There's a criticism of Card and Kreuger that seems devastating, though.
They assessed Full Time Employment (FTE), a statistic which counts full time work as 1 'unit' and part time work as 0.5 'units'. That's really alarming, because we want to assess hours worked, and they used a metric that can't differentiate 30 hours / week from 10 hours / week. It's even worse because cutting hours while hiring more workers is a known way to reduce benefit costs and increase flexibility.
Neumark and Wascher redo the analysis using hours worked, and find a nonsignificant downward trend instead.
I don't think this settles the minimum wage question, but I've never heard a defense of Card and Krueger's result against this charge.
They assessed Full Time Employment (FTE), a statistic which counts full time work as 1 'unit' and part time work as 0.5 'units'. That's really alarming, because we want to assess hours worked, and they used a metric that can't differentiate 30 hours / week from 10 hours / week. It's even worse because cutting hours while hiring more workers is a known way to reduce benefit costs and increase flexibility.
Neumark and Wascher redo the analysis using hours worked, and find a nonsignificant downward trend instead.
I don't think this settles the minimum wage question, but I've never heard a defense of Card and Krueger's result against this charge.
http://www.nber.org/papers/w5224