UC San Diego

The Democratic Party, once the party of southern slavery and farmers, has become the progressive party of the northern coastal states, while the Republican Party, once the party of Lincoln and northern industry, has become the party of southern and western states in presidential elections. Understanding how this happened promises to tell us something fundamental about American party politics. James Rowe left clues for us in his famous memorandum advising Truman in 1947. Rowe told Truman to explicitly “ignore” the South and push for liberal New Deal policies and civil rights. His policy prescriptions followed from studying the results of the 1944 election and determining that northern coastal states would hold the balance of power in 1948. Truman went on to win one the most improbable victories in history.

Give James Rowe his due for shrewd analysis, but even he acknowledged that it was “old hat” at the time. He was referring not to the content of the policies he recommended, but to the strategy of crafting regionally appropriate policies for battleground states. More recently, Donald Trump's victory may have hinged on the appeal of economic nationalism in the Rust Belt, where manufacturing jobs have been hit hardest. As with Truman in 1948, Trump jettisoned longstanding party orthodoxy in order to win a majority in the Electoral College. The presidents and the policies are different, and so are the historical contexts. Nonetheless, the underlying strategy is the same. What Rowe called “old hat” may in fact be a primary impetus for national party change.

Do Electoral College incentives (ECI) explain changes in party policies? This dissertation evaluates the relationship between ECI and party platforms to answer this question. It does so with unprecedented scope and rigor by considering all elections since 1832 and developing empirically valid approaches for identifying key states and translating them into policy signals for the parties to follow. Party platforms prove to be highly responsive to ECI in the statistical analysis. The results establish that ECI are a plausible driver of party change.