One of my formative political experiences, and one which almost single-handedly convinced me to pursue political behavior as an academic career, occurred in 1998 in a class I hated. More accurately it was in the class of a professor I hated, the resident expert on constitutional law and chasing 19 year old tail at the University of Wisconsin.

Madison is a remarkably liberal place, so it was no surprise that very liberal and openly lesbian Congressional candidate (now Representative) Tammy Baldwin was extremely popular on campus. Aside from her ideological compatibility with the average Madison resident her campaign was well-organized, hosting "Tammy Baldwin dance parties" that drew huge numbers of students with no particular interest in politics. So it was a safe bet that any classroom at UW in the Fall of 1998 was about 85-90% Baldwin supporters.

Professor X began class one day – a large Con Law lecture of about 180 mostly future law school students – and asked for a show of hands on the question, "Which one of the two candidates for Congress is more likely to vote to lower taxes?" A solid 90% of the hands went up for Baldwin, with only the remaining 10% choosing her Republican opponent (the forgettable Jo Musser, who I remember for some reason over a decade later). Now, Tammy Baldwin was and is many things. But by no stretch of the imagination was she an anti-tax hawk. In fact she explicitly stated that she was not opposed to increasing taxes if necessary to pay for expanded Federal programs. Musser was the typical cut-taxes-at-all-costs Republican who made lowering taxes the cornerstone of her campaign message. So this was not a subjective question. There was a wrong answer, and 90% of the class chose it.

This was the exact moment at which I realized that the mental calculus of the average voter goes something like:

Me like Baldwin.
Me like tax cuts.
Baldwin is for tax cuts.

People who study political behavior understand, from as far back as The American Voter in 1960, that party identification is the key to this kind of "reasoning." Partisanship is a screen through which all incoming information is filtered. Republicans will tell you that everything they like is a Republican idea and everything they hate is what Democrats stand for. After decades of contorting research in an effort to give American voters a little more credit for intelligence, the field has come full circle and once again embraced the idea of partisanship as a "fundamental part of an individual's identity" (Partisan Hearts and Minds) like religion or ethnicity. The overwhelming majority of us acquire a partisan identity in adolescence or early adulthood and interpret the rest of the world through it thereafter.

Why is this relevant? Well, as Krugman pointed out, this is relevant to understanding the folly of pursuing "deficit reduction" as a political strategy by the White House, to which I would add the strategy of appeasing conservative Democrats or any Republicans on health care. Republican voters, most of whom hate Obama, will believe that he made the deficit bigger and that his health care plan is a disaster – even if the deficit is demonstrably smaller. Democrats will say the opposite. Really. How do we know? Because this is exactly what happened to Bill Clinton, as Achen and Bartels demonstrate.


Voters don't actually know anything. This is widely understood and constitutes one of the few points on which political science research is nearly unanimous. If Obama, like Clinton, destroys the deficit, Republicans and "independent" voters will not realize it. Hell, even some Democrats won't realize it. What people DO realize is when their own circumstances improve or worsen. Anything nebulous or intangible like deficit reduction will be lost on them, but they understand unemployment or failing banks. So in political terms, if the President thinks he can jump start the economy by racking up an enormous deficit he should do so. No one will know if it increases or decreases – voters will simply fill in that information according to their partisan identification. But many of them will know if unemployment falls, interest rates for auto and home loans fall, or their 401(k) recovers. If a health care bill provides coverage to people who could not previously afford it, it will be a net benefit.

Do things that improve people's lives and you will find yourself politically rewarded. Achieve abstract goals in an effort to appease people who hate you and you will find the victory quite hollow.