Skip to content
SVERIGEFAKTA

What happened to Sweden?A data-driven storyChapter 4 of 61990–2025

Long-run outcomes — an overview

What were the consequences?

An editorial overview of the long-run outcomes: work, economy, municipalities, schools, crime, segregation and the second generation.

read time
14 min read
charts
7
primary sources
12

After this chapter you will understand

  • Why the same immigration volume can produce very different consequences.
  • What research actually shows about work, finances, schools, crime and segregation.
  • How the second generation has developed — and where the gap persists.
  • What is established, what is uncertain and what is political interpretation.

Why do the consequences look so different?

The same immigration volume can produce very different outcomes. What matters is not the number of people — it is the composition and the structures around them. Six factors explain most of the variance between cohorts, municipalities and neighbourhoods.

  • 01

    Migration category

    Labour vs asylum vs family — very different starting points.

  • 02

    Education level

    Share with tertiary education ranges from 20 % to 86 %.

  • 03

    Labour-market cycle

    Cohorts arriving in a boom establish faster.

  • 04

    Municipality

    Reception volume and local labour market vary sharply.

  • 05

    Neighbourhood

    Segregation and crowding affect school outcomes and safety.

  • 06

    Time in Sweden

    Employment rises; the native gap narrows but rarely closes fully.

This chapter walks through the consequences area by area. Deeper analysis, metrics and time series live on the in-depth pages linked from each section.

How to read this

Every claim is tagged by the type of evidence: observation (official statistics), statistical correlation (controlled or not), research (peer-reviewed or agency reports) and political interpretation (opinion — clearly marked).

Work and self-sufficiency

Employment is the single most important factor behind the whole chain of consequences: tax revenue, welfare dependency, housing, trust and second-generation outcomes. Without work the chain breaks early.

Observation

In 2024, 84.8 % of foreign-born people aged 20–65 with work as ground for settlement were employed, 68.9 % of refugees/persons in need of protection and 62.6 % of family members of refugees, against 84.2 % of native-born. Cross-section regardless of time in Sweden – not an effect of the ground. Source: Statistics Sweden TAB887.

  • Work84.8%
  • Other78.3%
  • Not stated/not applicable69%
  • Refugee/protection68.9%
  • Family of non-refugee68%
  • Family of refugee62.6%
  • Studies47.5%
  • Native-born (reference)84.2%
Share employed aged 20–65 by ground for settlement, 2024 (foreign-born regardless of time in Sweden). Cross-section, not arrival cohorts. Source: Statistics Sweden TAB887.

Public finances

The debate about the fiscal cost of migration is often confused because different studies measure different things. GDP and GDP per capita are not the same. Public revenue and public expenditure are separate sides. Cross-sections and life-cycle models answer different questions. All three matter — but only when the method is disclosed.

Ruist (2015)

Method
Cross-section · all foreign-born
Scope
Average foreign-born, per year
Result
−58,000 kr

Aldén & Hammarstedt (2019)

Method
Lifetime cohort · refugees 1980–2015
Scope
Lifetime, per individual
Result
−1,900,000 kr

ESO 2018:3 (Flood/Ruist)

Method
Life-cycle model · by ground
Scope
Lifetime, per individual, weighted average
Result
−600,000 kr
The methods give different answers. Lifetime models account for pensions, health care and tax progression; cross-sections do not. No study is definitive.
Research

Foreign-born residents as a group contribute less than they cost during their first decades in Sweden. The size of the net cost depends strongly on ground for residence, age at arrival and education. Ruist 2015; Aldén & Hammarstedt 2019; ESO 2018:3.

Statistical correlation

Aggregate GDP has grown with the population, but GDP per capita has stagnated since 2018. The correlation is not purely migration — productivity and the employment gap both matter. Source: Statistics Sweden National Accounts.

The municipalities

National averages hide the fact that municipalities carry most of the practical consequences. The gap between Danderyd and Filipstad is wider than the gap between Sweden and Denmark.

Observation

Municipalities' combined result in 2024 was SEK 15.7 billion, down SEK 10.3 billion on 2023, mainly due to higher operating costs. Source: Statistics Sweden, municipal accounts summary (news release 4 March 2025).

Statistical correlation

The share of welfare payments correlates strongly with the share of foreign-born in a municipality — but is not fully explained by it. Education, labour market and time in Sweden carry most of the effect. Source: Socialstyrelsen; SCB.

School

School results are the clearest early signal on whether integration works. A careful reading is needed: gaps between groups reflect much more than origin — age at arrival, parental education and the school's socioeconomic profile all matter.

OECD 476Native, native background493Native, foreign background460Arrived before age 6448Arrived 6–11405Arrived 12+358

PISA reading literacy 2022 by age at arrival, score (OECD average 476)

  • Native, native background: 493; Native, foreign background: 460; Arrived before age 6: 448; Arrived 6–11: 405; Arrived 12+: 358
PISA 2022, reading literacy, grade 9 students. Source: Swedish National Agency for Education.
Observation

The share of pupils leaving Year 9 without upper-secondary eligibility was 16.3 % in 2024 — on a par with 2018 (16.4 %), the highest share in the series since 2000. The gap between native and foreign background is large but narrows with years spent in Sweden. Source: Skolverket.

Research

School segregation has increased since the 1992 free-school reform. Researchers agree on the increase; they disagree on how much of the outcome gap it explains. Source: IFAU, Skolverket, OECD.

Crime and safety

This section is kept strictly factual. The statistics show clear differences between groups — but they also show that the differences narrow when controlling for age, sex and socioeconomic background, without disappearing entirely. Skipping the controls is misleading; controlling them all away is too.

Foreign-born

Unstandardised
2.51×
Adjusted: five factors
1.76×

Native, two foreign-born parents

Unstandardised
3.21×
Adjusted: five factors
1.68×

Native, one foreign-born parent

Unstandardised
1.86×
Adjusted: five factors
1.34×

Reference: native-born, two native-born parents

Unstandardised
1.00×
Adjusted: five factors
1.00×
Risk ratio of being registered as suspected of at least one offence in 2015–2018. Adjusted column: age, sex, disposable income, education and municipality type. Reference = 1.00. Source: Brå 2021:9, figure 7.
Research

The risk ratio for being suspected of at least one offence in 2015–2018 (persons aged 15+ registered on 31 Dec 2014, versus native-born with two native-born parents) is 2.51 unadjusted for the foreign-born and 3.21 for native-born with two foreign-born parents. After age and sex: 2.25 and 1.96. Also adjusting for disposable income, education and municipality type: 1.76 and 1.68. Source: Brå 2021:9, figure 7.

Observation

Deaths in shootings peaked at 62 in 2022 according to the Police and have since fallen to 54 in 2023, 44 in 2024 and 43 in 2025. The Police recorded 159 completed detonations in calendar year 2023 and 136 in 2024 (annual report 2025, table 4.20; the Police's separate 2024 annual file gives 139), and a revised 190 in 2025. Its initial annual statistics stated 189; the annual report later published 190. The separate first-half comparison fell from 108 in 2025 to 68 in 2026. Source: Police; Brå.

Political interpretation

Whether this should be read as an integration crisis, a recruitment crisis in vulnerable areas, or an effect of the structure of the drug market is a political and analytical question where researchers disagree.

Segregation

Segregation is the geographical side of everything above. Housing segregation, socioeconomic segregation and school segregation reinforce each other across generations.

0204060201553201761201960202161202359202565

Number of vulnerable and particularly vulnerable areas per year

  • Vulnerable areas (total): 2015: 53; 2017: 61; 2019: 60; 2021: 61; 2023: 59; 2025: 65
  • Particularly vulnerable: 2015: 15; 2017: 23; 2019: 22; 2021: 19; 2023: 17; 2025: 19
Vulnerable areas (total)Particularly vulnerable
Source: Swedish Police NOA, situational reports on vulnerable areas 2015–2025.
Observation

The number of vulnerable areas stayed around 59–61 between 2017 and 2023 and rose to 65 in the 2025 assessment. The number of particularly vulnerable areas has fallen from 23 (2017) to 19 (2025). The 2025 assessment dropped the risk-area category, which makes levels before and after 2023 only partly comparable. Source: Swedish Police NOA.

Research

Overcrowding is about four times higher in vulnerable areas than in the rest of the country. The share of children growing up in crowded housing affects school outcomes, health and crime risk. Source: Boverket; Public Health Agency.

The chapter's core question

The second generation

The second generation — children born in Sweden to foreign-born parents — is the decisive yardstick. If outcomes converge towards native background, the system has worked. If they get stuck between the first generation and native background, that points to structural barriers. The picture is mixed.

Upper-secondary eligibility, Y9

Gen 1
0.62
Gen 2
0.86
Native background
1.00
Trend
→ converging

Employment 25–34

Gen 1
0.72
Gen 2
0.90
Native background
1.00
Trend
→ converging

Median income 30–39

Gen 1
0.71
Gen 2
0.86
Native background
1.00
Trend
→ converging

Welfare dependency

Gen 1
2.80
Gen 2
1.70
Native background
1.00
Trend
→ converging
Outcomes as ratio to native background (=1). Welfare dependency is inverted — lower is better. Source: Statistics Sweden STATIV, Skolverket.
Observation

The second generation consistently outperforms the first on school, employment and income — but does not reach the native background level. Employment is about 10 percentage points lower; median income about 14 % lower; welfare dependency nearly twice as high. Source: Statistics Sweden STATIV.

Research

Convergence is stronger for pupils with parents from Europe and East Asia, weaker for pupils with parents from the Middle East and Africa. This holds after controlling for parental education. Source: Rooth & Ekberg, IFAU reports.

Political interpretation

Whether the remaining gap is due to discrimination, cultural distance, housing segregation, the school choice system, or a combination — this is where researchers do not agree. This is the chapter's most important open question.

Summary

What is established, what is uncertain, what is politics?

Well established

  • Establishment takes long and varies sharply by ground.
  • Foreign-born are on average a fiscal net cost in the first decades.
  • Municipalities carry the largest practical burden.
  • Vulnerable areas remain stable around 60.
  • The second generation converges but does not reach native levels.

Uncertain

  • The exact size of the lifetime net cost per ground.
  • How much of the school outcome gap is explained by segregation.
  • Whether the 2024–2026 decline in shootings is sustained.
  • Why convergence differs across origin regions.

Political interpretation

  • Is the net cost "too high" — compared to what?
  • Who bears responsibility for the remaining gap?
  • Which policy measures should be prioritised.

The next chapter — What drove the change? — takes these outcomes and asks which factors in the system determined that this became the result.

Primary sources

Sources in this chapter

Updated 2026-07-20.

Summary

Key takeaways

  • Consequences vary more with composition than with total volume.
  • Establishment takes long and varies sharply by migration ground.
  • Life-cycle studies indicate a fiscal net cost in the first decades; the size depends on method.
  • Municipalities carry the largest practical burden; national averages hide the local reality.
  • Differences in school outcomes and crime statistics narrow but do not disappear after controls.
  • Vulnerable areas have been stable around 60 since 2017; particularly vulnerable have decreased.
  • The second generation converges towards native background — but does not reach it.

Further reading

Sources for this page

The data comes from official primary sources.

Report an error on this pageSupport SVERIGEFAKTA
Share this page:XFacebookRedditWhatsAppEmail
Use this statistic

Copy a reference or the canonical link. The original source is included when registered for this page.

Short source
SVERIGEFAKTA, “What were the consequences? — long-run outcomes that can be observed”. Original source: SCB, Brå, Polisen, OECD, ESO, SOM-institutet. https://sverigefakta.com/en/what-happened-to-sweden/the-consequences
Full reference
SVERIGEFAKTA, “What were the consequences? — long-run outcomes that can be observed”. Original source: SCB (https://www.statistikdatabasen.scb.se/pxweb/sv/ssd/START__AM__AM0401__AM0401R/AKURLBefInrUtrAr/); Brå (https://bra.se/rapporter/arkiv/2021-08-25-misstankta-for-brott-bland-personer-med-inrikes-respektive-utrikes-bakgrund); Polisen (https://polisen.se/om-polisen/polisens-arbete/utsatta-omraden/); OECD (https://www.oecd.org/en/about/programmes/pisa.html); ESO (https://eso.expertgrupp.se/rapporter/tid-for-integration/); SOM-institutet (https://www.gu.se/sites/default/files/2026-03/Svenska%20f%C3%B6rtroendetrender%201986-2025.pdf). https://sverigefakta.com/en/what-happened-to-sweden/the-consequences
APA 7
What were the consequences? — long-run outcomes that can be observed. (n.d.). SVERIGEFAKTA. Retrieved 2026-10-27, from https://sverigefakta.com/en/what-happened-to-sweden/the-consequences Original source: SCB, Brå, Polisen, OECD, ESO, SOM-institutet.
Harvard
“What were the consequences? — long-run outcomes that can be observed” (n.d.) SVERIGEFAKTA. Available at: https://sverigefakta.com/en/what-happened-to-sweden/the-consequences (Accessed: 2026-10-27). Original source: SCB, Brå, Polisen, OECD, ESO, SOM-institutet.
BibTeX
@misc{sverigefakta_2026_en_whathappenedtosweden_theconsequences,
  title        = {What were the consequences? — long-run outcomes that can be observed},
  howpublished = {\url{https://sverigefakta.com/en/what-happened-to-sweden/the-consequences}},
  note         = {Presented by SVERIGEFAKTA; original source SCB, Brå, Polisen, OECD, ESO, SOM-institutet}
}
Canonical link
https://sverigefakta.com/en/what-happened-to-sweden/the-consequences
More citation formats and rules