Connecticut Men, Deadly on the Road

Gender Disparities in Fatal Crashes in Connecticut, 2003-2017

Author: Ilya Ilyankou, Transport Hartford Ambassador

Connecticut male drivers have a fatality rate that is four times that of women. This is much higher than the 2.4 ratio at the national level. What’s up with that?

Nationally, many more men die in motor vehicle crashes than women. According to the Highway Loss Data Institute, which analyzed traffic data by the US Department of Transportation, in 2017, there were 26,380 male deaths and 10,697 female deaths associated with motor vehicle crashes. For vehicle driver fatalities, the numbers were smaller, yet the disparity was just as pronounced with 12,500 male deaths, and 5,152 female deaths. Male drivers in the US die at 2.4 times the rate of women drivers.

This Connecticut driver smashed into the Supreme Court monument.

We used UCONN’s Connecticut’s Crash Data Repository to analyze traffic deaths in the state by gender, age, and DUI/speeding status of drivers to understand who are the people who die on Connecticut roads. This way, we know more about crash causes and can utilize that knowledge to decrease the number of traffic-related deaths statewide. You can view and download the raw data used for this analysis here.

To create visualizations, we looked at the fatal crashes in Connecticut between 2003 and 2017 and looked at the gender and age of those that died along with other factors.

[The charts are interactive. You can turn categories on/off by clicking on them. You can also hover over columns to get exact numbers.]

For nearly all ages, there are many more males than females who die in crashes. The peak of fatalities happens for ages 21-22 for males and 23-24 for females. In general, the peak for male fatalities happens in late teens/early 20s, and in late 40s/early 50s.

Note – ‘Unknown Gender’ primarily refers to passengers who were killed in crashes between 2003-2014. The gender of passenger fatalities was not reliably captured in this dataset during that time period. The 2015-2017 data that captured the gender of passengers killed did not reveal any disparities between genders of killed passengers. In other words, as many male passengers die in crashes as female passengers.

Now, let’s take a closer look at male crash fatalities. How were they getting around?

The vast majority of male crash fatalities were car drivers. Note, though, that passengers who were killed between 2003-2014 are not represented in this chart. The youngest driver fatality was only 7 years old. From ages 16 to 18, the death rate of male drivers nearly quadruples  from 16 to 56 deaths. At age 22, the driver fatalities count goes up to 97, only to drop by 30% at the age of 23.

Although we only have 3 years of consistently ‘gendered’ passenger data for 2015 through 2017, we can still see that the distribution of male passenger deaths is fairly flat among all age groups with a slight increase in early 20s, except for the age of 20, when the spike shows 9 male passengers killed.

One might assume that more pedestrians are killed when very young and very old, but the data does not show that. The distribution is fairly even among all age groups, with the peak at the age of 20 (13 fatalities) and 21 (12 fatalities).

When it comes to bicyclists, between 2003 and 2017 there were 12 fatalities, and all of them were male. Ten of twelve were males aged 42 and above.

Now, let’s take a look at female crash fatalities. How were they getting around?

Once again, remember that passenger gender was not tracked for years 2003-2014, so only those passengers killed in 2015-2017 are displayed.

First, notice that no female bicyclists died in 2003-2017. Excellent cycling infrastructure or simply no cycling infrastructure and few women cyclists?  Maybe women bicycle riders are very skilled?

Second, the proportion of passengers (green) and pedestrians (purple) is much more visible among females. At the age of 58, for example, there were 8 female drivers and 9 female pedestrian fatalities (47 and 53%, respectively). For 58-year-old males, the respective numbers were 29 drivers and 9 pedestrians (or 76% and 24%). This approximate ratio holds for most ages.

Third, the values themselves. At the age of 20, for 23 female fatalities there were 83 male fatalities. This is 3.6 times more male fatalities. Of those 20-year-olds, 60 were male drivers, and only 17 were female drivers (3.5 times). At the peak driver fatality age of 24, there are 23 drivers who are females. For male’s peak driver fatalities age of 22, there were 97 driver fatalities. The peak fatality age has  4.2 times more male fatalities than female.

There is roughly the same number of males and females in Connecticut. While it is true that males on average drive more than females by about 50% (according to Michael Sivak’s study at University of Michigan) – and we have no reason to believe that in Connecticut this figure is drastically different. Men account for 80% of driver fatalities, 64% of pedestrian fatalities, and 100% of the bicycle rider fatalities.

Comparing % of Male vs Female Crash Fatalities for Drivers and Pedestrians

Some may suggest that men are more likely to drive while drunk and choose to drive at higher speeds, so let’s take a look if more male drivers were intoxicated at the time of crash and if crashes were classified by the police as speeding-related.

For ages 19 to about 50, between a third to about half of all male drivers were under the influence at the time of crash. For the age of 22 (the peak death age for male drivers), there were 50 non-DUI and 47 DUI fatalities.  48% of 22-year-old male drivers who died were under the influence.

For female drivers at the peak death age of 24, 48% (11 of 23) of them were under the influence, a percentage of female drivers who were tested positively for DUI for most ages was smaller than that of males. Overall, 29.8% of female and 37.3% of male driver fatalities were driving under the influence.

Now, let’s take a look at speeding-related crashes that killed drivers.

Between 20% to 30% of male driver fatalities in their 20’s are speeding-related. The number of speeding-related fatal crashes falls as males grow older. For females, the number of speeding-related accidents are fairly low, and younger female drivers are more likely to die in a speeding-related fatal crash than older female drivers. Overall, 13.2% of female and 18.3% of male driver fatalities were speed related.

Due to the way the data is recorded in the database, we cannot conclude that the driver who was speeding or who was driving under the influence necessarily caused the crash. The correct interpretation of the data provided in these charts would be that “the driver who died in a fatal crash tested positive for alcohol/drugs”, and “the driver who died in a fatal crash was speeding.”


In Connecticut, males die in fatal motor vehicle crashes disproportionately more often than females as both drivers and pedestrians, even accounting for the fact that on average males travel around 50% more miles than females. The Connecticut gender disparity in crash fatalities is similar to the national trend.

The early 20s are the most dangerous ages for both male and female drivers, with a spike of alcohol and drug-related fatalities especially for males.  Parents may want to strongly reconsider buying their young adults in Connecticut that first car, especially those with sons. Bus transit is much safer.


What should we do with this interesting and powerful data?  Who are you going to share this with? Who should we share this with?  Send your comments, questions, and suggestions to Tony Cherolis at the Transport Hartford Academy (transporthartford@ctprf.org).

You can also get involved in the Transport Hartford – Talking Transportation Facebook discussion group.


SourceUConn Connecticut Crash Data Repository

Source Spreadsheet with CT Crash Fatality Data, Years 2003 to 2017

  1. These visualizations are powered exclusively by UCONN Crash Data.
  2. Data was deduplicated and then grouped by gender and age.
  3. Unknown gender is mostly passengers.