Study analysis · Journal of safety research · 2026
The U.S. and Canada drive side by side, but their traffic death trends went in opposite directions—and the gap is now 2.5-times higher in the U.S.
From 2007 to 2021, U.S. traffic deaths fell then rose 33% relative from 2011 to 2021, while Canada's kept falling; in 2020, about 118 per million Americans died in crashes versus 46 per million Canadians.
Overview
What the study found
The study in plain English — the bottom line, every takeaway we extracted, and what to do with them.
In simple terms
This study looks at patterns in car crash deaths in two countries over many years. It can show that some things happen at the same time, like certain laws and fewer deaths, but it can't prove that the laws caused the fewer deaths. It's like noticing that people who carry umbrellas also get wet less, but you can't be sure the umbrella is the reason.
What’s the bottom line?
A study compared car crash deaths in the United States and Canada from 2007 to 2021. U.S. deaths went down early then up, while Canada's kept going down. Driving exposure explains much of the U.S. pattern, and Canada's stricter safety laws may help explain its decline.
How strong is this study?
The study uses lots of data from entire countries over a long time, which is good for seeing big trends. But because it's not a controlled experiment, other things could be causing the differences, so we should be careful about trusting that one thing caused another.
40 / 100
- COI disclosure+40/40
- Data availabilitydata not shared
- Code availabilitycode not shared
13 / 100
- Randomizationnot randomized
- Blindingblinding unclear
- Control groupno control group
- Sample sizeno sample size reported
- Follow-up+10/10
100 / 100
0 / 100
- P-valuesno p-values reported
- Effect sizeno effect size reported
- Confidence intervalsno confidence intervals
- Pre-registrationnot pre-registered
Each component is scored out of 100 and then capped by the study design — a case series cannot reach the ceiling a randomised trial can, however well it is reported.
Where it sits
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 532 / 100
Probability of being correct
Groups of people are followed over time to see who develops an outcome. Strong for identifying risk factors and associations, but cannot prove causation as firmly as RCTs.
This design cannot establish causation — the findings describe an association, not a cause. This is an observational study using aggregate national data without randomization or a control group. It can identify associations and trends but cannot establish causal relationships due to potential confounding, ecological fallacy, and inability to control for all relevant variables.
No Conflicts
No conflicts of interest identified
No conflicts identified; authors declare no competing interests and no funding was reported.
Conflict Details
Traffic Injury Research Foundation: Director of Road Safety Engineering
Declaration of competing interest states no known competing financial interests or personal relationships. Acknowledgement: None. No funding statement is present. Author Craig Lyon is affiliated with the Traffic Injury Research Foundation, a non-profit road safety research organization; no industry funding or industry employment is disclosed.
Key takeaways
- 01
U.S.
- 02
crash deaths fell from 41,259 in 2007 to 32,479 in 2011, then rose to 43,230 in 2021—a 33% relative increase from 2011.
- 03
Canada fell from 2,238 in 2007 to 1,746 in 2020.
- 04
In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians—a 2.5-times higher relative rate in the U.S.
- 05
In 2020, the U.S.
- 06
had about 118 crash deaths per million people versus 46 per million in Canada—about 72 extra deaths per million people in the U.S.
- 07
The study did not report a single absolute risk increase for specific laws; it only shows associations, not proof that laws caused the difference.
Surprising findings
- U.S. fatality rate per 100 million VMT was relatively stable while Canada's consistently decreasedMany assume the U.S. has worse per-mile road safety than Canada. This suggests the U.S. total death rise is mostly about more driving, not a sharp per-mile deterioration.
- The 2020 U.S. rate was 2.5-times higher relative than Canada's: 118 vs 46 per million peopleA 2.5x relative gap between two similar countries is larger than many would guess. It suggests structural differences beyond just driving amount.
- Laws on distraction, seat belts, and speed cameras are associated with part of the divergence, but no effect sizes were reportedYou'd expect a study highlighting laws to quantify their impact. Here, the association is suggestive but statistically vague—no CIs or p-values.
Practical takeaways
When comparing road safety between places, look at fatalities per vehicle mile traveled, not just total deaths—total counts can rise simply because people drive more.
This study is observational and country-level; per-VMT rates can still hide differences in crash types, reporting, and road mix.
medium confidenceSupport proven or promising safety laws—seat belt enforcement, distracted driving bans, and speed safety cameras—as part of a broader strategy to reduce road deaths.
The study found associations, not causation, and did not quantify how much these laws reduce fatalities. Other factors (economy, demographics, vehicle safety) also matter.
low-to-medium confidenceIf you're a policymaker, focus on both exposure (how much people drive) and risk per mile—because the U.S. rise was largely exposure-driven, while Canada improved risk factors.
Exposure is influenced by fuel prices, unemployment, urban planning, and transit availability—all outside this study's causal scope.
medium confidenceUse the 118 vs 46 per million gap as a conversation starter about road safety priorities, but avoid claiming a single cause.
Canada data only through 2020; U.S. data through 2021. The study is ecological and cannot prove what would happen if the U.S. adopted Canadian laws.
medium confidenceWhy this study matters
Opposite directions: U.S. deaths rose 33% relative while Canada's fell
U.S. traffic fatalities dropped from 41,259 in 2007 to 32,479 in 2011, then climbed to 43,230 in 2021—a 33% relative increase from 2011. Canada fell from 2,238 in 2007 to 1,746 in 2020, continuing a downward trend.
It's a natural experiment: two similar countries with diverging road safety outcomes. People assume trends are driven by driver behavior alone, but this shows national-level differences.
More driving, not worse per-mile risk, explains much of the U.S. increase
When adjusted for vehicle miles traveled (VMT), the U.S. fatality rate per 100 million VMT was relatively stable from 2007 to 2021, while Canada's rate consistently decreased. That suggests the U.S. total rise was largely driven by increased driving exposure, whereas Canada reduced other risk factors.
It flips the common narrative that U.S. roads are uniquely dangerous per mile. The U.S. risk per mile didn't worsen much—people just drove more.
The 2020 gap: 118 vs 46 deaths per million people
In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians—a 2.5-times higher relative rate in the U.S. That's roughly 72 extra crash deaths per million people in the U.S. in that year.
A 2.5x relative gap is huge for two wealthy, car-dependent neighbors. It raises immediate questions about policy, road design, and vehicle mix.
Crash types driving the divergence: pedestrians, cyclists, alcohol, speeding, trucks
Several crash types showed opposite trends in the U.S. and Canada: pedestrian fatalities, cyclist fatalities, alcohol-involved fatalities, speeding-related fatalities, and truck-involved fatalities all contributed to the divergence, with increases in the U.S. and decreases in Canada (or vice versa).
It's not just one factor—multiple crash categories are moving in opposite directions. That suggests broad systemic differences, not a single cause.
Canada's stricter laws may explain part of the gap—but no effect sizes
Statistical models indicated that part of the difference may be due to Canada's more substantive and common laws related to driver distraction, seat belt use, and speed safety cameras. However, the study reported no effect sizes, confidence intervals, or p-values for these law associations.
It points to policy as a lever, but the lack of precision means we can't say how much laws matter versus other factors.
Want the whole report?
Detailed mode opens the full scientific breakdown — every score component, the methodology, conflicts of interest, the evidence analysis behind each claim, and the raw study data.
Overview
What the study found
The study in plain English — the bottom line, every takeaway we extracted, and what to do with them.
Not medical advice. For informational purposes only. Always consult a healthcare professional. Terms
A study compared car crash deaths in the United States and Canada from 2007 to 2021. U.S. deaths went down early then up, while Canada's kept going down. Driving exposure explains much of the U.S. pattern, and Canada's stricter safety laws may help explain its decline.
Research results
U.S. crash deaths fell from 41,259 in 2007 to 32,479 in 2011, then rose to 43,230 in 2021—a 33% relative increase from 2011. Canada fell from 2,238 in 2007 to 1,746 in 2020. In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians—a 2.5-times higher relative rate in the U.S.
What this means - more context
In 2020, the U.S. had about 118 crash deaths per million people versus 46 per million in Canada—about 72 extra deaths per million people in the U.S. The study did not report a single absolute risk increase for specific laws; it only shows associations, not proof that laws caused the difference.
To examine why U.S. traffic fatality trends diverged from Canada's from 2007 to 2021 and identify crash types and risk factors explaining the divergence.
U.S. total traffic fatalities declined from 41,259 in 2007 to 32,479 in 2011, then rose 33% relative to 43,230 in 2021. Canada declined from 2,238 in 2007 to 1,746 in 2020. After adjusting for vehicle miles traveled, the U.S. fatality rate per 100 million VMT was relatively stable, while Canada's rate consistently decreased. In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians, a 2.5-times higher relative rate in the U.S.
Methods Used
Used NHTSA Fatality Analysis Reporting System and Canadian National Fatality Database data from 2007 to 2021. Analyzed total and crash-type-specific fatalities and built statistical models with travel exposure, demographic, economic, and law-related variables.
Main Finding
U.S. total fatalities fell 2007–2011 then rose 33% relative from 2011 to 2021 (32,479 to 43,230), while Canada fell from 2,238 in 2007 to 1,746 in 2020. Per 100 million VMT, the U.S. fatality rate was relatively stable and Canada's decreased. Population and VMT had positive associations with fatality counts. Canada's more common laws on driver distraction, seat belt use, and speed safety cameras may explain part of the divergence; effect sizes for these laws were not specified.
Confidence Level
Medium: national surveillance data and statistical models, but observational/ecological comparison, no effect sizes, confidence intervals, or p-values reported, so causality is not established.
Study Flags
Red Flags
- •Observational country-level comparison; cannot establish causation.
- •No effect sizes, confidence intervals, or p-values reported for law associations.
- •Canada full data only through 2020; potential confounding by unmeasured factors.
No biological mechanisms were identified in this study. This may be an epidemiological, observational, or survey-based study that reports associations rather than proposing causal biological pathways.
Surprising Findings
U.S. fatality rate per 100 million VMT was relatively stable while Canada's consistently decreased
Many assume the U.S. has worse per-mile road safety than Canada. This suggests the U.S. total death rise is mostly about more driving, not a sharp per-mile deterioration.
Practical Takeaways
When comparing road safety between places, look at fatalities per vehicle mile traveled, not just total deaths—total counts can rise simply because people drive more.
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 532 / 100
Probability of being correct
Groups of people are followed over time to see who develops an outcome. Strong for identifying risk factors and associations, but cannot prove causation as firmly as RCTs.
Human Cohort Study
Subject
Lower probability
on the GRADE evidence scale
This study looks at patterns in car crash deaths in two countries over many years. It can show that some things happen at the same time, like certain laws and fewer deaths, but it can't prove that the laws caused the fewer deaths. It's like noticing that people who carry umbrellas also get wet less, but you can't be sure the umbrella is the reason.
No conflicts of interest were detected in this study. No score impact.
Strengths
- Uses comprehensive national fatality databases (FARS and Canadian National Fatality Database)
- Covers a 15-year period (2007-2021)
- Examines multiple crash types and exposure variables
Weaknesses
- Observational design without randomization or control group
- Aggregate (ecological) data limits individual-level inference
- Potential unmeasured confounding (e.g., cultural, enforcement differences)
Methodology
Evidence Keywords
Statistical Reporting
Not medical advice. For informational purposes only. Always consult a healthcare professional. Terms
A study compared car crash deaths in the United States and Canada from 2007 to 2021. U.S. deaths went down early then up, while Canada's kept going down. Driving exposure explains much of the U.S. pattern, and Canada's stricter safety laws may help explain its decline.
Research results
U.S. crash deaths fell from 41,259 in 2007 to 32,479 in 2011, then rose to 43,230 in 2021—a 33% relative increase from 2011. Canada fell from 2,238 in 2007 to 1,746 in 2020. In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians—a 2.5-times higher relative rate in the U.S.
What this means - more context
In 2020, the U.S. had about 118 crash deaths per million people versus 46 per million in Canada—about 72 extra deaths per million people in the U.S. The study did not report a single absolute risk increase for specific laws; it only shows associations, not proof that laws caused the difference.
To examine why U.S. traffic fatality trends diverged from Canada's from 2007 to 2021 and identify crash types and risk factors explaining the divergence.
U.S. total traffic fatalities declined from 41,259 in 2007 to 32,479 in 2011, then rose 33% relative to 43,230 in 2021. Canada declined from 2,238 in 2007 to 1,746 in 2020. After adjusting for vehicle miles traveled, the U.S. fatality rate per 100 million VMT was relatively stable, while Canada's rate consistently decreased. In 2020, about 118 per million Americans died in crashes versus 46 per million Canadians, a 2.5-times higher relative rate in the U.S.
Methods Used
Used NHTSA Fatality Analysis Reporting System and Canadian National Fatality Database data from 2007 to 2021. Analyzed total and crash-type-specific fatalities and built statistical models with travel exposure, demographic, economic, and law-related variables.
Main Finding
U.S. total fatalities fell 2007–2011 then rose 33% relative from 2011 to 2021 (32,479 to 43,230), while Canada fell from 2,238 in 2007 to 1,746 in 2020. Per 100 million VMT, the U.S. fatality rate was relatively stable and Canada's decreased. Population and VMT had positive associations with fatality counts. Canada's more common laws on driver distraction, seat belt use, and speed safety cameras may explain part of the divergence; effect sizes for these laws were not specified.
Confidence Level
Medium: national surveillance data and statistical models, but observational/ecological comparison, no effect sizes, confidence intervals, or p-values reported, so causality is not established.
Study Flags
Red Flags
- •Observational country-level comparison; cannot establish causation.
- •No effect sizes, confidence intervals, or p-values reported for law associations.
- •Canada full data only through 2020; potential confounding by unmeasured factors.
No biological mechanisms were identified in this study. This may be an epidemiological, observational, or survey-based study that reports associations rather than proposing causal biological pathways.
Surprising Findings
U.S. fatality rate per 100 million VMT was relatively stable while Canada's consistently decreased
Many assume the U.S. has worse per-mile road safety than Canada. This suggests the U.S. total death rise is mostly about more driving, not a sharp per-mile deterioration.
Practical Takeaways
When comparing road safety between places, look at fatalities per vehicle mile traveled, not just total deaths—total counts can rise simply because people drive more.
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 532 / 100
Probability of being correct
Groups of people are followed over time to see who develops an outcome. Strong for identifying risk factors and associations, but cannot prove causation as firmly as RCTs.
Human Cohort Study
Subject
Lower probability
on the GRADE evidence scale
This study looks at patterns in car crash deaths in two countries over many years. It can show that some things happen at the same time, like certain laws and fewer deaths, but it can't prove that the laws caused the fewer deaths. It's like noticing that people who carry umbrellas also get wet less, but you can't be sure the umbrella is the reason.
No conflicts of interest were detected in this study. No score impact.
Strengths
- Uses comprehensive national fatality databases (FARS and Canadian National Fatality Database)
- Covers a 15-year period (2007-2021)
- Examines multiple crash types and exposure variables
Weaknesses
- Observational design without randomization or control group
- Aggregate (ecological) data limits individual-level inference
- Potential unmeasured confounding (e.g., cultural, enforcement differences)
Methodology
Evidence Keywords
Statistical Reporting
Scoring
How strong is this study?
The study uses lots of data from entire countries over a long time, which is good for seeing big trends. But because it's not a controlled experiment, other things could be causing the differences, so we should be careful about trusting that one thing caused another.
40 / 100
- COI disclosure+40/40
- Data availabilitydata not shared
- Code availabilitycode not shared
13 / 100
- Randomizationnot randomized
- Blindingblinding unclear
- Control groupno control group
- Sample sizeno sample size reported
- Follow-up+10/10
100 / 100
0 / 100
- P-valuesno p-values reported
- Effect sizeno effect size reported
- Confidence intervalsno confidence intervals
- Pre-registrationnot pre-registered
Each component is scored out of 100 and then capped by the study design — a case series cannot reach the ceiling a randomised trial can, however well it is reported.
Where it sits
RCT reviewsReviews of RCTs (Meta-analyses)
Max 100Randomized TrialsRandomized Trials
Max 90Reviews of Cohort StudiesReviews of Cohort Studies
Max 85Cohort StudiesCohort Studies
Max 72Reviews of Case-Control StudiesReviews of Case-Control Studies
Max 63Case-Control StudiesCase-Control Studies
Max 58Cross-Sectional & Case SeriesCross-Sectional & Case Series
Max 50Expert OpinionExpert Opinion
Max 532 / 100
Probability of being correct
Groups of people are followed over time to see who develops an outcome. Strong for identifying risk factors and associations, but cannot prove causation as firmly as RCTs.
This design cannot establish causation — the findings describe an association, not a cause. This is an observational study using aggregate national data without randomization or a control group. It can identify associations and trends but cannot establish causal relationships due to potential confounding, ecological fallacy, and inability to control for all relevant variables.
No Conflicts
No conflicts of interest identified
No conflicts identified; authors declare no competing interests and no funding was reported.
Conflict Details
Traffic Injury Research Foundation: Director of Road Safety Engineering
Declaration of competing interest states no known competing financial interests or personal relationships. Acknowledgement: None. No funding statement is present. Author Craig Lyon is affiliated with the Traffic Injury Research Foundation, a non-profit road safety research organization; no industry funding or industry employment is disclosed.
Standing
The people behind it
The researchers who wrote the study this analysis is built on.
Authored by
3 researchersIf this is your work, this is how we attribute it on Fit Body Science. Craig Lyon is listed as the lead author.