College of Chaos

After the 2000 election fiasco, a Stanford University computer scientist named John Koza put aside his work on artificial intelligence to draw up a plan for fixing the Electoral College. Koza is no ivory-tower geek -- he's the co-inventor of the rub-off instant lottery ticket. He felt somebody had to make sure that what happened in 2000 -- the election of a president who didn't win the popular vote -- would never happen again.

By 2006, Koza and five colleagues had written a book, "Every Vote Equal," that presented what they called the "national popular vote" plan. Under it, individual states would agree to assign their electoral votes to the candidate with the most popular votes across the nation -- not just in that state.

The plan would take effect only when states representing at least half of the electoral vote total had enacted it. Four states -- Hawaii, Illinois, Maryland and New Jersey -- have adopted it so far, accounting for about one-fifth of the electoral votes the plan's supporters need. In nine more states, at least one chamber of the legislature has passed it. California's legislature approved it in 2006, only to have Governor Arnold Schwarzenegger veto it.

American democracy is much admired and copied around the world, but it's hard to imagine anybody copying the Electoral College. This institution emerged from the founders' special problems in getting the U.S. Constitution approved. They knew they could never achieve ratification if big states such as Pennsylvania and Massachusetts could muscle small ones such as Delaware and Rhode Island. So they set up an indirect system for electing the president, whereby a handful of electors in each state actually cast the ballots (to reduce the chances of mob rule by direct citizen voting). Votes were allocated according to the number of seats the state had in Congress, to give the small ones, with their two Senate seats each, a bit of a break.

It was probably the most awkward structure the founders created, and it has caused recurring problems. The framers seem not to have worried about the possibility that the popular vote winner could be denied election, or the turmoil such an event might cause. But it had happened three times already before the 2000 election made it four.

The Electoral College makes votes fundamentally unequal and leads candidates to focus most of their attention on a handful of swing states where a narrow margin can bring an electoral-vote bonanza. It discourages voters from showing up if their state isn't in play.

The national popular vote plan, Koza and his supporters contend, would change all that. Every vote would count the same, and candidates would be forced to court voters in states such as California and Texas where the statewide outcome is rarely in doubt. It also would end the possibility of mischief by electors, who can legally vote for any candidate they please, regardless of the popular result in their state.

Some Democrats and liberal good-government groups have embraced the plan. Republicans and conservative groups have countered that it's dangerous to tinker with the Constitution. But it's not only conservatives who have warned against making bold change. The late Democratic Senator Daniel Patrick Moynihan, for decades one of the nation's most provocative political thinkers, warned in 1979 that fundamental changes to the Electoral College would prove "the most radical transformation in our political system that has ever been considered." In creating our system, he concluded, "we built better than we knew."

Many Democrats, still stinging from the 2000 presidential result, find it hard to accept Moynihan's point. But behind the public debate is a strategic truth. Most political insiders of both parties prefer the current system, whose bizarre elements they know well, to a new one that would change the rules fundamentally and unpredictably.

So two things seem sure. First, John Koza will go back to making computers smarter instead of helping governments fix themselves. Second, if not in 2008, then some time, we'll face a repeat of the 2000 debacle. Then we'll rerun the whole debate about whether the founders were Rube Goldberg-smart or whether the process they created to elect our president was a dangerous mistake.