Network News

X My Profile
View More Activity

Research desk responds: How does unemployment affect voter turnout?

By Dylan Matthews

_SP_ asks:

Is there a correlation between unemployment rate and turnout? As we know, for some dumb reason election day is still a weekday that most people have to work which reduces turnout, so if there are a lot of people without jobs, can we expect them to vote in greater numbers (leaving aside how much they'd favor one party or the other.)

_SP_'s hypothesis is interesting, but isn't held up by the data. The Census recently issued a detailed
(PDF) breaking down voting and registration in the 2008 election by race, income, employment status and so forth, and found that employed people turned out more frequently (65.9 percent to 54.7 percent) than unemployed people:


The registration data also defies _SP_'s theory. The gap between employed and unemployed registration rates -- 72.8 percent to 64.1 percent -- is smaller than the gap between actual voting rates; 90.5 percent of employed people who had registered voted, while only 85.3 percent of unemployed who registered did. If _SP_ were correct, then unemployment should cause registered voters to turn out more often, since they have more time, whereas the data suggest the opposite effect, if anything.

The root of the gap appears to be in the fact that high earners turn out much more often than working-class people, from whose ranks the unemployed disproportionately come. Here's how voter turnout in 2008 broke down by annual family income:


The gap between turnout of people making less than $20,000 (51.9 percent) and those making more than $100,000 (91.8 percent) is staggering. The mechanisms involved here are numerous; as one example, the poor have less voting information. But the factor _SP_ focuses on -- time -- is probably the most important. The term "leisure class" does not exist for nothing; one thing money buys is time not spent worrying about its acquisition. That time, in this case, appears to translate into greater political involvement.

By Ezra Klein  |  July 23, 2010; 3:58 PM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Sen. Kent Conrad: 'The immediate threat is not the debt. It’s weak aggregate demand.'
Next: Shellenberger: 'This is the end of cap-and-trade for a long time'


Thanks for taking the question (incidentally, it's only _SP_ because the Post doesn't let you do 2 letter handles.) I hope they do a similar analysis this year because I wonder if the pattern will change in a recession, when more people who were formerly in high earning classes are now in lower classes. November 2008 was 6.7% which was considered unacceptable then and would be considered awesome now. They don't strip your registration when you become unemployed (although isn't that part of the Tea Party platform?) so at least the registration numbers should converge in a recession, and it will be interesting to see the turnout numbers.

Posted by: _SP_ | July 23, 2010 4:18 PM | Report abuse

I don't know if you could find a data set that would allow you to look at this, but i would wonder if there is a correlation one income levels are controlled for.

Posted by: ThePedro | July 23, 2010 4:33 PM | Report abuse

And of course, working poor, working class, and many others usually don't get time off from work to vote. Before and after work, many people are doing child care, cooking breakfast or dinner, going to that second job, etc. Why we don't have a national day off for voting boggles the mind. I would gladly give up 4th of july and celebrate it on voting day. I am sure the founders would approve.

Posted by: srw3 | July 23, 2010 4:47 PM | Report abuse

As expected. Just another factor in the equation why the unemployments benefit extension was in limbo after expiring.

What would be enlightening is slicing and dicing the data even further by age bracket. Anecdotally, everyone 'knows' the retired vote in larger numbers than younger voters.

Posted by: tuber | July 23, 2010 5:59 PM | Report abuse

Hmm. Seems like those graphs show there is a negative correlation between being unemployed and voting. But they don't really say whether an uptick in the unemployment rate can be expected to raise or lower voter turnout, do they? Maybe if you controlled for other attributes correlated with employment status, you'd find that unemployment leads to more voting because they don't have to leave work to vote. Or maybe you'd find it leads to less voting because unemployed people become depressed or something.

Posted by: yannigluthra | July 24, 2010 3:05 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company