With many issues, despite strong opinions, voters are short on the details.
For example, in a recent posting, Larry Bartels related survey evidence showing that there is a wide gap in voter enthusiasm between voters that favor renewing the Bush Tax Cuts and those that favor letting the tax cuts expire.
The hot button issue with the tax cuts is whether to renew the cuts for families earning more than $250,000 a year. The wrangling among politicians over this issue seems to mostly involve whether or not earning that amount of money qualifies somebody as wealthy.
What's amazing about the magic number of $250,000 is that, based on responses to a recent YouGov/Polimetrix poll, by and large, Americans have a very distorted view of how many people make that much money.
Any idea what proportion of American families make more than $250,000 a year? Or, to potentially make it easier, any idea what proportion of families in your state make more than $250,000 a year?
Don't feel bad if you don't know—most people don't. The actual number, nationwide is somewhere less than 3% of families earn more than $250,000 a year. What did the survey respondents say when asked this question? The average response was close to 17%!—meaning your typical survey respondent thinks that almost 1 in 5 families in America earn that kind of money, when the answer is closer to 1 in 50!
They were actually asked if they knew what proportion of people in their own state made more than $250,000—but that didn't bring many people closer to reality: there are only a few states where a guess of 17% comes close to being right—meaning that 17% is only about twice the actual number. These states are Maryland, Massachusetts, New Jersey, New York, and Virginia—and, perhaps ironically or appropriately, depending on who you ask, Washington, DC, at 15%. The people in these states were not as wrong as respondents in other states, but they still adjusted their answers up to about 21%.
However, it is not just that Americans are looking around at their own friends or neighborhood and making a guess from their peers: people making more than $150,000 were actually closer to being right, with an average guess of a little over 11%, than were people making less than $30,000, who, on average, guessed that the proportion making more than $250,000 is a whopping 21%. So people making less money actually believe that there are more wealthy people out there than wealthy people do.
This might lead you to ask if responses to this question have something to do with how educated a person is—and, yes, they do, but even among college graduates the average answer was 12%—or more than 4 times the correct answer. The most educated—those with post graduate degrees—do better, but not by much.
Did anybody get close to the truth? Well, of all survey respondents, a little more than 15% knew that the correct answer is less than 3% of families make more than $250,000. However, that includes about 5% of respondents that said the answer was 0%, meaning that absolutely nobody makes $250,000— which would make the entire debate about extending tax cuts to people making more than $250,000 seem a little pointless.
Is there a political element to all this? Do Republicans, who, more than Democrats, tend to favor leaving the tax cuts in place have a more distorted view of how many people are making that much money? No, not really, the average Republican gave pretty much the same answer as the average Democrat.
However, there does seem to be some political consequence: in statistical analysis that controls for other influences, among Independent voters, the more distorted their view of how many people make more than $250,000, the lower their opinion of President Obama and the more likely they are to say they will vote for a Republican for Congress next week. So, a person that says 20% of people make $250,000 is more likely to vote Republican than a person that says 5% of people make $250,000.
A note on my motivation for this post. I think it is good to be wary of claims about levels of ignorance in the American public. For example the oft repeated stories about some poll showing that Americans can't find the United States on a world map, or that school children can't recognize the Vice President, etc.. My guess is that most of these stories are apocryphal or based on questionable data sources. And I think these stories are also often employed to make political points, or to show that "things ain't what they used to be", or to make educated persons feel better about themselves. However, with careful consideration, a person might ask what reason an individual child has for knowing who the Vice President is—which isn't necessarily obvious—or why we might expect there to be accurate responses to what proportion of people make over $250,000, because there are many reasons to expect reasonable people to be wrong about this. However, the level of distortion on political issues can be consequential—for example, a tax cut that effects 1 in 5 or 1 in 6 families is very different than a tax cut that effects 1 in 50—and if these beliefs are affecting vote choice, then the public's distorted views are important.