You may not like the reality, but it is the reality. Forcing higher wages costs jobs. It’s not complicated.
The United States has an entire history of different areas raising their minimum wage differently, which offered very extensive opportunity to evaluate the impact of those policies. The problem here is that it seems that, as far as typical minimum wage increases are concerned, there is absolutely no sign of a negative effect on hours worked, or employment. If you want a classic example, you can look at David Card and Alan Krueger (1992). They focus on low skill workers in the fast food industry and they find no traces whatsoever of what you're talking about. So, either there was no negative effect on worker, or it was absurdly small. By the way, it's not an unusual result -- it's typical of cross-sectional studies using microeconomic data.
Of course, it doesn't mean that if you can extrapolate these kinds of findings to huge minimum wage hikes and never see a negative effect. On the other hand, it does suggest that you might have
some slack when it comes to enforcing a minimum wage. It might actually, up to a point, be doing more or less what people who push for it want it to do. This isn't exactly too much of a shock: labor market dynamics are pretty hard to model convincingly well, so I'm not falling off my chair that an ECON101 graph doesn't say everything.
So, it's a bit complicated. As for Thomas Sowell and Milton Friedman, though they clearly were very smart (and Sowell still is smart), they are people with a peculiar political bent. Even if he is also very smart, would you realy trust only Paul Krugman? If you want the insider point of view of an economist who actually reads the scientific papers, he usually makes a very good job of conveying arguments in business cycle analysis. It's usually up to date, to the point and he does give you the correct intuition. I know for a fact he read the relevant papers when he was commenting on low inflation, monetary policy and fiscal policy around 2010 and 2011 -- because I read those papers and he was going through the same arguments. I think he was ultimately somewhat wrong, but for reasons he could not have known before 2015 or 2016. Specifically, I know some of the theory behind his claims is wrong (from the articles that came out years after), but I also know that some empirical evidence goes in the direction of large crisis time multipliers and there is no way to tell if he might not be right or partly right, but for reasons he ignored.
