Who captures a wage subsidy?

Basically every benefit we give to the working poor ends up being an indirect subsidy for business - see, for example, employers telling their employees how to obtain food stamps.

One of the complaints about a wage subsidy over a higher minimum wage is that it will just be captured by employers, who will pay their employees less by that amount.  That’s also potentially true of a basic income, and with a minimum wage, employers may opt to gain non-monetary compensation (e.g. terrible hours).

Now, here’s where the limits of my economics education probably show a bit, in that I’m not familiar with the literature on how, empirically, this works out.  (Maybe @xhxhxhx can chime in.)

I realized that this is actually related to the marginal productivity of labor - how much revenue (and thus, potentially, profit) does each additional employee bring in, across the whole economy?  There are limits to this based on the amount of equipment/capital needed for a marginal employee or marginal hours, including facility size, as well as the potential customers it might bring in (e.g. why haven’t they hired additional labor already?).

The reason for this is that to determine the leverage of a low-wage employee under a wage subsidy system, we need to know how many potential jobs our wage subsidy can create, and at what quality.  How easy is it for an employee to just walk right out of the store, walk right in to another store, and get a new job?  Even if the pay is somewhat lower, this creates a much stronger incentive against bad hours, bad bosses, and unsafe practices, about which employees will then either demand higher pay, or just tell the employers to knock it off.

However, that increase in leverage only occurs if enough potential jobs emerge, and this is more or less an empirical question.

The greater the marginal increase in the number of jobs per marginal decrease in minimum wage prior to subsidy, the more of the subsidy that will be captured by the workers.  However, if cutting the minimum wage creates no new jobs, then leverage doesn’t change much at all and employers capture the majority of the subsidy.

If the leverage is high enough, wages may even be driven higher than they were prior to the subsidy, depending on employer margins that they were exploiting leverage over against employees.

However, since employers capturing part of the subsidy is potentially true for all subsidies for the working poor, even rental vouchers or healthcare, it has to be compared with other alternatives (such as basic income).

(For my preferred implementation, the accompanying decrease in minimum wage should be lower than the wage subsidy, and the wage subsidy should be paid directly to the employee, thus at least not resulting in a decrease in effective income even if the entire subsidy is captured.)