The Minimum Wage, will 15$ increase unemployment?
If you read my previous article on the minimum wage, you can skip to part 2.
Part1: Every student who studies microeconomics 101 sees the equilibrium of the labor market, which assumes that if policymakers will impose a minimum wage higher than the equilibrium point for certain workers, those workers will lose their jobs.
Right-wing politicians nowadays, rely on this theory whenever the idea of minimum wage is proposed to increase low-workers income and reduce the economic inequality.
In many countries the minimum wage hasn’t increased much over the years and especially in the U.S.
Actually, in the U.S, the real minimum wage in 1955 was equivalent to the minimum wage there is now.
In order to illustrate visually, hers is a graph of the minimum wage (in 2019 Dollars and Euros) in the U.S and France, taken from Thomas Piketty’s research:
As you can see, the US minimum wage has decreased while France (as a representative) increased dramatically.
Based on these assumptions of job loss, and most economists’ beliefs at that time. Two economists made a research in 1992 which changed a lot of economists view of the minimum wage and since then got a lot of data supporting their conclusion.
In Economics, as opposed to other science fields such as Biology, Chemistry etc. One cannot perform a study in a laboratory. In order to apply the same technique in Economics, we need to make the study in real life. If we examine the US for example, the researchers need to take 2 countries of US and for one country impose a minimum wage and in the other do not and see what the outcomes are going to be.
Nevertheless, as you probably imagined already, this type of study can’t be made on purpose and one cannot split the US into 2 countries.
Therefore, in order to apply that technique economists need a natural experiment. Which is an experiment that happens naturally, without the control of the researchers.
The same was applied for the minimum wage.
In 1992 there were two American Economists David Card (From Berkeley) and Alan Krueger’s who wrote a paper on the impact on fast-food employment of the 1992 increase in the New Jersey state minimum wage.
In their study, they checked the employment effect on the country that the minimum wage was raised in (New Jersey) and its neighbor (Pennsylvania) at its east area which is closer to the border.
They found, according to them: “no evidence that the rise in New Jersey’s minimum wage reduced employment at fast-food restaurants in the state.”
This study has got much attention for several reasons. First, the study had found a result which was not consistent with previous studies which are now condidered of lower quality. Second, an accusation of biased survey data from David Neumark and William Wascher. When they replicated their estimation using payroll data, theyfound a negative result. After their paper — https://pubs.aeaweb.org/doi/pdf/10.1257/aer.90.5.1362 was published Card and Krueger replicated it again using Bureau of Labor Statistics’s (BLS’s) employer-reported ES-202 data, and they found results consistent with their survey data and called into question therepresentativeness of the sample assembled by Berman, Neumark, and Wascher. https://www.nber.org/system/files/working_papers/w6386/w6386.pdf
Anyway, since then the minimum wage has been studied extensively. For instance, Arindrajit Dube from University of Massachusetts Amherst, T William Lester, Michael Reich generalized the Pennsylvania New Jersey with several advantages. First, they used a much larger number of cases. Second, studying the effect over time enabled them to test the possibility of longer-term effects. Using this large sample of border areas, the conclusion of Dube, Lester, and Reich find “no employment effects of minimum wage increases”.
They also further suggest why studies have found negative effects:
“By generalizing the local case studies, we show that the differences in the estimated elasticities in the two sets of studies result from insufficient controls for unobserved heterogeneity in employment growth in the national- level studies using a traditional fixed-effects specification. The differences do not arise from other possible factors, such as using short before-after windows in local case studies”.
The large negative elasticities in the traditional specification are generated primarily by regional and local differences in employment trends that are unrelated to minimum wage policies. This point is supported by our finding that neighborhood-level placebo minimum wages are negatively associated with employment in counties with identical minimum wage profiles”.
According to a different study of Dube: https://academic.oup.com/qje/article/134/3/1405/5484905
“jobs paying below the minimum decreased — since wages rose. But at least as many jobs were added at the new, higher wage — meaning jobs were upgraded, not destroyed”.
There are more studies of course, but in general it seems that the main conclusion of the empirical literature, from that studying past minimum wage increases, the effect on employment was either approximately zero or very small.
Part 2
How come minimum wage doesn’t cause job loss? The answer is that markets are not perfect and specifically the labor market. Another model that illustrates what could happen is if firms have “monopsony power”, meaning one buyer/firm in the market of labor that can set the wage independently.
In this market, without a minimum wage, the wage would be W2 with Q2 and after minimum wage of W3, wage would increase to W3 and employment would not change. Even more surprising, with these assumptions, increasing the wage to be between w3 and w2, for example w1, would decrease unemployment and the firm will actually hire more workers!
Even if there is no actually one buyer, firms can still have monopsony power due to information advantage, frictions etc. And therefore, the model is still valuable even if there is no actually 1 firm in the specific market.
This, however, doesn’t necessarily mean that a 15$ raise will replicate this result. A too much high minimum wage can definitely lead to higher unemployment, the monopsony power is limited and usually not as extreme as it is shown in the monopsony model.
In order to understand my point, take it to the extreme, in a 50$ minimum wage, of course unemployment will increase, no economist would doubt it.
How can we know if a minimum wage is too high? There is no perfect answer but there indeed there is some measurement that can indicate.
The Congressional Budget Office for example, thinks the 15$ minimum wage will cause 1.4 million Americans to lose their jobs by 2025.
However, in my opinion, this suggestion is not very convincing due to the current literature. A recent survey of the evidence in the Journal of Economic Perspectivesby Alan Manning — https://www.aeaweb.org/articles?id=10.1257/jep.35.1.3 suggests “the currently observed range of minimum wages apparently does not include the turning-point.”
A popular way among economists to measure if the minimum wage is high or low is to check what is the ratio of the minimum wage to the median wage. According to a different study of Dube and Attila Lindner, that studies minimum wage differences between cities. Cities like Seattle, Los Angeles and Chicago were around 63 percent of the median, sometimes stretching up to 80 percent. They found no significant change in the unemployment rate.
A 15$ will be higher than this ratio in some cities and therefore can definitely lead to employment loss. This point was also emphasized in the Chicago Booth economic council poll. In the council they were asked: “If the federal minimum wage is raised gradually to $15-per-hour by 2020, the employment rate for low-wage US workers will be substantially lower than it would be under the status quo”. The results were almost a normal distribution:
As you can see in the chart, a 15$ minimum wage effect on employment does not have a strong consensus, however, this poll was for 2020 and not 2025 as the current policy suggests. But since policymakers plan on raising it incrementally, they will be able to stop raising it after a certain point if large employment losses will be seen.
To conclude my view, although increasing the wage to 15$ would probably indeed cause employment loss, the current evidence suggests this loss will still be relatively small compared with the large benefits of lifting more than 900,000 workers from poverty according to CBO (which is a low estimate! due to the large employment effects they included in their calculation), and increasing wages for 17 million workers whose wages would otherwise be below $15 per hour.
.
