In a politically clever ploy, the California Republican Party has put forward an initiative that would apportion the state's electoral votes by congressional district rather than the current winner-take-all system. In the unlikely event that it were to pass, it would evidently give the Republican Party a huge (and unjustified) step up in subsequent presidential elections.
In a recent column in The New Yorker, Hendrik Hertzberg identified the two most obvious problems with the California initiative. First, even if apportioning electoral votes by congressional district made sense, it would be grossly unfair to do in only in one large, strongly Democratic state. (The same would hold true, of course, for a Democratic proposal to cherry-pick electoral votes in Texas.) The same electorate would be more likely to put a Republican in the White House, and the possibility of a Republican president being elected despite a decisive loss of the popular vote would increase.
Second, the proposed reform doesn't properly address the real problems with the Electoral College. Because the drawing of electoral boundaries is left to state legislatures, it would accelerate gerrymandering that could determine not only the composition of the House of Representatives but the presidency as well. However, even more worthy reforms -- such as the counterproposal by California Democrats to give the state's electoral votes to the winner of the national popular vote -- will be almost impossible to accomplish under the current system.
The California case illustrates the central problem with America's severely deficient electoral system: the fact that the administration of federal elections was largely left to the states. The Electoral College is an anachronism that distorts electoral outcomes (most recently, and with disastrous consequences for not only the country but the world, in 2000) and overrepresents small-state minorities that are already overrepresented throughout American political institutions. (As Yale law professor Akhil Reed Amar has pointed out, "the electoral college was designed to and did in fact advantage Southern white male propertied slaveholders in the antebellum era. And in election 2000, it again ended up working against women, blacks, and the poor, who voted overwhelmingly for Gore."
But privileging "states' rights" over people's rights not only constitutes a primary problem with the Electoral College but makes it nearly impossible to change. It produces the kind of collective action problems we can see in the California case (the states that act first will disadvantage their state's citizens) and gives small states a vested interest in maintaining the less democratic system. This is particularly irksome because, in a modern democracy, the decentralized administration of federal elections is "local control" fetishism at its least defensible.
The value of decentralized power in some contexts is that it can allow for experimentation and policies more attuned with local values. But such experimentation, while logical for a time period in which giving the franchise even to propertied white males was a fairly radical idea, could not be more inappropriate for a modern democracy. It should no longer be acceptable, of course, for states to "experiment" with which adults should get the franchise.
The most important bad consequence of state control of federal elections was the crucial role it played in upholding apartheid in the South for the century between the Civil War and the Voting Rights Act in 1965. The ability of state governments to administer elections, with the Supreme Court providing little opposition, allowed them to render the 15th Amendment's guarantee of race-neutral a dead letter. Although the states could not openly deny the franchise to black citizens, they could use arbitrarily administered literacy tests and poll taxes (as well as the tolerance of or collaboration with terrorist groups) to keep voting power in the hands of whites. The intervention of Congress and the Supreme Court was necessary to end the worst consequences of the decision to leave federal voting administration in the hands of states.
But, as the 2000 election demonstrates, "local control" of elections continues to have bad effects even in the post-Jim Crow era. The irrationality of letting individual districts determine their own ballot styles led to the butterfly ballot, which more than anything else led to George Bush winning Florida's electoral votes although a majority of Florida voters almost certainly intended to vote for Gore. The lack of uniform counting standards nearly led to a constitutional crisis. And further distorting the electoral outcome was the less reliable voting equipment in poor districts, which diluted the votes of the state's poor and African-American voters. (Before she removed it after being bullied by Antonin Scalia, Ruth Bader Ginsburg correctly noted in her dissent that the dilution of African-American votes because their districts couldn't afford cutting-edge electoral technology raised a much more substantial equal protection question that the claim the Court ultimately accepted.) Vote counts are either accurate or inaccurate; there's no justification for some districts having more accurate equipment than others (particularly since these effects of such differences are not random but will systematically overrepresent the votes of citizens in wealthier districts, who hardly need more representation in the modern political process.)
Although in many elections the outcomes aren't close enough to make the flaws of the electoral system as readily apparent, major reform remains necessary. The crucial principle is that the United States should, like most democracies, run federal elections with uniform ballots, uniform recount standards, and enough funding to ensure that poor districts are not disenfranchised by inferior technology. Ideally, steps would also be taken to ensure that the winner of the popular vote actually wins the presidency. (In 2004, Kerry was about 100,000 votes in Ohio from that election result being an even greater travesty than 2000.) But as the California case illustrates, such reforms often cannot be effectively accomplished from the state level up. So what can be done?
Some bad features of the American system, such as the Electoral College, we're probably stuck with. But some of the most egregious inequities can be remedied. First of all, Congress should use its spending powers to try to encourage relatively uniform ballot access and vote counting. While Congress cannot force the states to adopt these reforms, it can financially punish states who won't go along and also give states the resources necessary to make voting and vote counting more fair. Second, as the legal scholar Mark Tushnet has suggested (with tongue partly in cheek), we should pretend that Bush v. Gore actually articulated a legal principle. If vote dilution by counting ballots without a uniform standard violates the equal protection clause, this suggests (the attempt by the Supreme Court to cabin their decision to it would only apply to their favored litigant notwithstanding) that arbitrary reductions in voting power by using inferior voting technology in some districts also violates the Constitution.
None of this will be easy. But as long as the current irrational local-controlled system remains in place, the question is not if the next 2000-style crisis will come, but when.