Featured Post

Electoral College Concentration

The recent election has understandably drawn attention to the Electoral College’s role in selecting American Presidents. One argument presented in its favor deserves particular attention: that it rewards candidates for having geographically broad support.

This argument, it turns out, is not simply wrong but directly opposite the truth: the Electoral College grants an advantage to candidates with geographically concentrated support. This is clear from both a simple toy model and from the actual history of American elections

Theory

The current practice in the Electoral College is found nowhere in the Constitution; instead, it is the result of early 19th century practice. While the Constitution makes no demands on how electors are chosen and arguably forbids them from being bound by state law to support one candidate or another, in practice almost all electors are awarded to the winner of a plurality of the vote in their state. This creates obvious benefits to concentration.

Suppose a toy electoral college of three states, each containing 10 voters and each casting 1 electoral vote for the winner of the popular vote. In one state, candidate A wins 8 votes; in the other two, Candidate B wins with 6 votes in each. Candidate B wins the electoral college 2 votes to 1, in spite of losing the popular vote 16-14. But which candidate has the more geographically dispersed support?

While this is ultimately a subjective determination, I claim that candidate A’s support is more widely dispersed. First, his floor was higher, winning a minimum of 40% compared to B’s minimum of 20%. Second, fully half of his voters live in states won by candidate B, whereas only 14% of B’s voters live in states won by candidate A. Indeed, in this toy example, a popular vote-electoral vote inversion must rest on exactly this sort of distribution, with a candidate winning small majorities in two states while being crushed in the third.

The opposing argument is that victory in two states rather than one demonstrates dispersion, while winning a state 8-2 demonstrates excessive concentration. Considering the argument’s inverse, however, demonstrates its absurdity: this argument claims that candidate B’s support is shown to be dispersed by virtue of his weakness in the state he lost.

The actual system used in American elections is more complex than this toy in a number of ways, but an examination of actual Electoral College results shows exactly this pattern: the Electoral College boosts regional candidates and harms candidates with broadly dispersed support.

Case 1: The 1860 Election

The 1860 election is the most dramatic and straightforward case of geographic concentration in the Electoral College. Abraham Lincoln, running on a free soil platform, was unable to even get on the ballot in 9 states (all of which would go on to secede) and won single-digit percentages of the popular vote in three more; he won the election on the strength of 15 outright majorities, pluralities in California and Oregon, and a close fought race in New Jersey that split its electoral votes 4-3 in his favor. Lincoln won no electoral votes in a slave state, while taking 180 out of 183 in free states.

Lincoln’s opposition was fragmented between three major candidates: two Democrats, Stephen Douglas and John Breckinridge, and Constitutional Union nominee John Bell. Of these three, Breckinridge finished second in the Electoral College by winning outright majorities in 6 states and pluralities in 4 more. Bell finished third, with pluralities in three states.

Stephen Douglas, however, beat out both of these candidates in the popular vote but paid the price for dispersed support in the Electoral College: he only won electoral votes in Missouri, where he pulled a plurality of just over 35%, beating Bell by fewer than 500 votes, and New Jersey, where he won the 3 electors who did not go to Lincoln. He was the only candidate to win electoral votes from both free and slave states, and he finished with at least 10% of the popular vote in 19 states, only 1 shy of Lincoln’s mark in the same measure and ahead of Breckinridge (17) and Bell (15).

While Lincoln benefited from his opposition fragmenting between three candidates, this was not the primary cause of his electoral strength: he won outright majorities in states worth 164 electoral votes, well beyond the 150 needed to win the election. While some of these were close enough to plausibly be endangered had the Democrats been able to field a single candidate against him, the Republican base in the North would have made him competitive, at the very least, with a single national opponent.

Case 2: Third Party Candidates

In the postwar period, most third party runs have been minor sideshows; on two occasions, however, they have won electoral votes. Both of these candidates exploited the growing regional rift within the Democratic Party to win a few states, while mostly or entirely neglecting the rest of the country.

In 1948, Strom Thurmond, then Governor of South Carolina, walked out of the Democratic Convention due to a platform fight over civil rights. With the support of a number of Southern Democrats, he ran for President as the candidate of a splinter faction in the party. Thurmond was the archetype of a regional candidate: he was on the ballot in only 13 states, all former slave states, and he won only 2.4% of the total popular vote. However, as the official Democratic candidate in four states, Thurmond was able to win 39 electoral votes.

Twenty years later, another Southern candidate, George Wallace, followed Thurmond’s path to winning votes in the Electoral College. Wallace’s support was much more widespread than Thurmond’s: he was on the ballot in all 50 states, and he won more than 13% of the popular vote. However, in the Electoral College, he demonstrated somewhat less success than Thurmond, winning only 5 of the 6 states carried by Thurmond in 1948 (South Carolina went for Nixon) and 46 electoral votes. Wallace’s support, though broader than Thurmond’s, was still very heavily concentrated in the Deep South: his victory in Alabama with nearly 66% of the vote was a higher share of the popular vote than either major party candidate won in a single state (though Washington DC went for Humphrey with over 80% of the vote), but he failed to reach double digits in 27 states and DC.

Thurmond and Wallace’s success in the Electoral College stand in stark contrast to the most successful modern third party run in the popular vote, Ross Perot’s 1992 campaign. Perot led national polls at one point and finished with nearly 19% of the popular vote in spite of taking a hiatus from campaigning for a significant part of the summer; however, lacking the strong regional base of the Southern candidates, Perot did not manage a single electoral vote. In stark contrast to Thurmond and Wallace, Perot managed at least 10% of the vote in 49 states; however, his best state, Maine, only gave him a touch over 30%, leaving him 8 points behind Bill Clinton.
Case 3: The Solid South
The Republican Party was born a regional party of the North and West, and while it expanded into the South during Reconstruction, winning a number of southern states in the 1868, 1872, and 1876 elections, that same experience ensured that the white elites that reasserted control after the end of Reconstruction would overwhelmingly support the Democratic Party.

Yet the GOP’s lack of competitiveness in a significant region did not prevent the party from winning Presidential elections; beginning with the election of 1876, the first election after a significant number of states had begun rolling back black enfranchisement, and ending with the election of 1944, before Strom Thurmond’s Dixiecrat revolt marked the beginning of the end for the Democratic South, Republicans won 10 Presidential elections to the Democrats’ 8. Though the Republicans were able to make headway in the South by the 1920s, with Herbert Hoover winning 5 former Confederate states in his 1928 landslide, they never penetrated a core from Arkansas and Louisiana in the west to South Carolina in the east. However, with few votes wasted in safe Democratic states, they consistently ran ahead of their popular vote totals in the Electoral College, including winning the 1876 and 1888 elections in spite of losing the popular vote.

Conclusion

The Electoral College may have virtues; however, by granting a state’s full electoral heft to the winner, even if by a bare plurality, it favors regional candidates over national. This is ordinarily not a major issue: no election has been close enough to be thrown to the House by third party candidates, and even during the era of the Solid South Republicans contested enough of the country to win elections and claim broad support. Nevertheless, anyone concerned about rewarding candidates for geographically concentrated support should oppose rather than support the College.

Feature Image by mccun934


Guest Author

Andrew P Larson is everything wrong with libertarians on the internet. He is on Twitter at @applarson.

Please do be so kind as to share this post.
TwitterFacebookRedditEmailPrintFriendlyMore options

84 thoughts on “Electoral College Concentration

  1. Or instead of winner takes all, the electoral college could cast electoral votes in proportion to the share of votes in that state. Currently, there is no written rule that constrains how electors may cast their ballots. But, introducing such a rule would reduce the problem that you are concerned with. The advantage* over direct election of the president is that this prevents dense urban centres from overwhelming rural areas.

    *At least insofar as some states really don’t have dense urban centres, and those states have disproportionately more electoral votes. The caveat being that it is a feature and not a bug of the system that rural regions get a disproportionately large voice.

    Report

    • I think a case for the EC explicitly based on the premise that rural voters should count more than urban would be a better case than the one I’m arguing against here, insofar as it would focus on a real feature of the system.

      A proportionately allocated EC would closer track the popular vote, bringing the system even closer than it currently is to direct election; I don’t really see a point to retaining the EC at all at that point.

      Report

      • Depends. Suppose you have three areas A, B and C. A is urban and has 100 people while B and C are rural and have 20 each. Each has 7 electoral votes. Suppose a candidate got 40 votes in A, but gets 14 votes each in B and C. So she has lost the popular vote. However, she gets 3 electoral votes for A and 5 electoral votes each in B and C. This means she wins the electoral college. By your own criteria, she has broader based support than the opposition. That is to say, an electoral college which disproportionately allocated electors to states but which required electors to cast votes in proportion to the popular vote in that state will get you a system where the one with more broad based support wins (despite lacking overall popular support)

        Report

        • It’s probably worth clarifying that I don’t think geographic breadth of support should be a goal of designing a system; I’m just annoyed by people who profess to believe it and then support a system that works against it.

          Your example makes sense, though the rural weighting still seems to be where the action is, and the more you reduce that weighting the more the system approximates a simple popular vote.

          Any proportional allocation would presumably involve a threshold for receiving votes, which I suppose would mildly reward a candidate like Douglas for having at least some presence in most states.

          Report

      • A fact which eludes a surprising number of people. There’s also “coastal elites” as some sort of minority of America (something close to half of Americans live within a few hours drive of the ocean) and the weird belief that California represents some tiny fraction of America (as if they were South Dakota), instead of housing what — a tenth or more of the US population and almost 15% of the economy?

        Report

          • Both parties do about everything they can to tilt the odds in their favor. For whatever reasons, Republicans have just been better at it. If Democrats want to do better, the first thing they should do is figure out what they’re doing wrong and figure out how to do it better instead of complaining about how unfair it all is like some emo teenager.

            Report

            • Accident. The GOP had a big win in 2010 (backlash against single party control in Washington and the fact that the Democrats hadn’t fixed the worst recession in almost a century in 18 months or so), and that was a census year.

              The first census year with modern analytic methods and very, very powerful computers. It was the first time one could approach gerrymandering scientifically — it wasn’t a handful of well-trained experts that only had so much time, it was an engineered, mathematical approach to optimize district design on various priorities.

              Effectively, take the best gerrymander map-maker in all time, pair him with the best statisticians, and let him not only design every district in a GOP controlled state (Federal and State districts), but let him simulate millions of elections in each to test them.

              AFAIK, Democrats never even bothered to develop a tool like that. The GOP does, and it’s efficient as all hell.

              Report

                • Clearly analytics isn’t your field. You don’t think they optimized for just one election?

                  Or that they were strangely lacking demographic information, including trends?

                  I’m afraid you sell the GOP quite short here. This wasn’t a TV plot of the week — these were professionals planning for a ten year cycle.

                  Report

                  • morat20,
                    You sell the GOP too highly, as well. You CAN’T plan for trends. It doesn’t work.
                    55/45 is the split for classic gerrymandering. That’s enough that it’s too high to swing based on pretty much anything other than DeadGirls or LiveBoys.
                    Six years in, and the greedy bastards get 2006, when all of those fun trends are kicking in, and with a decent PR campaign, all the applecarts get upset.

                    Report

                  • You don’t think they optimized for just one election?

                    I actually mentioned two elections in my original post.

                    What I am saying is that things change over time, people move, people move back, people get born, people die, and, most importantly, new technologies show up and change things.

                    Just because Donald Trump can be easily analogized to The Mule doesn’t make the rest of the republicans easily analogized to Hari Seldon.

                    The Gerrymandering argument is very persuasive for the first few elections after a successful gerrymander.

                    It is not persuasive at all when it comes to the last election before it’s time for a new census.

                    Even if the Republicans were using computers.

                    Report

                    • What I am saying is that things change over time, people move, people move back, people get born, people die

                      “Or that they were strangely lacking demographic information, including trends?”

                      Report

                      • I’m down with guessing trends for a couple of elections.

                        100%.

                        You think that they can successfully guess where we’d be in 2018 and planned that into the 2010 plan?

                        I suppose, technically, they *KNEW* that they’d also be running against Hillary but they probably also *KNEW* that the Jeb wing of the party was going to be in charge rather than the Trump wing.

                        I mean, when Kid Rock unseats Debbie Stabenow in 2018, that’s going to be evidence of a kind of trend that 2010 would not have foreseen (not that Senate votes are gerrymanderable, of course, but the *TREND* is one that no one in 2010 would have foreseen).

                        You’re effectively saying that computers guessed the populism wave.

                        And I’m saying that, no.

                        I still don’t know what is going to happen in the 2018 election but I do know that if the Republicans win, it will be because the Democrats are doing stupid and avoidable thing rather than because the Republicans just had that many quality analytics people working for them following Obama’s election.

                        Report

                        • Rather than go into a lengthy explanation about basic gerrymandering and demographic voting patterns — something you’ve indicated you’ve understood in the past — let me simply say this:

                          Their maps neither attempted to nor needed to account for populist waves. That’s not how gerrymandering works. (If it did, Democrats wouldn’t have done so well in 2006 and 2008). It has nothing to do with Jeb Bush, Hillary Clinton, Kid Rock, or populism.

                          The fact that you think it does, that you think my point requires it does means you’re so far off the point that you couldn’t find it with a map and a telescope. Which means there is no point in continuing this.

                          Report

                          • That’s not how gerrymandering works.

                            But gerrymandering is thrown into chaos and turmoil when things like “waves” happen.

                            It’s all well and good to set up a handful of 55/45 districts until someone shows up and EVERY SINGLE ONE of the 45 people show up to vote and only half of the 55 people do.

                            When that happens?

                            Your gerrymandered district is worth two things: Jack and Squat. And Jack left town.

                            It results in the so-called “wave” election.

                            The Dems caught a bad break in 2010, sure. They caught the wrong end of a wave before a census.

                            But if they can’t overcome the 2010 gerrymandering in 2018? Their problem is *NOT* that they can’t overcome how the districts were drawn.

                            It’s because they don’t have a message to get every single one of their 45 people to the polls despite only half of the 55 people bothering to show up.

                            Report

                            • It’s like you have a mental block. It’s rather fascinating.

                              Like you can put the bits together, but there’s some weird bit that simply blocks the conclusion.

                              It’s like talking to a man who knows what a golf handicap is, but then objects to you saying that it indicates which player is better on average because “The other player beat him once!”.

                              Or perhaps someone who grasps a coin-flip is 50/50, but keeps saying “One time, I flipped heads five times in a row, so clearly math doesn’t work”.

                              It’s almost like the notion of gerrymandering having any effect on the last election is against your religion or something. You dance around the basics of gerrymandering — I mean so basic as in “that’s why people do it at all” — like you’re tap dancing on a really touchy point of religious doctrine.

                              Report

                            • Jay,
                              10points is DeadGirl or LiveBoy or WeVotedOurselvesARaiseon the Fourth of Fucking July territory.

                              That’s EPIC, not just a “wave” election.

                              By 2018, you’ve got enough districts that are at 5pt swing (which is changeable), to have some fun.

                              And, hell, without Clinton holding the reins, maybe someone‘s idea of fun won’t be spiting her.
                              (To be fair: I too would not like to go to Washington because she won).

                              Report

                              • Kimmi, the thing you claim is “EPIC” is something that I have seen twice in the last handful of elections.

                                In 2006, I saw a wave.
                                In 2010, I saw a wave.

                                (We could discuss whether 2008 should count because the Republicans got pantsed in the House in that one too, of course.)

                                2006 and maybe 2008 and definitely 2010 are two (and maybe three!) examples of this sort of thing happening when you’ve got every single one of the 45 showing up and half of the 55 staying home.

                                This happens.

                                It *COULD* happen again in 2018.

                                But it won’t if the Democrats run in Flyover on a platform that would be a surefire winner in Hipster Urban Enclave.

                                Report

                    • So gerrymandering is only useful for the first few elections after a census. Would that be three elections? A few sounds like three but i’m not really a number knower guy.

                      Report

                      • 2000, 2002,2004. Well, that seems to be empirically proven.
                        Of course, 2006 was when you had a… mastermind providing assistance to the Democrats. ;-)
                        2018 is probably a bad year for 55/45 districts circa 2010.

                        Report

                        • “The Gerrymandering argument is very persuasive for the first few elections after a successful gerrymander.”

                          12,14,16= three elections. Three seems to be pretty darn close to approximating a few.

                          I could also add that you have even said the large D losses might be 30-40% attributable to gerrymandering. Right? So these two things clearly prove gerrymandering is a non-issue.

                          Report

                          • Greg, I’m not arguing over whether gerrymandering is a non-issue in 2012, 2014, or, hell, even 2016.

                            I’m arguing that, in 2018, you’re past being able to point to the computer models they used in 2010 as being the reason you can’t flip an election in 2018 the way that you did in 2006 or 2008.

                            Report

                            • That’s nonsense also, If a gerrymander favors one party that party has an advantage. If a district is heavily weighted for D’s or R’s then there is almost no competition. There aren’t’ a huge number of competitive districts. But if a district is gerrymandered to be R +7 that makes it much harder for them to lose even in a wave year. If R’s are favored due to a gerrymander they continue to have that advantage. Most voters are predicable. R’s vote for R’s and D’s vote for D’s. It’s not like it wasn’t’ R’s voting for Trump or D’s voting for Clinton. It is some small subset of voters that are up for grabs. In some districts that subset might not even be enough to sway an election.

                              Report

                                  • I think that assuming the level of competence on the part of the Republicans that is being assumed is simplification to the point of laughability.

                                    But, hey. Maybe we’ll be able to point to the 2010 gerrymandering in 2020.

                                    The computers coming down the pike will be *EVEN BETTER*.

                                    “The Republicans used *QUANTUM COMPUTERS* to gerrymander! Democrats didn’t stand a chance!”

                                    “They ran Hillary Clinton on a platform of ‘I beat him last time, I can beat him again!'”

                                    “You’re sexist! And racist!”

                                    Report

                                  • At this moment, I wouldn’t be surprised if he claimed the partisan index of a district made no difference in elections, even though what it measures is basically pretty much exactly the point of gerrymandering.

                                    Report

                                    • “The partisan index is also a thing that stays the same for 8 years. Or it gets worse! They have computers that can predict such things with pinpoint accuracy now. Which is why the Democrats don’t have to change. They wouldn’t have won those elections anyway.”

                                      Report

                  • Or that they were strangely lacking demographic information, including trends?

                    Making a district voter-proof for the next election after the census is easy, but it’s increasingly harder the further you move forward in time.

                    A person in the United States is expected to move 11.4 times in his lifetime. https://fivethirtyeight.com/datalab/how-many-times-the-average-person-moves/

                    That doesn’t mean that 100% of the voters will have changed in 6 years (that would be a misuse of math), but it does mean making predictions is very hard, especially of the future.

                    Report

          • The lesson I would take from 1860 is that the Democrats had lost the House in 1858, but were still operating under the assumption that they were truly the popular party based upon history and their projections of how their success was reassured. Confirmation bias. It can last generations. Particularly if you think the electoral system is unfair.

            Report

          • LeeEsq:
            Republicans know that they would be screwed in a fair election system so they do all the things in their power to make it unfair.

            As recent as 2000 we saw the Dems arguing “fairness” required the changing of the election rules after the election had been run. The Supremes said 7-2 that it didn’t.

            Somehow I think the Dem (or GOP for that matter) definition of a fair election is one where they win.

            Report

      • Dude, I’m not making a normative claim, just engaging in speculative system design. You could allocate the number of electors each state receives in proportion to their land area instead of population size. It may not be particularly fair, but it is a possible institutional design.

        You could do it entirely in terms of population size, but if those electors also vote in proportion to the popular vote in each state, that would make the electoral college redundant.

        You could do a more mixed model which amplifies the rural vote more but not as much as it would if you allocated purely according to land area. Something like this would achieve the broad based support thing that Andrew is after.

        I’m not claiming that it is something we ought to do.

        Report

      • But the principle undergirding the EC is (badly stated as) “land gets a vote”. Is that an old and busted idea? Something that should be deprecated? I don’t really know.

        It’s encoded into the popular election of the Senate. The more-popularly-elected House is the demographic model for the Electoral College.

        Not to defend the results of the 2016 election, but the sheer inability of many people to come to grips with the outcome demonstrates that “land gets a vote” represents some sort of information.

        If this is to … I dunno, tip the scale? in favor of coastal population centers, then that’s another swing and a miss at what happened in 2016. What happened in 2016 was that a certain type of media usage won. This seems rather a 19th century solution to a 21st century “problem”.

        Report

        • Land gets a vote isn’t correct; small population is favored but not large area. Washington DC, Vermont, and Alaska each reported about 300,000 votes in the last election, and they each delivered 3 electoral votes to the winner. All three were favored over California, in which 14 million votes were cast and which delivered 55 electoral votes to the winner (about 250,000 votes per EV), Florida, in which 9.4 million votes were cast to allocate 29 electoral votes (324,000 votes per EV), or Texas, in which 8.9 million votes were cast to allocate 38 electoral votes (234,000 per EV).

          Report

          • When was the last time we had a candidate visit Alaska for the purposes of getting its electoral votes? Has that ever happened?

            No one pays attention to these “favored” places during the election because they’re insignificant. Their “favored” status is a rounding error at the bottom. If we’re going to count Alaska’s over representation as the big flaw in the system then there is no case.

            Report

            • This is a core of pretty good criticism of the EC. What it really does is give the power to big and medium sized electoral units on geographic/demographic borderlines. So the likes of Ohio and Florida count for a lot while both Alaska and California are disfranchised.

              This would be the core of my criticism of the EC as a system. It doesn’t really do the things its defenders say it does. It essentialy justs gives and takes away power to certain voters for essentially random reasons.

              The best criticism seems to be recount fears, but that’s more an indictment of how bad Americans are at running elections compared to other mature democracies, who can pull off a national popular vote without much hassle. Its effectively an admission that you’re collectively far worse at running elections than the French. That its a plausible argument at all should be a source of national shame.

              Report

  2. Andrew,
    So…..Are you recommending:

    1) Reforming the EC
    2) Scrapping the EC
    3) Something else ?

    Because all I’m seeing is “concerns”. I’m not seeing any major issues or flaws or any corrective actions suggested to remedy those flaws.

    Report

    • I don’t think EC reform should be a priority; the current system works well enough. If elections were consistently regionally polarized to the degree of the 1860 election, I might reconsider this.

      I do think that we should say true things about the EC rather than false things.

      Report

  3. Lincoln did not benefit from fragmented opposition though. If there had been one opposition candidate, Lincoln still would have won. My interpretation is that the Republicans won the North and the West, and the Dems were concentrated geographicaly.

    Report

  4. …it favors regional candidates over national…

    Yes. The EC is designed to enhance regions and hopefully prevent their abuse at the hands of majorities.

    Report

    • The fact that we divided it up by state rather than by religion or trade or ethnicity or financial class is one of those fun historical accidents that we like to pretend are important pieces of wisdom. If we the original constitutional coalition had been a little different, modern people whose party benefited would be saying, “But if Catholics don’t feel like they have a say, there would be war with the Protestants, so we need to make sure their voices are heard!”

      And because the Jews and the Muslims still wouldn’t matter in either system, the people saying it would be saying it for the same cynical reasons they use today to pretend that protecting (some) states is an important and holy duty.

      Report

      • The fact that we divided it up by state rather than by religion or trade or ethnicity or financial class is one of those fun historical accidents that we like to pretend are important pieces of wisdom.

        I think that’s a fair statement…

        However, the States were/are/remain political organizations by themselves, and have political boundaries and so forth even now.

        If the EC gave a certain number of electoral votes to the glass blowers, then it’s hard to see how that system would have remained stable over the centuries. Dividing up electoral votes by states (as opposed to by religion or guild or whatever) really does look wise, even with the judgement of history.

        Report

        • The states are political organizations by themselves because the constitution designed it that way, but that doesn’t really mean anything for these purposes. It takes a bit of a chain of reasoning to follow that statement up with, “And therefore we should aggregate and round their votes off and then reweight the results before combining them.”

          It’s just a weird historical accident that people spin as deep wisdom by claiming that it solves some arbitrarily framed problems in a Rube Goldbergian way, when the reality is that it just pushes those problems around and paints a target around the new result.

          If the 3/5 compromise still had some sort of weird effect that survived the abolition of slavery, whichever party benefited from it would be signing the praises of our founders’ foresight and painting a dire picture of what would happen if we ever abolished it.

          Report

          • The part that looks rwise is tieing the entire thing into the Census. New York loses Electoral votes every 10 years, it wasn’t that long ago that they had more than 50.

            Redistributing electoral votes in other systems is very, very painful. Similarly most of the other voting methods on the table that we’ve mentioned (vote by religion, vote by guild) instantly make the Census highly politicised, perhaps to the point where the census itself becomes unworkable.

            While there’s something to the argument that the system is a historical accident, there is also something to the idea that it’s held up well and it’s “wisdom” seems more than just “spin”. It’s trivial to think of other ideas from that era (i.e. slavery) which haven’t.

            There are arguments for different systems, but I’m deeply reluctant to replace a large, working system. The on-paper theoretical attractions of whatever we’d replace it with all too often fail either in real world practice or real world implementation.

            Report

            • I agree with you that the method they chose of arbitrarily reapportioning votes to solve no real problems has proved remarkably stable in that it has had real staying power. It’s not clear to me why that’s an actual recommendation of the system, though. It doesn’t appear to do anything except create weird distortions in campaign behavior and occasionally flip the results of an election in surprising ways, all while not solving any real problems.

              Our absurd employer-based health insurance system has proved remarkably stable and robust against change even thought it’s a stupid idea that creates more problems than it solves. I don’t think that says much about either its wisdom or its usefulness.

              I don’t really think that “one person one vote” is something that we should worry is a radical experiment. We manage it at the state level just fine, and other countries have no trouble aggregating their votes into a single number. It would certainly change the strategies of presidential candidates, but I’m not convinced that will be a bad thing.

              Report

              • Again, if the EC is so great, why isn’t anyone suggesting that states adopt it for gubernatorial elections? Counties are distinct political units and states have internal urban-rural divides, so why not over-represent rural counties in an arcane system that sometimes produces screwy results?

                Report

                • Very much this. It seems like it’s an algorithm that should work at multiple scales. But we don’t do it because it doesn’t work as advertised.

                  Let’s take an example: Apparently without the EC, the people of Montana would be completely ignored by presidential candidates, and the weighting of the EC gives them a voice. Let’s ignore the actual data that the EC hasn’t made the people of Montana worth visiting and pretend that the framing is 100% true. So now the people of Montana have a voice. But now there’s a good chunk of conservative northern and central California whose votes are basically pointless because they’re part of a reliably blue state.

                  So let’s try an experiment to see if the system makes sense at the margin: Let’s split California into two states–a populous Democratic middle/southern region and a lightly populated but strongly Republican north. According to the wisdom of EC, the Republicans in the new state whose votes we were perfectly happy to erase just a minute ago are suddenly tragically disenfranchised and will need to have their votes amplified to avoid some sort of political imbalance.

                  Keep turning the knobs and adding epicycles until you’re satisfied with the results.

                  Report

                  • Your example points out that while EC gives Montana a voice, we’re unjustly ignoring the voters in Northern California.

                    Your solution to this “EC created problem” is we should be unjustly ignoring the voters in Montana as well as Northern California.

                    Report

                    • No. If every voter’s vote matters, no voters are “ignored.” If a hundred thousand extra people vote for a candidate in either place, the candidate gets those votes, as opposed to those votes literally not mattering at all. If we bring in the EC because we’re sad that there aren’t more people in Montana and want to make their votes count for more, we disenfranchise a bunch of similarly minded people in NorCal for no apparent reason.

                      And as I alluded to and you and noted above, it doesn’t really even do anything for Montana. If anything it makes them less relevant because they’re not large *and* they’re not a battleground state. So we’ve chopped up simplicity and good sense for what, exactly?

                      The fundamental problem is that EC proponents can’t explain clearly what problem it’s set to solve and then compare those objectives with what really measurably happens. The feel-good story about “protecting” small states is clearly nonsense, so the claims get squishy.

                      The most coherent explanation still boils down to, “We don’t want specific big states to dominate, so instead we rejigger the system so some other big but not-so-big states dominate instead, and we choose those states totally by accident.” Which is great if your problem is with those big states, specifically. Not so great if your concern is based on a more general principle.

                      Report

                • Again, if the EC is so great, why isn’t anyone suggesting that states adopt it for gubernatorial elections?

                  We’re not suggesting it because of history, but the better question is whether we should. Some US states have less political freedom than some 3rd world nations.

                  A system which doesn’t let the powers-that-be control the outcome of the election by running up the totals in the populated areas (which their political machines control) has some attractions.

                  Additionally, one of the big problems in democracies is the majority abusing minorities, that’s especially been true in the US at a state level. That is the context or “screwy results” in which we need to evaluate the current system.

                  The advantage of one man one vote over the country is it’s easy to understand and intuitive. Having said that, maybe it’s a bad idea for California with all its masses to dominate decisions on appropriate land use regulations for the lightly populated outback. Or for the heavily populated cities to decide what is appropriate gun control for everyone else. That sort of thing seems like a majority abusing the minority.

                  Report

                  • “political machines” – really?

                    Outside of Philadelphia and some boroughs of New York, and Cuban neighborhoods of Miami, political machines are a thing of the past.

                    What you’re actually complaining about is that liberals have the temerity and arrogance to actually govern like liberals when they win elections.

                    Report

                    • What you’re actually complaining about is that liberals have the temerity and arrogance to actually govern like liberals when they win elections.

                      I think there’s an astounding lack of awareness of how diverse this country is from the masses of the city, and giving them the ability to cram down “it works for us and one size must fit all” solutions on the rest of the country would be a bad thing.

                      “Common sense gun control” makes sense if your neighbor is one wall away and the police are one phone call away. At the other extreme, a few years ago one of my friends posted pictures of the aftermath of him being attacked, on his front lawn, by a bear (he was armed but thought a fox was after his livestock). Similarly a 15 dollar min wage looks very different from the very high tax, high wage places like San Francisco than it does from outback Montana. 55 mph max speed limit is fine in some places, less so on roads built for 80 mph when you’re the only car on it for hours.

                      IMHO it’s a good thing to force big city liberals to make themselves appealing to something other than big city liberals.

                      Report

                      • Everything you wrote above has nothing to do with EC but whether the United States is governed as a unitary or federal system. You can have a direct popular vote for President and a much more fairly proportioned Congress and still have 80 MPH speed limits in some places and 55 MPH in some places. Nearly every country has at least some level of devolved political power even if it isn’t complete federalism including France, usually seen as the paragon of top down centralism.

                        Report

                        • Everything you wrote above has nothing to do with EC but whether the United States is governed as a unitary or federal system.

                          Sure, very true.

                          But how much support for a federal system is there among the big-city, big-gov, collectivists who this “reform” would make much more powerful?

                          And after we give these collectivists the ability to win elections without needing anyone outside of their fellows, will their unitary instincts increase or decrease?

                          Report

                    • Again, the problem appears to be that square acres don’t vote but pesky people do.

                      The wrong sort of people.

                      Defending the EC is a bit hard. It doesn’t actually do what it’s supposed to, and lately it’s taken to ensuring the 2nd place vote getter won, which seems a bit weird for a Democracy.

                      Is there a term for a government run by the single most popular person on the ballot?

                      Report

                    • Actually that’s what handed both houses of Congress to Republicans. Democrats are busy trying to see if they can get 85% of the vote in some city center, while losing all the surrounding rural districts.

                      Report

                        • It’s just math. If you have a slight majority you can spread it evenly over X districts and win all X, or you can concentrate one party in one district and win just the one district.

                          For example, a state with 10 seats, each representing 100,000 people, and each won by Democrats 53,000 to 47,000, a margin of 6,000 votes in each district.

                          Then get 7,000 Democrats from each district to move to the big city with the big state university and awesome night life.

                          Now you have 9 districts with 93,000 people that are won by Republicans 50.5 to 49.5, and one district with 163,000 people won by Democrats 71.1 to 28.8.

                          So you’ve just shuffled Democrats into the urban city, which always happens, and the state’s House seats went from 10:0 Democrat to 9:1 Republican, in a state that’s still majority Democrat, and all without persuading anyone of anything except moving to a district with more Starbucks.

                          Report

                          • I was challenging what I saw as an assumption that there’s urban and there’s rural, while ignoring the suburbs. Texas is a red state because enough of its suburbs vote red; California is a blue state because enough of its suburbs vote blue; Colorado has gone from red to purple-leaning-blue because of a shift in suburban voting (the urban and rural parts both still go >75% for blue and red respectively and cancel each other out, the fight is for the suburbs).

                            Rural and not-urban are not the same thing.

                            Report

  5. It would be instructive to look at sports. Modern analytics has told us that looking at the component factors of game outcomes is more predictive of team strength than the game outcomes themselves. When determining which of two teams are more likely to win a given context, run/point/goal differential is a much better predictor than record (given a sufficient sample size). This is because a team that has lost 100 game by a single point/run/goal and won 100 games by 10 points/runs/goals (record of 100-100, differential of +900) is better than a team that has won 150 games by a single point/run/goal and lost 50 games by 10 points/runs/goals (record of 150-50, differential of -350).

    Translating this to our Presidential elections, the EC is the game and the popular vote is the differential. Why *wouldn’t* we want to use the method that gives us a truer sense of which candidate is more preferable — that is, better — in the eyes of the electorate? This isn’t a rhetorical question… why don’t we want that? Go ahead… answer it.

    I recognize we don’t hand out the championship trophy based on differential* and it would only make sense to if the teams entered the season understanding this was the goal. Likewise, we should only really rely on the popular vote if/when this is the stated objective of the election. But, it sure as hell SHOULD be the stated objective of the election!

    * Even raw differential itself isn’t ideal, with most stat heads leaning on something known as “Pythagorean Record”. I believe the exact formula for this varies slightly from sport to sport but is roughly calculated as (Points For)^2 / ((Points For)^2 + (Points Allowed)^2). So a team that is +100 via 200 for versus 100 against has an expected winning percentage of .800 as compared to a team that is +100 via 1000 versus 900 against having an expecting winning percentage of .552. Which makes perfect sense given that the former doubled their opponents’ output compared to the latter being about 10%.

    If we took this in the other direction, it’d be similar to looking at the total number of Presidential elections won by a given party over the history of the nation and declaring whoever has the most victories as the “better” party. We’d never imagine doing this. Because it’s stupid.

    Report

    • The main reason we don’t hand out the championship based on differential is probably to keep one team from running away with it too early. That’s why if you’re a league owner, you want playoffs and post-playoffs and super-duper-post-playoffs and wildcards and a whole bunch of opportunities for statistically weaker teams to have a shot at the trophy rather than being eliminated with half the season to go.

      But I hope we all agree that “keeping things exciting” shouldn’t be the goal of an election. Like, a wildcard spot where Jill Stein could have won if she collected the most Twitter followers after October probably wouldn’t be a good addition to the system.

      Report

    • Likewise, we should only really rely on the popular vote if/when this is the stated objective of the election. But, it sure as hell SHOULD be the stated objective of the election!

      In this thread we have statements like this one. In another thread around here we’re talking about the white majority oppressing the black minority to the point of race riots only 50 years ago.

      And while using the power of the gov against racial minorities is now understood to be off limits, ideologically we’ve become a lot more “collective” and those who get in the way of “doing the right thing” are often viewed not only as wrong but evil. Our need to ensure minority rights has gone up, not down.

      Granted that the EC is deeply flawed as a mechanism for this, but it’s replacement seems worse.

      Report

      • That seems like a great argument for constitutional limits on what the majority can do, not for just allowing a minority to choose the same overly powerful officials. It’s not like handing a minority the reins prevents government overreach. It’s just a different, smaller group of people choosing the government.

        Much smaller, if you consider the ridiculously outsized effects the EC has in battleground states.

        Report

        • That seems like a great argument for constitutional limits on what the majority can do…

          (Sorry I’m late, wrote this then didn’t post it)
          We already have this. The Constitution was designed to prevent the federal gov from reaching into every nook and cranny and gaining the power to be abusive.

          And having started with a Constitution with that design, we have a gov which has the power to be seriously abusive if it chooses to be. This is a HARD problem and I don’t have good answers for solving it.

          It’s not like handing a minority the reins prevents government overreach. It’s just a different, smaller group of people choosing the government.

          The key word there is “different”. Go to pure majority rule and who is in charge is going to be a lot more consistent. It’s not so much that the Dems will consistently win elections, the GOP will change and pander to the city hoards as well.

          I view consistently in charge and pandered to as bad things. We have ethanol in gas because two small states’ delegates consistently go first. Having a large group decide the election, consistently, seems like it’d be worse.

          Report

      • Is it your argument that the Electoral College is the best means to protect minority rights? Or a better means to protect minority rights that election of the President by popular vote? You’re going to have to show your work on either claim.

        I am not arguing for making all decisions by popular vote. If you look at my words in the very quote you offered, you’d see I was talking about the Presidential election.

        Popular vote doesn’t guarantee we never end up with a Trump. We could elect someone far more vile than he if the will of the majority wants it. We protect against this not by some screwy system that doesn’t actually protect against it but by other systems, e.g., our system of checks and balances.

        I’m reminded of another example from the sports world…
        For a number of years in the late 90s and early 00s, a team other than the one most folks identified as “the best” won the World Series. Hey… these things happen. So Sports Illustrated did a weird thing in their baseball preview issue: they explicitly said, “Because of this recent trend, we are going to say Team X is the best in the league but pick Team Y to win the World Series.” (I don’t remember the specific year or teams.) Their argument wasn’t that, despite being inferior in some way, Team Y was better suited to win the playoffs than Team X. It was simply that the recent trend said the best teams don’t win and therefore it made sense to pick Not The Best Team. Hogwash. That’d be like betting on snake eyes in a dice game because it came up a couple times over the last few rolls and deciding it now had better odds than 7.

        The most popular candidate among the voters should be President. If that results in a monster becoming President, put other protection mechanisms in place. Deciding, “Let’s maybe just take someone who isn’t the most popular just in case the most popular person is a monster,” is just silly.

        Report

        • Is it your argument that the Electoral College is the best means to protect minority rights?

          I’ve heard this asserted before, but I have to ask another question: Which minority? It seems to me that we can divide the country up into a lot of different minority factions, but they generally don’t seem to break along state lines.

          Report

        • Frog

          Is it your argument that the Electoral College is the best means to protect minority rights?

          (Sorry I’m late, I wrote this then didn’t post it).

          Hardly. I’ve admitted it’s deeply flawed, call it a tenth of a loaf of bread rather than a half a loaf. But it’s a check and balance in a system where we’ve increasingly gotten rid of checks and balances.

          Trump as President has done great things in terms of showing the Dems why the Imperial Presidency is a bad thing and they should be backing Federalism and State’s rights.

          (Degenerate) Thought Experiment: The Presidency is given to a randomly chosen citizen. Sooner or later it’s going to end up in the hands of a member of a group you despise. My expectation is that NO ONE will think the Imperial President is a good idea, even if their guy is in there right now. The expectation is that if the President is given power to abuse that it will be abused.

          However that’s a very difficult lesson to keep in mind when you expect your group will be controlling the Presidency more or less forever.

          As for “protecting which minority” the better question is “which majority ends up in charge (and how consistently) if we go to pure majority rules”. California is going to need a bailout or bankruptcy at some point, if we have the EC I think it gets a bankruptcy and if we have majority rule then I think every Presidential candidate runs on giving it a bailout.

          Report

Comments are closed.