I am primarily a theoretical researcher in political economy. What interests me most are spatial models of voting in multiple dimensions.
My first main area of research is candidate differentiation on the basis of what positions may be taken. Since candidates come into political competition with party affiliations that restrain what positions they may hold, they can only promise from a subset of the overall policy space. This kind of restriction helps to reverse some of the classic “instability” results, allowing for equilibrium existence even when there is no total median or symmetry. It also delivers interesting comparative statics. For example, when voters become more extreme in their opinions, it is actually up to the candidate to choose whether to take advantage of this change- if their old position allowed them to win, then they can still win with that position. Electoral incentives alone are not enough to motivate extreme policy promises- that requires some ulterior preference on the part of the candidates.
You can look at my latest draft of this paper here: GSNS 2.pdf .
My second area of interest is the control of candidates in office. Many models assume that the only benefits that candidates get from office is the ability to siphon rents , or to implement some policy they really care about. Worse, they assume a “representative voter” who cares about the good of society and an ever-ready (but nameless and homogenous) challenger, rather than any sort of serious spatial election.
I think the original Downsian assumption, that politicians like getting into power and keeping it, has quite a bit of merit. If these politicians are really motivated by nothing but money and get no particular compensating differential from power, then there is no reason they shouldn’t go into finance or playing drums on the street or whatever the highest return is at the moment. Further, there is no reason to suppose that candidates actually take positions and campaign and need to get votes from a majority to get elected , and then assume that all voters merge into a public goods maximizing hive-mind 4 to 6 years later.
The issue is that most models see the incumbent as a sitting duck when it comes time for re-election- his positions are fixed, and so it will almost always be the case that a challenger can find a majority preferred position. This is at odds with the observation that incumbents are extremely hard to dislodge. However, the challenger has to actually find that winning position. If gaining valence by doing well in office makes those positions more scarce, making the probability the opponent finds it lower, then there is a clear incentive for incumbents to try hard as long as they believe it improves their odds the next time . Incumbents become lax in producing public goods only as the public places less emphasis on public goods and more on “divisive” issues, until in the limit, the incumbent has to incentive whatsoever to produce the public good.
This paper is under development, and will likely be my job market paper.