Great Graphic: Evidence of Minimum Wage Impact

Economics was never strong on experimentation. For the past generation, it has become even more abstract, constructing mathematic models of the complex world.   There is an opportunity here to a hypothesis about minimum wage without resorting to deducing it from first principles or using numerous equations.  

Danny Vinik walks you through it in an notable and short article in New Republic.    He cites studies that show three things.  First, 13 states have raised minimum wage (five via legislation and 8 through indexing) from the middle of 2013 through the middle of 2014.  Second, the lowest decile Americans saw an increase in wages that comes the 13 states that hiked minimum wage.  Third, those states that hike minimum wage did not experience slower job growth as would be the hypothesis of many economists and investors.  These Great Graphics come from Vinik's article.  

The first part of the argument comes from a study by Elise Gould at the Economic Policy Institute. She found that during the period under review, real wages of the bottom 10% of age earners rose 0.3% or 2 cents an hour. Every other deciles experienced a decline in real wages.  

Looking deeper at the  bottom decile, Gould found that in the thirteen states that hiked minimum wage, the real wages of rose 0.9%. In those other 37 states that did not hike minimum wage, real wages of the bottom 10% fell 0.1%  

So far, so good.  Nothing very controversial.  Now, however, we need to see what has happened to employment growth in those thirteen states.  For this Vinik draws on work by Jared Bernstein, a former chief economist for Vice President Biden.  Bernstein found that job growth in the control group, which are the 37 states that did not hike the minimum wage, was 1.5%.  

He found job growth in the 13 states in which minimum wage rose was 1.8%.    It is not that the difference is significant.  The power is that it refutes the orthodox (null) hypothesis.

Of course this is not the last word on the subject.  We always have to be willing to treat our ideas as hypotheses.  Our information set is always incomplete.   Over time conditions may change and the data may change.  No study is perfect.  None will end the debate.  We need to be particularly sensitive to the accumulation of data points such as these that go against inherited wisdom.  At the same time, it simply does not suffice to say that law of supply and demand dictates that the higher wages reduce demand for labor.