The United States has a presidential system of government. This simple–and constitutionally entrenched–fact has profound implications for any project to reform our electoral system. It is one thing to push a parliamentary system, such as New Zealand or the United Kingdom, toward proportional representation. It is quite another to push a presidential system in the same direction. I would argue that any such push should be limited–as is, for example, Richie and Hill’s current proposal of three-seat districts operating under the single transferable vote system.
The argument for limiting the move to PR goes as follows: first, PR tends to promote multipartyism; second, multiparty presidential regimes have in practice performed poorly. As the first point–that PR facilitates multipartism–is not in dispute, I shall focus on the second.
That multiparty presidential regimes have in fact performed poorly is evident from the Latin American record. Chile (1932-73) is the only case of a multiparty presidential system lasting more than a quarter century. All the other examples of long-lived presidential democracy–Colombia, Costa Rica, pre-1973 Uruguay, and Venezuela–come from systems with long spells of two-partyism.
The reason generally offered to explain why multiparty presidentialism works poorly is that such systems typically deprive the president of a working majority in the legislature, resulting in ineffective and gridlock-prone “divided government.” In Latin America, this has even contributed to attempts to govern by presidential decree (for example, Collor in Brazil) and coups d’tat (for example, Fujimori in Peru).
In the United States, divided government has not had such dramatic consequences but some of the same tendencies can be seen. I shall consider just two examples: budget deficits and “unilateralism.”
The enormous budget deficit of the early 1980s–one of the defining issues of US politics in the past twenty years–has been blamed in part on the fact that government was divided under Ronald Reagan’s presidency. A simple thought experiment suggests the line of argument: first, had Reagan had a firm majority of Republicans in Congress in place for the 1982 budget, the deficit might still have increased (due to the big increase in defense spending combined with tax cuts) but it would not have ballooned by as much as it did (because he would have made substantial cuts in social spending); second, had Jimmy Carter managed to defeat Reagan in 1980, again there might have been a deficit but it would not have been as large: the Democrats would not have increased defense spending by as much nor would they have cut taxes. Under divided government, however, the two parties’ opposing interests led to a triple threat to fiscal prudence–greatly increased defense spending, large tax cuts, no cuts in social spending–and hence the largest deficits as a percent of GDP in our history outside of wartime. This tendency of divided governments to produce larger budget deficits has been confirmed in studies of the US states and of bicameral institutions in Europe.
Another tendency that can be observed under divided government in the United States is for both parties to pursue their policy objectives unilaterally, to the extent that they are able. For the party controlling the presidency, this has meant attempts to expand the institutional powers of the presidency (e.g., Nixon’s attempt to expand the power of impoundment in the early 1970s) or to use executive orders rather than statutes to implement policies (e.g., Bush’s “gag rule” forbidding federally funded family planning counselors to mention the option of abortion). Perhaps the most extended example of unilateralism at both ends of Pennsylvania Avenue occurred during the 100th Congress, when President Reagan and Speaker of the House Jim Wright conducted what amounted to separate and incompatible foreign policies toward the Sandinista regime in Nicaragua–each side using powers and strategies that did not require the consent of the other.
The general point is this: a highly proportional electoral system to elect the legislature, when combined with presidentialism, will enhance the chances of getting divided government. While power-sharing governments of some forms are extolled as desirable–particularly in the literature on “consociationalism”1–divided presidential regimes have some undesirable features. The key problem from my perspective has to do with what happens when power-sharing politicians cannot agree on a compromise. Under a parliamentary regime, when coalition partners cannot agree on a sufficiently important matter, the government falls: the top political leaders lose their positions, or at least put them at immediate risk, and appeal to public opinion to resolve the impasse. Thus, part of the pain for failing to arrive at an acceptable compromise falls immediately on politicians. Contrast this to what happens in the United States when we fail to pass a budget before the close of the fiscal year. No politician immediately loses office. The electoral consequences may turn out to be substantial–arguably the budget impasse of 1995 hurt Bob Dole’s presidential bid–but they are not immediate. The only immediate pain that arose as a consequence of failure to agree on a budget compromise fell on various civil servants and users of government services. Not only can presidents and assemblies defy one another without fear of immediate loss of position in presidential regimes, but they both can claim a direct mandate from the people. The bargaining incentives in presidential regimes more often lead to delay and brinkmanship, public shouting matches, and the burning of bridges. Public dispute between elites is a precious thing, not to be done away with by any means. But there is such a thing as too much of a good thing.
Given that converting our system from a presidential to a parliamentary one is not at all likely at present, electoral reformers need to accept presidentialism as a given and react accordingly. Especially those who dislike the status quo (whether progressives or reactionaries) will want to make sure that a single reasonably cohesive majority coalition could in principle emerge under a proposed new electoral system. Otherwise, bold new departures in any direction are rendered unlikely.
1 See particularly Arend Lijphart, Democracy in Plural Societies (New Haven: Yale University Press, 1977).
