Winner Take All, State Take Less?

Cato's Pat Basham writes on research suggesting that winner-take all voting systems do (relatively) better at keeping the size of government under control than do proportional representation systems of the type common in Europe. I couldn't find a link to the original research, so I don't know what the authors control for (I can imagine electoral system serving as a proxy for other background features of the political culture), but the explanation Basham summarizes is plausible on face. One of the authors has an archive of interesting sounding papers in the same vein.

Comments

[thoreau]

I've been a fan of proportional systems, so this is dismaying news for me. Then again, one study is not the be-all and end-all on the effect of proportional systems on fiscal policy.

(Before people accuse me of supporting something that I don't support, let me make it clear that I have only supported the use of proportional systems for a single chamber of the legislature, be it state or national. I have never supported electing both chambers by a proportional system, and I have never supported choosing the executive via a parliamentary vote. I believe that the executive should be chosen separate from the legislature. The only thing I've ever suggested is that US House delegations should be chosen proportionally from within each state, and in state legislatures one house should be chosen from several large multi-member districts while the other house is chosen from single-member districts.)

One thing that the study didn't give much attention to (I looked it up) is the effect of bicameralism. The US is one of the few countries in the world where the national legislature has 2 branches of roughly equal power. In most countries there's either a single chamber, or a single powerful chamber and a weak second chamber. The rarity of systems like ours makes it more difficult to do a cross-country comparison with a statistically significant sample size, but one might ponder the effects of electing one chamber by a proportional system and a more powerful upper chamber from single-member districts (e.g. each state elects its House delegations at-large, or from several-large districts in states like CA and TX, but each state elects its Senators in the same way).

The authors propose that systems like ours vest power in a handful of swing voters and hence limit the amount of people who need to be courted with public spending, while proportional systems require appealing to more voters. If the more powerful Senate was elected by winner-take-all rules where only swing factions matter, one might consider the possibility that public spending would remain under control despite a proportionally elected House.

Then again, maybe I'm wrong.

[Dan]

It makes sense to me. The less choice voters have among candidates, the less it is necessary for candidates to bribe voters for their votes.

[Kevin Carson]

The more insulated candidates are from genuine expressions of popular opinion, the more closely they will reflect the common values of the class from which they come. That means that when the public is heavily in favor of things like single-payer health insurance and other soccer mom crap, the government's inertia will stand in the way of it. But it also means that if the public is ever in favor of libertarian and decentralist policies, the government's inertia on behalf of corporatism will impede that, too.

[thoreau]

It might be worth pointing out that even in a winner-take-all system there are other ways to do it than our current system. A single vote works fine if there's only 2 candidates. However, if there are 3 strong candidates, and each is polling about 1/3 of the vote, then the winner may be somebody with only weak support. In these scenarios, things like Approval Voting, Condorcet voting, runoffs, instant runoff, and other methods are good ways to get winners with broader support.

Of course, we rarely have 3 contenders (well, except in some party primaries). The reason is that if you start off with 3 contenders and one falls a little behind, rational voters will defect from the third-place candidate in most cases (McClintock's support in the CA recall being a notable recent exception). The result is that a 2-party duopoly is basically built into our current system.

On the other hand, with methods like Approval Voting and such, 3 or more parties could compete, giving us more meaningful electoral competition and more participants in the marketplace of ideas. Yet we might still enjoy the alleged libertarian benefits of winner-take-all.

Or so one might hope.

Then again, as long as we have such heavy use of gerrymandering, there will be no meaningful competition regardless of the method used.

[snarky]

Those damn Federalists again!

[kevrob]

I'd also like to see how countries with a separately elected executive and a PR legislature compare to a PM and cabinet responsible to the parliament. Except in the rare cases of minority govt.s, PM's have a majority in the house they are responsible to, or lead a coalition of parties that add up to a majority. The U.S. can have gridlock, France can have "co-habitation", and Italy's post-war history of revolving cabinets is legend, but those are outliers.

I have read comments from Canadians that they hope the Liberals get in with a minority, as they won't be able to enact much idiocy, while a Tory or Grit majority could do various kinds of damage.

PR schemes put a great deal of power in the hands of the party establishment, in the recruitment of candidates for the electoral list. It isn't impossible in a single-member district for party mavericks to win seats, and party stalwwarts to lose. PR lists protect the party regulars from being turned out.

Kevin

[thoreau]

kevrob-

Open-list systems give more power to individual candidates. The Swiss have taken it to an extreme where in a race for N seats you vote for up to N candidates from as many lists as you like.

I wouldn't recommend all aspects of the Swiss system, but some are worth contemplating. They've managed to incorporate a mix of direct democracy, highly decentralized federalism, and consensus politics. Not all aspects are necessarily worth borrowing, but there is much to learn from the Swiss.

[James Anderson Merritt]

Also, I wonder if anyone has done a study of how electoral results might have been different in the past few elections, had all states apportioned their electoral votes to candidates, based on their statewide popular vote totals. It would be interesting to see whether we would have ended up with any different Presidents from, say, 1948 onward. If anyone knows of such a study, especially if there's a weblink, please cite!

[thoreau]

Warren-

The thing about winner-take-all electoral votes is that no state has an incentive to divide them up proportionally.

Say that a state is guaranteed to go Republican. Why would the GOP majority in that state decide to toss 40% or whatever of that state's electoral votes to the Democrat? Same analysis holds if the state were guaranteed to go Democrat.

Now, say that the state is really close. With proportional allocation it would go from 3 or more votes being at stake to just 1 electoral vote being at stake. People whose interests were being heavily pandered to would suddenly become less pivotal. Even if you aren't a swing voter, core supporters still need love and attention from a candidate hoping to mobilize his base. So everybody in a swing state has an incentive to support winner-take-all.

So in each state the majority has an incentive to keep winner-take-all electoral vote allocation. This is the opposite of a cartel, because it's actually quite stable, and the result of a historical evolution toward a stable Nash equilibrium.

There are, of course, 2 states that use district allocation instead of winner-take-all (Nebraska and Maine) but those states have rather interesting political cultures. ME sends some fairly liberal Republicans to Congress and recently had an independent governor (Angus King) while NB has the nation's only unicameral and ostensibly non-partisan legislature.

Incidentally, the stability of the electoral college's arrangements also provides a clue to why popular election will never happen. A lot of people on both sides of the issue focus on the advantage it gives to small states. Whatever you might think of that, there's a second issue: Suppose that the number of electoral votes for each state was exactly equal to the number of people living in it (hypothetically). Suppose that each state had to choose between proportional allocation and winner-take-all. In this case proportional-allocation would be just like direct popular election (aside from issues of voter turn-out, but let's leave those aside for now). And the exact same analysis would apply.

Some may be pleased by this analysis, others may be dismayed, but either way, the electoral college's current arrangement is definitely a stable equilibrium.

[thoreau]

MEA CULPA!!!!

For some reason I referred to James Merritt as Warren in my previous post. I don't know what I was thinking.

At least I didn't decide that he's really Jean Bart, like I did with Gary Gunnels 🙂

[Karl Gallagher]

Thoreau, I really like your concept (I've had some thoughts along those lines for a while). The closest I've seen to that is the Australian system, with a PR upper house acting as a check on the Prime Minister's control of the lower house and executive. Has anyone looked on how the Aussies are doing relative to the other options?

[thoreau]

Hmm, as I think about the thesis of this report, it is (more or less) that PR systems require candidates to appeal to more people with more promises of spending, while winner-take-all requires candidates to appeal to fewer voters.

By that token, it would seem that a state with a heavily gerrymandered legislature should have higher spending per capita or as a percentage of GDP or something else along those lines. Now, I could point to CA with its heavily gerrymandered legislature and say "Ah! Your theory fails!" but the more scientific approach is to do statistical comparisons and look for overall trends. I wonder if anybody has compared the 50 state legislatures by some measure of gerrymandering or competitiveness, and checked whether the degree of political competition had an effect on the size of the state government.

If political competition leads to larger government in US states then that would give a bleaker outlook for electing one chamber of the legislature proportionally in the US. On the other hand, if political competition failed to lead to larger state governments, that would suggest that US political culture does not face the same factors that cause the (alleged) relationship between proportional elections and larger government.

Karl-

I think the Australian system is interesting, but it's the opposite of what I'm proposing (a proportional House and winner-take-all Senate, and a separately elected executive). A better model, at least for the legislature, is Switzerland with a proportional House and winner-take-all Senate, apportioned along the exact same "Great Compromise" as the US Congress. Interestingly, although the Swiss executive is selected by the legislature, the authors of the study categorized them as a Presidential system because their executive can't be removed at the drop of a hat by no-confidence votes (unlike, say, the Israeli and Italian systems).

Given how well the Swiss have fared with their system (a public sector that's fairly small compared with the rest of Europe, albeit too large to satsify purists on this forum), and given some of the similarities between their system and ours, a proportional House may be a reasonable thing for the US to have.

[James Anderson Merritt]

I don't know about everyone else here, but when people I know get bunched panties about the inequities of "winner take all," they are not worried about legislative elections, but rather, the electoral college system that selects the President and Vice President. Mr Basham's article covered only the legislative side, which seems of minor concern to most voters. Unfortunately, unless someone reads and appreciates the article in full, he may come away with the mis-impression that the "winner take all" problem he really cares about -- in the electoral college -- is intrinsic to our Constitution (as "winner take all" for legislative elections certainly is).

Of course, the "winner take all" assignment of electors to presidential candidates is NOT specified in the Constitution, and is a consequence of deliberately-engineered harmony between separate laws of most of the 50 states. Think of it as an OPEC-like cooperation between vote-producing states! If the cartel were someday to fall apart -- if more than a handful of states would change their method of assigning electors so that, say, a candidate's number of electoral votes would be roughly proportional to his or her share of the state's popular vote -- I don't think that the issue of "winner take all" in legislative elections would get much traction out on the street. For the most part, people seem cool with electing a single representative per district, as long as they can choose whomever they want, including independents and the occasional third-party exemplar. The study mentioned by Basham just underscores the general suspicion that, "it's not just the law, it's a good idea!"

[Jason Ligon]

The Most Depressing Thesis:

We get the government we want. No more and no less. Everyone wants free stuff. The act of increasing the breadth of political participation is exactly the act buying more voters into your coalition. Winner takes all is superior precisely because winning coalitions have to buy fewer people.

Voters interested in smaller government are so insignificant that their impact on the process in not measureable at all, so no one bothers trying to bring them into the tent.

Corollary:

Successful libertarian politics lies not in promoting an agenda of small government directly, but in attempting to align with the largest group of non-redistributionists possible. This means all manner of people you would not normally associate with - including christian activist conservatives. By placing all of the gimme's on the same team, you force their position to be either canibalistic or so horrificly expensive that they have to raise taxes at every opportunity. Success means making them raise taxes on themselves. The Republican party, with some reduction in spending is the only hope we have.

I don't necessarily believe this, as I just banged in out stream of consciousness. The big fear for me is that given broad coalitions, everyone would demand money in their own pockets, and every party becomes a redistribution party. If narrow coalitions don't work, if Republicans have to redistribute to get elected, we are well and truly screwed. It mwould mean that in a wealthy society, people want 'their fair share', and they will vote themselves goodies until everyone is destitute. Liberty is unsustainable in a society characterized by wealth and envy. We are all wasting our time.