The Grocery Stores Wage Distribution: A Semi-Parametric Analysis of the Role of Retailing and Labor Market Institutions

How and why has the wage distribution in U.S. grocery stores changed between 1984 and 1994? Unlike other industries in the time period, the important change in the wage distribution is not rising inequality, but the real wages fell across the entire wage distribution. Changes in labor market institutions explain more than half of the change in the wage distribution in grocery stores. Specifically, the decline in the real value of the minimum wage explains little of the decline in the mean real wage but much of the change in the shape of the distribution between 1984 and 1994, and 95 percent of the decline at the lowest 10th percentile. The decline in union coverage in grocery stores and the narrowing of the union-nonunion wage gap explains much of the decline above the 25th percentile. A third institutional change, the use of part-time employees, is not associated with the changes in grocery industry wage outcomes. One might think that the major changes in operation and technologies that occurred during this time period are at least contributing factors, but we find quite the contrary. If average store size, weekly operating hours, and the use of scanning technology had remained at their 1984 levels, the real wage decline would have been even greater than that actually seen, and for the entire wage distribution. Changes in grocery retailing prevented and even greater decline in real wages. Again unlike many other industries, skill-biased technological change does not appear important for grocery industry wage outcomes. The basis of our analysis is a statistical technique which combines nonparametric kernel density estimation with a parametric re-weighting, applied to Current Population Survey data supplemented with secondary data sources on the Grocery industry.