Institutional foundations of partisan change: A new look at critical elections in America

On five occasions during the first 150 years of American history (1800, 1828, 1860, 1896 and 1932), one of the two major political parties seized control of the Presidency, House and Senate, and controlled all three without interruption for an extended period of at least twelve years. Historically, these “partisan realignments” or “critical elections” occurred about once a generation. Yet one has not occurred since the 1930s. Why? How do we account for this apparent transformation in the complexion of partisan electoral history in America, and what significance should we attach to it?; These questions are important in light of the significant role critical elections have played historically in addressing major policy concerns and setting American politics and political institutions in new directions. The classic literature on realignments has not provided a complete answer in part because it has not provided a fully-developed explanation of the causes of critical elections. Combining historical analysis with new institutional economics, this dissertation offers a new theory of party behavior and policy change that accounts for both the genesis of critical elections historically in the U.S. and the absence of a classic partisan realignment since the 1930s.; Principal findings are threefold. First, under certain conditions, a party's prior commitments become self enforcing constraints on adjustment to new political circumstances. Past critical elections occurred when a dominant party's prior commitments to unresponsive policies became self-enforcing. Second, the rise of the administrative state has deeply affected the nature of the policy-making system, particularly its response to shifts in public opinion. I show theoretically why the kind of policy change that helped fuel past partisan realignments is less likely to occur in the post-New Deal administrative state, even in the face of self-enforcing party commitment to particular policies. Third, statistical data suggest that the uniformity of partisan electoral behavior that scholars have observed as a prelude to past critical elections is no longer unique to realigning moments. Critical elections are still possible, but, for institutional reasons, are less likely to occur, and are no longer necessarily an effective corrective for unresponsive policy-making.