Cheap labor leading to a technological slowdown

Some people have argued that a very large supply of cheap labor would lead to a slowdown in the development of labor-saving technology, like farm mechanization and bank ATMs, that ultimately benefit everybody. This argument has some surface merits, but suffers from some drawbacks.

Examples of the argument being made

Summary of counter-arguments

  • It’s important to distinguish between labor-saving “innovation” that merely substitutes capital for labor (essentially, moving along the production possibilities frontier) and innovation that expands the production possibilities frontier. While some of the innovation driven by labor scarcity is innovation that expands the production possibilities frontier, a lot of it is adjustment that substitutes capital for labor. More is discussed in Nathan Smith’s blog post and Vipul Naik’s comment on the post.
  • In Chapter 1 (PDF) of his book Let Their People Come, Lant Pritchett agrees with the factual diagnosis, but argues that there is something perverse about focusing on labor-saving innovation in a world where the main thing people have to sell is their labor. He argues that such innovation is a second-best response to a bad policy, not a first-best alternative. Pritchett also draws on the analogy between immigration and trade. A lengthy quote from his book is at the bottom of the page.
  • This does not mean that one must choose between open borders and labor-saving innovation. Rather, Pritchett is positing that with free labor mobility, many of the incentives for high-skilled workers to devote their energies to low-skilled labor-saving innovation would reduce. This still leaves plenty of room for them to focus their energies on other forms of innovation, including high-skilled labor-saving innovation. Instead of focusing their energies on mechanization of farm labor, engineers may have chosen to concentrate more on mechanization of high-skilled surgeries. In the long run, there would be innovation to save labor in all forms, but the short run priorities would be massively different, and more welfare-enhancing, in a world with free labor mobility.
  • Free labor mobility itself enhances innovation. For more, see the innovation case for open borders.
  • Scarcity of labor alone does not lead to labor-saving innovation. Rather, labor needs to be relatively more scarce than capital so that cheap capital can then be substituted in for expensive labor. If the capital is lacking, many of the big labor saving technologies we know about would not have been invented. Given that free labor mobility is expected to double world GDP, the capital stock will rise in the long run and enable more labor-saving innovation. Moreover, the labor-saving innovation will occur at the right time — when poverty is dead and people are genuinely wealthy and skilled enough to desire such innovations — rather than occurring prematurely in a world where unskilled labor is the main good many people have to offer.

A comprehensive collection of counter-arguments is available in Nathan Smith’s blog post titled innovation and open borders.

Quotations from Krikorian’s book where he raises the argument

In his book The New Case Against Immigration: Both Legal and Illegal, Mark Krikorian makes the argument:

The preceding discussion focused on the impact of mass immigration on workers—how it reduces their wages and crowds them out of the job market, and causes increased inequality. But is an artificially bloated low-wage labor market at least good for business? Spokesmen for the U.S. Chamber of Commerce, the National Restaurant Association, the California Farm Bureau, and others would have you believe the answer is yes, and in the short term, that’s likely to be the case. Employer organizations spend enormous resources lobbying the government to import a “reserve army of labor,” to use Marx’s phrase, so that they can hold down their labor costs and avoid unionization. But in the longer run, the superabundance of cheap labor harms the future competitiveness of industries where the immigrants are most heavily concentrated. The entire history of economic development— starting with the first ape-man to pick up a stick—is a story of increasing the productivity of labor, so each worker is able to create more and more output. But capital will be substituted for labor only when the price of labor rises, something the federal government’s mass-immigration program is specifically intended to prevent. A 2001 report by the Federal Reserve Bank of Boston highlights this problem by warning that a new wave of low-skilled immigrants over the course of this century may slow growth in U.S. productivity.44 By artificially holding down the natural process of wage growth in labor-intensive industries, mass immigration thus serves as a kind of subsidy for low-wage, low-productivity ways of doing business, retarding technological progress and productivity growth. In effect, mass illegal immigration is an unintentional, but very real, Luddite force in our economy. Germany experienced the same thing when it imported large numbers of Turkish and Yugoslav workers in the 1950s and 1960s; as two of the foremost scholars of immigration have written, “Economists began reporting that the program was slowing investments in automation and mechanization, so that ‘Japan [was] getting robots while Germany [got] Turks.’ ”45 That this is so should not be a surprise. Julian Simon, in his 1981 classic, The Ultimate Resource, wrote about how scarcity leads to innovation:   It is all-important to recognize that discoveries of improved methods and of substitute products are not just luck. They happen in response to scarcity—a rise in cost. Even after a discovery is made, there is a good chance that it will not be put into operation until there is need for it due to rising cost. This point is important: Scarcity and technological advance are not two unrelated competitors in a Malthusian race; rather, each influences the other.46   This is true for copper or oil, and just as true for labor—as wages have risen over the generations, innovators have devised new ways of substituting capital for labor, increasing productivity to the benefit of all. The reverse is also true; the artificial superabundance of a resource will tend to remove much of the incentive for innovation. Stagnating innovation caused by mass immigration is perhaps most apparent in the most immigrant-dependent activity—the harvest of fresh fruit and vegetables. Academic researchers have warned about the long-term viability of the industry: “New technologies and mechanization appear to offer the only solution to significantly reduce production costs and maintain competitiveness,” write three prominent agricultural economists.47 But immigration is an obstacle to adoption of such technologies, according to Orachos Napasintuwong, an economist at the University of Florida: Because of “the augmentation of labor supply through unauthorized foreign workers . . . the incentive for new labor-saving technologies is reduced from what it would be in the absence of international labor mobility.”48 It wasn’t always this way. The period from 1960 to 1975 was a time of considerable agricultural mechanization, precisely because it was a period of relative scarcity of agricultural labor, roughly from the end of the Bracero program, which imported Mexican farmworkers, to the beginning of the mass illegal immigration we are still experiencing today. 49 During hearings on the proposed termination of the Bracero program in the early 1960s, California farmers claimed that “the use of braceros is absolutely essential to the survival of the tomato industry.” But Congress ended the program anyway, causing harvest mechanization to accelerate; as a result, the production of tomatoes grown for processing (juice, sauce, and so forth) actually quintupled, demand for harvest labor dropped 89 percent, and the real price of tomato products fell.50 But a continuing increase in the acreage and number of crops harvested mechanically did not materialize as expected, in large part because the supply of workers grew artificially large due to the growing illegal immigration wave that the federal government was unwilling to stop. An example of a productivity improvement that “will not be put into operation until there is need for it due to rising cost,” in Simon’s words, is in raisin grapes.51 The production of raisins in California’s Central Valley is one of the most labor-intensive activities in North America. Conventional methods require bunches of grapes to be cut by hand, manually placed on trays for drying, manually turned, and manually collected. But starting in the 1950s in Australia (where the climate was suitable but there was no large supply of foreign farm labor), farmers were compelled by circumstances to develop a labor-saving method called dried-on-the-vine (DOV) production. This involves growing the grapevines on trellises, then, when the grapes are ready, cutting the base of the vine instead of cutting each bunch of grapes individually. This new method radically reduces labor demand at harvest time and increases yield per acre by up to 200 percent. But this high-productivity, innovative method of production has spread very slowly in the United States because the mass availability of foreign workers has served as a disincentive to farmers to make the necessary capital investment. In fact, fully half a century after their invention, DOV methods are still used for less than a third of California’s raisin crop. And it’s not just raisins. Florida citrus farmers have belatedly come to realize that they can never drive down their labor costs enough to match producers in the third world; as the New York Times writes, “Facing increased competition from Brazil and a glut of oranges on world markets, alarmed growers here have been turning to labor-saving technology as their best hope for survival.”52 Mass immigration enabled farmers to avoid making such a commitment for many years, meaning that in 2006, only about 5 percent of Florida’s orange groves used mechanical harvesting.53 Florida’s sugar cane harvest is a good example of how farming modernizes when immigrant labor is no longer cheap.54 In the 1930s, Eleanor Roosevelt decried the working conditions endured by sugar harvesters— using a machete, bending at the waist, dealing with heat, mosquitoes, and snakes—which had changed little since the Middle Ages. It was so bad that in 1942, U.S. Sugar was actually indicted on federal charges of slavery because of its treatment of black American cane-cutters. As a result, the sugar companies began to import West Indians through a federal guest worker visa program. But starting in the 1980s, the industry was hit by a persistent wave of lawsuits filed on behalf of farmworkers whose contracts had been violated by their employers, contracts that guaranteed a certain level of pay along with housing and transportation. Despite years of farmers’ claims that it was impossible to mechanize the harvest of sugar cane, these lawsuits raised the real cost of employing the foreign labor so much that the farmers finally concluded that it would be more profitable to mechanize than to honor all the legally required terms of the farmworker contracts. So, by the 1997-98 growing season, U.S. Sugar, the biggest producer, harvested 100 percent of its cane by machine, resulting in increased productivity, plus higher wages and more civilized working conditions for the remaining harvesters. The threat to the continued competitiveness of U.S. agriculture posed by mass immigration doesn’t come just from the inability to compete on the basis of wages with third-world countries; there is also the danger that the slowing of technological innovation brought about by artificial infusions of labor will allow our economic competitors in other developed countries to leap ahead of us. This is perhaps most disturbing in the field of robotic harvesting.55 Automated picking of fruits and vegetables by a robotic system is the third wave of agricultural mechanization (after labor aids, which facilitate harvesting work but don’t reduce labor demand, and labor-saving machines, which improve productivity and reduce labor needs). The development of viable robotic harvesting technologies is still in its infancy, but great progress is being made. Unfortunately, because of the mass availability of alien labor in the United States, the European Union is well ahead of us in bringing this potentially revolutionary technology to market. Mass immigration’s role in retarding economic modernization is not confined to agriculture, which is, after all, very different from the rest of the economy. Other parts of the economy experience the same phenomenon of a scarcity of low-skilled labor yielding innovation, while a surfeit yields stagnation. An example of the latter: A 1995 report on Southern California’s apparel industry warned of the danger to the industry of reliance on cheap immigrant labor:   While a large, low-cost labor pool has been a boon to apparel production in the past, overreliance on relatively low-cost sources of labor may now cost the industry dearly. The fact is, southern California has fallen behind both domestic and international competitors, even some of its lowest-labor-cost competitors, in applying the array of production and communications technologies available to the industry (such as computer aided design and electronic data interchange).56   As with agriculture, the limited academic inquiry that has been made into manufacturing has found that mass immigration is slowing the spread of labor-saving technology. An economist at the Federal Reserve Bank of Philadelphia has written that “plants in areas experiencing faster less-skilled relative labor supply growth adopted automation technology more slowly, both overall and relative to expectations, and even deadoption was not uncommon,” adding that the effect was even stronger when the growth in less-skilled labor came specifically from immigration. 57 That’s deadoption—some factories actually stopped using labor-saving technology once immigration drove down the price of labor sufficiently. A purer example cannot be found of the conflict between mass immigration and the goals of a modern society. Home construction is another field in which modernization is slowed by mass immigration. The form this modernization takes is modular construction. Modular, or prefab, homes are manufactured in pieces to the building site and assembled and finished off, resulting in a much higher-quality product than a “stick-built” home (one constructed from scratch on-site), and one that is completed faster with much less labor. Technological advances mean that such buildings are nothing like the trailer homes of the past, and the methods can be used even for luxury residences. According to Gopal Ahluwalia, director of research at the National Association of Home Builders, “In the long run, we’ll see a move toward homes built in factories.”58 But the home-building industry has moved very slowly to embrace this transformative technology. According to a 2002 report from the Massachusetts Institute of Technology, “The housing industry is fragmented, resistant to change, labor intensive, inefficient, unresponsive, and wary of new processes and technologies” and “is far behind other industries in the adoption of new process and technology innovations.” 59 As a result, only 3 percent of new homes are built using modular construction. The demand that does exist among builders for modular construction is driven by labor costs. As one modular home manufacturer put it, “With our systems, it’s almost always about labor when we sign up a new builder.”60 But although labor costs have been increasing, they’re lower than they would be otherwise—because of mass immigration. In the words of one building magazine, “Immigrant labor has for years seemed like a bottomless cornucopia of workers.”61 Even in the service sector, there is enormous potential for labor-saving measures that have been rendered less attractive because of the artificial glut of cheap foreign labor. After all, immigrants were not imported to pump gas, so now Americans pump their own gas, aided by technology that lets buyers pay at the pump—thus there are fewer attendants but more gas stations, and customers get in and out faster than before. Likewise with bank tellers, many of whose routine functions are now performed by ATMs, and telephone operators, most of whom were long ago replaced by automated switches. There are plenty of other innovations in the service sector that would spread more quickly if the low-skilled labor market were tightened through lower levels of immigration: Continuous-batch or “tunnel” washing machines can reduce labor demand for hotels, restaurants can install ordering kiosks, movie theaters can use ATM-style devices to sell tickets, the retail industry can adopt increasingly sophisticated vending machines as an alternative to hiring more immigrant clerks. As science fictiony as it might seem, many Veterans Administration hospitals are now using mobile robots to ferry medicines from their pharmacies to various nurse’s stations, eliminating the need for workers to perform that task.62 And devices like automatic vacuum cleaners, lawn mowers, and pool cleaners are increasingly available to consumers. These last examples point to perhaps the greatest competitive threat from mass illegal immigration: its inhibiting effect on the development and spread of robotics. Japan’s society is aging much more rapidly than our own because of its much lower birthrate, but it has decided not to import large numbers of foreign workers, investing instead in robots. Media coverage of this development has focused on cute robotic pets, but this is no laughing matter—Toyota in 2006 announced a major initiative to augment workers with robots at all its Japanese plants, robots much more sophisticated than the thousands of less-advanced devices it already uses for hazardous jobs like welding and painting.63 And the automaking giant plans to start retail sales of household robots by 2010, which the firm expects to become one of its major business units.64 Without a change in immigration policy, we run the risk of future observers noting that “Japan got robots while America got Mexicans.” Historian Otis Graham tied together the impact of mass immigration on American competitiveness and on its workforce: “The U.S. can either evolve towards a high-technology economy with a labor force of constantly advancing productivity, wage levels, and skills, or it can drift towards a low technology, low-skill, and low-wage economy, marked by widespread job instability and growing income disparity. Immigration policy will be important to the outcome.”65

Krikorian, Mark (2008-07-03). The New Case Against Immigration: Both Legal and Illegal (pp. 149-156). Penguin Group. Kindle Edition.

Quotations from Nathan Smith’s blog post and Vipul Naik’s comment, which make the distinction between adjustment and innovation

From Smith’s post:

Technological progress involves creating useful new ideas about how to capture natural phenomena for human ends. Expansion of the stock of such ideas is technological progress in the pure sense. For any given state of technology, that is, for any given stock of technological ideas, there will be many methods of satisfying various human needs, some more labor-intensive, some more capital-intensive. Technological progress is presumptively good as long as peace is maintained. There is no downside to having more options about what to do and how to do it. Capital-intensiveness, by contrast, is costly. To adopt more capital-intensive production methods requires a permanently higher investment rate, which cuts into consumption.

From Naik’s comment:

I think the distinction you draw between substitution of labor and capital for each other, versus innovation, is an important one. While one moves along the production possibilities frontier, the other expands it.

However, figuring out the distinction in practice can be tricky. I think that a lot of what appears to be “labor-saving innovation” may be a form of moving along the frontier rather than expanding it, i.e., an adjustment rather than an innovation.

For instance, suppose country X has closed borders and faces a huge labor shortage. People in country X may invent machines and robots to substitute for labor, thus ameliorating the bad effects of this labor shortage. One possibility is that these machines are technological advances. In that case, it should be possible to see other countries, even those that do not face acute labor shortages, to adopt these machines (though perhaps not to the same extent as country X) in order to take advantage of the expanded production possibilities frontier. The other possibility is that this invention is an adjustment, not an innovation. It’s a predictable invention that other countries already had access to but chose not to follow because they did not anticipate the labor shortage.

The tricky question is: are the labor-saving innovations that result from labor shortages created due to closed borders more likely to be true innovations, or adjustments? I think this is a difficult question, but my guess is that a lot of these are adjustments. The trick is to see if adoption happens in countries not facing the same shortage. In this respect, I think, for instance, that ATM machines were a genuine innovation, possibly in response to the high costs of bank tellers, but one that has spread throughout the world, including places that are not stereotypically associated with labor shortages. But if you pause and think about a lot of the automation-related true innovation, I suspect you’d see that much of it is innovation in areas that don’t very directly substitute for labor. For instance, I guess you could argue that Google search was an innovation due to the high cost of “ask-me-anything” services, but that would be quite a stretch. Mobile messaging (text/SMS) was also an innovation but it’s not clear that there was a specific labor shortage that prompted it.

On the other hand, expensive high-tech robots built in Japan to do tasks that humans are quite good at (like cleaning) and machines are bad at, are an example of an adjustment to a labor shortage, not a true innovation, considering that these innovations haven’t really spread to other countries.

Quotation from Riley’s book where he provides some counter-arguments

Jason Riley makes the point that low-skilled labor-saving innovation substitutes for, and detracts from, the potentially more valuable high-skilled labor-saving innovation. In a section titled “Automation Nation” in his book Let Them In: The Case for Open Borders, Riley spends a couple of paragraphs on this argument.

Some have argued that the availability of cheaper immigrant labor harms productivity by delaying automation in certain industries. “By not enforcing the immigration laws,” says agriculture economist Philip Martin, “the government is sending a signal to farmers that by hiring unauthorized workers they do not have to make a transition to a more mechanized, higher productive agriculture.” In a 2001 study, Martin said that the termination of the Bracero Program for Mexican guest-workers in 1964 led to the automation of tomato picking. “In the tomato case,” he concluded, “the end of the Bracero Program led to the mechanization of the tomato harvest, expanding production, and a reduction in the price of processed tomato products, which helped to fuel the fast-food boom.” Martin and others are right to note that immigration has slowed mechanization in certain sectors of the economy. But their argument presupposes that every activity that can be automated should be, as if the most efficient course is to keep all manual workers outside of developed countries. Does the availability of cashiers retard technological innovations in the retail sector that could produce universal self-service checkout? Not necessarily. To use Martin’s example, time and money spent trying to come up with machines to duplicate a low-skill human activity could have been directed at other improvements, such as developing healthier varieties of tomatoes (rather than varieties that can be recognized by machines). Furthermore, the tomatoes are still being picked by machines at a higher cost than if you allowed a sizeable number of guest-workers to do the job—and with no clear economic gain for Americans. Foreign workers who are less productive in their home countries could be more productive here. And unlike the machines, immigrants not only pick produce but also consume products and services, thus helping the U.S. economy expand.

Riley, Jason L. (2008-05-15). Let Them In: The Case for Open Borders (pp. 75-77). Penguin Group. Kindle Edition.

Quotation from Pritchett’s book

The quotation is from Page 28-30 of the PDF, which is Page 40-42 of the book.

However, the final insight from the technological advances of the lawn mower is that these advances required highly trained engineers working for years to make advances that made an owner-operated labor saving device better. This is nationally sound but globally perverse economics. Given the relative prices and endowments in rich countries, the incentives are to deploy very highly skilled labor to create innovations that reduce demand for low-skilled labor. In fact, there is substantial evidence that technical progress in rich countries has not been neutral between skilled and unskilled labor but rather has been skill enhancing. Moreover, this skill-biased technical change is induced by relative
prices and accounts for a substantial fraction of the rise in wage inequality (and/or unemployment) in industrial countries (Acemoglu and others 2003).

The development literature points out that research in specific areas—such as agriculture or medicine—is biased away from the concerns of the poorer countries, because of differences in willingness to pay. So, for instance, there are innovative proposals to induce pharmaceutical companies to address major health issues facing poor nations because their market incentives are to focus on conditions that disproportionately affect the rich. But the distortion in the research and development induced by restrictions on labor mobility gets almost no attention and almost certainly has an impact that is orders of magnitude larger. The current configuration of the “everything but labor” global economy
produces incentives for the invention of more and more unskilled labor saving devices in a world in which the key price for poverty alleviation is the wage of unskilled labor. Because of the artificially inflated price of labor in rich countries, the rich world is full of highly educated innovators dedicated, indirectly, to lowering the one price on which progress in poverty reduction depends.

Just think of the automated teller machine (ATM), which was invented and then diffused so as to reduce the labor content of handling routine banking transactions. There are almost certainly billions of people who would have been happy to take the jobs an ATM replaces, at wages that would make ATMs uneconomical. However, once the ATM had been invented, the fixed costs of its development borne, banking computing systems made consistent with it, and mass production begun so unit costs fell, then ATMs began to be present even where labor costs are extraordinarily low.

Once this perversity strikes you, it will strike you again and again if you live in a rich country (and particularly if one travels back and forth from poor to rich). In the cities of poor countries, it is not unusual for groceries to be delivered directly to your door. Even when I was a teenager (in the mid-1970s), many of my friends had jobs helping carry groceries to customers’ cars. Now, many retail stores (grocery, hardware, general merchandise) are introducing automated checkout, whereby customers use sophisticated technology and invested capital to ring up and pay for their own groceries. Why did people invent a technology to eliminate people working in retail when billions of the people on the
planet would be pleased to ring up your groceries? This labor-saving innovation was induced by distortions in the international market for labor.

Although something of an aside from labor projections, this is an important point, because one objection raised to allowing temporary labor mobility is that it creates “distortions” in the industries that survive on “cheap labor.” The further argument is that if importing labor were impossible, then industries would not move abroad but would survive by inducing innovations that reduce labor demand and substitute capital for labor. For instance, Martin (2004) tells the story of tomatoes in California and, to my mind, gets the real point exactly backward. In the 1960s, as part of the Bracero program
of allowing temporary migrant labor, tomatoes in California were picked almost entirely by seasonal migrants. When this program ended in the mid-1960s, farmers claimed the tomato industry would leave California. But by a combination of applying science to develop tomatoes whose shape and skin were more conducive to mechanization and developing a machine harvester, the California industry survived and even thrived—Martin emphasizes that it produces five times more tomatoes today than in the 1960s. But from an
economist’s point of view, what is the “distortion”—allowing seasonal workers (that is, more open labor markets across borders) or the induced-labor demand-reducing technological change from enforcing a restriction that willing employers and willing workers could not make a contract?

Any economist, when presented with the same scenario with trade in goods, would be able to give an easy answer—if an industry invents a new technology to displace an imported intermediate input because the price of the input is driven up by border restrictions like tariffs or quotas, this innovation is a response to a distortion, not that the lack of a tariff to induce that innovation would be a distortion.” From a global viewpoint, highly skilled labor devoted to research and development to reduce demand for labor (for example, machine-harvestable tomatoes, lawn mowers, ATMs, self-checkout at retail stores, robots that vacuum, pre-peeled carrots) is an inefficiency that is the result of the massive “distortion” in global labor markets.14 Because about the only thing known yet about “pro-poor” growth is that it is labor intensive, there is obviously a massive contradiction between rich countries pushing
“pro-poor” growth via their rhetoric about development assistance while at the same time promoting massively anti-pro-poor technological change via their policies toward labor mobility.