I'm getting a few people saying that I'm being unfair to Mulligan - he's just pointing out something everyone ignores: the labor supply effects of the safety net.
First (if you're tempted to say something like this), people don't ignore that. That's just marketing for the research agenda (which is fine - you always talk like you're doing something new... no big deal at all). I'd venture to say that the labor supply effect of the safety net is the most studied thing about the safety net by economists. In addition to the safety net literature, of course, the optimal tax literature is also obsessed with labor supply effects (and rightfully so!). Indeed, the labor demand effects are what are relatively neglected (labor demand in general is relatively neglected compared to labor supply because data on labor suppliers is a lot more plentiful than data on labor demanders).
I say demand "relatively neglected" because of course there are lots of multiplier estimates associated with social welfare spending, and that includes the labor demand effects along with labor supply effects. When you estimate a multiplier for food stamps, you're really getting the net effect of supply and demand.
And if we want to explore a "redistribution recession" that's what you want. You don't want to isolate labor supply (unless of course you just have a research interest in labor supply - but not if you want to know the impact on employment)!