the need for a theory of politics – that is, a theory of how policy proposals can be guided through the political process, and implemented without being completely undermined. And this is all the more important, because (on most plausible theories of politics) there are interaction effects between policy choices at time a and politics at time a+1. The policy choices you make now may have broad political consequences in the future.The comments to both posts degenerated pretty rapidly into name-calling and anti-"neoliberal" sloganeering. (Crooked Timber is famed for good comments threads, but I have never seen any.) However, Farrell's original point has some force: e.g., when one is thinking about optimizing policy in the American system, the correct procedure is to optimize subject to the constraint that the policy be sustainable, i.e., tailored to benefit enough powerful interest groups that it won't promptly be repealed, and not harmful to interest groups whose ends are broadly aligned with one's own. (I wrote some posts on this in 2010: e.g., on poverty, good faith, and means-testing.) The constrained optimum is usually less efficient than more vulnerable alternatives (the classic example here is means-testing, which should always be resisted); a shortcoming of much "technocratic" thought is that it fails to appreciate that these "improvements" are in fact harmful mirages. There are some cases in which the dynamic is transparent to everyone involved: so e.g. when a right-wing economist opposes cap-and-trade on the grounds that a carbon tax is better, the "advocacy" is clearly dishonest. However, there are also cases of cluelessness rather than mendacity; see e.g. Mark Kleiman:
Remember this the next time a conservative explains how we ought to voucherize public education. The minute that happens, the conservatives will come back and decide that we need to means-test the vouchers. That done, they’ll attack the remaining program as “welfare.”This example had a powerful impact on me when I read it. Going back, I see that Kleiman intended it as another case of conservatives moving goalposts; but it struck me at the time as something more interesting. One could honestly persuade oneself on technocratic grounds that education funded by means-tested vouchers would be superior to public education at equal levels of support; any liberal with a good theory of politics, however, would oppose the idea.
(An obvious problem is that a lot of people are dilettantish or incipient technocrats, and this fact should be part of one's theory of politics. In particular, it is difficult to bring up sustainability issues in a persuasive way in political debates without sounding patronizing. However, the questions of what positions one should hold and how one should advocate them are separate.)
Beyond the notion that one's advocacy should be informed by some notion of what policies are likely to stick, one might argue (as in CT comments) that neoliberal arguments are framed in the wrong "voice" -- i.e., they are policy advice in a disembodied sense, but not an agenda for anyone in particular. I think this is on the whole not very strong: people are interested in whatever they are interested in. MY has a blog that in principle gives him a platform for political advocacy, but presumably many of his readers -- e.g., ego -- read the blog for posts that are thinking-aloud rather than advocacy, and unless you think everyone ought to stop thinking aloud and start waving banners, it is stupid to object to this. Still, it is reasonable to complain that many technocratic left-wing policy debates are a purely academic game until one gets the political structure right.
Interestingly, however, the CT posts and esp. the comments are not really interested in fleshing out this critique (even to the extent that I have tried to), but are obsessed with one specific instance of neoliberal betrayal, viz. trade unions. And so it swiftly becomes clear that trade unions are by and large a proxy for communitarianism, and that (surprisingly enough) socialists dislike neoliberalism because it is a kind of liberalism. And while I can only speak for myself, I think it is wildly beside the point to accuse modern left-wing liberalism of "lacking a theory of politics" on the grounds that it is hostile to communitarian thought or "good" populism, of the George Scialabba variety. I oppose communitarian ends: a close-knit society, even if it were more egalitarian than an open one, would almost certainly be a more effective force for exclusion; it would be more racist and insular, more hidebound and suffocating and judgmental, than an open one. To the extent that one is a liberal, and values openness and freedom for their own sakes, one is unlikely to endorse (e.g.) Scialabba's shockingly cavalier view of the Jim Crow south.
If I were called in to construct a theory of liberal politics, I should treat it as the following task: beginning with the marginal figures of society, the extremely poor and those outside the cultural mainstream, to put together a functioning coalition that (a) had nontrivial political clout, (b) was bound together by mutual interest to some appreciable degree. (I don't have an answer but e.g. something that is obvious re: (a) is that an alliance of the poor and the lower-middle class generally won't work.) I would also ask what kind of institutional "division of labor" between the state and other institutions might be least bad for marginal populations. If one's answers to these questions are less emphatic than those of the Thomas-Frank-ish populist wing -- just get everybody really really angry and we'll burn teh corporationz!!! -- this is perhaps because it is harder to create institutional structures that safeguard the interests of minority groups than it is to do so for the cultural and economic mainstream, especially in a discourse that so insistently valorizes the "middle class."