Preventable road deaths in 72 countries, 2021

An analysis of 2021 data from 72 countries reveals that while alcohol laws showed no significant impact, adopting improved speeding laws and vehicle safety standards could potentially reduce road deaths by approximately 23% and 55%, respectively.

The Gist

Imagine the world's roads as a giant, chaotic game of "Risk," where the goal is to keep everyone safe. A researcher named Leon Robertson decided to look at the rulebooks of 72 different countries to see which rules actually stop people from dying in car crashes. He wanted to know: If countries that are currently playing with a weak rulebook switched to the "Gold Standard" rules recommended by the World Health Organization, how many lives could be saved?

Here is the breakdown of his findings, using some everyday comparisons:

The Three Main "Safety Shields"

The study looked at three specific types of rules, like three different shields a country can hold up against danger:

1. The Speed Limit Shield (Speeding Laws):

   • The Finding: This shield works very well. The study suggests that if countries with weak speed laws adopted the best ones, road deaths would drop by about 23%.
   • The Analogy: Think of speeding like driving a race car on a playground. If you slow down to a safe walking pace, you have much more time to react if a child runs out in front of you. The data shows that countries with strict speed limits are like playgrounds with clear "slow down" signs, and fewer people get hurt there.

2. The Car Armor Shield (Vehicle Safety Standards):

   • The Finding: This is the most powerful shield of all. If countries adopted the highest safety standards for the cars they sell (both new and used), deaths would plummet by about 55%.
   • The Analogy: Imagine buying a helmet. Some helmets are made of hard plastic and protect your head; others are made of flimsy cardboard. The study found that many low-income countries are being sold "cardboard helmets" (cars without safety features like airbags or strong frames) because manufacturers are shipping their older, less safe models there. If these countries switched to "hard plastic helmets" (cars that meet strict safety standards), they would cut the death toll in half.

3. The Sobriety Shield (Alcohol Laws):

   • The Finding: This one is tricky. The study found that having a law against drunk driving didn't show a statistically significant drop in deaths in this specific analysis.
   • The Analogy: Think of a bouncer at a club. A sign saying "No Drunks Allowed" is good, but if the bouncer doesn't actually check IDs or stop people from entering, the sign doesn't help much. The author notes that it's hard for police to catch drunk drivers without stopping them, and sometimes the laws don't even cover drunk pedestrians. So, while the rule exists on paper, it didn't seem to move the needle as much as the other two shields in this study.

The Weather and Crowd Factors

The researcher also looked at two other things that act like the "terrain" of the game:

• Temperature: Warmer countries tend to have more deaths. It's like how people are more likely to go out and play in the sun than in the freezing cold. More people on the road in warm weather means more chances for accidents.
• Population Density: Surprisingly, countries with more people packed into a small space (like a crowded city) actually had fewer deaths per person. It's like a crowded dance floor where everyone is moving slowly and carefully, versus an empty highway where people feel free to speed up and take risks.

The Big Takeaway

The study concludes that the biggest "leaks" in our safety boat are unsafe cars and speeding.

• The Car Problem: Many countries are buying cars that are essentially "time bombs" because they lack safety features. Fixing this is the single most effective way to save lives.
• The Speed Problem: Slowing down works wonders.
• The Data Problem: The author suggests that to fix things even faster, we need better maps. Instead of just knowing how many people died, we need to know exactly where and how (e.g., "pedestrians hit by cars at night" or "cars hitting animals"). If we know exactly where the danger spots are, we can put up a specific "stop sign" or "light" right there, rather than guessing.

In short: If countries swapped their "cardboard helmets" for "hard plastic ones" and enforced speed limits, they could potentially save more than half of the people who currently die on the roads. The rules to save lives exist; the challenge is simply putting them into practice.

Technical Summary

Technical Summary: Preventable Road Deaths in 72 Countries, 2021

Problem Statement
Despite decades of established research confirming the efficacy of specific road safety countermeasures, many countries have failed to adopt World Health Organization (WHO) recommendations regarding speeding laws, alcohol regulations, and vehicle safety standards. While global road death rates per 100,000 vehicles declined between 2010 and 2020, rates increased in 2021, with lower-income countries experiencing substantially higher per capita death rates. A critical issue identified is the disparity in vehicle safety: major manufacturers often sell new vehicles in low- and middle-income countries that do not meet the safety requirements of their home markets, and high-income countries export older, non-compliant vehicles to these regions. Furthermore, ambient temperature and population density are known confounding factors associated with road mortality that must be accounted for to accurately assess the impact of policy interventions.

Methodology
The study analyzed data from 72 countries for the year 2021, representing a total population of 5.4 billion and 779,161 road deaths. The research objective was to estimate the potential reduction in deaths if countries currently neglecting WHO recommendations had fully complied with them, while statistically controlling for average ambient temperature and population density.

• Data Sources: Road death rates were sourced from World Population Review. Countermeasure status (alcohol laws, speeding laws, vehicle safety standards) was derived from WHO assessments and Google searches. Environmental and demographic data (average temperature, population density, purchasing power) were obtained from global data repositories.
• Variable Coding: Alcohol and speeding laws were binary (1 if meeting WHO criteria, 0 otherwise). Vehicle safety standards were quantified as an index from 0 to 8 based on the number of adopted standards.
• Statistical Approach: Unlike traditional Poisson or Negative Binomial models that use population or vehicle counts as offsets, this study employed a least squares regression model. This decision was made to prevent the results from being unduly influenced by the massive populations of China and India, which constitute 52% of the total population of the 72 countries studied.
• Model Specification: The dependent variable was the logarithm of road deaths per 100,000 population. The independent variables included alcohol laws, speeding laws, vehicle safety standards, the logarithm of average temperature, and the logarithm of population density. A purchasing power index was initially considered but excluded due to high multicollinearity with other variables.
• Age Standardization: The study attempted age-standardization but found that differences in age distributions (specifically the proportion of children under 10) varied too widely between nations to produce a model with a good goodness-of-fit compared to the ordinary least-squares regression of logged unadjusted rates.

Key Results
The regression analysis yielded an R-squared value of 0.64. The model coefficients and odds ratios indicated the following associations:

• Speeding Laws: Improved speeding laws were significantly associated with lower death rates. The regression coefficient was -0.267 (Odds Ratio: 0.766).
• Vehicle Safety Standards: A higher number of adopted vehicle safety standards was strongly associated with reduced mortality. The coefficient was -0.095 per standard (Odds Ratio: 0.909).
• Alcohol Laws: While the coefficient was negative (-0.049), it was not statistically significant at the 95% confidence interval (Odds Ratio: 0.952, CI: 0.743–1.220).
• Environmental Factors: Higher average temperatures were associated with increased deaths (Coefficient: 0.483), while higher population density was associated with decreased deaths (Coefficient: -0.141).

Quantitative Estimates of Preventable Deaths
By applying the regression coefficients to countries that had not fully adopted specific countermeasures, the study estimated the following potential reductions in road deaths:

• Speeding Laws: If countries with suboptimal speeding laws adopted WHO-recommended standards, road deaths would likely decline by approximately 23.4%.
• Vehicle Safety Standards: If the 44 countries with fewer than eight safety standards adopted the full set of eight, road deaths would be approximately 54.8% lower.
• Temperature Impact: A one-degree Celsius increase in average temperature in these countries is predicted to result in a 2.7% increase in deaths.

The model demonstrated a good fit, with a predicted aggregate death count showing only a 3.9% overprediction compared to actual data.

Significance and Claims
The paper asserts that the failure to adopt vehicle crash avoidance and crashworthiness standards allows manufacturers to sell vehicles with inferior protection, particularly in low-income regions. The study concludes that adopting vehicle safety standards and improving speeding laws are likely to significantly lower death rates.

The authors highlight that while alcohol laws are a known factor, their statistical impact in this model was not significant, potentially due to enforcement challenges (requiring stops and testing) and the inclusion of inebriated pedestrians who are not covered by driver-focused laws. Conversely, vehicle safety standards and speeding laws offer immediate and substantial potential for mortality reduction.

The paper emphasizes that while replacing unsafe vehicle fleets takes time, targeted interventions based on specific collision clusters (e.g., pedestrian strikes in the dark, animal collisions, fixed object impacts) can yield immediate results. The authors argue that better data collection regarding specific collision types—facilitated by local public health workers and police using pin maps—can guide the adoption of effective countermeasures such as improved lighting, livestock removal, and energy-absorbing barriers. The study suggests that these data-supported measures are more likely to secure funding from charitable organizations and enhance economic growth by reducing the costs of injuries and deaths.