Friday, April 26, 2002
David Lewis Schaefer has a piece in NRO about how France's electoral system is to blame for Le Pen's victory. His point: direct elections for the Presidency, with a runoff to ensure that the victor has a majority, encourages the formation of splinter groups. Therefore, America should not have direct elections to the Presidency lest the same thing happen here.
The first point is true, but the latter doesn't follow. It's not direct elections that encourage the splintering of the vote by extremist parties; it's having runoffs. If you allow a plurality victory, the incentives are very strong to back the strongest horse you can stomach, rather than help the enemy by supporting a fringe candidate certain to lose. Another alternative to runoffs: preference voting. If there are five candidates, you designate first, second, third and fourth choices. If there's no majority, you divvy up the last-place finisher's votes by second-choice; if there's still no majority, you continue up the line until there is. Of course, preference voting has its own absurd effects. In France, it might have resulted in a Socialist victory. After all, most of the fringe groups were far-left, and most of their voters would probably have chosen Jospin for second place. It's not inconceivable that Jospin would have won the election in a preference-voting system. In general, preference voting will empower the extremes because the victor will to some extent be beholden to the votes cast for more extreme parties but who picked him second. But, while preference voting can result in some quite perverse outcomes, it is very unlikely that a fringe candidate could win under such a system, because, while such candidates can hold a core of strong supporters, they will get very few second-choice ballots.
The real reason why the electoral college is good is that it reduces the leverage in the system. If the Florida recount fiasco had happened under a system of direct elections, you'd have to recount all the ballots nationwide. If you want an electoral college reform that would really respond to Florida, you would have to move to congressional district-based allocation of electoral votes, the system in place in Maine and Nebraska. Under the current system, most electioneering energy is spent on swing states, and this pushes the system to the center. But that campaigning can take the form of base-oriented voting, as was in fact the case in the last Presidential election, at least on the Democratic side. Gore carried Michigan, for example, and nearly Florida because of an unprecedented turnout by black voters. But these voters were frequently concentrated in safe Democratic congressional districts. A district-based electoral vote allocation would have forced a different strategy, one focused on swing districts rather than swing states. It would also reduce the leverage in the system, since a close vote in one or another district would affect only the allocation of one electoral vote, not, in Florida's case, 25 (or, in a direct-election system, the entire election). The main downside of such a system is twofold. First, while it would have the salutary effect of pushing campaigns to the center, it would have the deleterious effect of focusing all campaigns on what would likely be demographically similar suburban districts. Rural and urban America, the former reliably Republican, the latter reliably Democratic, would get even less attention than they do now. Second, it would increase the leverage on the redistricting process, since this would now not only affect the outcome in Congress but potentially the outcome in Presidential elections as well. And this is probably not a good thing.