I like Larry Sabato, his Crystal Ball column and the fine work of the University of Virginia Center for Politics. But his post criticizing a proposal to reform the Electoral College falls well short of the standards ordinarily maintained by him and his organization. His thesis is easily unraveled, and it seems to be transparently based on concern that the change could hurt Democrats.
The proposal at issue would change the Electoral College by awarding one electoral vote to the winner of each individual congressional district and the state’s remaining two electoral votes to the statewide winner, the way Maine and Nebraska already award their electoral votes. The other 48 states and District of Columbia currently award electoral votes on a winner-take-all basis.
Crystal Ball‘s hyperbolic headline screams that the plan would “undermine democracy.” Sabato’s personal comment called the proposal “truly rotten” and claimed that it would “fix and game the Electoral College” to benefit Republicans.
The main analysis, penned by Senior Columnist Prof. Alan Abramowitz, claims the such a plan would have changed the result of the 2012 election because of gerrymandered congressional districts, specifically citing Republican-drawn plans in Florida, Michigan, Ohio, Pennsylvania, Virginia and Wisconsin, all states that Obama carried last year. But those states’ plans weren’t particularly gerrymandered or even unfair. Abramowitz failed to mention the most blatant gerrymanders, by Republicans in North Carolina, by Democrats in Illinois, and by a nominally non-partisan board in California. Since Romney won North Carolina statewide, the proposed plan would actually have given Obama some electors from that state that he did not receive under the current system. The reverse would have been true in Illinois and California.
Whining about gerrymandering ignores the fact that who benefits from gerrymandering changes over time. In Missouri, for example, the current Republican-drawn map, which caused the state’s loss of a seat from decennial reapportionment to come at the Democrats’ expense, represented the first time Republicans had drawn the state’s map since 1920. The liberally oriented main stream academics and media rarely objected when unfair redistricting favored Democrats.
The real reason that the proposed plan would have given the Republican nominee an electoral majority in 2012 even though the Democrat won the national popular vote by several million votes (and why Republicans control the House of Representatives even though Democratic congressional candidates outpolled Republican candidates nationally) is because of housing patterns. People who vote Democratic tend to congregate in metropolitan areas filled with people who think and vote like themselves, while people who vote Republican are more geographically dispersed. In the McKinley/Teddy Roosevelt era, the politics were exactly reversed, with Republicans dominating most cities and Democrats dominating most rural areas. (The red/blue state map of McKinley’s 1896 election is virtually a mirror image of the 2000 election, except for five states not then admitted to the Union.)
The rationale for the Electoral College is to give relevance to a broader geographical range of voters. When George W. Bush was criticized that he had won unfairly and against the will of the people in 2000, he explained that he would have campaigned differently if the popular vote had determined the outcome, but he campaigned according to the rules in place. Barack Obama did exactly the same thing, surgically concentrating on just ten “swing” states. Electing the president by the popular vote would put a premium on appealing to geographically concentrated voters who could be reached most economically, to the detriment of everyone else. That would shift attention away from most “swing” states, mostly to the benefit of single-party machine areas. Since Democrats are geographically concentrated, that plan would favor Democrats. And that bias would presumably be just fine for liberal academics.
Balkanizing presidential election returns into separate, independent elections (51 current jurisdictions, 486 under the proposal) insulates the election from fraud (whether by multiple or ineligible voting, voter suppression, or logistical problems with overseas and military ballots) by isolating its impact to a single area that is already likely to vote for the party that would benefit from the fraud. A few political machines (either urban Democrat or rural Republican) are in a better position to steal a national election in a larger election than in 51 or 486 smaller ones. For example, under the current system, the Chicago Democratic machine allegedly changed the outcome of the razor-thin 1960 election by swinging Illinois’ electoral votes to Kennedy with late reporting votes. (Democrats may prefer confused butterfly ballot voting in Florida in 2000 as an election-changing example.) Under a popular vote system, a few like-minded machines could manufacture millions of phony votes and steal an election that wasn’t close. Under the district vote proposal, they could have only affected congressional districts that they were going to win anyway, plus the two statewide votes.
The proposed reform would bring the election results closer to the people by giving relevance to their individual district, while still giving some attention to statewide results. The voices (and votes) of urban Democrats in Missouri and Texas and rural Republicans in Illinois and California would actually matter. Overall, that would currently favor Republicans. At other times (e.g., 1945-1980) it would favor Democrats. And the system that’s in place now isn’t really all that bad.