Home

Tue, Nov. 2nd, 2004, 08:21 am

Electoral College Effects

Here is an interesting question - Given the electoral college system, does it favor the large or the small states? The very short answer is that it ridiculously disproportionately favors the small states, because they're flat-out given disproportionately large representation. And if that weren't the case then switching to a simple electoral majority would be uncontroversial, following the trend of a simple majority deciding within each state.

But if we pretend that this is an issue of real interest, what are the effects? Well, that depends. If we had a country of two states, one larger than the other, than the chances of one's vote mattering in the smaller state would be just about nil. In a country of three states two of which were just slightly larger than half the size of the largest state the small states would be far disproportionately represented.

In practice there are enough states that statistical effects overwhelm the weirdnesses of specific enumerable outcomes. We can adopt a much more simplistic model of there being many small states, and we compare two, one of which is roughly double the size of the other. If we assume that the race is a dead heat across the entire country, (a completely unrealistic assumption, as I'll get to in a minute), the chances of a voter swinging the half size state is approximately 1.4 times that of the double size state (because a standard deviation is proportional to the square root of the number of voters) and the chances of it swinging the overall election are about half, so the chances of a single vote from the smaller state swinging the election are about a third less.

But we don't have a homogeneous voting population, and very few races are dead heats across the entire country. In practice state lines jerrymander quite heavily against New York and California, whose votes in recent elections have been such a foregone conclusion that nobody bothers campaigning in them. With the two coasts being very populous and the economic centers of the country, and getting more so, this effect is likely to become even more pronounced in the future.

And then there's the question of, in a close race, which states do you give out more candy to? The ones which are close races, obviously, and the ones which you can more likely affect the outcomes of. Small states are much easier to win by buying off votes, because a much smaller number of votes can change their outcome. The likelihood of their swinging the overall election is negated here because we don't go into elections blind - campaigns poll to find out what states are up in the air, and ignore the ones which aren't.

Which states are close races varies from election to election, so there's a random crapshoot which decides who gets the most resources each time. The result is inevitable arbitrary disparities, which the only consistent thing being their arbitrariness, and strong incentivization for local officials to make their states be close, or at least appear to be close.

If this all sounds stupid and unpleasant, it's because it is. The only clear effect of a truly voter weighted electoral college would be that New York and California would be (still) jerrymandered against. All the other effects are random and generally bad for everybody.

Unfortunately the chances of the electoral college getting fixed via an orderly political process are just about nil. Fixing it would require a constitutional amendment, which would have to be approved by 2/3 of the states, and most of them are, unsurprisingly, small. The smallest states get several times their proportional say in the electoral college, and many times their propotional say in the senate, with a flat two senators from every state, so any constitutional amendment cleaning up the mess would be dead in the water.

The rules favoring small states, by the way, were set up at the time of the formation of the United States to get the south to join. Back then, the rules were even worse because slaves counted towards representation, even though they couldn't vote. It took the civil war, which was caused by the political imbalances favoring the less economically productive parts of the country and the separation being on neat geographic lines, for the mess to get cleaned up. Kind of like the situation today, except that it hasn't gotten to the point of internal warfare, at least not yet.

On that note, I feel obligated, this being election day, to encourage everyone to vote. Unless of course you'll be voting with a diebold machine, or your registration got mysteriously lost, as mine did. [Update - I showed up to vote and they did manage to find my registration in some obscure place, but my wife, whose registration was sent in at the same time and on the same day, had to cast a provisional ballot.]

Thu, Nov. 4th, 2004 03:34 pm (UTC)
(Anonymous): Does it?

Proudly voted early Tuesday morning without even being late for class.

I'd say anyone's initial impression is that it favors the smaller states, but I question that. I also question what you can replace the electoral college with. As you stated, the reason it was setup to begin with was to give each state some kind of a say (whether or not it was balanced in the beginning or is entirely now is another matter). Going by popular vote alone you might as well turn the country into the United STATE of America when you eliminate that say. I'd argue though that regardless, at least for my arbitrary example (it's the only one I looked at), it's working alright even if the electoral college is a dinosaur.

Here's an example:
I'm from Massachusetts. One of the the smallest but with more population than say, Delaware. I compare my state with California:
California, which had 9,944,625 votes for presidential candidates, gets 55 electoral votes
Massachusetts, which had 2,888,083 votes for presidential candidates, gets 12 electoral votes.
MA had 29% of the votes of CA, but only gets 21% of the representation provided California. They got less per vote, but one can still argue it's even enough (I'm not complaining, I'm just trying to make a point).
Ball-parked, Delaware has 3.7% of the vote and gets 5.4% of the representation...a slight advantage but by no means severe.

Hopefully this is at least amusing food for thought, as that's all it's intended to be at best. :)
I've pondered alternatives and the only thing I've found that I do like is the Australian style of voting where you can select multiple candidates (I like candiate A before candidate B before candidate C but don't want a vote in any circumstance for candidate D) but that wouldn't require eliminating the electoral college, either.

Mon, Nov. 8th, 2004 05:21 pm (UTC)
[info]bramcohen: Re: Does it?

The geographic size of a state is completely irrelevant, what matters is number of people. When I say 'small' states I'm referring to ones of low population, so Alaska, for example, is small, even though it's one of the largest geographically.

The varying voter turnout between states is another source of complicated and arbitrary effects, mostly bad. For example, children and prisoners count towards population, even though they can't vote. For the sake of simplicity, I'll run some numbers ignoring voter turnout and just compare population versus representation.

California has the largest population, with about 35.5 million people, and 55 electoral college votes, giving it a ratio of about 1.55. DC has a population of about .56 million, and 3 electoral votes, for a ratio of 5.36, making it almost three and a half times as represented on a per capita basis. Alaska and the Dakotas are almost as bad. A factor of several times easily qualifies as severe.

The formula by the way is congressmen plus two, so once the number of congressmen gets up 10, like it is for Massachusetts, the extra is only 20%. You aren't getting an advantage because you aren't in a small state.

Mon, Nov. 8th, 2004 11:15 pm (UTC)
(Anonymous): Re: Does it?

Certainly population wise I'm not in a small state, yet if you think relative to California I still am. I didn't intend to base it on physical size and I'll call it partly coincidental.

You do bring up an interesting issue with extremely low populations and the severity is noted. I didn't factor in (ignorance on my part, I suppose) the formula for electoral votes. That certainly does change things, I just don't know what the alternative is.

DC - I don't even understand why they gave it electoral votes. I wholely agree the imbalance is nuts and would have no issue with it being corrected (though, I doubt it will be happening soon).

Alaska is another good example, but it is a state. Would an Alaska with 1 electoral vote have any say at all in federal elections? Despite it's population it is a state in our country. Hypothetically, say it did just have 1 vote. Of the 538, that's nothing. Say the Federal Government imposes some policies the state is strongly against. If it wanted to secede at some point, would the country have legitimate right to object? I guess here I'd side with my original argument that the each state should have a say, even if population-wise it is biased in some cases. I'm sure it was extremely biased at it's inception, but it's inherently part of our system of states.

Mon, Nov. 8th, 2004 11:58 pm (UTC)
[info]bramcohen: Re: Does it?

I probably should have given the formula for number of representatives in the electoral college. I just assumed (obviously foolishly, I shouldn't blog when half-asleep) that everybody knows it.

The way to fix the formula would be to make each state have the same number of representatives in the electoral college as it has representatives in congress, rather than adding two. If that change were made everybody would just think the electoral college was stupid and have it fixed though, the momentum for it is entirely based on the small states hanging on to their overrepresentation.

It's interesting how the formula ridiculously favors the very smallest states. Here's an interesting question: If the electoral college were made fair, which states would have a higher percentage of the total. Given how skewed everything is towards wyoming, it's probably more than half the states. If it's more than 2/3 of the states then there might be hope for a constitutional amendment fixing the situation after all.

DC was created specifically to be the nation's capital so that the states wouldn't fight over which one got to be the capital. It has electors because it has population, which is reasonable. The strange the about DC is that it has no senators, although the makeup of the senate is a complete disaster: each state gets exactly two senators, regardless of its population. Nobody even dreams of getting that one fixed.

Wed, Apr. 6th, 2005 12:57 am (UTC)
[info]dmarti: "Fixing" the Senate

Two senators per state is the one remaining Constitutional issue that actually can't be changed by the normal amendment process.

Article. V.

The Congress, whenever two thirds of both Houses shall deem it necessary, shall propose Amendments to this Constitution, or, on the Application of the Legislatures of two thirds of the several States, shall call a Convention for proposing Amendments, which, in either Case, shall be valid to all Intents and Purposes, as Part of this Constitution, when ratified by the Legislatures of three fourths of the several States, or by Conventions in three fourths thereof, as the one or the other Mode of Ratification may be proposed by the Congress; Provided that no Amendment which may be made prior to the Year One thousand eight hundred and eight shall in any Manner affect the first and fourth Clauses in the Ninth Section of the first Article; and that no State, without its Consent, shall be deprived of its equal Suffrage in the Senate.

The answer is to split up the large-population states.

Sat, Nov. 6th, 2004 08:34 pm (UTC)
(Anonymous): Voting

Voting should be standardized to the best means for those who VOTE, not vice versa.

Mon, Nov. 22nd, 2004 11:45 pm (UTC)
(Anonymous): one person one vote?

One problem I've noticed with most analyses of voting systems is the assumption of one person, one vote. I agree that a nice design goal for a voting system (I think of it as "influence") is to maximize the probability that a vote cast by a voter drawn at random from the entire electorate will swing the election. One can then compare the electoral college system to this ideal and ask how unfair it it? The problem with one person, one vote, IMHO, is that it is an idealized concept with fails to account for the fact that your vote may be miscounted or not counted at all. Thus, I would say that an additional design goal for a voting system (I think of it as "stability") should be to minimize the probability that the outcome of an election will change due to an error in recording a vote. Presumably, if the errors are uniform in a system where "influence" is uniform as well, it's a wash.
Since we live in a system where neither "influence" nor errors are uniformly distributed, do we at least get more "stability"? Given
the outcome of the 2000 election, I wonder...

Tue, Nov. 23rd, 2004 03:59 am (UTC)
[info]bramcohen: Re: one person one vote?

The average number of people whose vote could throw the election the other way is minimized with a single vote across the entire population. The electoral college adds a tremendous amount of volatility.

The chances of your vote on a purely technical level being counted in the US are among the lowest in the industrialized world. Suing to stop ballots whose intention is clear from being counted, as happened in the recent san diego elections, is simply unthinkable in most places which pretend to have democracy.

Sun, Feb. 13th, 2005 01:33 pm (UTC)
(Anonymous): So if you could change the world...

What would you do to change the world for the better? Einstein and other great minds thought some sort of informed socialism was the answer.