> we conclude that the second wage increase to $13 reduced hours worked in low-wage jobs [...] while hourly wages in such jobs increased by around 3 percent.
Sounds like some pretty intense statistical manipulation. The second wage increase was ~18% ($2/$11) not 3%. So they've fished for a larger pool that only had a 3% increase in wage and arbitrarily designated these as "low-wage jobs". Also very conveniently they've chosen to highlight only the second wage increase; it sounds like no data massaging could make the first one look bad. If they were honestly aiming at statistical significance the best approach was of course to bundle the two increases together. An overall 37% change in minimum wage should have much clearer effects than its individual halves.