Dylan Matthews of the Washington Post, who I think of as the most fair-minded of the left-of-center professional bloggers, linked to a blog post that made the following argument:

[P]eople care about what they care about. Telling them “No, do NOT care about that, care only about my arbitrary, rigid, and counterintuitive definition of liberty!” is not going to win your movement a lot of followers in the long term.

The advances made by social liberals across a number of domains, e.g., the acceptance of same-sex relationships, the increasing presence and status of women in economic sectors that had traditionally been dominated by men, etc., suggest that this is not quite true. The deployment of shame and sharp arguments can, over time, change public opinion, provided the underlying social change is consonant with underlying economic trends and isn’t too morally demanding.

Riots, for example, really take off when the likelihood of punishment decreases. People are opportunistic by nature, which is why social sanction is so important. If it becomes costly for people to hold anti-Bengali views because people with a lot of social and cultural capital project that anti-Bengali views are some combination of (a) immoral, (b) offensive, and, perhaps most importantly, (c) gauche, we can expect anti-Bengali views to fade, which is of course very good news for me personally as a person of Bengali origin. If, however, there is some countervailing pressure — say the people who are expected to abandon anti-Bengali views are locked in an intense economic competition with Bengalis they perceive to be zero sum — then we will see the persistence of a countercurrent.

The blogger Dylan cites seems to think that his view is obviously right, which I find sociologically interesting in itself. The rest of the post in question seems to mischaracterize an argument made by Bryan Caplan, a contrarian libertarian, though I don’t think the mischaracterization is intentional. I’ll let you judge for yourself.