a citizen’s journal by Thomas Nephew

The possible electoral college controversy ahead — and what *not* to do about it

Posted by Thomas Nephew on October 19th, 2012

Like most of news-following America, I suspect, I’ve been visiting a lot recently.  And while I don’t see a way to discern trends for the numbers, it’s my impression that Nate Silver’s multiple poll, multiple simulation-based estimates of the likelihood of controversial outcomes have been rising; as of today, those included…

  • Electoral College tie: 0.5%
  • Obama wins popular vote but loses electoral college: 1.9%
  • Romney wins popular vote but loses electoral college: 5.2%

As of mid-October, poll results suggest the chance of the US
presidential election winner not being the popular vote winner
is about the same as drawing a jack from a deck of cards.

Adding these up, the likelihood of the popular vote winner not winning the electoral college is 7.6 percent — about the likelihood of drawing a jack from a random deck of cards.

Of course, unlike in 2000, in 2012 it’s the Democratic candidate who is most likely to benefit from an electoral college override of the popular vote.  And while that may cause a little heartburn for some well-meaning people on the left, I’m going to argue it shouldn’t.

First of all, of course, that’s how George W. Bush won in 2000.*  A little historical balancing of the books isn’t a bad thing.

But it’s also high time to admit that in principle, the electoral college is a pretty good idea for a diverse, federal, continental-scale democracy like the United States.

First, it provides refuges for many voters to vote based on their convictions rather than for the so-called “lesser evil.”  Second, it preserves the need to compete for the majority of votes in actual political subunits of the country — the states — rather than merely in the aggregate national mass media market.  Finally, and perhaps most decisively, when states prove to be harming the exercise of the vote, the electoral college helps isolate democratic damage to the state or states responsible, and helps limit the remediation needed when corruption or irregularities occur.

Third party democratic refuge
The electoral college system of 51 state winners inevitably allows voters in some states greater flexibility than in others.  For example, in Maryland (where Romney has no chance at all of winning) disaffected but wavering left wing voters can more easily choose to protest against Obama and vote for a third party candidate, with much less concern than if they were in neighboring Virginia, where the race is closer.  This opportunity would diminish in a national popular vote election format.  Let’s say that voters tend to vote for their second choice rather than for their true first choice once the reported margin between the two most likely contenders in their voting zone is below, say, 2%.  Even when that’s the national margin — so that very few voters would take the risk —  inevitable state by state variance allows voters in some states to avoid that dilemma in the electoral college election format.

Bulkheads of federal democracy
The biggest problem with abolishing the electoral college is that you’re only half done. Without truly national, vigorously enforced voter registration, election, counting, and verification standards, a state or group of states could rig election processes to exclude or “underinclude” voting groups, fraudulently and undemocratically skewing the vote tallies they report. Other states, the ones running legitimate elections, would then face the perverse risk of designating presidential electors based not on their own legitimate results, but on fraudulent results elsewhere — and would have no recourse.  As I wrote in 2007 in opposing the National Popular Vote (NPV)** proposal  (emphasis added),

Under the NPV system, Maryland would routinely risk forfeiting its electoral votes to a candidate its voters didn’t favor — a candidate who necessarily only won elsewhere, in elections that were by definition completely unaccountable to Maryland voters.

This is no abstract, theoretical concern:  Florida was already known in 2000 for its voter list purges of ex-felons and, conveniently, people with similar names and addresses to ex-felons.  And there’s a systematic push to suppress the vote nationwide: the corporate-funded, right wing American Legislative Exchange Council (ALEC) is notorious for pushing voter suppression(/”identification”) laws nationwide to receptive, generally Republican state legislatures. The intent and the effect is to keep minorities and the poor off the voter rolls, making it easier for parties of the white or the rich to gain or keep their hold on power.

Like bulkheads in a ship that keep one flooding compartment from sinking the ship, the electoral college keeps serious flaws like these localized to the state in which they occur.  Given the arithmetic of a given electoral college tally, it may then turn out not to matter enough to tilt the election one way or the other.  By contrast, corrupt and undemocratic state vote counts will necessarily corrupt a single, national popular vote count.  Conversely, the electoral college also allows both fundamental remedies and one-time recounts to be pursued just for the states that need it.  With the electoral college, there was just a Florida recount in 2000.  With a national popular vote, there would be powerful incentives for recounts everywhere, despite the problems being clearly localized to Florida.

Mandate breadth
This seems counter-intuitive at first — don’t candidates routinely ignore states like Idaho or Alaska, and spend disproportionate time in relatively small- or midsized-population states like New Hampshire or Colorado?  Yes, they do — but some of that is due to the poor design of this electoral college rather than being an inherent feature; I’ll write more about that below.

Consider the alternative, though: with a simple national majority vote, candidates could — indeed would have to — simply compete in major media markets.  The country would shift from a corrosive red-state/blue-state divide to… an even more corrosive urban/rural divide.  In state by state races, rural areas can at least occasionally figure in the calculus of a path to victory.  In a national race, they never would.  And this point can and should be extended.  As I wrote in 2007, consider a scenario

“…where Candidate A wins 50 out of the country’s 51 states and jurisdictions by slim but consistent margins, say 51-49%. But Candidate B is a favorite in one large state, say… Texas. Or Florida. Or wherever. She wins by 60-40% there, and hence wins the popular vote. But just look at the map; personally, I would like Candidate A’s claim to the whole country’s leadership better than I like Candidate B’s — and so would quite a number of other voters, in quite a number of other states.”

The United States is a federation of states of continental scope; its electoral college imposes a reasonable, valuable, and I would argue very necessary breadth of victory requirement to become the president of that federal democracy.  The need for something like an American-style electoral college would actually be even clearer for the European Union, should it ever attempt to elect a single executive leader.

The real problem(s)
While the electoral college per se isn’t a problem, the design of the one we use certainly is.

One of the compromises in the original Constitution — needed to get small states to buy in to the result — was that the numbers of electors each state gets in the electoral college are derived from the sum of (1) its representatives in the House, allocated proportionally to each state’s population and, crucially, (2) its two senators, allocated to each state regardless of population size.  The number of electors is thus related to the state’s population, but not actually proportional to it.  It’s the skew from the “Senate part” of each state’s electoral count that gave Bush the election in 2000;  if each state had 2 electors fewer in 2000, Gore would have won the modified electoral college in 2000 (click for a spreadsheet showing this), and the Florida result would not have mattered.

Again, the point of a “House-based” rather than “Congressional delegation based” electoral college wouldn’t be to guarantee that a popular vote winner doesn’t lose the electoral college — though that result would probably happen less often.   The point would be to ensure that every American’s vote has an equal impact on the outcome, instead of having voters from Wyoming or Alaska (states so small they have only 1 representative, but therefore 3 electoral votes) have their votes count disproportionately far more than those from California or New York.  While the “solid South” poses a well-known political problem for the modern U.S, it’s the “empty West” of 1- and 2- and 3-representative states like Wyoming, Idaho, North Dakota, etc. that pose a democratic one by distorting the electoral map.

This post has been primarily about the electoral college, but in conclusion it’s important to return to some of the points alluded to above.  Far more than the electoral college itself — which, again, has real value, and could be made even better with a simple re-design — the real obstacles to democracy in the U.S are the lack of a clear, affirmative right to votevote secretly, and have that vote be verifiably counted.  Those issues are the ones to resolve first, while we preserve what’s best of the system we have.

FIVETHIRTYEIGHT UPDATES: 10/20: tie+Obama+Romney electoral-loss-despite-popular win estimate climbs to 8.2% (=0.4+1.8+6.0). 10/23: 8.3% (=0.6+1.8+5.9). 10/24: 9.0% (=0.7+1.8+6.5); 10/30: 8.0% (=0.4+1.9+5.7); 11/2: 6.7% (=0.3+2.0+4.4).
OTHER UPDATES, 10/22: Daniel Foster, at National Review Online, 10/19, The Old College Try : “Conservatives shouldn’t complain if Romney pulls a Gore” ; Jonathan Bernstein, 10/19, A Better Case for the Electoral College: “the interests which benefit from the electoral college are, on balance, those that are hurt by other factors in the overall system”;  both via Andrew Sullivan. 10/23: G. Magliocca, “Balkinization,” approvingly cites a Justice Jackson 1952 dissent: “To abolish [the Electoral College] and substitute direct election of the President, so that every vote wherever cast would have equal weight in calculating the result, would seem to me a gain for simplicity and integrity of our governmental processes.” The dissent was from a ruling upholding (but noting as nonbinding) Alabama Democrat efforts to prevent a”faithless elector.” 10/31: should mention that my friend and State Senator Jamie Raskin is squarely on the other side of this issue, most recently in a July article for ACSBlog: Singing the Electoral College Blues: Time for a National Popular Vote for President.

* Setting aside, for this discussion, the media consortium finding that Bush probably didn’t win the true Florida voter-intention vote count once spurious undervotes and overvotes were reviewed.
** NPV is a state-level, legislative compact to deliver participating states’ electoral college votes to the popular vote winner, regardless of the outcome in each participating state.  It has passed in Maryland and (I believe) in California.  Once an electoral college majority of states ratifies the plan, the electoral college as we know it would effectively be abolished, barring a successful court challenge to the process.

IMAGE CREDIT: Four jacks (22 January 2007, Enoch Lau, orig. filename IMG_1159.JPG), via

15 Responses to “The possible electoral college controversy ahead — and what *not* to do about it”

  1. oldgulph Says:

    With the current state winner-take-all system of awarding electoral votes, winning a bare plurality of the popular vote in the 11 most populous states, containing 56% of the population, could win the Presidency with a mere 26% of the nation’s votes!

  2. Thomas Nephew Says:

    It’s true, there are extreme scenarios for either side of this argument. I guess I see my scenario (1 state leans strong for a favorite son/daughter, the rest tilt the other way) as more likely than yours (TX, CA, NY, FL etc all tilt one way, the rest 100% the other?). But each example is intended to be illustrative, and of course something *like* what you describe could happen in less extreme form.

    But your focus is on the winner-take-all aspect, which you imply you’d like to do away with. NPV does that, of course, so I’ll assume you favor that over my “no Senate-based EC votes” suggestion — despite all my excellent points! 🙂

    But there are intermediate possibilities; ME and NE already kind of do this by (IIRC) assigning their “House” electoral votes based on the CD-by-CD results (winner take all, though) and the “Senate” ones based on the statewide results (also winner take all). So there’s a potential for a split electoral college vote from these states. It would be interesting to work out how Gore would have fared if all states did what ME or NE do.

    I could live with that approach, it preserves the values I see in the EC. Clearly, it’s a possibility, but just as clearly, only a few states have exercised it. The reason, I think, is that state governments prefer to maximize their potential value to the presidential winner by throwing all their support to one or the other. Many might prefer a “ME-NE” or a “no Senate EC votes” solution to an NPV one, though, should NPV approach critical mass.

  3. oldgulph Says:

    Dividing more states’ electoral votes by congressional district winners would magnify the worst features of the Electoral College system.

    If the district approach were used nationally, it would be less fair and less accurately reflect the will of the people than the current system. In 2004, Bush won 50.7% of the popular vote, but 59% of the districts. Although Bush lost the national popular vote in 2000, he won 55% of the country’s congressional districts.

    The district approach would not provide incentive for presidential candidates to campaign in a particular state or focus the candidates’ attention to issues of concern to the state. With the 48 state-by-state winner-take-all laws (whether applied to either districts or states), candidates have no reason to campaign in districts or states where they are comfortably ahead or hopelessly behind. In North Carolina, for example, there are only 2 districts (the 13th with a 5% spread and the 2nd with an 8% spread) where the presidential race is competitive. In California, the presidential race has been competitive in only 3 of the state’s 53 districts. Nationwide, there have been only 55 “battleground” districts that were competitive in presidential elections. With the present deplorable 48 state-level winner-take-all system, 80% of the states (including California and Texas) are ignored in presidential elections; however, 88% of the nation’s congressional districts would be ignored if a district-level winner-take-all system were used nationally.

    Awarding electoral votes by congressional district could result in third party candidates winning electoral votes that would deny either major party candidate the necessary majority vote of electors and throw the process into Congress to decide.

    Because there are generally more close votes on district levels than states as whole, district elections increase the opportunity for error. The larger the voting base, the less opportunity there is for an especially close vote.

    Also, a second-place candidate could still win the White House without winning the national popular vote.

    A national popular vote is the way to make every person’s vote equal and matter to their candidate because it guarantees that the candidate who gets the most votes in all 50 states and DC becomes President.

  4. Thomas Nephew Says:

    “A national popular vote is the way to make every person’s vote equal and matter to their candidate because it guarantees that the candidate who gets the most votes in all 50 states and DC becomes President.” I get that that’s your goal and that of NPV advocates — I just dispute that it’s as virtuous as you think it is, for reasons I’ve gone into but that I imagine you’ll continue to ignore. But I take your word for it that a CD-by-CD electoral college would have magnified the popular-electoral contradiction; magnifying the probability of that contradiction isn’t my goal, so I’m happy to drop the idea.

    I think it’s actually a good thing for candidates to focus on ‘battleground states’ and not have to spend time on ‘sure thing’ states. If Obama is ahead by 15 points in CA, that pretty much indicates his job there is done, he’s doing fine by CA. Of what benefit is it, even to Californians, for either him or Romney to campaign there? If Obama starts saying stuff elsewhere that makes Californians go “hey wait a minute”, that’s another thing, but then that will show up in CA polls and he may decide it wasn’t such a good idea.

    What I think you choose to overlook is that candidates will pick and choose where to spend their resources and efforts no matter which electoral system is adopted. A pure NPV system will simply shift campaigns from one battlefield — 50/50 states — to another: metropolitan areas. What you’ll “gain” is media blitzes in every major metropolitan area rather than just those in battleground states, because gaining 100 votes in Houston is suddenly just as appealing as gaining 100 votes in New York or Columbus, OH. But what you’ll lose is any interest whatsoever in appealing to areas with many rural voters, because the return on investment will always be lower.

    I think NPV presidential elections would become even more expensive, and would wind up focusing on major population centers. I prefer an electoral system that reflects the fact that this is a federal democracy, not one that pretends the US is a single, giant county with a single set of voter registration and election laws.

  5. oldgulph Says:

    With National Popular Vote, every vote would be equal. Candidates would reallocate the money they raise to no longer ignore 80% of the states and voters. […]

    [The remainder of this and 2 subsequent comments totaling some 1700 words moved to a Google Drive document, here, per this blog’s comment rules. This is not your blog, it is a place to leave manageable, succinct, conversational comments.]

    […] There is nothing incompatible between differences in state election laws and the concept of a national popular vote for President. The U.S. Constitution specifically permits diversity of election laws among the states because it explicitly gives the states control over the conduct of presidential elections. The National Popular Vote compact is patterned directly after existing federal law and preserves state control of elections.

  6. Thomas Nephew Says:

    Thank you for your comments, which I’ve largely parked as a separate document here; they were far too long to qualify as conversation, which is the goal of this comments section.

    If the NPV does what you say — preserves state control of elections — then I think it’s fatally flawed because groups of states (e.g., the former Confederate States) could well engage in systematic voter suppression to affect the popular vote, and my state’s electoral votes (and hence my own) would be burned on the bonfire of illegitimate results elsewhere.

    Your mantra is “every vote would be equal.” I’m far less concerned with absolutely precise equality of votes than with ensuring voting processes that are accountable to the voters involved. NPV by its very definition is not such a process: I have no effect whatsoever on voter registration or election policy in any state other than my own. My suggestion to remove Senate-based electoral votes makes votes substantially equal from state to state. One could further refine the EC so that state electoral votes were precisely proportional to population, meaning every vote would be more equal yet. But that is not the only goal of a fair election in a federal democracy.

    We disagree about the likely effect of NPV on campaign behavior; perhaps it’s unknowable in advance. I think the link you provide — Campaigns Blitz 9 Swing States in a Battle of Ads — demonstrates what I’m saying more than it does what you’re saying, because the reason the candidates are in those states in the first place is because of the EC effect. I’m not surprised the campaigns then gravitate to relatively small towns like Des Moines — that’s as big as it gets in IA.

  7. Nine Inch Bride Says:

    It occurs to me any winner-take-all scenario is destructive to the formation and growth of third-parties, which is to us the only way out of twin capitalist party stasis. And if the electoral college were truly representative, we would have proportional multiparty system as is common across the democratic globe, and there would then in fact be no need for it or benefit from it.

    The changes needed to the electoral system will never come about from the two party system which it serves.

  8. Thomas Nephew Says:

    Thanks for commenting — interesting web site! …and of course I’m curious about the name. 🙂

    As far as 3d party formation goes, I agree: winner-take-all is a huge obstacle for the presidency, state-wide office and even smaller-region office. As far as Congress or legislatures are concerned, I’d prefer proportional representation within states for exactly the reason you give, as a way of encouraging third parties, and diversity of political opinion. (I think there’s actually no textual prohibition against that in the Constitution; states get representatives based on population but are not directly commanded by Article 1, Section 2 of the US Constitution to assign those representatives to particular geographic areas.) I also agree that it’s moot in the short run; the two major parties work together and on their own to prevent 3d party formation.

    But I think electing a single executive/commander in chief, charged with 24/7/365 decisions defending and implementing the laws of a democratic, continental-sized federation is a different issue. One needs a selection process that satisfies the federation’s components, so that they willingly join the federation, and don’t later organize to leave, and that likewise gives their populations a sense of both electoral autonomy and say-so in the presidential selection.

    Obviously, the alternative is choosing the head of the majority legislative party, a prime minister or chancellor. It seems to me that usually works in the long run only if you either (1) have a relatively small, relatively homogeneous electorate that accepts the leader rather than finds him/her +/- a foreigner or (2) the executive is weak.

  9. Nine Inch Bride Says:

    There may be something to what you say about the presidency, were it not for the other problems besetting, and it seems NPV better paves the way for third parties there too. The idea of states not “joining the federation” was decided in the Civil War, which I can’t imagine over NPV vs. Electoral College or feeling disenfranchised somehow. The same people from the same states are voting either way, and the appearance, if not the reality, is that the middlemen serve no purpose apart from preserving the status quo ante multiparty proportional elections. I’m still not clear what you’re getting at.

    Thanks. Read the book if you’re curious about the name. If you’d care to review it, I’ll send along a complimentary copy.

  10. Thomas Nephew Says:

    “still not clear what you’re getting at.” My fault. Imagine there are 10 states in a country, and they like that, and they like running their own elections. 2 of them, however, run a massive voter intimidation and/or ballot-stuffing scheme sufficient to tilt the popular vote to a preferred candidate. Under NPV, it’s over. Under EC, not necessarily. This is essentially what happened in 1876: black voters were terrorized out of participation in the South, but the effect was limited to the electoral votes of those states (esp. SC, FL, LA, and esp MS iirc). SC, FL, LA still registered for Hayes, but with much smaller majorities than were truly the case, while MS was stolen outright. Hayes had a small legitimate EC majority and won, Tilden a stolen NPV majority and lost. Good.

    Put more succinctly, NPV absolutely requires a uniform, federalized voter registration and election process (in my view), EC does not (in my view).

  11. Nine Inch Bride Says:

    Thanks for clarifying. You have historical depth I am lacking, and a passion for the subject that it occurs to me were put to better use than constructing an argument against NPV because of sinister possibilities from 1876. Sure, if they can rig ballot machines to subtract votes instead of adding them, as in EC Florida, or via any other nefarious scheme, they can do that any which way. I think NPV wins in any tradeoff simply because we do not have a valid democracy in a one party system with two flavors, and the other problems, third-party making, proportional vote, etc. can only come through the establishment and victory of a third-party, or conceivably the implosion of one or both of the two. There are better arguments to pursue which do not detract from what’s necessary in the larger scheme, that’s all I’m getting at. The relative merits of EC is a small point in a much bigger question.

    I recommend you read the novel, “Nine Inch Bride,” contact through the site, or follow @nineinchbride. A slew of tweets from the book on “democracy” just this morning…

  12. oldgulph Says:

    The current state-by-state winner-take-all system of awarding electoral votes maximizes the incentive and opportunity for fraud, coercion, intimidation, confusion, and voter suppression. A very few people can change the national outcome by adding, changing, or suppressing a small number of votes in one closely divided battleground state. With the current system all of a state’s electoral votes are awarded to the candidate who receives a bare plurality of the votes in each state. The sheer magnitude of the national popular vote number, compared to individual state vote totals, is much more robust against manipulation.

    National Popular Vote would limit the benefits to be gained by fraud or voter suppression. One suppressed vote would be one less vote. One fraudulent vote would only win one vote in the return. In the current electoral system, one fraudulent vote could mean 55 electoral votes, or just enough electoral votes to win the presidency without having the most popular votes in the country.

    The closest popular-vote election in American history (in 1960), had a nationwide margin of more than 100,000 popular votes. The closest electoral-vote election in American history (in 2000) was determined by 537 votes, all in one state, when there was a lead of 537,179 (1,000 times more) popular votes nationwide.

    For a national popular vote election to be as easy to switch as 2000, it would have to be two hundred times closer than the 1960 election–and, in popular-vote terms, forty times closer than 2000 itself.

    Which system offers vote suppressors or fraudulent voters a better shot at success for a smaller effort?

  13. oldgulph Says:

    [repeats points made here — TN]

  14. oldgulph Says:

    More than 2,110 state legislators (in 50 states) have sponsored and/or cast recorded votes in favor of the National Popular Vote bill.

    The bill has passed 31 state legislative chambers in 21 states. The bill has been enacted by 9 jurisdictions possessing 132 electoral votes – 49% of the 270 necessary to go into effect.

  15. Thomas Nephew Says:

    @9 inch bride: (1)“The relative merits of EC is a small point in a much bigger question.” I quite agree, that’s why I closed with this: “Far more than the electoral college itself — which, again, has real value, and could be made even better with a simple re-design — the real obstacles to democracy in the U.S are the lack of a clear, affirmative right to vote, vote secretly, and have that vote be verifiably counted. Those issues are the ones to resolve first, while we preserve what’s best of the system we have.” (2) Of course 1876 is a long time ago. But you had claimed the Civil War settled things, I think I showed it didn’t. The methods used in 1876 are more extreme than those contemplated today such as voter ID and voter reg. purges, but the intent and effect are similar. With NPV, it becomes tempting to “run up the score” illegitimately in places willing/able to do so; EC avoids that temptation by making its success in changing election outcomes less likely (IMO).

    @oldgulph: (1)The NPV update was worthwhile and brief, thanks; looks like things are going well for you. (2)2000 wasn’t the closest EC vote in history, 1876 was. (3) Reasonable people, e.g. Mr./Ms. 9IB, can differ on whether EC or NPV are better for our federal system, I think. We’ve both used examples to illustrate our points; I think many of yours (e.g., “1 fraudulent vote deciding 55 electors”), while certainly possible, are much less likely than mine (e.g., 1876). Your point about the temptation for fraud and the need for smaller numbers of illegitimately cast/suppressed votes in swing states is well taken; there are bad scenarios for both EC and NPV. What I like with EC is that I have some influence on keeping *my* influence on the election intact (by insisting on clean elections in my state). I don’t with NPV. But others will decide for themselves.

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> -- (comment rules)