Are Blacks Really 3.5x Worse Drivers Than Whites?
Print Friendly and PDF

Are blacks really 3.5 times more likely than whites to have to go to the Emergency Room for motor vehicle crashes (including pedestrians being hit)? Are Hispanics 12% less likely to win up in the ER for car accidents? A couple of generations ago, Latinos were notoriously bad drivers. Have they really improved that much?

This graph is from a new CDC report:

Emergency Department Visit Rates for Motor Vehicle Crashes by Selected Characteristics: United States, 2017–2018

Danielle Davis, M.P.H., and Christopher Cairns, M.P.H.

Data from the National Hospital Ambulatory Medical Care Survey

Key findings:

● In 2017–2018, the overall emergency department (ED) visit rate for motor vehicle crash injuries was 5.3 visits per 1,000 persons.

● The ED visit rate was highest among patients aged 15–24 (9.1) and then declined with age.

● The ED visit rate for non-Hispanic black patients (15.2) was higher than for non-Hispanic white (4.3) and Hispanic (3.8) patients.

● ED visit rates for patients who had Medicaid, no insurance, or workers’ compensation insurance as their primary expected source of payment were higher than for patients who had private insurance or Medicare.

● The ED visit rate for motor vehicle crash injuries at hospitals located in the South (6.8 visits per 1,000 persons) was higher than for rates at hospitals in all other census regions of the United States.

In 2017–2018, an average of 3.4 million emergency department (ED) visits
for motor vehicle crash injuries occurred annually.

Okay, so if there 330 million residents in the US, that’s 1 vehicle crash emergency room visit per 100 residents per year or a little over 10 per 1000. But the summary talking points above say 5.3 per 1000 or half what the body text says.

Most persons injured or killed in motor vehicle crashes are occupants (3). Medical care costs and productivity losses associated with crash injuries and deaths exceeded $75 billion in 2017. …

Car crashes are bad.

● The overall ED visit rate for non-Hispanic black patients (15.2 visits per 1,000 persons) was
higher than the visit rate for non-Hispanic white (4.3) and Hispanic (3.8) patients.

● The ED visit rate for non-Hispanic black patients was higher than the visit rate for
non-Hispanic white and Hispanic patients in all age groups, particularly among patients
aged 25–44 (24.6 among non-Hispanic black patients compared with 5.7 among
non-Hispanic white and 4.2 among Hispanic patients).

● For persons aged 15–24, the ED visit rate for non-Hispanic white patients (9.2) was higher
than the visit rate for Hispanic patients (5.1).

Recently, my Taki’s Magazine column pointed out that the National Highway Traffic Safety Administration latest reported on traffic fatalities reported that black road deaths were a remarkable 36% higher during the Racial Reckoning (the last seven months of 2020) versus the same time time period in 2019: we can call this the Racial Wreckening.

But, in general, blacks don’t seem to get killed in car crashes 3.5 times as much as whites do. I pointed out:

Note that blacks are not terrible drivers. While blacks [at 1/8th of the U.S. population] were 55.9 percent of known murder offenders in 2019’s FBI crime statistics and 53.2 of homicide victims, they were a much more reasonable 16.9 percent of traffic deaths in 2019. So, blacks are somewhat more likely to be killed in car crashes, but the proportion is nowhere near as absurd as with murder. If blacks could get their share of America’s murders down to 16.9 percent, this country would be a racial utopia.

(Two methodological asides: While we can usually distinguish between perpetrator and victim in homicides, although they don’t differ much racially, guilt is murkier in traffic statistics. Also, like most police reports, the car-crash statistics don’t do a good job of distinguishing between whites and Hispanics, so in my graph I simply lumped the rest of the U.S. population together into “nonblacks.”)

But then the black share of road fatalities rose to 19.7 percent during the racial reckoning of June–December 2020: still not as awful as black gun violence, but moving in the wrong direction.

So, I’m not sure that I trust this new report. It could be that blacks do go to the hospital for car crashes a huge amount but don’t get killed all that much more, or at least not until George Floyd’s death liberated them to drive really, really bad.

More likely, this CDC report just isn’t that trustworthy for boring methodological reasons. Looking at the fine print, I see:

NOTES: Based on a sample of 917 emergency department (ED) visits for motor vehicle crashes, representing an annual average of 3.4 million ED visits.

A sample size of 917 incidents isn’t tiny, but it’s not big. (A typical Presidential election poll uses perhaps 1,400 respondents.) Dividing the sample into 3 races (plus miscellaneous) times 4 age groups stresses reliability further. And there’s the big question of how nationally representative this sample is.

Injured persons included motor vehicle occupants, motorcyclists, pedal cyclists, and pedestrians.

By the way, the Daily Mail article misinterpreted the fine print in this report, claiming, “Race was only known for 17 percent of patients and ethnicity was only known for 19 percent of patients.” No, the DM got it backwards:

Overall for 2017 and 2018, race data were missing for 17.1% and ethnicity data were missing for 19.2% of ED visits, and race was imputed for missing records.

Non-Hispanic other represents 3.1% of weighted visits. These visits are included in the total but not reported separately.

Anyway, the fact that I dug up a couple of weeks ago, that black road deaths were up 36% during the Racial Reckoning compared to 9% for everybody else is one of the more striking statistics from a year, 2020, full of unexpected numbers.

[Comment at]

Print Friendly and PDF