I have a brief exchange with Bill Bradford in the letters section of the February 2005 issue of Liberty magazine. Readers of L&P will find the discussion familiar; I argue that the evangelical and conservative Catholic vote in Ohio were crucial components of the Bush victory in that state. Bradford continues to argue that the anti-gay marriage ballot initiatives"actually reduced Bush's margin of victory." I still believe that Bradford is not giving enough credit to GOP strategists for getting out the socially conservative vote, and I don't see how Bush wins in Ohio without that group of voters.

As L&P readers know, I've never denied that other issues, especially the war, had an effect in shoring up Bush's winning coalition. But the point is that it takes coalitions to win votes. In my view, the anti-gay marriage ballot initiatives were promoted by GOP strategists to bolster one aspect of the winning Bush coalition.

Coalition-building among interest groups is the modus operandi in American politics. This is a point that our L&P contributing editor makes very clear in his fine Liberty essay,"Politics vs. Ideology: How Elections Are Won." Cox sees validity in the observation that"[t]here are such things as political 'bases,' communal sources of political identification." There are also"political ideas and social movements, and these can have noticeable effects on elections, occasionally dramatic effects." But, for Cox, most American presidential elections are"won by small margins." Cox maintains that Americans"are people of multiple social identifications." Thus,"[t]he task of the American political party is to exploit as many of these personal identifications as possible. This is not science," he argues,"and it cannot be." Indeed. We can argue over whether or not this group or that group, this bloc or that bloc of voters provided the crucial margin of victory. (I, myself, have not argued that social conservatives were the crucial bloc; but I have argued that Bush could not have possibly won without exploiting what Cox is here calling"personal identifications," in this instance, of a religious sort.)

Still, Cox is right:"Voting behavior is like other forms of human action, as explained by such economic theorists as Friedrich Hayek and Ludwig von Mises; it proceeds from individual, variable, nonquantifiable preferences." He continues:

What happens in American elections is that the party that lost the last one looks for a way to win the next one, knowing (if it's smart) that it cannot rely implicitly on any stable bloc of voters. Even the legendary strength of African-Americans' identification with the Democratic Party can easily recede sufficiently to keep most potential voters in that"bloc" away from the polls. The best that each political party can do is to go through its list of possible voters, trying to interest as many as possible, beginning with those most strongly identified with itself (at the moment) and proceeding as far down the list as its funds and energy permit. If the gay vote is sixth on the list, a party that has any possibility of getting it will try to do so, altering its own character and"ideas" when alteration is necessary to optimize its capacity for winning.

Precisely.

Cox argues, however, that

American elections are won not by stable power blocs but by shifts in party identifications among people who used to be in those blocs, until they escaped. Some of the shifts, which go on all the time, in every conceivable direction, coincide with major intellectual or social movements, the kind of movements that change large patterns of intellectual and social history. But electoral politics has its own more intricate, local, and self-adjusting patterns, the patterns of the marginal gains and losses that happen as parties hunt the all-important plurality of votes.

Again: precisely. From my perspective, Karl Rove and other GOP strategists did not take it for granted that socially conservative evangelical voters were a"stable bloc" that would vote for Bush. That's why the anti-gay marriage initiatives were so important to GOP strategy: they were a way of keeping that bloc stable. The GOP also targeted voters at the margins, which would explain how they bolstered the GOP"share" of the traditionally conservative Democratic Catholic and Hispanic-Catholic vote.

We can debate the effectiveness of Rove's strategy in terms of Bush's margin of victory. But I don't think it can be debated that, as Bradford himself puts it,"the Bush campaign followed a strategy that they hoped would exploit the ballot measure and that the ballot measure was quite popular with certain voters."

The major parties work hard to perfect the building and maintaining of winning coalitions of interest groups with which voters personally identify. It offers something to each group. As Cox states:"This is what supporters of minor parties usually do not understand. Almost every minor party is an ideological party, and that explains why such parties either remain minor or cease to exist." They don't learn how to play the game of coalition-building, an expression of what Theodore Lowi once called a system of"interest-group liberalism.""A minor party invariably has a well-disciplined set of ideological positions," Cox writes,"but it lacks the wide array of personal identifications that are necessary to unite a large proportion of American voters over a substantial period of time."

Cox does not believe that this translates into a glowing future for the Libertarian Party, since the major parties are obviously here to stay for the foreseeable future. But he argues that"[t]he libertarian idea really does offer something for rich and poor, black and white, male and female, gay and straight, Christian and atheist." Even if Libertarians can't get elected en masse,"being elected ... isn't the only way to affect the political system."

Amen.