The Thing about Surveys

Everyone knows that there are lies, damned lies, and then there are statistics.  Nowhere is this more obvious than when interpreting survey results.

In a March essay on America's political polarization in The New Republic, Bill Galston noted:

above and beyond their ideological disagreements, conservatives and liberals have come to understand the practice of politics differently. In a survey taken right after the Republican sweep in the 2010 midterm elections, 47 percent of American said that it was more important to compromise in order to get things done, versus 27 percent who thought it was more important for leaders to stick to their beliefs even if little got done. Liberal Democrats weighed in on the side of compromise, 58 to 16, moderate Democrats by 64 to 17. But conservative Republicans (the overwhelming majority of their party) favored sticking to their beliefs by 45 to 26. Ten months later, after the debt ceiling fiasco, an outright majority of adults favored compromise, including 62 percent of Democrats and 63 percent of liberals. But pluralities of Republicans and conservatives continued to favor leaders who stuck to their beliefs.

Republicans disdain compromise more than Democrats.  This is not surprising to anyone who has any familiarity with American politics.  There is, for example, no corresponding term for RINO on the left.  Republicans are much more concerned with ideological purity than Democrats.  Republicans tend to prize clarity and certainty more than Democrats.

But while a full 20% more Americans thought it was more important to compromise to get things done (a testament to Americans' once vaunted pragmatism), if you actually go to the USA Today/Gallup Poll survey, you find that 49% of these same respondents wanted Congress to have more influence over the direction of the country, while 41% wanted President Obama to have more influence over the direction of the country.  Now, the sampling error is +/-4%, which means this is almost a statistical dead heat, but that's not the point.  The point is that while Americans want our politicians to compromise, we also want to give more influence to the party that we ourselves perceive to be less inclined to compromise!  Galston's omission of this detail in his description of the survey results is misleading.  I know that politicans pick and choose the facts they use to score political points, but the aims of scholars should be different, even when they put on the pundit's hat.

Now the survey results do not necesarily mean that Americans don't know what they want, or want mutually incompatible things.  They could be the result of careful consideration of competing political values such as the value of the spirit of compromise as a political virtue against the prerogatives of institutional authority and separation of powers concerns.  I leave it to you to decide which is more likely.