Black Unemployment and the Minimum Wage

Summary: There is a lot of racism in the history of the US minimum wage – but the most important way that racism has been expressed is through exemptions to the minimum wage law which have kept workers of color from being protected as well as white workers. And evidence suggests that the minimum wage law does not increase unemployment amongst black workers.

On another thread, Jut wrote:

And, let’s face it: minimum wage laws were designed to price black people out of the labor market; why should we be surprised that they accomplished that goal?

It’s true that the history of the minimum wage is shot through with racism, including occasional examples of racists wanting a universal minimum wage in order to price non-whites out of the labor market. But the main way racists in the US have effected minimum wage law is by making exemptions for jobs that are primarily held by workers of color.

The first federal minimum wage law in the US was the Fair Labor Standards Act of 1938. But racist Southern Democrats lobbied heavily to exclude some classes of workers, especially agricultural workers, from the FLSA. For example, Florida Representative J. Mark Wilcox, debating the FLSA in 1937, said:

Then there is another matter of great importance in the South, and that is the problem of our Negro labor. There has always been a difference in the wage scale of white and colored labor. So long as Florida people are permitted to handle the matter, this delicate and perplexing problem can be adjusted; But the Federal Government knows no color line and of necessity it cannot make any distinction between the races. We may rest assured, therefore, that … it will prescribe the same wage for the Negro that it prescribes for the white man. … [T]hose of us who know the true situation know that it just will not work in the South. You cannot put the Negro and the white man on the same basis and get away with it. Not only would such a situation result in grave social and racial conflicts but it would also result in throwing the Negro out of employment and in making him a public charge.

As historian Juan Perria wrote (pdf link, long but excellent):

Specifically, southern congressmen wanted to exclude black employees from the New Deal to preserve the quasi-plantation style of agriculture that pervaded the still-segregated Jim Crow South. While they supported reforms that would bring more prosperity to their relatively poor region, they rejected those that might upset the existing system of racial segregation and exploitation of blacks.

President Roosevelt and his legislative allies recognized that in order to pass any New Deal legislation at all, it was necessary to compromise with Southern Democrats intent on preserving white supremacy. The compromise position was race-neutral language that both accommodated the southern desire to exclude blacks but did not alienate northern liberals nor blacks in the way that an explicit racial exclusion would. An occupational classification like agricultural and domestic employees, excluding most blacks without saying so, was just such race-neutral language.

In the decades since, anti-racist activists in the US have fought long and hard to reduce and eliminate those occupational classifications, with some success.

But what about Jut’s second claim – “why should we be surprised that they accomplished that goal?” Have minimum wage laws actually priced Black workers in particular out of the labor market?

Good empirical studies haven’t supported Jut’s assertion. From John Schmitt’s overview of the research (pdf link):

As Allegretto, Dube, and Reich note, however, a theoretical case can be made that minimum wages might instead improve the relative employment prospects of disadvantaged workers: “An alternative view suggests that barriers to mobility are greater among minorities than among teens as a whole. Higher pay then increases the returns to worker search and overcomes existing barriers to employment that are not based on skill and experience differentials.”62 A higher minimum wage could help disadvantaged workers to cover the costs of finding and keeping a job, including, for example, transportation, child-care, and uniforms.

Allegretto, Dube, and Reich’s (2011) own research on the employment effect of the minimum wage on teens looks separately at the effects on white, black, and Hispanic teens. For the period 1990 through 2009, which includes three recessions and three rounds of increases in the federal minimum wage, they find no statistically significant effect of the minimum wage on teens as a whole, or on any of the three racial and ethnic groups, separately, after they control for region of the country. Using a similar methodology, Dube, Lester, and Reich (2012) detect no evidence that employers changed the age or gender composition in the restaurant sector in response to the minimum wage. In a study of detailed payroll records for a large retail firm with more than 700 stores, Laura Giuliano (2012) found that teens from more affluent areas increased their labor supply (and employment) after the 1996-1997 increases in the minimum wage, while employment of teens in less affluent areas experienced no statistically significant change in employment.

We can also look at this in a cruder way: Does Black unemployment actually move up and down with the value of the minimum wage? Here are three graphs: A graph of Black unemployment, a graph of the minimum wage, and then an overlay showing the two graphs together.

Minimum wage and black unemployment in graphs

There are major periods (marked in blue in the overlay) in which black unemployment and the value of the minimum wage seem to be moving together. But there are also major periods (marked in pink) in which they seem to actually be opposites. Overall, it’s hard to argue, looking at this, that the correspondence is anything more than chance.

Furthermore, there are two big problems with arguing that the areas of correspondence shows that black unemployment is being drive by the minimum wage. First of all, Black and white unemployment rates virtually always move up and down together (although the Black unemployment rate is always higher). So it’s hard to argue that the minimum wage is causing unemployment amongst Black but not white workers.

Secondly, and more importantly, look at those gray bars in the chart of Black unemployment over time. Those mark recessions. And when you look at that, it becomes obvious that changes in unemployment rates are overwhelmingly driven by recessions. During recessions (including the time before and after the recession), and at no other time, unemployment rises steeply. After recessions, and at no other time, unemployment drops. The longer the time between recessions, the longer the drop.

Once you notice the link to recessions, it’s hard to see how the minimum wage can be the driving force behind black unemployment.

But maybe looking at Black unemployment is my mistake. If the idea is that higher minimum wages cause employers to favor white employees more, wouldn’t it make more sense to look instead at how the minimum wage corresponds with the ratio of Black to white unemployment? Yes, it would. But:


Again, no support for Jut’s theory there.

P.S. I just added a minimum wage category to “Alas,” for those who are interested.

This entry posted in Class, poverty, labor, & related issues, Economics and the like, Minimum Wage, Race, racism and related issues. Bookmark the permalink. 

9 Responses to Black Unemployment and the Minimum Wage

  1. 1
    Harlequin says:

    While I agree with your overall point, I’m not sure your graph overlays are the best way to show it. When you look at gross unemployment rate like that, it’s got some movement, obviously, but it’s constrained at the top and bottom by people who have stable good paying jobs and people who wouldn’t be able to keep a job with any nonzero wage. And the jump in recessions includes people who were well above the minimum wage before the recession. Only about 3 or 4% of black workers make the minimum wage, so you’re really looking at small changes in the distribution of less than 20% of workers, and this way you’re trying to show it doesn’t exist by looking at 100% of workers…there are just too many other factors there to draw conclusions. The research you linked to is a lot more persuasive because it looks at a group where the affected percentage is much higher.

  2. 2
    Ampersand says:

    Harelquin: Yes, I agree. :-)

  3. 3
    JutGory says:

    Why do none of your graphs pre-date approximately 1960?

  4. 4
    Ampersand says:

    Because I was lazy and grabbed the first tables with reliable sources I found. :-)

  5. 5
    JutGory says:

    Fair enough. I will try to respond tomorrow, but no promises. You deserve a response, but tomorrow is busy.


  6. 6
    Ampersand says:

    You deserve a response, but tomorrow is busy.

    No worries, and no hurry.

  7. 7
    gin-and-whiskey says:

    I am sure you know this, Amp, but for the posters who don’t:

    The unemployment rate is a tricky metric, because of who is classified as “unemployed” and who is not. Specifically:

    People who do not try to enter the labor force because of a bad job market; bad job skills; or lack of overall opportunities are not counted as unemployed. There may be only one employer in your town; she may have an ad requiring “3 years shop floor experience;” everyone there may be white; you may be a new high school graduate debating whether to try working or stay home… still, if you don’t spend the time to actively look for work, you’re not “unemployed.”

    Similarly, those people who try to enter the labor force, become discouraged, and cease looking, move into the above category. They’re not “unemployed” either.

    So when you say

    If the idea is that higher minimum wages cause employers to favor white employees more, wouldn’t it make more sense to look instead at how the minimum wage corresponds with the ratio of Black to white unemployment?

    I’m not so sure that is actually correct. A black employee who knows that only white employees will get hired is less likely to bother applying at all, which won’t actually show up in unemployment stats. This is also the reason that a lot of people with really poor job skills and qualifications don’t show up as “unemployed;” they have become tired of wasting time aiming for jobs they’ll probably never get.

    This can be difficult to track because can be a lot of assumptions involved–such as the issue of how you would expect black entry into the job market–but it is worth considering.

    Also, I squirm at things like graph 4 AD, which conveniently shows “blue” and “pink” to support its theory. But don’t trust them: print them out and look for yourself. Sure looks to me like from 1975 to 1980 the graphs both go “down, up, down” at about the same time. And if you look at 1995-2000, I’m not sure why that’s pink either–the whole of 1985-2000 seems like it should be blue. When they say “overall, looking at this, it looks like chance” that is because they made it look that way. Which is why I’d love to see the statistics on whether or not it numerically resembles chance.

  8. 8
    Glen Raphael says:

    Here’s Milton Friedman making the case back in 1966 using the data available at the time:


    “Before 1956, unemployment among Negro boys aged 14 to 19 was around 8 to 11 per cent, about the same as among white boys. Within two years after the legal minimum was raised from 75 cents to $1 an hour in 1956, unemployment among Negro boys shot up to 24 per cent and among white boys to 14 per cent. Both figures have remained roughly the same ever since. But I am convinced that, when it becomes effective, the $1.60 minimum will increase unemployment among Negro boys to 30 per cent or more.”

    FWIW, unemployment among black youth today is around 30%. (the rate for white youth is half that, it’s around 15%). Also note that reported “youth” categories changed at some point; it’s now 16-19 rather than 14-19.
    ( current numbers here )

  9. 9
    JutGory says:

    Amp, there is a lot in your post so my response is will be kind of long and piece-meal:

    To your larger point regarding the research and studies, I cannot necessarily speak to the studies and methodologies used. In addition, I don’t know all of the underlying facts they considered. They looked at New York minimum wages, as well as those in New Jersey and Pennsylvania, among other, I presume. That can get very dicey, though because you are dealing with potentially very different sets of data.

    For example, my state has four different wage classes: 1) Federal small employer; 2) State small employer; 3) Federal large employer; 4) State large employer. This could be contrasted with a state with no minimum wage (Mississippi?) that would only have the two federal classifications. In my state, the state minimum wage is greater than the federal one and let’s assume, by way of example, that the wages in the 4 classes are $4.50, $6.50, $7.50, and $8.50, respectively. A $0.50 increase in the federal minimum wage would have no effect on wages here, but would affect wages in Mississippi. It would only be when the federal wage exceeds the state wage that we would expect an impact. I do not know if or how those studies accounted for that. I presume they did, but I do not know. It looks like they considered many possible responses to changes in the wage, including reduction in employment rates for “less-skilled, less-educated New Yorkers.”

    Aside: did you review all of the underlying data, and the studies cited? I doubt it.

    I ask because I tend to be skeptical of things that I do not understand or am unable to review. So, I expect that, like you, in such instances, you look for voices you can trust. Glen Raphael, for instance, cited Milton Friedman. For my part, I tend to look at Thomas Sowell and Walter Williams, neither of whom could be considered academic slouches in the area of economics. Williams is even cited on Page 17 of your link.

    Looking at the racist background of minimum wage laws, Sowell does not start with the FLSA. He begins with the Davis-Bacon Act of 1931. He observes that: “The last year when the black unemployment rate was lower than the white unemployment rate was 1930, the last year before there was a federal minimum wage law.” Williams offers similar critiques and this link provides links to several of his articles on the matter.

    I could be accused of confirmation bias for trusting in conservative economists. But their critiques make sense to me and I presume that their analyses are just as fact-based as your cited study. As I said, academically, their reputations seem solid.

    So, moving on to your assertion regarding my “claim” or “theory” that increases in the minimum wages price Black workers out of the market, I have to give credit to Williams and Sowell. Sowell’s assertion about the effect of Davis-Bacon, above, could be dismissed on the correlation-causation basis. In any event, I may have stated it inaccurately. Williams described it as the minimum wage lowering the price of discriminatory conduct. If a Black guy wants to work for me $2.00/hour, but the White guy wants $3.00/hour, it costs me $1.00/hour to discriminate against the Black worker by hiring the White worker. That is the price of my discrimination. However, because my love of money is stronger than my dislike of Black people, the White guy will lobby to raise the minimum wage to $2.00. If that occurs, it no longer costs me any money to discriminate against the Black worker; so I hire the White worker. He presents a similar analysis here.

    [Snark: Of course, to address this problem, we have enacted laws to make it illegal to discriminate in this way, thereby eliminating any disparate impact experienced by Blacks as a result of increases to the minimum wage.]
    Now, in all fairness, Sowell acknowledges that inflation can mitigate the effects of the minimum wage, and Williams acknowledges that education, environment, and family situations can also negatively affect one’s employability. So, consistent with your study, they do not appear to deny that other factors can impact wages, but that, in isolation, raising the cost of labor will result in a decrease in demand, especially if the value of the labor is less than the cost.

    Your study appears to acknowledge that, but focuses on other ways to compensate for an increase in wages (e.g. reducing other benefits, raising prices, improving efficiency, cutting other wages). As I noted on the other thread, when confronted with an increase, we terminated the employees. Confronted with another increase, we would have to raise the pay of our least paid employee (or replace her with someone with more skills.

    Finally, and I know this is very long-winded response, not all increases are the same. A nickel or dime increase would likely have no noticeable effect on employment. An increase to $15.00/hour, on the other hand, could be substantial. That is why I, along with g&w, am excited to see certain cities experiment with this. The results could be telling.