Speed Camera Effectiveness: What the Research Actually Says

Few road safety interventions provoke as much debate as speed cameras. Supporters point to research showing large crash reductions. Critics dismiss them as revenue-raising tools that do nothing for genuine safety. The debate generates more heat than light, and both sides frequently cherry-pick evidence to support their position.

This article examines what the research actually says. Not what advocates claim, not what opponents assert, but what peer-reviewed studies, systematic reviews, and real-world programme evaluations have found about whether speed cameras reduce crashes, injuries, and deaths.

The Evidence Base: An Overview

The body of research on speed camera effectiveness is large. Decades of studies from multiple countries, using various methodologies, have examined the impact of different camera types on speed and crash outcomes.

The evidence falls into several categories:

• Systematic reviews and meta-analyses that synthesise findings across multiple studies
• Individual country programme evaluations (UK, Australia, Sweden, New Zealand)
• Site-specific before-and-after studies examining crash changes at individual camera locations
• Area-wide studies examining whether camera programmes affect broader regional crash trends
• Speed measurement studies examining the effect on actual vehicle speeds

The overall picture is more consistent than the public debate would suggest, but it comes with important caveats about research quality and methodology.

The Cochrane Review: The Gold Standard

The most authoritative assessment of speed camera evidence comes from the Cochrane Collaboration, an international organisation known for producing rigorous systematic reviews of healthcare and public health interventions.

What They Found

The Cochrane review on speed cameras, most recently updated in 2010, examined 35 studies that met their pre-set quality criteria. Their findings were:

Speed reductions:

• All studies reporting speed outcomes found reductions in average speeds after camera implementation.

Crash reductions near camera sites:

• All crashes: reductions ranging from 8% to 49%, with most studies in the 14% to 25% range.
• Injury crashes: reductions between 8% and 50%.
• Fatal and serious injury crashes: reductions between 11% and 44%.

Crash reductions over wider areas:

• All crashes: reductions from 9% to 35%, with most studies in the 11% to 27% range.
• Fatal and serious injury crashes: reductions from 17% to 58%, with most studies reporting 30% to 40% reductions.

The Cochrane Review's Caveats

The Cochrane reviewers were careful to note that the quality of included studies was "of overall moderate quality at best." But they concluded: "The consistency of reported positive reductions in speed and crash results across all studies show that speed cameras are a worthwhile intervention for reducing the number of road traffic injuries and deaths."

They called for higher-quality studies using controlled trial designs, longer follow-up periods, and more data collection points both before and after camera implementation.

This is an important nuance. The evidence is consistently positive, but its quality could be better. The direction of the effect is clear. Its precise magnitude is less certain.

New Zealand Research

New Zealand has produced its own body of research on camera effectiveness, which is particularly relevant because it reflects local conditions, road types, and enforcement practices.

Keall and Frith: Hidden vs Visible Cameras

Researchers Keall and Frith at the Land Transport Safety Authority conducted a study comparing hidden (covert) mobile speed cameras with the existing visible (overt) camera programme. Overt mobile speed cameras had been used in New Zealand since late 1993, confined to designated "speed camera areas," mainly road sections with a record of speed-related crashes.

A trial of hidden speed cameras began in mid-1997 in 100 km/h speed limit areas in one of New Zealand's four Police regions. The results were striking:

• The hidden camera programme produced an additional 17% reduction in injury crashes (not statistically significant) and a 31% reduction in casualties (statistically significant) over a two-year period, compared to the conspicuous camera programme alone.
• The generalised effects extending across the whole trial area were estimated at an 11% reduction in crashes and a 19% reduction in casualties, both statistically significant.

A follow-up study confirmed these findings, suggesting that hidden cameras create a broader deterrent effect because drivers can't predict exactly where enforcement will occur.

Mara, Davies, and Frith: Early Camera Evaluation

Earlier New Zealand research by Mara, Davies, and Frith (1996) estimated crash reductions of approximately 20% associated with speed camera installation. This figure, along with Keall and Frith's work, provided the evidence base for Waka Kotahi's estimate that fixed spot speed cameras reduce deaths and serious injuries by approximately 20%.

Waka Kotahi Operational Data

More recent operational data from Waka Kotahi provides real-world evidence of camera effects on driving behaviour:

• Across ten new camera sites with before-and-after tracking, speed limit compliance increased from 57% to 98%.
• At the Matakana Road average speed camera site (operational December 2025), compliance increased from 88% to over 99%.
• Waka Kotahi estimates that average speed cameras reduce deaths and serious injuries by approximately 48%, far more than fixed spot cameras.

These compliance figures are striking. A jump from 57% to 98% compliance represents a fundamental change in driver behaviour, not a marginal adjustment. I remember being genuinely surprised when I first saw those numbers.

UK Evidence

The United Kingdom has the most extensive body of speed camera research, reflecting decades of camera deployment across the country.

National Safety Camera Programme

The UK's national camera programme, which operated from 2000 to 2007, produced detailed evaluation data. The findings were consistent with the international evidence:

• Reductions in collisions at camera sites ranged from 5% to 69%, with most sites showing reductions of 15-25%.
• Injury reductions ranged from 12% to 65%.
• Death reductions ranged from 17% to 71%.

The London Study

Research by Gorell and Sexton examining 77 speed cameras in London found a 26% reduction in casualty crashes when all approach directions were considered. For right-angle and right-turn-against crashes (the types most affected by speed cameras at intersections), the reduction was 44%.

The estimated real-world benefit was the prevention of 17 serious or fatal crashes per year and 39 minor injury crashes, representing crash cost savings to the community of over $8 million.

Rural County Study

A study of mobile speed cameras in a rural English county found crashes decreased by 19% and fatal and serious crashes by 44% at camera sites.

Controlling for Regression to Mean

The London study controlled for regression to the mean (discussed in detail below), finding that after this statistical correction, the estimated reduction in all injury collisions was 12%, and the reduction in fatal and serious injury crashes was 21%. These are lower than the raw figures but still large and statistically meaningful.

Diffusion of Benefits

Some UK studies found evidence of benefits extending beyond the immediate camera site. One study reported a marked 21% reduction in crashes on a motorway section without camera enforcement adjacent to an enforced section. Another found real crash reductions extending one kilometre upstream and downstream of camera locations.

Australian Evidence

Victoria: The Most Studied Programme

Victoria, Australia, has one of the world's most extensively evaluated speed camera programmes, primarily through research conducted by Monash University's Accident Research Centre (MUARC).

Fixed Camera Effectiveness

MUARC evaluated Victoria's fixed digital speed and red-light cameras across 77 intersections and found:

• A 27.7% reduction in all casualty crashes (statistically significant).
• A 30.2% reduction in fatal and serious injury crashes.
• A 44% reduction in right-angle and right-turn-against crashes.
• An estimated 17 serious or fatal crashes prevented per year and 39 minor injury crashes, representing annual crash cost savings exceeding $8 million AUD.

Fixed Freeway Cameras

A separate evaluation of Victoria's fixed freeway speed cameras found:

• A statistically significant 41% reduction in fatal crash outcomes associated with high levels of camera activity.
• Crash reductions that increased monotonically with increasing levels of speed camera enforcement.

This dose-response relationship (more enforcement leading to greater reductions) is important evidence that the observed effects are genuinely caused by the cameras rather than being coincidental.

Mobile Camera Effects

Victoria's mobile speed camera programme was associated with a 3.8% reduction in casualty crashes across the state and a 4.8% reduction in metropolitan Melbourne. While these percentages appear modest, applied across the entire state road network they represent a large number of prevented crashes.

The Interaction Effect

An interesting finding from Victorian research was the interaction between speed camera enforcement and media publicity. The combination of cameras and public awareness campaigns produced greater speed reductions than either intervention alone, suggesting that cameras are most effective when accompanied by public communication about their purpose and location.

How Effectiveness Is Measured

Understanding the research requires understanding how effectiveness is typically assessed.

Speed Reduction Studies

The most straightforward measurement is the effect on vehicle speeds. These studies use speed measurement equipment (radar, loops, tubes) to record vehicle speeds before and after camera installation, and at camera sites versus control sites. Speed reduction is the most direct measure of camera effect, but it doesn't directly measure crash outcomes.

Crash Reduction Studies

Crash studies compare the number and severity of crashes at camera sites (and sometimes surrounding areas) before and after camera installation. These are more directly relevant to safety outcomes but face real methodological challenges:

• Before periods may not be representative if they include unusual events.
• After periods may be affected by other changes (road modifications, traffic changes) that coincide with camera installation.
• Crash data quality varies, and changes in reporting practices can create artificial trends.

The Regression to Mean Problem

This is the single most important methodological issue in speed camera research, and it deserves detailed attention.

Speed cameras are typically installed at locations with a history of crashes. Crash numbers fluctuate randomly from year to year. A location that experiences an unusually high number of crashes in one period is statistically likely to experience fewer crashes in the next period, simply due to random variation. This is called regression to the mean.

If cameras are installed at high-crash locations (which they almost always are), some of the observed crash reduction after installation would have occurred anyway, without the camera. Studies that don't control for this effect will overestimate camera effectiveness.

Research has addressed this in several ways:

• Control groups: Comparing camera sites to similar sites without cameras. If the camera sites show greater improvement, the difference can be attributed to the cameras.
• Empirical Bayes methods: Statistical techniques that estimate the expected crash rate at a site based on its characteristics and the overall crash trend, providing a more accurate baseline for comparison.
• Long before-periods: Using multiple years of pre-installation data reduces the influence of a single high-crash year that may have triggered camera installation.

The general finding is that after controlling for regression to the mean, the estimated crash reduction from speed cameras is smaller than raw before-and-after comparisons suggest, but remains positive and statistically meaningful. Studies that apply these corrections typically find reductions in the range of 10-25% for all crashes and 15-35% for serious and fatal crashes, compared to raw estimates that may reach 30-50%.

The Halo Effect: Do Benefits Extend Beyond the Camera?

One of the key questions about speed cameras is whether their effects extend beyond the immediate camera location. Researchers distinguish between two types of halo effect.

Distance Halo

The distance halo is the distance over which speed reductions persist upstream and downstream of a camera site.

Research findings on distance halo:

• Fixed cameras: The Norwegian Public Roads Administration found speed reduction effects extending 2.5 to 3.0 km from camera sites, though the largest effect was at the site itself.
• Mobile cameras: One study found a 6 km/h reduction in mean speed immediately adjacent to the camera, but effects had completely disappeared by 1,500 metres downstream.
• Urban enforcement: A study found speed reductions extending 500 metres upstream and downstream of enforcement locations.

Time Halo

The time halo is the duration for which speed reduction effects persist after enforcement operations cease (relevant to mobile cameras).

Research on time halo has found less encouraging results. A study on overt mobile speed cameras found no measurable time halo effect, meaning speeds returned to pre-enforcement levels relatively quickly after cameras were removed.

This finding has important implications. It suggests that permanent fixed cameras or average speed cameras, which provide continuous enforcement, are likely to produce more sustained speed reductions than mobile cameras that operate intermittently.

Average Speed Cameras: What the Evidence Shows

Average speed cameras (point-to-point enforcement) represent a newer technology that addresses several limitations of fixed spot cameras.

Why Average Speed Cameras May Be More Effective

Average speed cameras measure speed over a distance rather than at a single point. This eliminates the "kangaroo" effect where drivers brake for a camera and then accelerate, and it enforces compliance along the entire monitored section.

Research Evidence

Research has consistently found average speed cameras to be more effective than fixed spot cameras:

• Average speed enforcement lowered the ratio of unsafe driving by 10.97%, far outperforming fixed cameras which reduced it by only 0.20-0.57%.
• Point-to-point cameras can detect 95% of drivers' speeds compared to spot cameras which only detect 10-20% of speeders.
• Drivers in point-to-point zones are more likely to maintain a constant speed throughout the monitored section, rather than showing the speed variations typical near fixed cameras.

Speed Behaviour Changes

Research from New York City found that automated speed cameras reduced traffic crashes by 14% and decreased speeding violations by 75% over time, with most cameras achieving their safety purpose within six months. Violations dropped and stayed low, indicating genuine and lasting behavioural change rather than temporary compliance.

The ITF Assessment

The International Transport Forum (ITF/OECD) assessment concluded that average speed enforcement "offers a greater network-wide approach to managing speeds" and represents "a highly reliable approach producing considerable returns on investment through reduced crash costs," though it's comparatively more expensive to deploy than fixed cameras.

Waka Kotahi's Estimate

Waka Kotahi's estimate that average speed cameras reduce deaths and serious injuries by approximately 48% is much higher than the 20% estimated for fixed spot cameras. While the specific methodology behind this estimate isn't publicly detailed, it's broadly consistent with international research showing average speed cameras achieving roughly double the crash reduction of spot cameras.

The Economic Case

Cost-Benefit Analysis

Speed cameras are consistently shown to be cost-effective in formal economic analyses:

• The net benefit of speed cameras was estimated at five times the initial investment after one year and more than 25 times after five years in one major evaluation.
• Fine income covered operational costs in all but one area studied.
• In Victoria, Australia, the estimated annual crash cost savings from fixed cameras exceeded $8 million AUD from just 77 intersection installations.
• In Sweden, the government reported a 20% reduction in traffic-related fatalities after implementing a nationwide system, with revenue exceeding operational costs within three years.

The Revenue Question

Speed camera revenue is a lightning rod for criticism. But the economic case for cameras doesn't depend on fine revenue. The primary economic benefit comes from reduced crash costs: fewer emergency responses, fewer hospitalisations, less rehabilitation, less lost productivity, and less human suffering.

A speed camera that generates no fine revenue because everyone obeys the limit is the most successful speed camera of all. The criticism that cameras are "revenue raising" implicitly acknowledges that people are speeding, which is the problem the cameras are designed to address.

Arguments Against Cameras

A fair assessment of the evidence requires engaging seriously with the arguments against speed cameras. Some are stronger than others.

Revenue Raising

The argument: Cameras are primarily deployed to generate revenue for government rather than to improve safety. Camera placement is optimised for ticket volume rather than crash reduction.

What the evidence says: The incentive problem is real. When fine revenue flows to the agency that operates cameras, there's a structural conflict of interest. Some jurisdictions have addressed this by directing fine revenue to road safety initiatives or general funds rather than the operating agency's budget. In New Zealand, speed camera fines go to the Crown, not to Waka Kotahi, which reduces (but doesn't eliminate) the conflict.

But the argument doesn't address the effectiveness question. Whether cameras are motivated by revenue or safety, the research consistently shows they reduce crashes. The motivation matters for policy design, but it doesn't negate the safety benefit.

Distraction

The argument: Drivers who brake suddenly for speed cameras create a hazard by disrupting traffic flow and distracting attention from the road.

What the evidence says: The brake-and-accelerate pattern is real and is one reason why average speed cameras are considered superior to fixed spot cameras. But the research doesn't support the conclusion that camera-induced braking causes more crashes than it prevents. The net crash reduction at camera sites is consistently positive, meaning any crashes caused by braking are more than offset by crashes prevented through lower speeds.

Displacement

The argument: Cameras don't reduce speeding overall. They merely displace it to unmonitored roads.

What the evidence says: Some displacement may occur, but the area-wide studies reviewed by Cochrane found crash reductions across broader regions, not just at camera sites. The net effect across the road network is positive. And average speed cameras address this concern more effectively because they enforce compliance over longer distances.

Civil Liberties

The argument: Automated enforcement represents excessive government surveillance and violates privacy and due process.

What the evidence says: This is a legitimate values-based argument that can't be resolved by evidence alone. Different societies balance surveillance and safety differently. New Zealand's current framework requires cameras to be signposted in advance and limits the data collected and retained. Whether this balance is appropriate is a political and philosophical question, not a scientific one.

Speed Is Not the Problem

The argument: Speed cameras target drivers exceeding the limit by small amounts while ignoring the real causes of crashes: drunk driving, fatigue, distraction, and poor road design.

What the evidence says: Speed is rarely the sole cause of a crash, but it's almost always a factor in crash severity. The relationship between speed and crash energy is quadratic: a 10% increase in speed produces a roughly 20% increase in crash energy and a 40% increase in fatal crash risk (Nilsson's Power Model). Reducing speed doesn't prevent all crashes, but it makes crashes more survivable. And speed cameras aren't deployed instead of other interventions. They're one part of a broader road safety system.

Limitations of the Research

Transparency about the limitations of the evidence base is essential for an honest assessment.

Study Quality

The Cochrane review rated the overall quality of studies as moderate at best. Few studies used randomised controlled trial designs (which are difficult to implement for road safety interventions). Most relied on before-and-after comparisons with varying degrees of statistical control.

Publication Bias

There may be a tendency for studies showing positive effects to be published more readily than those showing null or negative effects. While the consistency of findings across many countries and contexts provides some reassurance, publication bias can't be ruled out.

Confounding Variables

Speed cameras are often implemented alongside other changes: road improvements, increased policing, public awareness campaigns, or changes in traffic patterns. Isolating the specific effect of the camera from these concurrent changes is methodologically difficult.

Context Dependency

Research from one country may not directly apply to another. Road conditions, traffic patterns, enforcement practices, driver culture, and vehicle fleets all vary between countries. While the direction of the effect (cameras reduce crashes) appears universal, the magnitude may vary depending on local conditions. I've tried to focus on New Zealand and comparable countries here, but even then the transferability of results isn't guaranteed.

Long-Term Effects

Most studies cover relatively short follow-up periods (2-5 years). Whether camera effects are sustained over decades, or whether drivers eventually adapt and revert to higher speeds, is less well understood. The limited evidence on time halo for mobile cameras suggests that sustained enforcement is important, but more long-term research is needed.

A Balanced Conclusion

The evidence on speed camera effectiveness is more consistent and more positive than the public debate suggests. Across multiple countries, research designs, and camera types, the finding is the same: speed cameras reduce vehicle speeds, and reduced speeds lead to fewer and less severe crashes.

The magnitude of the effect varies. After controlling for regression to the mean and other methodological issues, the best estimate is that fixed speed cameras reduce crashes at their locations by approximately 10-25%, with larger reductions for serious and fatal crashes (15-35%). Average speed cameras appear to achieve much greater reductions, in the range of 30-50%.

But cameras aren't a panacea. They work best as part of a complete road safety system that includes good road design, safe vehicles, effective driver licensing, and policing of impaired driving. Cameras deployed as a substitute for infrastructure investment or as a primary revenue source will be less effective and less publicly accepted than cameras deployed as one component of a genuine safety strategy.

The most productive framing of the debate isn't "are speed cameras effective?" (the evidence says yes) but "how should speed cameras be deployed to maximise safety benefit?" The evidence suggests the answers are: at locations selected on safety grounds rather than revenue potential, as part of a visible and transparent programme, in conjunction with other safety measures, and increasingly using average speed technology that enforces compliance over distances rather than at single points.

New Zealand's expanding safety camera programme, with its emphasis on average speed cameras and transparent site selection based on crash data, is broadly consistent with what the evidence recommends. Whether it goes far enough, or whether the resources could be better spent on infrastructure improvements, remains a legitimate debate. But the claim that speed cameras don't work is not supported by the evidence.

Sources

• Do Speed Cameras Reduce Road Traffic Crashes? - Cochrane Review
• Speed Cameras for the Prevention of Road Traffic Injuries and Deaths - Cochrane Library
• Effectiveness of Speed Cameras: Systematic Review - BMJ/PMC
• The Relative Effectiveness of Hidden vs Visible Speed Cameras - Keall, Povey, Frith
• Further Results from a Trial Comparing Hidden and Visible Cameras - Keall, Povey, Frith
• Evaluation of Victoria's Fixed Digital Speed and Red-Light Cameras - MUARC Report 307
• Evaluation of Victoria's Fixed Freeway Speed Cameras - MUARC Report 350
• Do Speed Cameras Reduce Road Traffic Collisions? - PLOS One
• Speed Cameras - UK College of Policing
• Reducing Road Traffic Injuries: Effectiveness of Speed Cameras in an Urban Setting - PMC
• The Effectiveness of Average Speed Cameras - ITF/OECD
• Time and Distance Halo Effects - EU Road Safety Observatory
• Investigating Distance Halo Effect of Fixed Speed Cameras - ScienceDirect
• Speed Cameras, Regression to Mean - Understanding Uncertainty
• Safety Cameras - Waka Kotahi
• Speed Safety Camera Enforcement - NHTSA
• Speed Limit Enforcement Evidence Brief - NZ Ministry of Justice