnics, on their way into the labour market.
The question is why labour is recruited from abroad while one in ten of the labour force is out of work. The counterargument is mismatch: open jobs, the skills of the unemployed and where people live do not always meet.
On the statistics, though, recruiting labour from abroad does not settle the matter by itself — what decides it is selection. In this section's own figures, work-based migration is the only group clearly positive for the public finances, and the spread to the weakest is roughly €19,500 a year depending on the grounds for entry. These are a single-year (2023) cross-section, not net effects computed over a lifetime: the lifetime estimates on the background-country pages are a different quantity.
One qualification applies to every figure computed by grounds: it says what one person pays and receives under their own permit, not what the family arriving with or after them pays and receives. A work-permit holder appears in the statistics under work, their spouse and children under family. The same data shows the order of magnitude: attaching one family-ground person at that group's average turns the work-based surplus of €971 negative. The calculation is rough, because the family-ground average covers every family reason including the reunification of people granted protection, and not every work-permit holder brings a family. The combined effect of work-based migration together with its attached family has not been measured in Finland, so neither direction can be read from these figures: they neither establish that work-based migration is positive once families are counted, nor that it is negative.
This section treats fiscal effects at a general level. Closer figures on what the integration system and the reception system cost are in this knowledge base, each in its own section.
A bar to the right means the group pays more in taxes than it receives in transfers, a bar to the left the opposite. The figure is an average per person per year.
Net income transfers of the foreign-background population by reason for immigration (2023)
Average € per year, aged 20+
Work
+971 €
Born in Finland
−728 €
EU residence right
−3,593 €
Study
−3,623 €
Citizenship
−4,040 €
Permanent residence permit
−5,874 €
Other / unknown
−6,852 €
Family
−9,989 €
Protection (humanitarian)
−18,537 €
Compared with native-background residents of a similar age and sex distribution. Positive = net contributor (pays more in taxes than it receives in transfers). Negative = receives more in transfers than it pays. Work-based migration is the only clearly positive group.
The figure is taxes paid minus transfers received, on average per year, for people aged 20 and over. A positive figure means the group pays more than it receives.
The work-based plus (+971 euros) is close to zero, and work-based migration typically brings family reunification with it, which is recorded in its own clearly negative group (−9,989 euros). Protection-based migration comes to −18,537 euros.
This is a single-year (2023) cross-section, not a net effect computed over a lifetime.
A bar to the right is a net contributor, a bar to the left a net recipient. The figure is an average per 20 to 62 year old per year.
Net public-finance effect by country of birth (2011)
Average € per year per 20–62-year-old
Somalia
−13,850 €
Iraq
−13,360 €
Thailand
−6,560 €
Former Yugoslavia
−5,680 €
Turkey
−4,240 €
Former Soviet Union
−3,750 €
All foreign-born
−3,620 €
China
−3,280 €
Sweden
−1,690 €
Estonia
−1,230 €
Finland (native-born)
+80 €
Germany
+2,340 €
Includes collectively consumed public services, so the figures are more negative than net income transfers alone (excluding collective services: Somalia ≈ −€7,900, native-born ≈ +€3,400, Germany ≈ +€5,100). For Finland the 95% confidence interval is shown: −€173…+€339. A negative figure means the group received more in public spending than it paid in taxes. The publisher, Suomen Perusta, is the Finns Party's think tank.
The data year is 2011 and the report was published in 2015. It is by far the oldest figure in this section, and it is not directly comparable with the VATT 2023 calculation above: different year, different grouping (country of birth vs. reason for immigration) and different method.
The publisher, Suomen Perusta, is a think tank associated with the Finns Party. The calculation uses public register data, but it is not official statistics and not peer-reviewed.
No newer figure has replaced it, which is a different thing from its being out of date. Suomen Perusta has not repeated the calculation: since then it has published a lifecycle projection for two countries (2019) and municipal statistics on 2022 data (2025), but no new net effect by country of birth. VATT's Datahuone project for the Ministry of Finance promises a breakdown by reason for migration and country of origin, with results due in early 2027. Until then this is the only Finnish calculation that puts a net figure on an individual country of birth and counts collectively consumed public services in it.
VATT Datahuone: incomes and transfers received by people with a foreign background (report 1/2026)20262 findings
Commissioned by the Ministry of Finance, register data for 2023. Aged 20 and over, standardised against people of Finnish background with the same age and sex distribution. The euro figures by ground of residence permit are read off the report's chart, not its body text
The euro figures in the chart show how much each group's net transfers deviate from the average for comparable native Finns of the same age and sex, not the group's absolute net transfer position. In absolute terms, both the work- and EU-residence-based groups had positive net transfers according to the report, while the native-Finnish-background average was negative.
The difference arises above all from the reason for migration: work-based migration is fiscally close to zero or positive, while humanitarian and family-based migration are clearly negative. The overall picture is less negative once the younger age structure of the foreign-background population is taken into account.
Why the cost estimates for immigration differ from one another20266 findings
Comparison: Salminen's calculation (Suomen Perusta) and Sarvimäki's alternative calculation in the VATT blog post “On the costs of immigration”, plus Pauli Vahtera's calculation (2019, 2017–18 data)
The figures in public circulation do not contradict each other. They measure different things. The differences come from three choices: which population is counted, which costs are attributed to immigration, and whether the calculation is a single-year snapshot or a whole life cycle.
Broad attribution (Salminen, Suomen Perusta): public-goods costs are assumed to rise by EUR 3,181 per immigrant per year — the total cost of public goods divided by Finland's population (about EUR 3,100 per head) plus the Immigration Service's expenditure divided by the number of foreign citizens (EUR 81 per foreigner). On that assumption the total cost is some EUR 700 million a year.
Narrow attribution (the alternative calculation in Sarvimäki's VATT blog): the cost of public goods (defence, infrastructure, interest) is not attributed to immigrants at all. Only the Immigration Service's expenditure is counted (EUR 81 per foreigner). That yields an average net cost of EUR 520 per person per year and a total of some EUR 107 million a year.
NOTE: EUR 3,181 and EUR 520 are not comparable with each other. The former is an input assumption about the cost of public goods. The latter is an output — the average net cost. The comparable figures are the totals, EUR 700 million and EUR 107 million.
Vahtera: about EUR 3.2 billion a year. It covers humanitarian immigration only, not immigration as a whole, and includes items such as the costs of crime. The calculation was made in 2019 from 2017–18 data.
All three are cross-sectional calculations. VATT itself notes that a dynamic life-cycle calculation would be more sensible: an immigrant arrives with the costs of childhood already paid elsewhere, may leave before old age, and integration takes time before full labour-market participation.
ETLA (Kauhanen 2026): economic effects of immigration — structure and integration are decisive20266 findings
Kirjallisuuskatsaus: Pesola ym. 2024a (Migri + Tilastokeskuksen rekisteriaineisto 2011–2021, 18–59-vuotiaat) ja Ruotsin Konjunkturinstitutet 2025a/b
In Finland (Pesola et al. 2024a) the residence-permit basis is decisive for employment: work-based immigrants have an employment rate above 80% already in year one. EU citizens go from about 60% to 80% over ten years, students from about 25% to 70%, family-tie immigrants from about 10% to 70%, and those granted international protection from about 10% to 60%.
Income gaps outlast the employment gaps: those granted international protection earn only about 3% of the native Finnish-background income level in their first full year, still under half after ten years.
About 17% of immigrants leave Finland again within ten years — nearly a quarter of students, but only about 8% of those granted protection.
Finland's own register data (Salminen 2015) confirms the direction of the Suomen Perusta report: in 2011 the net fiscal impact of German-born immigrants was clearly positive — even higher than for the native-born — while Somali- and Iraqi-born immigrants had a markedly negative impact.
Sweden's long-run data (Konjunkturinstitutet 2025a) puts the same pattern in numbers: refugees' net fiscal impact averages SEK –27,500 per person per year (SEK –22bn for the group as a whole) and turns positive only after about 15 years of residence. Only about 40% of working-age refugees are ever net contributors, versus over 50% for other foreign-born and over 70% for the native-born. In the same table the average net effect for Sweden-born residents is also slightly negative (SEK −700 per person), so in a static calculation a negative net effect is not something particular to immigrants.
Integration policy has a measurable payoff: Finland's 1999 individual integration plan raised participants' cumulative ten-year earned income by 47% and cut paid social benefits by 13% (Sarvimäki & Hämäläinen, 2016).
VATT — the fiscal impact of immigration on public finances (ongoing)20254 findings
A study commissioned by the Ministry of Finance, begun in February 2025
VATT (Finland's Government Institute for Economic Research) is currently studying the overall fiscal impact of immigration.
Results are expected no earlier than 2027.
The study covers direct costs (reception, integration, social security) as well as tax revenue and labour-market effects.
A figure that has circulated publicly: about €–520 per person per year, or ~€107M a year in total. It is not VATT's own estimate but an alternative calculation presented by a VATT researcher (Sarvimäki) on his blog from Salminen's (Suomen Perusta) data, and it does not cover lifecycle effects.
Pauli Vahtera (authorised public accountant) – cost estimate for humanitarian immigration20194 findings
A calculated estimate. The author is an authorised public accountant and columnist, not a research institute. 2017–18 data, calculated in 2019.
Vahtera's estimate for the annual total cost of humanitarian immigration: about EUR 3.2 billion. The figure does not cover immigration as a whole.
The calculation covers social security, health care, education and administrative costs against the taxes paid by immigrant groups, plus the costs of crime.
Vahtera has himself said the available data was incomplete and that he chose a conservative approach. For how this compares with VATT's calculation methods, see the previous item.
He also gives cumulative figures: EUR 46 billion for humanitarian immigration over 1990–2018, and EUR 79 billion — 76% of the state debt — once development aid and an imputed 2% interest rate are added. Both rest on the assumption that spending rose evenly from zero to the 2018 level, and the interest calculation assumes all of it was debt-financed.
Immigration and Finland's public finances - lifecycle effects20198 findings
Municipality-level register data. The lifecycle figures are discounted estimates and are sensitive to the assumptions made about employment and return migration.
Somalia: net contribution excluding all services €–7,900 per person. Total net effect as low as €–13,850 per person. NOTE: these annual figures are not in the linked lifecycle summary but come from an earlier Suomen Perusta analysis based on 2011 data. From the same source, a native Finn contributes +€3,400 a year and a person born in Germany +€5,100.
Iraq: similarly strongly negative.
Immigrants from Western countries (Germany, Sweden) are close to zero or positive.
Discounted lifecycle calculation: an immigrant from Somalia €–951,000 (without children), with children up to €–1,343,000.
An immigrant from Iraq: €–690,000 (without children), with children €–844,000.
The report also gives standard deviations for its projections, and they are large: Somalia ±€801,000 (±€1,769,000 with children) and Iraq ±€623,000 (±€1,227,000). The figures therefore say nothing about any individual's lifetime effect.
Labour-based immigration (from Western Europe) is close to zero or positive.
The publisher, Suomen Perusta, is the Finns Party's think tank and has its own position on immigration policy. The calculation still rests on register data, and its results point the same way as those of the Danish Ministry of Finance and ETLA.
Immigration threatens Finland's pension system
The Helsingin Sanomat guest column and article ask whether immigration endangers the long-term sustainability of the pension system.
An HS interview: Suvi-Anne Siimes, chief executive of Tela, the earnings-related pension providers’ interest group, argues that immigrants qualify for Kela pensions too easily and that tightening the conditions would encourage people to come to Finland to work at a younger age rather than arriving in old age to draw a pension. Entitlement to the guarantee and national pension arises after three years’ residence, and the full guarantee pension is about EUR 990 a month — the same amount for immigrants and the Finnish-born alike. This is the position of an interest group’s chief executive, not a research finding: the piece presents no calculation of immigrants’ pension contributions.
The most expensive reception centre, in Helsinki's Punavuori, cost €130 per person per night in 2024. The cheapest (Salo, Halikko) cost €25. The national average has fallen to €35, from €66 at the 2015 peak. Food at Punavuori costs about €25 per person per day, or about €36,000 a year for a family of four. The unit costs €5 million a year in total: €1m rent, €1m food, €1m staff and about €1m in outsourced health and social services. The Finnish Immigration Service plans to move the unit by 2028 to improve cost efficiency.
AN OPINION COLUMN, not reporting and not research: the author, Simon Hankinson, is a senior research fellow at the Heritage Foundation and the piece ran on the Telegraph's comment pages. It relays Suomen Perusta's figures at second hand — the same ones this site's own chart uses directly — and the article itself discloses that Suomen Perusta is linked to the Finns Party. Its checkable factual claims hold: Finland receives about 50,000 new arrivals a year, and per the OECD about half of the 2024 arrivals came to join family members. It also cites Minnesota figures (81% of Somali-headed households use some federal welfare, against 21% of native-headed households) — a different country and a different welfare system, not comparable with the Finnish numbers. The article does not present Finland as Europe's clearest example. It says that Finland's findings echo those from Denmark and the Netherlands.
HS on a report by VATT, the Finnish government's economic research institute, comparing income taxes and transfers in 2023 with age and sex held constant. Working-age Finnish-background residents paid clearly more income tax and received less in transfers than a comparable foreign-background population, and foreign-background residents received clearly more unemployment benefits and social assistance. At the same time, working-age foreign-background residents paid more income tax on average than they received in transfers. Lead researcher Matti Sarvimäki says that, for direct transfers, the effect of all immigrants on public finances is currently roughly zero. The spread between groups is wide: people on work-based residence permits had better net transfer positions than comparable Finnish-background residents, while those who arrived on protection or family grounds were clearly the most negative. Note that the study covers direct transfers and income taxes, not public services and not the whole life cycle.
Denmark
Average net contribution to public finances by age · Denmark 2018
The figures cover Danish public finances and the Danish tax and benefit system, not Finland's. They do not transfer to Finland as they stand. The shape of the age profile is what the chart is for.
The unit is DKK 1,000 per person (DKK 1,000 is roughly 134 euros).
The upper bar is the actual net effect, the lower the same age-standardised, that is with the group's age structure set to the whole population's. A bar to the left is a cost.
Actual and age-standardised average net contribution · Denmark 2018
DKK 1,000 per person
ActualAge-standardised
Western immigrants
+38+2
Western descendants
−72+8
MENAPT immigrants
−85−118
MENAPT descendants
−110−58
Other non-Western immigrants
−4−60
Other non-Western descendants
−108−33
Age-standardisation recomputes each group's net contribution as if it had the same age distribution as people of Danish origin. For Western descendants the negativity is an age effect (positive once standardised), whereas for MENAPT immigrants the figure is even lower after standardisation (−85 → −118) — so the gap is not explained by age structure. Values read from the report's figure 2.8.
Age standardisation shows what the net contribution would be if the group had the same age structure as the native population. The gap between actual and standardised is the part that comes from age structure alone.
As above: the Danish system and Danish kroner, not Finland's.
A bar to the right is a net contributor, to the left a net recipient, in thousands of Danish kroner per person per year.
Average net contribution to public finances by country of origin · Denmark 2019
DKK 1,000 per person per year
Somalia
−130
Syria
−118
Lebanon
−102
Iraq
−90
Afghanistan
−79
MENAPT countries, average
−74
Eritrea
−71
Morocco
−46
Former Yugoslavia
−45
Pakistan
−34
Turkey
−32
Bosnia and Herzegovina
−26
Iran
−24
All immigrants
+5
Sri Lanka
+6
Philippines
+10
Vietnam
+12
Bulgaria
+12
Other non-Western, average
+14
Hungary
+15
Thailand
+15
Latvia
+18
Danish origin
+22
Iceland
+24
Romania
+31
Russia
+34
Norway
+34
Poland
+38
Germany
+42
Lithuania
+46
Ukraine
+48
China
+51
Western countries, average
+52
Spain
+57
Sweden
+69
Italy
+70
India
+94
Netherlands
+110
United States
+110
France
+114
United Kingdom
+114
The group averages and the Danish-origin level come from table 2.1 of the same publication, so every country reads against its own group and against the native population. Western, MENAPT and other non-Western are the ministry's own grouping: Denmark counts EU countries as Western, so Poland, Romania and Bulgaria are Western while Russia and Ukraine are not.
The figures cover Danish public finances and the Danish tax and benefit system, not Finland's. They do not transfer to Finland as they stand. The unit is DKK 1,000, which is roughly 134 euros.
Immigrants only. The ministry counts descendants separately and they are not in these figures. That matters, because in the same table the descendants' average is −102 against the immigrants' +5, largely because of their young age structure.
Only countries with at least 5,000 resident full-year persons in Denmark in 2019 are included. The list is therefore the 36 largest origins, not all of them.
This is a single year's cross-section, not a lifetime figure. Age structure feeds straight into it: a group with many working-age people looks better than one with many children or pensioners. The same publication's age-standardisation is in the chart above.
Danish Ministry of Finance: immigrants' net effect on public finances in 201820186 findings
Finansministeriet (the Danish Ministry of Finance), register data (Danmarks Statistik), 2018 levels
In 2018, non-Western immigrants and their descendants were a net cost to the public finances of about DKK 31 billion (some EUR 4.2 billion). Western immigrants and their descendants were by contrast net contributors, at about +DKK 7 billion.
Of the non-Western net cost, the MENAPT countries (Middle East, North Africa, Pakistan, Turkey) accounted for about DKK 24 billion and other non-Western countries about DKK 7 billion.
Average net contribution per person in 2018 (DKK 1,000): Danish origin +8, Western immigrants +39, MENAPT immigrants −85. The raw figures for descendants (efterkommere) are negative largely because of their young age structure (school pupils and students). Age standardisation changes the picture (see figure 2.8).
Of working-age (25–64) MENAPT immigrants, 35% were net contributors, against 53% of other non-Western immigrants.
Even after standardising for age and sex, the average net contribution of MENAPT immigrants remains clearly below that of people of Danish origin — so the gap is not explained by age structure alone.
Finansministeriet has produced the same analysis annually since 2015. The non-Western net cost fell by DKK 2 billion from 2017, reflecting among other things a rise in the employment rate.
United Kingdom
Dustmann and Frattini 2014: the fiscal effects of immigration to the UK20146 findings
The Economic Journal 124, University College London. The whole population over 1995–2011, in 2011 pounds.
Over the whole period from 1995 to 2011, immigrants from EEA countries paid 4 per cent more into the system than they received from it, a net GBP 8.8 billion. Over the same years the native-born paid in 93 per cent of what they received, a net GBP −604.5 billion.
Immigrants from outside the EEA were a net cost across the whole period. The authors put part of that down to age structure: they have had more children than the native-born, and the education spending on those children is charged to them, while their own education was paid for by the country they came from.
For those who arrived after 1999 the picture differs. Between 2001 and 2011 the EEA arrivals among them paid 34 per cent more than they took out, a net GBP 22.1 billion, and the non-EEA arrivals 2 per cent more, a net GBP 2.9 billion. Some 25 billion together, over a period in which the state itself ran a deficit.
The result turns on how shared spending is divided, and the authors compute it both ways. Pure public goods such as defence are not used up in consumption, so one more resident adds nothing to their cost. Congestible services such as fire protection, waste management and water supply are used up, and those are charged at average cost.
When defence and the like are charged to the native-born alone, immigration saves them money by spreading a fixed bill across more people. The saving was GBP 4.7 billion in 1995 and GBP 15.8 billion in 2011.
The authors say themselves that they take a cautious stance, one that is likely to overstate the net fiscal cost of immigrants. The figures cover the British tax and benefit system and do not transfer to Finland as they stand.
Europe
The bar is one immigrant's lifetime net effect on public finances including children, in thousands of euros. Left is a cost, right a gain.
Lifetime net contribution to public finances by region of origin · Netherlands
EUR 1,000 per immigrant, children included
Horn of Africa and Sudan
−606
Morocco
−542
Afghanistan, Iran, Syria and Iraq
−418
Central Africa
−382
West Africa
−348
Turkey
−340
Caribbean
−321
North Africa (excl. Morocco)
−319
Aruba and former Netherlands Antilles
−254
Pakistan
−238
Arabian Peninsula, Jordan and Lebanon
−224
Suriname
−185
Former Soviet Union (excl. Baltic states)
−177
Former Yugoslavia and Albania
−161
Thailand, Indochina and Myanmar
−159
Central America and other South America
−118
East Africa
−98
Poland and the Baltic states
−71
Bulgaria and Romania
−70
Philippines, Malaysia, Brunei and East Timor
−66
China, Mongolia and North Korea
−47
Brazil, Argentina, Paraguay, Uruguay and Chile
−34
Indian subcontinent (excl. Pakistan)
−27
Portugal
−27
Indonesia
−24
Greece and Cyprus
−13
Hungary, Czechia, Slovakia, Slovenia and Croatia
−6
South Korea, Taiwan, Hong Kong and Singapore
+14
Germany and Austria
+23
Spain
+39
Italy and Malta
+50
Israel
+58
Belgium and Luxembourg
+63
Native Dutch reference
+95
Southern Africa
+158
France
+165
Oceania
+166
EFTA, microstates and crown dependencies
+182
United Kingdom and Ireland
+191
Japan
+194
North America
+203
Denmark, Sweden and Finland
+208
The figure is the taxes and contributions an immigrant pays minus the public spending that falls on them, over the whole remaining life course from arrival. It also includes the life course of their children born in the Netherlands and the probability that the immigrant leaves again. The calculation covers the immigrants living in the country in 2016, with future years discounted to their 2016 value. The native Dutch reference is the authors' hypothetical immigrant with the schooling, labour-market position and health-care costs of the average native Dutch person. It is therefore not the native population's average. A native Dutch person born in the Netherlands has, according to the authors, a lifetime net contribution of roughly zero on average.
The figures cover Dutch public finances and the Dutch tax and benefit system, not Finland's. They do not transfer to Finland as they stand.
The regions of origin are the authors' own 42-region division, and even the smallest group has at least 5,000 people. The former Yugoslavia group excludes Slovenia and Croatia, which are grouped with Hungary and Czechia. The Brazil group also includes French Guiana.
People who moved from Finland are in one group with Denmark and Sweden. They are not broken out.
The figures are from the baseline scenario: a real discount rate of 2.5%, productivity growth of 1% a year and a state pension age rising to 70. Shared spending such as defence is divided equally between all residents. The same table has four variants. Dividing shared spending partly by income narrows the differences, and a discount rate one percentage point lower widens them. The order stays almost the same in every variant. No region moves more than two places, and the Horn of Africa and Sudan is last in each one.
IZA discussion papers are preliminary and not peer-reviewed. The authors are the same as those of the Borderless Welfare State report, and the data are the same Statistics Netherlands register data.
The Danish chart above is a single year's figure per person, this one is a lifetime figure per immigrant. The values cannot be compared directly, only the ordering.
Borderless Welfare State20233 findings
Anonymous registry data for 17 million Dutch residents (1995–2019)
Immigration cost the Dutch public sector about EUR 400 billion net in 1995–2019, averaging about EUR 17 billion per year and peaking at EUR 32 billion in 2016.
Immigrants use more social security, education, and other benefits than the native population and pay less in taxes and social-security contributions.
Work and study migration from most Western countries and from a number of non-Western ones, East Asian above all, was a net gain for the public finances. All other immigration was at best roughly budget-neutral, and family and asylum migration were the most clearly negative.
United States
NASEM 2016: the economic and fiscal consequences of immigration20166 findings
A panel of the US National Academies of Sciences, Engineering, and Medicine, a 508-page report. Lifetime calculations over a 75-year horizon, in 2012 dollars.
A new immigrant's net effect on the public finances is USD 259,000 when the calculation runs on the Congressional Budget Office's long-term path and shared spending such as defence is not charged to them. Of that total, 173,000 is the immigrant's own and 85,000 their descendants'.
The same person turns into USD −36,000 if the budget path is switched to one with no fiscal adjustment at all and shared spending is charged to them at average cost. In between lie 173,000 and 77,000. It is the same calculation and the same person, only two assumptions.
The marginal cost of pure public goods from one more resident is, in the report's words, “at least in the short run, zero or close to it”. Shared spending is almost a third of federal spending, so the report treats the choice between average and marginal cost as quantitatively very important.
The choice also reverses the order of the generations. When shared spending is divided among everyone at average cost, the first generation looks slightly weaker than the third-plus generation at the federal level. When it is charged to the native-born only, the order reverses.
As a cross-section, the first generation costs the public finances more than the native-born, mostly at state and local level, because that is where the education of their children is paid for. The children of immigrants, the second generation, are in the report's words among the strongest economic and fiscal contributors in the population.
The site's own lifetime calculator uses this same split and puts the choice in front of the reader as a switch rather than making it for them.
Employment by background countryEconomySocial cohesion
Key figures
In 2024 the employment rate of the foreign-born of foreign background was 56.0 percent, against 74.5 percent for the Finnish background together with the Finland-born of foreign background. For the African background it was 49.9 percent.
Employment differences between background countries are wide, and employment is the central measure of integration. The figures cover everyone with a foreign background, by background country — they do not separate work, family and protection arrivals, so the differences between countries largely reflect which countries people arrive from on which grounds. The split by ground is under grounds for arrival, and where employment leads in earnings under income level.
Statistics Finland's integration indicators break employment down by region of background. The gap to the native population has narrowed over thirty years without closing, and the differences between regions of origin are larger than the average gap to the native population.
Employment rate by origin and region of background · Finland 1995–2024
% of 20 to 64-year-olds
Finnish background and Finland-born foreign backgroundForeign-born, foreign background, totalBackground in other EU27 countriesBackground in AsiaBackground in Africa
Details
The group “Finnish background and Finland-born foreign background” combines the native population with the second generation, because the statistic does not separate them.
Region of background is a coarse split: Africa, for example, covers both refugee-background people and those who moved for work.
A register-based employment rate is not the same figure as the labour force survey's, so it is not comparable with the site's other employment figures.
Show the figures
Year
Finnish background and Finland-born foreign background (% of 20 to 64-year-olds)
Foreign-born, foreign background, total (% of 20 to 64-year-olds)
Background in other EU27 countries (% of 20 to 64-year-olds)
Occupation group says where in the wage distribution a group works. “Elementary occupations” is the lowest-paid major class in the classification: cleaning, kitchen assistance, refuse collection and labouring. Of the Finnish-background employed 5.3% work there, of the Philippine-background 43.5%.
Share of the employed in elementary occupations, by background country · 2023
Occupation group is a proxy for pay, not pay itself. Statistics Finland publishes no earnings by background country: neither the structure-of-earnings statistics nor the index of wage and salary earnings has an origin dimension at all.
The share is of the group's own employed, so it says nothing about how much of the group is in work. Employment rates are in the other figures in this section.
Groups with fewer than a hundred employed are left out.
The pay gap and the occupational structure
The pay gap between foreign-background and Finnish-background employees is clear overall but small within the same occupation group.
The median hourly wage was EUR 16.75 for foreign-background employees and EUR 19.61 for Finnish-background employees in 2022, 14.6 per cent lower. Within an occupation group the difference is small and in places reversed: among managers the foreign-background earned 8.3 per cent more (EUR 47.07 against 43.46), among specialists 3.7 per cent more, and among professionals pay was about equal. Among service and sales workers the gap was 3.4 per cent in favour of the Finnish-background. The overall gap therefore comes mostly from which occupations the groups work in, not from pay for the same work.
Employed share by background country and sex
In 2023 the employed share of 20–64-year-olds was 79% among people of Finnish background and 84% among those who moved from the Philippines, against 46% for Afghanistan, 42% for Iraq and 29% for Syria. The figures are by background country and do not separate the grounds a person arrived on, so the same group holds both work-based and protection-based arrivals. The gap between the sexes is wide in many groups: for those who moved from Afghanistan the employed share was 57% for men and 30% for women. The largest gaps concern the employment of women from Muslim-majority countries, whereas women who moved from the Philippines or Thailand, for example, are employed at the same rate as men.
The figures in this section and its chart are employed persons relative to the population aged 20–64, by background country and sex (tables 115t and 11rs). They are not official employment rates, because the employed include people of all ages. Comparison across groups is still meaningful, as the same definition applies to every background country.
The labour shortage and the grounds for arrival
The employment figures sit in tension with the labour-shortage argument. Work-based and humanitarian migration are different things, and the difference shows up precisely in employment: people from the large refugee source countries are employed well below the native rate, while those who arrived on a work permit reach it and in the early years exceed it – at the same time as Finland has about 297,000 unemployed. Immigration does not fix the problems of domestic education and sectoral workforce planning, and it can also mask them in a way that leaves them unfixed.
Työ- ja elinkeinoministeriö (TEM 2024:9) · Pesola, Sarvimäki & Virkola2024
Summary
The study links the Finnish Immigration Service's residence-permit decisions for 2011–2021 to Statistics Finland's registers, so it can follow employment by the grounds on which a person arrived. That answers directly whether work-based migrants are employed as well as the native population. They are, and at first better: in the early years of residence, employment and earnings among work-permit holders are higher than among Finns of the same age and sex, though the difference disappears fairly soon. After ten years the employment rate is about 80% for work-permit holders and EU citizens, about 70% for those who came on family or study grounds, and about 60% for those granted international protection. Earnings diverge more than employment does: at the ten-year mark work-permit holders sit slightly below their native counterparts, EU citizens and former students reach a little over 80% of native earnings, while protection and family arrivals remain below half. Two numbers are worth reading separately. The first is the starting point: at the end of the first year of residence only one in ten of those who arrived on protection or family-reunification grounds was in work, and protection arrivals earned 3% of Finnish-background earnings in their first full year in the country. The second is the closing of the gap: after ten years the difference in employment between work-permit and protection arrivals is down to 20 percentage points, against more than 80 in the first year. The groups converge from both directions, because work-permit holders' employment also declines over the decade. The authors' own most surprising result is that people who arrived to study do not reach the income level of Finnish-background people of the same age, even though they study mostly at universities and live in the country's most prosperous regions.
Emigration is dealt with in the study and does not explain the differences. Within ten years 24% of study-permit arrivals leave the country, 19% of work-permit arrivals, 12% of family-reunification arrivals and 8% of those granted protection — that is, the groups with the best figures are the ones that shed the most people. The authors tested this by repeating the calculation on those who stayed for the whole period, and the results were very similar. What does belong with the figures is that the ground is the one a person first arrived on: a student may move to a work permit and a work-permit holder to permanent residence, so the groups are not the same people in the same positions a decade later. The data also misses 15% of permit holders who have no personal identity code in the registers, and the gap is largest precisely among study and work permits.
Work-based applications fell 25% and study-based applications 4% in 2025 from the year before, attributed to Finland's weak economy. The largest applicant countries were India, the Philippines, China, Vietnam and Thailand. About 180,000 applications were received in total, and 80% were approved.
According to an Etla report, the shrinking working-age population and public finances would require annual net migration of about 44,000, roughly a quarter more than the 2025 outcome of 34,852. The report names high taxation, negative attitudes and international competition for skilled workers as obstacles. The economic effects of work-based immigration arrive faster than those of humanitarian migration.
Discrimination explains part of it, but not the ranking
Three explanations are offered for the employment gap: discrimination, the level of benefits, and differences in skills. Discrimination has been measured. In a field experiment where identical CVs were sent to a thousand vacancies with only the applicant's name changed, a Somali name had to send 3.9 times and an Iraqi name 2.9 times as many applications as a Finnish one to win the same number of callbacks, and every applicant was presented as educated in Finland. It is not sufficient on its own, though: measured discrimination does not order countries the way employment gaps do – France records the sharpest hiring discrimination in Europe, yet the OECD puts its employment gap between the foreign-born and the native-born at 7 percentage points against 16 in the Netherlands – and a name does not change with the applicant's sex, although the gap is widest precisely among women.
A meta-analysis of 97 field experiments and over 200,000 applications across nine countries. Discrimination was found everywhere, but its level varies: in France and Sweden a minority applicant has to send 70–94% more applications than a native one, in Germany and the United States 25–40% more. The ordering does not match the employment gaps: in the OECD's Settling In 2023 the gap between the foreign-born and the native-born is 7 percentage points in France and 16 in the Netherlands. The authors themselves note that the results contradict several established macro-level explanations.
A field experiment in which five equivalent applications were sent to each of a thousand vacancies, the only difference being the applicant's name. Winning the same number of callbacks as a Finnish name required 1.45 times as many applications for an English name, 1.71 for a Russian, 2.91 for an Iraqi and 3.94 for a Somali one. For men of Somali background the multiplier was 4.97. Every applicant was presented as educated in Finland, so language or the origin of the qualification does not explain the difference. The experiment measures the first step of hiring, not what happens after a job is obtained.
45 % of respondents had experienced racial discrimination within five years, against 39 % in the 2016 survey. The highest figures were in Germany and Austria, above 70 %. Those same countries measure as low-discrimination in hiring experiments, so perceived discrimination and measured employer behaviour are not the same thing and do not put the countries in the same order. Only 9 % reported the discrimination they experienced to an authority, so official statistics do not measure the extent of it.
Differences between systems order the countries better
In 2024, 65% of the people who had fled Ukraine were in work in Poland and about 27% in Germany, though the group is the same. When Denmark cut benefits for newly arrived refugees by about 40% in 2002, employment rose from 10% to 19% within a year. It came at a price: the effect faded in about five years, and poverty and subsistence crime rose alongside it.
People who fled Ukraine were spread across European countries close to at random, so comparing their employment says something about the systems rather than the group. In 2024, 65% were in work in Poland, 61% in the United Kingdom, 53% in Lithuania, 51% in Czechia and about 27% in Germany. IAB attributes the difference to ease of labour market access, language requirements, how long qualification recognition takes, the availability of childcare, and the level of benefits. The German model puts language training first, which pushes employment later. Three in four arrivals are women.
American Economic Journal: Economic Policy · Dustmann ym.2024
Summary
In July 2002 Denmark cut social assistance for newly arrived refugees by about 40%, leaving earlier arrivals at the old level. The design separates the benefit level from everything else, because the country, the labour market and the employers were the same. Employment rose from 10% to 19% in the first year, but the effects on earnings and employment faded after about five years. Poverty rose at the same time, and subsistence crime, mainly shoplifting, increased over the first five years by 5 percentage points among adults and 12 among adolescents. The cut was repealed a decade later.
The gap arises above all from women
The employment shortfall of people who moved from outside the EU arises primarily from women. That bounds how much employer prejudice can explain: prejudice does not select by sex within the same name. The same pattern shows in this section's own figures for Finland, and the mechanism is documented there as incentive traps.
An unemployment trap is defined as a participation tax rate above 80 %, and an income trap as an effective marginal tax rate above 70–80 %. The highest rates fall on single parents, on families with children where one or both parents are unemployed, and on people living alone. The effect comes from unemployment benefits, general housing allowance, daycare fees and the marginal tax rate acting together. The figures cover the whole population rather than immigrants separately. The mechanism reaches immigrant families more often than average because those families are larger and have lower incomes, not because the rules differ for them.
In the EU, people who moved from another EU country are employed more often than natives and people from outside the EU less often, but the gap arises primarily from women. In Poland, Czechia and Italy, non-EU immigrant men are employed more often than native men. The largest gaps are in the Netherlands, Switzerland, Norway, Sweden and Belgium, countries with comprehensive benefits and a high wage floor.
Entrepreneurship
Immigrants are entrepreneurs more often than the majority population. In 2024, 12.7 per cent of the foreign-born in work were entrepreneurs, against 9.8 per cent of the majority. The largest numbers come from Russia and Estonia, but entrepreneurship is most common among people from Turkey: 29 per cent of those in work were entrepreneurs in 2022. Among people from Nigeria and Iraq it was one in five.
Immigrants' firms are smaller than majority firms, though. In 2021 they made up a tenth of all entrepreneurs' firms but had 7 per cent of the staff and 5 per cent of the turnover. Together they employed about 27,000 people. Before the pandemic their micro and small firms created more new jobs relative to their size than majority firms of the same size. The firms are concentrated in sectors where growth and productivity are typically low. The most common are restaurants, construction, taxis and cleaning. One sector runs the other way: in 2021 nearly a third of the entrepreneurs in scientific research and development were immigrants.
The growth in entrepreneurship since 2018 has come from men, and much of it is platform work. 10 per cent of people of foreign background had done platform work in the past year, against 3 per cent of people of Finnish background. In postal and courier services, which include food delivery, 60 per cent of entrepreneurs' firms belonged to immigrants in 2021. One in five men from Africa in work was an entrepreneur in 2024, against under four per cent of women. A growing number of immigrant entrepreneurs earn no business income at all but a wage instead, which points to umbrella-company self-employment. Such an entrepreneur employs no one else, so part of the high entrepreneurship rate reflects the platform economy rather than business activity.
Tilastokeskus, Tieto&trendit · Anna Pärnänen2023-06-29
Summary
According to the 2022 labour force survey, 10 per cent of people of foreign background and 3 per cent of people of Finnish background had done platform work in the previous year. Courier services, and to a large extent taxi services, rely largely on workers of foreign background.
Register-based entrepreneurs as a share of the employed by origin, region of background and sex, 1995–2024. The majority here means people of Finnish background and people of foreign background born in Finland. In 2024, 21.0 per cent of men and 3.5 per cent of women from Africa in work were entrepreneurs.
Pellervon taloustutkimus PTT (PTT Raportteja 296) · Busk & Alasalmi2025-01-29
Summary
A register study of immigrants' entrepreneurship and firms in 2010–2022. Immigrant means a person of foreign background born abroad, and the comparison group is the majority population. The firm figures cover firms with employees and exclude sole traders, jointly owned firms and primary production. A little over 70 per cent of immigrant entrepreneurs could be linked to a firm, so the figures are on the low side. The 2021 staff figure is not fully comparable with earlier years because the business register changed its method. The study was funded by the Federation of Finnish Enterprises, the Finnish Business School Graduates, the Academic Engineers and Architects, the Finnish Pension Alliance Tela and the Union of Professional Engineers.
Employment by grounds for arrivalEconomySocial cohesion
Key figures
Those who arrive on a work permit are over 80% employed in their first year. Among people granted international protection the rate starts near 10% and is about 60% after ten years (register data 2011–2021).
Of arrivals from the Philippines, 53% got their residence permit on work grounds and 84% are employed. Of those from Syria, 1.7% came on a work permit and 29% are employed (permits 2015–2025, employment 2023).
The grounds a person arrived on explain employment better than almost any other single factor. After ten years of residence the employment rate is about 80% for work-permit holders and EU citizens, about 70% for those who came on family or study grounds, and about 60% for those granted international protection. The figures are register-based, from a Ministry of Economic Affairs and Employment study summarised in Employment by background country.
There is no country-level version of this: no published statistic cross-tabulates the permit ground with employment by country. Both halves exist separately, though, and the chart sets them side by side. The two orderings are nearly the same — of arrivals from the Philippines 53% came on work grounds and 84% of that group is employed, while of arrivals from Syria 1.7% came on a work permit and 29% is employed.
One thing stays out of the permit ground's reach: the gap between the sexes. Of arrivals from India 39% came on work grounds, and the gap between men's and women's employment is still 24 percentage points. The widest gap on the list belongs to Pakistan at 34.5 points, and Pakistan is not a refugee source country. The country-level figure mixes two different groups, though: in India the work permit typically goes to a man and the spouse arrives on family grounds, so a country's work-permit share says nothing about the grounds its women arrived on. In register data, where each person appears under their own permit, the picture sharpens. Ten years after the permit, 83% of the men who arrived on a work permit are employed and 79% of the women, against 83% and 83% for EU citizens, 75% and 64% for study, 79% and 67% for family, and 70% and 49% for international protection. The first year is the sharpest: 14% of the men granted protection had earnings that year, against 4% of the women. The grounds someone arrived on do say something about the gap between the sexes, then – it is small for work migrants and widest for protection – but only once they are read from the person's own permit rather than from their country's average. The policy of work-based immigration began with the 2003 government programme.
The two bars are two different statistics about the same country, not two measurements of the same set of people.
A cross-section mixes the newly arrived with those long settled. This follows the same permit cohorts by years of residence, so it shows what happens to a group over time. Work-permit holders start high and rise little, the others rise a great deal and still do not catch them.
Employment rate by grounds for arrival, first year and tenth · Finland (permits 2011–2021)
% in employment
First yearTenth year
What the figures measure
The figures are the study's rounded summary values. The intermediate years are drawn in the study's own charts and are not published as a table.
The grounds are those of arrival, not the person's current status. Someone who came to study may later work, and someone who came as a family member may enter the labour market, and each stays in the group they arrived in.
The tenth-year point comes from the oldest permit cohorts, those who arrived around 2011, while the first-year point comes from all of them.
For work-permit holders the study gives the first year as “above 80 per cent”, so that bar is a floor.
The reason for coming divides experiences clearly. This is the only data on the site that breaks a result down by why a person came. The average across all immigrants was 43.1 per cent.
People who experienced discrimination, by reason for entry · Finland 2022
% of respondents
Details
This is perceived discrimination, not established cases. The survey measures what the respondent interprets as discrimination.
The high figure for those who came to study is partly explained by the group being young and having lived in the country for a short time.
The survey was run once, in 2022, so there is no time series.
Work-permit share of arrivals and the group's employment · by country
Per cent
Work-permit share of arrivals 2015–2025Employed share of 20–64-year-olds 2023
Details
First bar: first residence permits granted on work grounds, divided by all first permits, asylum decisions and temporary protection decisions in 2015–2025 (Finnish Immigration Service, by citizenship).
Second bar: employed share of 20–64-year-olds by background country in 2023 (Statistics Finland).
Ukraine's figure is low because most arrivals received temporary protection: 18,176 work permits against 89,937 temporary protection decisions.
The chart shows the 13 countries with both figures and more than 2,000 arrivals. The correlation of 0.72 is computed across all 26 countries, not only those shown.
The same calculation for the other grounds gives the mirror image: the protection share correlates with employment at −0.60, study at 0.29 and family at −0.18. Weak employment does not follow family migration as such, it follows protection. Study is nonetheless the fastest-growing ground of all.
The comparison is ecological: it sets two country-level numbers against each other, not the same people. The permit grounds cover those who arrived in 2015–2025, while the employment rate covers the whole background population, including people who moved long ago and people born in Finland. The chart therefore says that groups from countries sending work-based migration are better employed in Finland — not how many individual work-permit holders are in work.
No country-level cross-tabulation of permit ground against employment has been published: the ministry study reports by ground without a country split, Statistics Finland's labour force survey does the same, and Eurostat's 2021 survey module divides the reason for migration only into EU-born and non-EU-born.
The employment gap between the sexes and the work-permit share · by country
Percentage points / per cent
Gap between men's and women's employment 2023 (pp)Work-permit share of arrivals 2015–2025 (%)
Details
The chart shows the ten countries with the widest and the narrowest gaps. They are ordered by the size of the gap, not by the work-permit share. If the grounds explained the gap, the blue bars would rise from left to right. They do not.
India (39% work permits, a 24 pp gap) and the Democratic Republic of the Congo, shown as Congo (3% work permits, a 0.9 pp gap), are the two ends of the same observation.
The gap is the difference in percentage points between the employed shares of men and women aged 20–64, by background country in 2023 (Statistics Finland). Vietnam's gap is −0.1 pp, meaning women are employed slightly more often than men.
The correlations of 0.72 and −0.34 are computed across all 26 countries carrying both figures.
The family share does not predict the gap at all: the correlation is −0.07. The widest gap on the list belongs to Pakistan at 34.5 points, and Pakistan is not a refugee source country — 4% arrived on protection grounds.
Show the figures
Group
Gap between men's and women's employment 2023 (pp)
Employment ten years after the permit · by ground and sex
Per cent
MenWomen
Details
A person counts as employed if they had earnings during the year. The figures cover people who were granted a permit at ages 18–59 between 2011 and 2021, and in each year only those still living in the country.
The gap between the sexes is not the same on every ground: it is 4 percentage points for work permits and 0 for EU citizens, against 21 for protection.
The reference is the average for Finnish-background people in the same data, 83% for men and 84% for women. Those two bars therefore do not move with the years.
The values were read off figures 20 and 21 in the ministry study's appendix, where they are not tabulated, so they are accurate to about a percentage point. The study's own text reports both sexes together only.
The Work and Wellbeing of Foreign-Origin Population survey is the only Finnish dataset that asked the same respondent both the reason for migrating and their country of origin. It has six groups. Family was the most common reason in almost every group. Estonians were the exception, where work outranked family. Two thirds of arrivals from Russia and the former Soviet Union moved for family reasons, and for those from the Middle East and North Africa the need for protection was almost as common a reason as family. The data is from 2014 and has not been repeated, so it does not describe migration after 2015.
Statistics Finland publishes the employment rate by reason for migration from the labour force survey, which is an interview survey. The article finds that after ten years of residence the employment rate is close to the Finnish-background level regardless of the reason for migrating, and that in 2023 people who came to look for work were slightly more often employed than those who had a job arranged. There is no country split. Statistics Finland has itself noted that the employment rate for foreign-background people is systematically higher in the interview survey than in the registers, so these figures cannot be set alongside the register-based country figures used elsewhere on this site.
The 2021 module of the EU Labour Force Survey mapped migrants' labour market position and their reason for migrating across all member states. The reason is crossed only with broad country-of-birth groups, however: 18.5% of those who moved from another EU country had found a job before migrating, fewer among those from outside the EU. No country-level combination of reason and employment is published. The module is repeated every eight years.
Income levelEconomy
Key figures
In 2024 the median income of working-age people was €33,615 for those of Finnish background and €24,060 for those of foreign background. The gap is 28%, exactly what it was in 1995.
The median income of the foreign-background population has fallen three years running: €26,265 (2021) to €24,060 (2024). The reason is a change in the composition of the group, not a collapse in individual incomes.
The bottom income decile holds 24.8% of the foreign-background population and 9.7% of the Finnish-background one. In the top decile the ratio reverses: 5.1% and 12.9% (2024).
Debate about the economic effect usually turns on what immigration costs. The other half is the income level: what the foreign-background population earns.
Among the working-age population (18–64) the median equivalised disposable income in 2024 was €33,615 for Finnish-background and €24,060 for foreign-background residents. The gap is 28% — practically the same as in 1995, when it was 28.0%. Both groups' incomes have risen over thirty years. The distance between them has not narrowed.
Since 2021 the foreign-background median has fallen three years running (€26,265 → €24,060), considerably faster than the Finnish-background one. Composition explains much of it: new arrivals start lower, and from 2022 the group has taken in a large number of people who fled Ukraine. Behind the income level is employment, under employment by background country, and its other side under welfare use.
Finland
Median income by origin, aged 18–64 · Finland (1995–2024)
€ per year
Finnish backgroundForeign backgroundForeign background, born abroad
Details
Equivalised means the household's income is scaled to its size and composition. The figure is not a wage but the household's disposable income divided by consumption units.
The chart covers ages 18–64. Across all ages (€31,374 for Finnish background and €22,612 for foreign background in 2024) the comparison is partly a comparison of age structures: the foreign-background population has more children and fewer pensioners.
Only 17,507 working-age foreign-background people were born in Finland in 2024, so their figure swings from year to year. The chart therefore shows those born abroad separately (410,257 people), who make up the great majority of the group.
The median is a single number from the middle of the distribution. This shows where the groups sit: if a group were spread evenly across the whole population’s income distribution, every decile would hold 10 % of it. The Finnish-background population is close to that and leans slightly upwards. Half of the foreign-background population is in the lowest third, and the lowest decile holds 24.8 % of it, two and a half times the Finnish-background figure.
Income deciles by origin, aged 18–64 · Finland (2024)
% of the group
Finnish backgroundForeign background
What the figures measure
The deciles are cut from the whole population’s income distribution, not from the group’s own. Each bar is the share of that group falling in that tenth, so each group’s bars add up to a hundred.
The measure is equivalised disposable income, so household size is accounted for and transfers are included.
Foreign-background people born in Finland are distributed almost identically to those born abroad, 49.1 % and 49.9 % in the lowest third. That group is young even within the 18–64 range, and earnings are low at the start of a career, so the figure is not a measure of mobility between generations.
The at-risk-of-poverty rate (60% of the median) was 31.4% for the foreign-background population and 12.7% for the Finnish-background one.
It bottomed out at 23.6% in 2020 and has since climbed back to its late-1990s level.
In the income distribution, 24.8% of the foreign-background population falls in the lowest tenth, against 9.7% of the Finnish-background one.
In the top tenth the ratio reverses: 12.9% of Finnish-background residents against 5.1% of foreign-background ones.
The group is not uniform: the median marks the middle, not a claim that everyone is poor — three in ten are above the whole population's median.
Welfare useEconomySocial cohesion
Key figures
Foreign-language speakers are about 11% of the population but 29.4% of basic social assistance recipients (2024). In 2017 the share was about 19%.
Foreign-language speakers receive about 37% of all basic social assistance paid out, over €300 million of a roughly €825 million total (2024).
Kela paid EUR 9.2 billion in core income-security benefits in 2025. A fifth, about EUR 1.8 billion, went to foreign-language speakers. In 2017 the share was 15 percent.
In the immigration debate welfare use is usually treated as one question: do immigrants draw more in benefits than the native population. Kela's figures answer yes, and at the same time show the question is not well put. In 2024 basic social assistance went to 52.6 per cent of Arabic speakers and 8.4 per cent of Estonian speakers, against 4.5 per cent of Finnish or Swedish speakers. The gap between the highest and lowest foreign-language group is sixfold, while the gap between the lowest and the native population is twofold.
Welfare use is the flip side of employment: it shows how much of a group depends on benefits for its livelihood. The foreign-language share has grown since the late 2010s, and the gap shows in the euros paid out, not only in the number of recipients. The two directions still part ways: the share of foreign-language adults receiving social assistance has fallen since 2017, yet as the group grows its share of the money paid has risen — to 41% of basic social assistance in 2025. Behind the use of benefits are employment by background country and earnings under income level.
Each row is one benefit in two years: the grey dot is 2017, the blue one 2025 and the line between them is the change. Both figures are printed at the end of the row. The dashed lines are the foreign-language share of the whole population in the same years, 6.8 and 11.4 percent, so a dot to the right of the line means overrepresentation. The button restates the figures as multiples, each year divided by its own year's population share: basic social assistance was 3.5 times the share in 2017 and 3.6 times in 2025, so the share of the money rose while the overrepresentation barely moved.
Share of each benefit's expenditure paid to foreign-language speakers · 2017 → 2025
Kela published these two years only, so there is no series for the years in between.
For four benefits the expenditure share was given for 2025 only, so they carry a single dot: labour market subsidy, guarantee pension, sickness allowance and the national pension.
Foreign-language status is the mother tongue recorded in Kela's benefit register, and the category also holds people whose mother tongue is missing from the register. That inflates the national pension figure in particular (*): in 2025 the unknown-language group halved, which cut counted recipients from 32,000 to 22,000.
A share of expenditure rises from the group growing alone. There were 373,000 foreign-language speakers in 2017 and 646,000 in 2025, which is why each year's population share is drawn as a dashed line.
The population share is calculated on the whole population. Some of these benefits are paid to adults only, of whom 506,000 were foreign-language speakers in 2025.
The euro figures are nominal and not adjusted for inflation.
Labour market subsidy and basic unemployment allowance were replaced by the general allowance on 1 May 2026.
The same data the other way round. The chart above asks how much of a benefit's money goes to foreign-language speakers. This one asks how many foreign-language adults receive it. The blue dot now sits to the left of the grey one, meaning the share has fallen: it fell for six of the eight benefits and held steady for two. Their share of the money still rose, because the group nearly doubled over the same period.
Share of foreign-language adults who received the benefit · 2017 → 2025
The denominator is foreign-language adults, not the whole foreign-language population and not the benefit's recipients. That is why this chart is not read against the one above: the two measures answer different questions.
The sickness allowance figure is rounded in the release, about 5 percent in both years.
(*) The national pension's fall from 9.4 percent to 4.4 is partly a register artefact: the unknown-mother-tongue group halved by 2025, leaving 22,000 counted recipients instead of 32,000.
There were 506,000 foreign-language adults in 2025. As the group grew, the number of recipients could rise even where the share fell.
Labour Force Survey table 15bm (ages 15–64): the unemployment rate among people of foreign background was 17.4% in 2024 against 7.3% among people of Finnish background, a 2.4-fold difference. Foreign-background unemployment was at its lowest in 2022, when it fell briefly to 12.1%. By 2025 the figures were 19.6% and 8.3%.
According to Kela's official statistics, foreign-language speakers made up 29.4% of recipients of basic social assistance in 2024 (2017: about 19%). They received about 37% of the money paid out — over EUR 300 million, against some EUR 825 million of basic social assistance paid in total. Kela reports foreign-language speakers as a single group and does not officially publish shares by language. The breakdown by language is based on a separate data delivery from Kela.
54% of Arabic speakers receive basic social assistance — Kela's last-resort benefit. Among Finnish or Swedish speakers the rate is only 3.7%. Foreign-language speakers make up roughly 11% of the population but account for around 30% of all social assistance recipients.
Kela's 2017–2025 statistics release shows two opposite directions: the share of foreign-language adults receiving basic social assistance fell from 23% to 18% and of labour market subsidy from 24% to 16%, while the group's share of paid social assistance rose from 24% (€172m) to 41% (€408m). Kela paid €9.2 billion in core income-security benefits in 2025, a fifth of it — about €1.8 billion — to foreign-language speakers. In 2017 the share was 15%. Both directions have the same explanation: the foreign-language population grew by over 70% (373,000 → 646,000), and per Kela the growth in recipients since 2023 is mostly Ukrainians. €597 million of labour market subsidy went to foreign-language speakers, a third of the total. Persons whose mother tongue is missing from the register are counted as foreign-language, and the euro amounts are nominal. The labour market subsidy ended on 1 May 2026, replaced by the new general allowance.
Last year Kela detected more than 1,000 suspected benefit misuses worth about €7 million. 471 investigation requests were filed with police. In Sweden Försäkringskassan prevented incorrect payments worth €877.5 million in 2024. Finland recovers about €110 million in incorrect payments. Minister Grahn-Laasonen: 'what is possible in Sweden is possible in Finland' — organised crime systematically exploits social security.
Remittances abroadEconomy
Key figures
Remittances paid from Finland grew from about USD 16 million (1990) to USD 1.57 billion (2024), almost a hundredfold.
Part of the income earned in Finland does not stay in Finland. Remittances sent abroad are the only regularly measured part of that.
Remittances are a billion-scale outflow from Finland. According to World Bank balance-of-payments data, remittances paid from Finland have grown from about USD 16 million (1990) to USD 1.57 billion (2024), almost a hundredfold. That growth belongs to the 1990s and 2000s: since 2010 the series has been essentially flat.
Official statistics cover only formal channels. Informal systems such as hawala fall outside them.
On 8 Feb 2024 the Finnish Government adopted a resolution (VM/2024/27) partially updating the national money-laundering and terrorist-financing risk assessment and action plan. Unregistered hawala operators were added to the action plan for the first time, with the associated money-laundering and terrorist-financing risk assessed as significant. The Financial Intelligence Unit, supervisors, reporting-obligated entities and other authorities are cooperating to identify unregistered hawala operators and pursue criminal liability. This is Finland's own official assessment, unlike the page's other hawala sources, which cover Swedish authorities' risk classification.
Remittances paid from Finland were USD 15.7 million in 1990 and USD 1,574 million in 2024, a hundredfold increase. The growth happened before 2010, though: the figure was already USD 1,411 million in 2010, and the series has been flat since (lowest USD 1,245 million in 2016, highest USD 1,574 million in 2024). The change over the last fifteen years is a few per cent, not an order of magnitude. The statistic covers formal channels only — banks and registered money-transfer services — so cash, sums carried by relatives and hawala fall outside it: the figure is a floor.
Cost estimatesEconomy
Key figures
This site's own calculation by the Danish method: the non-Western background population, 529,000 people in 2025, is an estimated annual cost of €5.3…9.2 billion to public finances at average cost, or €10,000…17,300 per person. The range is wide because service use is modelled rather than measured, and the scope includes people who came for work: the protection background alone is €1.4…2.0 billion.
Figures put on the cost of immigration to Finland run from under a billion to over ten billion euros a year. All of them are called calculations, but they measure different things by different methods, and the spread comes more from the method than from the data.
Finland has no single official total. The most precise figures are per-person net effects computed from register data, and they have not been summed to a national total. The annual billion-euro figures are either sums of budget lines or derived from other countries' results. Below the estimates are set side by side by what they rest on and what they leave out. The most precise figure is neither the largest nor the smallest.
This site's own calculation is included too. It combines VATT's register figures, Statistics Finland's age structure and unit costs of public services the way the Danish Ministry of Finance does. It is run for three different scopes, because the scope moves the figure more than any choice of method. Last come two items other calculations claim are large: the justice system and dwelling prices.
Three annual figures, two of them in circulation and one this site's own. Vahtera's is the gross spending on one group and Virén's a ratio borrowed from another country. Between them sits a calculation whose every step is set out in a source box.
9.2€bn/yrThis site's highest estimate: non-Western background, at average cost
10.5€bn/yrVirén: all immigration, scaled from the Dutch figure
No authority has published an annual net effect of all immigration in euros for Finland. Of these three only the site's own deducts taxes paid in full.
Details
€9.2bn is the top of this site's range for the non-Western background population, 528,567 people, at average cost and by the Danish method. It stands here because the other two headline figures are likewise from the top of their calculations, and a mid-point would not be comparable beside them. The full range is €5.3…9.2bn, and its mid-point of €7.2bn is the figure the section's other charts use. The range is wide because the services are modelled from age structure and unit costs rather than measured.
The state accounts' €0.7bn used to be one of these headline figures. It is not an estimate of the same thing but the outturn of two budget lines, so it is now only in the comparison table.
€3.3bn is Pauli Vahtera's calculation for about 100,000 people from the largest humanitarian countries of origin. It includes imputed items such as a €225m fall in property values, and deducts only €121m in taxes.
€10.5bn is Matti Virén's rough estimate derived from the Dutch total of €400bn for 1995–2019. The method of derivation has not been published.
Sources: mamutus.info September 2026, steps in the source box, Vahtera (Iltalehti 5 Mar 2019), Virén (Suomen Uutiset).
Seven figures that all answer the same question: what immigration costs public finances in a year. They differ more in method than in data, and the quality tag says which is which.
Annual cost for the whole country
Annual costTaxes deductedPublic services includedChildren and lifecycleOfficial or peer-reviewed
Mamutus 2026non-Western background, 529,000 people, the Danish methodestimate€5.3…9.2bnestimate€5.3…9.2bn✓✓~✗
Public services at same-age use€5.4…6.5bn
Health, education and other services at the use of same-age natives, raised by the group's measured extra use.
Other spending, ten small items€0.7…0.8bn
Interpreting, the justice system, the Immigration Service, grants to organisations and other named items.
Minus taxes and employer contributions€2.2…4.3bn
The group's taxes net of transfers, plus employer contributions. The low end pairs the smallest spending with the largest revenue, the high end the other way round.
Marginal cost€1.7…5.1bn
Share of collective spending€3.6…4.1bn
Defence, administration and interest on the national debt divided equally over every resident.
State accounts 2024reception and integration compensation onlypartial€0.7bnpartial€0.7bn✗~✗✓
Realised cost of the reception system 2024€358m
Integration compensation to municipalities and counties, 2024 budget€384m
Total€742m
Vahtera 2019budget lines and imputed items, about 100,000 peopleupper bound€3.3bnupper bound€3.3bn~✓✗✗
Items from the state budget€865m
Housing, imputed rents on 50,000 dwellings€546m
The item is not housing allowance paid out but an imputed rent on 50,000 dwellings. The same dwelling would earn rent from any tenant, so the item is not public spending in the sense that a paid benefit is.Vahtera, Iltalehti 5 March 2019.
Labour-market support, assuming 80% of adults€352m
The share is not measured, it is an assumption of the calculation. It is higher than the measured unemployment rate of any foreign-background group, so the item is an upper bound.Vahtera, Iltalehti 5 March 2019. Unemployment rates: Statistics Finland employment statistics.
Social benefits€267m
Municipal spending€261m
Other items, including a €225m fall in property values€976m
A fall in property values is not public spending but a change in private owners' wealth. This site's own analysis found no price gap that would support the item. For the whole country the difference is small enough that its confidence interval contains zero, and the within-city difference is explained by the share of rented dwellings and does not survive controlling for it.Vahtera, Iltalehti 5 March 2019. The price comparison is this site's own: Statistics Finland 13mu and Paavo 2024.
Total, with €121m of taxes deducted€3,267m
€121m is a fraction of the taxes a group of about 100,000 people pays, and indirect taxes are missing entirely. This is why the row is tagged an upper bound rather than an estimate.Vahtera, Iltalehti 5 March 2019.
Sources
Vahtera, Iltalehti 5 March 2019.
Vahtera, Iltalehti 5 March 2019. Unemployment rates: Statistics Finland employment statistics.
Vahtera, Iltalehti 5 March 2019. The price comparison is this site's own: Statistics Finland 13mu and Paavo 2024.
Kanniainen 2025a synthesis of foreign lifecycle studiesderived€3bnderived€3bn✓✓✓✗
Lifecycle cost of a humanitarian arrival, derived from Dutch and Swedish results€0.9m
The Dutch register study puts the lifecycle cost of an asylum seeker at €800,000, and Sweden's at −SEK 987,000 over the first seven years for the average refugee. No Finnish data of its own is used.Kanniainen 2025, citing van de Beek et al. 2021 and a Swedish register study.
Divided over about 60 years of residence€15,000/yr
A lifecycle figure is not an annual cost. To turn it into one it has to be divided by the length of stay. Sixty years is the remaining life of someone who arrives young, and a shorter span would raise the annual figure.The division is this site's, not Kanniainen's. The book has no arithmetic between the lifetime and annual figures.
€3bn a year therefore requires a group ofabout 200,000 people
€3,000m divided by €15,000 is 200,000 people.This site's own check.
People of the four refugee-origin backgrounds in Finland in 202588,102 people
Statistics Finland table 11rt. Holders of a protection permit alone number about 40,000. Reaching a group of 200,000 would mean counting family migration and origins beyond the four refugee-origin countries.Statistics Finland 11rt, population by background country 2025. The number of protection permits: Finnish Immigration Service.
The annual figure Kanniainen gives, with no published step betweenover €3bn/yr
The book gives both the lifecycle figure and the annual figure, but no arithmetic between them and no population the annual figure rests on. The review in Kansantaloudellinen aikakauskirja (Kotamäki 2026) finds the survey weighted towards the costs of humanitarian migration and passing over studies in which work-based migration is net positive.Kanniainen 2025. Review: Kotamäki, Kansantaloudellinen aikakauskirja 2026.
Sources
Kanniainen 2025, citing van de Beek et al. 2021 and a Swedish register study.
The division is this site's, not Kanniainen's. The book has no arithmetic between the lifetime and annual figures.
This site's own check.
Statistics Finland 11rt, population by background country 2025. The number of protection permits: Finnish Immigration Service.
Virénthe Dutch figure scaled by populationderived€10.5bnderived€10.5bn✓✓✓✗
Dutch net cost 1995–2019 (van de Beek et al.)€400bn
Divided by 25 years€16bn/yr
Finland's population against the Netherlands', 5.65 / 17.9m× 0.32
Scaling by population assumes the volume and composition of immigration are the same in both countries. The Netherlands has a larger foreign-background population of a different composition, so the ratio is rough in both directions.Populations: Statistics Finland and Statistics Netherlands. The ratio is this site's.
Scaled by population€5.1bn/yr
The figure Virén gives, with no published step between€10.5bn/yr
The gap to the population-scaled figure is a factor of two. No published calculation explains it. The figure also covers all immigration, not humanitarian immigration.Virén, Suomen Uutiset. The comparison figure: van de Beek et al. 2021.
Sources
Populations: Statistics Finland and Statistics Netherlands. The ratio is this site's.
Virén, Suomen Uutiset. The comparison figure: van de Beek et al. 2021.
Benchmarks from other countries
Denmark, Ministry of Finance 2021register data 2018, non-Western and descendants, 494,000 peoplemeasured€4.2bnmeasured€4.2bn✓✓✓✓
Net effect of non-Western immigrants and descendants 2018−DKK 31bn
of which MENAPT countries −24bn and other non-Western −7bn
At 7.45 DKK/€÷ 7.45
The Danish krone is pegged to the euro in ERM II within a narrow band, so the rate is effectively fixed and the choice does not move the figure.European Central Bank, ERM II. The central rate is 7.46038 DKK/€ within a ±2.25% band.
In euros€4.2bn
Sources
European Central Bank, ERM II. The central rate is 7.46038 DKK/€ within a ±2.25% band.
Netherlands, van de Beek et al. 2021register data 1995–2019, all immigrationmeasured€16bnmeasured€16bn✓✓✓✗
Net cost of those who arrived 1995–2019, Statistics Netherlands microdata€400bn
Divided by 25 years€16bn/yr
The average smooths large year-to-year swings. In the peak year of 2016 the net cost was €32bn, twice the average.van de Beek et al. 2021, Grenzeloze Verzorgingsstaat.
Sources
van de Beek et al. 2021, Grenzeloze Verzorgingsstaat.
This site's calculationBudget and accounts linesDerived from other countriesBenchmark from abroad
Every figure is a cost per year for the whole country, and positive means spending for public finances. A figure with two ends is the calculation's own range, a single figure is the middle of a published range. Each row opens onto a table of the steps. ✓ = yes, ~ = partly, ✗ = no.
Details
The quality tag says what the figure rests on. Measured = computed from whole-population register data. Estimate = modelled from register figures and unit costs. Projection = one year's age profile stretched over a whole lifetime. Partial = only some items are included. Upper bound = includes items that are not public spending, or taxes are not deducted. Derived = scaled from another country's result with no data of its own.
Only this site's row carries a range, because its every step is known and the range is the honest answer. The other rows show the middle of the range their author published, so that the rows stay readable. The €5.3…9.2bn range is wide because the services are modelled from age structure and unit costs rather than measured. The €7.2bn used in the other charts is the middle of this range.
The Finnish figures are not comparable with one another in scope. The state accounts' €0.7bn is the outturn of two budget lines, Vahtera's €3.3bn the gross spending on about 100,000 people, and Virén's €10.5bn covers all immigration.
Taxes deducted: the group's taxes and contributions are netted against the costs. Vahtera deducts €121m, a fraction of the group's taxes, hence partly.
Public services included: education, health care and other services are priced in, not only transfers.
Children and lifecycle: the calculation spans the arrival's whole lifetime and their children, not a single cross-section year. For the annual figure the mark is partly, because a cross-section is a cross-section: the children already here are counted as children, but the taxes they pay as adults are not credited. The lifecycle row, and the lifecycle chart below it, put that right by running the same population forward to death.
Official or peer-reviewed: the publisher is a research institute, a ministry or a peer-reviewed journal. Suomen Perusta is the Finns Party's think tank and the van de Beek report was commissioned by a party, though both use register data.
The same methods per person per year. These are not the rows of the table above divided by a population, but calculations of their own for groups of their own, which is why they cannot be added to the annual totals.
Per-person figures per year
Cost per person per yearTaxes deductedPublic services includedChildren and lifecycleOfficial or peer-reviewed
VATT Datahuone 2026protection permits, gap to natives, transfers onlymeasured€18,500/yrmeasured€18,500/yr✓✗✗✓
Gap in net transfers between protection-permit holders and the comparison group€18,500/yr
Register data for 2023, people aged 20 and over. The euro figure is read off the report's chart rather than its body text, so the last few hundred euros rest on the reading.VATT Datahuone 1/2026, the report's chart. Register data 2023.
Comparison group, people of Finnish background with the same age and sex distributionstandardised
Standardising removes the effect of age structure. Without it the gap would be smaller, because the group is younger than the native population and pensions are the largest single transfer.VATT Datahuone 1/2026, method description.
Taxes and transfers only, no public services€18,500/yr
Education, health care and other services are not in the figure, so it is not comparable with this site's figures or Denmark's. Nor is it the group's absolute net position, but a gap to a comparison group whose own net position is negative.VATT Datahuone 1/2026, the report's scope.
Sources
VATT Datahuone 1/2026, the report's chart. Register data 2023.
VATT Datahuone 1/2026, method description.
VATT Datahuone 1/2026, the report's scope.
Suomen Perusta 2015Salminen, Iraq and Somalia, register data 2011measured€13,700/yrmeasured€13,700/yr✓✓✗✗
Arrivals from Iraq, net effect per year€13,400
Arrivals from Somalia, net effect per year€13,900
Middle of the range€13,700/yr
Mamutus 2026protection background, at average cost, services includedestimate€15,500…22,800/yrestimate€15,500…22,800/yr✓✓~✗
Public services at same-age use€10,200…12,500/yr
Health, education and other services at the use of same-age natives, raised by the group's measured extra use.
Other spending, ten small items€2,200…2,800/yr
Interpreting, the justice system, the Immigration Service, grants to organisations and other named items.
Minus taxes and employer contributions€100…3,600/yr
The group's taxes net of transfers, plus employer contributions. The low end pairs the smallest spending with the largest revenue, the high end the other way round.
Marginal cost€8,700…15,200/yr
Share of collective spending€6,800…7,700/yr
Defence, administration and interest on the national debt divided equally over every resident.
Denmark, Ministry of Finance 2021non-Western and descendants, 494,000 peoplemeasured€8,400/yrmeasured€8,400/yr✓✓✓✓
Net effect 2018, −DKK 31bn at 7.45€4.16bn
Non-Western immigrants and descendants on 1 Jan 2018494,248 people
Divided by the population€8,400/yr
Finnish register dataThis site's calculationDerived from other countriesBenchmark from abroad
All figures are costs per person per year, positive = spending for public finances. A figure with two ends is the calculation's own range, a single figure is the middle of a published range. Each row opens onto a table of the steps. ✓ = yes, ~ = partly, ✗ = no.
Details
The quality tag says what the figure rests on. Measured = computed from whole-population register data. Estimate = modelled from register figures and unit costs. Projection = one year's age profile stretched over a whole lifetime. Partial = only some items are included. Upper bound = includes items that are not public spending, or taxes are not deducted. Derived = scaled from another country's result with no data of its own.
Only this site's row carries a range, because its steps are known. The other rows show the middle of the range their author published.
VATT's €18,500 and the middle of this site's range, €19,200, are close by coincidence and do not confirm one another. VATT's figure is the gap in transfers alone to age- and sex-matched people of Finnish background. This site's is the whole cost including services at average cost.
Denmark's €8,400 per person is the only measured figure in which the whole population, all spending and revenue and the descendants are included. It is smaller than the figure for Finland's protection-background group because the Danish group holds a great deal of long-settled work and family migration.
Taxes deducted: the group's taxes and contributions are netted against the costs. Vahtera deducts €121m, a fraction of the group's taxes, hence partly.
Public services included: education, health care and other services are priced in, not only transfers.
Children and lifecycle: the calculation spans the arrival's whole lifetime and their children, not a single cross-section year. For the annual figure the mark is partly, because a cross-section is a cross-section: the children already here are counted as children, but the taxes they pay as adults are not credited. The lifecycle row, and the lifecycle chart below it, put that right by running the same population forward to death.
Official or peer-reviewed: the publisher is a research institute, a ministry or a peer-reviewed journal. Suomen Perusta is the Finns Party's think tank and the van de Beek report was commissioned by a party, though both use register data.
A whole lifetime as one figure. A lifetime figure is not an annual figure multiplied by a life span but a calculation of its own, which carries one year's age profile decades ahead, which is why it is not comparable with the annual figures above.
Lifetime cost per person
Cost over a whole lifetimeTaxes deductedPublic services includedChildren and lifecycleOfficial or peer-reviewed
Suomen Perusta 2019Salminen, arrival aged 20–24, children includedprojection€1.0m lifetimeprojection€1.0m lifetime✓✓✓✗
Arrival from Iraq aged 20–24€690,000
the same, with children included€844,000
Arrival from Somalia aged 20–24€951,000
the same, with children included€1,343,000
Middle of the €0.7…1.3m range€1.0m
The ends of the range are Iraq without children and Somalia with children included. The projection stretches one year's age profile over a whole lifetime, so it assumes employment stays where it is.Salminen, Suomen Perusta 2019.
Sources
Salminen, Suomen Perusta 2019.
Mamutus 2026, lifecycledifference to natives of the same age, the people already hereprojection€0.18…0.54m lifetimeprojection€0.18…0.54m lifetime✓✓✓✗
Population by age band, sex and generation in 202588,102 people
Statistics Finland table 11rt. 64,441 born abroad and 23,661 born in Finland. The generation split is what allows those born here to carry an adult profile of their own.Statistics Finland 11rt, background country, age, sex and generation 2025.
Each cohort is run to age 100 on the life tableStatistics Finland 12ap 2024
Death probability by single year of age and sex. Each year's cost is weighted by the share of the cohort still alive that year.Statistics Finland 12ap, life table 2024.
The price of a year of life: this section's unit coststhe same as in the annual figure
Early childhood education €14,304 at 55% participation, basic education €11,650, health care €4,138 with the age coefficients, other adult services €3,500…4,500, old age €31,703 and a services uplift of 10…25%. Collective spending of €6,759…7,684 per person and employer social contributions of 19.25…20% of payroll are the same as in the annual calculation. No new unit cost is introduced, so the model and the annual figure cannot drift apart.This section's own unit costs, sourced on their own rows.
The same run for natives of the same age and sex distribution€0.92…0.96m
An absolute lifetime figure is not an interpretable quantity. On this method any resident of Finland costs close to a million, because services and collective spending are counted while the whole tax base is not measured. Only the difference says anything. The native run is conservative here: it leaves out the fact that the native population goes onto a disability pension almost three times as often, 6.7% against 2.4% of those aged 40–64.Model: repo_support/model_lifecycle_projection.js.
The group's own figure on the same run€1.14…1.45m
Born abroad €1.04…1.36m and born in Finland €1.39…1.71m. The figure for those born here is larger because they have a whole life ahead of them, old age included, while the median age of those born abroad is already over thirty.Model: repo_support/model_lifecycle_projection.js.
The difference per person over the whole lifecycle€0.18…0.54m
This is the scenario in which the gap of those born in Finland stays at their parents' level, the top of the range. Derived from education, which is the realistic figure, the gap is €0.14…0.46m per person and €12.3…40.3bn for the whole group. Discounted at 3%, €0.12…0.21m and €10.5…18.9bn. The low end comes from the smaller earnings-related pension, which returns much of the working-age gap in old age.Model: repo_support/model_lifecycle_projection.js.
Kanniainen 2025from foreign studies, no Finnish dataderived€0.9m lifetimederived€0.9m lifetime✓✓✓✗
Netherlands, lifecycle cost of an asylum seeker€800,000
Sweden, net cost of the average refugee over seven years−SEK 987,000
Kanniainen's figure for Finland, with no published step between€0.9m
The Dutch figure is computed from that country's own register and the Swedish one covers only the first seven years. The Finnish figure is an estimate placed between them with no published step. For comparison, the lifecycle figure Suomen Perusta computes from Finnish register data is €0.7…1.3m.Kanniainen 2025. The comparison figure: Salminen, Suomen Perusta 2019.
Sources
Kanniainen 2025. The comparison figure: Salminen, Suomen Perusta 2019.
Finnish register dataThis site's calculationDerived from other countries
All figures are costs per person over a whole lifetime, positive = spending for public finances. A figure with two ends is the calculation's own range, a single figure is the middle of a published range. Each row opens onto a table of the steps. ✓ = yes, ~ = partly, ✗ = no.
Details
The quality tag says what the figure rests on. Measured = computed from whole-population register data. Estimate = modelled from register figures and unit costs. Projection = one year's age profile stretched over a whole lifetime. Partial = only some items are included. Upper bound = includes items that are not public spending, or taxes are not deducted. Derived = scaled from another country's result with no data of its own.
Only this site's row carries a range, because its steps are known. The other rows show the middle of the range their author published.
Lifetime figures are not comparable with annual ones. A lifecycle calculation makes one year's age profile permanent for decades, which is its largest weakness.
Taxes deducted: the group's taxes and contributions are netted against the costs. Vahtera deducts €121m, a fraction of the group's taxes, hence partly.
Public services included: education, health care and other services are priced in, not only transfers.
Children and lifecycle: the calculation spans the arrival's whole lifetime and their children, not a single cross-section year. For the annual figure the mark is partly, because a cross-section is a cross-section: the children already here are counted as children, but the taxes they pay as adults are not credited. The lifecycle row, and the lifecycle chart below it, put that right by running the same population forward to death.
Official or peer-reviewed: the publisher is a research institute, a ministry or a peer-reviewed journal. Suomen Perusta is the Finns Party's think tank and the van de Beek report was commissioned by a party, though both use register data.
The cost is the difference between two stacks of similar size. Spending is public services at same-age use, the small items row by row and a share of collective spending. Revenue is taxes minus transfers and employer contributions. The same chain is computed for the protection background and for the non-Western background population, and every row opens onto its note and source.
The site's calculation: two scopes, one chain
Public servicesOther spendingCollective spendingTaxes minus transfersEmployer contributions
Protection background88,102 people
Non-Western background528,567 people
€bn per yearProtection background 88,102 peopleNon-Western background 528,567 people
Revenue
Taxes minus transfers−€0.1…+0.2bn+€1.2…3.2bn
Protection background
Close to zero, because VATT's figures are read as gaps to the native population and indirect taxes are counted in. The group's cost therefore does not come from transfers but from services.VATT Datahuone 1/2026, net transfers 2023.
The figure is a per-person gap to the native population, not the sum of taxes the group pays. The plus sign means the group is better than its comparison group on taxes minus transfers, and it is netted off the cost of services. The range is the calculation's own, not a VATT confidence interval, and it covers the fact that indirect taxes are not measured by background country.VATT Datahuone 1/2026, net transfers 2023. The native level and indirect taxes are the calculation's own addition.
Employer social contributions+€0.11…0.12bn+€1.0…1.2bn
The employer's pension and sickness insurance contributions are general government revenue that the register's net transfer does not include. The employed are the sum of the four origin countries, and the earnings are wage and self-employment income per earner for people of African and Asian background, weighted by the employed. The plus sign means the item is netted off the spending.Statistics Finland employment statistics 115t 2023. Integration database 11q4, 2024. Finnish Centre for Pensions 2025.
The register's net transfer is the taxes and contributions the employee pays minus the transfers received. The employer's share is left out of it although it is general government revenue: the earnings-related pension contribution of 17.38% and the sickness insurance contribution of 1.87% of payroll, about 20% with unemployment and accident insurance. The employed are counted, and the earnings are wage and self-employment income per earner: €27,466 for people of African background and €32,530 for the whole foreign-background population. The plus sign means the same as on the line above: the item is netted off the spending.Employed: Statistics Finland employment statistics 115t, by background country 2023, Western countries removed on the Danish definition. Earnings: integration database 11q4, 2024. Rates: Finnish Centre for Pensions 2025.
Spending
Public services at same-age use€0.9…1.1bn€5.4…6.5bn
Protection background
The largest item is school and day care, because 38% of the group are children. Age structure, not employment, is the single most important driver of the figure.Age structure: Statistics Finland 11rt 2025. Unit costs: this section's own, sourced on their own rows.
Non-Western background
The uplift covers services that are not itemised, such as disability services and employment services, and child welfare's use above the population share, whose base level has a line of its own. It is applied to the first four items only, because the education items are measured from student counts and need no allowance for services nobody named. Its size is calibrated against the Danish Ministry of Finance calculation, in which individual services do not stop at four items. Without it the sum would be €3.9…4.2bn.Calibration: Finansministeriet 2021, the non-Western group's service use in Danish register data.
Early childhood education–€0.36bn
Non-Western background45,215 children × 55% × €14,304
55% is the group's estimated participation rate, which is lower than the native population's. €14,304 is the net public cost of one full-time place. If participation rose to the native rate, the item would grow to about half a billion.The unit cost is the net operating cost of one full-time early childhood place. The 55% participation rate is the calculation's own assumption, not a measurement of this group.
Basic education–€0.70bn
Non-Western background59,735 pupils × €11,650
The pupil count is the school-age children of the group, and every one of them is priced at the full unit cost. Finnish as a second language is not an item missing from here: it is taught inside ordinary basic education, in the mother-tongue lessons, so its cost is already in the unit cost. Preparatory teaching is a separate cost, and it is recorded separately: €74.3m nationally in 2024, which over 10,792 pupils is €6,884 per pupil. The item is in practice entirely immigration-related, and it has a line of its own below. The pupil count has been falling since the 2022 peak of 14,757.Unit cost: Statistics Finland 12g7, operating cost of basic education per student €11,085 in 2024. The calculation uses €11,650, about five per cent above the published figure. Preparatory teaching: Vipunen, education costs, teaching form Perusopetus, valmistava opetus 2024, and pupils in the 2024/2025 school year. Pupil counts for the group: Statistics Finland 11rt 2025.
Upper secondary–€0.50bn
Non-Western background6,970 in general upper secondary × €9,054 and 52,425 in vocational × €8,289
Basic education ends at 15, but compulsory education runs to 18 and upper secondary is free. Before this line the calculation priced the group's 16–19-year-olds at nothing. The student counts are measured, and allocated to the group by citizenship: 72.0% of foreign citizens in general upper secondary and 82.6% in vocational education are non-Western. Vocational education also holds adults, not only 16–19-year-olds.Students: Statistics Finland 14fd, by generation 2024. Non-Western share: Statistics Finland 13sn, by citizenship 2024. Unit costs: Statistics Finland 12g7, 2024.
Tertiary–€0.39bn
Non-Western background19,023 in universities of applied sciences × €7,060 and 15,135 in universities × €16,837
Tertiary education is free as well and therefore public spending. Of foreign citizens, 81.5% in universities of applied sciences and 76.6% in universities are non-Western, so the item is about residents rather than exchange students. It is different in kind from the others, though: a graduate pays the education back in taxes and is a better than average taxpayer afterwards. A cross-section charges the spending this year and credits none of that return, just as it credits none of the future taxes of native students.Students: Statistics Finland 14fd, by generation 2024. Non-Western share: Statistics Finland 13sn, by citizenship 2024. Unit costs: Statistics Finland 12g7, 2024.
Health care–€1.63bn
Non-Western background138,359 × 0.44, 363,251 × 0.73 and 26,957 × 2.53, all × €4,138
€4,138 is health spending per person for the whole population. The coefficients scale it down, because the group is young and health spending rises steeply with age. They are measured rather than assumed: THL allocated €18.1bn of health and elderly care spending by age group, and 8.9% of it fell on 0–17-year-olds who were 20.1% of the population, 45.2% on 18–64-year-olds who were 61.8%. Without the coefficients the same calculation would give €2.2bn. The third coefficient is 2.53 and applies to people aged 65 and over, of whom the group has 26,957. They were previously priced at the working-age 0.73, which understated the item by €201m. Old age is not a separate item in the annual calculation, so nothing is counted twice. The 0.73 is a national average, and this group exceeds it for one reason: mothers from the four refugee-origin countries had 1,508 children in 2024, 3.4% of all births, while the group is 1.6% of the population. For deliveries and maternity care the item is therefore an underestimate, and its size can be derived from the same table. Health spending for women aged 18–40 was €2,304 against €1,602 for men, an excess of €702 a year. Over the country's 749,000 such women that is €526m, and over the births of 2011 it is €8,771 per birth. That is an upper bound, because women's excess spending is not all maternity care. At half of it, and 827 excess births, the additional cost is €3.6…7.3m a year, under one per cent of the group's services item.Health spending per person: Statistics Finland, general government expenditure by function 12a6. Age coefficients: Kapiainen & Eskelinen, Miesten ja naisten terveysmenot ikäryhmittäin 2011, THL Report 17/2014, with Statistics Finland population structure 2011.
Other individual services for adults–€1.37…1.76bn
Other spendingnine items, the rows below€0.19…0.25bn€0.66…0.78bn
Child protection, base level€48m€0.20bn
Protection background33,248 children × 1.6% × €89,532
Child protection had no line of its own: the source box read it into the 20…35% services uplift. The uplift is calibrated on Danish data in which the MENAPT group uses individual services a third more than natives, so it covers excess use, not the base every child population carries. The base is the group's children times the national placement rate of 1.6% (17,100 children placed outside the home in 2024) times the cost of child protection per placed child, €1,064m of institutional and foster care and €467m of open care in 2023 over the placed. The excess is real and large: in the child-welfare section's register study, 7.5…10.4% of refugee-background children had been placed at least once against 3.6% of the majority, 2.1…2.9 times, and that stays inside the uplift.THL, Lastensuojelu 2024 (statistical report 23/2025): placements and costs. Statistics Finland 11rt: the group's 0–19-year-olds in 2025. Placement ratios: this site's child-welfare section.
Non-Western background138,359 children × 1.6% × €89,532
The base at the national placement rate, as on the protection-background row. The group's excess use stays inside the services uplift.THL, Lastensuojelu 2024. Statistics Finland 11rt 2025.
Adult basic education€8…40m€40m
Protection background
€49.8m nationally in 2024. Adult basic education is for people who hold no basic education certificate, so it is in practice entirely the teaching of immigrants. The group's share is not measured: the low end is its share of the non-Western population, 16.7%, and the high end 80%, because the lack of basic schooling is concentrated in the refugee-origin countries.Vipunen, education costs, teaching form Perusopetus, aikuisten perusopetus 2024.
Non-Western background
€49.8m nationally in 2024. Adult basic education is for people who hold no basic education certificate, so it is in practice entirely the teaching of immigrants. The share allocated to the group is 80%, the same as its share of the foreign-background population.Vipunen, education costs, teaching form Perusopetus, aikuisten perusopetus 2024.
Preparatory teaching€18m€59m
Protection background
€74.3m nationally in 2024, in practice entirely immigration-related. Allocated to the group at its share of non-Western 0–19-year-olds, 33,248 of 138,359 or 24.0%. Children from the refugee-origin countries are probably over-represented among new arrivals, but the share is not measured.Vipunen, education costs, teaching form Perusopetus, valmistava opetus 2024, and pupils in the 2024/2025 school year.
Non-Western background
€74.3m nationally in 2024, 10,792 pupils and €6,884 per pupil. Preparatory teaching for basic education is for newly arrived pupils, so the item is in practice entirely immigration-related. 80% is allocated to the group, the same as its share of the foreign-background population.Vipunen, education costs, teaching form Perusopetus, valmistava opetus 2024, and pupils in the 2024/2025 school year.
Own-mother-tongue teaching€2.8m€9.4m
Protection background
About €11.8m nationally. No euro total is published, so it is built from the state grant's own rule: 24,827 pupils in the autumn of 2025, a grant for at most two weekly hours per calculated group of four pupils, a calculated cost of €25 an hour and a 38-week school year, which is €475 per pupil a year. The state pays at most 86 per cent as a grant and the municipality the rest, so the whole sum is public money. The hourly rate is administrative and a teaching hour costs more, so the figure is a floor. Allocated to the group at its share of non-Western 0–19-year-olds, 24.0 %.Finnish National Agency for Education, participants in own-mother-tongue teaching 2025 (from the grant applications), and the 2024 general grant's terms.
Non-Western background
About €11.8m nationally. No euro total is published, so it is built from the state grant's own rule: 24,827 pupils in the autumn of 2025, a grant for at most two weekly hours per calculated group of four pupils, a calculated cost of €25 an hour and a 38-week school year, which is €475 per pupil a year. The state pays at most 86 per cent as a grant and the municipality the rest, so the whole sum is public money. The hourly rate is administrative and a teaching hour costs more, so the figure is a floor. Allocated at 80 per cent, as the other teaching items are. The language table supports the share: in the autumn of 2025, 23 per cent of the pupils studied an EU or other Western language.Finnish National Agency for Education, participants in own-mother-tongue teaching 2025 (from the grant applications), and the 2024 general grant's terms.
Integration training€14m€68m
Protection background
Integration training bought by the employment administration €68.4m and the language and integration courses of liberal adult education €16.8m, €85.2m in all in the 2024 budget. Allocated to the group at its share of the non-Western population, 16.7%. Participants are new arrivals, and in 2023 and 2024 the largest group were people who fled Ukraine, so the population share is an assumption rather than a measurement.State budget 2024, moments 32.30.51 (public employment and business services, estimated use of the appropriation) and 29.10.31 (state grants for liberal adult education).
Non-Western background
Integration training bought by the employment administration €68.4m, 1.3 million student days, and the language and integration courses of liberal adult education €16.8m, €85.2m in all in the 2024 budget. From the start of 2025 the responsibility moved to municipalities and the money runs inside the state share, so 2024 is the last year in which the item shows as a line of its own. 80% is allocated to the group.State budget 2024, moments 32.30.51 (public employment and business services, estimated use of the appropriation) and 29.10.31 (state grants for liberal adult education).
Old-age care€13m€0.16bn
Protection background2,153 × €6,022
The group's over-65s are priced above at the health-care coefficient of 2.53 but with no care. The in-kind spending of the old-age function of general government, consumption €6,022m and purchased market production €2,093m, was €8,115m in 2024. Over the country's 1,347,494 people aged 65 and over that is €6,022 per person. Pensions are not here but in the transfers. Use is assumed to be the same as for people of the same age.Statistics Finland, general government expenditure by function 12a6, function G1002 old age, transactions P3K and D632K, 2024. Population: Statistics Finland 11rt, people aged 65 and over in 2025.
Non-Western background26,957 × €6,022
The group's over-65s are priced above at the health-care coefficient of 2.53 but with no care. The in-kind spending of the old-age function of general government, consumption €6,022m and purchased market production €2,093m, was €8,115m in 2024. Over the country's 1,347,494 people aged 65 and over that is €6,022 per person. Pensions are not here but in the transfers. Use is assumed to be the same as for people of the same age.Statistics Finland, general government expenditure by function 12a6, function G1002 old age, transactions P3K and D632K, 2024. Population: Statistics Finland 11rt, people aged 65 and over in 2025.
The justice system above the population share€21…49m€48…112m
Protection background
Police, courts and prisons cost €440m for the foreign-background population at its share of the population, and that is already inside collective spending. The part above the population share is €140m, or €60m standardised for age and sex. People of West Asian and Sub-Saharan African background are 35% of foreign-background suspects, and that is this group's upper bound, because those background groups also hold Turkey, Iran and many other origins.Statistics Finland crime statistics 13z4, and the justice appropriations of the state budget. Calculation: mamutus.info.
Non-Western background
Police, courts and prisons cost €440m for the foreign-background population at its share of the population, and that is already inside collective spending. The part above the population share is €140m. The low end of the range is the age- and sex-standardised €60m, because standardisation explains about three fifths of the gap, and the high end the unstandardised €140m. 80% is allocated to the group.Statistics Finland crime statistics 13z4, and the justice appropriations of the state budget. Calculation: mamutus.info.
How the 140 million is worked out
Public spending on these came to €2,710m in 2024. People of foreign background were 16.3% of crime suspects and 11.1% of the population. The grey bar is the cost attributed by that share, the purple one the part above the population share.
The spending is the whole public sector's, not the state's alone. Fire and rescue, €770m, is left out because it is not a cost of crime.
A suspect is not a convict. Suspicion records are the earliest stage of the process and overstate the level, but they are the only source that covers the police workload. Measured by prisoners the share would be higher: 22.9% of prisoners are foreign citizens, so on that count this calculation is conservative.
About three fifths of the part above the population share is explained by age and sex structure. Standardised, the excess would be about €60m. The bars are not standardised, because the bill that was paid does not standardise.
Only part of police and court spending arises from crime. The same proportional split is still the best available, because costs are not recorded by a suspect's background.
In the average-cost calculation the population-proportional part is already inside collective spending. What is added there is the purple bar, not the grey one.
The Immigration Service and voluntary return€22m€16…72m
Protection background
The Immigration Service costs €88m and voluntary return €2m. The agency handles every permit type, work and study included, so the group is charged the humanitarian share of its decisions, 22.3% in 2024, and voluntary return in full.State budget, moments 26.40.01 and 26.40.22. Distribution of decisions: Finnish Immigration Service statistics 2024.
Non-Western background
The Immigration Service costs €88m and voluntary return €2m. The agency handles every permit type, work and study included, so reading the whole sum against this group would be wrong and leaving it out would be wrong too. The low end of the range is the humanitarian share of the agency's decisions, 22.3% in 2024, and the high end the whole agency. Both are allocated to the group at 80%.State budget, moments 26.40.01 and 26.40.22. Distribution of decisions: Finnish Immigration Service statistics 2024.
Interpreting€46m€37m
Protection background
The reception system and reception allowance, €239m, used to be on this line. They do not belong in this calculation: the money is spent on asylum applicants while their applications are processed, before they are part of the resident population whose cost the calculation measures, and those refused never are. Interpreting stays, because it is spent in the services residents use. State compensation for integration, €377.7m in 2025, is not added on top: it is the way the state pays municipalities and wellbeing services counties for the very services the services item already prices.Moments 26.40.21 and 26.40.63, state final accounts. Integration compensation: moment 32.50.30, 2025 outturn.
Non-Western background
The reception system and reception allowance, €239m, used to be on this line. They do not belong in this calculation: the money is spent on asylum applicants while their applications are processed, before they are part of the resident population whose cost the calculation measures, and those refused never are. Interpreting stays, because it is spent in the services residents use, and 80% of it is allocated to the group. State compensation for integration, €377.7m in 2025, is not added on top: it is the way the state pays municipalities and wellbeing services counties for the very services the services item already prices.Moments 26.40.21 and 26.40.63, state final accounts. Integration compensation: moment 32.50.30, 2025 outturn.
Organisation grants and integration programmes–€19…23m
Non-Western background
STEA awarded €273.9m in grants for 2026, of which €17.3…20.4m went to work with immigrants, 6.3…7.5% of the total. The low end counts grants whose target group is named in the organisation's name or in the opening clause of the purpose, the high end every grant that mentions it anywhere. On top of that comes the Ministry of Economic Affairs and Employment's own programme appropriation for promoting integration and work-based immigration, €2.1m in 2025. The item is real but small, four hundredths of the services item, and it is counted in the marginal cost. The Ministry of Education and Culture's budget has no line earmarked for immigration or multiculturalism at all. Its money runs inside general appropriations, the largest of which are the state share of pre-primary and basic education €622.8m, arts and culture grants €93.9m and the promotion of youth work €32.1m in 2025. The share would show only in individual grant decisions, which are not published as data. The education appropriation would not be an addition in any case, because it funds the same teaching the education item already prices.STEA grant register, avustukset.stea.fi, application year 2026, filtering repo_support/measure_org_grants.js. Moments 32.50.03, 29.10.30, 29.80.50 and 29.91.50, 2025 outturns, budget proposal 2027, Ministry of Finance position.
Marginal cost, spending minus revenue€0.8…1.3bn€1.7…5.1bn
Non-Western backgroundall the spending above minus taxes, transfers and employer contributions
Marginal cost answers the question of how much public spending grows because of the group. The ends of the range in full: the low end 5.38 + 0.04 + 0.06 + 0.07 + 0.16 + 0.05 + 0.02 + 0.04 + 0.02 + 0.20 − 3.16 − 1.18 = €1.7bn and the high end 6.48 + 0.04 + 0.06 + 0.07 + 0.16 + 0.11 + 0.07 + 0.04 + 0.02 + 0.20 − 1.21 − 0.96 = €5.1bn, in order services, adult basic education, preparatory teaching, integration training, old-age care, the justice system, the Immigration Service, interpreting, organisation grants and the child-protection base, and finally the gap in taxes and transfers and the employer contributions deducted. The range is wide because its ends combine the least and most favourable assumptions: at the low end services are at their cheapest and the tax take at its largest, at the high end the reverse. Collective spending such as defence and administration is not in it but on the next line.The calculation's own combination of the rows above.
Collective spending, shared across the whole population€0.6…0.7bn€3.6…4.1bn
Non-Western background528,567 people × €6,759…7,684
The item is measured by function rather than assumed. General government expenditure per person in 2024: general public services €3,513, economic affairs €2,397 of which transport €1,370, defence €761, public order and safety €648, housing and community amenities €258 and environmental protection €107. That is €7,684 in total, the top of the range. The low end of €6,759 leaves out public debt service €810 and foreign aid €115, neither of which grows with one more resident. The individual functions — health care, education and social protection — are not here but on lines of their own. The average-cost calculation shares collective spending evenly across the whole population, as the Danish Ministry of Finance does. In the marginal-cost calculation it is zero, and this is the single largest difference between the two readings.Statistics Finland, general government expenditure by function 12a6, functions G01–G06, total consolidated expenditure per person 2024. Method: Finansministeriet 2021.
Average cost€1.4…2.0bn€5.3…9.2bn
How the marginal and the average cost are computedopens the formula and a chart
How much does public spending grow because of the group?
Spending that grows with the groupPublic services+Other spending−Revenue the group paysTaxes minus transfers+Employer contributions=Marginal cost
And if the group also carries its share of collective spending, such as defence, administration and interest on the debt?
Marginal cost+The group's share of collective spendingCollective spending=Average cost
Protection background88,102 people
Non-Western background528,567 people
Per person per yearthe whole group, children included, as the divisor€15,500…22,800/yr€10,000…17,300/yr
Protection background
The denominator is the whole group, children included, not only adults. Per adult the figure would be larger, but then children's services would be charged to a group that does not use them.Statistics Finland 11rt, 88,102 people in 2025.
Non-Western background
Average cost of €5.3…9.2bn over 528,567 people.
The same calculation over time
At the native age structurenon-Western background, €7.2bn at average cost now€8.8…10.2bn/yr
A cross-section measures the group as it is now, and now it is young. Of the non-Western background population 5.1% are 65 or over, against 26.2% of the native-background population. Old-age spending is Finland's largest single public expense, so the difference is large. Here the same group is reweighted to the native age structure to see what happens to the figure.
The method: the group, 528,567 people, is reweighted to the native age distribution. It would then hold 111,896 more people aged 65 and over and 35,417 fewer under 20. Each is priced at the national spending per head, €31,703 for old age and €15,229 for education.
The low and high bars come from pensions. The earnings-related pension scales with lifetime earnings, and this group's earnings are lower, so it would not draw the national average. The high bar assumes the national level, the low bar 60% of it. The guarantee pension sets a floor, so it does not go below that.
The figure is a floor, not a ceiling. Health care and elderly care services are not counted, though both rise steeply with age. With them the adjustment would be larger.
This is not a lifecycle calculation. It does not follow one arrival from arrival to death and does not model the second generation's own contribution. It answers one question: how much of the current figure is explained by the group being young.
A true lifecycle calculation does not follow from this data. Statistics Finland publishes income and employment by origin in three age bands only, 0–17, 18–64 and 65–, and three bands do not make an earnings arc. VATT's project for the Ministry of Finance promises a finer breakdown in 2027.
The adjustment also runs the other way in time. Today's over-65s paid their contributions somewhere, and part of this group's future pensions is accruing only now. A cross-section sees neither.
Lifecycle, difference to natives of the same ageprotection background, the people already here, €1,000 per person€140…457k per person
A cross-section charges a child for school and day care and credits nothing for the taxes that child pays as an adult. Here the same group of 88,102 people is run forward year by year with mortality until the last of them is gone. The figure is a difference to a native population of the same age and sex, because a euro amount on its own says nothing: on the same method any resident of Finland costs close to a million over a lifetime.
The method: the population is split into single-year cohorts and each is followed to age 100 on Statistics Finland's mortality rates. Every year of life is priced with this section's unit costs and the measured gap in taxes minus transfers. The same run is made for a native population with the same age and sex distribution, and the figure is the difference between the two.
The difference is the only interpretable figure. Collective spending and the part of the tax base this calculation does not measure cancel between the two runs, so four measured differences remain: the gap in taxes minus transfers, the gap in employer contributions, the uplift for heavier service use, and the smaller earnings-related pension.
Employer social contributions are revenue in both runs. The native population's employment is 74.5% and earnings €38,901 per earner, the group's 37.0% and €28,985, and the contribution is 19.25…20% of payroll. The gap is about €3,500…3,650 per working-age person a year, and it is the main reason the difference is larger than before the September 2026 correction.
The range is wide because its ends move in opposite directions over time. The high end assumes a full earnings-related pension and a 25% services uplift, the low end a 60% pension and a 10% uplift. A lower pension is a late saving that eats much of the working-age gap.
There is no Finnish measurement for the adult years of those born here. Statistics Finland puts the Finland-born second generation in the same class as the native population in its employment and income statistics, so the group has no employment rate of its own. The figure here is derived from education instead: of the Finland-born foreign-background population aged 25–34, 19.1% hold no post-basic qualification against 9.0% of the native-background population. Priced at each level's measured employment rate, that is a gap of 4.3 points. The same derivation run on the first generation gives 16.8 points against a measured 18.5, catching 91% of it, and corrected by that the second generation's gap is 4.8 points. That is 21% of their parents' gap. The chart therefore carries three bars: the derived gap, with the scenarios where the gap persists or closes entirely on either side of it.
The working-age gap is held constant across the working years. That is no longer an assumption but a choice the data supports. Employment does rise through the first decade: for the foreign-born from 48.0% in the year of arrival to 54.6% at one to four years and 63.5% at five to nine, against 74.5% for the comparison group. No figure is published for ten years and over, but it can be solved: the overall rate of 56.0% is the weighted average of the duration classes, so the fourth follows from the three that are published. Under any plausible distribution it lands between 46% and 55%, below the five-to-nine figure. The rise does not continue, and the gap after ten years is 20…28 points rather than 11.
The shift in the duration mix has also been worked out. The model applies today's average gap, but the present population ages into the long-resident class, whose employment is below today's average. Moving the first generation to that level raises the difference from €140,000…457,000 to €145,000…521,000 per person, and from €12.3…40.3bn to €12.8…45.9bn. Holding today's average in place is therefore a mildly optimistic choice rather than a cautious one.
This is an observation about cohorts, not about lives. The ten-years-and-over group is dominated by refugee-origin arrivals of the 1990s and of 2015, while the five-to-nine group currently holds Ukrainians and work migrants whose employment is high. The figure does not say that an individual's position deteriorates after ten years. It says the observed path cannot be read as one person's career, and that the calculation must therefore not narrow the gap year by year.
The model has no emigration and no future births. It measures what the people already here cost, not what the group costs as it grows. Emigration would lower the figure and future children would raise it.
Interpreting is set to zero, because it falls mostly in the arrival years and this calculation looks forward from today. The early cost of integration has already been paid. The reception system is not in the annual calculation either, because it is a cost of asylum applicants rather than of the resident population.
Earlier retirement was looked for and was not found. In 2024, 2.4% of the foreign-born aged 40–64 were on a disability pension against 6.7% of the native population. The gap has run the same way through all 30 years the statistic covers, so it is not a passing feature of a young group. Part of it is entitlement rather than health, though, because an earnings-related disability pension requires a work history in Finland. To that extent the cost does not disappear but moves into working-age transfers, which the gap in taxes minus transfers already measures.
The retirement age is the same for everyone, but the financing of the pension is not. The earnings-related pension accrues only from work done, and in the model the group's is 60…100% of the native level. The floor is the guarantee pension, which unlike the national pension is not scaled to years lived in Finland: three years of residence qualify, and it is paid in full to someone who has never paid a pension contribution. It is not a theoretical item: in 2025 foreign-language speakers received 33% of guarantee pension spending, €104m to 17,000 recipients, against an 11% share of the population (Kela). The lifecycle model runs the population to death, so these pension years are in the figure, and they are why late employment does not turn the lifetime balance around: a working life of 20–30 years does not pay back the years before it or after it.
The register supports this directly. In the integration database's table 14t6, adults of African background aged 18–64 still receive, after more than ten years in the country, €2,510 a year more in transfers than they pay in direct taxes, while people of Finnish background pay €6,718 more than they receive. The gap of about €9,200 per adult a year does not close with time, which is why the model does not narrow it. For the dependency ratio this means the head-count ratio improves while the ratio of workers can worsen: an added person of working age whose lifetime balance is negative is not a financer but someone financed.
The model gives both runs the same health and old-age spending, even though the native population goes onto a disability pension almost three times as often. On this channel the model therefore understates the native cost and overstates the difference. Correcting it would move the figure down, not up.
Kanniainen's €0.9m and Suomen Perusta's €1.0m are not comparable with this. They are absolute lifetime figures for one hypothetical arrival, not a difference to the native population, and they start from arrival rather than from today.
Marginal cost answers the question of how much public spending grows because of the group. Average cost also shares collective spending, such as defence and administration, evenly across the whole population, as the Danish Ministry of Finance does. For the whole foreign-background population only the totals are computed, €5.1…9.1bn at average cost, and they are in the comparison table.
Details
The low and high bars are the calculation's own ranges, not confidence intervals. The largest uncertainty is in the services, which are modelled from age structure and unit costs, not measured.
Employer social contributions, 19.25…20% of payroll, are revenue that the register's net transfer does not include. The group's 20,307 employed bring in €0.11…0.12bn of them. The reception system is not included: it is a cost of the asylum applicants' application year, not of the resident population. Interpreting, adult basic education, preparatory teaching, integration training, old-age care for the group's own over-65s, the justice system above the population share and the Immigration Service's humanitarian caseload are in other spending.
The child-protection base, €48m, has been in other spending since the second audit of September 2026. The group's excess use, a 2.1…2.9 times placement share, stays inside the 20…35% services uplift.
The applicant phase, that is reception, the Immigration Service's processing and voluntary return, €261m a year in all, is kept outside the total because it is not the resident population's cost. If the question is what the humanitarian channel costs, it is added: €1.0…1.6bn at marginal and €1.6…2.3bn at average cost.
Marginal cost counts only spending that grows with the number of residents. Average cost adds a share of collective spending such as defence and administration, €6,759…7,684 per person.
The group definition is a choice. Protection-permit holders alone would be about 40,000 people and the figure about half of this. Counting every family reunification regardless of origin would raise it.
The figure is a single-year cross-section of a young, recently arrived group. Employment rises with time in the country, so the same people cost less each year.
The scope moves the figure more than any choice of method. There are 88,102 people of protection background, 528,567 of non-Western background and 660,800 of foreign background in all.
The non-Western group's lower bound, €5.3bn, is higher than the lower bound for the whole foreign-background population, €5.1bn. People of Western background are net positive, so removing them raises the figure. This is why summing only the negative groups is not the net effect of immigration, but a figure for what the loss-making groups cost.
The Danish bar is one measured figure, not a range, so both of its bars are the same height. It is the −DKK 31bn for 2018 computed by the Danish Ministry of Finance from whole-population registers, at 7.45 DKK/€.
Per person the figures are: protection background €15,500…22,800, non-Western background €10,000…17,300, all foreign background €7,700…13,700 and Denmark €8,400. Denmark's measured figure falls below Finland's non-Western range. Its group is almost six times the size of Finland's protection-background group, and it holds a great deal of long-settled work and family migration.
In all three scopes employer social contributions are counted as revenue and the reception system is left out. Without those two corrections the figures would be €0.3…0.4bn higher for the protection background, €1.2…1.3bn higher for the non-Western background and €1.7…1.8bn higher for the whole foreign-background population.
The non-Western estimate is less certain than the protection-background one. The ground of the permit is not known at background-country level, so the weighted gap in taxes and transfers is derived from the figure for the whole foreign-background population rather than measured.
None of the three scopes is non-work-based immigration. The non-Western background also holds people who came for work from Asia and 25,065 people of Ukrainian background, which is why that scope's taxes minus transfers come out positive. The scope is drawn by background country rather than by permit basis because, according to VATT, 47% of the foreign-born are citizens or permanent residents whose original reason for entry the data does not carry. The nearest proxy for non-work-based is the protection background, and its broader counterpart is Denmark's MENAPT scope (Middle East, North Africa, Pakistan, Turkey), which held 118,000 people in Finland in 2025 and 155,000 with the Horn of Africa. In Denmark the net effect of the MENAPT background was −24bn DKK in 2018, about €11,900 per person.
This site's calculation: three parts by the Danish method20263 findings
mamutus.info, September 2026. Population: background country Iraq, Somalia, Afghanistan or Syria, 88,102 people in 2025, of whom 33,248 children (37.7%) and 54,854 adults (Statistics Finland 11rt). Every input is on the site or public
Spending: public services at same-age use €0.9…1.1bn, nine small items €0.19…0.25bn and a share of collective spending €0.6…0.7bn.
Revenue: taxes minus transfers €−0.1…0.2bn and employer contributions €0.11…0.12bn.
Marginal cost €0.8…1.3bn and average cost €1.4…2.0bn, or €15,500…22,800 per person per year. This box is the calculation's source reference, the rows and notes are in the ledger.
Spending in 2024, whole public sector: police €1,762m, courts €660m and prisons €288m, €2,710m in all. Fire and rescue, €770m, is excluded.
Suspects in 2024, counted by the most serious offence of the year: 152,323 people, of whom 24,757 or 16.3% of foreign background. Of the population, 623,949 or 11.1% were of foreign background.
The cost attributed by that share is €440m a year. The part above the population share is €140m. Standardised for age and sex about €60m of the latter remains, because standardisation explains around three fifths of the gap.
The order of magnitude matters: €140m is under a tenth of what services cost for the protection-background group. The public cost of crime is real but small beside education and health care.
By background region, 4,962 suspects were of West Asian and 3,706 of Sub-Saharan African background. Those two groups contain the four refugee-origin countries but also Turkey, Iran and many others, so their 8,668 suspects are an upper bound for the humanitarian group, not its figure.
Left out are the losses to victims of crime, insurance payouts and private security. They are not public spending, and they are not recorded by an offender's background.
This site's calculation: dwelling prices and an area's foreign-background share20266 findings
mamutus.info, September 2026. 786 postal areas that have at least eight dwelling sales as well as Yle's foreign-background figure and Paavo's economic data
The price per square metre is weighted by the number of sales across studios, two-room, three-room-plus flats and terraced houses, for the latest year in which the area had at least eight sales.
The raw correlation between price per square metre and the foreign-background share is +0.28. With income the correlation is +0.62, and between income and the foreign-background share −0.08.
The crude comparison usually used in debate: within a city and an income third, the areas with the highest foreign-background share are on average 16.5% cheaper across 26 comparisons. In Turku the gap is −38% and in Pori −44%, in Espoo +8%.
In the regression the sign flips at the rented share. Without it +25.0% per ten percentage points, with it −1.7%, and under the full set of controls −2.1%, confidence interval −5.7…+1.4. The R-squared rises from 0.53 to 0.87.
In the capital region, 147 areas, the same full set of controls gives −10.1% per ten percentage points, confidence interval −14.9…−5.2. In cities above 20,000 residents −7.6% and in areas with at least 20 sales −4.2%.
The conclusion: at national level no price effect is found, and within cities there is an association this data cannot separate from the quality of the housing stock. Neither is public spending.
Kotamäki 2026: what the survey leaves out20265 findings
Mauri Kotamäki, Is immigration nothing but a cost to the economy? Kansantaloudellinen aikakauskirja vol. 122, 1/2026. The author is chief economist at Finnvera and a senior researcher at the Population Research Institute of Väestöliitto
The book's cost estimates are mostly about humanitarian migration rather than immigration as a whole. The boundary should have been stated at the start, because a reader may take the results to cover immigration in general.
From the UK Office for Budget Responsibility's report the book takes only the negative net contribution of a low-paid immigrant. In the same report a median-wage immigrant is a net contributor from the start of the lifecycle, and high-wage ones considerably more.
The book cites a newspaper article claiming that Skilled Worker visa holders are a net cost over their lifetime. The UK Migration Advisory Committee's assessment is the opposite: their fiscal impact is over £16,000 a year positive.
On Sweden the book deals at length with Aldén and Hammarstedt 2019 on the negative net effect of refugee migration. Konjunkturinstitutet's 2025 report, which finds the net effect of the foreign-born improved in recent years and in some groups already more positive than the native population's, is given no comparable weight.
In Pesola, Sarvimäki and Virkola's 2025 register study the employment rate of refugees is about 10 per cent in the first year and rises to 60 per cent in ten years, while the initial advantage of work-based migrants narrows.
Valkonen 2026: there is no single general figure20265 findings
Tarmo Valkonen, On researching the public finance effects of immigration. Kansantaloudellinen aikakauskirja vol. 122, 1/2026. The author is a Senior Fellow at the Research Institute of the Finnish Economy
There is no single general figure for immigration's effect on public finances. Immigration changes in both volume and structure, the observations are incomplete, and different methods give different results.
The book's comparison between Etla's analyses of immigration as a whole and the selected calculations about refugee migration is not meaningful. The closest in method to Etla's work is Holler and Schuster's 2020 analysis of Austria, and even that is about the effects of the 2015 refugee wave.
In Finland the lifecycle observations large enough to be representative are mostly of children and young adults, because immigration is young. Nearly half of what a lifetime figure needs therefore rests on forecasts.
In Pesola, Sarvimäki and Virkola's 2025 data the net transfers of immigrants turn positive after six to eight years of residence.
In Vaalavuo et al. 2025 immigrants use health services less than native Finns of the same age.
Kanniainen 2025: a lifecycle cost synthesised from foreign studies20253 findings
Vesa Kanniainen, Julkisen talouden vaje: pelastaako maahanmuutto? (BoD, October 2025). A literature review of register studies from seven European countries, no Finnish data of its own
The lifecycle cost of a humanitarian arrival is put at about €900,000 and Finland's annual cost at over three billion. The figures are derived from foreign results: a lifecycle cost of €800,000 per asylum seeker in the Netherlands, and a net cost of −SEK 987,000 over seven years for the average refugee in Sweden.
The Swedish data gives a time profile Finland lacks: the net cost falls from SEK 192,000 in the first year to SEK 95,000 in the seventh.
The review in Kansantaloudellinen aikakauskirja (Kotamäki 2026) finds the survey weighted towards the costs of humanitarian migration and passing over studies in which work-based migration is net positive.
Virén: 10.5 billion a year scaled from the Dutch figure20252 findings
Matti Virén, professor emeritus of economics. According to Suomen Uutiset a rough estimate mirroring the Dutch study's result onto Finland
The Dutch €400bn over 1995–2019 is €16bn a year. Scaled by population that gives about €5bn for Finland, not 10.5. No published intermediate calculation explains the gap.
The figure covers all immigration, not humanitarian immigration, and the structure of Dutch immigration differs from Finland's.
Suomen Perusta 2019: lifecycle effect for arrivals from Iraq and Somalia20193 findings
Samuli Salminen, Statistics Finland register data. The 2011 cross-section by age group stretched over the whole lifetime of someone arriving aged 20–24. The publisher is the Finns Party's think tank
The lifetime net effect on public finances of someone arriving from Iraq aged 20–24 is −€690,000, and −€844,000 once the effect of children is included.
For an arrival from Somalia the corresponding figures are −€951,000 and −€1,343,000.
It is the only lifecycle estimate computed from Finnish register data. Its weakness is shared by every lifecycle calculation: one year's age profile is assumed to hold for decades.
Vahtera 2019: 3.2 billion from budget lines and imputed items20193 findings
Pauli Vahtera, Iltalehti blog 5 Mar 2019, three parts. State final accounts 2016–2017 and the 2019 budget, about 69,000 adults and 31,000 children from the largest humanitarian countries of origin
The largest items are the state budget 865, housing 546, labour-market support 352, social benefits 267 and municipal spending 261 million euros. Total 3,267 million.
Housing is computed from imputed rents on 50,000 dwellings, not from housing allowance paid. Labour-market support is assumed for 80% of adults. A €225m fall in property values is included, which is not public spending. Taxes deducted come to €121m.
Ministry of Finance officials told Yle the calculation was overstated, imprecise and resting on shaky assumptions, with figures counted in that do not belong there. The itemisation was not disputed line by line.
International
Danish Ministry of Finance: net effect of non-Western immigrants −DKK 31bn (2018)20215 findings
Finansministeriet, Økonomisk Analyse October 2021. Whole-population register data, all taxes and public spending including collective consumption, by origin and descent
The net effect of non-Western immigrants and their descendants was −DKK 31bn in 2018, of which MENAPT countries −24bn and other non-Western −7bn.
The figure has improved: −42bn in 2015 and −33bn in 2017. The same calculation is repeated yearly with the same method, which Finland lacks.
This is the method Finnish annual estimates should be compared against: one year, the whole population, all spending and revenue, descendants on their own line.
The figure looks large beside Finland's, but the difference is in the size of the group, not the method. On 1 January 2018 Denmark had 494,248 non-Western immigrants and their descendants (Statistics Denmark, FOLK1C), so −DKK 31bn is −€8,400 per person per year. This site's calculation gives €15,500…22,800 per person for the protection-background group, which is more. Denmark's group is simply almost six times the size of Finland's.
Denmark's group holds a great deal of long-settled work and family migration, the largest origins being Turkey, Lebanon, Pakistan, Bosnia, Vietnam and China, whose employment is closer to the native population than that of Finland's four refugee-origin countries. That pulls the Danish per-person figure down. Finland's non-Western background population, 528,567 people, is of almost the same size, and the same calculation gives it €5.3…9.2bn at average cost, or €10,000…17,300 per person, and Denmark's measured figure falls below that range.
Netherlands: €400bn over 1995–2019 by ground of entry20213 findings
Anonymous registry data for 17 million Dutch residents (1995–2019)
Immigration cost the Dutch public sector about EUR 400 billion net in 1995–2019, averaging about EUR 17 billion per year and peaking at EUR 32 billion in 2016.
Immigrants use more social security, education, and other benefits than the native population and pay less in taxes and social-security contributions.
Work and study migration from most Western countries and from a number of non-Western ones, East Asian above all, was a net gain for the public finances. All other immigration was at best roughly budget-neutral, and family and asylum migration were the most clearly negative.
Criminality
This site documents crime statistics and the overrepresentation of immigrant-background groups in them extensively. The graph below shows the overrepresentation of sexual and violent crime among immigrant groups: it brings together the overrepresentation found in the crime statistics of five comparison countries and calculates an average for each country of origin. Combining the averages across all countries, Syria, Iraq and Afghanistan show the highest overrepresentation in sexual offences, while Tunisia, Somalia and Iraq top the list for violent crime. Finland's legal definition of sexual offences differs somewhat from the other countries, so the true overrepresentation may be even more significant.
Sections in this group and their figures
Violent crimeIn Germany, 43.1% of violent-crime suspects in 2024 were foreigners.
Sexual crimes63% of those convicted of rape in Sweden have an immigrant background (2025).
Property crimeForeign-background overrepresentation in crime statistics is partly a young, male-heavy age structure. Standardised for age and sex, the whole foreign population falls below the native level in property crime (×0.9), but violent crime stays at ×1.6 and sexual offences at ×1.8. The average hides the spread between groups: in property crime too, seven nationality groups remain above the native level after standardisation, highest among them Romania ×6.6, Sweden ×6.2 and Somalia ×3.9.
Domestic violence — perpetrator backgroundAccording to the West Uusimaa police, nearly half of domestic-violence suspects in Espoo were born abroad, while foreign-language speakers make up over a quarter of the city's population (2026).
FGM and honour-based violenceOver 10,000 women and girls living in Finland are estimated to have undergone female genital mutilation.
Street gangsAccording to the police situation report (autumn 2023), Finland has 10 street gangs with about 260 people linked to them. They are mostly men aged 18–30 and predominantly of immigrant background.
The second generation in crime statisticsIn 2025 Finland-born 18–29-year-olds of foreign background were suspected of offences more often than either comparison group: 70% more often than people of Finnish background and almost twice as often as those born abroad. The age range is restricted because the second generation is much younger than the native population.
Foreign prisonersThe number of foreign prisoners has grown 75% in ten years: in 2025 an average of 839 foreign citizens were in prison daily, 23.6% of all prisoners, against a 7.7% population share.
Key figures
The age-standardised rate of non-traffic offences in 2024 was 839 suspects per 10,000 for the foreign-background population and 465 for the Finnish-background one. The gap survives standardisation.
In serious drug cases investigated by the National Bureau of Investigation, the foreign share of suspects rose from 53% (2024) to 79% (2025).
Per head of population, people of foreign background are victims of assault about twice as often as the native population. Sexual-offence victimisation, by contrast, is roughly as common in both groups (2009–2025).
Crime is one of the side effects most often raised in the immigration debate. This section measures it primarily through suspicion records. Suspicions are the earliest stage of the process and overstate the level, but the overrepresentation does not disappear on moving to convictions, at most it shrinks.
People of foreign background account for a clearly larger share of offences than their share of the population. The age and sex distribution explains about three fifths of the gap: in violent and sexual offences the gap remains severalfold after standardisation, while in property crime the standardised foreign-national level falls slightly below that of the native population. Differences between background countries are large. The victim statistics look different: relative to population, assault victims are about twice as common among people of foreign background, while being the victim of a sexual offence is roughly equally common in both groups. Neither figure is age-standardised, and the foreign-background population is younger.
Some figures concern foreign citizens and others people of foreign background, which includes Finnish citizens. Each entry states which, and the standardisation can be switched on in the crime chart of the Charts view. By offence type the subject continues under violent crime and property crime, and the Finland-born generation has its own section under the second generation.
The figure is over-representation: 1 means a share equal to the group's share of the population, 3 three times that. Each panel is one offence type, the bar is the countries' median and the markers are the individual countries' figures for the same origin.
Overrepresentation in crime statistics by country of origin · Finland, Germany, Switzerland, Denmark
Details
Bar = median across the countries that report suspects (Finland, Germany, Switzerland). Symbols = per-country value. A row is shown only when at least three suspect-reporting countries are available.
Every ratio is computed inside one country: (suspects per resident of citizenship X) divided by (suspects per resident of the host's own citizenship). Only the ratios are comparable across countries, never the levels, because offence definitions and recording practice differ.
Numerators: Finland Statistics Finland 13je (2024), Germany BKA PKS 2024 table T62 (key 100000 sexual offences, 220000 assault), Switzerland BFS PKS 2024 (StGB title 1 Leib und Leben and title 5 Sexuelle Integrität, permanent residents only), Denmark DST STRAFNA4 (2024).
Denominators: Eurostat migr_pop1ctz, population by citizenship (Finland, Germany), the Swiss BFS STATPOP permanent resident population by citizenship, and Denmark's DST FOLK2 population by country of origin. Eurostat publishes citizenship detail for some countries only, which is why Switzerland and Denmark come from their own registers.
These figures used to be compiled here by hand, with the population denominators cited nowhere although half of every ratio rests on them. Both halves are now computed from a named series by repo_support/fetch_overrep_crime.js.
Denmark (DST STRAFNA4) reports convictions rather than suspects, and classifies by country of origin rather than citizenship. Its figures are therefore systematically lower and cover a different group: Denmark is plotted as a symbol but is excluded from the bar.
Austria is out for now. Its figures are in parliamentary answer 17582/AB, which covers foreign nationals only, so the suspect level among Austrian citizens — the denominator of the ratio — is not in it. Without a sourced baseline the row is not computed.
The Swiss crime table groups Serbia, Montenegro and Kosovo together while the population table does not, so no Swiss value is computed for Serbia.
The German figures are PKS suspects, and suspects include people who do not live in Germany — which is why the BKA itself warns against deriving population-adjusted crime rates from the PKS. Table T61 bounds the effect: in 2024, 2.2% of non-German sexual-offence suspects and 2.4% of violent-crime suspects had no right of residence or no residence in Germany at all. Transit therefore inflates the figures most for countries with few residents in Germany.
Finland's value is the highest on several rows (Iraq, sexual offences: Finland 13.8 against 3.7-6.3 elsewhere). The median keeps one country from moving a row. The gap points to definition: Finland's sexual-offence categories are broad and the figures are suspects, not convictions.
Four different population concepts — and why the figures are not interchangeable
Statistics use four different concepts, and the source decides which one you get. They describe the same phenomenon with different boundaries, and their sizes differ substantially (2025): foreign citizens 432,654 = 7.7% of the population. Foreign-language speakers about 10%. People of foreign background (by parents' birth country) 660,800 = 11.7%. Foreign-born about 10%. The rule followed on this site: when computing overrepresentation, the numerator and the denominator must use the same concept. Prison statistics, for example, are based on citizenship, so they must be compared with the 7.7% figure for foreign citizens — not the 11.7% for foreign background. Kela's statistics are based on native language, so they are compared with the share of foreign-language speakers. A ratio computed across two concepts is not a real quantity, and the error can run in either direction. This is why every figure states which concept it uses.
The number of second-generation people of foreign background passed 100,000. This group was born in Finland and is often Finnish by citizenship, so it is included in the foreign-background figure but not in the foreign-citizen one — a concrete illustration of why the two concepts are not interchangeable.
At the end of 2025 there were 432,654 foreign citizens living in Finland, 7.7% of the population. People of foreign background numbered 660,800, or 11.7%. The difference between the two figures is the group that holds Finnish citizenship but has foreign origins.
Age and sex standardisation: what it does to crime statistics
A common counter-argument is that foreign residents' crime overrepresentation is merely an artefact of a young, male-skewed age structure. Standardisation narrows the gap markedly, but does not remove it. Using Statistics Finland data (suspects 13yy + population 11rg, permanent residents, 2024), directly standardising for age and sex brings foreigners' overall crime rate down from a raw ratio of about x1.5 to about x1.2 against the native-background population - age and sex explain roughly two thirds of the gap. In property crime the standardised rate for the whole foreign population falls even below the native level (about x0.9). In violent and sexual offences the gap remains many-fold after standardisation: about x1.6 for violence and x1.8 for sexual offences across the whole foreign population, and considerably more for the main refugee-origin countries - Iraq, for example, at about x6-7 for violence and x9 for sexual offences. The same standardisation can be switched on in the Charts view's crime chart via the 'Age- and sex-standardised' option. What standardisation does not tell you: it answers whether an individual from this group is more crime-prone than a native-background person of the same age and sex. It does not answer whether the total number of offences rises. If the arriving population is heavily skewed towards young men, aggregate crime rises even when the age-specific rates are exactly the same as in the native population - the population structure is itself part of the effect. These are two different questions, both valid, and standardisation answers only the first. Suspects or convictions: suspect statistics are the earliest stage of the process and include cases that never lead to charges or conviction, and they are sensitive to reporting propensity and police targeting - so they overstate the level. Conviction statistics are firmer but narrower and lag behind. Overrepresentation does not disappear when moving to convictions, it shrinks: in this section's cross-country chart the Danish conviction-based ratios are systematically lower than the suspect-based ratios of the other countries, but still many times the native-background level. The figures on this site are suspects unless stated otherwise. Permanent residents: the standardisation above uses permanent residents only. That matters, because according to 2015 data more than half of crime suspicions against foreign nationals concerned people who do not live in Finland but were staying temporarily. Unstandardised foreign-citizen figures - prison statistics, for example - include that group even though it is not part of the population denominator, and are therefore upper bounds.
The source table from which the age and sex standardisation is calculated: suspects by nationality × age × sex, combined with population table 11rg. Direct standardisation to the age and sex structure of the country as a whole.
European Journal of Criminology (DOI 10.1177/1477370815587768)2015
Summary
The data is the 2012 wave of the Finnish Self-Report Delinquency Study (N = 8,914). Because the offending is self-reported rather than drawn from police records, the result is not exposed to reporting-propensity or police-targeting bias, unlike suspicion statistics. Immigrant youth showed a higher offending rate in 14 of the 17 offence types measured. The gap was smallest in shoplifting, vandalism and bullying, and there was no significant difference in substance use. The authors' central result is explanatory: the gap was tied to routine activities and parental control, and once a wide set of background variables was controlled for, the elevated risk among immigrant youth shrank but remained statistically significant. The same pattern as in the site's own standardisations: background explains part of the gap, not all of it. NOTE: the 14/17 split does not appear in the abstract, which speaks only of "several" offence types. It needs checking against the full article.
In 2018, 297,600 crime suspects were identified, of whom 34,200 were foreign nationals. Age- and gender-standardised suspicion index: foreign nationals were suspected 1.2x more often than Finnish nationals. Iraqi men were suspected of sexual offences 12.8x more often than Finnish men (2017–2018 combined). Somali men were suspected of drug offences 3.4x more often. Swedish women of property crimes 3.1x more often. Iraqis overall 2.0x, Swedes 2.3x. Figures represent suspicions, not convictions.
Four nationally representative self-report studies, N = 21,504 Swedish adolescents (mean age 15). Both first- and second-generation immigrants report committing more offences than natives. Because the data are self-reported rather than police-recorded, the result cannot be explained by reporting propensity or policing focus — and the gap appears in the second generation as well. The authors stress, however, that the association is weak and accounts for only a marginal amount of the variance in total offending. The link between delinquent friends and offending is stronger for both immigrant generations than for natives. The same authors note that overrepresentation is markedly larger in register data than in self-reports, and that the register gap is at least partly due to how likely a young person is to be recorded by the police in the first place — worth keeping in mind when reading the suspect statistics in this section.
Helsingin yliopisto, Kriminologian ja oikeuspolitiikan instituutti2026-02
Summary
The Rikollisuustilanne 2024 review standardises crime-suspect rates using Eurostat's European Standard Population (a different method from the site's own direct age/sex standardisation of StatFin tables). In 2024 the age-standardised non-traffic crime rate was 839 per 10,000 for the foreign-background population vs. 465 for the Finnish-background population — a ~1.8× gap that persists after standardisation. For assault the standardised rate was about 70% higher and for sexual offences more than 1.5× higher, but for theft it was lower for the foreign-background population, and the sexual-offence gap has narrowed because the Finnish-background rate has risen faster — which the review links to the 2023 consent-based law reform and a lowered threshold for reporting. As shares: in 2024 people with a foreign background made up 19% of non-traffic crime suspects and 17% of traffic offence suspects, highest by crime type in smuggling (32%), rape (32%) and tax offences (30%). Non-traffic crime committed by immigrants has grown roughly 70% since 2015, mostly explained by growth in the foreign-background population: the age-standardised rate rose about 7% for the foreign-background population and 12% for the Finnish-background one. Victims: between 2015 and 2024 police recorded over 77,000 crime victims with a foreign background, the annual count growing from 5,400 (2015) to 11,439 (2024), about 14% of all police-recorded victims. The victim share was highest in violence and threat offences: assault and unlawful threats averaged 15%, stalking 15%, robbery and rape 12%.
The share of foreign suspects and offenders in serious narcotics cases investigated by the National Bureau of Investigation (KRP) rose from 53% in 2024 to 79% in 2025, KRP intelligence chief Jaakko Christensen said in a Rikospaikka interview. The figures cover the most serious cases KRP investigates, often linked to organised crime — not all drug offences. By nationality, Swedes have risen to the top of KRP's drug cases and are the second-largest group across all serious drug crimes, after Estonians. KRP's priorities include organised crime, street gangs and Swedish organised crime. Youth violent crime has also been rising since 2015.
Violent crimeSafetySocial cohesion
Key figures
In Germany, 43.1% of violent-crime suspects in 2024 were foreigners.
Violent crime is among the most persistent topics in the immigration debate, and the usual response to the over-representation is age structure: arrivals include many young men, and young men commit more violence whatever their background. For violent crime that does not account for it. The age- and sex-standardised level for the foreign population as a whole is about 1.6 times the native one.
In Germany, well over two fifths of violent-crime suspects are foreign nationals, a share that has risen from the previous year (BKA crime statistics). The whole crime picture is under criminality, and its organised part under street gangs.
The figure is the ratio to Finnish-background men of the same age: 1 is the same level, 2 twice it. The bar is a ten-year average.
Violent crime rate of young men (18–29) by region of origin relative to Finnish-background · 2015–2024 avg.
Details
Both rates are unweighted 10-year (2015–2024) averages for men aged 18–29.
Not separately age-standardised, since the comparison is already restricted to one age and sex group.
Background-country groups are internally heterogeneous and vary greatly in size. In small groups individual cases can shift the whole group's rate substantially.
The East Asia figure is omitted due to too few cases (avg. ~5 suspects/yr).
Figures are suspects, not convictions, and include people only temporarily present in Finland.
A review of Nordic homicide by Martti Lehti, researcher at the Institute of Criminology and Legal Policy (Krimo): in Norway, Sweden and Denmark, immigrants now account for 25–40% of homicide perpetrators. In Finland and Iceland the share is 10%. In all Nordic countries the immigrant homicide rate exceeds the native rate, but the gap is larger in Scandinavia than in Finland. Finland's peculiarity: its immigrant homicide rate is the highest in the Nordics — but so is the native Finns' rate among the highest.
The National Bureau of Investigation (KRP) report on youth violent crime highlights an increase in aggravated robberies and homicides committed by minors after 2015. Between 2015 and 2020, youths with a foreign background were suspected of violent crimes 2.8 times more often than those with a native background (7.0 vs. 2.5 suspects per 1,000 residents). For 15–17-year-olds, the suspect rate was 28.3 for the foreign-born, 21.1 for Finland-born youths with a foreign background, and 7.9 for native-background youths per 1,000 residents. The share of foreign-background suspects among all minors suspected of violent crimes rose from 18% in 2015 to 23% in 2020.
Violent crimes committed by persons with a foreign background are common in Helsinki, with the group being clearly overrepresented in police statistics. In the 2000s, suspects with a foreign background accounted for over 40% of rapes and nearly a third of robberies and extortions reported to the Helsinki police. The article is from February 2008, so “the 2000s” covers 2000–2007. The figures are old, but they still give an indication of a phenomenon that more recent reviews show has persisted.
The article is from 2008, but the pattern it describes has continued to hold true ever since – see also the more recent National Bureau of Investigation report below.
Statistics Finland reports that approximately one in five (20%) suspects in aggravated robberies under the age of 21 has a foreign background. According to the police, a subset of foreign-background youths in the Helsinki metropolitan area carry weapons and are quick to use them. Yle also discusses the criteria for reporting the ethnic background of suspects.
A National Police Board analysis (Polstat data 2007–2024): homicides and attempted homicides by under-21s have increased since 2019, driven mainly by attempted murders. In 2020–2023 the rate for Finnish-background youths was about 1.9 and for foreign-background youths about 5.3 homicides or attempts per 10,000 people — nearly triple. Over half of suspects were men aged 18–20, about 28% were minors, and nearly 80% had prior violent offences.
In Mecklenburg-Vorpommern there are 1,497 open arrest warrants. 1,134 (75%) concern foreign nationals. For violent crimes, foreigners account for roughly 90% of wanted suspects (74 of 83). Top nationalities sought: Poles (364), Georgians (107), Romanians (75), Ukrainians (68), Tunisians (58). Foreign suspects can more easily evade justice by fleeing to their home countries or into parallel communities.
43.1% of violent-crime suspects in Germany in 2024 were foreign nationals, and roughly 40% of all suspects lacked German citizenship. Welt documents the gap between media portrayal (where foreign suspects appear in up to 94.6% of TV crime reports) and the actual statistical share.
Sexual crimesSafetySocial cohesion
Key figures
63% of those convicted of rape in Sweden have an immigrant background (2025).
Sexual offending is among the most contested topics in the immigration debate. In Finland 33.7 per cent of those suspected of rape or aggravated rape in 2025 were of foreign background, against 11.7 per cent of the population — an over-representation of about threefold. The same pattern recurs across Europe: a University of Lund study (2025) finds that a clear majority of those convicted of rape in Sweden have an immigrant background, and Eurostat data show police-recorded sexual offences, and rape in particular, rising steeply across the EU over the past decade.
The usual response is that the gap reflects age and sex structure: arrivals include many young men, and young men commit more crime whatever their background. For sexual offences that does not account for it. The age- and sex-standardised level is clearly above the native one for the foreign population as a whole, and several times higher for the major refugee-origin countries.
The figures on this page are unstandardised and mix suspects with convictions, see the standardisation note in the Criminality section.
The bars are police-recorded sexual violence offences and the line the rapes among them, in thousands a year from the same 26 EU countries.
Sexual violence and rape in the EU · 2014–2024
Sexual violenceRape
Show the figures
Year
Sexual violence
Rape
2014
127.2
39.2
2015
126.6
39.2
2016
138.7
44.3
2017
143.8
48.9
2018
162.5
54.8
2019
173.2
60.6
2020
172.4
63.7
2021
202.5
77.1
2022
224.2
85.6
2023
237.5
91.9
2024
248.9
98.2
Source: Eurostat · crim_off_cat · thousands of police-recorded offences · the 26 EU countries reporting both series in every year · ec.europa.eu/eurostat
Khoshnood, Sundquist & Sundquist (2025) – Immigrant background and rape conviction: a 21-year follow-up study in Sweden20255 findings
Peer-reviewed register study, Journal of Interpersonal Violence (DOI 10.1177/08862605241311611). Nationwide register data 2000–2020, case-control design. Open access — this source can be opened and checked directly.
Covers everyone convicted in Sweden between 2000 and 2020 of rape, aggravated rape or an attempt at either, where the victim was a woman aged 18 or over: 4,032 convicted and 20,160 matched controls.
Of those convicted, 36.9% were Swedish-born with two Swedish-born parents, against 69.5% of controls. This is the source of the widely repeated figure that 63% of those convicted had an immigrant background.
The unadjusted odds ratio for those born abroad who arrived at 15 or older was 6.25 (CI 5.76–6.78), and 4.57 (4.16–5.03) for those who arrived before 15.
The authors' central result is what adjustment does. Controlling for socioeconomic status, substance-use and psychiatric disorders and prior criminality nearly halved the risk for those who arrived before 15 (4.57 → 2.37) but left it essentially unchanged for those who arrived at 15 or older (6.25 → 6.22).
The authors stress that the raised risk is concentrated among those resident less than five years and those who arrived at 15 or older, which they read as pointing to the role of acculturation.
Sexual offences in recorded statistics across the EU — and what perpetrator background shows
Eurostat statistics show a substantial rise in recorded sexual offences in several European countries: reported rapes in Spain rose 322% over a decade, and the EU-wide increase was about 150%. The number of recorded offences does not by itself say anything about perpetrators, however: over the same period several countries widened their legal definition of sexual offences (Spain in 2022, Sweden in 2018) and reporting propensity rose, both of which raise the figures without the number of acts changing. There is better evidence on perpetrator background than growth in totals: according to the Spanish Interior Ministry's own statistics, foreign nationals accounted for 37.3% of those arrested or investigated for sexual offences in 2023, while foreign nationals are about 13–14% of Spain's population. Other years' shares appear in each year's own report, which is not linked here. Note that Spain's tourist volume means some suspects do not live in the country and are not part of any population denominator.
Ministerio del Interior, Informe sobre delitos contra la libertad e indemnidad sexual (2023)2023
Summary
The Spanish Interior Ministry's annual report breaks down people arrested and investigated for sexual offences by nationality. In 2023, foreign nationals accounted for 37.3% (5,136 of 13,767) and Spanish nationals for 62.7%. The most common countries of origin were Morocco (1,001, or 7.3% of the total), Colombia (600, 4.4%) and Romania (344, 2.5%). Foreign nationals are about 13–14% of Spain's population, so the share is well above the population share. This is direct perpetrator-background data, unlike growth in the total number of recorded offences. NOTE: each year is published as its own report. The one linked here is the 2023 report and does not contain other years' shares.
IS investigation of 2018 sexual offence convictions in Helsinki and Oulu: out of 82 convicted individuals, 43 (52.4%) were of foreign background. In Helsinki, 40 foreign-background and 31 native Finns were convicted. In Oulu, 3 foreign-background and 8 native Finns.
Suspected sexual offences by taxi drivers in Helsinki multiplied after taxi deregulation (from 2 reports in 2017 to nearly 50 in 2023–2024). According to Helsinki police, the suspected drivers have without exception been of foreign background.
Per Statistics Finland's crime and coercive measures statistics, 364 of 1,080 (33.7%) rape/aggravated rape suspects in 2025 were of foreign background, up from 281 of 893 (31.5%) in 2024. Over the same period the foreign-background share of the total population was 11.7% — roughly a threefold overrepresentation among suspects, a share that has risen steadily throughout the 2000s.
The Finns Party's party newspaper compiled the gang rapes against women that reached publicity in 2015–2026, over ten cases, in every one of which the perpetrators were of foreign background. In the compilation's only case where one of the perpetrators was of Finnish background, the victim was a man. The cases were found in news reports and public court documents with the help of an AI search, and the paper itself states that Finland has no comprehensive statistics on the backgrounds of gang rape perpetrators. In the paper's own search the perpetrators were of foreign background in practically every case it found.
Property crimeSafetyEconomy
Key figures
Foreign-background overrepresentation in crime statistics is partly a young, male-heavy age structure. Standardised for age and sex, the whole foreign population falls below the native level in property crime (×0.9), but violent crime stays at ×1.6 and sexual offences at ×1.8. The average hides the spread between groups: in property crime too, seven nationality groups remain above the native level after standardisation, highest among them Romania ×6.6, Sweden ×6.2 and Somalia ×3.9.
Property crime is among the most persistent topics in the immigration debate. It is also the one offence category where age and sex standardisation turns the picture around: the standardised rate for the whole foreign population is about ×0.9, below the native-background level. The corresponding figure is about ×1.6 for violent crime and ×1.8 for sexual offences.
That whole-population average hides wide differences between groups, though. Taken separately by citizenship, seven sufficiently large groups remain above the native level even after standardisation: Romania ×6.6, Sweden ×6.2, Somalia ×3.9, Estonia ×2.8, Iraq ×2.1, Syria ×1.7 and Iran ×1.6. Standardisation therefore explains the overrepresentation at the level of the whole population, but not within every group.
Unstandardised, youths with a foreign background are significantly overrepresented among robbery suspects, with particularly high rates for sub-Saharan African and Western Asian backgrounds. The figures on this page are unstandardised and concern young men — they answer a different question from the standardised total.
The figure is the ratio to Finnish-background men of the same age: 1 is the same level, 0.5 half of it. The bar is a ten-year average.
Crime rate of young men (18–29) by region of origin relative to Finnish-background · 2015–2024 avg.
Details
The standardised per-group ratios are computed on this site from Statistics Finland tables rpk 13yy and vaerak 11rg for 2025, covering groups with at least 5,000 residents and 100 cases.
Both rates are unweighted 10-year (2015–2024) averages for men aged 18–29.
Not separately age-standardised, since the comparison is already restricted to one age and sex group.
Background-country groups are internally heterogeneous and vary greatly in size. In small groups individual cases can shift the whole group's rate substantially.
The East Asia violent-crime figure is omitted due to too few cases (avg. ~5 suspects/yr).
Figures are suspects, not convictions, and include people only temporarily present in Finland.
Helsingin yliopisto, Kriminologian ja oikeuspolitiikan instituutti2026-02
Summary
Table 2 of the Rikollisuustilanne 2024 review breaks down the crime rate of men aged 18–29 by region of origin (2015–2024 average, suspects per 10,000 population). The lowest property- and violent-crime rates are found among those with East Asian backgrounds, the highest among those with North African backgrounds. For sexual offences, the lowest rates are among East Asian backgrounds and the highest among those representing Sub-Saharan Africa and Western Asia. Differences in traffic offences are generally small, except for men with Western Asian backgrounds, whose rate is clearly higher than that of other groups. The review cautions that background-country groups are internally heterogeneous and that figures for small groups can be swayed by individual cases.
Statistics Finland's offences and coercive measures statistics record 439 robberies in 2019 across the 0–14, 15–17 and 18–20 age groups, with foreign-background people making up 24.2% of suspects, against 16.4% in 2018. The total number of robberies fell over the same period (458 to 439), so the share rose while the count shrank. The share was highest in the youngest groups: 34.3% (0–14) and 34.0% (15–17). Of the robberies committed by foreign-background youth, Sub-Saharan Africa accounted for 32% (0–14) and 36.4% (15–17), Western Asia for 16.0% and 24.0%. The numbers are small — 24.2% of 439 is about a hundred cases — so year-to-year variation is large. Statistics Finland publishes figures only for suspects in cleared offences, and being a suspect implies neither prosecution nor conviction. A person counts as foreign-background if both parents were born abroad. Dual citizens holding Finnish citizenship are counted as Finnish.
According to the West Uusimaa police, nearly half of domestic-violence suspects in Espoo were born abroad, while foreign-language speakers make up over a quarter of the city's population (2026).
About 13,000 intimate-partner and domestic violence offences were recorded in 2024. 18% of the victims were of foreign background, against a population share of 11%.
Domestic violence is often presented in debate as something that falls evenly regardless of background. The Finnish data does not support that. According to the West Uusimaa police, nearly half of domestic-violence suspects in Espoo were born abroad. According to Statistics Finland, 18 per cent of domestic-violence victims nationwide in 2024 were of foreign background, against 11 per cent of the population. The over-representation thus covers victims as well as suspects.
These are unstandardised shares of suspects and victims, not age- and sex-standardised offence rates. The Espoo figure was given by the police in an interview and is not a published statistic.
In 2024 about 13,000 intimate-partner and domestic violence offences with 12,500 victims were recorded. 75% of suspects were men. Foreign-background people made up 19% of victims in 2023 and 18% in 2024, against an 11% share of the population.
Published on the day of the Sello mall shooting in Espoo, with Martti Lehti of the National Research Institute of Legal Policy giving background on violence among immigrants. Lehti says family violence is the largest single category among homicides committed by immigrants, that jealousy is a more prominent motive in them and that the perpetrator is typically sober — Russians are the exception, their homicides resembling those committed by Finns. The article's headline generalises this to violent crime as a whole, although Lehti's assessment is specifically about homicide. He also says immigrants' overall crime rate is substantially higher than the native population's, though the absolute number of offences is small. From the victim's side: an immigrant woman's risk of being killed is nearly twice that of a native-born Finnish woman, with no difference for men, and in 2002–2007 4% of women killed in Finland were foreign-born — but more than a third of their killers were Finns.
According to Markku Heinikari, deputy police chief of the West Uusimaa police department, nearly half of domestic-violence suspects in Espoo were born abroad, although foreign-language speakers make up over 25 per cent of Espoo's population. He says about 40 per cent of violent offences in the area are linked to domestic violence. The figure was given in an interview and is not a published statistic.
FGM and honour-based violenceSafetyValues and attitudes
Key figures
Over 10,000 women and girls living in Finland are estimated to have undergone female genital mutilation.
In a Statistics Finland survey (2024), 3% of men considered the genital mutilation of girls acceptable in certain circumstances. Among men under 25 the share was 10%, and among foreign-background men 8%.
About 23,000 people living in Finland have been pressured or forced into marriage (Statistics Finland, 2024). The assistance system for victims of human trafficking took on 66 forced-marriage victims in 2024.
Female genital mutilation originates in countries where the practice is common, and it came to Finland with migration. Honour-related violence is also discussed here in connection with immigration, but the survey evidence on it covers the whole population. Neither is recorded as an offence category of its own, so what is known rests on estimates and surveys.
FGM is a crime in Finland. THL estimates that about 10,000 women and girls living in Finland have undergone genital mutilation, and that 650–3,000 girls are at risk of it. In Statistics Finland's 2021 survey, 8% of 16–74-year-olds had witnessed or experienced honour-related violence, about 325,000 people. The same survey also measured forced marriage and attitudes to mutilation: about 23,000 people living in Finland have been pressured or forced into marriage, and 3% of men consider the mutilation of girls acceptable in certain circumstances, rising to 10% among men under 25. Survey data on values is under views of Europe’s Muslims and the overall picture of violence under violent crime.
Finland
Neither is recorded as an offence of its own, so these are estimates and survey findings rather than reported crimes.
Mutilation and honour violence in figures
10 000Women and girls in Finland who have undergone mutilation (THL estimate)
3 000Girls at risk of mutilation, upper estimate (THL)
8Of those aged 16-74 who have seen or experienced honour violence, %
Details
The THL figures are estimates, not measurements: mutilation is not recorded in health data as a category of its own.
The honour-violence survey covers the whole population, not only people of foreign background.
Nobody has measured what Muslims living in Finland think about mutilation. This is the closest measurement, and it separates respondents by origin: foreign background means both parents were born abroad. The shares among women are lower throughout, on average 2%.
Men who accept the mutilation of girls in certain circumstances
3Of all men aged 18-74, %
10Of men under 25, %
8Of foreign-background men, %
Details
The question was whether mutilation can be acceptable in certain circumstances. It measures whether a respondent rejects the practice outright, not any intention to carry it out.
Statistics Finland warns that young men and people of foreign background are a small group in the data: 1,423 men answered in all. The figures for these subgroups are indicative.
For comparison: circumcision of boys is accepted by 18% of men and 16% of women.
Mutilation is not a practice that arose in Finland but one that came with migration, so the scale of the question follows from which countries people came from. About 42,000 people living in Finland have a background country where more than half of women have been cut. Of them 28,000 are of Somali background, and in Somalia the practice is close to universal. The largest background group on this list by headcount is nonetheless the Iraqi one, and Iraq's national figure is low.
Prevalence of mutilation in background countries · share of women aged 15-49
The figure is the prevalence in the country of origin, not among people living in Finland. THL estimates that about 10,000 women and girls living in Finland have undergone mutilation.
Survey years differ by country because each figure is that country's most recent: Somalia 2020, Iraq 2018, Egypt 2015, Eritrea 2010, Nigeria 2024.
The practice follows region and ethnic group more than religion. In Ethiopia it stands at 82% among Muslims and 10% among the Orthodox, but in Nigeria at 14% among Muslims and 27% among followers of traditional religions, and in Kenya at 46% among Muslims and 68% among the traditional. In most Muslim-majority countries the practice does not exist at all.
Included are the countries that practise mutilation and have a background group of at least 200 people in Finland.
The population survey and the assistance system measure different things. The survey says how many people living in Finland have at some point been married against their will. The assistance-system figure says how many came to the authorities' attention during one year as victims of human trafficking. The second is a fraction of the first, and the gap is a gap in what comes to light, not in what happens.
Forced marriage on two measures
23 000People in Finland who have been pressured or forced into marriage
1Of ever-married under-75s, married against their will, %
66Forced-marriage victims taken on by the assistance system in 2024
Details
Most of those married against their will were married as a result of persuasion or pressure. Direct coercion is reported by 0.2% of those who have been married.
Statistics Finland notes that a population survey does not reach narrowly defined groups, and that forced marriage may be more common in some population groups than the population-level figure suggests.
The assistance-system figure shows no trend: 72 victims in 2022, 50 in 2023 and 66 in 2024. The relative share has still risen, and in the first half of 2025 forced marriage was the most common form of exploitation at 34%.
In 2024 forced marriage was written more explicitly into the human trafficking provisions of the criminal code. It is not an offence category of its own, so court statistics yield no series.
About 10% of Finns aged 16–74 have witnessed or experienced honour-based violence. More than one in ten people under 35 know at least a victim or perpetrator or have experienced it themselves. About 1% have been married against their will. The figure covers the whole population and includes both witnessing and experiencing the violence, not victims alone. These are survey answers, not population-adjusted or age- and sex-standardised offence rates, so the figure is not comparable with the standardised ratios in the Criminality section.
THL estimates that more than 10,000 women and girls living in Finland have undergone female genital mutilation. THL puts the number of girls at risk at 650–3,000. MAAMU study (THL Report 61/2012, ch. 8.2 · Koponen & Mölsä): about seven in ten women of Somali background (69%) and almost a third of those of Kurdish background (32%) said they had been circumcised. These are estimates, not population-adjusted offence rates.
EU Institute for Gender Equality report on the situation and trends of FGM in Finland — includes risk-group estimates, legislation and prevention measures.
A Statistics Finland report that measured, for the first time across the whole population, forced marriage, honour-related violence and attitudes to genital mutilation. About 1% of ever-married under-75s were married against their will, some 23,000 people. Honour-related violence has been witnessed or experienced by 8% of 16-74-year-olds, about 325,000 people, 80,000 of whom experienced it themselves. The mutilation of girls is considered acceptable in certain circumstances by 3% of men, rising to 10% among men under 25 and 8% among foreign-background men.
Annual review 2024 of the assistance system for victims of human trafficking: forced marriage accounted for 32% of new clients and became the second most common form of trafficking after forced labour (45%). There were 66 victims, against 50 in 2023 and 72 in 2022. There is no clear upward trend, but the relative share has risen. A forced marriage may have taken place wholly or partly abroad even when the victim lives in Finland.
An Yle MOT investigation into the practice known as dhaqan celis. A young person is sent abroad to a disciplinary institution under the guise of a holiday. Reporter Wali Hashi posed as a father looking for a placement and documented facilities in Somalia and Kenya: residents in chains, an electrified barbed-wire perimeter, and fees of several hundred euros a month. A report by Kriisivalmiusjärjestö, funded by the Ministry of Justice, had already confirmed that young people have been sent from Finland to such institutions. How many are affected is not known.
Street gangsSafetySocial cohesion
Key figures
According to the police situation report (autumn 2023), Finland has 10 street gangs with about 260 people linked to them. They are mostly men aged 18–30 and predominantly of immigrant background.
Street-gang violence in Sweden has featured heavily in the immigration debate, and in Finland the question has been whether the same phenomenon has reached the other side of the Gulf of Bothnia. There do appear to be signs of it: the police's national street-gang situation report (autumn 2023) identifies 10 street gangs in Finland and about 260 people linked to them, roughly half core members and half other associates.
Gang members are mostly young men and the majority have an immigrant background. No more precise official share has been published, and the higher figures aired in public are individuals' estimates rather than statistics.
The phenomenon is rooted in the Helsinki region, with signs of street-gang crime also in Turku, Tampere and Oulu. The overall picture of violence is under violent crime, and what is known about those born in Finland under the second generation.
Police street-gang situation report · autumn 2023
10Street gangs in Finland
260People linked to the gangs (approx.)
95Foreign background, % (2022 estimate)
Details
Of the people in the situation report, police assess about half to be core members and half other associates, such as accomplices.
The police national street-gang situation report: in autumn 2023 the police identified 10 street gangs in Finland, with about 260 people linked to them – about half core members and half other associates, such as accomplices. Gang members are mostly men aged 18–30, the majority with an immigrant background. The areas involved are socioeconomically disadvantaged with relatively large foreign-background populations.
In October 2022, Jonne Rinne, chair of the Finnish Police Federation, estimated that 11 street gangs with about 200 members operate in the Helsinki region and Turku. Over 90% of the members are adults and 95% have a foreign background.
The 95% foreign-background share is a public estimate by the police union chair, not an official police statistic.
According to the police, about 10 street gangs with roughly 200 members operate in Finland. Around 20 gang-related criminal cases from 2021–2022 were under investigation, with 23 suspects aged 18–29. Charges include attempted murder, aggravated robberies, extortion, firearms offences and witness intimidation. Helsinki police set up a new 16-person investigation unit against the phenomenon.
The second generation in crime statisticsSafetySocial cohesion
Key figures
In 2025 Finland-born 18–29-year-olds of foreign background were suspected of offences more often than either comparison group: 70% more often than people of Finnish background and almost twice as often as those born abroad. The age range is restricted because the second generation is much younger than the native population.
The Finland-born second generation of foreign background is almost always Finnish by citizenship, so citizenship-based crime statistics cannot separate it at all. In those statistics it is part of the native population. Statistics Finland also records suspects by origin, and in that split the group appears on its own from 2010 onwards.
The picture does not run one way. The overrepresentation of the foreign-born has many possible explanations that cannot apply to people born here: language, interrupted schooling, selection by the grounds for entry, time in the country and conditions in the country of origin. The Finland-born generation went to Finnish school and speaks the language, so those explanations do not hold for it. That is exactly why the figures are interesting, whichever way they point. The citizenship-based overall picture is under criminality and the same generation at school under education.
Citizenship-based crime statistics cannot see the Finland-born second generation at all, because it is almost always Finnish by citizenship. Statistics Finland also records suspects by origin, and there the group stands out. The chart shows each group's suspect rate relative to the Finnish-background rate, which is the baseline of 100.
Suspected offenders relative to the Finnish-background rate, aged 18 to 29 · Finland 2010–2025
Index, Finnish background = 100
Finnish background, born in FinlandForeign background, born abroadForeign background, born in Finland
Details
The ratio is used because the offence classification changed in 2021 and cut every group's absolute level at once. A ratio stays comparable across that break, absolute figures do not.
A suspect is not a convict. The figure counts suspicions recorded by the police, not guilt.
The age band is restricted to 18 to 29 because the second generation is far younger than the native population. Whole-population figures would compare groups of different ages.
In sexual offences the second generation is below the native level, as are its minors across all offences. The chart covers all offences and infractions together.
Show the figures
Year
Finnish background, born in Finland (Index, Finnish background = 100)
Foreign background, born abroad (Index, Finnish background = 100)
Foreign background, born in Finland (Index, Finnish background = 100)
One figure does not describe the group. When suspects are split by type of offence and age, the two generations fall on opposite sides: in sexual offences the overrepresentation is among the foreign-born and the Finland-born stay below the native level, while in overall crime among 18 to 29-year-olds the order reverses. The baseline of 100 is the Finnish-background rate in the same age band.
Suspected offenders by type of offence relative to the Finnish-background rate · Finland 2025
Index, Finnish background in the same age band = 100
Born abroadBorn in Finland
Details
The blue bar is foreign-background people born abroad, the red one those born in Finland.
The sexual-offence cells are small: 25 suspects among the Finland-born aged 18 to 29 and 56 among minors, so these figures move a lot from year to year.
The group is still a children's population. At the end of 2025 there were 102,178 of them and the median age was 9. Those under 18 numbered 79,599, or 78 per cent, and those aged 18 to 29, the band the overrepresentation concerns, only 17,381. The group doubled in ten years.
Foreign-background people born in Finland by age group · Finland 2015 and 2025
People
20152025
Details
The light bar is 2015 and the dark one 2025.
The age structure says how many will move into the ages where offending peaks over the next decade. It does not say what their offending level will be then.
The group also grows through births, so the under-18 cohort does not empty out as it ages.
The number of foreign prisoners has grown 75% in ten years: in 2025 an average of 839 foreign citizens were in prison daily, 23.6% of all prisoners, against a 7.7% population share.
The number of foreign prisoners has doubled in four years: an average of 434 a day in 2021 and 839 in 2025. In 2025 foreign citizens were 23.6 per cent of the prison population against 6.6 per cent of the population at large.
The ratio has to be computed with the same measure at both ends, however: prison statistics record citizenship, so they must be compared with the foreign-citizen share of the population rather than the foreign-background share.
Foreign background is not measured in prisons at all. The statistic knows only citizenship, so anyone granted Finnish citizenship appears in it as Finnish. The same holds for crime statistics more widely: they too record citizenship rather than country of background. Figures based on origin are under the second generation and the whole crime picture under criminality.
Details
The figure is an upper bound: prison counts include remand prisoners and people who do not live in Finland and are therefore part of no population denominator.
It is not standardised for age and sex, so it is not comparable with the standardised ratios in the Criminality section.
The two shares are from different years: the prison share is from 2025 and the population share from 2023, when there were 368,410 foreign citizens. The rise in the population share since then narrows the ratio a little.
In 2025 nearly half (46%) of foreign prisoners were on remand, and the largest citizenship groups were Estonians, Iraqis and Swedes.
Foreign prisoners, daily average · Finland 2014–2025
Prisoners, daily average
Foreign prisoners, daily average
Details
Foreign citizens (sentenced, fine-default and remand prisoners) as a daily average. The share view = foreigners' share of the average prisoner population.
Citizenship excludes naturalised citizens and foreign-background Finnish citizens.
Show the figures
Year
Foreign prisoners, daily average (Prisoners, daily average)
Criminal Sanctions Agency statistics 2025: 839 foreign prisoners per day (23.6% of the prison population), 78 nationalities. Growth +22% on the previous year. Largest groups: Estonia (112), Iraq (65), Sweden (64), Albania (44), Lithuania (43), Romania (42), Somalia (29). Narcotics offences 49.1% and sexual offences 20.6% as the principal offence.
Vantaa prison sees 60 nationalities a year. About 50% of remand prisoners are foreign. National trend: about 20 foreign prisoners in 1990 → 839 in 2025. 75% growth in ten years.
Yle, 17 September 2026: Finland's prisons hold about 3,900 prisoners, about 950 of them foreign citizens. The number of foreign prisoners has risen 76 per cent in ten years, and that rise accounts for about four fifths of the whole increase in the prison population. The most common principal offences are drug offences, sexual offences and homicides. The prison population is projected to reach 4,000 in 2028. For cost: a prison year costs an average of EUR 80,000 in Finland, while Sweden pays Estonia at least EUR 30.6 million a year for a minimum of 300 places, or EUR 102,000 per place.
Diversity and cohesion
Social cohesion should not be taken for granted. Several studies find a negative relationship between ethnic — and above all cultural — diversity and trust. The values held in Europe's Muslim communities, the prevalence of traditions such as cousin marriage, and the frequency of honour violence are examples of the deep cultural differences between people from the main refugee-sending countries in particular and the Finnish or European native population.
Sections in this group and their figures
Foreign-language men and military serviceAt call-ups in Uusimaa, 75% of Somali-speaking and 57% of Arabic-speaking conscripts failed to appear. The Defence Forces deny classifying conscripts by mother tongue (2026).
Views of Europe's MuslimsIn European surveys Muslim minorities' values differ markedly from the native population's, and the gaps have not narrowed. The data is from elsewhere, because the values of Muslims living in Finland have never been measured, but it is still indicative.
Religious communities and the state grantIn 2025 the state paid registered religious communities a general grant of 6.87 euros per member, 524,000 euros in all. The Evangelical Lutheran Church is not part of this allocation.
Islamic religious education in schoolsThe number of comprehensive-school pupils studying Islam grew 2.6-fold from 7,995 (2012) to 20,957 (2024). Islam is now the third-largest religious-education subject.
Consanguineous marriage worldwideIn Bradford, 62 percent of mothers of Pakistani background were married to a cousin in the 2007–2010 cohort and 46 percent in the 2016–2019 cohort. Among mothers born in Britain the share fell from 60 percent to 36.
Key figures
In ethnically homogeneous American communities, 70–80% of people trusted their neighbours a lot, against about 30% in the most diverse cities (2007).
1,959 stamped letters were dropped on the streets of 52 Berlin neighbourhoods (2011). The more diverse the neighbourhood, the less often a letter was posted on: a one standard deviation rise in local ethnic diversity nearly halved the return rate and more than tripled stamp theft.
1,001 estimates from 87 studies (2020): the association between ethnic diversity and social trust is negative and statistically significant.
In public debate it is common to present diversity as strengthening a community in itself. The meta-analysis that draws the whole literature together — 1,001 estimates from 87 studies — finds an association, but the other way: ethnic diversity goes with lower trust, most strongly for trust in neighbours and where diversity is measured at the level of the residential area.
The direction is settled, then, but the average effect is small. The same meta-analysis says so outright: the result does not support reading a large effect into it.
The individual studies below, from 2007–2016, represent the part of the evidence that finds a negative association. The meta-analysis heads the list because it draws in the whole literature rather than only this selection. Survey data on values is under views of Europe’s Muslims and residential sorting under residential segregation.
The meta-analysis gathers the whole literature into one figure, and Putnam's survey shows how large the gap is. The direction is confirmed, but the average effect is small, and the meta-analysis says so itself. In Putnam's data about 30 per cent of the people of Los Angeles and San Francisco say they trust their neighbours a lot, against 70 to 80 per cent in the ethnically homogeneous communities of North and South Dakota.
The size of the evidence base and the trust gap Putnam measured
1 001Estimates in the meta-analysis
87Studies they are drawn from
30Trust their neighbours a lot, the most diverse cities, %
70Trust their neighbours a lot, the most homogeneous communities, %
A meta-analysis covering the whole literature finds a statistically significant negative relationship between ethnic diversity and social trust.
The relationship is strongest for trust in neighbours and when diversity is measured at the neighbourhood level.
The average effect is small. The result confirms the direction of the negative relationship, but does not support reading it as a large one.
Ziller (2015) - Ethnic diversity and social trust in European regions20154 findings
European Social Survey 2002–2010 combined with immigration figures from the EU Labour Force Survey. The unit of analysis is regions within countries, not countries
As immigration increases, social trust declines. Both cross-sectional and longitudinal analysis confirm the result.
Economic recession and ethnic polarisation strengthen the negative effect further.
The association is not negative everywhere, however: the author reports signs of the opposite, a positive association, in areas of low ethnic polarisation. The effect is conditional on context rather than universal.
Published in Social Forces (Oxford University Press).
Koopmans & Veit (2014) - Ethnic heterogeneity and neighborhood trust in Germany20145 findings
EDCAS telephone survey in 55 German cities and regions, October 2009 – May 2010, over 7,500 interviews (4,200 respondents in the analysis). A priming experiment embedded in a survey
Ethnic and religious diversity in the neighbourhood lowers trust in neighbours.
The effect applies to both groups, not only the native population: diversity is associated with lower trust among native-background and immigrant-background residents alike.
The mechanism differs, though. For the native population trust is lowered by distant encounters and negative experiences with immigrants, while for immigrant-background residents it falls because positive contact with the native population is scarcer in diverse areas. The authors measured contact unusually precisely, separating friendships, chatting with neighbours, chance encounters, and experiences of discrimination and harassment.
The method is a priming experiment: the treatment group was experimentally made aware of the ethnic or religious heterogeneity of their own area, and their trust in neighbours came out lower than the control group's — so this is not merely a cross-sectional correlation.
Published as Koopmans & Veit, Social Science Research 47 (2014), 91–107: Ethnic diversity, trust, and the mediating role of positive and negative interethnic contact — A priming experiment.
Koopmans & Veit (2014) — a lost-letter field experiment in Berlin: diversity and cooperation20149 findings
Field experiment: 1,959 letters across 52 Berlin neighbourhoods, 17 January 2011. The sender's ethnicity and religion (German/Turkish, Christian/Islamic) were randomised experimentally.
A field experiment, not a survey: 1,959 stamped and addressed letters were dropped on pavements across 52 Berlin neighbourhoods on a single day (17 January 2011). The return rate was 63%, ranging from 32% to 88% by area. The measure is an actual act — picking a letter up and posting it — not an attitude, so social-desirability bias does not apply to it.
A one-standard-deviation rise in ethnic fractionalisation (0.19) almost halves the probability that a letter is returned (risk ratio 0.56). The effect survives controls for unemployment, population density, the number of tourists and businesses, the time of the drop, and East Berlin.
The same change more than triples the probability that the stamp is stolen (risk ratio 3.30). Theft is a direct measure of defection, and it gives the same result as the return rate — two different measures, one direction.
NO IN-GROUP FAVOURITISM WAS FOUND. Letters sent by a Turkish or Islamic organisation were returned as often as those from a German or Christian one, and the effect of diversity was the same for every letter type. The one exception was a lower return rate for Islamic letters in East Berlin. The result matches Putnam's finding that diversity lowers trust in one's own group as well as in others.
Segregation did not explain the result: the dissimilarity index had no significant effect on the return rate, nor did its interaction with diversity. The effect therefore appears in low-segregation settings too — Berlin's segregation is modest by international standards.
Ethnic diversity predicts cooperation better than the share of immigrants does: substituting the latter for the former weakened the effect and clearly worsened the fit of the whole model.
The effect is smaller in West than in East Berlin, and the difference is statistically significant. The authors read this as supporting the idea that long coexistence and contact dampen the effect over time, in line with Putnam's own expectation. The negative effect nevertheless persists in the West as well.
Because in-group favouritism does not explain the result, the authors consider differences in social control the most likely mechanism: in denser networks people have a stronger sense that their actions are noticed. They stress themselves that this remains a hypothesis without direct measurement.
Published as Koopmans & Veit, Political Psychology 35(3) (2014), 379–400. Same pair of authors as the priming experiment above, but a different study and a different method.
United States
Robert Putnam (2007) - Diversity weakens social capital2007About 30,000 interviews, United States11 findings
Data: the Social Capital Community Benchmark Survey (2000) — about 30,000 interviews across 41 US communities. Ethnic diversity measured with a Herfindahl index.
Trust in neighbours collapses with diversity: in ethnically homogeneous communities (e.g. rural South and North Dakota) 70–80% said they trusted their neighbours 'a lot', whereas in the most diverse cities (Los Angeles, San Francisco) only about 30% did.
The more ethnically diverse a residential area is, the less people trust one another — including members of their own racial/ethnic group. Yet the gap between in-group and out-group trust stays completely uncorrelated with diversity (R²=0.000): only the overall level of trust falls. This refutes both 'contact theory' (contact should raise trust) and 'conflict theory' (diversity should strengthen in-group trust). Putnam calls the pattern 'constrict theory'.
Scale of the effect: Putnam's own comparison finds the neighbour-trust gap between homogeneous Bismarck, North Dakota and diverse Los Angeles is roughly as large as the gap between areas with 7% versus 23% poverty rates, or between 0% and 36% college-educated residents.
In diverse areas, people vote less, do less volunteer work and give less to charity, participate less often in community activity, have fewer close friends, and report lower happiness and quality of life. Confidence in local government and local news media, and belief in one's own influence, are likewise lower, and TV-watching is higher — but political interest and participation in protests are, conversely, higher. Organizational and religious participation, by contrast, is not reduced by diversity once other factors are controlled for.
"Hunker down" effect: people withdraw into themselves and social isolation increases.
The effect appears across all racial groups, not only the majority.
The result holds after controlling for age, education, income, homeownership, poverty, crime, income inequality, population density and commuting time — so it isn't simply poverty or crime in disguise. The effect was negative in 39 of the 41 communities studied, and was independently confirmed using a separate dataset covering all 3,111 US counties. The pattern showed no generational difference: Americans raised in the 1970s were affected by diversity just as much as those raised in the 1920s.
Putnam is one of the best-known social scientists and the author of Bowling Alone.
Putnam's paper included an optimistic hypothesis: in the long run, societies can forge new, more encompassing shared identities that overcome diversity's negative effects ('long-run amelioration'). This was a theoretical prediction, however, with no empirical support presented in the paper.
Subsequent longitudinal research has not validated this optimism. Laurence & Bentley (2016), using an 18-year panel, found that sustained exposure to diversity reinforces withdrawal rather than reducing it. Replications of the Putnam effect across multiple countries have consistently confirmed the negative trust impact without the predicted recovery.
Australia
Wickes et al. (2014) - Ethnic diversity and social cohesion in Australian suburbs20144 findings
n=4,000 residents in 148 suburbs of Brisbane (147 in the analysis, those with complete register data), multilevel modelling. Australia serves here as a replication of Putnam's thesis outside the United States and Europe
In ethnically diverse areas social cohesion and neighbourhood interaction weaken (Putnam's "hunkering down" effect).
For neighbourhood interaction, diversity matters less to immigrants than to the population as a whole. Note that the source compares immigrants with the whole population, not with the native-background population specifically.
The authors state the result cautiously themselves: it gives Putnam's thesis "at least partial support", not confirmation as such.
Data: British Household Panel Survey, three waves (1991, 2001, 2011), 13,236 person-year observations. Diversity measured with Simpson's diversity index (0–1), using 1991-census wards as the community unit (mean population about 5,300).
In fixed-effects models, an increase in local ethnic diversity lowered the probability that a resident said they liked their neighbourhood — coefficient −3.92 (p<0.001) across the full sample, and −3.85 (p<0.05) among those who stayed in place.
The largest effect was among people who moved from a more diverse to a more homogeneous area: their attitude improved markedly (coefficient −12.32, p<0.01). No significant effect was found for people who moved from a more homogeneous to a more diverse area (coefficient −0.49, not significant) — suggesting this group had already self-selected into tolerance of diversity.
Attitudes were generally stable across the dataset: only 14% of respondents reported a change in their answer between waves.
The longitudinal, fixed-effects design minimises selection bias and supports a stronger causal interpretation than cross-sectional data.
Published in European Sociological Review (open access, CC-BY).
International
Foreign-language men and military serviceSocial cohesion
Key figures
At call-ups in Uusimaa, 75% of Somali-speaking and 57% of Arabic-speaking conscripts failed to appear. The Defence Forces deny classifying conscripts by mother tongue (2026).
Military service applies to all male Finnish citizens regardless of background. It is therefore one of the few directly measurable places where it can be seen whether society's obligations are shared evenly.
Foreign-language men complete military service at significantly lower rates than Finnish-language men. The other measures of integration are under language proficiency and education.
Finland
The figures are from call-ups in the Uusimaa regional office and rest on what Captain Lauri J. Mattila told Iltalehti. The Defence Forces deny that they break conscripts down by mother tongue, so there is no official statistic for these shares.
Failure to attend the call-up, by language · Uusimaa region
75%Somali-speakers who did not attend the call-up
57%Arabic-speakers who did not attend
70%Of these groups, exempted on health grounds
Details
The figures are for the Uusimaa regional office, not the whole country.
They come from one officer rather than from published Defence Forces statistics, and the Defence Forces deny keeping such a breakdown.
Ilta-Sanomat article on the avoidance of military service among foreign-language men in Finland. Foreign-language men miss call-ups or seek exemption more often than Finnish-speakers.
Sotilasaikakauslehti article examines the situation of foreign-language conscripts, call-up participation and integration challenges from the perspective of the Defence Forces.
According to Captain Lauri J. Mattila, 75% of Somali-speaking and 57% of Arabic-speaking conscripts failed to attend call-ups in the Uusimaa region, and 70% of these groups received exemptions on health grounds. The Defence Forces deny breaking down conscripts by mother tongue. The Conscripts' Union considers the underrepresentation worrying for both national defence capability and fairness.
Views of Europe's MuslimsSocial cohesionValues and attitudes
Key figures
In European surveys Muslim minorities' values differ markedly from the native population's, and the gaps have not narrowed. The data is from elsewhere, because the values of Muslims living in Finland have never been measured, but it is still indicative.
Muslim minorities' values are a recurring topic in the immigration debate, argued over in Finland too, although the values of Muslims living in Finland have never been measured. The closest thing is a Statistics Finland population survey that asked about attitudes to the genital mutilation of girls. It separates respondents by origin rather than by religion, so it is not a measurement of Muslims, but it is the only Finnish measurement in this section's subject area. It is the first box here. In European surveys, Muslim minorities' values differ markedly from the native population's, and the gaps have not narrowed so far.
The data comes mostly from British-Pakistani, French-Maghrebi and Austrian-Turkish communities, while Finland's Muslim population is mainly of Somali and Iraqi background. The results do not transfer to Finland as such, but they can still give an indication of the views of Muslims living in Finland. The size and growth of the group are under Muslim population growth, and the research on diversity and trust under ethnic diversity and trust.
This is the only question asked in the same form in several countries. The other surveys in this section ask about support for sharia in wordings that differ from one another, and are therefore not comparable.
Share who place religious rules above the law of the country
% of respondents
Details
THE SAMPLES ARE NOT THE SAME. The French figure covers all Muslims aged 15 and over, the Austrian one Viennese of immigrant background aged 14 to 24, and the German one young Muslims. The country levels are therefore not read directly against each other.
The wording varies slightly by country. France asked about placing religious rules above the law in certain situations, Germany about the rules of the Quran taking precedence over the law.
3% of men and 2% of women consider the genital mutilation of girls and women acceptable in certain circumstances.
Among men under 25 the share is 10%. It falls with age: 4% of men aged 25–34 and 2% of those aged 35–74.
Among foreign-background men aged 18–74 the share is 8%.
Circumcision of boys is accepted by 18% of men and 16% of women, so the mutilation of girls is rejected far more sharply than that of boys.
The survey separates respondents by origin, not by religion. Foreign background means both parents were born abroad. There is no separate figure for the views of Muslims living in Finland.
Statistics Finland notes itself that young men and people of foreign background are a small group in the data and that the figures should be read with caution.
United Kingdom
British Muslims' attitudes: the Israel conflict, antisemitism and demands on society202413 findings
26% consider the Israel-Palestine conflict the most important election issue, compared with 3% of the general population.
47% think Jews have too much power over British government policy, compared with 13% of the general population. The figure is 53% among men and higher among men aged 18-34.
Only 25% believe Hamas committed murder and rape on 7 October 2023. Among 18-34-year-olds, 53% believe Hamas did not.
46% sympathize more with Hamas than with Israel, compared with 27% of the general population and 53% among 18-34-year-olds.
25% believe Israel has a right to exist as a Jewish homeland, while 57% disagree.
63% want prayer rooms in public non-religious spaces, compared with 10% of the general population.
65% want Eid al-Fitr to be an official public holiday, compared with 21% of the general population.
57% want mandatory halal food in schools and hospitals.
52% want to ban showing images of the Prophet Muhammad, compared with 16% of the general population.
Only 23% consider sharia law undesirable, compared with 60% of the general population.
35% consider legalizing polygamy undesirable, compared with 70% of the general population.
28% consider banning homosexuality undesirable, compared with 62% of the general population.
Younger Muslims aged 18-34 and more educated Muslims are more radical on most questions.
British Muslims' values: sharia, terrorism, sexual morality and gender roles20168 findings
The French Interior Ministry's May 2025 report documents the activities of organisations linked to the Muslim Brotherhood (Ikhwan) in France.
It maps mosques, schools, associations, and political networks connected to the Brotherhood.
Authorities regard the Muslim Brotherhood as a significant vector of Islamism in Europe, though it is not formally banned.
The report is the French state's first systematic public mapping of Brotherhood-linked structures in the country.
Ifop - Muslims' relationship with Islam and the current state of Islamism20256 findings
1,005 Muslims drawn from a nationally representative sample of 14,244 people (aged 15 and over, mainland France). Fieldwork 8 August – 2 September 2025
65% of Muslim youth hold that Islamic rules apply to all areas of life and must be strictly observed.
53% hold that women should wear a headscarf in public.
41% place religious rules above Austrian law.
43% blame the West for the problems of the Islamic world.
The study's central — and largely overlooked — finding: algorithm-driven social media use is the single strongest predictor of antidemocratic attitudes, stronger than ethnicity, education or religion alone. Intense religious online engagement correlates with political-religious authoritarianism (r = 0.46), acceptance of violence (r = 0.33) and patriarchal gender views (r = 0.28).
Muslims in Vienna schools (Bildungsdirektion Wien, via Die Presse / heute.at)20264 findings
All primary, middle, special, and polytechnic schools (excluding gymnasiums), Vienna
32.8% show some degree of discriminatory attitude toward other religions.
10% consider their own religion superior. Among Somalis the figure is dramatically higher at 56.1%, compared with 11.1% among Sudanese and 7.5% among West Africans.
52.3% consider Quran teaching more important than ordinary schooling: 63.3% of men and 38.3% of women.
52.8% accept that a Muslim woman may choose her own spouse. Men are clearly more reserved at 36.6%.
Muslims in Austria (Peter Ulram / Ecoquest, Integration Fund)20174 findings
45.1% of Muslims under 40 show open or latent Islamist attitudes: 11.5% clearly Islamist and 33.6% latent.
29.1% expressed open antisemitism.
25.1% consider that the rules of the Quran should take precedence over German law.
23.8% regard an Islamic state as the best form of government.
Background: the median age of the Muslim population is about 32 (whole population about 44). Germany is estimated to have 2.5–3.5 million Muslims aged 18–40. MOTRA is an extremism-monitoring network funded by federal ministries. The same figures on Islamism and antisemitism have also been reported as a BKA study. They come from the monitor's chapter by Brettfeld, Wetzels, Fischer, Farren and Endtricht, “Menschen in Deutschland 2021 bis 2025” (the MiD survey series, 21,899 respondents 2021–2025).
A BKA study found that 45.1% of Muslims under 40 in Germany harbour latent or open sympathy for Islamism (Sharia over the constitution). 11.5% hold manifestly Islamist attitudes. Antisemitic attitudes among young Muslims rose sharply from 11.3% (2021) to 29.1% (2025).
Religious communities and the state grantEconomyValues and attitudes
Key figures
In 2025 the state paid registered religious communities a general grant of 6.87 euros per member, 524,000 euros in all. The Evangelical Lutheran Church is not part of this allocation.
The state pays registered religious communities a general grant, which the Ministry of Education and Culture divides in proportion to membership. In 2025 every community received the same EUR 6.87 per member, and the whole appropriation was EUR 524,000. The Evangelical Lutheran Church is outside this division and is compensated separately for its societal duties, EUR 117 million in 2021.
The Islamic communities' share of the grant is therefore their share of registered members, not a choice by the ministry. Registered membership is a fraction of Finland's Muslim population, which has a section of its own. The series shows how many people have joined a community and how much money that has meant over the years.
The state pays a yearly general grant to registered religious communities, that is to everyone other than the Lutheran and Orthodox churches. The Ministry of Education and Culture divides the whole appropriation between the applicants in proportion to registered membership, so in 2025 every community received the same EUR 6.87 per member. A community's share of the money is its share of the members, not a judgement about the community.
The state's general grant to registered religious communities · 2025
524 000The appropriation, EUR per year
25.9Islamic communities' share in 2025, %
19 750Registered members of Islamic communities, 2025
986 714To Islamic communities in 2011-2025 in total, EUR
Details
Membership is the register figure from the end of the previous year, and it decides the whole division.
The appropriation was EUR 200,000 a year in 2008-2015, EUR 532,000 in 2016 and EUR 524,000 from 2017 on.
The years 2013 and 2021 are missing from the cumulative figure, because no breakdown was published for them, so it is an underestimate.
For scale: state financing for the Lutheran church's societal duties was EUR 117.0 million in 2021 (act 430/2015). It pays for burial services, population registration and historically valuable buildings rather than for a community's own activity, so the two sums are not the same thing.
The share rose from 12.4 percent to 25.9 percent while the appropriation stayed the same. Because the division follows membership, the series measures registered membership: Islamic communities had 5,639 members in 2011 and 19,750 in 2025.
Islamic communities' share of the general grant · 2011-2025
% of the appropriation
Details
The years 2013 and 2021 are missing because no breakdown was published for them. No bar is filled in with an estimate.
Registered membership of Islamic communities is a fraction of Finland's Muslim population: most do not belong to any registered community, so this does not measure the number of Muslims.
The 2016 appropriation was EUR 532,000 as a one-off and in 2019 EUR 518,940 was paid out.
All 30 communities that received a grant, with the Islamic ones highlighted. The order is the order of membership, because the grant is membership multiplied by the same number for everyone.
The general grant by community · 2025
Katolinen kirkko Suomessa
109 779
Suomen Vapaakirkko
96 762
Suomen Helluntaikirkko
90 815
Suomen Islamilainen Yhdyskunta
31 178
Ortodoksinen Pyhän Nikolauksen Seurakunta
25 292
Helsinki Islam Keskus
23 794
Suomen Adventtikirkko
19 680
Tampereen Islamin yhdyskunta
19 027
The Islamic Rahma Center in Finland
14 260
Turun Islamilainen Yhdyskunta
10 482
Pohjois-Suomen Islamilainen Yhdyskunta
10 200
Resalat Islamilainen Yhdyskunta – Suomen Ähl-e Beit
9974
Suomen Baptistikirkko
9823
Helsingin juutalainen seurakunta
6210
Espoon Islamilainen Yhdyskunta Suomessa
5571
Finlands svenska baptistsamfund
5323
As-Salaam Islamic Center & Masjid As-Salaam
4918
Suomen Metodistikirkko
4801
Suomen vietnamilaisten buddhalaisten yhdyskunta
4692
Finlands svenska metodistkyrka
4101
Helsingin Islam Seurakunta
3998
Suomen evankelis-luterilainen seurakuntaliitto
2878
Ortodoksinen Pokrovan seurakunta
2384
Suomen Bosnjakkien Islam-yhdyskunta
2260
Missionskyrkan i Finland
1848
Helsingin Debre Amin Abune Tekle Haymanot -seurakunta
1374
Suomalais Vietnamilainen Buddhadamma Yhteisö
941
Harvest Church Finland
886
Danakosha Ling
385
Uuden toivon seurakunnat
364
Show the figures
Group
EUR
Katolinen kirkko Suomessa
109,779
Suomen Vapaakirkko
96,762
Suomen Helluntaikirkko
90,815
Suomen Islamilainen Yhdyskunta
31,178
Ortodoksinen Pyhän Nikolauksen Seurakunta
25,292
Helsinki Islam Keskus
23,794
Suomen Adventtikirkko
19,680
Tampereen Islamin yhdyskunta
19,027
The Islamic Rahma Center in Finland
14,260
Turun Islamilainen Yhdyskunta
10,482
Pohjois-Suomen Islamilainen Yhdyskunta
10,200
Resalat Islamilainen Yhdyskunta – Suomen Ähl-e Beit
9,974
Suomen Baptistikirkko
9,823
Helsingin juutalainen seurakunta
6,210
Espoon Islamilainen Yhdyskunta Suomessa
5,571
Finlands svenska baptistsamfund
5,323
As-Salaam Islamic Center & Masjid As-Salaam
4,918
Suomen Metodistikirkko
4,801
Suomen vietnamilaisten buddhalaisten yhdyskunta
4,692
Finlands svenska metodistkyrka
4,101
Helsingin Islam Seurakunta
3,998
Suomen evankelis-luterilainen seurakuntaliitto
2,878
Ortodoksinen Pokrovan seurakunta
2,384
Suomen Bosnjakkien Islam-yhdyskunta
2,260
Missionskyrkan i Finland
1,848
Helsingin Debre Amin Abune Tekle Haymanot -seurakunta
The two national churches are not included: they are financed differently and do not apply for this grant.
The same series as a cumulative total per community: EUR 986,714 to 17 communities over 13 years. It is an underestimate, because no breakdown exists for 2013 and 2021.
Grants paid to Islamic communities in total · 2011-2025
A community's name is spelled differently from year to year, and the cumulative figure merges the spellings of one community.
A community appears only for the years it received a grant. Some were founded or registered during the period.
Islamic religious education in schoolsValues and attitudes
Key figures
The number of comprehensive-school pupils studying Islam grew 2.6-fold from 7,995 (2012) to 20,957 (2024). Islam is now the third-largest religious-education subject.
In Uusimaa 8.8% of pupils in grades 1–6 study Islam, against 4.2% nationally and 0.4% in Central Ostrobothnia (2024).
The mix of religious-education subjects is one measure of how demographic change shows up in school. It follows from what guardians request, so it also reflects families' own choices.
Some Islamic teaching happens outside school and is not in these figures at all. Islamic congregations teach the Quran — several dozen congregation premises, most of them, by Yle's 2012 survey — and parents teach their own children at home before the formal teaching that starts around age seven. How common this is nobody knows: Muslims in Finland are not counted as such, so there is no figure for how widespread Quran teaching is either. Professor Jaakko Hämeen-Anttila's assessment is that far from all parents consider it important.
The school subject and the congregation's teaching are also not measuring the same thing. Islamic religious education at school is knowledge about the faith rather than its practice — it does not include praying, for instance — and religious upbringing stays with the family (Putkonen, University of Helsinki 2023).
The Basic Education Act obliges municipalities to arrange instruction in a pupil's own religion when guardians request it.
The mix of subjects has shifted fast: the share studying Evangelical Lutheran religion has fallen markedly while secular ethics and Islam have grown. The number studying Islam has multiplied, making it the third-largest subject.
Regional differences are large: in Uusimaa the share of primary pupils studying Islam is many times what it is in the regions that have received the least immigration.
Enrolment in Evangelical Lutheran religion fell from 495,279 (2010) to 484,140 (2012) and 428,926 (2024) — from 92.9% to 79.3% of all pupils. Over the same period secular ethics rose from 3.2% to 13.5%, Islam from 1.5% (2012) to 3.9% and Orthodox religion from 1.3% to 2.2%.
Pupils studying minority religious-education subjects · Finland 2012–2024
Pupils
Secular ethicsIslamOrthodox
Details
The Islam figure is a floor, not a count of Muslim pupils. Instruction in a pupil's own religion is arranged only where at least three pupils belonging to the same registered religious community attend the schools and their guardians ask for it. A pupil whose family does not ask, or who belongs to no registered community, is counted under some other subject.
Instruction given outside school does not appear in the statistic at all. Karvi's table does not separately show the pupils who take no religious-education subject at school and study their religion in their own religious community's teaching instead.
The statistic is not annual, so the bar groups are unevenly spaced.
No separate Islam figure exists for 2010 (it was included in other religions), so the series starts in 2012.
The dominant subject, Evangelical Lutheran religion, is omitted from the chart for scale.
The subject reflects participation in instruction, not a pupil's or family's religious conviction: some pupils of Muslim background study secular ethics or the majority religion.
Niina Putkonen, Helsingin yliopisto, kasvatustieteellinen tiedekunta2023
Summary
The number of children and young people taking part in Islamic religious education in basic education has grown in recent years. The teaching is knowledge about the religion rather than its practice: it does not include prayer, for example, and religious upbringing is left to the family. As an alternative to the school’s teaching, a child may at the guardian’s request be taught by their own religious community.
The Quran is taught in Islamic congregations. Finland has a few dozen congregation premises and most of them teach the Quran. Teaching typically begins at about seven, and before that children are taught by their parents at home. There is no figure for how common it is: "There is no exact information on how common Quranic teaching is, because not even the number of Muslims living in Finland is recorded." Professor Jaakko Hämeen-Anttila estimates that far from all parents consider the teaching important.
Juttu on vuodelta 2012 ja arvioi Suomen muslimien määräksi noin 40 000. Tämän sivuston oma luku on nykyisin noin 140 000, joten mittakaava on vanhentunut. Lähteestä käytetään kuvausta opetuksen luonteesta ja siitä, ettei sen yleisyyttä tilastoida.
In 2016 Helsinki spent €1.2 million a year on own-religion instruction, of which Islamic education accounted for about €800,000. One weekly lesson cost about €2,000 a year. 5,400 pupils attended own-religion instruction, with 14 teachers of Islam. Nationwide, about 9,000 pupils studied Islam at the time.
In Helsinki, 4,746 basic-education pupils (9%) studied Islam in 2018. In Espoo 2,349 (8%) and in Vantaa 2,268 (9.6%) in 2020. The share in Evangelical Lutheran instruction is declining: in Helsinki grades 1–6 from 72.4% (2012) to 60.8%. Secular ethics was studied by 21.3% of Helsinki pupils.
Consanguineous marriage worldwideValues and attitudes
Key figures
In Bradford, 62 percent of mothers of Pakistani background were married to a cousin in the 2007–2010 cohort and 46 percent in the 2016–2019 cohort. Among mothers born in Britain the share fell from 60 percent to 36.
Cousin marriage is not an incidental habit but an old family institution. Across parts of the Middle East, North Africa and South Asia, marrying within the extended family keeps property, alliances and the choice of spouse inside the kin group. Where the practice is the norm, that choice belongs to the family more than to the individual — which is what makes it relevant to integration: the custom does not disappear on changing country, it changes over generations.
Change is real but slow, and it starts high. In Bradford, 62% of mothers of Pakistani heritage were married to a cousin in the 2007–2010 cohort and 46% in the 2016–2019 one. Among mothers born in the UK the share fell from 60% to 36%. In the short run that decline settles nothing: nearly half is still nearly half, the marriages entered into now set the health and coercion questions of the next generation, and in the countries of origin the level is still 40–65%, so each new arrival brings the practice at its source level. Change measured in generations is not the same thing as a problem solved.
The health effect is a second, separate reason to follow the phenomenon: how common cousin marriage is in a country of origin predicts a raised prevalence of congenital disorders, and therefore of care needs, in the receiving country too. Several European countries have moved to legislation. The research findings and the details of that legislation are in the source boxes below.
The map compiles country-level research estimates, a risk map, not a population register and not an estimate about individuals.
International
Sweden bans cousin marriage from 1 July 20262026Decision of the Swedish Riksdag, 26 May 20265 findings
Among mothers of Pakistani heritage, 62% were married to a cousin in the first cohort and 46% in the second.
The fall was steepest among mothers born in the UK: from 60% to 36%. The practice changes fastest where the second generation grew up in the receiving country.
Education separates the same way: among those educated beyond A-level the share fell from 46% to 38%.
John Wright, the study's chief investigator, summed it up as cousin marriage shifting in under a decade from a majority activity to a minority one.
The figure therefore moves in both readings: the practice does not vanish on changing country, but neither does it stay fixed.
Newer national estimates of cousin marriage20185 findings
National surveys and peer-reviewed compilations. The Consang.net compilation behind the map is partly older
Pakistan: in the 2017–18 national health survey (PDHS) about 64% of married women were related to their husband. The map's Consang.net value of 61.2% is of the same order.
India: the national family survey NFHS-5 (2019–21) put the share at about 11%, against 9.9% in NFHS-4 (2015–16). The map's 7.5% is therefore likely an underestimate.
Arab states: Tadmouri et al. (Reproductive Health 2009) compile national ranges — Iraq 47–60%, Sudan 44–63%, Saudi Arabia 42–67%, Yemen 40–45%, Syria 30–40%, Egypt 21–33%. A single number on a map cannot show that spread.
Turkey: the statistics office TÜİK reports that 3.2% of marriages contracted in 2023 were between first cousins (2010: 5.9%). That figure is not comparable with the map: it covers marriages contracted in one year and first cousins only, where the map describes all existing consanguineous unions.
The trend is downward in several countries with urbanisation and women's education, but the change shows first in newly contracted marriages and only later in the population-level stock.
Born in Bradford: cousin marriage and congenital anomalies20135 findings
Sheridan et al., The Lancet 2013;382(9901):1350–9. The Born in Bradford birth cohort, recruited 2007–2011: 13,776 babies, of whom questionnaire data are available for 11,396. Of these, 386 (3%) had a congenital anomaly – 305.74 per 10,000 live births against a national level of 165.90
Cousin marriage roughly doubled the risk of a congenital anomaly: adjusted risk ratio 2.19 (95% CI 1.67–2.85). The association held after adjusting for deprivation — the authors found no association with increasing deprivation at all. (The 6.2% vs 2.6% figures often quoted in public are consistent with these numbers but do not appear in the abstract, so they need checking against the full article.)
The absolute increase is modest: the great majority of children born to cousin couples are born healthy. Public debate often overstates the risk.
At population level the effect is still substantial where the practice is common: about 31% of congenital anomalies among children of Pakistani origin were attributable to consanguinity. The practice was common in this cohort: 18% of all babies and 37% of those of Pakistani origin (1,922 of 5,127) were the children of cousin unions.
The risk stems from the inheritance of recessive gene variants, not from ethnicity as such. The same risk applies to any cousin couple. A comparison point comes from the same data: among white British mothers over 34, risk rose by roughly the same amount (risk ratio 1.83, CI 1.14–3.00). Age and consanguinity are therefore risk factors of the same order.
The authors' other conclusion is usually left out of public debate: a mother holding a degree was protective across every ethnic group (risk ratio 0.53, CI 0.38–0.75). They recommend sensitive risk counselling both for at-risk groups and for couples in cousin unions to support reproductive decisions — not prohibitions.
Consang.net — international consanguinity tables and dynamic map7 findings
The table metric is Consanguinity(%): the share of unions where spouses are biological relatives. In practice the source usually compiles estimates up to second cousins or closer.
The material is not a uniform current statistic. It includes registries, hospital samples, household surveys, and community-level studies from different decades. Study periods in the source's own tables run from the 1940s to the 2010s: Nagasaki, Japan 1948–53, urban Beijing 1949–67, rural Tamil Nadu 1969–74 and all-Pakistan 2012–13, for example.
Some country figures rest on a sample from one region or one city rather than on national data. A country's colour therefore reflects, in varying degrees, both its level and where and when a study happens to exist.
Seven countries carry a named national source and year on the map (visible on hover): Pakistan's and India's figures are replaced with recent national surveys (PDHS 2017–18: 64%, NFHS-5 2019–21: 11%), and for Iran, Afghanistan, Iraq, Syria and Turkey the source and its limits are stated as they are.
The map shows the 72 countries for which the source has an estimate. The rest are grey: the source's map layer records them as zero, but that stands for a missing study rather than a measured zero — Finland, Denmark, Austria and Greece, for instance, have no row in its tables at all.
The highest country-level compiled values cluster in the Middle East, North Africa, and South Asia. Most European and American countries are low.
Map geometry is generated locally from Natural Earth's 1:110m country polygons. The browser does not load an external map library or map data.
The asylum system
The asylum system is justified as protecting the persecuted and helping the world's poorest and most vulnerable. Most applicants, however, do not meet the criteria for international protection, the system does not reach the world's poorest, and brain drain leaves source countries worse off than before. Population growth in Africa and the Middle East is explosive. Even if Europe took in only a small share of those countries' people in need, climate migrants or those living in miserable conditions, it would place considerably greater strain on the receiving countries than it does now, while doing practically nothing for the countries they leave.
Sections in this group and their figures
Recognition rates and protection need39% of the EU's first-instance asylum decisions in 2025 were positive (325,765 / 832,360), and 21% of appeal decisions.
Establishing ageIn 2025, 58.1 percent of unaccompanied minor asylum applicants in Finland and 73.1 percent in the EU27 declared an age of 16 or 17.
Migration and global povertyThe EU receives about a million asylum applications a year, under 0.15% of the world's 700 million people living in extreme poverty (2024). Nor are the arrivals those people: reaching Europe costs thousands of euros, which someone in extreme poverty cannot pay.
People smuggling and boat arrivalsSmuggling fees range from under $20 to over
5,000 depending on the route, the mode of travel and the client's background. Payment is usually cash or a hawala-type system (2026).
Illegal stay and returns in FinlandIn 2025 Finland ordered 5,070 people to leave the country and recorded 1,915 returns. About two in five were carried out.
Safety at asylum reception centres43% of reception-centre staff had faced violence or threats at work in the previous 12 months, most often from a resident of the centre (survey 2016).
Reception-system costsReception-system costs track applicant numbers steeply: they peaked at €607 million in 2016, against about €100 million in 2021.
Recognition rates and protection needSafety
Key figures
39% of the EU's first-instance asylum decisions in 2025 were positive (325,765 / 832,360), and 21% of appeal decisions.
Of those who arrived in Italy by sea in 2016, only 2.65% were granted refugee status under the Geneva Convention.
The EU asylum system rejects a majority of applications. The share of positive decisions is well under half at first instance and lower still after appeal, so an appeal rarely changes the outcome.
Differences between nationalities are stark: in the same year most applicants of one nationality are recognised and only a fraction of another. A recognition rate measures an administrative outcome: it says who met the standard of proof, not directly who was in need of protection.
Finland's own figures say the same. In 2016, 27.5% of decisions were positive. By 2018 the share was 43%. It also moves with the line taken: among Iraqi applicants aged 18–34, 85% received a positive decision in 2015 and 33% in 2017, after the service reassessed the security situation in Iraq.
No case is known of a person returned from Finland having been persecuted, and nobody follows returnees systematically.
What happens after a rejection is not comprehensively recorded: about 2,100 people used assisted voluntary return in 2016 and about 1,400 in 2017, but no figures are published on interrupted removals, and forced return does not work to some countries of origin. The Ministry of the Interior concludes that this has created a new group of people staying in Finland unlawfully. How this came about is on the timeline from 2015 onwards.
Details
32,477 asylum applications were lodged in Finland in 2015. In 2016 the Finnish Immigration Service decided 28,208 international-protection cases, of which 7,745 were positive.
Of the 8,854 appeals the administrative courts decided in 2017, 2,976 — about a third — succeeded. The most common reason was new evidence (1,639 cases) rather than an error by the service. A procedural error was recorded in 360.
People who fled Ukraine are not in these shares at all: they were received under the temporary protection directive, outside the asylum procedure.
The UN Human Rights Office has documented the detention, torture and intimidation of Afghans that Iran and Pakistan returned to Afghanistan. The risk falls above all on women, people who served the former government or its security forces, journalists and civil-society workers.
For people returned from Europe the cases are few in all, though the individual ones draw exceptional attention in the press and the courts.
At first instance 325,765 of 832,360 decisions were positive, and on appeal 35,565 of 170,175. An appeal therefore overturns a decision less often than a first hearing accepts an application.
Share of positive asylum decisions in the EU · 2025
% of decisions
Details
The shares are of decisions, not of applicants. The same applicant can receive both a first-instance and an appeal decision, so the figures do not add up.
The recognition rate varies sharply by nationality: it is high for Syrians and under 20 per cent for many North and West African countries.
Both international protection and protection granted on national humanitarian grounds are included.
Recognition rates: most applicants do not meet the criteria for protection
In 2025 the EU issued 832,360 first-instance asylum decisions, of which 325,765 (39%) were positive. At the appeal stage 170,175 decisions were issued, of which 35,565 (21%) were positive. Taking both stages together, about 43% receive protection and about 57% do not — this is the most honest single figure. The extremes: an appeal overturns the decision in only about one case in five, and recognition rates vary widely by nationality — high for Syrians and Afghans, under 10% for many countries in Africa and Asia. The figures are decisions taken in the same year rather than a follow-up of the same cohort, so the combination is an approximation.
A full review of the asylum process from application to return, commissioned by the Ministry of the Interior. In 2015 there were 32,477 asylum applications. In 2016 the Immigration Service decided 28,208 international-protection cases, of which 7,745 or 27.5% were positive. In 2018 there were 4,548 applications and a 43% positive share. The administrative courts decided 8,854 appeals in 2017 and upheld 2,976 of them, and in 2018 upheld 2,370 of 6,821. The most common ground for a successful appeal was new evidence (1,639 cases in 2017) rather than an error by the service. A procedural error at the Immigration Service was recorded in 360 cases and an error of law in 17. The review also states that no statistics are available on interrupted removals or their reasons, and that legislative changes have created a new group of people staying in Finland unlawfully, because returns do not succeed to some of the main countries of origin.
Comprehensive academic analysis of European asylum decisions 2003–2017. The average overall recognition rate was ~30% — but formal refugee status under the Geneva Convention was granted to only ~15%. Outside the crisis years 2015–16, more than half of applications were rejected at first instance. Hatton himself notes that these are first-instance figures: recognition rates would be higher, on average by around 7 percentage points, if successful appeals and reviews were counted — but even then more than half of all applications are ultimately rejected. Recognition rates vary sharply: high for conflict countries (Syria, Eritrea), below 10% or even below 5% for many African and Asian countries. Hatton notes that many applicants do not meet the strict persecution criteria — findings suggesting a significant share of applications are economically motivated.
EU member states granted protection status to around 361,000 applicants in 2025 (an 18% fall from the previous year). The first-instance recognition rate was 39.1% — meaning more than 60% of applicants were rejected at first stage. In final decisions after appeal, the recognition rate fell to 20.9%. A large share of those rejected file appeals, clogging the system for years.
Eurostat's official statistics on EU asylum decisions. Recognition rates vary sharply by nationality: Syrians (~90%) and Afghans (~70%) are recognised at high rates, whereas applicants from many West African, North African, and South Asian countries fall below 10–20%. There are also significant differences between member states — the same nationality can receive very different outcomes in different EU countries.
MPI's interactive data visualisation of asylum recognition rates in EU and EFTA countries 2008–2017. Shows how sharply approval rates vary both by country and by nationality: the same applicant can receive a completely different decision in different EU countries. This inconsistency suggests decisions are not based solely on the applicant's objective protection need but also on the receiving country's political and administrative practices.
Italy 2016: how many of the sea arrivals received protection
In 2016, 181,436 people arrived in Italy by sea (Italian Ministry of Interior statistics). Only 4,808 received formal refugee status under the Geneva Convention — just 2.65% of all arrivals. Of the 123,600 who applied for asylum, 60% (54,254) were rejected. About 57,800 did not apply for asylum in Italy at all but continued to other EU countries. The share compares Geneva-Convention refugee status with all arrivals, not only applicants. Broader measures that include subsidiary and humanitarian protection produce higher rates.
Fact-check of Italian Ministry of Interior (Viminale) statistics. The article works through different calculation methods: strict refugee status vs. all forms of protection, applicants vs. all arrivals. In 2014, about 38% of all sea arrivals applied for asylum in Italy. Of those applicants, about 10% received refugee status and about 50% received some form of protection. At the same time, about 62% of applications were rejected. Open Migration criticises the narrow 2.65% figure — yet it confirms the same underlying data.
In 2016, 2.65% of sea arrivals received formal refugee status. Il Sole 24 Ore reported the figure from these statistics in April 2017. The daily cruscotto statistico contains sea arrival counts, asylum applications and decisions.
Gender and age profile: arrivals are predominantly young men
Refugee flows registered in UNHCR camps typically have a broadly balanced sex and age profile: families, women and children. Those crossing the Mediterranean to Europe show a clear male majority. The route itself selects young men regardless of motive, because crossing the Mediterranean and the Sahara is life-threatening and costs thousands of euros in smugglers' fees, while UNHCR camps sit next to the conflict and are within reach of families. Migration research also documents the household strategy in which a family sends the member most likely to earn — typically a young man — as a way of diversifying risk. The same distribution therefore arises from both economic and humanitarian motives.
Frontex (European Border and Coast Guard Agency)2016
Summary
The EU border agency Frontex's annual risk analysis. Reports that in 2014 about 74% of those crossing the Mediterranean were adult men, 11% women, and 15% children. During the 2015 crisis peak the composition shifted as the Syrian civil war brought more families, but adult men remained the majority afterwards. Frontex emphasises that mixed flows contain both people genuinely needing international protection and economic migrants.
A survey of about 22,000 people who arrived in Europe in 2015–2016 (IOM Flow Monitoring, Gallup World Poll). The study asks who leaves: whether those who go are better educated than the average in their home country or not. Refugees and women who arrived without permission are better educated than that average. Men who arrived without permission are not. What decides the direction is the situation at home: from war-torn countries education raises the likelihood of leaving, from peaceful ones it lowers it. This means there is no single answer to whether arrivals are refugees or people moving for economic reasons. The answer depends on the country of origin. Of those from Afghanistan, Eritrea, Iraq, Somalia, Sudan and Syria, more than 90% gave conflict or persecution as their reason for leaving, while most of those from Morocco and Algeria gave economic conditions. Only 3% named a lack of basic services, food or shelter as their main reason.
UNHCR's official operational data portal for maritime arrivals in Europe, with counts by route and receiving country and the numbers dead or missing. Its monthly situation reports also break arrivals down by sex and age: in 2015, 58% of arrivals across the Mediterranean were adult men, 17% women and 25% children. In Q1 2016 the shares shifted to 47% adult men, 20% women and 34% children as the Syrian refugee flow grew.
Eurostat's official database of asylum applicants by age and sex (migr_asyappctza). In December 2025, men aged 18–34 accounted for 38% of all first-time applicants — by far the largest single age-sex group. Historically, the male share was substantially higher during the 2015–16 crisis. The database enables comparison by country and year.
Age assessments: most of those examined were assessed as adults (2011–2016)
Status as an unaccompanied minor brings advantages: a lighter reception procedure, a guardian and better prospects of family reunification. Age is verified with a forensic examination (dental and wrist X-rays). In 2011–2015, 56–79% of those examined were assessed to be adults: of the 149 examined in 2015, 90 (60%), and of the record 630 examinations in 2016, about 65%. The shares cover those examined, not everyone who claimed to be a minor: an examination is ordered where there is reason to doubt the stated age, so the share for the whole group cannot be read from these figures. The method is accurate only to about ±2 years and has been criticised on medical grounds. The number of examinations collapsed after 2017, so the figures describe 2011–2016.
In 2016 a forensic age assessment was performed on 630 asylum seekers who had claimed to be minors — a record number. About 65%, roughly 410 of those tested, turned out to be adults. The test is based on dental and hand/wrist X-rays.
A review by forensic dentists: in 2011–2015, 56–79% of examined persons who had claimed to be under 18 were assessed to be over 18. In 2016, 642 examinations were performed (630 ordered by authorities). From the start of 2017, responsibility for the examinations moved from the University of Helsinki to THL.
Asylum shopping: applicants move on from the first country of arrival
A significant share of asylum seekers move within the EU beyond the first country of arrival. The Commission describes this as a search for the best conditions in which to apply. What those conditions consist of is where the sources disagree. A legal analysis puts the weight on labour-market prospects and family networks and treats welfare as one factor among several. A Danish natural experiment shows benefit levels moving application numbers, though the size of that effect is disputed. This so-called asylum shopping does not itself shift the burden onto border states — that burden already follows from the Dublin Regulation's first-country-of-entry rule. Shopping instead strains the system as a whole: when an applicant lodges claims in multiple countries, transfers back to the country of first entry often fail in practice.
A peer-reviewed study of when generous benefits pull migrants in and when they do not. It compares two migration routes. Movement within the EU, Norway and Switzerland is free: the migrant picks the country and nobody screens them. Migration from outside is restricted: the destination decides who is let in. On the free-movement route the magnet effect is there: the more generous the welfare, the less educated the arrivals. On the restricted route the result reverses, the more generous the system the better educated the arrivals, because voters do not want to admit net recipients and permit policy selects accordingly. Asylum is neither of these. The applicant picks the country as under free movement, but entry is decided by the destination as under restricted migration, and that decision is a legal test of the need for protection rather than the fiscal selection the study describes. The study therefore does not measure asylum, and neither of its results carries over to it as such.
Politico analysis of EU efforts to stop so-called asylum shopping — the phenomenon in which asylum seekers move from country to country to apply where benefits are highest or approval odds are greatest. The Dublin Regulation requires claims to be processed in the first country of arrival, which already burdens Mediterranean border states by virtue of their geography. In practice, however, the system does not work: applicants continue their journeys and lodge applications in multiple countries, and transfers back to the country of first entry often fail. This strains the system as a whole and undermines mutual trust among member states.
A peer-reviewed natural experiment on whether the level of benefits affects migration. In 2002 Denmark cut benefits for arrivals from outside the EU by about half, reversed the cut, and later reintroduced it. That is three turns of direction, and they can be compared with each other. According to the authors the cut reduced net immigration by about 5,000 people a year and the reversal restored the effect almost exactly. Asylum applications fell from about 10,000 to under 5,000 within a few years and rose again after the reversal. The elasticity of migration with respect to benefits was 1.3. This is the strongest single piece of evidence for benefits as a pull factor. Its size is disputed nonetheless: in replication checks the elasticity comes out a fraction of the original.
A review of the research on how asylum seekers choose a destination country. There is a positive correlation between the generosity of welfare and the number of applications, but it is weak. The link between the networks formed by earlier applicants and application numbers is about five times the effect of social spending, and the effect of deterrents such as bans on working is about a quarter of the effect of networks. According to the review, applicants typically do not even know the destination country’s rules that apply to them, and follow the footsteps of family and acquaintances instead.
Legal analysis of the EU Dublin system. Secondary movements — applicants relocating from the first country of arrival to other member states — are a direct consequence of the Dublin Regulation's failure to build genuine burden-sharing solidarity. Countries with better benefits or higher recognition rates attract more applicants. The problem is structural and cannot be solved by enforcement alone.
What happens to those returned — the monitoring gap
The fate of those refused is not followed up, so claims of danger and claims of safety alike rest on individual cases. Between 2015 and 2017 the decision line shifted sharply: positive decisions for Iraqi applicants fell from 85% to 33%. The best-known individual case led to Finland being condemned by the European Court of Human Rights, but the National Bureau of Investigation later came to suspect the death had been fabricated, and it was never confirmed.
Yhdenvertaisuusvaltuutettu, Turun yliopisto ja Åbo Akademi2018
Summary
A pilot study comparing the Immigration Service's international-protection decisions from 2015 and 2017. Among Iraqi applicants aged 18–34 in the material, 85% received a positive decision in 2015 and 33% in 2017. The change showed most concretely in how the service assessed whether an applicant's stated fear of violence was well founded. The study does not show which year judged the situation correctly: the security situation in Iraq did change in between, and the administrative courts upheld the great majority of negative decisions. The finding still matters for reading recognition rates: the rate is partly a variable of policy and interpretation, not only a property of the applicant pool. The study led the Immigration Service to review the quality of its own decision-making.
The European Court of Human Rights found that Finland had breached the Convention in a case where an Iraqi man who received a negative decision in 2016 returned to Baghdad and was reported to have been killed weeks later.
The National Bureau of Investigation opened a criminal inquiry on suspicion that the man reported killed in the ECtHR case was still alive. According to the Foreign Ministry the death was never confirmed and no compensation was paid. The case shows how difficult it is to verify what happens to those who are returned.
The UN Human Rights Office calls on states to halt involuntary returns to Afghanistan. It cites the UN’s No Safe Haven report, which documents individual cases, gathered in 49 interviews, of returnees facing arbitrary arrest, detention or torture. The report covers mainly people returned from Iran and Pakistan rather than from Europe, so it is not an observation about what happens to people returned from here. It is so far the only call by an international body to suspend returns that this site carries.
Establishing ageSafety
Key figures
In 2025, 58.1 percent of unaccompanied minor asylum applicants in Finland and 73.1 percent in the EU27 declared an age of 16 or 17.
Age decides which law applies to a person. Someone under 18 falls under child protection and compulsory education, is given a representative, is treated more leniently in criminal law and is in a stronger position for family reunification. The authorities, however, do not observe age. They record it. Government bill 240/2009 listed the reasons age can remain unclear: the country of origin may have no civil register, documents may be lost or out of reach because of persecution, and not every applicant knows their own age. The same bill noted that some children state an age above their real one and some applicants state an age below it, and that smugglers advise in both directions.
This section is not about individual cases. Three series can be measured: the age unaccompanied minors state, the results of forensic age examinations, and the age structure of the route into school. None of them says how many people state a false age. Together they say how narrow the age range in question is, and how blunt the instruments for answering it are.
About two thirds of unaccompanied minor asylum seekers state that they are 16 or 17. Those are the last two years of childhood before the threshold of adulthood. The share under 14 stays small even though that band covers fourteen years of age. The same share holds across the EU and Finland does not stand apart from it, so this is not one country's phenomenon.
Share aged 16 to 17 among unaccompanied minor applicants · Finland and EU27, 2012–2025
% of unaccompanied minors
FinlandEU27
Details
Eurostat rounds to the nearest five, so the shares in small years move about. In 2025 the whole group was about 155 people in Finland and 17,935 in the EU.
The age bands are of different widths. Under 14 covers fourteen years of age, 14 to 15 covers two and 16 to 17 covers two.
The figure is the age stated at registration, not a proven age. Travelling alone is more common for older teenagers than for children, so the concentration would be expected even without any misstatement.
An examination is ordered only once there is reason to doubt the stated age, so the share does not describe everyone who claimed to be a minor. It describes the group the authorities had already picked out on suspicion. With that caveat, the result has been the same in two countries and over fifteen years: a majority of those examined were assessed to be adults.
Share assessed as adults in forensic age examinations · Finland and Sweden
% of those examined
FinlandSweden
Details
The Finnish 2009 figure is from government bill 240/2009 and covers 89 examinations ordered by four police departments between January and July, of which 58 turned out to be adults. The law on establishing age only entered into force in 2010.
The Finnish 2015 figure is 90 of 149 examined, and the 2016 figure is about 65 per cent of 630 examinations, so it is rounded.
Of Sweden's 9,617 examinations in 2017, 8,007 pointed to an age of at least 18, 1,425 to a possible age under 18 and 119 to possible adulthood, and in 66 cases no assessment could be made.
The method is not precise. The government bill described the reference atlas most commonly used in wrist-bone examination as compiled in 1935 from Americans of European descent, and not for age assessment. The biological range allowed for is two years in the Netherlands, one and a half in Belgium and one in Austria.
Preparatory education is a newly arrived pupil's route into basic school. Most of its pupils are of primary-school age, but every year it also includes pupils who have turned 16. In 2022 there were 520 of them, 5.6 per cent of all 9,307 pupils, and in 2025 there were 291, 4.6 per cent of 6,392. The figure does not measure misstated age. It measures how many begin the basic-school path at the top end of compulsory education.
Pupils in preparatory education for basic education by age · Finland 2022 and 2025
Pupils
20222025
Details
Preparatory education is meant for pupils of compulsory-school and pre-primary age, so a pupil aged 16 or 17 is the rule there rather than the exception.
There is a separate route for adults. Adult basic education had 9,444 students in 2025, more than the number of pupils in preparatory education.
Age is a register entry, that is, the age at which the pupil is recorded in the population information system.
The high pupil numbers in 2022 and 2023 are partly explained by the arrival of children fleeing Ukraine.
The bill that added sections 6 a to 6 c on establishing age to the Aliens Act. An examination requires written consent from the applicant and their representative, and a person who refuses is treated as an adult unless there is an acceptable reason for the refusal. Refusal alone cannot be a ground for rejection. The bill records that only a few dozen examinations were carried out from the mid-1990s to the end of 2008, and that of 89 examinations between January and July 2009, 58 turned out to be adults. Unaccompanied minor applicants numbered 220 in 2005, 108 in 2006, 98 in 2007 and 706 in 2008. Of the 473 who had arrived by mid-November 2009, 417, about 90 per cent, said they were 15 or older. The bill reviews the limits of the methods and the practice of other European countries.
Table 14wv breaks preparatory-education pupils down by single years of age from 2022 onwards. It is the only basic-education statistic with age as a variable of its own: the basic-school pupil tables carry the year group but not the age, so a pupil's age relative to their year group cannot be computed from these statistics.
The statistics published by Sweden's National Board of Forensic Medicine covering every medical age assessment carried out. In 2017 it issued 9,617 statements, of which 8,007 pointed to an age of at least 18. In 2018 there were 1,252 statements of which 896 pointed to adulthood, and in 2019 the figures were 199 and 125. The number of examinations collapsed as requests from the Swedish Migration Agency fell away. It is the most comprehensive single body of evidence on age assessment results in Europe, because Sweden carried out an order of magnitude more examinations than anywhere else.
Annual statistics on unaccompanied minor applicants in three age bands: under 14, 14 to 15 and 16 to 17. The data covers EU countries from 2012 onwards. It is the only continuous series from which the distribution of stated ages can be read by country and year. The figures are ages stated at registration and are rounded to the nearest five.
Migration and global povertyValues and attitudesEconomy
Key figures
The EU receives about a million asylum applications a year, under 0.15% of the world's 700 million people living in extreme poverty (2024). Nor are the arrivals those people: reaching Europe costs thousands of euros, which someone in extreme poverty cannot pay.
Humanitarian migration does not reach the world's poorest, and structurally cannot: for nearly all arrivals getting to Europe costs thousands of euros, which people living in extreme poverty cannot afford.
The EU receives, in a year, a fraction of the number of people living in extreme poverty. Those who leave are often among the most educated in their countries of origin, so migration drains human capital from precisely the countries that need it most. Remittances home are a larger income flow for source countries than development aid, but the shortage is not primarily one of money. Money does not build what migration takes away: development needs capable people who get things done. Nor are remittances aimed at building anything — they go mostly to the consumption of the households that receive them.
The scale is clearest in births. About 7.6 million children were born in Nigeria in 2024, against about 3.5 million in the EU and 1.3 million in Russia: one African country has more births than the EU and Russia combined (UN World Population Prospects, Eurostat). The EU's entire annual asylum caseload amounts to only a small part of that, so migration cannot level a difference of this magnitude in either direction. How many applications lead to protection is under recognition rates, and what the journey costs under people smuggling.
The third bar is so low it is barely visible. That is the section's argument as a picture: the asylum system is sized to handle a margin, not to solve global poverty.
Scale: poverty, forced displacement and EU asylum applications · 2024
million people
Details
The bars are not subsets of one another. Someone in extreme poverty is usually not displaced, and a displaced person usually does not apply for asylum in the EU.
Extreme poverty is under 2.15 dollars a day at 2017 purchasing power parity.
An asylum application is not the same as a person: one person can apply in several countries.
The scale of asylum migration compared with world poverty
The number of asylum applications received by the EU (~1 million/year) is less than 0.15% of the world's 700 million people living in extreme poverty. Globally there are 123 million forcibly displaced people — European asylum channels reach only a fraction of them.
About 997,000 asylum applications were filed in the EU in 2024 (first-time ~912,000). Those two figures come from the Parliament's page. The comparisons come from elsewhere: about 700 million people live in extreme poverty worldwide (World Bank) and 123 million are forcibly displaced (UNHCR Global Trends 2024). Set against those, the European asylum system reaches only a fraction of global need.
Comprehensive overview of global poverty figures by region. Poverty is highly concentrated: the vast majority of those in extreme poverty live in sub-Saharan Africa and South Asia — regions from which asylum migration to Europe is proportionately low.
Around 700 million people live in extreme poverty worldwide (below $2.15/day in 2017 PPP). The EU receives about one million asylum applicants per year — less than 0.15% of those in extreme poverty. Immigration cannot mathematically solve world poverty.
The EU Asylum Agency's annual report: the EU receives around one million asylum applicants per year — a negligibly small share of the world's 123 million forcibly displaced. The system is designed to handle the margin, not to solve global poverty or displacement.
At the end of 2024 there were 123.2 million forcibly displaced people worldwide (refugees, asylum seekers, internally displaced). The vast majority stay in nearby regions or their own country. Only hundreds of thousands reach Europe each year — the Western asylum system covers barely 1% of global forced displacement.
People smuggling and boat arrivalsSafety
Key figures
Smuggling fees range from under $20 to over
5,000 depending on the route, the mode of travel and the client's background. Payment is usually cash or a hawala-type system (2026).
Some asylum seekers reach Europe from North Africa across the Mediterranean in people smugglers' boats. The journey is not travel people arrange for themselves but a paid service, and its price raises the question of who can afford to leave: for many the journey is an investment in getting to Europe.
Illegal migration into Europe is organised through smuggling routes run by criminal networks. UNHCR tracks maritime arrivals officially. The Mediterranean and the English Channel are key routes. Smuggling networks use violence against both migrants and local residents.
The demand does not come from nowhere. Most asylum applications are rejected, but a rejection rarely means leaving: only a small share of return decisions in the EU is ever enforced, and the Commission itself uses that as the argument for its new Return Regulation. When arrival is in practice a permanent outcome regardless of how the application ends, paying for the journey is worth it. That is exactly what the smugglers sell. Countering smuggling is therefore as much a question of return policy as of border control. There is a route for those in the greatest need that does not run through a smuggler's boat: quota refugees are selected in refugee camps on UNHCR's proposal, and their need for protection is examined before the journey. The route is narrow next to the need. In 2024, 188,759 people were resettled worldwide against 31.0 million refugees, about 0.6 per cent. EU countries took 13,790 of them and Finland 810, while 2,275 first-time asylum applications were made in Finland that year. What happens at the end of the journey is under recognition rates, and what follows a rejection under illegal stay and returns.
Third-country nationals found illegally present in the EU · 2008–2025 (Eurostat migr_eipre)
The bar is the number of third-country nationals found to be illegally present in the EU during the year, according to Eurostat.
The figure counts detections, not the number of people staying in the country. These are findings by the authorities, so it measures enforcement activity as much as the size of the phenomenon. Eurostat counts each person once per year per country, so in the EU total someone found in Greece and later in Germany appears twice — which matters most in transit-heavy years.
People who apply for asylum after being found are included: an asylum application makes the stay lawful while it is processed, and some are granted protection. The 2015 peak is mostly crisis-year arrivals, not 2.1 million people residing illegally. The only asylum-related exclusion Eurostat names is Dublin transfers between EU countries.
The figure covers both those who crossed the border without authorisation and those who arrived legally but stayed after their permit expired. These are different phenomena and the statistic does not separate them.
UNHCR reports that 188,759 people were resettled worldwide in 2024, a year with 31.0 million refugees and 8.4 million asylum seekers, so resettlement reached about 0.6 per cent of the refugees. Eurostat has EU countries resettling 13,790 people in 2024 and 10,220 in 2025. Finland's figures were 810 and 620. Finland received 2,275 first-time asylum applications in 2024 and 1,970 in 2025.
UNODC's Observatory on Smuggling of Migrants estimates the financial value of smuggling by sea on the Central Mediterranean route at US$290–370 million in 2023, and 254,000 people departed the North African coast, 79% more than in 2022, of whom an estimated 175,000–228,000 were smuggled. 151,000 reached Italy or Malta, North African coastguards intercepted 102,000, and 1,800 died or went missing.
The Commission's proposal for a new Return Regulation, 11 March 2025. Its reasoning states that successful returns of people staying illegally currently run at around 20 percent. The proposal would replace the 2008 Return Directive and make one member state's return decision recognised in the others.
According to UNODC's research brief, smuggling fees vary widely, between under US$20 and over
5,000. The price depends on the smuggled person's nationality or ethnicity and economic status, the route, the mode of smuggling (land, sea, air), demand and levels of enforcement, among other factors. Fees are most often paid in cash or through hawala-type systems that are hard to trace.
Parliament and Council reached political agreement on the Return Regulation on 1 June 2026. The Commission notes the return rate rose to 28% in 2025, but says the number of effective returns is still too low.
Illegal stay and returns in FinlandSafety
Key figures
In 2025 Finland ordered 5,070 people to leave the country and recorded 1,915 returns. About two in five were carried out.
The number of people found illegally present has fallen throughout the 2020s: from 1,305 (2018) to 660 (2025), the lowest in the series.
When an asylum seeker receives a negative decision and has no other ground to stay, the right of residence ends and remaining in the country becomes illegal. Not everyone leaves. What follows is a group with no right to be in the country that is here all the same, and there is no count of it: nobody follows what happens to the people whose application was refused.
Staying has also carried a financial incentive. Residence-based social security has required a valid residence permit only since April 2019, when the requirement was written into the act on residence-based social security. Before that the right was decided on permanent residence, and a permit running out did not by itself end the benefits. Urgent social assistance and urgent care remain available to everyone regardless of status.
Three Eurostat series describe illegal stay in Finland, and they measure different things. The number of third-country nationals found to be illegally present has fallen through the 2020s, from 1 305 in 2018 to 660 in 2025, the lowest in the series. The 2015 figure, 14 285, is the crisis year's arrivals rather than a count of people living here without permission.
The second and third series are the decision and its execution. In 2025 Finland ordered 5 070 people to leave and recorded 1 915 returns, about two in five. Across the EU the same year it was 486 705 orders against 156 000 returns, about one in three. The gap was widest when the 2015 applications were decided: 17 975 orders in 2016 against 6 005 returns.
What happens after a rejection is not fully counted in Finland. The Interior Ministry's own review of the asylum process states that returns do not succeed to some countries of origin and that the changes in the law have created a new group of people staying in the country illegally. Assisted voluntary return was used by about 2 100 people in 2016 and about 1 400 in 2017, and no statistics are published on interrupted returns. Acceptance rates and the need for protection.
Found illegally present, orders to leave and returns · Finland 2014–2025
people
Found illegally presentOrdered to leaveReturned following an order
What the figures measure
The number found is what the authorities detected, not the number of people staying in the country, so it measures enforcement effort as well. An asylum application makes the stay lawful while it is processed, so some of those found become applicants and some are granted protection.
A return can follow a decision made in an earlier year, so a single year's share is an approximation rather than one cohort followed through. The 95 per cent of 2014 reflects that lag, not that year's enforcement.
Eurostat rounds these figures to the nearest five, and 2025 is the most recent year published.
An order to leave does not mean a person leaves. The chart sets side by side how many third-country nationals in Finland were ordered to leave and how many actually left following an order.
Orders to leave and enforced returns · Finland 2008–2025
People per year
Ordered to leaveReturned following an order
Details
Eurostat rounds the figures to the nearest five.
An order and a return do not concern the same people in the same year: processing and appeals run past the turn of the year, so the annual ratio is indicative rather than a completion rate.
Returns count both voluntary departure and removal enforced by the authorities.
The 2016 peak follows the decisions on the 2015 backlog of applications.
Eurostat's return statistics for Finland. Orders to leave numbered 3 050 in 2021, 3 200 in 2022, 3 320 in 2023, 3 865 in 2024 and 5 070 in 2025. Returns recorded in the same years were 1 195, 1 170, 1 475, 1 340 and 1 915, a share of 35 to 44 per cent in recent years. Across the EU in 2025 there were 486 705 orders and 156 000 returns, about 32 per cent. A return can follow a decision from an earlier year, so the share is an approximation.
660 third-country nationals were found to be illegally present in Finland in 2025. The figure has fallen every year since 2022, when it was 1 190. In 2015 it was 14 285 and in 2016 only 2 130, which follows the crisis year's arrivals moving into the asylum process. The statistic counts each person once per year per country.
Safety at asylum reception centresSafety
Key figures
43% of reception-centre staff had faced violence or threats at work in the previous 12 months, most often from a resident of the centre (survey 2016).
Asylum seekers were suspects in 2,057 offences between 1 January and 30 September 2016. The commonest were assault (596) and petty theft (330).
The accommodation provided for asylum seekers has raised questions about how safe the centres are, both for the people working in them and for the residents themselves. The question is hard to answer, because up-to-date aggregate statistics are not published openly: what exists is from 2016–2018, the years of the large asylum influx. The same system’s spending is under reception-system costs and the comparison with the native population under criminality.
A total of 2,057 suspicions were recorded. Two thirds of them are assault or petty theft, and the police say the assaults arose mainly from fights between applicants inside the centres.
Asylum seekers suspected of an offence by offence type · 1 Jan to 30 Sep 2016
The figures are suspicions per offence type, not convictions and not individuals. One person can appear under several types.
There is no comparison group in this picture. Tens of thousands of applicants were in the country at the time and the group was overwhelmingly men aged 18 to 34, so comparing with the native population requires standardising for age and sex.
The data covers nine months of 2016. No up-to-date compilation is published openly.
A safety survey of reception-centre staff carried out in May 2016 by the National Council for Crime Prevention and the Finnish Immigration Service: 876 respondents from 86 municipalities, response rate about 30%, so those with experience of the problem may be overrepresented. 43% had experienced violence or threats at work during the previous 12 months, the perpetrator most often a resident of the centre (77%, an outsider 9%, a colleague 4%). The most common acts were throwing objects and breaking things (24%), psychological violence such as verbal threats (22%), physical violence (7%), and threats with a blunt, edged or firearm weapon (3%), which is five times the municipal-sector average (0.6%). Fear of violence was more common (36%) than among employed Finns generally (16%, national crime victim survey 2015). Instructors were most exposed: 32% had had objects thrown at them. Idleness was seen as the main cause of disruptive behaviour, but 40% of respondents judged religion-related conflicts to have a large or very large effect and over 30% considered substance use a major cause. The centres were nonetheless experienced as safe: 96% considered working in the centre safe, only 3% saw bullying between residents as a big problem, and levels of violence matched social- and health-care work generally. Threats also came from outside: 45% of directors reported damage to centre property and 10% reported arson, and a fifth of respondents saw outsiders' harassment of residents as a serious problem.
Poliisin rikosilmoitusjärjestelmä / OMSO 1/2017, liite 22017-01
Summary
A compilation from the police crime-report system, published as an annex to the Ministry of Justice report: asylum seekers were suspects in 2,057 offence records between 1 January and 30 September 2016. The most common offence titles were assault (596), petty theft (330), petty assault (149), unlawful threat (133) and facilitation of illegal entry (95). Sexual-offence titles included sexual harassment (38), sexual abuse of a child (25), rape (23), aggravated rape (10) and attempted rape (9), homicide titles attempted manslaughter (7) and manslaughter (2). For the whole of the preceding year, 2015, 926 asylum seekers were recorded as crime suspects. Serious violence involving knives has also occurred inside the reception centres, and probably only part of the offences committed there comes to police attention. The figures are suspicions per offence title, not convictions and not individual persons. At the same time there were tens of thousands of asylum seekers in Finland: about 32,500 applications were lodged in 2015, and the reception system housed some 30,000 people at its early-2016 peak. That group was overwhelmingly men aged 18–34, the most crime-prone age-and-sex group in every population, so a meaningful comparison with the native population requires age-and-sex standardisation, which narrows the gap substantially but does not remove it.
According to Finnish Immigration Service data, in early 2018 reception centres recorded 10–15 reports per month of violence or threats against staff, and 20–60 reports per month of violence between residents (including self-harm). The figures covered both verbal threats and physical assaults.
Reception-system costsEconomy
Key figures
Reception-system costs track applicant numbers steeply: they peaked at €607 million in 2016, against about €100 million in 2021.
The Finnish Immigration Service bought more than €550 million of reception services without competitive tendering in 2022–2023, as direct awards justified by the emergency.
The two-year contract for a family group home in a villa on Villilänsaari in Tampere is worth about €1.4 million. It has places for seven unaccompanied minors and eight staff (2024).
Keeping an asylum system running is not free. The state pays for applicants’ accommodation and subsistence while their applications are processed, and spending tracks applicant numbers, so it swings sharply from year to year.
In the state's final accounts, purchased reception services and allowances paid to clients peaked in the years following the 2015 spike in applications and bottomed out in 2021. The level then rose again, mostly because of temporary protection for people who fled Ukraine. The scale of reception between 1997 and 2015 is on the timeline.
The cost of the reception system tracks applicant numbers with sharp swings. The annual cost depends heavily on the number of people received: at its highest, costs rose to as much as €607 million in 2016, while in 2021 the cost was about €100 million.
Reception-system expenditure in the state's final accounts · Finland (2014–2024)
The chart covers purchased reception services (moment 26.40.21) and allowances paid to clients (26.40.63). It does NOT include the operating costs of the Immigration Service and state-run reception centres (26.40.01, ~€76–104M/yr in 2022–2024), since that moment also funds the agency's other administration, nor the integration compensations paid to municipalities (covered in the Integration section).
Costs lag applicant numbers: the autumn 2015 asylum wave shows up mainly in the 2016 accounts (€530M + €78M).
The 2022–2024 level (~€340–360M/yr) is driven mostly by temporary protection for those fleeing Ukraine. The caseload peaked at ~44,000 clients (2023).
The 2025 budget (~€239M) and 2026 proposal (~€99M) anticipate a steep decline as reception capacity is wound down — they are budgeted appropriations, not actuals, and are therefore not drawn in the chart.
The unit costs in the introduction (€/person/day) come from the indicator table for moment 26.40.21, a different table from the moment-level spending in the chart. They do not multiply out to the total: the 2024 figures agree (€357.8m / 28,571 clients ≈ €34/day), but the 2023 indicator of €30/day is higher than the moment spending implies, because the indicator covers the reception of asylum seekers rather than every temporary-protection client.
A night of reception accommodation cost EUR 45–61 on average in 2012–2016, but the price ranged by type of accommodation from EUR 18 in private accommodation to EUR 208 in a group home for unaccompanied minors. The report gives no total price tag and does not claim to: it says very little is known about the long-term net effect on public finances because the reason for immigration is not recorded in statistics, and that at the time the fiscal effects of refugee-background immigration had not been studied in Finland at all. On the scale, the ministry says that against an annual budget of over EUR 55 billion a phenomenon confined to hundreds of millions has a marginal effect, and that these arrival numbers are no answer to the sustainability gap even if everyone granted a permit found work. Other figures: processing an asylum application cost EUR 1,584, the calculated compensation paid to a municipality was EUR 6,845 a year for a child under 7 and EUR 2,300 for anyone older, and the employment rate of 20–64-year-olds with a refugee background was 39% in 2014 against 74% for people of Finnish background.
A place in a reception centre came to EUR 43 a day in 2023 and EUR 41 in 2024, about EUR 15,000 a year. An applicant in private accommodation cost EUR 10–12 a day, and an unaccompanied minor in a group home EUR 231. The average total cost per person across the reception system was EUR 30 a day in 2023 and EUR 34 in 2024, excluding VAT, about EUR 11,000 and EUR 12,400 a year. The unit price does not fall as the system shrinks: the 2026 budget puts a reception-centre place at EUR 72 a day as the client count is expected to drop from 28,571 to around 9,000. Standby capacity costs much the same whatever its occupancy.
Ylen MOT-toimituksen selvityksen mukaan Maahanmuuttoviraston kilpailuttamatta ostamien vastaanottopalvelujen menot olivat yli 550 miljoonaa euroa vuosina 2022–2023. Hankintoja tehtiin suorahankintoina poikkeustilanteeseen vedoten Ukrainasta paenneiden vastaanoton kasvaessa.
The two-year contract for a seven-place family group home is worth about €1.4 million, and the wider procurement of four such homes up to about €5.5 million a year excluding VAT. A family group home has operated since June 2024 in a villa of over 300 m² on the shore of Lake Pyhäjärvi in Tampere — the former home of footballer Ari Hjelm, last listed at €1.2 million and sold in early 2024. It has places for seven unaccompanied minors holding a residence permit or temporary protection, and eight social- and health-care staff work there in round-the-clock shifts. The Pirkanmaa wellbeing services county buys the service from the private operator Solum, which rents the property. According to Yle's follow-up, the unit's two-year contract is worth about €1.4 million, and the wider procurement covering four comparable group homes is worth up to about €5.5 million a year (excluding VAT). These costs do not appear in this section's chart: a family group home is not a reception centre but aftercare for minors who already hold a permit, reimbursed to the wellbeing services county by the state separately under the Integration Act, from moment 32.50.30 (state compensation for promoting integration, €349.7 million in the 2026 proposal). The reimbursement follows actual costs, but unaccompanied minors have no line of their own in that moment's breakdown, so no national total for family group homes is published. Nor is the contract price the same as the realised cost per child, since the unit has been running below full occupancy.
Opinion and politics
Immigration divides opinion. Although parties that have broadly supported and pursued immigration have won a substantial majority of votes in parliamentary elections, criticism of immigration is more common than immigration-critical parties themselves. An often-overlooked side of immigration is its political effects: voting behaviour among people of foreign background differs markedly from that of Finnish-background voters. Immigrants particularly favour parties that push for a loose immigration policy and a high level of social security. The political effect is weakened by the fact that voter turnout among people of foreign background is markedly lower than among the native population.
Sections in this group and their figures
Public opinion on immigration60% of Finns feel that public debate does not dare to discuss the problems of immigration (2023).
Political effectThe most popular party among foreign-language speakers, by a wide margin, is the Social Democrats at nearly 50% (2025). A second poll (2026) reached the same result.
Public opinion on immigrationValues and attitudesSocial cohesion
Key figures
60% of Finns feel that public debate does not dare to discuss the problems of immigration (2023).
In 2023, 81% of Finns wanted Finland to actively recruit skilled foreign professionals, and 51% thought Finland takes in too many refugees. Both figures come from the same survey.
Attitudes vary by country of origin: 73% viewed people from EU countries favourably, 60% those from the Middle East and 50% those of African background (2019).
Immigration policy is a political choice. Finnish attitudes toward immigration are divided: a majority wants stricter immigration policy and believes public debate avoids discussing the problems. At the same time, labour-based immigration is broadly supported. How opinion shows at the ballot box is under political effect.
Finland
Three figures from the EVA values and attitudes survey of 2023. They do not point the same way, and that is a result rather than a contradiction: a majority think Finland takes too many refugees and a majority feel the problems cannot be discussed openly, yet four in five support recruiting educated professionals. In the same survey 55 % would make it easier for foreigners to move to Finland, the highest reading in 25 years of the series.
Opinions on immigration · three figures
51%Think Finland takes too many refugees
60%Feel the problems of immigration cannot be discussed openly
81%Want Finland to actively recruit educated professionals
Details
Survey findings on attitudes, not a measure of behaviour or of policy.
Although Finland's Muslim population is growing, Finnish attitudes toward Muslims have remained essentially unchanged — researcher Hanna Salomäki: 'no habituation to Islam has occurred'.
Only about 25% held an at-least-fairly-positive view of Muslim women's hijab. Fewer than one in ten (under 10%) viewed the burqa positively.
Nearly 60% felt that everyone should be free to wear religious symbols.
More than a third (over 33%) viewed mosques fully or fairly negatively — the same share in 2015 and 2022.
Among under-30s, about one in three (33%) viewed mosques positively.
EVA Values and Attitudes Survey — immigration20236 findings
n=2,043 respondents, age 18–79, margin of error 2–3%
A majority of Finns (57%) favour tighter immigration restrictions.
30% consider the current level of immigration appropriate.
Only 8% would like to see more immigration.
Attitudes have hardened notably compared with surveys from the early 2010s.
Political effectValues and attitudes
Key figures
The most popular party among foreign-language speakers, by a wide margin, is the Social Democrats at nearly 50% (2025). A second poll (2026) reached the same result.
Immigration eventually affects election results too, because some of those who move here gain the right to vote. The more arrivals there are, and the more uniformly the group votes, the larger that effect is. Foreign-language voters far more often support parties that advocate a more permissive immigration policy. What the population thinks about immigration is under public opinion.
In a Taloustutkimus survey (2025) their most popular party was the Social Democrats (SDP) at nearly 50%, and Helsingin Sanomat's own Taloustutkimus poll (2026) reached the same result.
Party preference of foreign-language speakers if elections were held now · Finland (Taloustutkimus 2025)
The shares are calculated from the 555 respondents who stated a preference. The sample was 1,232, so a little over half of those asked named no party at all — the shares describe those who expressed a view, not all foreign-language speakers.
The survey was commissioned by Suomen Perusta, a think tank associated with the Finns Party. The margin of error is ±4.5 percentage points.
Political participation nevertheless remains markedly lower among people with a foreign background. The big question for Finland's political future is this group's support potential, should its turnout rise. The two observations have to be read together: there are relatively few voters, but those who do vote, vote in a concentrated way. The electoral effect is therefore clearly smaller at present than the support figure alone suggests, and it grows as both the number of eligible voters and turnout rise.
The turnout gap between Finnish- and foreign-background voters in the 2025 municipal elections was 33.0 percentage points.
Voter turnout in the 2025 municipal elections by background · Finland
A survey carried out by Taloustutkimus for the think tank Suomen Perusta (May–July 2025, sample 1,232, of whom 555 or 45% stated a party preference). Support among foreign-language speakers split as follows: SDP 49.2%, National Coalition 12.3%, Left Alliance 11.6%, Swedish People's Party 7.8%, Greens 6.6%, Centre 5.0%, Finns Party 3.3%, Christian Democrats 2.3%, Movement Now 1.0%. Left-wing and immigration-friendly parties dominate. The publisher, Suomen Perusta, is the Finns Party's think tank, but the survey itself was conducted by Taloustutkimus, an established polling firm.
Helsingin Sanomat (kysely: Taloustutkimus)2026-07-20
Summary
A poll commissioned by Helsingin Sanomat from Taloustutkimus (July 2026): the SDP is the most popular party among foreign-language speakers too. 21% of all foreign-language speakers would vote SDP, rising to 45% among those who state a preference. The Greens were second at 7%. Of respondents, 29% could not name a preference, 17% declined to say, and 10% would not vote. Taloustutkimus CEO Jari Pajunen attributes the support to the SDP's campaign work and the government's immigration policy. The article frames immigrant-majority neighbourhoods as a potential vote reservoir for the SDP. This HS-commissioned poll confirms the same picture as the earlier Suomen Perusta survey — this time ordered by a mainstream outlet.
An HS follow-up to the July 2026 poll in which the SDP was clearly the most popular party among foreign-language voters. Taloustutkimus managing director Jari Pajunen judges that the government's strict immigration line has increased the party's support in this group. Tuomas Finne, executive director of the Helsinki SDP, stresses that this trust cannot be taken for granted. The piece is reported from Itäkeskus, where the party periodically runs a campaign tent. It is an explanatory feature rather than a new poll.
History
Public debate on immigration has produced target figures and promises. Some of them can now be checked: the target year has arrived, or the number demanded has been reached.
This part collects the claims as they were stated and places the measured outcome beside them. The point is not to catch individual speakers out but to look at what followed from the targets. It opens with a map that shows, one year at a time, which countries people moved to Finland from and on what grounds.
Where people have moved to Finland from
In 1990, 1,186 foreign-born people moved to Finland. The largest origins were Sweden with 372 and the former Soviet Union with 341, and the third largest, Germany, sent 36 people. In 2024 the figure was 56,110, forty-seven times as many, led by the former Soviet Union with 8,048, Ukraine with 7,200 and the Philippines with 3,706.
The map shows this one year at a time. The thickness of an arc is the number of people who moved to Finland, the thin arc in the other direction is the people who moved from Finland to that country, and the colour says which grounds were the most common among that country's applicants that year.
The balance of grounds shifts more within a decade than the totals suggest. In 2015, 31,707 people applied for international protection and 5,976 on the basis of work. In 2019 work was the largest ground with 11,813 applications. In 2024 the largest was family with 22,267, and in 2025 work-based applications were down to 10,752 against 22,592 family-based ones.
Details
A line's thickness compares countries within one year. The reference is that year's largest movement, so a thickness cannot be read across two years and the year's own figures sit beside the map instead.
The grounds filter changes the measure. With a ground picked, an arc is the number of applications to the Finnish Immigration Service rather than registered migration from Statistics Finland, the years narrow to 2015–2025, and the outbound arcs go, because nobody files an application in order to leave.
The thickness of an arc and its colour come from different sources. The thickness is registered migration by country of birth from Statistics Finland, the colour is applications by citizenship from the Finnish Immigration Service. Country of birth and citizenship are different classifications, so the two halves of one arc do not describe exactly the same group of people.
Grounds are available from 2015. For earlier years the arcs carry a single colour, because the grounds were not published rather than because they are unknown.
Citizens of EU and EEA countries and of Switzerland need no residence permit, so they generate no applications. Those countries' arcs are coloured for free movement rather than left without a colour.
The international protection figure is applications rather than permits granted, and it covers both asylum applications and temporary protection. Applications always outnumber the people who stay.
The map covers the foreign-born. Finnish-born people returning from abroad do not appear, although in the 1990s they were a large share of all immigration.
The former Soviet Union, the former Yugoslavia, former Serbia and Montenegro and former Czechoslovakia are classes of the classification rather than current states. People born before the break-up are recorded under them, and on the map the arc leaves from the territory that state used to cover.
The map draws each year's 42 largest movements among countries with at least 60 moves. Smaller ones are left undrawn so that the large ones stay legible.
The thickness of the map's arcs comes from this table: immigration and emigration by country of birth, whole country, from 1990 onwards. It covers every registered move, including those by EU citizens, who do not appear in residence permit statistics.
The map's colours come from this source: first residence permit applications on work, family and study grounds, plus asylum and temporary protection applications, by citizenship from 2015 onwards. The figures are applications lodged.
Year by year
Below is the history of immigration to Finland one year at a time, from the year Finland acceded to the Refugee Convention. Each year carries its own figures and what was said about the arrivals at the time. The quotations are verbatim and each has its source beside it.
The emphasis is on speech. How presidents, ministers, party papers and letter-writers talked about arrivals has varied more than the numbers have, and often in the opposite direction. What is known about the same subjects today is collected in the knowledge base.
Filter by type
Filter by stance
1968
Law
Finland acceded to the Geneva Refugee Convention. The implementing decrees were issued on 5 December 1968.
A military junta seized power in Chile on 11 September 1973 and the first asylum seeker soon arrived at Finland's legation in Santiago. Finland took a group of one hundred.
By the end of 1975 Finland had received 151 refugees, of whom 32 to 42 had already moved on. The Chilean group stood at 118 in March, 92 of them Chilean citizens.
On 3 September 1976 the Ministry of the Interior appointed a committee to prepare a reform of aliens legislation. Finland had no Aliens Act, only the decree of 1958.
It would be more sensible to consider the forms and extent of financial aid than to daydream about bringing boat people to Finland. The number of refugees Finland could place would in any case be so small that it would make no difference to the refugee problem.
The governing party's paper rejected taking boat people. Five days later, after the government decided to take a hundred, the same paper wrote: "Now that this road has been taken..."
The cabinet's foreign affairs committee refused the UN's request to take boat people on 21 June 1979 and decided on 4 July to take a hundred. The first arrived in the autumn.
Finland's first Aliens Act (400/1983) was passed on 26 April 1983 and entered into force on 1 March 1984. Its decree was late, so March 1984 ran without one.
At the end of the year the cabinet's foreign affairs committee decided in principle to take about a hundred new refugees a year. This is the origin of the refugee quota. That same year the government had admitted nine refugees and ten family members.
In March 1986 Ilta-Sanomat measured support for relaxing immigration policy. It stood at 15 per cent of citizens. The Immigration Service's own history uses the figure to show that four years earlier, in the 1982 uproar over the turning back of a Salvadoran priest, public opinion had been measured "only in the corridors of parliament and among journalists". Between what the politicians said and what the public thought there was a considerable gap.
There are 12 to 15 million refugees in the world. Let us use the larger figure. The Finnish people are one per mille of the world's population. Our responsibility would then be one per mille, that is 15,000 refugees. [...] Those would be the limits, 15,000 to 75,000, depending on the point of comparison.
An interview in Helsingin Sanomat's monthly supplement. According to the book, Sorsa meant directing development aid to UNHCR for 15,000 refugees, but the statement was at once read as a demand for 15,000 quota refugees a year.
It is quite possible that our population will not develop the way many consider desirable. We once had a debate that labour should be brought into this country, and it met a lot of resistance then. It is quite possible that this is the very same phenomenon under a slightly different name. And it is quite possible that it would be to our advantage to take in people of an advantageous age.
An A-studio interview. According to the book it irritated the Swedish People's Party's papers and MPs, who held that fitness for work should not be a criterion in selecting refugees.
The first wave of Magdalena Jaakkola's series of attitude surveys. Of the respondents 41 per cent would have admitted more refugees than Finland was taking. The series put the same questions five times up to 2007, and this was its most receptive reading on refugees.
Magdalena Jaakkola's partial study came out in early October 1987, from interviews made between May and July. It overturned myths: Finns related to foreigners "considerably more positively than has been presented in the media". The most hospitable were highly educated Swedish-speaking young women, and sympathy depended on where the arrival came from. The most interesting finding was a collective misperception: most Finns believed others to be more prejudiced than they actually were.
Helsingin Sanomat counted the letters its own readers sent in the summer of 1987. Half of the writers opposed refugees coming to Finland, a smaller group supported them and some took no clear view. The writer of the summary could find no arguments among the opposing majority beyond unemployment, the housing shortage and safety. In the autumn twice as many letters arrived, of which the paper published a third.
The starting point has been that the Ingrians are Finns. Their forefathers were moved to that region under Swedish rule and they have been through a great deal there. It can be compared to return migration. No decisions have been made in this sense regarding other Finnic peoples.
An interview with the foreign correspondents' association. According to the book it took the Aliens Office by surprise and was read as an order to proceed. The return migration of Ingrian Finns began here.
Somali asylum seekers began arriving via the Soviet Union. Almost 200 came through Vainikkala on 26–28 November 1990, and 230 by train on 30 December. From the start of the year to November the arrivals numbered about 1,500.
The Somalis should be stopped at the border. Take down their names and promise to consider. State that the Soviet Union is a safe country. If someone disagrees, then we argue. If these are not stopped, how would any others be.
From Koivisto's own memoirs. On 30 November 1990 the Holkeri government prepared a decision under which arrivals via the Soviet Union could have been turned back at the border. Prime Minister Holkeri opposed it and the decision was never used. A form often repeated in public, in which Koivisto is said to have spoken of shooting the arrivals, appears in none of these sources.
The height of it was when Somalis arrived from Leningrad on a Finnish bus and at the border it was merely noted: right, refugees, fine. Off to Lahti. Not even the names were taken down.
Two expressions from the speech. According to the book Ilaskivi meant the Somalis and the Finns who granted residence permits and social security too readily. The speech won him popularity that did not carry the 1994 presidential election.
EVA's autumn 1990 measurement (n = 2,426) is the oldest run of immigration questions it has. Two figures give the setting of the time. The claim that it is "better to pay Finns to have children than to increase refugee immigration" was accepted by 69 per cent and rejected by 17. The claim that "the growing labour shortage can only be solved with foreign labour" was accepted by 14 per cent and rejected by 68.
199113,118 movers · largest origin Former Soviet Union 5,833
It wears on me that there is some elite telling the people what to think..., that one should somehow grovel before other people.
An Ilta-Sanomat interview. According to the book the press pressed Koivisto six times in January 1991 to condemn popular hostility to foreigners, and he repeated that he had seen no evidence of such an attitude.
There is a great deal of sanctimony here that the border must not be closed. Anyone who sets foot across the border and says I request asylum should be taken in, in this world situation. [...] And what if large numbers really do start moving from there? It would be far better to support their efforts rather than hinder them, so that no new migration of peoples takes place.
The reformed Aliens Act (378/1991) entered into force on 1 March 1991. The Ombudsman for Aliens and the Asylum Board began the same day. Reception centres were put on a statutory footing on 31 December 1991.
199210,182 movers · largest origin Former Soviet Union 4,053
Event
Refugees from the wars of Yugoslavia's break-up began arriving at the start of the year, mostly Kosovo Albanians. In June the Aliens Centre took part for the first time in selecting quota refugees, in Turkey, from where Iraqi and Iranian Kurds came. The last Vietnamese boat refugees were selected in August.
The image of a closed society and an isolated people is largely a myth of the post-war era.
The article was written in the middle of the first Somali arrivals and the recession. According to it Finland probably had over 40,000 foreigners in 1914 and 24,451 at the end of 1920, and before the Second World War more foreigners lived in Finland than in Sweden. At the end of 1991 there were 37,579 foreign citizens.
199310,768 movers · largest origin Former Soviet Union 2,613
Law
An act (14/1993) provided for exceptional handling of certain Yugoslav citizens' asylum applications. On humanitarian grounds Finland took patients from Sarajevo's hospital with their companions, 111 at once and 140 more in 1994.
The second wave of the series, at the bottom of the recession. Only 20 per cent would now have admitted more refugees, and six in ten would have admitted fewer foreign job-seekers than Finland was taking. The six-year fall on the refugee question was 21 percentage points.
EVA's spring 1993 measurement (n = 2,367) asked whether respondents believed racism and xenophobia would rise in Finland over the following ten years. It was thought likely by 65 per cent. In the same measurement 61 per cent held that capable Finns moving abroad is a bigger problem for Finland than foreigners arriving.
19947,747 movers · largest origin Former Soviet Union 2,149
Figure
There were 839 asylum seekers, against 3,634 two years earlier. The government later noted: "In 1994 to 1997 fewer than 1,000 asylum seekers a year arrived." There were 3,538 Somali citizens living in Finland at the end of the year, against 1,882 in 1992.
The number of Somalis grew throughout the decade even though few asylum seekers arrived. The growth came from family reunification and births, not from new applicants.
The refugee quota had been 500 since 1989. For 1995 and 1996 an additional quota of 500 refugees from the former Yugoslavia was confirmed, so the quota was 1,000 in those years. In practice 221 quota refugees were received in 1995 and 817 in 1996.
The quota went unfilled almost every year. In 1997 it returned to 500.
The Joensuu skinheads came to national attention when Darryl Parker, the American signing of the Kataja basketball team, left the city after skinheads attacked him. In November 15 skinheads were charged with assaulting and threatening Somalis.
Yle wrote in 2023 that the skinhead problem only reached the public then, although the violence had gone on for years.
19967,909 movers · largest origin Former Soviet Union 2,132
Figure
There were 711 asylum seekers, the fewest of the decade. According to the Ministry of Labour, immigrants' unemployment rate at the end of the year averaged 47 per cent. For Bosnians it was 92, for Iraqis 89, for Vietnamese 76 and for Somalis 75 per cent.
The figures are from the 1998 integration bill, which used them to justify the law. They are the baseline against which later employment figures can be read.
19978,444 movers · largest origin Former Soviet Union 2,402
Figure
In 1997 the Directorate of Immigration decided 742 asylum cases. Four people were granted asylum, 16 a residence permit on grounds of need for protection and 231 on humanitarian grounds. In 1995 to 1998 asylum in the strict sense was granted to 4, 11, 4 and 7 applicants a year.
The figures are from the table in the government's 2002 Aliens Act bill. They show that in the late 1990s people stayed almost entirely on humanitarian permits, not on refugee status.
On 16 October 1997 the government adopted its immigration and refugee policy programme, the first of its kind. According to it "quickly adopting the language and the rules of society is a precondition for succeeding in Finland". The programme proposed raising the refugee quota by one hundred a year up to a quota of one thousand.
The quota was still 500 in 1997. In 2019 Rinne's government promised to raise it to at least 850, and in 2023 it was 1,050. In June of that year the next government cut it to 500.
19988,568 movers · largest origin Former Soviet Union 2,550
Figure
The government submitted the integration bill. According to it "at the end of 1997 about 80,600 people, about 1.5 per cent of Finland's population, were foreign citizens. The corresponding figure at the start of 1990 was about 0.4 per cent." The refugee population was about 15,500. Those who had come through the work permit procedure were estimated at under 10,000.
The 1998 budget had provided for 600 quota refugees and 500 family members. In the event 308 quota refugees were received.
The same bill admitted the problems of reception: "The average length of accommodation is over a year." About one asylum seeker in ten was detained, and detention took place in facilities intended for criminal suspects, "although under international agreements this should not be done".
In December 1997 there were 11 reception centres with 1,250 places. In 2015 there were eventually about 140 centres. What reception costs today is collected under the costs of reception.
19998,167 movers · largest origin Former Soviet Union 2,057
Law
The Act on the Integration of Immigrants and Reception of Asylum Seekers (493/1999) entered into force on 1 May 1999. Integration had a law of its own for the first time.
The act tied support to obligations. For the first three years labour market support was paid only as integration support, and the basic amount of income support could be cut by 20 per cent if an immigrant refused an integration plan, and by 40 per cent for repeated refusal. That structure is still the basis of integration support. What integration costs and how far it reaches is under the cost of integration.
On 23 April 1999 Finland announced it would receive a thousand evacuees from Kosovo within two weeks. Five hundred came in the first week, 164 of them to the Joutseno reception centre. According to the Ministry of Labour some of the arrivals were in poor health.
The UN refugee agency put the reception request to Finland the same day. The same spring also brought Finland more than a thousand asylum applications from Slovak Roma.
There were 4,410 Somali citizens living in Finland at the end of the year, almost a thousand fewer than a year earlier. Somali speakers, however, numbered 6,251, more than ever. The gap is explained by 1,208 Somalis receiving Finnish citizenship this year and 476 the year before, as the first arrivals of the early 1990s met the residence requirement.
From here on the citizenship statistics show a different picture from the language statistics. Every later claim about the number of Somalis depends on which figure is used.
In July 1999 Finland imposed a visa requirement on Slovak citizens after about 1,100 asylum applications from Roma from that country. Yle reported on 8 July 1999 that dozens were already returning home.
There were 1,516 Slovak applicants in 1999. In 2001 there were 88. The visa requirement lapsed when Slovakia joined the EU on 1 May 2004.
EVA's long population question begins in spring 1999, and it begins at the floor of its own series. The claim that an ageing and shrinking population requires making it easier for foreigners to move here was accepted by 18 per cent and rejected by 61. It is from this level that the 29-point shift EVA referred to in 2021 is measured.
20009,466 movers · largest origin Former Soviet Union 2,270
Law
An accelerated asylum procedure was enacted in the summer of 2000. An application from a safe country of origin had to be decided within seven days. The bill recorded that 3,106 asylum applications were made in 1999, of which 1,516 came from Slovakia, and that 29 people were granted asylum.
In 2001 there were 88 applicants from Slovakia and three from Poland. The 2003 Aliens Act bill concluded that the law's aim had been well achieved.
Our historical ethnic minorities and immigrants bring Finland language skills, cultural competence and other resources that we have so far learned to use for the common good only on a very small scale. Reporting that makes it possible to condemn a whole ethnic group, or to brand all immigrants as criminals and benefit abusers, must always be avoided.
A speech at the European Cultural Foundation seminar on immigrants between two cultures, a year after the Slovak Roma asylum row. In the same speech the president noted that the foreign-citizen share of 1.8 per cent was the lowest in Western Europe.
200111,200 movers · largest origin Former Soviet Union 2,439
Figure
Statistics Finland's population projection of 1 October 2001 stated: "Annual net migration has been assumed at 5,000 people." It projected the population to grow until 2023 and to decline after that.
Realized net migration in 2001 was 5,802 and exceeded the assumption every year after that. From 2007 on it was more than double the assumption. The assumption applied to the whole projection period, not to a single year.
The transitional periods must be flexible, and their possible shortening can be considered on a tripartite basis in due course.
A transitional period meant the right to keep workers from the new EU countries out of the labour market for up to seven years after accession. Finland applied one when Estonia and Poland joined in 2004 and dropped it on 1 May 2006. SAK backed the restriction but warned that it is circumvented by sending workers as service providers. That was exactly the reason the Minister of Labour Filatov gave five years later for abandoning it.
A society growing multicultural is a richness and an opportunity. It requires tolerance and openness to interaction. New cultures need not take anything away from Finnishness, they can widen the room to live and act.
A speech at the Academy of Finland seminar on Finland in a multicultural world, two weeks after the 11 September attacks. The president observed that although the shrinking labour force had put the need for immigrants on the agenda, they were still felt to be a threat.
200210,224 movers · largest origin Former Soviet Union 2,055
Law
On 1 November 2002 the government submitted a bill for a new Nationality Act. Under it "a Finnish citizen would retain their citizenship even if they acquired the citizenship of a foreign state by application or in a comparable way". At the same time naturalisation would as a rule require six years of residence instead of the previous five.
The bill noted that it had earlier been thought undesirable for Finland to acquire minorities holding the citizenship of two states. The act took effect on 1 June 2003. There were 18,041 dual citizens at the end of 2002 and 197,944 at the end of 2025. The residence period was cut to five years in 2011 and raised to eight in 2024.
On 20 December 2002 Paavo Lipponen's second government submitted a bill for a complete reform of the Aliens Act. Its justification stated: "The number of foreigners residing in Finland has risen in ten years from about 25,000 at the start of the 1990s to over 100,000." Long processing times were described as a problem that showed up in complaints to the Parliamentary Ombudsman.
The bill lapsed when parliament ended its work before the March 2003 election. The next government resubmitted it in June 2003, and the act took effect on 1 May 2004. Parliament had called for the complete reform as early as 1998.
Interior Minister Ville Itälä is pressing for an official immigration policy to be created for Finland. According to Itälä, who spoke at the Council of Europe ministerial meeting on Monday, foreign labour will be needed in the care sector within the next few years.
Yle's report from the Council of Europe migration ministers' meeting in Helsinki. In the same report Johannes Koroma of the employers' confederation TT said it was high time to decide on Finland's immigration policy. The new Aliens Act was being drafted.
200310,069 movers · largest origin Former Soviet Union 1,655
Law
The programme of Matti Vanhanen's first government of 24 June 2003 promised: "Where the dependency ratio so requires, the government will promote a work-based immigration policy." It also promised a flexible transitional period at EU enlargement and a separate immigration policy programme.
All three were delivered on schedule: the Aliens Act 301/2004, the transitional act 309/2004 and the immigration policy programme of 19 October 2006. How the ground for entry shows in employment two decades later is under grounds for entry.
Statistics Finland reported on the Ministry of Labour's Työvoima 2020 forecast in September 2003: "When the labour outflow exceeds the inflow to the labour market by a good 10,000 people a year at the end of this decade, the annual difference rises to nearly 20,000 in the next decade."
The same article warned: "There is a risk that we will have a large number of unemployed people and a large labour shortage at the same time." The working-age population began to fall around 2010, but after the 2009 recession unemployment did not fall durably and the labour shortage was widely discussed only in the 2020s.
Finland cannot have an immigrant and migration policy that departs from our population policy. These must be woven into our general population policy.
The party won three seats. Its leading vote-getter Tony Halme took 16,390 votes in Helsinki, five times what the party chair Timo Soini had guessed before the election.
Immigrants are a richness for Finland. Finland needs a clear immigration policy that separates refuge from work-based immigration.
The programme for the March 2003 parliamentary election. It called for work-based immigration to be made easier. The government's bill for a new Aliens Act was tabled in June.
200412,381 movers · largest origin Former Soviet Union 1,945
Law
The new Aliens Act (301/2004) entered into force on 1 May 2004, the day Estonia, Latvia, Lithuania and Poland joined the EU. Finland kept transitional restrictions on workers from those countries until May 2006.
Migration forecasts for the enlargement were described the same year: "Estonia would then send 15,000 people a year to Finland and the other current candidate countries 1,000 a year in total." Lower estimates put migration from Estonia at 4,000 to 5,000 a year.
There were 14,000 Estonian citizens living in Finland at the end of 2004 and 25,510 at the end of 2009. The increase was 2,000 to 3,000 a year, the level of the lower estimates.
The Family Federation of Finland's population policy programme set an annual net migration target of 7,000 people. It proposed that Finland move to active immigration.
The target was exceeded every year from 2005 on. The figures put forward ten years later were five times as large.
In Soini's view the act must include a real fast-track deportation procedure for manifestly unfounded asylum applications, as is already in use in many other EU countries.
Parliament was finalising the new Aliens Act and had just rejected faster fast-track deportations. In the same report Suvi-Anne Siimes of the Left Alliance demanded that applications continue to be handled individually.
Foreigners working in Finland are not a cost to society, nor do they create any particular competitive threat to Finns
A speech at the industry days in Raisio two days before EU enlargement and the entry into force of the new Aliens Act. Employers opposed the two-year transition period imposed on workers from the new member states.
200513,546 movers · largest origin Former Soviet Union 2,087
Whether it is because of a generous reception system with high daily allowances or something else, asylum shopping must be curbed by joint effort in the EU.
Rajamäki said about 70 per cent of applicants in July 2005 were unfounded. There were 3,574 asylum seekers in 2005 and only 1,505 in 2007. The cash income support was replaced by a reception allowance only in 2011.
I see no obstacle to increasing the number received. Most immigration, though, takes place on other grounds, and it must be looked at within the common rules of the EU countries.
An interview in the centenary issue of the church weekly Kotimaa, asked whether Finland could take more refugees. Given weeks after the French suburb riots.
200614,586 movers · largest origin Former Soviet Union 2,165
Finland did not at the time apply a transitional period to services but does apply one to the movement of labour. This has led to work being disguised as services.
The labour market organisations and the state agreed the same day not to extend the transitional period for workers from the new EU countries. The restrictions ended on 1 May 2006. There were 17,599 Estonian citizens in Finland at the end of 2006 and 25,510 three years later.
On 19 October 2006 the government adopted its first immigration policy programme as a resolution. The programme set out that Finland should make its attractions better known, simplify the residence permit system for workers and increase the number of foreign students.
The programme was written in anticipation of a labour shortage. Three years later its successor was found partly premature because of the recession, and the residence permit reform was postponed.
Researchers at the Migration Institute pointed out that people coming temporarily for work do not enter the migration statistics at all. As an example they noted that 12,000 Estonian construction workers worked in Finland in 2006.
There were 17,599 Estonian citizens resident in Finland at the end of the same year. The gap shows how much labour mobility falls outside the statistics.
The programme of Matti Vanhanen's second government of 19 April 2007 promised two things: "The Directorate of Immigration will be renamed the Finnish Immigration Service" and "Labour market testing will be phased out."
The agency was renamed on 1 January 2008. Labour market testing is still in force in 2026, so that promise has been open for almost two decades.
The Ministry of Labour's Työvoima 2025 report used an annual net migration of 7,500 in its calculations. The Ministry of Employment and the Economy's own report noted a year later that net migration in 2007 was almost 13,000.
The realized figure was 13,586. The report concluded that immigration can offset the fall in the labour force only if both net migration grows substantially and immigrants' labour force participation rises.
Finland is slowly beginning to understand that immigration can be a good thing for Finnish society.
The programme's opening sentence in the year of the labour-shortage debate. That spring the country's first migration minister, Astrid Thors, told Yle that Finland needed at least 10,000 new workers a year from abroad.
In immigration policy the Finns Party Youth support the principle "in the country, the country's way". We see the cultivation of multiculturalism at the expense of our own Finnish and Western culture as short-sighted.
The party paper's election issue. The author stood in the election and entered parliament in 2011. The party won 4.1 per cent and five seats in March 2007.
The last wave of Jaakkola's series, and its permissive extreme on labour migration. Only one in four would now have admitted fewer foreign job-seekers, and more people would have admitted more than fewer. On refugees the change was smaller: 27 per cent would have admitted more.
In the October municipal elections the Finns Party increased its support more than any other party. Statistics Finland put its support at 5.4 per cent, 4.5 percentage points more than in the previous municipal elections, on 137,758 votes.
Three years later the party took 19.1 per cent in the parliamentary election. By then immigration had become an election theme for the first time.
The word maahanmuuttokriittinen, immigration-critical, became established in public language around 2008. It was introduced by the people it describes, and the media adopted it after the 2008 municipal elections.
In December 2017 some newsrooms briefly switched to the word anti-immigration and went back to the old one in 2018. Which word is chosen has itself been contested throughout.
Since the purpose of refuge is not economic migration and staying in Finland but returning to the country of origin, residence in a wholly foreign culture is apt to create frustration and various side effects that often cause insecurity for the native population.
Adopted by the party council in Järvenpää for the October 2008 municipal elections, in which the Finns Party rose to 5.4 per cent and Jussi Halla-aho was elected to Helsinki city council.
Integration is a precondition for the growth of immigration. Learning the language has an important role in the integration process, but it cannot be made a binding obligation.
The informal meeting of EU justice and home affairs ministers in Cannes, where France presented its immigration pact. The minister rejected a binding language requirement.
There were 5,988 asylum seekers, the most up to then. Total reception costs were about 84 million euros, against about 33 million the year before.
The number fell the next year and stayed at a little over three thousand until 2014. In 2015 there were 32,476 applicants. How many applications lead to protection is under the need for protection and recognition rates.
On 5 November 2009 the government adopted an action programme on labour immigration. It stated that "the supply of labour will start to contract rapidly early in the next decade, so a labour shortage lies ahead once the current recession has passed". It also noted that the number of foreign workers needed cannot be derived directly from forecasts.
After the recession unemployment did not fall durably, and the wide debate on a labour shortage began only in the 2020s. The programme set no annual target.
On 8 September 2009 the Helsinki District Court fined MP Jussi Halla-aho 30 day-fines for breaching the sanctity of religion. The charge of ethnic agitation was dismissed. The court of appeal upheld the judgment in 2010.
The Supreme Court changed the outcome in 2012 and convicted him of ethnic agitation as well.
Magdalena Jaakkola's series of attitude studies was published as a whole. According to them, in 1993 six out of ten Finns would have admitted fewer foreign job-seekers than at the time. By 2007 only one in four thought so, and more people would have admitted more than fewer.
The series began in 1987 and asked the same questions five times. Opinion on refugees moved differently: those wanting to admit more stood at 41 per cent in 1987, 20 per cent in 1993 and 27 per cent in 2007.
Active support for work-based immigration could be invested in the same way as the promotion of Finnish exports of goods and services.
The report opened by noting that only a year earlier the talk had been of a labour shortage and now it was of unemployment. The proposed state recruitment organisation later took shape as the Talent Boost programme.
A Coalition Party working group proposed in May 2009 that age testing be written into law, that family reunification be limited to the nuclear family and that unfounded applications be reduced.
Age testing entered the Aliens Act on 1 August 2010. The limitation of family reunification to the nuclear family was not carried out then.
All the old Finnish parties and the Greens support, without restraint, a sharp increase in immigration in the name of a so-called labour shortage. The Finns Party supports a nationally responsible immigration policy whose starting point is the principle "in the country, the country's way", a principle that also includes an active will to find work.
The programme for the June 2009 European election, in which Timo Soini won 130,432 votes after leaving Halla-aho off the list.
Bad immigration is not a force of nature. It is a political choice.
The manifesto demanded among other things an end to cash income support in reception centres and a conditional citizenship. A reception allowance replaced income support in 2011 and the residence requirement for citizenship rose to eight years in 2024. The signatories later included the party's chair, a Speaker of Parliament and ministers.
A joint forecast by Helsinki, Espoo and Vantaa in March 2010 estimated that the foreign-language population of the Helsinki region would grow from 100,000 to 250,000, and its share to 15 per cent, by 2030. In Helsinki the share was then 9.6 per cent.
Three years later the same cities raised the 2030 estimate to nearly 300,000. The forecast was already too low before 2015.
An amendment to the Aliens Act took effect on 1 August 2010. Age assessment by forensic examination was written into law, and family reunification for a person granted international protection now required sufficient income if the family had been formed only after arrival in Finland.
More than 2,000 family-tie applications were lodged in 2011 and slightly under 600 a year in 2012 and 2013.
Yle's MOT programme The Price of Immigration reported in December 2010 that about 5,000 new family members were expected in Finland that year, mainly from Somalia and Iraq.
Family-tie applications from people granted protection collapsed after that year's amendment, from over two thousand to under six hundred a year.
But if the pull factor is that we have a fair processing system, then that is a factor I do not want to give up.
An interview on Yle's morning programme after the record asylum year of 2009 and threats against the minister. She called for broad condemnation of hate speech and described Finnish racism as refined.
The programme of Jyrki Katainen's government of 22 June 2011 stated: "Immigrants are a permanent and welcome part of Finnish society." It set the aim of raising immigrants' employment rate and halving their unemployment.
The government's own review in January 2015 reported that foreigners' unemployment rate had varied between 25 and 30 per cent in 2008 to 2013 and stood at 28.6 per cent at the end of 2013. Unemployment was not halved.
In the parliamentary election the Finns Party took 19.1 per cent of the vote and 39 seats. Its support grew by 15.0 percentage points on 560,075 votes.
The party's support was 17.7 per cent in 2015, 17.5 per cent in 2019 and 20.1 per cent in 2023.
In the coming years we will need immigrants as labour, as our population ages and the number entering working life shrinks. Immigrants bring with them varied human, social and cultural capital, language skills included.
The programme, titled Work is the best integrator, was published six weeks before the election and opened under the heading Difference is richness. The party's chair Päivi Räsänen became interior minister after the election.
In the spring of the Finns Party's election win, EVA's series was at its post-financial-crisis floor. Easing immigration for demographic reasons was supported by 25 per cent and rejected by 56, the sharpest rejection in the series since 1999. Ten years later the same figures were 48 and 34.
On 8 June 2012 the Supreme Court also convicted Jussi Halla-aho of ethnic agitation. Its reasoning held that the claims in the text "are apt to arouse intolerance, contempt and possibly even hatred towards the group they target". The fine rose from 30 to 50 day-fines.
The ruling is still the precedent on where the limit of political speech runs. The person convicted became his party's chair in 2017 and Speaker of Parliament in 2023.
Statistics Finland's population projection of 28 September 2012 stated: "Annual net migration has been assumed at 17,000 people in the projection." It put the population in 2060 at about 6.23 million.
That is the highest figure any Statistics Finland projection has given for 2060. Six years later the same office projected the population to start falling as early as 2035.
A Green municipality is open and welcomes people of all kinds. We make use of the opportunities of multiculturalism, support the integration of immigrants and take enthusiasm from subcultures.
The programme for the October 2012 municipal elections, chapter What a Green municipality is like.
Since some of Finland's Somalis who have been granted refugee status or subsidiary protection holiday in Somalia, they can no longer be considered to need the protection Finland offers
A written question demanding that Somalis' protection status be reassessed as Mogadishu calmed. Pekka Haavisto replied the same day that it was still too early to say whether peace had really been reached.
EVA's Values and Attitudes survey found the hardening of immigration attitudes had stopped. Because of an ageing population, 32 per cent would have made immigration easier, and 46 per cent thought immigration enriches the culture. One in two, 51 per cent, considered Finns' reserve towards foreigners wise.
On 13 June 2013 the government adopted the strategy Maahanmuuton tulevaisuus 2020. Its vision opened with the sentence "Immigration is anticipated and managed". The strategy's own chart estimated 345,300 foreign citizens in 2020 and 525,300 in 2030.
There were 278,917 foreign citizens at the end of 2020, about 66,000 fewer than estimated. At the end of 2025 there were 432,654. The year 2015 came two years after the strategy.
Helsinki, Espoo and Vantaa updated their forecast for the foreign-language population in February 2013. In Helsinki the number was projected to grow from 68,000 to 146,000 by 2030 and the share from just under 12 per cent to 21 per cent.
Three years earlier the same cities had estimated 250,000 for the whole region in 2030. Now the estimate was raised to nearly 300,000.
The problem is that these people do not have the kind of capacities for entering the labour market that post-industrial countries like Finland, Sweden and the countries of Western Europe require
Halla-aho and Vesa-Matti Saarakkala opposed taking Syrian refugees to Finland. Petteri Orpo of the National Coalition called the stance complete indifference.
It is in everyone's interest that each person finds their place in Finland. Immigrants' skills must be put to use, their particular needs must be met, and more and more must be invested in integration
A blog post on the Maahanmuutto 2020 strategy the government adopted the same day. The minister called immigration an opportunity Finland cannot leave unused.
In a poll TNS Gallup ran for Helsingin Sanomat (n = 1,002), 52 per cent thought immigration must be restricted for as long as Finland has unemployment. Two years earlier the share was 45. Those fully agreeing that anyone who wants to come should be allowed to were 17 per cent, against 26 in 2011. Taking more immigrants was opposed by 44 per cent, the same share as in 2011.
The Confederation of Finnish Industries reported in January 2014 that 32,000 foreign employees worked in its member companies at the end of 2013, a third more than in 2011. One member company in ten expected to increase their number.
This is the baseline against which the organisation's 2020s targets can be read. In 2023 it proposed 40,000 work-based residence permits a year and in 2026 at least a net 45,000 immigrants coming into work.
Integration into society is a lofty aim into which hundreds of millions of euros are pumped through various channels without even asking whether they do any good. What is being maintained is a large official apparatus and the illusion that immigrants can be integrated into society as future taxpayers and labour
A demand to audit every form of immigration support after Pia Kauma of the National Coalition had raised the Espoo pram controversy.
The central aim of the action programme is to raise the employment rate of immigrants. Finland needs immigrants' labour, new innovations and international networks
The launch of the action programme of the government's Maahanmuuton tulevaisuus 2020 strategy.
The programme of Juha Sipilä's government of 29 May 2015 promised: "An independent study will be prepared on the costs and effects of immigration in our society, enabling a fact-based debate and better integration policy and decision-making." It also promised tuition fees for students from outside the EU and EEA.
The study was made by the Ministry of Finance itself and was completed on 22 December 2017. Tuition fees took effect on 1 August 2017. The programme was written four months before the number of asylum seekers rose tenfold.
On 29 January 2015 the business think tank EVA published the analysis Tulevaisuuden tekijät, which said Finland needs 34,000 net immigrants a year, up from a level of about 18,000, because the working-age population is shrinking.
Total net migration in 2015 was 12,441. The figure of 34,000 was first exceeded in 2022.
About 32,500 asylum seekers arrived in Finland, almost 29,000 more than the year before. The Finnish Immigration Service's figure is 32,476, of whom 20,484 were Iraqi citizens.
Sea rescue operations are a direct transfer of income to people smugglers.
In the same broadcast he estimated that two hundred thousand people a year come to Europe. The interview was three months before the number of asylum seekers in Finland rose tenfold.
I hope this becomes some kind of popular movement that voluntarily bears the burden of housing asylum seekers.
Sipilä announced live that he would give his Kempele home to asylum seekers at the turn of the year. In the same interview: "I appeal: what is needed now is caring."
Well under half of the asylum seekers will be granted asylum.
Orpo said it was impossible to judge whether the figure was 30 or 40 per cent. According to the Ministry of Finance's report, about a fifth of the 2015 applications led to permission to stay at first instance. In October 2015 the ministry had cut its forecast for the year from 50,000 to between 30,000 and 35,000, and the final figure was 32,476.
From the centres they will be returned efficiently to their home countries, either voluntarily or by force. This of course also requires return agreements, which are being negotiated intensively with Iraq among others.
No agreement with Iraq was reached. Iraq's migration ministry announced in January 2017 that it would not agree to any deal on involuntary returns.
When the new people entering the labour market start to grow the economy, we will understand: a new Nokia was given to us as a gift from abroad. It is a lottery win to get more child-bearers to Finland.
The column appeared three weeks after the number of asylum seekers began to increase tenfold. About one in five of the 2015 applicants was allowed to stay.
Finland's short-term objective is to cut off the uncontrolled flow of asylum seekers into our country, bring asylum costs under control and integrate effectively those granted asylum.
The programme's first sentence, three months after the Ykkösaamu interview. The programme removed humanitarian protection from the law, made family reunification harder and shortened appeal periods.
In a poll Taloustutkimus ran for Yle (n = 1,001, October 2015), 51 per cent believed that most of the asylum seekers who stay would remain unemployed and live on benefits, while 31 per cent disagreed. In the same poll 65 per cent held that asylum seekers are not a threat to Finnish culture and identity. As a threat they were seen by 26 per cent.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.
The same scale in every year.
201627,050 movers · largest origin Iraq 3,238
The offer stands. I have asked security experts to assess whether it is safe for a family with children to move there. Because of the great publicity, the assessment is that it is not sensible now.
No asylum seekers ever moved into the Kempele house. Sipilä sold it in March 2016. In 2019 he said the security authorities had judged that the family would face hostility in Finland.
In February 2016 the Finns Party estimated that if the approval rate stayed at the previous year's level, 17,500 Iraqis would receive a residence permit. Päivi Nerg, Permanent Secretary at the Ministry of the Interior, replied that applications would lapse and that a large share of applicants would be returned as Dublin cases.
The Finnish Immigration Service made 28,200 asylum decisions in 2016, of which about 27 per cent were positive. For Iraqis the share was 18 per cent.
On 17 May 2016 the Finnish Immigration Service announced it would stop granting humanitarian protection and issued new country guidelines on Afghanistan, Iraq and Somalia. According to the agency, people could be returned to those countries regardless of which part of the country they came from.
Humanitarian protection had been removed from the Aliens Act by an amendment the same spring. It had been granted 119 times in 2015. The share of positive asylum decisions fell from about 60 per cent in the second half of 2015 to about 30 per cent in the second half of 2016.
The income requirement for family reunification was extended on 1 July 2016 to the family members of people granted international protection. During preparation it was reported that, for example, the parents of a family with two children would have to earn a combined 2,600 euros a month net.
In 2019 Rinne's government promised to stop applying the requirement to minors acting as sponsors. According to the Ministry of Finance's report there were then about a thousand positive family reunification decisions a year. Employment and income for those who came on family grounds are under grounds for entry.
You may come to Finland to work, but not to loaf. In this employment situation Finland needs no extra labour beyond highly educated people in the rarer fields.
A governing-party MP defending the bill that cut refugees' integration allowance by about a tenth, part of the post-2015 package to reduce Finland's pull factors.
The day before yesterday Barcelona, yesterday Turku. We are no longer an island.
Ten people were stabbed in Turku on 18 August 2017 and two of them died. The perpetrator had arrived as an asylum seeker in the spring of 2016 and had a negative decision under appeal. He was sentenced to life on 15 June 2018 in Finland's first conviction for a terrorist offence.
On 22 December 2017 the Ministry of Finance published a report on the costs of asylum seekers. Its conclusion was: "In an annual budget of over 55 billion euros, a phenomenon limited to hundreds of millions still has only a marginal effect."
The report was the independent assessment promised in the government programme, made by the ministry itself. It is still the official benchmark for cost figures put forward in public.
The countries are not willing to send the necessary documents very quickly.
Risikko said there were 11,000 people with negative decisions in reception centres and 1,400 refusing to leave. In 2017 the police escorted 154 people to Iraq. In June 2018 Baghdad police sent three escorted returnees back to Helsinki.
So in 2003 dual citizenship was not in fact ”allowed”, since it had never before been completely forbidden either.
Dual citizenship became a national talking point at the turn of January and February 2017, when the question was whether a citizen of two countries could hold posts in the defence forces and the foreign service. According to Leitzinger the media and many politicians did not understand what the concept means. The debate led to act 564/2019, in force from 1 May 2019, under which citizenship can be lost for treason, high treason or a terrorist offence. There were 197,944 dual citizens at the end of 2025.
Promoting, even attracting, the immigration of skilled people will be very important for Finland in the coming years. The population structure is turning less favourable fast, and without immigration the labour force loses 300,000 people by 2030.
Made during the debate on a residence permit for start-up entrepreneurs, seven weeks after the Turku stabbings. The speaker drew a line between work-based and humanitarian migration.
Most of those on the move were there for economic reasons. Not because they were fleeing war or personal persecution.
Said on Yle's Ykkösaamu about the arrivals of 2015 and 2016. Yle checked the claim: EU countries made about two million asylum decisions in 2016 and 2017, of which more than a million were positive.
The government's immigration policy programme Töihin Suomeen of 4 January 2018 quoted a three-year-old EVA figure: "annual immigration should grow to 34,000 immigrants so that the labour force does not contract in the coming decades". The programme set no annual target of its own.
According to the programme's own figures, 5,436 work-based residence permits were granted in 2015, 5,770 in 2016 and 6,751 in 2017. Net migration in 2018 was 11,965, the lowest of the decade.
Statistics Finland's population projection of 16 November 2018 assumed net migration of 15,000 people a year. It projected the population to grow until 2035 and then decline.
The next year's projection moved the turning point forward to 2031. Six years later the assumption was raised to 40,000 and no decline was projected within the period at all.
Technology Industries of Finland reported on the basis of a member survey that companies in the sector need more than 53,000 new skilled people by 2021. Three in five of those needs concerned graduates.
The figure is part of a series. In 2021 the organisation put forward 130,000 skilled people over ten years and in 2022 raising net migration to 50,000 a year.
In September 2018 the Finnish Immigration Service reported that the cases of a good two thousand of the 2015 asylum applicants were still pending. A few individual cases were still awaiting a first decision.
The 2015 government programme had promised that decisions and returns would be made without delay. Three years later some of the 2015 applications were still unresolved.
But it is also a fact that Finland needs immigration, work-based immigration above all, and of course a very well functioning integration system.
A governing-party MP in the first reading of the new Integration Act. In the same speech he said people seeking only a better standard of living should not be admitted as asylum seekers, and closed by saying racism must be given no room.
201924,426 movers · largest origin Former Soviet Union 1,811
Everyone who comes here must obey Finnish law and respect the bodily integrity of every person. Those who commit crimes will be held responsible under the principles of the rule of law. Serious crimes also weigh against the granting of a permanent residence permit. The asylum system cannot protect criminals.
The statement was given after the Oulu sexual offence investigations. On 15 January 2019 the parliamentary groups agreed to expedite three bills, and they were passed before the April election.
In March 2019 the Finns Party told a question hour that immigration costs as much as 3.2 billion euros a year. Yle's fact check rated the claim false: officials at the Ministry of Finance called the calculation overstated and shaky in its underlying assumptions.
The ministry's own report had arrived at hundreds of millions in 2017. In the April election the Finns Party took 17.5 per cent of the vote.
The programme of Antti Rinne's government of 6 June 2019 promised: "The refugee quota will be raised to at least 850 in 2020." It also set a target of an average one-month processing time for work-based residence permits.
The quota was raised and stood at 1,050 in 2023. In June of that year the next government halved it to 500.
On 14 November 2019 the European Court of Human Rights found that Finland had breached the Convention by returning an Iraqi man who was reported to have been killed after his return. The state was ordered to pay compensation.
The National Bureau of Investigation opened an inquiry in 2020, and according to the Iraqi authorities the death certificate was forged. The court annulled its own judgment entirely on 13 July 2021 and rejected the application as an abuse of the right of petition.
Two surveys measured attitudes in the election year. According to the Foundation for Municipal Development 47 per cent of Finns viewed immigration positively and 41 per cent negatively. In a survey by Yle and Taloustutkimus 41 per cent wanted more labour migrants, 38 per cent found the pace right and 19 per cent would slow it.
The same survey separated the types of immigration sharply. A clear majority supported admitting foreign students and EU workers, while about half would have reduced the number of refugees and asylum seekers.
We will also need work-based immigration here in the future. In practice that means that if the foundation of our welfare is to be kept, there have to be the makings of growth here, and those makings come through labour, which also comes from abroad.
Answering the National Coalition's interpellation on employment policy four months after the government took office.
In EVA's Values and Attitudes survey, 39 per cent held that an ageing population requires making immigration easier and 36 per cent disagreed. At the same time 52 per cent considered immigration at its current level too much. The data was collected both before and after the Oulu sexual-offence suspicions, which according to EVA's research director did not move general attitudes.
202023,035 movers · largest origin Former Soviet Union 1,835
We must help these people and above all the children. This is a humane decision and it shows that Finland respects human rights.
Finland decided to receive at most 175 especially vulnerable asylum seekers from Greece, Cyprus, Malta and Italy. Greece had asked for 150. In the autumn of the same year Finland took 11 minors from the burnt-out camp at Moria.
On 22 January 2020 the Finland Chamber of Commerce proposed that the target for work- and study-based residence permits granted could initially be about 25,000 a year. About 15,000 were granted in 2019.
The figure covers residence permits, not net migration. Total net migration in 2020 was 17,814.
Too large a share of immigration to Finland is harmful immigration. It need not be so. The situation is the result of the wrong kind of policy. Abusing the asylum system is easy and tempting.
Second reading of a bill raising the age limit of integration services for unaccompanied minors to 25, at the end of the first pandemic spring.
In EVA's measurement 40 per cent would make moving to Finland easier because of the ageing population and 35 per cent would not. Attitudes moved towards a more permissive position for the fourth consecutive wave. There were 2,060 respondents.
Chamber of Commerce residence permits, not a net figure25,000
Actual 202017,814
The Chamber of Commerce's 25,000 means work- and study-based residence permits granted, not net migration. It is included because permit targets are quoted alongside net figures in public.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.
The same scale in every year.
202127,554 movers · largest origin Former Soviet Union 2,404
Hybrid influencing as a practice is state-led illegal activity in which the victims are the people used as its instruments.
Said when thousands of migrants had gathered at the Belarus-Poland border. The same definition was Finland's position when the eastern border was closed in November 2023 and when the temporary act on instrumentalised entry was passed in 2024.
On 9 September 2021 the Government published its roadmap for education- and work-based immigration to 2035. The aim was at least to double work-based immigration by 2030 so that the total increase of at least 50,000 work-based immigrants required by the sustainability roadmap is achieved, and after that an increase of at least 10,000 a year. The number of new foreign degree students was to be tripled to 15,000.
The total increase is a cumulative figure for 2021–2030, not an annual level, so it cannot be shown as a bar alongside the other figures. Total net migration in 2021 was 22,905.
At its spending-limits session in April 2021 the government confirmed the aim of doubling work-based immigration by 2030. The goal was 10,000 work- or study-based immigrants a year.
There were 15,012 first work-based residence permit applications in 2021 and 20,960 in 2022. After that the number fell three years running: 16,999, then 15,024, then 11,324 in 2025.
On 19 April 2021 the Finland Chamber of Commerce proposed a target of 30,000 people moving to Finland for work or study as early as the following year, and doubling that by the end of the decade.
About 29,000 first work and study applications were filed in 2022, close to the target. In 2025 the figures were 11,324 and 13,565.
In October 2021 the Finns Party and Movement Now tabled an interpellation on immigration policy and the economy. It was the first interpellation on immigration in at least fifteen years. The government won the confidence vote by 89 to 68.
Between 2007 and 2018 no interpellation was tabled on immigration, even though the subject dominated election debates. The next one came in December 2022 and concerned street gang crime.
Statistics Finland's population projection of 30 September 2021 stated: "The projection assumes that Finland gains 20,000 people through net migration this year and 15,000 a year after that."
In the following years realized net migration was 34,363, then 57,914, then 47,051. Three years later the assumption was raised to 40,000.
This much we do know: Finland unambiguously needs immigration. Humanitarian immigration rests on the fact that people are fleeing danger.
Replying to the same interpellation. The speaker added that a person who meets racism and hatred does not integrate, and that fixing the attitude costs nothing.
Most of the costs of immigration arise from social security spending and services. The direct costs, such as reception or integration, are a fraction of that.
The Finns Party and Movement Now interpellation on immigration policy and the economy. The speaker said the costs were still something of a taboo in political debate.
In EVA's measurement the share rose to 48 per cent, with 34 per cent disagreeing. According to EVA the attitude had moved 29 percentage points towards permitting immigration since 1998, meaning that at the start of the series about a fifth of respondents agreed.
202241,587 movers · largest origin Former Soviet Union 4,295
Event
On 12 December 2022 Technology Industries of Finland set a target of raising Finland's net migration to 50,000 people a year by 2030, citing slowing population growth, ageing and a shrinking working-age population.
Total net migration in 2022 was 34,363. It passed the target in 2023, seven years early.
More than four million people have already fled Ukraine. The current estimate is that as many as 40,000 to 80,000 refugees could come to Finland too.
The appropriation was based on an estimate of 60,000 applications in 2022. In the event 47,302 applications for temporary protection were filed, near the bottom of the range. In the following years the figures were 19,426, then 12,617, then 12,018.
In December 2022 the Finns Party tabled an interpellation on tackling youth and street gang crime by immigration and criminal policy means. The government won the confidence vote by 100 to 70.
It was the third interpellation touching on immigration in little over a year, after none had been tabled for at least fifteen years.
The importance of immigration to a greying Finland grows all the time. Finland needs immigrants so that we can secure the funding of the welfare state and so that employers find enough skilled people for open jobs.
The minister presenting the new Integration Act as the government pushed to raise work-based immigration.
The 2022–2024 peak is explained in large part by temporary protection for people fleeing Ukraine, which none of the people setting targets had in mind.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.
The same scale in every year.
202366,462 movers · largest origin Former Soviet Union 11,904
Discrimination cannot be accepted. Racism is to be given no room in Finland.
The statement followed a summer in which Minister of Economic Affairs Vilhelm Junnila resigned after 11 days on 30 June 2023 and Finance Minister Riikka Purra's writings from 2008 became a public dispute. The leaders of the governing parties signed an anti-racism declaration on 11 July 2023, and the statement was given to parliament at the start of the autumn session.
On 3 February 2023 Etla calculated that permanently stabilising the size of the birth and working-age cohorts would, with the current age and sex structure, require net migration of as much as 44,000 a year, almost three times Statistics Finland's projection assumption of 15,000 at the time.
On 8 March 2023 the Confederation of Finnish Industries (EK) proposed doubling work-based immigration to 40,000 a year by the end of the electoral term and tripling international students to 15,000.
The figure covers work-based residence permits, not net migration, so it is not comparable with the other targets.
Net migration was 57,914, more than any target that had been put forward. The level demanded was reached and exceeded years before the earliest target year.
The programme of Petteri Orpo's government of 20 June 2023 set out: "Finland's refugee quota is 500 people a year." On the residence requirement for citizenship it said: "The period of residence in Finland will be extended to eight years." On work-based permits: "A person must leave Finland if their employment ends and they have not entered a new employment relationship within three months."
The quota was 1,050 in 2023. The eight-year residence requirement took effect on 1 October 2024 and the three-month rule on 11 June 2025. The euro limits for family reunification income had still not been decided by September 2026.
The government closed the entire eastern border from 30 November 2023. Minister of the Interior Mari Rantanen said a full closure was necessary in the name of national security, and Prime Minister Orpo said Russia had caused the situation and could also end it.
More than 1,300 applicants arrived across the eastern border in 2023. The border was still closed in September 2026. First asylum applications across the whole country numbered 2,405 in 2024 and 2,047 in 2025.
In EVA's values and attitudes survey 55 per cent of respondents agreed with the statement that the ageing and threatened decline of the population require making immigration to Finland easier. Thirty per cent disagreed.
It was the first time in 25 years of measurement that a majority agreed with the statement. Agreement stood at 31 per cent in 2016, 39 per cent in 2019 and 48 per cent in 2021. In the same survey 85 per cent regarded asylum seekers and work-based immigration as two entirely different things.
If we think only of the costs that go into income transfers, I would point out that problematic immigration is a very large economic factor.
Debate on the Orpo government's programme a week after it took office. The speaker called the government's change of paradigm in immigration policy a significant economic measure as well.
The other figures from the same EVA measurement: 82 per cent would make it much easier for skilled and educated foreigners to move to Finland, 52 per cent held that too large a share of arrivals are refugees, and 39 per cent judged immigration to do the Finnish economy more harm than good. The same survey therefore backed labour migration and rejected humanitarian migration.
In spring 2023 EVA's oldest immigration question drew as much support as opposition for the first time. The claim that "the growing labour shortage can only be solved with foreign labour" was accepted by 42 per cent and rejected by 41. In 1990 the same figures were 14 and 68, a shift of 28 percentage points over thirty-three years.
EK's 40,000 is work-based residence permits, not net migration. It is included because it is quoted alongside net figures in public.
The 2022–2024 peak is explained in large part by temporary protection for people fleeing Ukraine, which none of the people setting targets had in mind.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.
The same scale in every year.
202456,110 movers · largest origin Former Soviet Union 8,048
A state under the rule of law has the right to defend itself.
Said on the day parliament passed the temporary act on countering instrumentalised entry. The Greens and the Left Alliance voted against, as did six Social Democrats and Eva Biaudet of the Swedish People's Party.
On 8 November 2024 Etla put forward an alternative to Statistics Finland's population projection. Statistics Finland had raised the projection's net migration assumption from 15,000 to 40,000 a year, and Etla argued the long-run level should be set at 24,000.
On 24 October 2024 Statistics Finland published a projection assuming net migration of 52,000 in the current year, 45,000 in 2025 and 40,000 a year after that. It projected the population to grow to 6.5 million by 2070.
The assumption was 2.7 times the one made three years earlier. Within two months the Ministry of Finance was calculating with 25,000, the Bank of Finland with 27,000 and Etla with 24,000. Realized net migration in 2025 was 31,233.
The amendment to the Nationality Act took effect on 1 October 2024. The residence required for naturalisation rose from five to eight years, and only residence on a permit counts. In July parliament had passed, by 167 votes to 31, a temporary act on countering instrumentalised entry.
A third fewer citizenship applications were filed in 2025 than the year before. The temporary act had not been applied once by September 2026, and its validity was extended to the end of 2026. Naturalisation numbers and the make-up of the population are under demographic change.
In its alternative budget in November 2024 the Centre Party proposed halting the deterioration of the dependency ratio by raising net migration to 40,000 a year.
The party's deputy chair Antti Kaikkonen publicly withdrew the target in June 2026, saying there is no particular need for work-based immigration right now.
A Ministry of the Interior memorandum estimated in February 2024 that without restrictions at least 10,000 to 20,000 asylum seekers a year would arrive across the eastern border.
The estimate cannot be tested because the border stayed closed. Across the whole country there were 2,405 first asylum applications in 2024 and 2,047 in 2025.
Finland is a safe country, but Finland is unsafe especially for women and especially for immigrants. And Finland is unsafe for immigrants because racism occurs in Finland.
A debate on equality and non-discrimination after the racially motivated stabbing of a child in Oulu, a year into the racism controversy around the government.
Statistics Finland projection assumption, not a target40,000
Etla proposed projection level24,000
Actual 202447,051
Statistics Finland's 40,000 is a calculation assumption of the population projection, not a target. It is included because Etla's proposal was an alternative to exactly that.
The 2022–2024 peak is explained in large part by temporary protection for people fleeing Ukraine, which none of the people setting targets had in mind.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.
The same scale in every year.
202544,607 movers · largest origin Ukraine 4,673
Finland is not as attractive a country for work-based immigration as it used to be.
In the same article the Director General of the Finnish Immigration Service, Ilkka Haahtela, said the reducing effect of the current government's measures cannot yet be read from the statistics. There were 15,024 first work-based residence permit applications in 2024 and 11,324 in 2025.
Net migration was 31,233, 46 per cent below the 2023 peak.
The 2022–2024 peak is explained in large part by temporary protection for people fleeing Ukraine. The preliminary figure for 2025 was about 35,000, and the figures put forward in spring 2026 were still compared with that.
The government programme's commitments took effect in three stages. The duration of international protection permits was shortened on 2 January 2025, the three-month rule for holders of work-based permits began on 11 June 2025 and new conditions for family reunification on 16 June 2025.
Asylum is now granted for three years instead of four. A person who loses their job has three months to find new work, specialists six. About 2,800 removal decisions were issued to permit holders in 2025, 41 per cent more than the year before.
If you want growth, you had better understand the realities. We need immigration, we do not need the belittling of people who come here from elsewhere.
Debate on the fiscal plan the day after finance minister Purra wrote on X that 38 per cent of mothers giving birth in Vantaa and Kerava were non-Finnish, and called it worrying.
Immigration is also, to the greatest degree, a security question. We have heard that from Sweden too. Immigration, poor integration and crime go together, and that should be recognised and admitted here as well.
The minister presenting the report on internal security and replying to the opposition's claim that Finland needs immigration.
On 23 March 2026 the Confederation of Finnish Industries (EK) published its targets for the 2027–2031 electoral term: the target for immigration leading to work is set at a net 45,000 immigrants a year at least, and EK wants the target written into the next government programme.
Three years earlier EK's target was 40,000 work-based residence permits. The new figure is a net figure, so it is more demanding than the earlier one. What arrivals on different grounds cost the public finances and bring to them is under the economy.
On 23 March 2026 Etla's report 176 estimated that securing a sufficient workforce for public finances would require net migration of about 44,000 a year. Research director Antti Kauhanen called the target very challenging.
Yle first wrote that net migration should be tripled and corrected it the same day: the need would be a good quarter more than the preliminary 2025 figure of about 35,000.
Annual net migration of about 25,000 would keep the number of the elderly relative to the working-age population roughly stable for the next 20 years.
In the same week EK demanded 45,000 and Etla calculated 44,000. The figures answer different questions: Toikka's stabilises the dependency ratio, Etla's the public finances.
On 30 January 2026 the Finnish Immigration Service published the 2025 statistics. There were 11,324 first work-based residence permit applications, a quarter fewer than the year before, and 13,565 study-based ones. Director General Elina Hirvelä said the fall in work-based immigration is primarily due to the weak development of the Finnish economy in recent years.
The targets put forward the same spring were 44,000 and 45,000. The agency's own estimate for 2026 was 11,000 to 13,000 work-based applications.
In March 2026 the Central Organisation of Finnish Trade Unions replied to EK's target of 45,000 that it does not take account of high unemployment. The Coalition Party's platform adopted in June set no figure but promised to process specialists' residence permits in under a week and to abolish labour market testing.
Abolishing labour market testing was already promised in a government programme in 2007. It was still in force in September 2026.
In June 2026 the Centre Party's deputy chair Antti Kaikkonen withdrew his party's target of 40,000 net immigrants. He said there is no particular need for work-based immigration right now, although there are needs in the longer term.
The target had been put forward in the party's alternative budget a year and a half earlier. It is so far the only one of the figures put forward that its proponent has publicly withdrawn.
Our growing immigration has also caused us internal security problems, which show in our prisons as well.
Debate on the prime minister's announcement of government policy for 2026. The minister added that, on the forecasts, more than half of prisoners might be foreigners within a few years.
That is why the situation changes. Immigration must be limited to 10,000 people a year. Other countries are going to do the same. We have to take this decision now.
The first number on the timeline that is a ceiling rather than a floor. All sixteen earlier ones concern how much immigration would be needed at a minimum: nine are targets, three are calculations of the need, one criticises a target and one withdraws one. Harkimo's limit would cover humanitarian immigration too. For scale: Statistics Finland recorded about 51,000 arrivals last year and just under 190,000 in 2023–2025 together, and the Confederation of Finnish Industries wants at least 45,000 work-based net immigrants a year. Harkimo argues from unemployment and artificial intelligence: "We have 400,000 unemployed and AI is taking jobs. We have no work to offer them."
Minister for Foreign Trade and Development Ville Tavio (Finns Party) proposed capping immigration from outside the EU and EEA by law at 0.1 per cent of the population a year, which he put at ”in the order of about 5,600 people a year”. He argued for ”a new social contract that recognises Finland's carrying capacity as limited”.
The second ceiling in four days, and drawn around a different group than Harkimo's. Harkimo's 10,000 covers all immigration, Tavio's limit only arrivals from outside the EU and EEA, so movement by EU citizens would fall outside it. The figure is tied to the population, so the ceiling moves with the head count. For scale: Statistics Finland recorded 51,121 immigrations to Finland in 2025 and the Finnish Immigration Service granted 49,785 first residence permits the same year. A first residence permit goes to people from outside the EU and EEA, which is exactly the group the ceiling would cover, so the ceiling would be about a ninth of last year's level. The two figures still do not measure the same thing, because a permit is a decision rather than a move.
EK's 45,000 is immigration leading to work, in net terms, so only one part of total net migration. Family members, students and people granted protection come on top of it.
Etla calculated its figure from the point of view of public finances and VATT from that of the dependency ratio. A different question yields a different number, though both are net migration.
Net migration is immigration minus emigration across all countries of birth, and it does not distinguish the grounds for entry.