Last Wednesday the New York Times published a house editorial that was feebly reasoned even for the Times, a cry for “fairness” that in fact cared only about the fortunes of the Democratic party. Titled “Stacking the Electoral Deck,” it began this way: “The Electoral College should be abolished, but there is a right way to do it and a wrong way.” Three paragraphs later, the Olympian voice of the Times intoned, “The Electoral College should be done away with, but in the meantime, any reforms should improve the system, not make it worse.”
What was The Newspaper of Record complaining about? Well, as you may have heard, there is an initiative afoot in California to change the way the state allocates its electoral votes in presidential elections. Currently the state gives all 55 of its votes to the winner of the statewide popular vote. This “unit rule” or winner-take-all system, originating in the Jacksonian era as mass political parties came into being, is used today by 48 states (Maine and Nebraska being the only exceptions). With just 270 votes needed to win, a victor in California is one-fifth of the way there already.
The proposed change would allocate electoral votes to the popular vote winner in each of the state’s 53 congressional districts, with the remaining two votes going to the statewide winner. Since no second-place finisher in California gets zero popular votes, it seems on the face of it to be a movement in the direction of a more democratic system to permit the second-place finisher to carry home a portion of the electoral votes greater than zero — maybe even something approximating his percentage share of the popular vote. In 2004, when John Kerry took California, George W. Bush nevertheless won in 22 of the state’s congressional districts. Give Kerry two votes for his statewide victory, and he would have won 33 votes to Bush’s 22 under the current proposal — a perfect 60-40 split in percentage terms, and a “fairer” approximation of Bush’s share of the California electorate. Fairer than zero anyway, if the electoral vote is supposed to reflect popular vote results.
So what is the cause of the high dudgeon at the Times? Why do the editors call this plan a “sneaky initiative” by a “shadowy group,” an “elaborate dirty trick” that is “anti-democratic” and would “do serious damage to American democracy”? Why, because it “would rig elections in a way that would make it difficult for a Democrat to be elected president, no matter how the popular vote turns out.” Thus the plan is “anti-democratic” (or is that anti-Democratic?) because it is a “power grab on behalf of Republicans.”
Well, politics ain’t beanbag, as someone once said. The Constitution permits state legislatures to choose any method they like for the allocation of electoral votes. (The fact that this plan would, in California, be created by initiative and referendum, rather than by the state legislature, casts serious doubt on its constitutionality, as I recently told CNS News. But that’s another matter.) It is the least surprising thing in the world that one party wants electoral rules that advantage itself. The unit rule was first adopted by parties that controlled majorities in their states, and wanted to increase the weight of their contribution to the electoral vote tally. The party losing of late in California — the Republicans — wants to change the rule in that state. There’s nothing inimical to democracy in that.
How clear is it that such a change would advantage Republicans, especially if adopted nationwide? Well, in 2004, George W. Bush, with 51 percent of the popular vote to John Kerry’s 48 percent, took 286 electoral votes (53 percent) to Kerry’s 252 (47 percent). I spent a little quality time with Michael Barone’s 2006 Almanac of American Politics (the 2008 edition is available this week), the only source I know with results of the most recent presidential election by congressional district. If the proposed California change had been used nationally in 2004, with electoral votes awarded by House district and two votes to each statewide winner, George W. Bush would have won 317 electoral votes (59 percent) to Kerry’s 221 (41 percent). Maybe the Times did the math too.
This arithmetic is somewhat surprising, though. It is a well-known effect of the statewide unit rule that it usually magnifies somewhat the winner’s margin of popular vote victory (and can even, very rarely, give a victory where the nationally tallied popular vote would spell a defeat, as in 2000). One might expect a more finely gauged allocation of the electoral votes, by congressional district rather than whole states, to diminish this effect rather than magnify it further. So what’s going on?
A glance at the district-by-district analysis reveals the answer. In states that Bush won in 2004, he won more districts too — all or nearly all in many cases. In states that Kerry won, there were often a goodly number of districts won by Bush, many of them the districts of Democratic members of Congress. Bush’s margins of victory were often narrow, in both districts and states in which he was victorious. Kerry’s victories at the district level were usually overwhelming, with strongly Democratic voters concentrated there and thinner on the ground elsewhere, even in the states he won — and he won fewer Republican-held House districts than Bush did of Democrat-held ones.
The most striking case is Michigan, which Kerry took statewide while winning in only five of the fifteen congressional districts. Bush would have taken the majority of the state’s electoral votes, ten to Kerry’s seven, under a district-tally system. This is the only state that would have “flipped” in this way, but Bush would also have made serious inroads in other states, such as Illinois (9 of 21 votes), Minnesota (5 of 10), New Jersey (6 of 15), New York (9 of 31), Pennsylvania (9 of 21), Washington (3 of 11), and Wisconsin (4 of 10) — in addition to the 22 he would have gained in California. Kerry would have picked up fewer comparably substantial portions of Bush states, such as in Colorado (3 of 9 votes), Florida (7 of 27), Georgia (4 of 15), Missouri (3 of 11), North Carolina (4 of 15), Ohio (5 of 15), and Texas (7 of 27). In no other states would either candidate have gained more than two electoral votes under district-tally rules, but the changes in favor of each candidate would have been a pick-up of 51 votes by Kerry in Bush states, as against a pick-up of 82 votes by Bush in Kerry states, for a net gain of 31 by Bush.
A final statistic: Bush won eleven states in which Kerry took no district (in some cases could not have done, in the smallest states), for a total of 53 votes. By contrast Kerry won just eight states in which Bush took no district, for a total of 40 votes. This included the sizable prize of Kerry’s home state of Massachusetts, with 12 votes — the largest state either candidate took in toto on a district tally.
So it looks like the Times is right that the proposed change in California would favor Republicans, if adopted either there alone or nationwide. At least it would favor them in the short run. There is nothing in the nature of things, however, that makes Democratic presidential candidates weaker competitors in more of the nation’s congressional districts. The party’s strength is presently concentrated the most in urban districts held safely by Democratic House members, the product in part of gerrymandering (often more cooperative between the parties than competitive nowadays) to maximize the safety of those seats. The Republican party has its own hard core of safe districts (and whole safe states, usually smaller ones), but controls many districts by narrower margins and is more competitive in districts it does not even hold in the House — in presidential politics at least. In short, the GOP has its strength more geographically distributed, which accounts for the advantage it would have in a district-tally system.
Would redrawn congressional districts change the electoral map in presidential contests? Surely to some extent, and the district-tally system, if instituted in every state with at least two congressional districts, would create powerful incentives to rethink the virtue of gerrymandering for the sake of incumbent House members’ reelection. Our parties are famously adaptable to new institutional realities, and there is no reason to expect that the preceding “what-if” analysis of the last election, under rules that were not then in place, is a good predictor of what would happen one or two or three elections from now, if the new rules were actually adopted. Redistricting experts in both parties would quickly start to fiddle with the map to maximize geographically distributed strengths, yielding small but marginally predictable district-wide majorities. This would risk more House seats but help in presidential elections. Could the Democrats find more voters for their candidates on a more distributed basis? They would surely try.
So the Times editors’ complaint that the California change “would make it difficult for a Democrat to be elected president” is based on a prediction of what could turn out to be either a very short-term effect, or wholly chimerical. But what about their flat statement that such a change is “anti-democratic”?
Horsefeathers. Until the Times defines what it means by “democratic,” this charge cannot be taken seriously. The unit rule commonly used to allocate electoral votes by whole states is perfectly democratic; since it honors plurality rule in the states, it can be called federally democratic. The proposal pending in California, by honoring plurality rule in House districts, would be locally democratic. The Times would like to do away with the Electoral College altogether, presumably moving to some form of nationwide tally of the popular vote, which would be nationally democratic. None of these ways of electing a president has more legitimate democratic credentials than another. It’s fairly obvious, from their bleating about “fairness,” that the editors of the Times are more concerned about the fate of the Democratic party than about the fate of democracy.
The more serious questions about any change in the framework of the electoral college are, what would be the institutional and behavioral effects? How would each of our parties adapt, or fail to adapt? How would presidential campaigning change to take in the new situation? And what about the future of our two-party system itself?
A shift to the system the Times prefers — no electoral votes, just count all the individual votes of citizens — would reinforce the present strengths of the Democrats, who do best in the country’s most dense population centers. Do we want a president of our urban enclaves? But that would be only a short-term effect. We know that the state-unit rule, and the majority requirement in the Electoral College (270 of 538 votes to win), together lend powerful support to the two-party system’s dominance over our politics, with minor parties effectively frozen out. No one has ever proposed an elimination of the Electoral College that would not predictably weaken the two-party system, and destabilize our presidential politics by subjecting it to the bargaining power of minor parties seeking coalition rights. (Thirty years ago last month, the great political scientist Martin Diamond made this argument to a subcommittee of the Senate Judiciary Committee, just before dying of a heart attack in the hearing room. His argument is still irrefutable.)
As an approximation of its dearest wish, the Times backs a “rival initiative” in California, also proposed in other states, which would “ensure that the winner of the national popular vote becomes president.” The idea is that a law in each state would mandate the award of all the state’s electoral votes to the party of the national popular-vote winner, regardless of who led among the state’s own voters. How this might affect the two-party system over time is harder to say, but it has other problems. In a close election, a national recount might be necessary, while the present system confines recounts to particular states. Imagine the drama, tension, and shenanigans of Florida 2000 on a nationwide scale.
The district-tally system proposed in California is also hard to assess for its impact on the two-party system. To have much of an impact, a minor-party candidate would have to roll up some House districts, and that won’t soon happen. But eventually the decay of the two-party system might set in, and for some people, myself included, that’s reason enough to oppose the California initiative.
But this argument requires some thought, not just a knee-jerk fear that one’s favored party will be disadvantaged in the near term. And that’s all the Times had to offer.