Mancur Olson has suggested that the rate of economic growth declines as stable polities grow older, a proposition he tested using data from the United States in the 1960s. We reexamine Olson's hypothesis, using a large panel data set covering the twentieth century in the U.S. We find, first, that younger states typically had relatively large governments, rather than the smaller governments one might suspect if Olsonwere correct. Second, we fail to find convincing evidence to support Olson's specific claim: that older states have, ceteris paribus, lower rates of economic growth. The only specification that supports the hypothesis is one in which population movements are attributed wholly to Olsonian forces. The obstacles to testing the Olson hypothesis are formidable because of the difficulty of separating migration into Olson and non-Olson components.