The Electoral College is kind of like the BCS (Bowl Championship Series) for college football. For both, the system is often more controversial than the winner. Opponents of both systems claim that these institutions preserve an outdated way of doing things, while supporters claim that the alternative could be anarchy.

Since the origins of the U.S. Constitution, we’ve had an Electoral College that picks a president. States are assigned a number of electors based upon how many members they have in the House of Representatives and Senate. For example, Georgia currently has 13 congressional districts and a pair of senators, giving the state 15 votes in the presidential election.

Win the state and you win all of its Electoral College votes, no matter how close the results in the state actually are. Whether you win by a vote or a million of them, it is a “winner-take-all” system except for a few states like Maine and Nebraska, who award them by congressional district.

One problem that critics claim is that the Electoral College could depress voter turnout. If your state is strongly Republican or Democrat, there’s clearly less incentive to vote. The weaker side will get discouraged, and the stronger side may not have everyone show up, given the expected margin of victory.

It’s the same with college football; turnout is likely to be higher at games where the teams are well matched (like LSU and Alabama) instead of poorly matched (like Auburn vs. Samford).

To see if the critics are right, my students and I looked at a close election: the 2004 election, where George W. Bush and John Kerry ran a neck-and-neck race that was decided on the national level by only a pair of percentage points (51 percent to 49 percent). Turnout as a result was pretty good for an American election (60.1 percent). But couldn’t it have been higher? Or did a number of uncompetitive states hold us back?

The data seems to indicate that this is the case. Our research of states’ electoral turnout and the margin of victory shows a significant relationship between the two factors. We found that the closer the election, the greater the turnout. States where the electoral contest wasn’t very close had, on average, lower turnouts than expected. This is all happening while polls are showing the national race to be neck-and-neck, both in the popular vote and the Electoral College itself!

Nowhere is this more apparent than the South, where the region had a lower-than-average turnout. Only one Southern state had a turnout more than a percentage point ahead of the national average: Florida. Only one Southern state had a close race, and that was in (you guessed it) Florida, where Bush prevailed 52 percent to 47 percent.

Twelve states were decided by five percentage points or less. Of these, nine had a turnout exceeding the national average. The others (like Nevada and New Mexico) were only a few percentage points from the national average. For those who pay attention to this sort of thing, the results of the crosstabs are statistically significant.

Despite the excitement over the close 2004 election, many found themselves stuck in a state with a lopsided race, and chose to stay home. It’s like the BCS game where biased pollsters or computer glitches provide a mismatch, like USC-Oklahoma or Miami-Nebraska. Like fans and TV viewers boycotting the big game, voters are doing the same to the Electoral College, to the detriment of democracy in America.

Brock Morris, Tray Roux, Cody Pringle and Sabrina Leverette contributed to this report.

John A. Tures is associate professor of political science at LaGrange College.