Monday, February 25, 2008

Gaming the Electoral College

Somehow I didn’t realize until this election year that the Republicans and Democratics don’t have exactly as many delegates for their nominating convention as there are electors in the electoral college. There are 538 electors (found by adding the number of reps and senators from each state, plus three for D.C.). But the Democrats have over 4000 delegates for their convention, the Republicans almost 2500 (I’ve gone with ballpark figures here because the actual figures seem to be in flux).

Nevertheless, the number of delegates is approximately proportionate to the number of electors, at least once you ignore delegates from territories (which get no electors but which have delegates in the Democratic Convention). My quick calculations for the D’s (I was too lazy to do it for the R’s) showed that the percentage of delegates and the percentage of electors from a state never differed by more than one percentage point.

That means that there’s still room, in future elections, for one or both parties to alter their convention rules to game the electoral college. The basic strategy would be to put extra weight on states that could easily go either way in the general election, and to put minimal weight on states that are clearly in one camp or the other. For example, it would make sense (for both parties!) to reduce the number of delegates allocated to California, which will almost assuredly go Democratic, and also to reduce the number allocated to Texas, which will almost assuredly go Republican. The delegates should be reallocated to states like Ohio and Florida, which have demonstrated their willingness to go either way.

Note that it does not make sense to give more weight to a state that your party “owns,” because those electors are essentially yours anyway. The Democrats will probably get California regardless of their nominee, so Californians’ opinions are not very informative in choosing the candidate with the best likelihood of winning the general election.

Now, why haven’t the parties already done this? One answer is that it would seem undemocratic. But that doesn’t really fly, because (a) the electoral college itself is undemocratic in some sense, yet both parties now allocate delegates approximately in line with the college’s electors; and (b) the Democrats have a large number of super-delegates (about 20% of the total) who are free to vote however they want, regardless of primary and caucus results, and the Republicans have a smaller number of delegates with similar freedom (though they aren’t usually called “super-delegates”).

The existence of super-delegates (or their Republican equivalent) points to another answer: the parties already have a mechanism for incorporating strategic considerations. Delegates not bound by the states’ primaries or caucuses can throw their votes to candidates favored in those states that should strategically have more weight. But why not change the allocation of regular delegates so that it will already reflect good electoral strategy? Then you can let the super-delegates focus on other considerations that are not as easy to include mechanically.

A third possible answer is that the parties have other purposes for their delegates. The Democratic Party (and maybe the GOP) allows its delegates to determine its national party platform – which makes sense of the otherwise bizarre decision to grant delegates to territories that cannot vote for president. But why not simply have a different set of delegates for that purpose? Yes, it would be complicated – but the status quo is also mindbogglingly complex for the average person. In general, the parties don’t shy away from complex strategies that will improve their electoral odds; consider, for instance, the widespread practice of gerrymandering.

So what gives? Are there dollar bills on the sidewalk in the presidential market?


Larry Sheldon said...

Something that seems always to escape attention.

If the new rules (for any set of "new rules") get in the way of something a lot of power (lot of people with little power, or a few people with a lot of power) want to do, the new rules _will_ be gamed to accomlish the desired result.

See The Hours of Service rules for truck drivers for an example. Or see the IRS regulations over time.

Ran said...

A few thoughts:

* California has had Republican governors before — indeed, doesn't it currently have one? I wouldn't expect it to vote Republican in a Presidential election, but I have to imagine that the Democrats taking delegates away from it probably wouldn't help them at all.

* You seem to be assuming that partisans in swing states will tend to choose the candidates that swing voters will prefer; I'm not sure that's a valid assumption. Granted, there are some states (such as New Hampshire and Wisconsin) where independents can and do vote in primaries, but that's not the case in most places — and except possibly in the very special case of New Hampshire, I don't think the independents really decide the primaries anywhere.

* These things change over time. Since parties have their National Conventions only once every four years, they couldn't react quickly to political trends that would affect what states merited better weighting. Further, states that are favored under one system might not be so quick to help fix the system toward one that would disfavor them. (I'm assuming here that a subjective system is more susceptible to well-intentioned gaming than a fairly objective one that approximates electoral representation.)

Protonk said...

Interestingly enough, the democratic (and republican, I think)caucus and primary rules are often the reverse (by county) of what you are proposing. More delegates are awarded by county for areas carried by the same party in the last general election.

My explanation for this one is non-economic. Apart from the democratic "superdelegates", pledged delegates are low level party officials who are chosen to go to the convention. For 99% of the convention goers, it is an opportunity to network and get excited/involved in campaigning. In this case, going to the convention is a reward for staff members, precinct captains and other lower level operations. Campaign operations that reliably get the vote out (as measured imperfectly by the last general election) get rewarded.

daniel john said...

It is nice to know, I really like this, I mean it really looks interesting, for the info.