The Problems of Regional Primaries

Written Statement on Federal Regional Primary Legislation
Submitted to the Senate Rules Committee

By William G. Mayer
Associate Professor of Political Science
Northeastern University
19 September 2007

Finally, I would like to call the Committee's attention to a number of problems with regional primaries, however they are adopted and enforced. First, though regional primaries have recently been proposed primarily as an antidote to front-loading, it is by no means clear that a regional primary system would actually reduce front-loading. It depends on how the system is designed. The proposal formulated by the National Association of Secretaries of State (NASS) provides a good example of the problem. The NASS calendar allots separate weeks to Iowa and New Hampshire -- and then, one week later, the first region would vote. In other words, one week after New Hampshire, delegates would be selected in twelve different states on the same day. By comparison, most recent presidential nomination calendars have started up more slowly. Immediately after New Hampshire, there have typically been several weeks in which only one or two states held their primaries or caucuses. The NASS calendar, to be sure, would be less front-loaded after that: There would be a month off before the next region voted. But this is small consolation to all the candidates who cannot afford to campaign in twelve states, even twelve contiguous states, just one week after the race begins and who will therefore not be around when the second region goes to the polls.

Another major problem associated with regional primaries is that they would confer a significant advantage on any candidate who happened to be particularly strong in whatever region went first.27 As Table 1 indicates (it is located at the end of this statement), region is a very important variable in explaining primary outcomes. Almost every recent presidential candidate has done significantly better in one region than in the others. In 1976, for example, Gerald Ford won 60 percent of the vote in the average northeastern primary, as against 35 percent in the average western primary. In the same year, Jimmy Carter won, on
average, 62 percent of the vote in the South, 35 percent in the Northeast, and 21 percent in the West. In 1980, Edward Kennedy won 53 percent of the vote in the average northeastern primary, but only 18 percent in the southern primaries.

In the contemporary presidential nomination process, the order in which primaries are held matters. Indeed, that is why front-loading developed in the first place. And thus, which region goes first could have very important implications for which candidate gets nominated. In 1992, for example, Bill Clinton's candidacy would likely have been doomed if the southern states had voted last: for the first five weeks of that year's delegate selection season, Clinton didn't win a single primary or caucus outside the South. Supporters of regional primaries implicitly acknowledge this problem, for regional primary proposals invariably include a provision that rotates the order in which regions vote or determines that order by lot. But rotation and lotteries do not eliminate this problem -- they merely ensure that the direction and recipient of the distortion will vary, in a random manner, from one election cycle to the next.

FULL STORY