But perhaps third parties don't siphon off more votes because they're undercapitalized. It's hard for an upstart domestic auto company to challenge General Motors, or other established automakers. Remember DeLorean? He was a "third party" automaker. Democratic politics appear to be (inherently?) oligopolistic. (Funny, I just remembered that the Soviet political system was often described by western observers as an oligopoly--although they described themselves as a "democracy." More support for my pet theory that differences between Communism and social democracy, while they do exist, are in many ways less striking than the parallels.)
~Alypius Skinner > > I've never really studied the Median Voter Theorem. > Recently I read where someone claimed that the U.S. > political system was designed to keep the two parties > nearly identical by keeping other parties out. I > assumed that the reason they Dems & Reps seem so close > may be because of the MVT--they want the middle guy's > vote. So then I thought, suppose a third party were > let into the race, does the MVT still hold w/ for 3 or > more candidates? Does it weaken as more candidates > are added, or do they all bunch toward the center for > for any n>2, where n is the number of candidates? > Does anybody know of a good discussion of it online? > ---------------------------------------------------- Well look at the 1992 presidential race. You had Bob Dole, the tax-collector > for the welfare state who never met a tax hike he didn't like and the > architect of affirmative action, Bill second-biggest tax hike in history, and > Ross let's fix what's broke by raising taxes Perot. You essentially had > three mushy-moderate statist candidates running for office, and nobody openly > advocating either mainstream conservatism or mainstream liberalism (if there > is still such a thing). We needed Perot's brand of mushy-moderate statism > like Al-Queda needs a new form of explosive. > > John Anderson in 1980 likewise offered "fiscal conservatism and social > moderation," in other words, warmed over Jimmy Carter, although since Reagan > won, and would have won even had Carter gotten all of Anderson's votes > (unlikely in the extreme based on exit-polling) it would seem we had two > candidates rather far from the media voter. > > Still, most third party candidates in America (and perhaps in some of the > parliamentary democracies) seem to offer platforms that are determinedly away > from the median voter's squishy preferences. I think of candidates like > Strom Thurmond, who probably captured the median white voter in the South, > but fared poorly with most other voters. Green Party and Libertarian Party > candidates, offering platforms well away from the median voter, fare even > more poorly, at least in all but small local races. (I recall a bar owner in > Denver, registered as a Libertarian, getting elected to the Denver Election > Commission while I lived out there.) > > >From the little I know about the MVT--and it's little indeed--it seems to > assume that the candidates have no ability to influence the median voter, so > as to move it more or less in one direction or the other. If so I'd have to > say that it makes a more-than-heroic assumption. I think few people would > have guessed that during what appeared to be the heyday of unabashed > statist-liberalism and in the wake of Watergate that a strongly-conservative > Republican candidate would win by a large majority in 1980. It's remarkable > how quickly attitudes appeared to shift on a wide variety of issues from > busing to taxes, to welfare programs to abortion to defense. > > While it's undoubtedly true that many people secretly agreed with Ronald > Reagan's positions throughout the 1970s but feared to admit it to avoid > social condemnation, it must also be true that Reagan and his supporters > persuaded others who had not previously agreed, thus shifting the median > voters toward the "right" across a spectrum of issues. By focusing on the > median voter, the MVT seems to give credence to the mushy moderate's election > creed--"pander to me or lose when I vote for your opponent"--but oftentimes, > as we've seen in recent elections with Libertarians pulling votes from > Republicans and Greens pulling votes from Democrats that not pandering to the > extremes loses elections too. > > Indeed, it's not clear that the median voter theorem actually describes the > process by which candidates typically win in highly-publicized elections. > Presidents don't typically win by persuading all the mushy moderates, who > tend to break both ways and can't generally be relied upon by a major party > no matter what it does, but rather by building coalitions of voters > highly-motivated by various issues. Put together a coalition of blacks, > Jews, Northern WASP elites and labor union members and you can win even if > you're too liberal (or too statist) for the median voter. Put together a > coalition of defense hawks, right-to-bear-arms advocates, tax-cutters, > budget-balancers, welfare-cutters, deregulators and pro-life advocates and > again you can win without appealing too much to that mushy moderate in the > middle. > > Voters tend to vote based on how they feel about candidates rather than what > they think about candidates. The highly-ideological voter bases that feeling > on the candidates' position on issues. But the mushy moderate median voter > (that has a nice assonance to it, doesn't it?) based that feeling on things > like how well-spoken the candidates seem, or whether the candidate came to > the voter's house for the Iowa Caucus or has a funny accent made from a Texas > accent overlaid on a Georgian one, or whether the candidate seems nice, or > smart, or well-dressed, or like the voter's dream date, or mother-in-law, or > whomever doesn't raise the voter's least-favorite issue (typically abortion) > in an uncomfortable way, or whatever. The median voter in real life by > definition typically doesn't give as much weight to political issues as do > voters out from the mean. > > The median voter, furthermore, being mushy on issues, oftentimes wants to > vote for whomever he thinks will win--a product of the herd instinct, the > fact that humans are social animals--so that if one candidate appears to have > put together a winning coalition, the median voters has a greater likelihood > of voting for that candidate regardless of positions on issues. What this > suggests to me is that candidates actually struggle to put together > coalitions of the ideologically-motivatated non-median voters and then split > the median votes partly by appearing personally appealing, partly by > appearing to have the winning coalition, partly by persuading some of them > (at least temporarily) to move in the candidate's direction, and partly by > making revealing the unpleasantly-ideologically-motivated coalition of the > opposing candidate. > > I've never sat down and explicitly written about the median voter in this > fashion, but I've long been sceptical of the role of the political center. > This little impromptu thought-experiment reinforces my scepticism, and makes > me wonder whether I actually know anything at all about the median voter > theorem. :) > --------------------------------------------------------------
