That still isn't the right question because it's a false dichotomy.
At any interesting wage, a minimum wage increase will increase unemployment. An increase from $.10/hr to $.50/hr does nothing because approximately zero people make less than $.50/hr to begin with. Once you get into wages that people are actually paid, with high probability there will be some jobs that can't be justified at a higher wage.
Imposing a minimum wage is taking from the people at the very bottom (who become unemployed) in order to give to the people at the almost-but-not-quite bottom. Which is why the whole exercise is folly.
It doesn't matter whether the higher wages outweigh the increase in unemployment in any particular scenario because there is a third option -- increase income without imposing the entire cost on employers who hire people at the bottom.
The simplest way to do that is a UBI. The UBI supplements everyone's wages, which brings up the lowest paid workers to the minimum, and is paid for from general taxes, which minimizes the economic damage by spreading out the burden (and having the rich pay their proportionate share).
This is not obvious. Increasing the income of the almost-bottom could act as a stimulus which drives down unemployment. It is possible that this effect is greater than the unemployment caused by the minimum wage directly.
Having said that, I agree that the UBI is a generally better way of accomplishing this.
You're essentially arguing the butterfly effect. It could also be the exact opposite, if the employees use the money to pay down debts when the employers would have spent it on new consumption. Or various completely arbitrary outcomes that are always possible whenever making any change to a complex system.
But the real flaw in the stimulus theory is that it assumes the conclusion. If the net loss from people losing their jobs to begin with is more than the net wage gain to the people who kept their jobs to begin with then there is no extra money to create a stimulus to recover from that situation. If it isn't then you didn't need the stimulus effect to come out ahead regardless. (Though you're still even better off with a UBI.)
If a company pays 65% of their employees the minimum and wages go up 5% the company doesn't fire most of its workers it grudgingly pays more.
In a lot of cases the existing low wage is actually below market now and subsidized by benefits like food stamps.
It depends on what the company's alternatives are. Many companies have only 5% margins to begin with. And most people who suggest minimum wage increases argue for more than just 5%.
> In a lot of cases the existing low wage is actually below market now and subsidized by benefits like food stamps.
That is one of the problems a UBI fixes. Since the money isn't contingent on employment, employers can't pay less against the leverage that without a job the employee would lose their food assistance.
But it doesn't always hold in small markets or markets with little variety. Consider towns where there are only a few important employers.
I still agree that imposing the costs on the employers is arbitrary.
Obviously the question then becomes how much the UBI should be, but at least it gets us away from the folly of a minimum wage.
------
Every minimum wage increase is different!
Even the staunchest free market economist wouldn't claim that a hypothetical minimum wage increase from $.10/hr to $.50/hr in Seattle would increase unemployment. And even the staunchest socialist economist would not claim that a minimum wage increase from $8/hr to $800/hr wouldn't increase unemployment.
Similarly, how can you compare an increase from $5-6 in New jersey in 2005 to one from $11 to $13 in Seattle in 2017? You can't just make blanket statements about this stuff.
The question is not "Are minimum wage increases good in general?" the question is "Given the potential tradeoffs, is this specific minimum wage increase good?"