The second argument, which I suspect Yglesias (and many of his readers) simply take for granted, is that moving "a little bit to Obama's left or a lot to his left" is "unlikely to help Democrats down-ballot"; the danger is that "the party is marching steadily to the left on its issue positions...even though existing issue positions seem incompatible with a House majority or any meaningful degree of success in state politics."
Unspoken in this analysis is a major premise of centrist politics that deserves our attention: the belief that Democrats will attract fewer voters as it moves "to the left". This is actually an extraordinarily complicated claim based on all kinds of assumptions that seem less than certain given minimal scrutiny:
- Policy positions map logically onto a left-to-right ideological spectrum. When Yglesias says that Democrats are moving "to Obama's left," this has some definite and coherent meaning that's independent of voter opinion. So for example, we should be able to assess whether a ban on drone strikes is "to Obama's left," and on that basis we can then ask whether more or less people tend to support it.
It's true that relative positions along this supposed spectrum often seem intuitive, but our very reliance on intuition should be a clue that this is not an objective issue. This is particularly obvious when there is controversy. Compare for instance the Clinton plan of providing means-tested funding for college education with the Sanders plan of universal funding. One can argue that Clinton's position is the "left" position in the sense that it is operationally redistributive; but one can also argue that Sanders' position is the "left" position since it is effectively redistributive. Here the argument seems mostly definitional; there is no objective or consensus way to resolve it.
- Popular support forms a bell-curve along the ideological spectrum. Voter support will tend to congregate around one point along the ideological spectrum, and from there it will decrease whether you move to the left or to the right. You can never have a situation where support for a position decreases and then increases as you move in one direction.
It is not at all clear why this should actually be the case; there is plenty of evidence to suggest otherwise, and it's easy to understand why. Obamacare, for example, is usually understood as a policy that's "to the left" of the status quo ante, and single-payer healthcare positioned "to the left" of that; but polling has clearly and consistently demonstrated that single-payer is far more popular than Obamacare, even though Obamacare is arguably less popular than the status quo ante. The reason for this is simple: Obamacare is a compromise between everything we dislike about the leftist position (higher taxes) and everything we dislike about the rightist position (privatized health care). Obviously the popularity of a compromise between any two policies will depend not only on their relative positions, but also on what is being compromised.
- The bell-curve of popular support reaches its apex between the Democratic and Republican positions. This is why it's self-evidently dangerous for Democrats to move "to the left": popular opinion will always, by definition, peak to their right.
Supposedly, this positioning is maintained because of competition between the two parties, which inclines them to move towards the popular center, and ideological orientation, which draws them in opposite directions. Clearly this is only true if we accept the second point, since other distributions of support could lure the parties into competing for seperate constituencies, or into a race towards one increasingly popular end of the spectrum. But even if we do accept the second point, we're still assuming that American democracy is facilitating a functional competition between the two parties. Suppose for example that some third factor - say, the imperatives of capitalism - has artificially forced both parties to track further to the right than competition would otherwise dictate. If this is true, then the apex of popular support would be located to the left of both parties, and electoral success would be defined by their ability to defy capitalist incentives / deterrents and move towards more popular positions.
Crucially, all three of these premises have to hold for Yglesias' analysis to work. What is interesting here is that while there is little question that politicians need to advance popular positions in order to win elections, that is precisely the point Yglesias insists on defending - while ignoring entirely the more controversial and consequential question of where precisely popular support actually lies.
All of this, I should add, is just a further elaboration on the standard leftist theory of change I noted recently - the one that Jeet Heer seems to think does not exist. If we suppose that Yglesias, too, is simply not taking the leftist analysis into account, that would explain why he assumes that the votes are to the Democrats' right. Heer and Yglesias are both relatively smart guys, so when we see them relying on such stubbornly unexamined assumptions, we have good reason to suspect the hand of ideology at work.