Monday, September 19, 2011

One Cheer for the Electoral College

I had a piece up over the weekend at TNR talking about changes in the effects of the Electoral College over time. Short version: the electoral college used to have a major big-state bias, but recently that's dissipated quite a bit.

Still, overall there's still a slight big state bias in the EC, and there's at least some theoretical reason to believe that it would be that way. Basically, the states that matter per the EC system are those that are big and close. And that's good -- because Congress rewards small states (in the Senate) and one-party states and districts (to the extent that seniority matters). Which means that large cities, in particular, have traditionally been helped by the EC.

That is, I strongly disagree with Scott Lemieux's claim that "electoral college is a particularly egregious example of status quo bias; there’s no real defense on the merits to be made." Perhaps in the abstract there's no defense to be made -- although I'm not really sure about that. But in practice, what the EC did during the 20th century -- and still does a bit today -- is to balance out other biases in the system, and that's not nothing.

What I'd say more generally is that in my view it's a mistake to assume that there's some perfect electoral system out there that would perfectly reflect the true views of the electorate. What we know both about voters' preferences and about the math of electoral systems suggest that there is no such system. Of course, that doesn't mean that all systems are equally suspect, but it does mean that we shouldn't assume that either a simple plurality or a runoff/majority system would be some sort of obvious ideal. And the US is hardly the only democracy that uses convoluted methods to choose heads of government; after all, it's not as if the British PM is chosen by a direct popular vote.

I would say that the Electoral College would be far less justifiable if it regularly returned the "wrong" winner (if we suppose that the "correct" winner is the plurality winner), but in fact that's rare and likely to stay that way. I also do think that the EC is much less supportable if the current trend continues of big states becoming less competitive. But overall, I'm not at all convinced that moving away from the Electoral College is a good idea, at all.


  1. What does this mean? "we shouldn't assume that either a simple plurality or a runoff/majority system would be some sort of obvious ideal."

    Those are obvious ideals to me... that is to say I've never actually questioned the stupidity of the electoral college. Moving closer towards "one person one vote" as an ideal sounds good - and you haven't really changed my mind on this by reminding me that the senate is also pretty undemocratic.

  2. How does the EC have a "big-state bias"? That sounds incredibly counter-intuitive. A Wyomingite's presidential vote counts WAY more than a Californian's. (In 2000, CA had 615,848 voters per elector; WY had only 164,594.) In my book, that's small-state bias. Which just reinforces the Senate's small-state bias.

    Are you saying that there's a big-state bias because the results of elections are more likely to align with the preferences of big states? That's coincidence; not bias.

    And I understand that candidates won't want to campaign in small states (even if they are swing states) because there isn't enough at stake. But that has nothing to do with "bias" either, does it? I'd rather be in the position of a Wyoming voter (i.e., greater say in who the electors are but no visits from candidates) than a Florida voter (candidates campaigning constantly in the state but voters have less say).

  3. This comment has been removed by the author.

  4. It's hard for me to see what kind of argument there is in favor of a system where who becomes President can turn on bad ballot design in a single large state during a close election.

    Since you can count on one hand the number of elections where the EC and the PV didn't agree, why not just get rid of the EC so we don't ever have to repeat the nightmare of those few times like 2000 when the popular vote loser won the electoral college? It's far more likely that we'll repeat the instance where one large state is virtually tied leading to random factors or who has the most Supreme Court justices deciding the election than that the overall national popular vote would be so close as to not be decisive.

  5. Bad link. And bad headline by TNR editor.

  6. Rare is in the eye of the beholder, I think. I can count at least three elections where the EC unequivocally picked the popular vote loser (1876, 1888, 2000); depending on where you start counting, that's an error rate of at least 6% (there have been 52 presidential elections since the 12th amendment). Not to mention the number of times that came close--2004, 1960, etc. It's a pretty poor performer.

  7. There's a big problem with the comparison here between the EC and the "convoluted" systems of other nations. To my knowledge, no other Western nation has a system in which a national popular vote for head of government is totaled up, but then possibly set aside in favor of some other count. In multi-party systems, it's possible for, say, the leaders of two or more minority parties to join together in a majority coalition against the party that comes in first, but only if that party didn't get a majority itself. And in such a case, the majority coalition would still be representing, if imperfectly, the wishes of the national majority, i.e. the popular-vote "winners" (inasmuch as, together, they did win most of the votes, and if they're in coalition it means they must agree politically on key policies).

    Even in the UK, where first-past-the-post MP constituencies make it theoretically possible for the leader of a second-ranked party to become prime minister, I don't think this has actually happened -- mainly because those constituencies are so small (smaller even than US congressional districts, let alone states) that the party winning the most votes is also very likely to be the party winning the most seats. And anyway, there is no separate vote taken for prime minister. So, in the unlikely event that (say) the Conservatives got the most votes but Labour won the most seats and took office, this fact would be noted and might weaken the new Labour government, but you wouldn't be able to say that the people had chosen the Conservative leader and then been overruled by the system. It might be, instead, that Labour had simply done a better job than the Conservatives of fielding attractive local MP candidates (because that's what people vote for in Britain, not the prime minister or the party as such).

    What we have is much more dangerous. In our system, people vote directly for president, and these votes are (unofficially) totaled up nationally as they're coming in. But then a different calculation actually determines who's president. And because there are only two significant parties, it is entirely possible (as we know) for the people, acting directly, to choose Candidate A and his/her policies over Candidate B, yet have Candidate B take office nonetheless and implement the diametrically opposite policies. This poses problems of legitimacy that other systems don't face, especially in an era when questioning the legitimacy of even a president who uncontestably did win the election has become a standard political tactic. If you think things are ugly now, imagine what would happen if Barack Obama lost the popular vote next year to Rick Perry, yet still won the EC and stayed in office. While I would be grateful to the EC in that case for saving us from Perry, I don't think reactions up to and including armed revolt are out of the question. There's just no reason to keep a system that runs that risk, especially since (for reasons I won't explain now because I've already gone on too long), the EC does not fulfill the Framers' original intent at all, and was originally designed as a solution to problems that no longer exist.

  8. Because of the state-by-state winner-take-all electoral votes laws (i.e., awarding all of a state’s electoral votes to the candidate who receives the most popular votes in each state) in 48 states, a candidate can win the Presidency without winning the most popular votes nationwide. This has occurred in 4 of the nation's 56 (1 in 14 = 7%) presidential elections. The precariousness of the current state-by-state winner-take-all system is highlighted by the fact that a shift of a few thousand voters in one or two states would have elected the second-place candidate in 4 of the 13 presidential elections since World War II. Near misses are now frequently common. There have been 6 consecutive non-landslide presidential elections (1988, 1992, 1996, 2000, 2004, and 2008). 537 popular votes won Florida and the White House for Bush in 2000 despite Gore's lead of 537,179 popular votes nationwide. A shift of 60,000 voters in Ohio in 2004 would have defeated President Bush despite his nationwide lead of over 3 million votes. Some insider Republicans believe under the current system in 2012, President Obama could win the electoral vote without winning the popular vote.

    The National Popular Vote bill would guarantee the Presidency to the candidate who receives the most popular votes in all 50 states (and DC).

    Under National Popular Vote, every vote, everywhere, would be politically relevant and equal in every presidential election. Every vote would be included in the national count. The candidate with the most popular votes in all 50 states would get the 270+ electoral votes from the enacting states. That majority of electoral votes guarantees the candidate with the most popular votes in all 50 states wins the presidency.

    National Popular Vote would give a voice to the minority party voters in each state. Now their votes are counted only for the candidate they did not vote for. Now they don't matter to their candidate.

    With National Popular Vote, elections wouldn't be about winning states. No more distorting and divisive red and blue state maps. Every vote, everywhere would be counted for and directly assist the candidate for whom it was cast. Candidates would need to care about voters across the nation, not just undecided voters in the current handful of swing states. The political reality would be that when every vote is equal, the campaign must be run in every part of the country.

  9. Your argument doesn't really make sense to me. A candidate would obviously care more about winning all of Florida's votes than all of New Mexico's. But that means he will campaign harder in absolute terms, NOT in per capita terms. Florida has more delegates available, but also way more voters you need to convince in order to get those delegates. In fact, the electoral delegate payoff for a given amount of campaign money and effort is much higher in New Mexico because it has fewer voters per delegate.

    Here's another way to think about it: if the election was decided by the popular vote, where would candidates focus their efforts? They'd focus overwhelmingly on the cities because a) that's where most of the people are and b) the city-folk are conveniently grouped close together for you to talk to thousands at a time. No one's going to wander around doing appearances in New Mexico and Iowa when he can reach ten times more people with one appearance in Chicago.

    Finally, I haven't studied those correlations between winning big states and winning the electoral college very closely, but I have to say, OF COURSE that would happen! If you've just won 7 of the 10 biggest states in the nation, that's a huge portion of the votes already in your column. Winning the nation while losing most major states/cities would be extremely rare under almost any system. You'd have to completely monopolize the rest of the country.

  10. Under National Popular Vote, when every vote counts equally, successful candidates will find a middle ground of policies appealing to the wide mainstream of America. Instead of playing mostly to local concerns in Ohio and Florida, candidates finally would have to form broader platforms for broad national support. Elections wouldn't be about winning states.

    Now political clout comes from being a battleground state.

    Now with state-by-state winner-take-all laws presidential elections ignore 12 of the 13 lowest population states (3-4 electoral votes), that are almost invariably non-competitive, and ignored, in presidential elections. Six regularly vote Republican (Alaska, Idaho, Montana, Wyoming, North Dakota, and South Dakota), and six regularly vote Democratic (Rhode Island, Delaware, Hawaii, Vermont, Maine, and DC) in presidential elections.

    Support for a national popular vote is strong in every smallest state surveyed in recent polls among Republican voters, Democratic voters, and independent voters, as well as every demographic group. Support in smaller states (3 to 5 electoral votes): Alaska -- 70%, DC -- 76%, Delaware --75%, Idaho – 77%, Maine -- 77%, Montana – 72%, Nebraska -- 74%, New Hampshire --69%, Nevada -- 72%, New Mexico -- 76%, Oklahoma – 81%, Rhode Island -- 74%, South Dakota – 71%, Utah - 70%, Vermont -- 75%, and West Virginia – 81%, and Wyoming – 69%.

    Nine state legislative chambers in the lowest population states have passed the National Popular Vote bill. It has been enacted by the District of Columbia, Hawaii, and Vermont.

    Under the current system, the 11 most populous states contain 56% of the population of the United States, and a candidate could win the Presidency by winning a mere 51% of the vote in just these 11 biggest states -- that is, a mere 26% of the nation's votes.

    With National Popular Vote, big states that are just about as closely divided as the rest of the country, would not get all of the candidates' attention. In recent presidential elections, the 11 largest states have been split -- five "red states (Texas, Florida, Ohio, North Carolina, and Georgia) and six "blue" states (California, New York, Illinois, Pennsylvania, Michigan, and New Jersey). Among the four largest states, the two largest Republican states (Texas and Florida) generated a total margin of 2.1 million votes for Bush, while the two largest Democratic states generated a total margin of 2.1 million votes for Kerry. 8 small western states, with less than a third of California’s population, provided Bush with a bigger margin (1,283,076) than California provided Kerry (1,235,659).

    Under National Popular Vote, every vote is equal.

    16% of Americans live in rural areas.

    With National Popular Vote, big cities would not get all of candidates’ attention, much less control the outcome.
    The population of the top five cities (New York, Los Angeles, Chicago, Houston and Philadelphia) is only 6% of the population of the United States and the population of the top 50 cities (going as obscurely far down as Arlington, TX) is only 19% of the population of the United States.

    Suburbs and exurbs often vote Republican.

    If big cities controlled the outcome of elections, the governors and U.S. Senators would be Democratic in virtually every state with a significant city.

  11. Evidence as to how a nationwide presidential campaign would be run, can be found by examining the way presidential candidates campaign to win the electoral votes of closely divided battleground states, such as in Ohio and Florida, under the state-by-state winner-take-all methods. The big cities in those battleground states do not receive all the attention, much less control the outcome. Cleveland and Miami certainly did not receive all the attention or control the outcome in Ohio and Florida in 2000 and 2004.

    Because every vote is equal inside Ohio or Florida, presidential candidates avidly seek out voters in small, medium, and large towns. The itineraries of presidential candidates in battleground states (and their allocation of other campaign resources in battleground states) reflect the political reality that every gubernatorial or senatorial candidate in Ohio and Florida already knows–namely that when every vote is equal, the campaign must be run in every part of the state.

    Even in California state-wide elections, candidates for governor or U.S. Senate don't campaign just in Los Angeles and San Francisco, and those places don't control the outcome (otherwise California wouldn't have recently had Republican governors Reagan, Dukemejian, Wilson, and Schwarzenegger). A vote in rural Alpine county is just an important as a vote in Los Angeles. If Los Angeles cannot control statewide elections in California, it can hardly control a nationwide election.

    In fact, Los Angeles, San Francisco, San Jose, and Oakland together cannot control a statewide election in California.

    Similarly, Republicans dominate Texas politics without carrying big cities such as Dallas and Houston.

    There are numerous other examples of Republicans who won races for governor and U.S. Senator in other states that have big cities (e.g., New York, Illinois, Michigan, Pennsylvania, and Massachusetts) without ever carrying the big cities of their respective states. It is certainly true that the biggest cities in those states typically vote Democratic. However, the suburbs, exurbs, small towns, and rural parts of the states often voted Republican. If big cities controlled the outcome of elections, the governors and U.S. Senators would be Democratic in virtually every state with a significant city.

    Under a national popular vote, every vote everywhere will be equally important politically. There will be nothing special about a vote cast in a big city or big state. When every vote is equal, candidates of both parties will seek out voters in small, medium, and large towns throughout the states in order to win. A vote cast in a big city or state will be equal to a vote cast in a small state, town, or rural area.

    The main media at the moment, namely TV, costs much more per impression in big cities than in smaller towns and rural area. So, if you just looked at TV, candidates get more bang for the buck in smaller towns and rural areas.

    If the National Popular Vote bill were to become law, it would not change the need for candidates to build a winning coalition across demographics. Any candidate who yielded, for example, the 16% of Americans who live in rural areas in favor of a “big city” approach would not likely win the national popular vote. Candidates would still have to appeal to a broad range of demographics, and perhaps even more so, because the election wouldn’t be capable of coming down to just one demographic, such as voters in Ohio.

  12. If national popular vote ever came about, presidential candidates would spend their time and money campaigning in the 15 most populous urban areas of the country.

    This is both intuitive and pretty obvious, I'd say.

  13. It's not meaningful to talk about the 6 or 50 largest cities. What matters is the metro areas, and the large metro areas really do have a large share of the national population. So when I said candidates would focus on big cities I technically meant densely populated areas.

    When Obama makes an appearance in LA, he doesn't have a bouncer kick out the people from Pasadena.

    Also, with regard to your point about Ohio and Florida, I agree that candidates would visit suburban areas somewhat under either system. You have to remember though, that with the electoral college, and with senators and governors, they are only focusing on one or a few states, so they can cover them more extensively. In a true national campaign a presidential candidate would have to cover the whole country, so he'd have less time to visit lightly populated areas.

  14. The main media at the moment, namely TV, costs much more per impression in big cities than in smaller towns and rural area. So, if you just looked at TV, candidates get more bang for the buck in smaller towns and rural areas.

    Under the current system, they spend two-thirds of their time and money in just six closely divided battleground states; 80% in just nine states; and 99% in just 16 states. That's precisely what they should do in order to get elected under the current system, because the voters of two-thirds of the states simply don't matter. Candidates have no reason to poll, visit, advertise, organize, campaign, or care about the concerns of voters of states where they are safely ahead or hopelessly behind. Over 85 million voters are ignored.

    Charlie Cook reported in 2004:
    “Senior Bush campaign strategist Matthew Dowd pointed out yesterday that the Bush campaign hadn’t taken a national poll in almost two years; instead, it has been polling 18 battleground states.”

    Former White House Press Secretary Ari Fleischer acknowledging the reality in the Washington Post on June 21, 2009 said:
    “If people don’t like it, they can move from a safe state to a swing state.”

    If every vote mattered throughout the United States, as it would under a national popular vote, candidates would reallocate the money they raise. They could not afford to ignore any voter. Candidates would have the incentive and survival instinct to poll, visit, advertise, organize, campaign, and care about voters everywhere in direct competition with their rival candidate, as in every other election in the U.S. where the candidate with the most votes among all voters wins.

  15. Wow, if you were hip, I'd call you the Village Voice: "so if the polling in both is equally close, campaigns are going to focus on the Sunshine State. On the other hand, campaigns would rather focus on close New Mexico than GOP-lock Texas"

    You are defining "EC benefits" as campaigns spending time and money in your locale. That's ridiculously insider savvy-ness. I can't believe it, actually. Let's see, WashPo, TNR, are you shooting for Politico now?

  16. Sorry I haven't had a chance to get to all the comments...but I did want to jump back in for this one.


    Huh? It's not insider savvy-ness; it's standard variety political science. And, I'd say, common sense. I actually didn't say "spending time and money", although that's true; what's far more important is that campaigns will adjust their policy promises to the states that matter most. And *that* is a big deal; it (at least partially) explains why presidents from both parties in the mid-20th century tended to care a lot more about cities than presidents over the last twenty years have.

    Which gets to Andrew's comment,

    The question is what matters. For me as a voter, it doesn't really matter what the math shows about how many EV/people there are in my state; why should it? What matters is whether the incentives are for candidates to offer goodies to my state, or city, or directly to me. It's obvious that campaigns don't care at all about people in Wyoming, so they're screwed by the EC, while people in Florida and Ohio do very well from it.

  17. Okay, name some policy promises that Obama made to battleground states. That were kept.

    Next, provide some evidence that Eisenhower "cared" about cities more than Clinton and Bush 43.

  18. PS, on the reply to Andrew, I'm fine with you saying the actual electoral value of one vote doesn't matter - WY, CA, whatever, it's just one vote. But saying candidates offer goodies to states, cities, or directly to a voter to get one vote at a time is hogwash. If you weren't thinking retail politics, good, and that brings up a related point.

    Individual states, cities and voters are not interest groups. Goodies are offered to trade groups, unions (ha), industries, ideological groups, et al. for money or election support. What the hell would MN want from a candidate? Who would do the asking? What would MN give in return?

    And what are these Presidentially controlled goodies you're talking about? Any appropriation comes from the House and the local Rep. is far more significant. What could they be?

  19. Mark,

    I'm hardly a supporter of the idea that presidents control everything, but of course their policy views matter. Including, by the way, appropriations -- it certainly matters which things presidents include in their budgets and fight for.

    Hmmm...don't have a good Obama PA/FL/OH/NM/etc example that comes to mind, but as far as mid-century presidents: the obvious one is civil rights. When African Americans became a swing vote in big, close states, the presidential wings of both parties shifted to support civil rights, and were considerably more supportive than Congress. Of course, the complication with all of this is that nomination politics also matters, and often pushes presidents in different directions than general election considerations.

  20. "When African Americans became a swing vote in big, close states, the presidential wings of both parties shifted to support civil rights"

    This is the kind of evidence-free assertion that pundits make. Thankfully, you follow it up by saying things are pretty complicated, so who knows? Wait, I mean "worse" not "thankfully." As to the quote above:

    What big states (and when) did AA's become a swing vote? - meaning a decisive constituency that will decide the state's EC votes, not that the votes were up for grabs.

    What is a "presidential wing" of a party?

    Eisenhower never treated civil rights - pro or anti - as an election strategy, and besides, his record is decidedly mixed. I'm still trying to think of how you can even say the GOP shifted to support civil rights. Up to now, I mean.

    Kennedy didn't want to make pro-Civil rights his issue, because he knew, just like Johnson, that coming out strongly for Civil Rights meant losing the South for at least a generation. The Voting Rights and Civil Rights Acts were huge election losers for Democratic presidents.


Note: Only a member of this blog may post a comment.

Who links to my website?