One unfortunate consequence of the median voter theorem is the notion that the “magical median” always wins. This is often taken to imply that, if you can successfully capture the “middle,” you automatically hold an electoral advantage.
There is a critical assumption undergirding this: that electoral politics is unidimensional. We often believe this: all too often, we take it for granted that there are such things as “liberals” and “conservatives” who hold well-defined views across all manner of different issue areas. In practice, the evidence for this in the electorate is varied: as long as there have been survey research into voting behavior and political attitudes, those who pay close attention to politics (and participate frequently) have been shown to have fairly unidimensional ideology, while those who do not have been consistently found to lack such an ideology. An old (and, in my opinion, an unfortunate) tradition holds that this is a problem of the politically “unsophisticated”: that they don’t know what goes with what and, since they don’t know, they don’t hold consistent “ideology” and keep changing their mind. The problem with this viewpoint is that it places the conventionally defined “ideology” on a pedestal as the “normal” state of the universe, without any reason being supplied, other than that that is how things are.
In practice, the bundling of positions on different issue areas into an “ideology” is the product of often quite complex political maneuverings among the elites: the Democratic Party in US, in the post World War II era, especially, came to bundle the positions associated with the “economic liberalism,” in favor of the working class, and “social liberalism” in favor of the underrepresented minorities through a series of internal bargains. Democratic politicians, as part of this bargain, came to support a set of reasonably consistent set of positions (even if they might look to be all over the place in retrospect, given today’s politics), and even without knowing why or how that was achieved, this practice drew linkage between these positions among those who paid close attention to politics but did not have particularly strong views on the individual issues themselves.
Among those who had strong position on particular issues and/or did not pay close attention to politics generally, exceptions arise. From the beginning, farmers were deemed to be an exception to the ideological/partisan mindset, since their political behavior could be far more reliably predicted by particulars of farm policy and agricultural economy rather than by any “ideology” or “partisanship,” and such behavior was so consistent that it was clearly not driven by “ignorance.” But it also reveals an important omission in the theory of ideology: the politicians might have entered into a compact within the framework of a party to jointly pursue a more or less consistent “ideology.” The voters were not party to that compact. Why should we expect the voters to abide by that compact, whether they are wise to the ways of politics or not?
Indeed, we can extend the thought experiment a bit further. Suppose there are two groups of voters: economic liberals, who are deeply interested in economic dimension and favor the positions associated with the “liberal” position along this dimension, and the social liberals who are deeply interested in social dimension and favor the “liberal” positions along this dimension. Neither cares much for the dimension not of their interest.
If the Democratic Party, as a group, has committed to both social liberalism and economic liberalism, it ensures that it would enjoy support from both. Economic liberals may have all sorts of views on social issues, but they are subsumed by their support for economic liberalism. The same for social liberals. If, however, the Democratic Party seeks to become “moderate” by loosening its commitment to economic liberalism, it forces the economic liberals to reassess their support for the Democrats. Since the Democrats are no longer economic liberals, they need to choose based on other dimensions, those that are not the most salient for them. Those who are social liberals might opt to remain, those who are social conservatives might defect to the Republicans, while those who don’t care much would stop participating in politics.
DW Nominate type measures would show the movement by the Democrats on the economic dimension (while remaining at the same position on the social dimension) as becoming more “moderate,” as, in effect, the unidimensional ideology it measures is essentially the average of the two dimensions. In so doing, however, the Democrats who become more “moderate” will have lost support.
I have mentioned before (if I could find which post it was) that something like this is what afflicted the Democrats in 1980: Kennedy was, in the larger scheme of things, both a social and economic liberal. Carter was, on the other hand, largely a social liberal (at least, relatively speaking) but an economic moderate or even conservative (who was committed to deregulation and deficit reduction). In course of the primaries, Kennedy drew support from union workers, many of whom, subsequently, defected to Reagan. While the central theme of this oft-told story is that many union workers were drawn to Reagan’s social conservatism, I think an important dimension that is overlooked is that they were repelled by Carter’s economic conservatism. While it is certainly not the case that Reagan was less economic conservative than Carter, they were not all that different from the perspective of the union workers on that dimension. Tweedle Dum and Tweedle Dee they may have been on the economic dimension,, however, but at least one was better on other, secondary issues for many union workers–whether it was Reagan’s social conservatism, personal charm, or whatever else.
I think the same logic is at work in 2016. Clinton and Sanders, I think, do equally well with social liberals that dominate the present Democratic Party. Clinton’s economic conservatism and hawkish internationalism, however, makes her distasteful to many who do not align neatly with the current ideological divides that define the parties. To the degree that vast majority of Sanders supporters are social liberals as well as economic liberals, there is no question that large majority of them will side with Clinton (if they remain active in politics at all–which should worry the Democrats more), but for the sizable minority to whom social issues matter little but economic issues are preeminent, support for Clinton is improbable. While their defection to Trump is somewhat unlikely–if they liked Trump so much, they would have supported him already–they have, at minimum, some potential to be recruited by the Trump campaign, as his speech on June 7 indicated.
Will there be enough of them? I don’t know. The size of the politically inactive population in United States has always been quite huge, for good reasons too. What makes them tick has been difficult to measure precisely because their political inaction generated so little data to be studied, which, in turn, shaped conventional wisdom that systematically dismissed their significance (justifiably so in most elections). 2016 primaries have dumbfounded many because so many previously inactive voters participated–not necessarily leading to an increase in political participation broadly, but rather different and more unpredictable compositions of the electorates. How this set of voters will behave (if they do bother doing something–and it has usually been the safe bet that they’d do nothing) in November is not something one should make too many predictions on if the goal is to get right answers. Still, this is all the more reason to pay closer attention to this election–and pay especially close attention to the ways that data shows theories wrong.