The question that most American minimum wage research addresses is whether raising the minimum wage reduces employment or not.  For the last several decades until the mid 1990s, the answer was, yes, a wage increase means some loss of jobs and wage hours.  Then in 1995, David Card and Alan B. Krueger’s book Myth and Measurement: The New Economics of the Minimum Wage combined several of the authors' studies that showed no significant employment impact following an increase in a minimum wage. (A later paper of theirs is included on this Research page.) Since then, criticisms of Card and Krueger’s work, together with renewed reassertions of the traditional assessment, have been numerous, and the controversy continues. 

            One reason that the controversy persists is that the welfare-to-work reforms since the late 1990s are shifting the profiles of both the American poor (who are less likely to be on welfare than they used to be) and low-wage workers (who are more likely to be poor).  How well does the traditional analysis of the minimum wage’s impact fit the new profile of the working poor?  Since minimum wage studies require many years’ worth of statistics to work with, returns that are definitive enough for a new consensus to emerge are only now beginning to come in.  

            Among the critics of Card and Krueger's revised approach, the most prolific scholars has been David Neumark and William Wascher, one of whose papers appears on another site page.  Their critiques range from demonstrations that a minimum wage increase helps some families but impoverishes others, to a study of the wage’s impact on the enrollment of teenagers in school.

            Among the approaches economists use, some studies have drawn on data from one of two “laboratories” where contrasts in the impacts of different wages can be observed.  One has been the states in the U. S. that continue to raise their state minimum wages above the federal level; comparisons can be drawn before and after an increase within a state, or between states, or between state minimums and the federal minimum wage.  The other laboratory is England, where a national minimum wage was instituted for the first time in 1999 (although individual industries set minimum wages prior to that date).  In the past few years, the state studies have yielded optimistic reports but few detailed studies, while more detailed research, sampled here, has come from England.

             A skeptical non-economist might scratch his or her head over why the employment impact of a minimum wage increase is so difficult to measure conclusively and might conclude that political leanings must be playing a large role. Be that as it may, economic researchers, as they set up each study,  commit themselves to certain assumptions and certain methods and reject others, and the intense debate is often about those choices more than about the small economic stakes of minimum wage change.  Is it a sound starting point to expect that the price of human labor will behave in the market in the same ways that the prices of commodities do?  Should one assume that the employment impact on teenagers of a minimum wage change will also hold true for low-wage adults (the standard model) or not (the alternative model)?  Should one assume a fixed economic pie that will require employers to cut back somewhere when a wage increases (standard model) or should one allow for business growth following a wage increase?  Are small changes in the levels of benefits or on-the-job training worth considering as significant effects of minimum wage change?  And how long after a minimum wage increase should one expect to be able to reliably identify the results, good or bad, of that increase?  If one looks ahead only a few months, the results (such as short-term changes in employment) may be accurate but not especially significant.  If one looks at the results after a couple of years, they may be more significant but other factors in addition to the wage change have been at work by that time. 

         So the skeptical reader of the research literature should be cautious: partisanship may be a reality, but so are the difficulties of economic measurement and so are the uncertainties in the field of economics about how much and what kind of evidence are sufficient to refute a theory.

                                -- Brock Haussamen; revised August 2007