Analysis · 2 min read
SFI, jobs and the years that passed — students, completion and work
Skolverket’s report 2024:3 separates participation in 2022, results for SFI entrants in 2018 and labour-market status in 2021 after leaving in 2019. The measures concern different student populations, not verified tracking of the 2015 refugee cohort or SFI’s causal effect.
Guiding question:What does SFI statistics show about participation, completion and work — and which groups are followed?
Short version in the chapter: Work and benefits
Published 2026-07-15
Participation is not completion
During calendar year 2022, 123,500 students attended SFI; 62% were women. This is the report’s student count, not new refugees or residence permits. One person may count as several course participants when taking different courses.
The previous claim of 71,700 refugees granted permits in 2015 and associated employment, education and municipality shares lacked an identifiable matched register extract. They are not used here. The question of SFI and work is instead answered using the report’s explicitly defined student populations.
Entrants in 2018 followed through 2022
Skolverket follows 52,800 individuals who started SFI in 2018. After five calendar years, through 2022, 63% had passed at least one course; 37% had passed none. Having no passing grade must not automatically be equated with being without work.
The proportion passing final course D was 12% on study path 1, 33% on path 2 and 44% on path 3. The paths have different educational backgrounds. This is not the population of 15,400 who completed course D in 2022 used in the report’s study-duration table.
Chart observations in a table
Entrants in 2018 followed through 2022. Different study paths, not immigration cohorts. 52,800 individuals in the study-results follow-up.
Study path 1
- Value
- 12
Study path 2
- Value
- 33
Study path 3
- Value
- 44
| Measure and period | Value |
|---|---|
| Study path 1 | 12 |
| Study path 2 | 33 |
| Study path 3 | 44 |
Work after SFI: a different student population
The labour-market follow-up covers 56,900 individuals who left SFI during calendar year 2019, with or without passing grades, and their status in 2021. It does not measure time since immigration.
The report’s figure 9 shows established status for 26% of those with course D, 28% passing a lower course and 36% without a passing grade. In the last group, 64% overall were in work when established, uncertain and weak status were combined. These terms are not synonymous.
In the report’s definition of established status, earned income above a threshold is central; in 2021 the threshold was SEK 18,092 per month averaged over the year. Uncertain and weak status involve lower income or unemployment among other criteria. Students may also have worked during part of the year.
Association is not an observed effect
Students without a final grade being more frequent in some work categories does not establish that SFI worsens employment. Interviewed providers describe students sometimes leaving to work and parallel work making studies irregular. These are interview accounts, not a causal effect estimated by this statistical comparison.
Follow-up needs the same group, period and definition. Grades, work, studies and sustained income answer different questions. Group proportions do not determine an individual student’s ability or outcome.
What this shows
- Participation 2022, results 2018–2022 and labour-market status 2019–2021 concern different student populations.
- Passing a course and established status are different outcomes. Statistics do not show SFI’s causal effect.
Definitions
- Student and course participant
- A person may attend several courses and count as several course participants.
- Established status 2021
- The report’s income-based category; threshold SEK18,092/month averaged over the year. Not every form of work.
Method and limitations
- Register analyses in Skolverket2024:3 use different populations; they must not be joined into a 2015 cohort.
- Providers’ interview explanations are not causal effect estimates.
Primary sources
FAQ
- Does this page follow the 2015 refugee cohort?
- No. Results concern SFI students: entrants2018 followed to2022 and leavers2019 followed to2021. A matched register extract for the previously claimed refugee cohort is missing.
- Does leaving SFI mean being without work?
- No. The report shows64% without passing grades were in work2021, two years after leaving2019, with differing income and labour-market status. This does not prove a causal effect of leaving.