Economica (1999) 66, 473–88 Minimum Wage Legislation, Productivity and Employment By GIANNI DE FRAJA University of York Final version received 18 August 1998. The effects of minimum wage legislation are analysed with the assumptions that firms are able to alter the working conditions of their employees, and that workers have different prefer- ences about the characteristics of their job. The main findings tally with Card and Krueger’s (1995) recent somewhat puzzling empirical evidence about the effects of changes in minimum wage. I find that the effects of changes in the minimum wage on employment are limited, that there is a positive spillover on high-wage workers, and that there is bunching of workers at the minimum wage. INTRODUCTION Conventional wisdom that the imposition of a minimum wage reduces employ- ment has been challenged by recent controversial research: Card and Krueger (1995: CK hereafter) find that the overall empirical effect of minimum wage on employment is small or negligible and, in some cases, even positive. Their results are confirmed for the United Kingdom by the work of Machin and Manning and their associates (see the references cited in Machin and Manning 1996). Conventional wisdom relies on a model of competitive labour markets. In this model, workers are paid the value of their marginal product, and when a minimum wage is imposed, all the workers are dismissed whose marginal product is worth less than this minimum. This view of labour markets is incon- sistent with two further effects detected by CK: the spillover effect, according to which the wage of workers whose wage is above the new minimum increases as a consequence of a minimum wage hike; and the bunching effect, by which a large number of workers are employed at the minimum wage. It is of course well known that, if the assumption of a perfectly competitive labour market is relaxed, then a minimum wage floor may increase employment: this happens if the employer is a monopsony (Robinson 1933; Stigler 1946; or, in a model of efficiency wages, Manning 1995). However, the monopsony model does not seem an appropriate explanation of the empirical findings for the markets stud- ied, where, typically, employers do compete with each other for the labour force. (The archetypal industry is fast food restaurants.) CK themselves do not feel satisfied with the existing theoretical explanations and call for further research (ch. 11). I build a model on the simple, and in my opinion very plausible, foundation that the employment contract is multidimensional, that is that jobs differ in more than the wage attached to them: some jobs are harder, more tiring, less pleasant, require longer or more unsociable hours, and so on. I also, again plausibly, assume that workers are willing to work harder if they receive a higher wage. (Thus, in line with the efficiency wage literature, a worker’s pro- ductivity is endogenous and positively related to his wage. 1 ) Related to these  The London School of Economics and Political Science 1999 ps327$$p69 08-11-99 07:40:43