I wanted to follow up on some of the questions and comments left in the seedbed of my earlier post about increasing the size of the US House via the ‘Wyoming Rule.’ This rule would say that the size of the House would be based on the citizens-to-Representative ratio of the smallest state (Wyoming).

First, two references to visuals: See my graph of how the US used to readjust the size of the House periodically but by no longer doing so has fallen well off the cross-national trend in the relationship between population and lower (or sole) chamber size; then also see Steven Taylor’s calculations of how many seats this rule would result in for each state.

In my previous post I suggested this should be an issue that Democrats ought to take up, and I said it would never get on the agenda of the Republican party. Chris asked why I thought Republicans would oppose it. My reasoning was that small-state overrepresentation benefits the Republican party. I do not have House state-by-state votes data immediately available, but we can get a good (perhaps better) approximation by looking at presidential votes by state. And, sure enough, the size of the state (here proxied by votes cast, because that was what I already had in a spreadsheet) is a significant predictor of the Republican vote share for president in 2004. The variance explained is small, so maybe Chris is right to question my assumption that Republicans would automatically oppose increasing the size of the House. But the relationship is significant, and thus Republicans will derive some benefit from long-term increases in the disparity of district populations–which will get worse as states further diverge in population because of the entitlement that even the smallest states get one representative apiece.

(I regressed Republican vote percentage for president in 2004 on the log of the number of votes cast in that election for each of the 50 states. The regression suggests that a state in which there are 350,000 votes cast would provide the Republican with 58.6% of the votes, with a 95% confidence interval of 53.4-63.9; the Republican would get 54.1% in a state with one million votes cast, 50.2% in a state with five million votes cast, and 48.5% in a state with ten million. The 95% interval on the latter estimate is 43.3-53.6).

But Chris went a step farther and suggested that Republicans could get a nice side effect of larger House size by being able to pack Democrats into even more compact districts than now in states where Republicans control the redistricting. I do not know if that is true, but it could be. But let me be clear: It is an even more high-priority democratic reform cause to eliminate gerrymandering (i.e. to establish non-partisan redistricting processes) than it is to expand the House (and I mentioned this in the previous post, and that was also brough up by bindare4u in the comments to the linked post at Poliblog).

Holding constant both the degree of geographic concentration of party voters and the ability of crafty line-drawers to advantage one party or another, a given state should return a more proportional seat allocation as its total number of seats to be allocated increases. This would benefit the minority party in any given state that sees additional seats. It is entirely possible that there would be more potential Republican seats added net, because the biggest states are somewhat more likely to be Democratic overall and the largest states, obviously, will gain more of the total new seats in an expanded House.

The bottom line for me on this point is that this is a question of democratic (small ‘d’) fairness of representation, and I personally do not care which party would gain the net benefit. I would prefer that the increase be combined with MMP (or another proportional system), but even within the paradigm of all single-member districts, the House should be increased to provide better representation and eliminate the long-term increase in malapportionment that will result from keeping the fixed chamber size.

On the question of MMP, Alan says it’s a bad idea for a presidential system. I must admit I do not follow the point. He seems to think that all elected offices should be elected by the same rules, and I don’t buy it, especially when we are talking about presidentialism, in which the very idea is that different agents of the electorate have different constituencies and must work together to accomplish anything. Even if STV (of which IRV or alternative vote is indeed the single-member district subtype) were used for all offices, the fact that the president’s constituency is vastly larger means it is a different electoral system, as the scope of the district is as much a part of the electoral system as is the formula for allocation within the district.

Bob asked whether increasing the size of the House is consistent with Supreme Court precedent on one-person, one-vote. Well, sure, though the Baker v. Carr decision was specifically about state legislatures. Besides, what the Supreme Court jurisprudence is on this, or many other topics is, shall we say, up for grabs. Current Court nominee Samuel Alito is on the record as being opposed to Supreme Court rulings against malapportionent, as I have noted in a couple of posts at the subdomain, judiciary.fruitsandvotes.com.