Analysis
Analysis — long-form deep dives
Data-driven essays with a clear method and traceable sources. Each card shows analysis type, data quality and period so you quickly see what the text can — and cannot — establish.
Guiding question:Which analyses and deep dives are on sverigefakta.com, which questions do they answer and which methods do they use?
What you'll find here
- The analysis library collects long-form texts grounded in official statistics and research. They differ from the fact pages by interpreting patterns — not just reporting numbers.
- A well-supported analysis need not prove causation. Many analyses here are descriptive or trend-based; a few use scenarios or counterfactual estimates.
- The method markers on each card show analysis type, data sources and period. Always read the method section of the analysis before quoting a conclusion.
Crime and safety
Analysis · 3 min read
The shootings that stopped — and what still explodes
Question: Which Swedish places have durably reduced gun violence, and which measures coincide?
Conclusion: A few areas show a multi-year decline that clears a cautious 'stopped' threshold. Interventions coincide — but single causes are rarely provable.
Before/afterResearch synthesisSubstantial uncertainty2012–2025Sources: Swedish Police, Brå, Malmö University, Prosecution Authority
Analysis · 2 min read
The camera surveillance that arrived — 2016–2026
Question: How has police camera surveillance expanded, and what do impact studies show?
Conclusion: Sharp expansion 2019–2025; impact studies show mixed results.
Trend analysisOfficial statisticsMedium confidence2018–2025Sources: Police, IMY, Brå
Analysis · 2 min read
Antisemitic hate crimes during the Gaza war
Question: How have reported antisemitic hate crimes developed after October 7, 2023?
Conclusion: Sharp rise 2023–2024; still above prior average.
Trend analysisOfficial statisticsMedium confidence2018–2025Sources: Brå, SÄPO
Analysis · 2 min read
Emergency services under attack — stones, violence and police escort
Question: How have attacks on police, ambulance and fire services developed?
Conclusion: Sharp increase 2015–2024; 2020 sentencing hikes have not broken the trend.
Trend analysisOfficial statisticsMedium confidence2015–2024Sources: MSB, Police, SKR
Analysis · 3 min read
Overrepresentation in crime — what Brå 1996, 2005 and 2021 actually show
Question: What do Brå's three overrepresentation studies over 25 years show?
Conclusion: Differences persist after socioeconomic controls, but individual attribution is inadmissible.
Cohort analysisOfficial statisticsMedium confidence1985 → 2018 (cohorts)Sources: Brå 1996:2, 2005:17, 2021:9, 2024:5
Labour market and integration
Analysis · 2 min read
SFI, jobs and the years that passed — a cohort analysis
Question: What does employment look like seven years after residence permit for the 2015 refugee cohort?
Conclusion: Employment is below 60% after seven years; self-sufficiency lower. Differences reflect several co-acting factors.
Cohort analysisOfficial statisticsMedium confidence2015–2022Sources: SCB, Audit Office, Skolverket, ESO, IFAU
Public finances
Analysis · 2 min read
Parliament salaries in reality — 81,400 SEK vs the median
Question: How do MP salaries compare to the median wage over time?
Conclusion: The MP salary is about 2.1× the median in 2026 — lower than in 1994, higher than in 2004.
Trend analysisOfficial statisticsHigh methodological confidence1994–2026Sources: Parliament's remuneration board, SCB
Analysis · 3 min read
What did 2015 cost? Lifetime net eleven years later
Question: What is the estimated lifetime net for the 2015 asylum cohort?
Conclusion: Estimates vary widely; outcomes are highly sensitive to employment assumptions.
Life-cycle calculationIn-house calculationSubstantial uncertainty2015 → lifetimeSources: NIER, ESO, Ruist, SCB
Housing and municipalities
Analysis · 3 min read
The housing shortage and the pace of migration
Question: How are population growth, migration, housing construction and shortage connected?
Conclusion: Today's shortage reflects a production collapse and the 2015–2022 cohort — not the low 2024 net immigration. The relationship is co-variation, not a single cause.
Trend analysisOfficial statisticsMedium confidence1993–2025Sources: SCB, Boverket (BME), Migration Agency, FI, Riksbank, NIER
International comparisons
Analysis · 3 min read
What Denmark succeeded with — the 2015 paradigm shift
Question: Which Danish policy choices and outcomes differ from Sweden's?
Conclusion: Lower volumes and faster labour-market requirements — but comparable outcomes hard to isolate.
International comparisonInternational comparisonMedium confidence2001–2025Sources: DK Ministry of Employment, Rockwool, SCB
Institutions and power
Analysis · 7 min read
The new matriarchy? Gender, representation and power in Sweden
Question: What does the gender balance actually look like in education, welfare, public administration, academia and business — and does the talk of a 'matriarchy' hold?
Conclusion: Women dominate numerically in education, welfare occupations and municipal management; men dominate listed-company top posts and the professoriate. Power is divided by sector — not reversed.
Descriptive analysis and thesis testOfficial statistics + supplementary sourcesMedium confidence1977–2025Sources: UKÄ, SCB (occupational register, gender statistics), SKR, Gender Equality Agency, AllBright, Agency for Public Management
Analysis · 12 min read
Suicidal empathy? When compassion replaces judgement
Question: Does Gad Saad's thesis of ”suicidal empathy” — that compassion without consequence-testing harms society — hold when tested against research and Swedish official sources?
Conclusion: The partiality of empathy is well documented and Swedish control failures are on record, but no study links the two. The concept works as a question, not as an explanation.
Literature reviewResearch synthesisMedium confidence1937–2026Sources: Schools Inspectorate, National Audit Office, ESO, Agency for Public Management, PNAS, PLOS ONE, APSR
Definitions
- Analysis
- Systematic interpretation of data or texts to understand patterns or mechanisms.
- Descriptive statistics
- Description of how data look — without explaining why.
- Causal analysis
- Attempt to establish cause–effect relations, usually with experimental or quasi-experimental design.
- Correlation
- Two variables move together. Does not say one causes the other.
- Causation
- A causes B. Requires more than co-variation — e.g. temporal order, mechanism and excluded alternatives.
- Hypothesis
- Testable assumption about a relationship.
- Scenario
- A calculated 'if–then' case. Not a forecast but an illustration of an assumption.
- Forecast
- Statement of what is expected to happen, given current trends and assumptions.
- Estimate
- A value computed with uncertainty — often reported with intervals.
- Counterfactual outcome
- What would have happened if something had not occurred. Model-based.
- Sensitivity analysis
- Test of how the result changes when an assumption is varied.
- Cohort
- A group sharing a defining trait, e.g. year of arrival.
- Time series
- Observations of the same variable over time.
- Comparison group
- Group without the intervention used to estimate a counterfactual.
- Control variable
- Variable held constant to isolate the effect of another.
- In-house calculation
- Calculation performed by the editors on open data. Assumptions are reported.
- Primary source
- Original source of a datum (e.g. SCB, Brå).
- Secondary source
- Source that cites or summarises a primary source.
Method and limitations
- Analyses may use different definitions and periods. Always compare definitions before stacking conclusions.
- In-house calculations depend on chosen assumptions. Change the assumptions and the result changes — this is disclosed in each analysis.
- International comparisons can be affected by different statistical systems, collection methods and legal definitions.
- The absence of an analysis does not mean the absence of a relationship. We do not cover everything.
- A clear trend shows co-variation — not automatically causation. Words like 'caused' or 'effect' are used only where evidence permits.
- Headlines and summaries do not replace the method sections. Read the method box before quoting.
Primary sources
- SCB — Statistics Sweden
- Brå — Swedish National Council for Crime Prevention
- National Board of Health and Welfare
- SALAR — Swedish Association of Local Authorities and Regions
- NIER — National Institute of Economic Research
- ESO — Expert Group for Public Economics
- OECD — Data
- Eurostat
- Swedish Migration Agency — Statistics
- Swedish Public Employment Service — Statistics
FAQ
- What is the difference between a fact page and an analysis?
- The fact pages report numbers with source and definition. The analyses interpret patterns over time or across groups — and take clear responsibility for their conclusions.
- Does 'analysis' mean the conclusion is proven?
- No. An analysis can be well-supported without proving causation. We mark analysis type and method confidence on each card.
- Why do you sometimes use in-house calculations?
- When official statistics do not publish the ratio or scenario the question requires. Assumptions are always disclosed — anyone should be able to reproduce.
- How are sources chosen?
- We start from primary sources (SCB, Brå, Boverket etc.) and cite research where it complements. Secondary sources are used sparingly and marked.
- Why is there no analysis on topic X?
- Either we do not yet have sufficient material, or the topic sits in the fact bank. Absence is not a position.

