Don Boudreaux lists out some various possible effects of an increase in the minimum wage, all of which hurt low skilled workers:
First, neither I nor any other economist insists “that the minimum wage must raise unemployment.” Raising unemployment is a possible – indeed, likely – effect of a higher minimum wage; it is not, however, a necessary effect. The argument is that a higher minimum wage reduces the full range of the attractiveness of employment options open to low-skilled workers. The effects of a higher minimum wage might play out exclusively in the form of more-intense demands on workers – that is, a demand by employers that workers work harder and produce more output per hour. The effects of a higher minimum wage might play out exclusively in the form of reduced fringe benefits for workers. The effects of a higher minimum wage might play out exclusively in the form of fewer paid hours of work per year for low-skilled workers who nevertheless all keep their jobs (and who continue to be hired). The effects of a higher minimum wage might play out exclusively in the form of fewer jobs for low-skilled workers.
In a previous post, he explains why these effects (especially disemployment) might not be visible in the data:
Bryan is correct. And not only might a minimum-wage hike that was anticipated not result in any observable layoffs at the time it kicks in, but also the unemployment rate of low-skilled workers might not jump noticeably or detectably as a result of the higher minimum wage. The reason is that, once employers anticipate the looming hike in the minimum wage, they then – before the actual hike – start trimming their workforces (probably mostly through attrition and reduced hiring). Some, perhaps many, low-skilled workers, discouraged by the increasing difficulty they face when searching for jobs, respond by dropping out of the workforce before the minimum-wage hike actually happens. By the time the minimum-wage hike does occur, the resulting increase in the measured rate of unemployment of low-skilled workers might well be genuinely small.
And a skilled commentor explains why the most famous pro-Minimum Wage study might fall prey to these reasons:
This would apply to Card and Krueger as well, the legislation in NJ was passed in 1990 but the increase didn’t take place until 1992.