I know this is an academic blog (as opposed to a “mixed” blog). Thus, we generally don’t discuss politics. But several weeks ago, I was party to some lively dinner conversation that raised an interesting issue of the limitations of empirical research for political consulting. So, despite the political context, this is a serious—very “serious”—blog post, assuming there is such a thing. Further, it is probably right in the wheelhouse of some political scientists out there who specialize in voting and election strategy.
Here is my question: Is it possible that the constructs of leadership or sincerity, which cannot be easily measured, are part of the error term in the voting models used by political consultants? In other words, does the conventional political wisdom rest upon a series of mis-specified regressions? The context of my dinner conversation might throw this issue into sharper relief.
My companions included an economist, a historian, and two political scientists.
Someone asked whether the Democrats could capitalize on the country’s political malaise (e.g., Iraq, deficit spending, healthcare woes, etc). The economist—who is hardly a lefty—opined that the country was indeed in trouble, but he also commented that it was hard to underestimate the Democrats ability to “snatch defeat from the jaws of victory.”
I piped up and suggested that it might help if the Democrats (or a single Democrat) could articulate a series of clear, honestly held policy positions. For example, Republican candidates are generally associated with lower taxes, fewer regulations, restricted access (or a ban) on abortions, conservative judges, and a willingness to fight a more successful war on terror, even at the expense of civil liberties. It is hard (perhaps impossible), I argued, to rattle off a similar list of political first-principles for Democrats.
One of the political scientists suggested that the Democrats would lose if they ran campaigns based on ideas. Apparently there is empirical research (done by consultants or political scientists) that suggests that highlighting your opponents failures is a more effective strategy that offering bold policy initiatives. Roughly paraphrasing his comments, “There is no evidence that offering new ideas and policies have a political payoff.”
Perhaps it is possible to code for attack strategies and “new” policy ideas or initiatives in a campaign. But certainly it is difficult to quantify leadership or sincerely held policy positions that are not the product of a political calculus.
Here is a concrete example. According to a new book by journalist Joe Klein, when the Abu Ghraib torture scandal story broke in the spring of 2004, the Kerry campaign vetted appropriate reactions with a focus group, which apparently revealed sympathy for using whatever means necessary to extract information from the "terrorists." There is some scuttlebutt on how this scenario actually played out (e.g., here). But I clearly remember that Kerry said nothing about the scandal during the first week after Seymour Hersh’s story appeared in The New Yorker because I documented my disbelief in a letter to the New York Times (which I still have).
It is impossible to know what would have happened if John Kerry had (a) immediately and unambiguously condemned this conduct as shameful and counterproductive (according to Klein, that was what Kerry actually believed), and (b) proceeded to draw upon his experience to make this case to the American people. In other words, Kerry could have tried to shape rather than follow public opinion. It is also possible that a large number of voters can intuit sincerely held beliefs and those made for political gain. Thus, despite the country's discontent, the Kerry campaign went down in flames.
But how can a researcher code sincerity or leadership? They can’t. In any regression model of past elections, these intangible but arguably very important qualities become part of the error term. In turn, success or failure runs the risk of being attributed to substantive policy positions rather than the sincerity with which those convictions were held. As a result, data-driven political advice can backfire when it completely smothers a candidate’s political or moral instincts. I wonder if Democrat consultants have figured this out.