You are doing a wrong comparison. To see if factor X has an influence, you need to compare two similar situations differing only in factor X and see the outcome, not compare two outcomes from different situations both not having factor X and conclude since there's a difference factor X does not matter.

So, if a unionized company in SF with higher salaries would be consistently hiring more people than essentially similar company with no unions and lower salaries, in the same environment - that might be a counter-example or at least a beginning of one (one still needs to consider the confounders - maybe CEO of the former one is a genius and of the second one is an idiot and had run the company into the ground, and that explains hiring differences).

Of course in SF and Jackson salaries are different - those are different markets. That has nothing to do with my argument.

> things are a lot more complex and unions don't "kill jobs".

Making a claim is not proving it. I've provided an argument why I think they do. If you want to argue against it, the least you could do is providing your own counter-argument. And no, "it's complicated" does not count as one. You have to point out which part of the complication, in your opinion at least, makes the conclusion wrong.

BTW the article itself essentially admits my point:

> Since workers for most gig economy companies are considered independent contractors, not employees, they do not qualify for basic protections like overtime pay and minimum wages. This helped Uber, which started in 2009, quickly grow to 700,000 active drivers in the United States, nearly three times the number of taxi drivers and chauffeurs in the country in 2014.

In other words, Uber could overtake taxi & chauffeur industry and provide jobs to 700K people because they weren't saddled with onerous regulations and union rules. That's not my words, that's NYT's words.