Doucouliagos and Stanley published a great review in 2009 with the key figure being a scatterplot of study size vs elasticity (elasticity = change in # jobs / change in min wage) [1]. This looks like a funnel centered around the true value. The funnel is wide where N is small and narrow where N is large. Two conclusions from the figure:
1. The true elasticity is ~0 (maybe ever so slightly negative). Only small increases in minimum wage were studied, so this isn't an argument in favor of drastic raises, just an argument that small raises don't hurt under prevailing conditions.
2. The funnel should be symmetrical because random error does not favor higher vs lower elasticity. Instead, the funnel is bent, which is evidence of strong publication bias in favor of results of the form "increases in minimum wage decrease the number of jobs."
FWIW I don't think minimum wage increases are a great way out of this mess because I think reduced workweek and basic income / minimum income / EITC are all better options. Still, I think this figure is pretty convincing evidence that economists place too much faith in their neoclassical model.
That's a pretty misleading statistic there. US unemployment in that time period has tended to increase in spikes on the order of a few months then fall rapidly in most cases but slowly since the 2007 recession. So you'd expect that in any randomly selected state with an unemployment rate over 7% it'll have decreased after one year in 90+% of cases.
If you're willing to be creative you can come up with a statistic that seems to support nearly anything. There are plenty of decent arguments and studies you can use in support of a minimum wage rather than something like this.
It comes from academic material, typical neoclassic economics textbooks teach this so I assume its from The Wealth of Nations. By increasing the minimum wage it locks out a portion of employees from being employed where the marginal product of labour is less than the minimum wage.
Because of this there are actually tonnes of papers studying its effects:
https://cdn.americanprogress.org/wp-content/uploads/2013/12/...