Increasing Returns and Immigration

Will income per capita tend to rise or fall as the population increases? There are two forces at play. The first is the ratio of capital to labor, with capital defined broadly as anything which can be used to produce another thing. As the number of people falls, they each get to use more capital. If the amount of capital is fixed, such as in the case of land, a fall in population leads to a rise in living standards, and a rise in population leads to a fall in living standards. There are decreasing returns from having more people in such a world.

There are also, however, gains from specialization and ideas. These have increasing returns with the number of people. An idea produces a lot more benefit when there is a billion people than when there is a million people. Since ideas require a fixed cost to invent, and then have basically no marginal cost to be used by more people, you get far more ideas with more people, and this does not even include the possibility of ideas being randomly generated by people.

Imagine a job which requires no training, and in which there is no improvement over time. Someone in their first day on the job would do just as well as someone in their 20th year. Because each person is just as capable as any other, the cost of finding a worker is close to zero. A rise in population would result in no change in per capita income. Each person would enter the job, and produce the same as someone who has worked there a long time.

Other jobs require more specialization. Imagine a software engineer. In order to be employed, they need companies to make considerable investments into identifying possible roles. In order for the companies to invest in finding possible roles, they need for there to be many people who have invested in specialized skills. If we assume that having more people and more jobs makes it easier to find a job, the industry will either converge to the optimal level, or to nothing at all, depending on the number of people. The idea here is essentially Diamond’s, from his 1982 paper “Aggregate Demand Management in Search Equilibrium”. Externalities in search costs means that the total number of people with those skills matters, and that the total size of the population also matters.

So what will the effect of immigration be on wages? We must distinguish between short run and long run effects, which may be completely different from each other. It is unfortunate that our studies are best able to detect short run effects. If the number of jobs are the result of investments which were made in the past, a surge in immigration would lead to a fall in wages in the short run. If the job is specialized, though, in the long run there will be more jobs and higher wages.

I do not think it is reasonable to believe that technology is an industry typified by decreasing returns. So when we are discussing the H-1B visa — or any visa, really — we should not expect restricting immigration to lead to higher wages for American workers in the long run. On the contrary, we should expect it to lower them. For all the conflicting arguments about the effect of immigration on the wages of unskilled workers, there can be no argument about the effect of specialized workers. And those who oppose high-skilled immigration are either misinformed about the consequences of immigration, concerned only with the immediate future, or are possessed by an appalling racial prejudice.