Paul Craig Roberts writes, in "A minimum-wage study with minimum credibility" (Economic Viewpoint, Apr. 24): "Michigan State University Professor David Neumark and Federal Reserve economist William Wascher acquired the actual payroll data from fast-food establishments in New Jersey and Pennsylvania. The payroll data show that fast-food employment did not increase in New Jersey after the minimum wage increase. [Instead, it declined 4.8% relative to the control group in neighboring Pennsylvania.]"
On the contrary, Neumark and Wascher did not acquire the payroll data. Rather, the data were acquired from a selected group of franchisees by the Employment Policy Institute (EPI), an organization funded by business contributions and opposed to minimum-wage increases. EPI then provided data to Neumark and Wascher.
The data set used in my study with Alan Krueger of the effect of the New Jersey minimum wage includes 410 restaurants from four fast-food chains: Burger King, KFC, Roy Rogers, and Wendy's. Our sample includes both company-owned and franchise outlets. The payroll data set used by Neumark and Wascher includes 71 restaurants in the Burger King and Wendy's chains, all owned by franchisees. Neumark and Wascher have not shown that results based on "the actual payroll data" differ from the results reported in our study. They have not even analyzed data for the same restaurants, or the same restaurant chains, or for stores operated by the same groups of owners.
Professor of Economics
Editor's Note: New Jersey's experience is hotly debated in the economic profession. The study by Card and Krueger was based on a telephone survey of outlets of four fast-food restaurant chains. It concluded that employment in New Jersey outlets rose after the state raised the minimum wage in 1992, while employment fell in neighboring Pennsylvania, where the minimum wage did not change. Neumark and Wascher, using payroll data from two chains, found that employment rose in New Jersey but rose even faster in Pennsylvania. Neumark says he verified by phone that the data provided by EPI were complete. He says his study, still collecting data, has compared results only with the similar portion of Card's restaurants.
Roberts claims that the data used in the Card-Krueger study of minimum wages are obviously flawed and that the publication of this paper casts serious doubts on Card's competence as an economist as well as on the standards at The American Economic Review.
Roberts dramatically overstates the flaws. He notes that the data contain large variations hard to explain. In economics, data often look like this--and well-understood ways exist to assess the consequences of such variation for empirical results. One of us did a statistical re-analysis of the Card-Krueger data and found the basic findings unaffected by this variation. The study by Neumark and Wascher cited by Roberts performs a service by collecting additional data on this point. That their findings contradict Card's and Krueger's is evidence of healthy discourse on a difficult empirical problem--not, as Roberts asserts, evidence of incompetence.
Anyone who follows economics literature knows that Card is among the most careful and conscientious people doing empirical work in labor economics. Roberts' attempt to portray Card otherwise is a blatant effort to intimidate dissenting voices.
Sloan School, MIT
Carnegie Mellon University
Paul Craig Roberts says that to argue that an increase in the minimum wage may raise employment denies "the law of demand, the cornerstone of economic science." Nonsense!
Economists have long known that under the widespread imperfect competition in our economy, raising the minimum wage can lower the critical marginal or incremental cost of hiring additional workers. Where there is no effective minimum wage, hiring more workers may require offering higher wages not only to new workers but also to existing ones, making the additional costs prohibitive. If the minimum wage is set high enough to be effective, increased hiring may entail an incremental cost equal to only the wages paid to the additional workers, since existing workers have already had their wages raised. Hence, companies may indeed increase employment when the minimum wage is raised.
The study was properly received as ingenious and outstanding. But it was only one of a number of recent studies showing that the minimum wage did not significantly reduce employment. Roberts' attacks on the American Economic Assn. and on Card, who is to receive the AEA's highest award for an economist under 40, are nasty and unjustified. I have served on the committees that make the awards; they are not given to those who do shoddy work.
Department of Economics
Editor's note: Eisner is a past president of the American Economic Assn.