Skip to content
SVERIGEFAKTA.com

Review · Safety & Crime

Sweden is expanding surveillance. Does it work?

The number of police cameras is growing and permit requirements have been loosened. Brå's own meta-study shows where the cameras work — and where they don't.

Sweden has expanded camera surveillance sharply since 2018, and since 2020 the police no longer need a permit. But does it work? Brå's own meta-study of around 80 evaluations worldwide gives a clear and partly uncomfortable answer: the cameras reduce planned property crime — vehicle crime, theft, drug offences — in car parks and residential areas. On violent crime and public-order offences there is no demonstrated effect. Below we show what the evaluations find, where researchers disagree, and what the build-out costs — with a source at every step.

Aktiva polis-kameror i Sverige, 2018 → 2024

Från ~700 kameror (2018) till ~3 800 (2024). Uppgången accelererar efter att tillståndskravet slopades för polisen 2020. Värdena är uppskattade utifrån Polisens verksamhetsredovisningar — läs som storleksordning.

Method: Exakt nationell tidsserie publiceras inte — läs som storleksordning.

Source: Polismyndigheten — Verksamhetsredovisning & IMY tillsyn (accessed 2026-01-15)

Senast uppdaterad:

The build-out — three legal steps

Exakta nationella siffror på antal poliskameror per år är ej offentligt aggregerade — grafen ovan bygger på delsummor från Polisens verksamhetsredovisningar och ska läsas som storleksordning.

  1. 2012

    The police's first permanent cameras

    The police begin permanent camera surveillance at locations such as Medborgarplatsen and Stureplan in Stockholm, after a permit from the then county administrative board.

    Polismyndigheten, lägesredovisningar

  2. Aug 2018

    New Camera Surveillance Act (2018:1200)

    The Camera Surveillance Act enters into force. The permit requirement is narrowed — fewer actors need a permit, and law-enforcement purposes weigh more heavily in the balancing test.

    SFS 2018:1200

  3. Jan 2020

    Permit requirement removed for the police, Customs and the Coast Guard

    An amendment removes the permit requirement entirely for law-enforcement authorities. The balancing test against personal privacy is now done more internally. IMY (the data protection authority) retains supervisory responsibility.

    Prop. 2018/19:147 · IMY

What the meta-study actually shows

In 2018 Brå published the meta-study Does camera surveillance prevent crime? It is based on Welsh & Farrington's international review of around 80 evaluations.

Effect by crime type

Vehicle-related crime

Effect
Works
Source
Brå 2018

Property crime (theft, burglary)

Effect
Works
Source
Brå 2018

Drug offences

Effect
Works
Source
Brå 2018

Violent crime

Effect
No demonstrated effect
Source
Brå 2018

Public-order offences

Effect
No demonstrated effect
Source
Brå 2018

Effect by location

Car parks

Effect
Works
Source
Brå 2018

Residential areas (general)

Effect
Works
Source
Brå 2018

City centres

Effect
No demonstrated effect
Source
Brå 2018

Public transport

Effect
No demonstrated effect
Source
Brå 2018

Specific multi-family housing areas

Effect
No demonstrated effect
Source
Brå 2018

Brås metastudie 2018 sammanfattar: kamerabevakning förebygger främst planerade egendomsbrott. På våldsbrott och ordningsstörningar syns ingen påvisad effekt.

Redaktionens sammanfattning: det är alltså inte den brottslighet som driver otryggheten i de mest utsatta miljöerna som kamerorna dokumenterat påverkar. Slutsatsledet är vår tolkning av Brås resultat — inte ett citat från rapporten.

Källa för sakuppgifterna: Brå 2018 (Fungerar kamerabevakning brottsförebyggande?).

Where the research disagrees

ESO 2015:4 — cameras can be economically justified

I ESO-rapporten Verksamma insatser mot brott? (2015:4) konstaterar författarna att flera äldre studier i Brås metaanalys har metodproblem. De få utvärderingar som använder exogen variation och gör en kostnads-nyttoanalys indikerar enligt ESO att kamerareformer faktiskt har varit samhällsekonomiskt motiverade — särskilt mot planerad brottslighet som fickstölder och bilbrott.

ESO lutar alltså mer positivt än Brås metastudie. Det är en genuin forskaroenighet — inte en samlad slutsats — som läsaren behöver känna till.

Local Swedish evaluations diverge

  • Södra Sofielund, Malmö (Brå/Malmö University, 2024): The cameras were perceived to disrupt open drug dealing. But the effect is entangled with other measures — closed streets, BID work, increased police presence — so the camera's isolated contribution is hard to determine.
  • Helsingborg, Landskrona, Medborgarplatsen: Brå's evaluations found no support for a general reduction in crime at the locations studied.

Cost vs benefit

Camera surveillance has investment and running costs: equipment, operation, storage, footage review and staff. Exact national figures are not publicly aggregated — we therefore give no total.

The empirical question becomes: is the scale of the build-out proportionate to the documented effect? Strong evidence exists for a subset (planned property crime in car parks and residential areas). On violent crime — what drives insecurity — there is no demonstrated effect.

Privacy — the actual trade-off

Removing the permit requirement for law-enforcement authorities in 2020 means the balancing test between surveillance needs and personal privacy is to a greater extent made internally by the police, instead of being reviewed ex ante by IMY. IMY retains supervisory responsibility after the fact.

This is a documented goal conflict — expanded state surveillance capacity vs. privacy protection — that the legislature and IMY handle themselves. The page presents the conflict neutrally and does not speculate about future regimes or systemic shifts.

Sources in the original

The question this page answers

How many cameras watch Sweden — and what does the research say about when camera surveillance actually reduces crime?

At a glance

Key figures — the most important numbers on this page

Active police cameras 2024

≈3,800

Swedish Police annual report — order of magnitude, no consolidated official time series

More police cameras 2018→2024

5.4×

From ~700 to ~3,800 active units

Average crime reduction

≈16 %

Welsh & Farrington 2009 meta-analysis of 44 studies — mainly property crime

Effect in car parks

≈51 %

Welsh & Farrington — strongest effect, combined with improved lighting

Effect in city centres

≈7 %

Weakest effect; no statistically significant effect on violent crime

Camera Act replaced permit regime

2018

Preventive surveillance in public space made easier for police

Municipal cameras 2024

≈1,100

IMY (Data Protection Authority) supervision reports — annual reporting by municipalities

Swedes supporting expanded camera surveillance

72 %

SOM Institute 2024 — highest level ever recorded

What does the data show?

Objective observations — not interpretations

  • The number of active police cameras has more than quintupled since 2018. The growth reflects new legislation (the Camera Act 2018), the Tidö Agreement 2022 and the government's 34-point programme (2023), plus targeted deployment in vulnerable areas.
  • The evidence base is clear on one point: cameras work best against planned, rational property crimes in delimited environments (car parks, transit, shops). They work worst against impulsive violent crime and alcohol-related disorder in city centres.
  • Cameras combined with improved lighting and police presence produce a larger effect than cameras alone. Brå (2018) stresses that cameras are a tool, not a policy in themselves.
  • IMY (the Data Protection Authority) has repeatedly sanctioned municipalities and private actors for excessive or unlawful camera surveillance — especially of schools, workplaces and public places without sufficiently strong grounds.
  • No consolidated national statistics on the total number of cameras in Sweden are published. Orders of magnitude rest on Swedish Police annual reports, IMY supervision statistics and industry estimates.

Method note
Camera surveillance is legally regulated by the Camera Act (2018:1200) for public spaces and by GDPR for private actors. Police operational cameras require a permit from IMY, with limited exceptions for urgent needs. Private cameras on one's own property are generally exempt. National totals are estimates — Sweden has no central camera register, unlike the UK.

Definitions

How the numbers are counted — and what they do not cover

Camera surveillance (legal)
Per the Camera Act (2018:1200): monitoring of a location using a camera that is permanently installed and directed at a location to which the public has access. Surveillance of private property or closed workplaces is instead governed by GDPR.
Permit requirement
Since 2018 a permit from IMY is required for camera surveillance of places accessible to the public — with exceptions for police, coast guard and customs. Permits weigh benefit (crime prevention) against privacy intrusion.
Recording vs retention
The camera captures in real time; material may only be retained for as long as needed for the stated purpose, generally no more than 30 days without special grounds.IMY has repeatedly sanctioned actors that retained material longer than necessary.
Displacement
The phenomenon that crime moves from a monitored to a nearby unmonitored location instead of disappearing. Measured as the difference between crime reduction in the studied zone and total reduction in the area.Welsh & Farrington find moderate or no displacement for property crime — cameras generally appear to yield net effects, not merely redistribution.
Diffusion of benefits
The opposite of displacement — preventive measures also reduce crime in adjacent unmonitored areas, e.g. via increased perceived risk of detection.
Active monitoring
Camera feeds are watched in real time by an operator who can call for a response. Effects are typically larger than passive review after incidents — but resource-intensive.
Passive surveillance
The camera records, but images are reviewed only after an incident. Works mainly as signposted deterrence and as post-hoc evidence.
AI analysis and ANPR (number-plate recognition)
Cameras linked to image analysis for e.g. licence plates, facial recognition or behaviour detection. Regulated more strictly in the EU (AI Act 2024) and requires a separate legal basis beyond the Camera Act.
Crime reduction (measured effect)
Difference in reported crime in the monitored zone before/after installation, versus a control area. Standard method in the Welsh & Farrington meta-analyses. Requires at least 12 months of post-period to capture seasonality.
Vulnerable areas (camera context)
Police prioritise camera deployment in the areas listed in the national situational report. The legal basis is the Camera Act itself — not the area status — but the necessity test may be more easily met.

Primary sources

Agencies and research institutions behind this page

4

agencies/institutions

5

reports & studies

2

primary datasets

Frequently asked questions

Short answers to what is most often discussed

The evidence is weak. Welsh & Farrington (2009) find no statistically significant effect on violent crime overall. Shootings and bombings are typically planned and carried out by people who expect to be filmed — cameras then provide mainly evidentiary value after the act, not prevention. Police use cameras in vulnerable areas as one of several tools, partly to raise clearance rates.

Related topics

Logical next steps if you want to understand the background

Sources for this page

The data comes from official primary sources.

Report an error on this pageSupport Sverigefakta