Congress reconvenes next week, and once again it seems likely to consider and reject the McCain-Feingold campaign finance reform bill. Even if such legislation did make it through Congress, it would likely run into trouble in the courts. Just two weeks ago, a federal judge struck down as unconstitutional most provisions of California's Proposition 208, a 1996 initiative governing state campaigns.

But amid the debate, one of the most important questions seldom gets asked: Just what difference does money make in congressional elections? At least for general elections in November, the answer appears to be: Not much.

Our organization recently conducted a comprehensive analysis of 1996 U.S. House election. In reviewing the contests for open seats -- perhaps the best measure of money's impact -- we found that how a district voted in that year's presidential race was a much stronger predictor of its congressional choice than relative campaign spending was. Nearly one-third of Republican winners in open-seat races in 1996 were outspent by Democrats -- but all of them were from districts where Bill Clinton received less than 50% of the vote.

Where Mr. Clinton ran poorly, Republicans consistently won easily. Where he ran well, Democrats won easily. In districts where Mr. Clinton ran close to his national average, nearly every race was close. These results occurred no matter how much money candidates spent. And in most districts the 1996 presidential results were very similar to those four years earlier, suggesting that voters tend to be consistent between elections as well.

All of which strongly suggests that most elections are decided during the decennial redistricting process. That is when Democrats and Republicans, now aided by polling and sophisticated computer software, carve up the political map to protect incumbents, creating noncompetitive districts "safe" from changing parties. In 1996, only two out of 171 incumbents first elected before 1990 were defeated; 162 won by at least 10%. Politicians are choosing the voters, rather than the other way around.

Campaign contributors respond to high incumbent re- election rates more than they cause them. Most big donors seek to buy influence, not elections. And come this November we will again have mostly noncompetitive elections in one-party districts, where voters merely ratify the decisions made by redistricting committees. A disproportionate amount of money will continue to be spent on the small number of undecided swing voters who decide the small number of close races in swing districts.

Campaign finance reformers should focus more energy on campaign financing in party primaries, where money indeed has a major impact because voters lack the guide of partisan labels -- and where parties have the power to set new internal rules immediately without blaming each other for the failure of legislation.

To make general elections truly competitive, it won't be enough to change campaign finance. We will need to take the redistricting process out of the hands of the incumbents and their parties, and give it to an independent, nonpartisan commission that uses nonpolitical criteria for line drawing. This has been done in Iowa since 1981, and has resulted in somewhat more competitive elections.

There may even be a case for considering more drastic reforms, such as cumulative voting or proportional representation. But at a bare minimum, any serious attempt at election reform must address the redistricting process, the least democratic aspect of our winner-take-all system.

Commentaries Index