Skip Out On The Elks Lodge, Die In A Traffic Accident

Well, not exactly. But new research suggests that communities with low "social capital" have 50 percent more traffic fatalities.

By Colin Lecher Posted 01.04.2013 at 2:00 pm 9 Comments

Highway 401 Traffic In Canada Benson Kua/Wikimedia Commons

Do you live around a cranky neighbor? Are you the cranky neighbor?

These are important questions because, according to a new study, areas with low "social capital"--lack of neighborliness, I guess would be one way of putting it--are associated with higher rates of traffic fatalities. Love thy neighbor, or stay off the roads.

All the usual caveats here are in full effect. Correlation doesn't mean causation (maybe people are on edge because of the higher-than-normal traffic fatalities?), but it's still an interesting find: not only were the numbers statistically significant, but the chances of dying were a full 50 percent greater in areas with low social capital.

The City College of New York study, published in the Eastern Economic Journal, looked at 10 years of data for the 48 contiguous U.S. states. The study's author, Matthew Nagler, pored over survey responses to how "honest" people in each state viewed other people, assuming that was an indicator of trust. He also created a social capital "index" based on four different community activities: election turnout, church attendance, club meeting attendance, and volunteer activity. Looking at those parameters, Nagler found the correlation between lack of social capital and higher rates of traffic fatalities. Here are the tabulated results:

Traffic Fatalities V. Social Capital: Matthew Nagler

His hypothesis on why: Because people are more courteous to each other when they have ties to the community.

Well, maybe. There was some work done to eliminate other state-by-state causes--"real gross state product per capita, vehicle miles traveled per capita, unpaved roads as a percent of local road mileage, percentage of the population over 65 and the state maximum speed limit"--but it sounds like there's a lot of potential factors at play. At least it's a good excuse to finally join that book club.

[CUNY]

Previous Article: Scientists Send A Cloud Of Atoms Plunging Below Absolute Zero

Next Article: UK Police Launch Website Asking Public To Identify Over 1,000 Found Bodies

9 Comments

Link to this comment

Robot

01/04/13 at 2:54 pm

And our government leaders live in the White House Palace of their positions, so are out of reality and compassion of their decisions or their lack their of.

Link to this comment

Tardbot

01/04/13 at 4:03 pm

Hastily considered responses.

Link to this comment

laurenra7

01/04/13 at 5:16 pm

I'm going to give you the benefit of the doubt here. The data in the tables you displayed must be from something else other than this study because if you graph the data, there is ABSOLUTELY NO CORRELATION between the "social capital index" and the "fatality rate". The only possible (however slim) correlation that can be found is when you compare the averages of the top 10 to the averages of the bottom 10. This is as useful as correlating it to how many carrots are consumed in each state. It doesn't explain why Utah, with a "social capital index" (whatever that means) of -0.2; far lower than any of the top 10 (average 4.75); has a lower accident rate than all but 3 of the top 10. In fact Utah's accident rate is also lower than the AVERAGE accident rate of the top 10 (11.25 vs. the average of 13.276). Ring up Matthew Nagler and see if he has an explanation. And ask about carrots.

Link to this comment

laurenra7

01/04/13 at 5:23 pm

Correction: Utah (and California) has an average accident rate lower than half of the top 10, also lower than the average of the top 10, and yet the "social capital index" is much lower. It goes without saying that there is no causation identifiable here. Why? Because there's no correlation either. "Social capital index," sheesh. How about a "good driver index"? That might actually show a correlation.

Link to this comment

Husbarn

01/05/13 at 3:56 pm

People die everyday in car accidents.

Link to this comment

African Rover

01/06/13 at 4:27 am

I can not see how you can draw a conclusion from that study.

Link to this comment

annjanet254

01/06/13 at 9:56 am

Carson. if you think Eric`s postlng is great... I just got a brand new Lotus Esprit after having earned $9969 this-past/four weeks and just a little over ten-k this past-month. it's by-far the most-comfortable work I've had. I started this four months/ago and almost straight away started making more than $77 per/hr. I went to this website, http://www.bit90.com

Link to this comment

Ellis_Simon

01/07/13 at 10:20 am

I wrote the news release that was the basis for this story. Laurenra7 is correct: The correlation occurs when you look at average results for the states with the highest and lowest social capital scores. You find a huge difference in highway fatality rates for incidents involving more than one vehicle or occurring a junction points. There are outliers, such as South Dakota, which has the highest social capital score and a high death rate. However, the trend emerges when you compare data across all 48 contiguous states.

Professor Nagler conclusion is that your chances of being involved in a fatal collision are greater in states where social capital is low. However, it doesn't necessarily mean people who don't vote or go to church are more likely to die. Clearly, further research is needed, and I hope to report on additional findings in the future.

Ellis Simon
Director of Public Relations
The City College of New York

Link to this comment

mnagler

01/15/13 at 10:45 pm

I am the author of the study. Laurenra7 is correct in that there is something other than the data displayed here in the study - a lot really. What I do in the study, and this might not have come through in the PopSci article, is examine the statistical relationship between two social capital measures and the fatality rate by state by year for 48 states in 10 years, so that is 480 data points. All you see here is 20 of them. Also the study included a long list of variables as "controls." So while you might not see a visible correlation when looking at some of the data, there is a statistically valid relationship accounting for control factors, etc.

I'd also like to point out that this study was specifically designed to make causal claims - not mere correlation - so in that sense the PopSci article is incorrect. I have used a statistical instrument to create what is often called a "natural experiment" in the variation of social capital due to an exogenous factor. Thus I am able to observe, as statistically significant, that social capital at a state level "causes" highway safety (again, at the state level). This is not necessarily true at the INDIVIDUAL level, of course, as one might think I am implying, based on the PopSci article title!

I invite you to read the full article I have written. Google my name, go to my website, and you can download it (the pre-published version). Or read the published version on the Eastern Economic Journal's website.

Cheers, and safe driving!
Matthew G. Nagler
Associate Professor of Economics
The City College of New York