I think the argument isn't "let's not improve things because people will use things more" but "let's not spend resources for little improvement". The induced d

I think the argument isn't "let's not improve things because people will use things more" but "let's not spend resources for little improvement". The induced demand argument suggests that the pain and utility of traffic are in equilibrium, and that there's so much latent demand that no feasible amount of pavement will reduce the pain appreciably. It seems to me of a similar genre to studies showing that safety equipment increases risky behavior.