Rosenberg’s theory of hollow hope has severe methodological errors — he cannot prove that court decisions cause setbacks in social change.

I have two methodological worries about the book. First, in his view, causation can only be demonstrated when an event (a court case) revs up massive social transformation from a dead stop, with nothing else around that could pos- sibly account for the change. He finds little court influence, because (1) the world was changing before the court decision and the trend simply continued afterward, or (2) the change that occurred after the court decision was argu- ably caused by the same social movement that produced the court challenge in the first place and so the unique contribution of the court's decision cannot be determined, or (3) there was little immediate change (though there was later change), and so the causal chain is broken by the intervening years and intervening events. Given Rosenberg's strict picture of what would count as causation, it's not surprising that he finds little influence of the court in any area of social policy; causation probably doesn't hap- pen that way in the social world. For one thing, the cases that he discusses were accompanied by social movements that were trying to push all the levers of power at once. Under circum- stances like that, it's always hard to disentangle causes. For another thing, large-scale social change often happens because many forces per- colate through a complicated social order and are altered in the process of achieving change. The relatively blunt measures Rosenberg uses are not well suited to picking up these influ- ences, although he makes a major effort to as- sess influence in a number of different ways. From the evidence he presents, Rosenberg can convince us that there is not enough for us to believe that there is causation in any rigorous sense (and this is a very important contribution to the debate over judicial impact), but he can't convince us that there is no causation going on here at all. Unfortunately, he tries to argue the latter case. Second, Rosenberg never considers counter- factuals. How is it possible to assess the influ- ence of the court on anything without consid- ering what would have happened if the decisions he studied had gone the other way? Since Rosen- berg considers only liberal successes, we can't tell. Of course, studying counterfactuals empir- ically is a chancy matter in investigating singu- lar real-world events, so the best social scien- tists can do is to sample cases to study that present variation in outcomes. But Rosenberg doesn't do this. If he had also studied cases where liberal causes were not victorious (e.g., in Bowers v. Hardwick, where gay rights groups were dealt a setback) and then looked to see what happened when the social movements didn't have the court decision on their side, his argument would have been more convincing. But the final conclusion that he draws, that lib- erals have wasted their time by trying to win court cases, is simply not sustainable on the evidence he presents.