Self-Milking Cow Machines

And more from this week…

Note: The application is open for this summer’s American Compass internship program. Check it out.

Another month, another jobs report beating expectations, another inflation report beating expectations, another set of economists working on explanations for how really this just goes to show that economists had it right all along, and also things would definitely be even better if everyone had listened to them.

And another journalist stumped by how the agricultural sector could ever possibly create jobs Americans would do, even as he stares right at them. Daniel explains:

In a recent New York Times column, Binyamin Appelbaum argues that immigration restrictionism is a fool’s errand for the agricultural sector. The argument is familiar: these are jobs Americans won’t do; robots can replace some workers, but not all; and if farms cannot import foreign labor, our nation will be forced to import foreign food instead. Policymakers, he concludes, should accept second-class immigrant labor as a permanent feature of American agriculture. But as happens so often with these narratives, the story he tells points precisely the opposite way.

Appelbaum’s central exhibit is Dale Hemminger, a dairy farmer in upstate New York who installed milking robots in 2007—after immigration authorities arrested one of his workers. His farm now produces 2.5 million pounds of milk per worker per year, up from 800,000 before mechanization. His remaining employees—about half the headcount he would otherwise need—earn more, work shorter hours, and do less grueling work.

Appelbaum presents Hemminger as a vision of the future without asking what created the conditions for his success. The answer is obvious: immigration enforcement drove the investment. The result was not the collapse of his dairy farm, but a transformation: higher productivity, better jobs, and a more resilient enterprise. This is precisely the outcome a serious agricultural policy concerned with feeding the nation and respecting its workers should replicate at scale. However, Appelbaum describes the solution and then spends the rest of the piece arguing that we should preserve the conditions that prevent it.

Consider what Hemminger’s example implies about untapped opportunity. A milking robot is a fun piece of technology. (Check out this video from Hemminger’s farm where cows enter the barn when they want to be milked and have their udders cleaned and pulled by a robotic arm.) But it’s not new or exotic. The Dutch company Lely commercialized the first such system in 1992, more than three decades ago, and the technology is common across Northern Europe. About 25% of dairy farms in Denmark and the Netherlands use automated milking systems, compared with about 5% in the United States. That adoption delta is not about engineering capabilities but incentives. American dairy farmers have access to cheap, abundant immigrant labor, which makes investment in mechanization economically irrational. The technology exists and is proven, but most farmers have not deployed it because they have not needed to. The difference between what is available and what farmers have deployed represents enormous potential for productivity growth that immigration enforcement and a reduction in temporary visas would unlock, by altering the firm-level incentive structure.

The deeper problem is Appelbaum’s static analysis of a dynamic market. “There are no robots to do that work,” he writes of crops on Hemminger’s farm, like cabbage, which are still planted and harvested by hand. But the Danish company ASA-LIFT has been manufacturing commercial cabbage harvesters for decades (watch a demonstration here). And the frontier is advancing: a 2024 paper from researchers in South Korea describes the development of an autonomous self-propelled cabbage harvester with AI-based attitude control for cutting on uneven terrain, while a separate team is developing an electric harvester using computer vision to detect and sort cabbages onto a conveyor. The machines exist, or soon will. They are simply not standard on American farms, for the same reason milking robots are not: cheap labor makes them unnecessary. We should wonder what other such innovations American firms might have developed, had demand existed here all along—and what they might develop if demand emerges.

America’s enormous supply of cheap labor is neither a demographic inevitably nor an accident: it was created through distortion of the labor market for the benefit of employers. Visa issuances under the H-2A Temporary Agricultural Workers program have grown from roughly 50,000 in 2005 to more than 385,000 today. And a large share of the broader agricultural workforce lacks legal work authorization, a reality sustained by decades of lax enforcement that functions as a de facto subsidy. Each additional worker, legal or otherwise, reduces the pressure to mechanize and create jobs Americans will do.

Appelbaum concludes by writing that Hemminger’s dairy farm is “a vision of a better future,” but that current policy won’t get us there. On that last point, he is right, just not for the reason he thinks. Progress towards a high-wage, high-tech agricultural economy requires immigration restriction. But it also requires an agricultural industrial policy to facilitate the transition away from labor-intensive production processes, one that provides public investment in research and development, capital subsidies for farms that cannot absorb the upfront costs, and institutional capacity to move technology from the laboratory to the field. Restriction creates pressure. Industrial policy ensures our nation can transition while not losing domestic capacity to import competition.

The phrase “jobs Americans won’t do” is an admission that businesses within certain sectors of our economy have broken business models, premised on working conditions and compensation we do not expect our fellow citizens to tolerate. This framing treats a job’s characteristics as fixed. It suggests incorrectly that agricultural work is inherently incompatible with decent wages, when the real culprits are policy choices, the market structures they create, and the rational ways employers respond. For too long, the United States has organized segments of its agricultural sector around labor access rather than labor productivity. It’s somewhat bizarre that journalists like Appelbaum, usually quick to advocate better for workers and demand more of employers, choose in this instance to shrug and look away. “A vision of a better future” for American agriculture requires policy that creates necessity and provides the conditions and the means. — Daniel

STAYING ON THE LABOR MARKET…

…and woefully bad journalism therein, the Wall Street Journal’s editorial board does not disappoint. Commenting on “Mass Deportation and Florida Jobs,” the Journal laments:

In May 2023, Florida Republicans passed legislation aimed at countering Joe Biden’s porous border policies. The law’s centerpiece requires private employers with 25 or more employees to use the federal government’s E-Verify system to confirm the work authorization of new hires ... .U.S. Labor Department data suggest that Florida’s E-Verify law has harmed job growth. In the year before the law took effect, Florida led the country in job creation. But employment growth fell by half in 2024 and has been flat since last February.

Just so we have this straight: Florida moved aggressively to reduce the presence of illegal workers, and so the number of workers stopped rising as quickly? Gasp. If only there were some measure of job quality for those legal Floridians still working, maybe a measure of their wages? There is! Average wage isn’t the best measure, but it’s what we’ve got at the state level, and in the past 32 months, since May 2023, the average real wage in Florida is up 6.4%. Strong, to quite strong.

How about in the 32 months before passage of E-Verify? Down 3.0%. OK, but that was during COVID. How much did Florida’s average wage rise from pre-COVID to the passage of E-Verify? 0.0%. In fact, it took from mid-2015 to mid-2023 for wages to rise as much in Florida as they did in the two-and-a-half years since passage of E-Verify. So there’s that.

Elsewhere, Howard University’s Ron Hira highlights a new paper from Harvard University’s George Borjas that finds “on average, H-1B workers earn 16 percent less than comparable natives.”

Comparable natives? But H-1Bs are supposed to be for job openings where the employer can’t find a native worker. So how could Borjas even… OK, when you’re done laughing, move to the next item.

One group that might include many comparable natives would be young college grads. Which might help provide Alyssia Finley’s effort in the Wall Street Journal with an answer to: Why Unemployment Is Rising Among Young College Grads. She focuses on another issue, surely part of the problem as well, that “their skills, experience and ability to function are increasingly out of step with employers’ needs.” As she sees it:

Artificial intelligence isn’t taking their jobs. Young grads’ struggles started before AI went mainstream. Between 1990 and 2014, unemployment for young college grads was generally 1 to 3 percentage points lower than for all workers. The gap started to tighten around 2014 and reversed in late 2018. Unemployment for young college grads is now about 1.4 points higher than for all workers. The real problem is a mismatch between labor supply and demand.

Supply and demand! That sounds like a job for Mr. Market, if only policymakers were not instead filling the mismatched demand with alternative, foreign supply.

GOOD READS FOR THE WEEKEND

In the Financial Times, Martin Sandbu has an excellent profile of philosopher Michael Sandel, “the pessimist who became a prophet.”

In the Wall Street Journal, Gavin Bade goes into “The Chinese Factory That Opened in the U.S. and Clobbered Its Rivals.” Or rather, he doesn’t go into it, because he was not allowed in. So all the photos are from the rivals getting clobbered. Wonder why? Could it have anything to do with illegal labor practices so blatantly illegal that the Biden administration was sending in raids?

In RealClearInvestigations, Nancy Rommelmann tells the heartbreaking story of how “Caring for Mom Is an Education in Scams and Fraud.” If we could take 5% of the “serious thinker” attention dedicated to pondering AI, and redirect it to addressing the extraordinary social challenges we will encounter as history’s wealthiest generation continues to age and then passes on, America would be a much better place.

And your academic paper of the week is Work from Home and Fertility, from a bevy of researchers who find “estimated lifetime fertility is greater by 0.32 children per woman when both partners WFH one or more days per week as compared to the case where neither does.”

• Bonus read: One of the first features ever published at Commonplace was on this topic, from EPPC’s Patrick Brown: Remote Work Created a Baby Boom. Can We Keep It Up?

DEPRESSING READS

The New York Times reports, “In South Korea, Questions About Cram Schools, Success and Happiness. Academic pressure has become so intense that even preschoolers are taking private extracurricular classes, raising worries about children’s rights.”

The Times also profiles “Clavicular” (we didn’t know who that was either), in “Handsome at Any Cost.” This one’s pretty crazy. But so is the Times’ conclusion. “Guns, drugs, misogyny, body dysmorphia: The miasma of nihilism swirling around Clavicular has made him an irresistible symbol of social decline — a freakish avatar for the hopelessly fallen, social-media-addled state of the young American man. At the same time, he may simply be ahead of the curve.”

This echoes back to the recent Times profile of professional prediction-market gamblers, where it suggests, “making a living betting on prediction markets just might be one of those era-defining occupations.” The helpless resignation is an inevitable function of a progressive culture that precludes judging any set of choices as worse than any other, but the airy and ironic tone with which the stories are thus told is badly out of step with the challenges we face.

Sticking with the Times, it’s worth reading the editorial board’s far-too-little-too-late mea culpa on marijuana legalization: “This editorial board has long supported marijuana legalization ... .At the time, supporters of legalization predicted that it would bring few downsides….It is now clear that many of these predictions were wrong.” Go figure.

From ABC News, “MLB Players Strike Deal to Be Turned Into AI Characters That Can Chat with Fans.”

From the Financial Times, “Prediction Market Kalshi Seeks US Approval to Offer Margin Trades.” Thank goodness. Only “$24 million Was Bet On Whether Mark Wahlberg Would Attend the Super Bowl” (NBC Sports). But if people could bet borrowed money, we might have a truly liquid and predictive market. Then they’d also have more money to bet in derivative markets predicting how high the prices will go in the prediction markets.

FREE TRADE, FREE MARKETS, CHOOSE ONE

This week’s illustration comes from the OG example of drug pricing. As Oren wrote in the introduction to The Once and Future Worker, nearly a decade ago:

Things become even more complicated when we introduce an international boundary and conflicting legal regimes. We protect patents on new drugs, but what should we do when drugmakers voluntarily sell those patented drugs in Canada at prices far below what they charge in the United States—because the Canadian government requires the lower price? Should someone be allowed to buy the drug in Canada and then resell it in the United States, undercutting a drugmaker’s U.S. price? We call this drug reimportation, and we prohibit it, again on the basis of bolstering the free market, again with strong support from libertarians. Some politicians will offer a rationale of “safety,” as if we can’t trust Canadians to monitor their drug supply as well as we do. The actual rationale is that we wish to insulate what we consider to be our freer market in drugs from contamination by the more controlled Canadian market.

Canada is hardly the archetypal case of market distortion. China, for instance…

Indeed, when one allows multinational corporations to operate across legal jurisdictions that intentionally distort markets to their benefit, one becomes quickly disoriented while trying to stay true to “free market principles.” Like, for instance, the “more than 50 conservative leaders sign[ing a] letter against Trump drug pricing policy codification,” who warn that a Most Favored Nation drug pricing regime (i.e., you can’t charge more for your drugs in America than you charge in other developed countries) would “import socialist price controls and values into our country.” In fact, it is the free trade advocated by said leaders that has done just that, leaving us no choice but to respond.

AS FOR OTHER TRADE NEWS

Wall Street Journal: “The American and Chinese Economies Are Hurtling Toward a Messy Divorce.” Good.

Rhodium Group: “Over the past year, German political leaders, central bankers, business associations, and unions have begun talking openly about the risks of a ‘China shock.’ The rhetoric, however, has not been matched by decisive policy action.” Some brutal charts. Bad.

Reuters: “French Advisers Urges EU Tariffs or Weaker Euro to Counter China.” Good.

Bloomberg: “Ford Falls Behind China’s BYD in Global Sales For the First Time.” Bad.

Politico: “Warren and Banks to Introduce AI OVERWATCH Act.” Good.

Semafor: “NSA Pick Warns of China Pursuing AI Chips for Weapons.” Bad.

If the U.S. and Europe teamed up to exclude imports from China, restrict exports of advanced technology to China, and rebuild industrial supply chains, much would be achieved. Doing what we are doing right now may end less well.

And finally, from an interesting paper, Industrial Policies, Global Imbalances and Technological Hegemony:

We provide a framework that connects industrial policies to global imbalances and technological hegemony, and describe some empirical facts consistent with our model. We study the international spillovers triggered by industrial policies promoting high-tech sectors. Since high-tech goods and services are typically traded internationally, these policies boost the supply of tradable goods. Moreover, industrial policies lead to trade surpluses if the government pursues an unbalanced policy mix, such that domestic demand does not rise as much as supply. These surpluses are absorbed by the rest of the world, which in response runs trade deficits. Absent policy interventions, trade deficits reduce the competitiveness of the domestic tradable sector, stifling innovation and productivity growth. Innovation policies can help the rest of the world to mitigate these negative spillovers.

Tl;dr thread from one of the authors, and one from Michael Pettis.

Enjoy the weekend!