The cost of bias lingers.

Photographer: Charles 'Teenie' Harris/Carnegie Museum of Art/Getty Images

Bigotry Is Expensive

Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.
Read More.
a | A

It seems like every day there’s a new battle being fought over discrimination in the U.S. There was the Ellen Pao trial and its claims of sexual bias at Silicon Valley's leading venture capital firm, the continuing revelations of endemic racial discrimination in Ferguson, Missouri, and the so-called religious freedom law in Indiana that many believe is a thinly veiled cover for anti-gay discrimination.

If racial and gender discrimination were purely matters of fairness, ending them would still be a worthy cause. But there is another reason to combat discrimination -- it boosts the economy.

The basic reason is misallocation of resources. In an economy functioning at tip-top efficiency, people do the job that they’re most efficient at doing, relative to other jobs they could be doing, and relative to other people in the economy. That’s called the principle of comparative advantage, and it dates back at least to the famous 19th century economist David Ricardo. Notice that nowhere does the principle of comparative advantage leave room for race or gender. These things, in and of themselves, are simply not important for economic efficiency -- the only thing that matters is how well people can perform their jobs.

So if a society bases its decisions of who gets which job on race and gender, it’s going to be sacrificing efficiency. If women aren’t allowed to be doctors, the talent pool for doctors will be diluted, and wages will be pushed up too high, choking off output. This would be true even in a bizarro world where every man was a better doctor than every woman! Of course that’s not even remotely true, but the point is, the theory of comparative advantage doesn’t care about average differences in absolute ability. If you’re making rules about which type of people are allowed to do which type of job, you’re hurting the economy.

Just how big of a difference does this make? A team of top economists has recently studied the question, and their results are pretty startling. In “The Allocation of Talent and Economic Growth,” economists Chang-Tai Hsieh and Erik Hurst of the University of Chicago Booth Business School and Charles Jones and Peter Klenow of Stanford estimate that one fifth of total growth in U.S. output per worker between 1960 and 2008 was due to a decline in discrimination. From their abstract:

In 1960, 94 percent of doctors and lawyers were white men. By 2008, the fraction was just 62 percent. Similar changes in other highly-skilled occupations have occurred throughout the U.S. economy during the last fifty years. Given that innate talent for these professions is unlikely to differ across groups, the occupational distribution in 1960 suggests that a substantial pool of innately talented black men, black women, and white women were not pursuing their comparative advantage.

The authors’ approach is clever. They treat discrimination as a tax. If you think about it, this makes sense -- just as taxes discourage business activity, discrimination in employment or education discourages its victims from taking certain jobs or getting training for certain skills. The authors explicitly allow for the possibility that different groups might have different average ability levels with respect to different occupations. They also count home production -- child care and housework -- as a real contribution to gross domestic product. They put these assumptions into a Roy model, which is a very flexible general equilibrium model that economists typically use when evaluating people’s occupational choices. 

In 1960, as you might expect, occupations were very segregated by race and gender. In 2008, they were far, far less segregated. Plugging these numbers into their model, the authors find that about 15 to 20 percent of the growth in output per worker since 1960 can be attributed to this reduced sorting. Since the percentage of Americans with jobs grew substantially between 1960 and 2008, this means that more than a fifth of U.S. economic growth during that period can be attributed to reduced sorting by race and gender. The model also predicts falling wage gaps between different races and genders, which is exactly what we have seen in real life. 

In other words, reducing discrimination made us a much richer nation.

 The authors’ model also allowed them to estimate just how far discrimination has fallen, by examining the changes in the labor market and working backward. Here are some graphs showing how their measure of barriers has fallen for various occupations. First, white women:

And for black men:

 The authors also estimate what would happen if discrimination were to fall all the way to zero. Output per worker, according to the model, would rise a further 10 percent to 15 percent! 

The implications of this result couldn’t be clearer. Eliminating race and gender discrimination continues to be an economic priority for the U.S.

One way to do that is to promote more economic competition. The late Nobel-winning economist Gary Becker showed that competition can exert powerful pressure on companies to hire and promote workers based on efficiency instead of on race and gender. We probably did a lot of that unwittingly in the late 1970s and early 1980s, with deregulation and free trade. But competition will always be imperfect, and we need other tools as well.

Simply reducing individual prejudice would be a wonderful way of attacking discrimination at the source. That’s why the continuing social movements for greater racial and gender equality are valuable -- not just for reasons of fairness but for our economic future. 

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Noah Smith at

To contact the editor on this story:
James Greiff at