Why Don't Democrats Run On Soaking The Rich?

Political scientist Jonathan Bernstein wonders. Andrew Sullivan offers the typical conservative viewpoint on why this approach doesn't resonate with voters - that "it's just not American to bash the successful". There may be something to this - the American political culture does appear to me more hostile to redistributive schemes than those of, say, Europe - but I doubt that it's the only reason, or even the most important one. Rather, I think that the seeming contradiction here - polls showing high support for taxing the rich more heavily, with actual voter behavior not reflecting that preference - arises from 1.)a misreading of what the polling results mean, 2.)a failure to account for the difference between stated preferences (poll responses) and revealed ones (actual voting), and 3.)a failure to attend to the fact that many polls (and particularly those which pose questions on general policy preferences such as these) do not use samples that are representative of the portion of the population that actually votes.

Firstly, I think it's likely that a very large number of voters do not interpret the phrase "tax the rich" the same way the Democrats do. To the Democrats, "the rich" has a statistical definition - it means "people in the top 20% of earners" or "people who make more than $250,000 a year" or whatever. I suspect, however, that to the majority of voters, even well-off voters, "the rich" means "people who make more money than I do". It's a well-known fact that a most Americans - even those who by any objective socioeconomic measure would have to be considered rich - consider themselves some variety of "middle class". Ask white-collar professionals making $200,000 a year (particularly those living in an expensive city like New York) whom they think of as "rich", and they'll probably name someone like Warren Buffett or Bill Gates. They're unlikely to think of themselves. So when they hear the question "should we tax the rich more heavily?", they are very likely to interpret it as "should we raise Warren Buffett's taxes?", and unsurprisingly, many are in favor. I expect, however, that if you phrased the question differently, more along the lines of specific proposals of the sort Democrats like to make - say, "should we raise taxes on households that make more than $250,000 a year?" - you'd get much less enthusiastic support.

Even if we grant that voters are not interpreting the question differently, and that the type of voters I mentioned in the previous paragraph realize that the tax hikes in question will apply to them as well as to Warren Buffett, we have to consider the possibility that many of them may be more willing to support tax hikes in theory than they are in practice. It's not exactly news that people frequently say one thing and do another, particularly when there is social pressure to appear a certain way involved. Respondees who want to appear conscientious and socially responsible might very well say yes, I would pay higher taxes if it meant better government services for those less well-off than I am, or what have you. Whether such people are still willing to say that when the taxman is actually at the door is a test of conviction that I suspect not all of them will pass. Anyone who wants to can pay higher taxes doesn't need an act of Congress to do so - the fact that few people choose to pay more than they have to makes me thinks there's more than a bit of empty self-administered back-patting going on among these poll respondents.

Finally, there's the fact that poll results on a question like this may not and quite likely do not represent the opinions of the people that actually show up on election day. General preference polling like this often samples registered voters, who as a group are distinctly more liberal than likely voters, i.e., the people who generally turn up on election day to vote. There is also a strong positive correlation between personal wealth and regular voting, and as such the wealthy are likely to comprise a disproportionately heavy portion of any given electorate compared to their numbers in the population as a whole. As a result, politicians who run on raising taxes on the rich do not garner as many votes because of that position as polling might lead one to believe.

This is essentially the same conundrum that the Democrats faced on healthcare. Why, they wondered, was the Affordable Care Act so unpopular, given that the more ambiguous notion of "health care reform" polled well? It's because, as is always the case with public policy, the devil is in the details. Any policy is going to poll better when described vaguely and abstractly than it will when it's described in detail, because when you get into detail it becomes clear whose ox is being gored and opposition starts to coalesce. This is why politicians are so fond of campaigning in airy generalities and so averse to substance, and it's why adopting policy positions based solely on what the public tells pollsters (something Democrats are all too prone to do) is a stupid idea.