Tag Archives: states

Do Americans Vote With Their Wallets? State Unemployment Rates and Support for the Incumbent

Americans vote with their wallets, or at least that’s the conventional wisdom. Economic issues are supposed to be of primary concern to American voters. Or, as Bill Clinton’s presidential campaign famously put it in 1992, “It’s the Economy, Stupid.”

I avoid politics here – it’s not that kind of blog – but sometimes it’s impossible to separate economy from political economy, particularly in an election year. So I decided to take a look at this piece of conventional wisdom. In a non-partisan sort of way.

If the economy really is key to American voters, then one might expect that – all else being equal – states with high unemployment rates should be less likely to support the incumbent (Obama), and states with low unemployment rates should be more like to support him.

Put another way, there should be a negative correlation between support for the incumbent and state unemployment rate. A graph of states with support for Obama on one axis and state unemployment rate on the other should tend to slope downward.


The economists’ union requires that I preface this with caveats: There are various technical issues with analyzing aggregate data this way, and this is not intended to be a scientific study. That said, back-of-the-cocktail-napkin analyses can still be interesting and enlightening.


Correlations are measures of how much two sets of numbers tend to move in the same or opposite directions. A correlation of 0 means there’s no relationship, close to -1 means the two move in opposite directions, and close to 1 means there’s a strong positive relationship. In this case, we’re expecting a negative number.

The correlation between state support for Obama and state unemployment rate actually turns out to be positive 0.27. This means that – all else being equal – the higher the unemployment rate in a given state, the higher the support for President Obama. The opposite of what we’d expect, if economic issues are driving voter preferences.

(Note: I used FiveThirtyEight’s unadjusted average poll numbers by state, simply because they were available in an easy-to-use format. The numbers were as of 3:00 pm on 28 October. The unemployment rates are those for September 2012, from the Bureau of Labor Statistics, bls.gov.)


Correlations are rather unexciting to many people. After all, they’re just numbers. A mentor of mine used to recommend an alternative technique that he called, “Look at your data”. So here’s a – yes, I am an economist – graph of state preference for Obama versus state unemployment rate.

As suggested by the positive correlation, the data points do tend to slope upward. The best fitting line is shown, and it has a positive slope too.

States such as Rhode Island, California, New Jersey, and Nevada have relatively high unemployment rates and yet have relatively large percentages supporting Obama. States such as North Dakota, Nebraska, and Oklahoma have low unemployment rate yet have smaller percentages supporting Obama.


What gives? Is the conventional wisdom about Americans voting with their wallets wrong? Is it not the economy, stupid? Does high unemployment actually cause people to prefer the incumbent?

That’s hard to say. Political preferences are complex, and many factors combine and interact to determine an individual’s political views. It gets even more complicated when you consider things at the state level.

It’s possible that higher unemployment rates really do cause people to support this incumbant. Perhaps people who are unemployed are more likely to prefer Obama because they believe, for whatever reason, that they personally will be better off if he is re-elected. On the other hand, it’s also possible that other factors are affecting both unemployment rates and political preferences, making it appear that higher unemployment rates are causing people to support Obama.

Whatever the details, it’s clear that saying money is all that matters to American voters vote is too simplistic. Maybe Americans do vote with their wallets, but that could mean voting for the candidate they believe is most likely to fill them.

More on this Topic: Voting With Your Wallet II: Education, Liberalism, and Unemployment