This discussion concerns a number of methodological difficulties in essays by Clark Nardinelli, Myles S. Wallace and John T. Warner and by Erich Weede in testing some ideas of Mancur Olson to explain the differential economic growth of nations. A critical problem is the selection of proxy variables to measure the strength and encompassing nature of interest groups. A number of other problems including the selection of the sample and the nature of the test are also designated. The basic conclusion is that we are far from a satisfactory test of Olson's ideas.
Mancur Olson is well known for the fertility and originality of his ideas and his book, The Rise and Decline of Nations has many striking propositions. However, Olson illustrated most of his propositions by examples and did not discuss in detail how his ideas could be tested empirically in a systematic fashion. This has raised problems for those who have tried to follow up and to develop some of his themes.
Clark Nardinelli, Myles S. Wallace and John T. Warner (hereafter NWW) have taken as their starting point some evidence of Kwang Choi which Olson did discuss in his book, namely the differential growth of U.S. states. Because it is difficult to measure the strength of interest groups, they have used as a proxy variable the length of time since admission into the United States (or, for the former members of the Confederate States of America, the time elapsed since the end of the Civil War). Other explanatory variables include a catch-up variable (i.e., the per capita income at the beginning of the period examined) and a variable indicating the change in state and local governmental activities, measured by salaries of state and local employees, a causal variable proposed by Gordon Tullock (in a conference paper, August 1986). Although I have been told that NWW carried out this exercise in order to prove the correctness of Olson's propositions, they found in their regression experiments that the catch-up variable explained most of the differential growth of states and that the interest group proxy suggested by Olson did not perform as expected.
This exercise is troubling. If we focus our attention on the South, we know that for the first half century following the Civil War, the per capita