"Suppose that the two lowest-cost options for Acme Co. to produce Q amount of output X are as follows (and reckoned on an hourly cost basis):First, I want to give kudos to Don for spelling this out when many people (myself included) were confused by a previous post where he said that increased supply from a minimum wage might increase overall employment. Actually that still confuses me and it's not at all what he has here (here it's a demand shock, not a supply shock that's occurring). This provides a sensible (likelihood is another question I'll get to in a moment) explanation of a result that empirical analyses seem to point to - and a result that is not one that is particularly amenable to Don's own views on the minimum wage as policy.
1) 10 hours of low-skilled labor combined with 50 dollars of capital expenses;
2) 11 hours of skilled labor combined with 30 dollars of capital expenses.
If the prevailing hourly wage for low-skilled workers is $7.25, then Acme Co.’s hourly production costs will be $122.50 if it goes with option 1. ($72.50 for ten hours of low-skilled labor plus $50 of capital expenses.) If the prevailing hourly wage for skilled workers is $8.41 or higher, then Acme will use option 1; it will produce X using low-skilled rather than skilled labor. (If Acme employs 11 skilled workers at $8.41 per hour, and uses with these workers $30 of capital every hour, Acme’s hourly production costs are $122.51 – higher than the total costs of hiring ten low-skilled workers at $7.25 per hour along with $50 worth of capital each hour.)
Now let the minimum wage be raised to (say) $8.41 per hour. If Acme continues to produce Q amount of X each hour by employing ten low-skilled workers, along with $50 worth of capital, Acme’s hourly production costs would rise from $122.50 to $134.10. (Ten low-skilled workers at $8.41 per hour = $84.10; adding $50 of hourly capital expenses sums to $134.10 per hour.) But by instead employing 11 skilled workers at $8.41 per hour, along with $30 worth of capital, Acme’s hourly production costs will rise only by one cent, to $122.51."
It's also interesting, though, because reswitching situations like this are typically highlighted by left-heterodox economists and Sraffians, most notably during the Cambridge capital controversy. For all Don complains about fairly standard models with turnover, fixed hiring costs, and monopsony power as explanations of the minimum wage, this invocation of reswitching is a much bigger departure from received neoclassical economics.
So what do we make of it? In practice I doubt it's a major contributor to the lack of a disemployment effect in the best empirical analyses. A lot of the studies have focused on low-skill service sector work where the opportunity for this sort of reswitching seems like it would be minimal. It seems like it would be much easier as a manager to make low-skill workers more productive than it would be to change a production process, particularly because the reswitching usually involves a capital-labor substitution of some sort. Add in the fixed costs of making the switch in the first place and it just doesn't seem likely. But if there are any good examples where something like this is going on that would be really interesting and I'd love to hear about it. The result is so anomalous I think it's likely that lots of things are going on to drive the result and this might be part of the puzzle.