The Economic Policy Institute (EPI) has been getting a lot of media attention for its estimate that 4.5 million American workers will benefit collectively by more than $5 billion in additional earnings this year from minimum wage hikes in 18 states that took effect yesterday. For example, see media reports here, here, here, and here that cite EPI's $5 billion estimate.
But it seems like there might be a little problem with EPI's estimates. Based on an analysis using the data in the table above from the EPI website, it would appear that the $5 billion in estimated higher earnings is at least several billion dollars too high, maybe more. For example, the minimum wage in Alaska increased yesterday by $0.04 per hour from $9.80 to $9.84 as a result of an inflation adjustment, and that will affect 12,000 minimum wage workers.
Even if each of those minimum wage workers is unrealistically assumed to be working full-time 40 hours per week for 50 weeks per year (2,000 hours annually), the $0.04 cent per hour increase would increase the earnings of Alaska's minimum wage workers by only $960,000. And yet EPI reports a nearly $19 million increase in earnings for minimum wage workers, which is almost 20 times too high.
According to the most recent BLS data on the characteristics of more than 2 million minimum wage workers nationally, a minority of those workers (41%) are employed full-time (defined by the BLS as 35 hours per week or more), and a majority (59%) are employed part-time (one hour per week or more). If we assume that the average full-time minimum wage worker is employed 35 hours per week and the average part-time minimum wage worker is employed 20 hours per week, that would mean that the average minimum wage worker is employed an average of 26 hours per week.
Based on that assumption, the minimum wage workers in Alaska would gain additional earnings this year of only $624,000, or about 97% less than the EPI estimate of nearly $19 million (see table above)! In total, the increase in earnings this year for minimum wage workers in the 18 states affected, using the assumption that the average minimum wage worker is employed 26 hours per week, would be $2.845 billion (see red text at the bottom of the second to last column above). That's nowhere close to EPI's $5.17 billion estimate above and in fact is about 45% lower!
Even if we assume that the average minimum wage works full-time at about 40 hours per week and the average part-time minimum wage workers is employed at about 24 hours per week, so that the average minimum wage worker is employed an average of 30 hours per week, the increase in earnings would only be $3.3 billion this year (see the last column in table above). Again, the EPI estimate of more than $5 billion is almost several billion dollars too high compared to the actual increase, which would be about 36.5% lower.
A preliminary analysis of EPI's reported data would suggest that its estimates of higher earnings this year for minimum wage workers at $5 billion is way too high, possibly by a factor of almost 2X. One issue is that EPI doesn't provide details of its analysis. But given the data it provides on: a) the number of workers affected in each state by minimum wage increases this year and b) the hourly increase in each state, there's no way that $5 billion in higher earnings can be accurate, even if we unrealistically assume that ALL minimum wage workers are employed full-time working 40 hours per week.
And as I pointed out last week on CD, whether the increase in earnings this year from 18 minimum wage hikes around the country is $1 billion or $5 billion, there's still the reality that it's an ironclad law of economics that to stimulate one group with public policies like minimum wage hikes (workers), there's another group in the economy that has to be as equally "un-stimulated" (businesses).
In the case of the 18 increases this week in state minimum wages, the EPI's estimates of additional earnings that will supposedly stimulate low-skilled workers this year - whether it's $1 billion or $5 billion - will unfortunately "un-stimulate" merchants, businesses, business owners and their families in those 18 states by the exact same amount - whether it's $1 billion or $5 billion.
Perhaps another lesson from the analysis above is that the media - and even respectable financial news organizations like CNBC, which reported the $5 billion figure two different times - apparently blindly accept what might be described as "fake research" from groups like EPI and the AFL-CIO (see CD post here) and never seem to question the methodology or details behind what are frequently questionable research methodologies.