Many people have commented in the last few years on the apparent oddness of the situation that voters in two small U.S. States, New Hampshire and Iowa, have had a seemingly disproportionate influence on the U.S. Presidential primaries compared to much more populous states.
Recently, some states (e.g. California) have wanted to have a greater impact on the results of the Presidential primaries by moving the dates of their primary elections closer to the start of the year. The date of each State’s primary is determined (I think) by state legislation, and here is the current calendar of state primaries.
Many states will have held their primaries by Tuesday 5 February 2008, barely 9 months before the actual election – and by the end of that day a cumulative majority of delegates to both the Republican Party’s and Democratic Party’s conventions will have been elected. By the end of Tuesday 4 March 2008, 83.6% of delegates to the Democratic Party convention and 76.4% of delegates to the Republican Party convention will have been elected (source), by which time the it seems quite likely the Presidential Democratic and Republican candidates will be known.
I can understand each U.S. state wanting to have a greater impact on the course of the Presidential nomination. But the cumulative effect of each state moving the date of its primary earlier in the year is that the primary elections may end up being held a very long time before the election proper and the states may compete with each other to have earlier and earlier primaries. It is conceivable that a state could move its primary to the year before the election (this was discussed in a recent edition of the Economist). A question then arises as to when it is desirable to hold the primaries and what format the primaries take.
If, as is likely, the Democratic and Republican Presidential candidates are known in early March 2008, the campaign will take place over three-quarters of a year. I don’t know what U.S. voters think of this – to my mind a nine month campaign (on top of the primary campaigns) is very long – a six month campaign is long enough.
Instead of thinking about the length of the formal campaign, let’s look at the structure of the primaries. While the primaries are in theory concerned with the selection of delegates to the parties national conventions, they are seen as a rolling election-guide to the popularity of the candidates and how they perform on the campaign trail – does candidate X do well in the north-east, does candidate Y appeal in the mid-west and so on. After a poor performance in one primary, a candidate might pull out of the race, and after the initial couple of primaries, usually only a few candidates for each party remain in the race.
Unfortunately, the rolling-primary-election effectively deprives voters of input into the choice of a candidate in the states that hold their primaries last, which is possibly why the primary calendar has changed so that most primaries have been held by early March, with 14 jurisdictions choosing delegates to the Democratic convention and 15 jurisdictions choosing delegates to the Republican convention after 4 March 2008.
Perhaps an alternative scheme could be to hold a rolling primary for two or three months to tease out the candidates and how they perform on the campaign trail, and then choose all the remaining delegates on a single day. For instance perhaps 15-20 jurisdictions (e.g. states and territories) could select their delegates in a rolling fashion – perhaps two jurisdictions per week or fortnight – and the remaining jurisdictions could choose their delegates in one big bang. The 15-20 jurisdictions in the initial part of the process could be chosen randomly, or could be chosen to reflect a mixture of different parts of the country – e.g. at least one north-east state, one southern state, one western state, one mid-west state, one low population state and one high population state.
It’ll be interesting to see the impact on the new primary calendar on the selection process, and it’d be interesting to see how different primary procedures would work. Readers are welcome to suggest different schemes or timetables for the primaries in comments.