Our predictions are based on a straightforward idea, but one that still can surprise some observers: most people are grounded in their political views and consistent in their voting patterns over time, particularly at any given level of election. If one can find an accurate measure of what we call "district partisanship" -- the partisan leanings of voters in a district -- then most districts tilt toward one major party or the other.
- Monopoly Politics 2000
Conceived in the mid-1990s, district partisanship has let insiders approximate "the partisan leanings of voters in a district."It is the lynchpin of Monopoly Politics, a spreadsheet we hope anyone can use to predict the landscape of U.S. House seats well in advance of a Congressional election. It has the added benefit of allowing the user to model entrenchment and vulnerability in different "partisan worlds." What if the Republicans win 52% of votes nationwide? What if the Democrats win 52%?
An operating assumption is that we live in, more or less, a "50/50 world." The major parties have cobbled relatively equal shares of voters into their respective coalitions. The battle is for a narrow group of swing voters in so-called swing districts. As a concept, "swing voter" encapuslates ticket-splitters, increased turnout on the winner's side and decreased turnout on the loser's side.
Using data on the presidential vote in each district and nationwide, MP constructs district partisanship. Then it uses data on (1) incumbency (2) candidates' past performance relative to partisanship and (3) the national, two-party vote in order to project a margin of victory in each district. If it can, it projects the winner. Later on, we use certified election results to test our predictions.
In the 1990s, readers used MP to show how partisanship was a better predictor of election outcomes than candidate spending. In this decade, it underscores the advantage accruing from incumbency and the destiny handed down by geography.
Use the links on the left to browse Monopoly Politics.
"One interesting statistical study was released last month by the non-partisan Center for Voting and Democracy.... The report says candidate performance has nothing to do with how much money a challenger has and everything to do with the racial, economic and political makeup of the constituents in that district."
- Charlie Cook, August 1997