Minimum wage impacts from a pre-specified research design

In calling for an increased use of this approach, Levine (2001, p. 161) states: “On the one hand, the limits of the research design are clear – only data after the policy change can be included in the analysis… On the other hand, the strengths of this method are avoiding unconscious and conscious biases of the authors, editors and referees.” The purpose of this study is to utilize Neumark’s (2001) [...] Variation in minimum wage indexes in the U. S. tends to come from differences in state ‘top-ups’, coverage or the slow erosion of the real value of the minimum wage as its infrequent changes do not keep up with changes in the average wage of the state – the latter forming the denominator of the minimum wage index. [...] The coefficient estimates on the minimum wage and adult wage variables for these specifications are presented in Appendix 3. The theoretical prediction is that breaking the minimum wage index into the minimum wage and adult wage should produce an estimate on the adult wage that has the same magnitude and opposite sign of the minimum wage variable. [...] The increase in the minimum wage may induce spillovers that increase the wages above the minimum wage, which may shift the distribution of all wages to the right. [...] Table 3 provides our estimates of the third specification of the pre-specified research design, based on a minimum wage measure that utilizes the fraction of workers in each youth age group that is below the minimum wage in the period when the minimum wage was increased.