Jobs Americans Would Do

The labor shortage afflicting American employers nationwide is nothing new in the San Joaquin Basin of California’s Central Valley, which produces 40% of the nation’s fruits and nuts. For Alex, who asked that we use only his first name, the trouble started eight years ago, with an epiphany that came while driving down the I-5 from his office in San Francisco to a conference in Los Angeles. “Our team of programmers was expensive, and demanding,” he explained, “our lease costs were spiking. As I looked out my window at the toiling farmhands, I thought maybe I saw an answer.”

Alex relocated his app development firm to a small patch of farmland outside Fresno, setting up crates on which workers could rest their laptops in the dusty field, and lowered pay to $16 per hour. Not a single member of the team joined him. Ever since, getting apps developed has proved nigh impossible. “Occasionally someone will apply,” he explains, “but they never know Java. It’s what I call the ‘skills gap.’”

Alex is not alone. Short-staffed coffee shops and machine tools without operators have prompted economists and business groups to raise the alarm. An “imbalance in labor supply and demand” is dragging down the nation’s economic prospects and fueling inflation, according to the Wall Street Journal. “We hear from businesses every day that the worker shortage is their top challenge,” explains Neil Bradley, chief policy officer at the U.S. Chamber of Commerce. The labor market is “overheated” and “much too tight,” says Michael Strain, director of economic studies at the pro-business American Enterprise Institute. In his view, we need “a pretty big increase in the unemployment rate.”

What exactly is an “overheated” labor market or a “worker shortage,” how would we know if one existed, and what would be an appropriate response? Colloquially, a shortage refers simply to having less than a desired amount of something. But in a market mediated by price signals, that concept falls flat and an “imbalance in supply and demand” is incoherent. Supply and demand are not levels; they are relationships that describe the various quantities of a good supplied and demanded across a range of prices. These quantities are, by definition, imbalanced at all prices except the equilibrium point, and a key function of markets is to find that point. Most people experience a private-jet shortage, insofar as they would like one and have none, but this “imbalance in supply and demand” means only that they are not willing and able to pay the market-clearing price.

As with the masses glumly flying commercial, Alex’s claimed programmer shortage seems easy to dismiss. The problem comes in distinguishing his complaint from that of his neighbor, Bernie, who operates a more conventional Central Valley farm. Like Alex, Bernie also wishes to pay $16 per hour to workers who would labor all day in a hot field, and he too reports that hiring is a major challenge. Unlike Alex, Bernie finds his complaint taken quite seriously by economists and policymakers. The government has expanded the H-2A temporary visa program from fewer than 50,000 agricultural guest workers in 2005 to more than 250,000 in 2021, in an effort to fill, in Bernie’s words, “jobs Americans won’t do.”

If Alex’s claim is outlandish, why isn’t Bernie’s? Markets do experience shortages in response to shocks, before time has allowed for adjustment, but in such situations prices go haywire. The most salient fact about the current labor market is that real wages are slowly falling: down 0.7% in March, compared to a year ago, when wages were down 2.4% compared to the year before that. Median weekly earnings were no higher at the end of 2022 than at the end of 2019. The Chamber’s Bradley made his comment in May 2022, in the midst of a two-month period when real wages were falling at an annualized rate of nearly 8%. Employers claiming to be in dire straits are reducing the real compensation they offer. In aggregate, the prices they charge are rising at the level of inflation, and yet they are not willing even to raise the wages they offer at that same rate. They are all Alex.

The outlandish, panicked response of economists and business leaders to a labor market in which workers are not readily available at the price employers wish to pay illuminates the problem of wage stagnation that has now bedeviled America’s economy for 50 years. From 1972 to 2022, real corporate profits per capita rose 185%. GDP per capita rose 141%. Productivity rose 135%. The average hourly wage for production and nonsupervisory workers rose 1%. How is that even possible?

It is possible because employers will tend to raise wages under one, and only one, condition: when they cannot hire the workers they need at the existing wage. All of labor economics turns on that simple fact. If you are happy with the workers you have at $16 per hour, you will not offer $17 per hour. If open positions at $16 per hour go unfilled, or workers earning that wage start to leave or threaten to strike, $17 per hour will be on the table. The average wage went from $28 in 1972 (in 2022 dollars) to… $28 in 2022, because employers did not have to offer $29. They could offer $29—after all, the corporate sector had roughly tripled its profit even after accounting for population growth. But they did not have to, and so they did not.

Rather than follow recommendations from pro-business groups to raise the unemployment rate, policymakers should row hard in the opposite direction, increase the pressure, and make clear that no relief from a tight labor market will come. What tools are at their disposal to make employers offer real wage increases to lower wage workers now and continue doing so for, say, the next 50 years? We have a lot of catching up to do. A wide range of policy options hold promise, but any discussion should start with the straightforward one: limiting the supply of workers available. Of course, this may not be as easy as it sounds. The goal is to ensure good jobs for American workers, so excluding some such workers for the sake of boosting the wages of others would be a pyrrhic victory. But there is another set of workers, not Americans, whose access to the labor market American policymakers could limit.

Immigration Policy as Active Labor Market Policy

A few statistics are helpful in illustrating the magnitude of the effect that immigration has on the American labor market. Immigrants accounted for more than half of the increase in the labor force over the past 15 years—a total of 7.0 million additional workers. By comparison, the total shortfall in workers caused by a depressed labor-force participation rate at the end of 2021 was 2.9 million. Immigrants account for 19% of workers overall, but 32% of those in occupations with median earnings below $30,000 per year (compared with 16% in occupations paying more than $60,000 per year). Such government figures also presumably underreport the total impact of illegal immigration in the labor market’s lower-wage segments. Immigration has provided the margin between a labor market in which employers would feel constant pressure to find and retain workers—especially lower-wage ones—and the labor market as it has operated, in which they can offer the same low wages and poor conditions for decades on end.

Both libertarians and progressives, with their different ideological commitments to open borders, offer a peculiar counterargument to the common-sense case for restricting immigration to boost worker power and wages. Workers are also consumers, they note, and thus immigrants increase both the supply of labor and the demand for it. Make any effort to reduce the flow of immigrants, and employers will need fewer workers to serve them as customers.

This argument fails as theory, because immigrants adding their labor to some segment of the labor market prompts a concentrated increase in supply within that segment but a diffuse increase in demand across all segments. Guest workers picking lettuce do not spend their wages primarily on lettuce. The evidence for the argument’s failure in practice is that, quite obviously, no one believes it. The business groups and economists who have been vocally and pretextually advocating immigration to relieve labor shortages or reduce inflation rather give the game away. If an increase in immigration expanded demand by as much as supply, the “shortage” would remain acute. If it didn’t suppress wages, inflation would not abate.

In (limited) defense of progressives, they are less inclined to make the pretextual economic argument for their preferred policy than simply to remain silent. A classic example is the Roosevelt Institute’s recent whitepaper on “A Whole-of-Government Approach to Increasing Worker Power,” which mentions immigration only once, in a footnote explaining that the issue will go unaddressed due to “space constraints.” So much for whole-of-government. One might make a range of cases, on humanitarian or other grounds, for admitting large numbers of immigrant workers to the labor market’s low-wage segments. But it is hard to take seriously anyone who professes a commitment to enhancing worker power and dignity while refusing even to discuss the most direct tool for enhancing them. There is no complicated trilemma here, only a simple dilemma. As I have written, Worker Power, Loose Borders: Pick One.

The kernel of truth in the case that supply will not affect wages is that immigration generally, with skills composition similar to that of the existing population, need not distort labor-market conditions. When immigrants enter all segments of the labor market simultaneously, their cumulative effect on broadly distributed demand will offset their various smaller and more concentrated effects on supply. Another way to see this is to recognize that, economically speaking, immigration is a form of population growth. The American labor force’s expansion as the next generation enters it has little implication for worker power because the economic characteristics of that cohort will generally mirror the characteristics of the market’s existing participants. The same would hold true for a cohort of immigrants whose skills mirrored those of the population it was entering.

For policymakers, then, the skills composition rather than level of immigration is the main economic issue. The problem with American immigration policy has not been that it generously welcomes newcomers to our nation but that, as the statistics show, it has concentrated the increases in labor supply at the market’s low-wage end. Conversely, immigration could increase worker power for lower-wage workers if that immigration were predominantly into other (i.e., higher-wage) segments. Thus, the need for a skills-based immigration system. Maintaining the current immigration level, but skewing its composition toward workers who will compete in the labor market’s high-wage segments, will tend to strengthen worker power in the market’s low-wage segments even more quickly than would a policy of restricting immigration broadly. It will increase demand for what is today low-wage labor, create strong incentives to invest in improving the quality of those jobs, and have distributional effects that shift income back toward the lower and middle classes.

Corporations and their spokespeople understandably dislike any constraint on their supply of labor, which would put them in the position of having to raise wages. So it is important to remember that capitalism does not succeed when profits are highest, but rather when those activities that will yield the highest profits are also those most in the public interest. As Adam Smith warned in The Wealth of Nations:

[T]he rate of profit does not, like … wages, rise with the prosperity, and fall with the declension, of the society. On the contrary, it is naturally low in rich, and high in poor countries, and it is always highest in the countries which are going fastest to ruin. The interest of [employers], therefore, has not the same connexion with the general interest of the society.

Capitalism works when capital in pursuit of profit must find ways to expand output with the labor present, and when it must share the rising proceeds of that joint project. These conditions are more likely to hold, and will hold more strongly, when workers act as the limiting reagent in the production process. Business leaders will always chafe at this constraint and insist they need more labor, emphasizing the additional production they could unleash. The pleas must fall on deaf ears.

Flooding the low end of the domestic labor market with foreign workers may boost production but it erases the incentives to produce in ways beneficial to the lower-wage workers already here, or to share the proceeds of rising output with them. Total output may increase, but output per person will not. The nation might simultaneously become “wealthier” in aggregate and place itself on a lower trajectory for economic progress—for instance, with 50 years of declining construction-sector productivity. Owners of capital and managers of firms might see their profits and wages rise, even as the typical worker’s wage stagnates or falls. In many respects, the impact of reliance on foreign labor entering the United States mirrors the impact of offshoring American production to foreign labor abroad. Capitalism will not deliver on its promise so long as American capital can choose workers in those markets over the workers in its own.

By contrast, when employing the nation’s workers is a competitive imperative, and increasing productivity is the key to growth, that is what firms will do. This principle is so obvious to capitalism’s proponents in other contexts, as they celebrate the power of competition in free markets to solve any problem, overcome any scarcity, innovate around any obstacle. Yet when labor is the problem to solve, scarcity to overcome, or obstacle inviting innovation, all is suddenly lost. If complaints about “jobs Americans won’t do” elicited only laughter, and creating jobs that Americans would do were a non-negotiable prerequisite to generating profit, imagine what capitalism’s awesome power might achieve.

Rather than set their hair ablaze and run circles around Washington screaming for help, how might firms respond when their best laid plans are dashed by a labor market that does not offer the desired workers with the desired skills at the desired price? They might make investments in automation or training, allowing them to meet their needs with the workers they have and allowing them to offer better wages, both to those workers and others they might need to hire. They might alter working conditions or reengineer processes to make the jobs they offer a better fit for the workers they can find. They might reduce pay for higher-wage workers, at least in real terms, by no longer raising their wages to keep ahead of inflation. They might revisit their business models, changing the goods and services they provide or the customers they serve, thereby reallocating capital to uses better aligned with labor. 

Prices would adjust, consumption would adjust, low wages would rise. Markets can work for workers if policymakers don’t step in with the “solution” of bypassing the existing labor force with new and more pliable options. 

An Unproductive Conversation

Pose such scenarios to Alex and Bernie, other CEOs, their lobbyists, or the “free market” economists they sponsor, and the objections are a muddled mix of superficial references to competition, productivity, and consumer welfare that belie a willful ignorance of what markets are for and how capitalism generates prosperity.

The superficially coherent argument underpinning the labor-shortage lament begins with a simplistic claim about productivity. A core tenet of the market fundamentalism that has shaped so much economic thinking in recent decades is the idea that all workers are paid a wage reflective of their individual output (the “marginal revenue product,” to be precise). If this productivity does not rise, then wages will not rise—and if productivity appears to be rising while the median wage stagnates, this must mean that all the productivity gains are being achieved by higher-wage workers, justifying their more rapidly rising wages. If workers do not accept jobs at the wages and in the conditions that their productivity will bear, then a “shortage” exists. From this perspective, the question of how to find more workers willing to accept lower pay is an entirely reasonable one.

The fundamental problem with this view is its conflation of two different concepts of productivity. In the hypothetical market of Economics 101, where individuals produce and exchange various goods, measuring each person’s productivity is easy enough and showing that his income will reflect that productivity is trivial. The problem emerges when the yeoman farmer takes a job installing windshields at a multinational car company. The company’s labor productivity is still objectively measurable: value of cars coming off the assembly line less value of inputs purchased divided by employees. But how can the marginal productivity of various workers be disaggregated, especially seeing as many are not involved directly in any production process? Consider the windshield installer and the industrial engineer. If the engineer is developing new processes that allow the installer to secure windshields more quickly, whose productivity is rising? Who should be paid more? And what wage should the HR executive claim, for overseeing the benefits program used by them all? If one of their cars is a hit and moves off the lot with fewer discounts and thus higher prices, their marginal revenue product will be higher. But which of them has become more productive?

Such questions may seem tendentious, but they are inevitable in a model that pretends an objective measure of productivity validates the wage paid to each worker. The questions vanish painlessly with acknowledgment that, while a firm’s productivity determines how much income it has to allocate amongst its workers and as profit, the allocation therein is a function of supply and demand in relevant labor markets and the power exercised by the various claimants. Windshield installers are paid what they are because that is what must be offered to employ the required number. The same is true of marketing executives. The total cost of producing the car is a function (in part) of how much all these workers must be paid, and the costs that manufacturers incur to produce their vehicles in turn affect vehicle prices, and thus manufacturer revenue, and thus the measure of a worker’s value. An assessment of how many cars consumers will buy at a given price in turn informs how many cars to produce and thus how many windshield installers to hire. A central function of markets is to find equilibrium amongst these many variables.

The questions become even more interesting across industries. How do we compare Alex’s programmers and Bernie’s lettuce pickers? The price of a head of lettuce or a downloadable app determines the productivity of the firm growing or publishing it. But the relationship between those prices is itself a function of the relationship between the wages paid to workers in those industries, which depends upon supply and demand in labor markets. Which firm is more productive, the publisher of Candy Crush, or the grower of half the lettuce consumed on the Eastern Seaboard? Who should be paid more, a Candy Crush programmer or a lettuce picker? This depends upon how much lettuce to exchange for a download, which might change dramatically if food became scarce—or if farmworkers did.

Farmer Bernie seems to need the government’s help addressing his labor shortage because his competitors are all relying on the lowest-wage labor too, yielding low prices in the store, thus creating an economic rationale for the jobs to be low-paying ones. The circular logic of productivity measurement makes the status quo view of jobs self-fulfilling. But if none of the growers had access to low-wage labor, paying higher wages would suddenly appear viable for them all. Picking lettuce for an hour might then appear more productive than spending an hour debugging a new loot box, and worthy of a higher wage. The relevant factors are the number of people available to do each type of work, the wages they demand, and the prices that customers will pay. Isn’t this what markets are for?

Markets for Me, But Not for Thee

Bernie, and his lobbyists at the U.S. Chamber, have only two answers available: one commonly deployed but backward in its economic reasoning, the other clearly implied but for good reason rarely said aloud. The first answer is all about consumption. If low-wage workers gain substantial raises, lettuce will get more expensive; so will eating out, completing a home renovation, and so on. Lower-income households and those on fixed incomes will suffer. Consumer welfare will decline, and that is the yardstick by which economic progress is measured.

As a preliminary matter, the scale of the prospective price increase requires consideration. For instance, down on the farm, labor costs are typically less than 20% or for specialty crops close to 40% of total operating costs, and the price from the farm is about one-third the price on the shelf. People colloquially worry about strawberries costing $50 per quart, but the reality is that even if wages quadrupled, farms made no investments to increase output per picker, and the full cost were passed on to consumers, prices might rise 40%. A $3.00 quart of strawberries might cost $4.50, a $1.70 head of lettuce might cost $2.50. Nor are farm products a significant share of the American family’s budget. The U.S. Department of Agriculture estimates that 7.4% of American food spending goes to farm production, while the Bureau of Labor Statistics estimates that the typical household spends roughly $1,000 on fruits and vegetables consumed at home, suggesting only about $100 attributable to farm labor. Quadrupling those wages might cost the typical family $300 in a year.

The agricultural example is illustrative, but not the best lens for considering the economy-wide implications of rapidly rising wages at the labor market’s low end. A $300 increase in food costs could be quite painful for many families, while only a very narrow category of farmworkers would receive the very substantial benefit. The broader lesson is that low-wage labor is only a small share of the total cost embedded in most products. Especially with an adjustment period, no systemic shocks occur, which is good news for everyone.

The news for the low-wage workers themselves is even better. Because low-wage labor is only one part of what low-income households consume, a broad-based increase in low-end wages raises the earnings for low-income households faster than it can raise their costs. When low wages increase but other costs do not, the total cost increase and thus the final price increase is relatively lower. Wage increases may fuel some inflation, but those receiving the wage increases see real gains. Indeed, a peculiar feature of the present alarm about wages driving inflation is that this is not happening. With real wages falling, they represent a drag on the overall rate at which prices are rising—one wonders why free-market economists do not see fit to focus more attention on, say, the decision of firms to raise prices faster than they apparently need to. Imagine the screaming to come, if and when low-wage workers find themselves with sufficient power to achieve real gains at the expense of higher-wage workers and profits.

If higher wages prompted no investments to increase output per worker (a poor assumption, but a worst case worth considering), real gains for lower-income workers would mean real losses for higher-income ones, who would experience the price increases without the higher wages—in real terms, their own wages would fall. That shift would be an entirely natural one in a well-functioning market. The engineer is as worthless without the windshield installer as the windshield installer is without the engineer, perhaps more so. When forces of supply and demand led to higher-wage workers earning a rising share of firm output, and capital earning a handsome return, while the median worker shared little of the gain, this was “the market working.” Economists developed elaborate theories of “skill-biased technological change” to justify the result. If the market now begins to work the other direction, as HR executives find themselves rather dependent on workers to make products and provide services, surely that deserve at least as much celebration. And how wonderful an illustration of the market’s power if, as prices adjust, HR executives find themselves heading down to jobs on the shop floor.

The celebration is not forthcoming. For all its talk about productivity, the business community’s enthusiasm for price signals and efficient market outcomes turns out to be rather selective. The implication of Bernie’s thinking is a second answer: that some jobs are better than others and should pay more. The programmer could pick lettuce if he wanted, but the farmworker cannot code. Past market outcomes were so readily celebrated because they reinforced this worldview. The phrase “jobs Americans won’t do” is so potent because, in setting out certain classes of work as beneath the dignity of the people issuing the judgment, it makes clear exactly what they really think. The irony is that, to maintain his faith that wages fairly reflect productivity, and advance his preference for a supply of low-wage labor that eases his own life and elevates his own status, the “free market” enthusiast must embrace the role of central planner overriding price signals with his own judgment of what prices should be.

Substantively, though, classifying which jobs Americans will do makes no sense. It wrongly assumes that jobs have fixed characteristics—picking crops or cleaning rooms or shingling roofs is a job Americans won’t do because it pays poorly, makes difficult demands, and must be performed under harsh conditions; programming apps or suing for patent infringement is a job Americans will do because it pays well and comes with great perks. But the only fixed element of the job is the function performed—the picking or cleaning or building; the programming or document review—and even that can be mediated in many ways by technology. Described in the abstract, harvesting fruit outdoors or repairing a house is not an inherently inferior task to sitting in a cubicle (or, perish the thought, an open floorplan) staring at a rectangle of glowing pixels while moving fingers up and down on a piece of plastic. Described as a vocation, feeding the nation or providing it with shelter is rather more edifying than targeting it with underwear ads.

Bernie could solve his problem quickly if he and his competitors all had to offer a starting wage of $30 per hour, an engaging experience, flexible scheduling, career paths to management, and 30 minutes in a catered, air-conditioned tent after each 30 minutes in the field. Trucking would be more popular if hauls were shorter, weeks away were followed by weeks off, and pay were more like that of investment banking. Conversely, no one will work for Alex.

For 280 of the first 282 months in the Federal Reserve Bank of Atlanta’s “Wage Growth Tracker” dataset, workers in “high-skill” occupations saw their nominal wages grow as fast or faster than workers in “low-skill” ones. In June 2021 the picture flipped and, for the nearly two years since, every month saw higher (nominal) growth for low-skill wages. The near-perfect overlap of that period with the one in which experts have proclaimed the labor market in crisis is, perhaps, not a coincidence.

The choice to treat these types of jobs and workers so differently, and the power that choice holds over public policy, is cultural. Certainly, it has no moral basis. It has no economic basis, either, beyond the circular logic now breaking down at the first sign of labor’s insufficient pliability to capital’s demands. Good riddance.

High-Wage Immigration Policy

A major benefit of immigration policy is that it acts directly on the relevant variable: the supply of workers. Other tools, from education and training to trade and industrial policy to tax reform and public spending all have the potential to improve labor market conditions as well, but they require a chain reaction to reach the wages offered to workers, and greater demand for workers accomplishes little if swamped by constant increases in supply. One way to understand the labor market’s outcomes in recent decades is that these policies did work, demand did rise significantly, productivity did rise significantly… and the supply of workers was allowed to rise at the rate needed to counteract it all. As a result, the benefits accrued primarily to a narrow set of high-wage workers and to businesses whose profits surged. That certainly seems to be the story that the data tell.

If policymakers want to slow the growth in labor supply through changes to immigration policy, they can do so. The catastrophic failure to enforce immigration law in recent years has led to skepticism that enforcement is possible—indeed, efforts have been so inept that one might reasonably conclude the promotion of such skepticism was the goal. Piled atop these failures comes the rhetoric that setting and enforcing an immigration policy for the benefit of Americans—of all races and ethnicities, including the millions of legal immigrants already in the country and the millions more that we will welcome—is somehow immoral or inconsistent with the nation’s values. But setting the terms of entry into the community is the right and obligation of every nation, and applying the law firmly and fairly is the nonnegotiable prerequisite to a functioning democratic republic. What we have lacked is the will. What is unbefitting our nation and inconsistent with our values is the lawlessness that we have tolerated.

Enforcing the Law

In isolation, support might be quite broad for a skills-based system, which welcomes large numbers of immigrants and skews the group’s composition heavily toward highly skilled newcomers who are likely to compete in the labor market’s high-wage segments. But implementation requires a robust system of enforcement that limits labor supply to legally authorized workers. On a prospective basis, the nation must be able to control border crossings, ensure that temporary residents depart when their visas expire, and hold employers accountable for hiring only legal workers. And any effort to prevent future illegal entry into the labor market must grapple with the more than 10 million illegal immigrants living in the country and often participating in the labor market already.

The best chance of making progress is to recognize that putting future immigration policy on sound footing is more important for the economy than dealing with past illegal immigration in one way or another. A compromise on the past, where there are no good answers, coupled with a strong policy for the future, where the right answer is clear, is a far superior outcome than the current impasse. The one caveat for any bargain is that the promise of future enforcement must be credible. If the world believes, based on how the United States deals with its existing population of illegal immigrants, that future illegal entry will be rewarded as well, the nation will never have control of its labor market. Policy reform is impossible without secure borders and effective laws.

Mandatory E-Verify should be the cornerstone of labor market enforcement. U.S. policymakers already have a system, called “E-Verify,” that validates the legal status of American workers. What’s lacking is the political will to ensure its consistent use. Employers should have no choice, they should face penalties for employing illegal workers even inadvertently, and penalties for intentional and repeated offenses should be catastrophic and include criminal prosecution. The law should recognize that the employer who opts for illegal and exploitable labor and undermines the power of American workers is committing a far more serious and less excusable offense against the community than the illegal immigrant pursuing a better life. Aggressive deterrence of employer malfeasance constrains the labor supply directly and, by eliminating the job opportunities and thus much of the incentive for illegal immigration, offers an efficient means of addressing that challenge as well.

For immigrants already in the country illegally, policymakers need a solution that will neither send sudden shocks through the labor market and millions of lives nor signal lackluster commitment to enforcing the law. They should pursue an approach that has received little consideration from either side of the immigration debate: last in, first out (LIFO). Under LIFO, the more recently an illegal immigrant has arrived in the country, the sooner he must leave. Those who have only just arrived, migrate frequently across the border, or have a criminal record would be barred from the labor market immediately. Those who have already resided in the country for up to five years would be issued a work permit that lasts for the same number of future years as the number of past years for which residence can be established. And those the United States has allowed to settle in the country over five years or more would be granted a five-year work permit that could lead to permanent legal status after payment of a substantial fine, on condition that the more transient groups have left as scheduled and the nation has demonstrated firm control over its borders and its labor market.

Eliminating Guest-Worker Programs

Perhaps most obviously—and easily, with robust enforcement in place—temporary visa programs for low-wage work should be immediately capped at current levels and then phased down to zero over a decade. Both the H-2A and H-2B programs, for temporary agricultural and other seasonal work, respectively, provide employers with an alternative to offering jobs to Americans at good wages in decent conditions. The only economic rationale for doing so is to artificially suppress wages and prices below levels where the market would otherwise settle, benefiting consumers and employers at the expense of American workers, and disproportionately benefiting higher-income households that enjoy the lower prices without seeing their own labor undercut.

The H-1B program, designed for high-skilled workers unavailable in the domestic market, would be a more difficult case if it operated as designed. At least hypothetically, employers could seek skillsets that no American workers possess—jobs Americans can’t, not won’t, do—and expanding the supply of such specialized and high-wage labor would redound to the benefit of lower-income households whose workers would not be in competition. In practice, the program is a boondoggle. Roughly 40% of the nearly 100,000 H-1B visas issued each year are for employment at 30 companies, more than half of which are top outsourcing providers like Infosys, Tata Consultancy Services, and Cognizant, which use the visas to replace incumbent American workers with temporary foreign workers paid less than market wages, unable to change jobs, and subject to abuse. Roughly 85% of visas are awarded to entry-level and junior workers.

In the short run, rather than assign H-1B visas via lottery, the federal government should award them based on wage level, with all visas going to those positions for which employers are offering the highest wages. Ironically, those highest paying and most attractive seeming jobs, not the typical programming roles, and certainly not the farm work, are truly the jobs Americans won’t do; the ones where employers might credibly claim an inability to fill the role and policymakers might serve the common good by providing assistance. But one benefit of shifting permanent immigration toward highly skilled workers is that H-1B visas can be phased down like other temporary programs. If just an additional 10% of the roughly 1,000,000 green cards issued each year went to the most highly skilled workers, the influx would exceed the total number of H-1B visas issued annually.

Policymaking with Purpose

A temporary visa program may still make sense for the United States, if it operates explicitly as an element of industrial policy aimed at building capabilities and capacity in strategically important sectors. The CHIPS and Science Act’s subsidies for domestic semiconductor producers provide a case in point. The issue has arisen that firms hoping to participate are struggling to find the engineers and construction specialists necessary to develop a multi-billion-dollar fabrication plant.

At first glance, this appears to be another lame complaint of “labor shortage,” to which the answer should be that the real failure is in the business plan. But note what’s different: policymakers have chosen a specific need that the market was not meeting, chip capacity, and they are seeking to override price signals and private capital allocations. The unavailability of labor to support the objective should be neither surprising nor discouraging. The projects are ones that private actors did not consider economical and were not pursuing. That’s what the industrial policy is for.

Unlike at Facebook, whose prowess in targeting ads and immiserating teens does little to advance national imperatives or the common good, using temporary workers makes sense to support the rebuilding of domestic semiconductor capacity. That project’s immediate purpose is not a jobs program—though in the long run, the industry and the ecosystem supporting it could provide vast numbers of good jobs far from already-booming regions. Clearly identifying other priorities and shaping policy to meet them is necessary and proper. Further, a well-designed policy should not only pair such temporary visas to specific initiatives, but also establish predictable time limits, creating the incentive for employers to invest in building domestic labor capacity for the future.

The question of purpose should always be central for economic policy. Policymakers and business leaders who have viewed the economy’s purpose as providing cheap stuff and investor returns have understandably favored an immigration policy that kept wages low. But rising wages, especially for lower-wage workers, are not a threat to broad-based prosperity, they are its source. The “supply-side” theory that prioritizing the conditions for capital’s success will eventually redound to the benefit of all is a bastardization of capitalism rightly understood and, as an empirical matter, a failure. Good and dignified jobs that allow workers to support their families and communities are not a hoped-for byproduct of a healthy economy, they are its purpose. Gains in consumption and material living standards are desirable as well, but cheaper prices through lower wages is a losing proposition for affected families and the nation as a whole.

The American economy has not failed in recent decades with respect to the material standard of living. The failure has been the creation of insecure jobs that do not meet workers’ needs, a shift in the economy-wide distribution of income that has left working families struggling, and a decay in social solidarity as the economy’s winners declare themselves most valuable and the losers expendable. Only the power for workers that comes from being needed will reverse those trends.