“Dey Took Err Jerbs”: Immigration and the Lump of Labor Fallacy

Why the economy is not zero-sum

Zixuan Ma is a PhD student in economics at George Mason University. You can keep up with him by both subscribing to his Substack and following him on X.

On December 22, then President-elect Trump announced the appointment of Sriram Krishnan, a successful venture capitalist, as a senior adviser on artificial intelligence. A model immigrant, Krishnan came to the US at 23 to work at Microsoft, later becoming a naturalized citizen. What should have been a discussion of AI policy and a celebration of the success of high-skilled immigration soon turned into a civil war on the right, after nationalist influencers highlighted Krishnan’s past comments endorsing more high-skilled immigration. The H-1B visa program became the focus of the debate, which exposed a deep division among the Trump coalition, between the techno-libertarians led by Elon Musk and the national-populists led by figures like Steve Bannon. Trump eventually came out in favor of the H-1B visa, but it remains to be seen what his administration does on this issue.

At the heart of the objection by the national-populists are arguments based on what economists have called the “fixed pie fallacy,” but in this case can be referred to as the “lump of labor fallacy.” “It’s about taking American jobs,” Bannon has claimed. “A tech company used the H-1B visa to replace [American workers],” JD Vance bemoaned when criticizing the visa as he was running for Senate in 2022. From the other side of the political spectrum, Bernie Sanders chimed in against the program, arguing it replaces “good-paying American jobs with low-wage indentured servants from abroad,” forming an unholy alliance of socialists and nationalists.

According to this logic, there is a set, or at least highly inelastic, number of jobs in the United States. This is true in tech, just like it presumably is in other fields like retail or oil extraction. If more people are allowed in the country, then more Americans are going to either be unemployed or expect much lower wages for their work.

Season 8, Episode 7

This fallacy has long been known to economists. Adam Smith in The Wealth of Nations (1776), challenged zero-sum assumptions prevalent in mercantilist theories. Later, David Ricardo and John Stuart Mill further explored how trade, innovation, and productivity could expand the economic “pie.” And in the 20th century, President Reagan’s economic guru Milton Friedman famously taught that “most economic fallacies derive from the tendency to assume that there is a fixed pie, that one party can gain only at the expense of another.”

Yet this is not how the economy works. Newcomers do in a sense compete for jobs within one industry. But they also buy homes, food, clothing, and shelter within the United States. If employers prefer the labor of immigrants, whether due to them being better workers, cheaper, or some combination of the two, then it must mean that firms are likely to become more efficient after hiring them, further stimulating economic growth. If a tech firm pays less money for labor, even assuming no gain in efficiency otherwise, they do not simply burn the money but invest it or put it back into the economy by returning it to shareholders. Even if all the savings go to CEO compensation, and executives spend that money rather than place it in the stock market, this still increases demand in the economy.

The lump of labor fallacy assumes that the demand for labor is static: there is a set number of jobs and that introducing more workers must reduce employment opportunities for those already in the workforce. American history is filled with counterexamples. When the US gradually allowed women into the workforce, it didn’t cause massive job losses among men; when segregation was abolished in the South, black workers didn’t cause mass unemployment among whites; when the 13 colonies established free movement of people among the states with the Articles of Confederation in 1781, this did not cause economic upheaval. The vast majority of native-born Americans are descended from immigrants, including the national-populists vehemently opposed to H-1B visas. Did their ancestors take away American jobs? Or did they contribute to the economy by working with the natives?

Even setting aside their roles as consumers, what restrictionists fail to realize is that high-skilled immigrants not only take existing jobs, but create new ones by becoming entrepreneurs themselves. The most salient statistic is perhaps the fact that 45% of Fortune 500 companies were founded by immigrants or children of immigrants. In AI, a report from the National Foundation for American Policy concluded immigrants also founded or co-founded 65% of the top companies. Elon Musk, maybe the greatest entrepreneur in history, was once on an H-1B visa. All of this has happened in spite of the restrictiveness of the immigration system. In many ways, high-skilled immigrants are the lifeblood of American innovation. Even low-skilled immigrants, by taking the most unglamorous and unpleasant jobs, enable natives to be productive in other ways by lowering the costs of labor. One may have other arguments against immigration, but the economic case based on a fixed number of jobs makes no sense.

The fixed pie fallacy fails not only in the context of the labor market. As alluded to above, immigrants raise demand for goods and services, including those that natives are uniquely positioned to provide, such as education, legal services, entertainment, and management. If the pie were truly fixed, a virus that randomly killed half of Americans would make everyone twice as rich.

Housing is an area where the pie is close to being fixed in many regions, but this is only the result of zoning laws that restrict supply. There are an endless number of reasons why we should make housing easier to build, and the payoffs to doing so are potentially massive. But to restrict immigration on that basis alone would mean responding to one broken policy by hurting the economy in another.

What all this means is that not only are H-1B workers highly-paid, with median earnings at around the 90th percentile of wages, but each H-1B worker is estimated to create 1.83 jobs for native-born Americans. This is not surprising at all. As high earners, H1-B recipients must be a net benefit to the economy. In no other context would anyone argue that workers in the top decile of productivity are making Americans poorer.

Callum Williams, senior economics editor at The Economist, has argued that having a much larger common market goes a long way towards explaining why Americans are so much wealthier than Europeans. In a tiny village where individuals only trade with members of their own community, compared to a large developed country, it is less likely there will be enough demand to make being the owner of a boutique bookstore, a Substacker, or an exterminator a realistic career path. This is why, for example, American restaurants will commonly have different jobs for hostesses and waiters, while in Europe one employee will take on both roles. More people means greater economies of scale. When newcomers are well above average in productivity, as H1-B recipients are, opening the borders to them becomes a no-brainer.

What if you only wanted to focus on wages for tech workers? One might acknowledge that highly productive workers making businesses more efficient are good for the economy as a whole, but can we at least say that a certain class of worker they compete with is worse off?

That is far from certain. Noah Smith has an excellent recent summary of papers arguing that H-1B visas don’t even hurt workers that are in direct competition with new arrivals. We have a strong basis for believing this because H-1B visas are distributed based on a lottery system, so we can treat the process as a random experiment and see what happens to the wages of individuals working for firms that win the lottery versus those that don’t. Workers in the firms that get immigrant labor end up hiring more tech workers and paying them more, because they become more efficient and sometimes scale up.

Of course, immigration policy shouldn’t be determined on the basis of what is good for workers in one industry, but what is good for the nation as a whole. So even if you do not find such studies convincing, you should want more productive workers to come into the country because a welcoming policy still makes society better off. This is not controversial among economists, as the question gets to the basic assumptions of their field. A 2016 poll of economists showed exactly 0% disagreeing with the idea that letting employers hire more immigrants with STEM degrees would raise per person income in the US in the long run.

To be fair, an overwhelming majority also said that such a policy would at least temporarily lower the premium earned by workers with similar degrees. However, it must be noted that this survey was conducted before most of the studies cited by Noah Smith in his recent article. It would be interesting to see what economists who are familiar with the literature that has come out over the last few years think. But, again, even if it is true that immigrants directly suppress the wages of workers they compete with, public policy being designed to help subsets of society rather than the entire country is the path to stagnation. If we try to “protect” the workers in every industry from competition, we will all end up poorer.

Unfortunately, zero-sum thinking tends to be common when individuals consider economic issues. People often misjudge market interactions as win-lose rather than win-win, especially when foreigners are part of the equation. Irresponsible politicians and pundits take advantage of such biases, demagoguing bad policies into existence, to the detriment of society.

Yet letting in more immigrants is not a form of charity on the behalf of Americans. It’s an arrangement that expands the pie, benefiting both newcomers and natives. America is blessed with being the top talent magnet in the world. It would be a shame to waste one of its greatest assets.