Of course, Borjas's comparable workers exist. But they are all employed in similar jobs and therefore not available for the comparable vacancies filled by lower paid H1B workers. Oren got out a little over his skis here and was too quick to criticize. Besides, the wage differential cited helps make his point. US employers need to pay the going wage and stop gaming the system.
ChatGPT reports the payback period for automated milking robots for 2,000 cow herd is 10 to 20 years, at a cost of $6mm plus which is a huge financial consideration beyond prevailing wages. I’m not an expert on the finances of dairy farms and I suspect Kishi isn’t either. PS I bet those milking machines are built in a country that is not USA.
And why have those machines been developed and are now being manufactured elsewhere? Might it be that by ignoring and even encouraging corruption here in America, we created perverse incentives? That's the point of the article.
You get what you reward. We've been rewarding farmers and an ever increasing number of businesses for exploiting cheap and often illegal foreign labor, both at home and abroad. Which means we have been unintentionally punishing those that try to develop more productive labor saving technologies.
I went to whatever AI Google uses and it shows that net revenue would increase (driven by increasedproduction and decreased labor costs) by about $1.6 million annually or a break even point of 4.6 years. As others point out, if the US market moved towards stronger use of the equipment prices would drop and technology improve (admittedly after a temproary increase until production catches up). PS - same size herd.
If the numbers were compelling I assume more dairy farmers would move faster to robots. If things go south you can fire manual labor and keep the farm, but you can’t fire the debt on robots, and the lender gets the farm.
Hmmm, why were the numbers compelling enough for Europe to move so much faster to robots? Could it be that illegal aliens were not readily available there to work far below market rates and. in turn, drive down the market rate?
BTW, a dairy herd of 2000 requires capital of about $20 million depending on location and farming method. A business this size will usually see gross revenue of $10 million plus and net of $2 million or so (without robotics). With the increased revenues paying for the cost of converting to robotics, the net would remain the same for about five years and then jump significantly thereafter.
The claim that illegal workers are needed for agriculture is simply not true. It is simply that too many farmers are not willing to adapt and/or not good enough business people to make better decisions.
No one has claimed illegal workers are needed, but ones that will accept the wages that the employer can afford to pay, based upon what the market will bear for their product. A lot of things affect milk production and pricing. All I'm saying it takes a HUGE investment in a business with razor thin profit margins. Too many farmers aren't willing to take the risk.
1) The entire point of the article is to refute claims that illegal aliens are needed.
2) If legal applicants won't accept the pay, then the business model is wrong and must be changed (or the business can just go under and someone else will fix it).
3) As with many industries, profit margins vary upon size and efficiency of the business. Profit margins for mid to large dairy herd actually run higher than average for most US businesses. Across all industries, the average is roughly 7.7% to 8.9%.while mid size dairies are 5-15% and large 15-25%.
What we are looking at in the case (and in a lot of agribusiness) is reality catching up on an outdated tradition of farms run on small scale to accomodate a lifestyle choice. Its been declining for a while but - when we eliminate the exploitation of illegal aliens with below market wages - those low margin farms will be taken over by the big ones (and, BTW, a lot of those farmers will go to work for the large dairies for a better income and working conditions).
You're missing the distorting effect of what is a combination of both legalized and illegal corruption that reduces the economies of scale that would have otherwise brought the cost of milking machines down while improving the efficiency of those machines. Just imagine the economies of scale if rather than 5% of dairy farmers in America were using those machines, but more like 75% of dairy farmers?
This is yet another example of basic economics being blatantly ignored when it's convenient to the existing order.
The price probably reflects the capitalized value of the future payments for the manual labor replaced. As adoption becomes more widespread competition among suppliers inevitably drives prices down and performance up.
For sure Frankie. However, the author vastly simplifies switching manual labor for robots. There’s a ton of risk for first movers. As most milk robots are deployed in Europe I’m assuming the robots are designed and built there too, that itself is a risk these days.
The arguments of slaveholders never go away. Slavers always argued that free men would never do the work of slaves. In his Treatise on the Patriarchal, or Co-operative System of Society, Zephaniah Kingsley argued that « the labor of the negro, under the wholesome restraint of an intelligent direction, is like a constant stream; that of the white man is economically measured out by his urgent necessities, or dissipated by his expenses. » Here we are two hundred years later and the same arguments are repeated.
So the solution for a farmer is to take on the fixed expense of single function capital equipment and jettison the variable expense of multifunction labor. Someone has never had to run a business where the product selling price is highly variable.
If the argument is that innovation is systematically under-invested because innovators can't capture all the gains from what they invent, and therefore immigration restriction is good because it shifts the amount of resources spent on innovation closer to optimum, why not just subsidize innovation directly?
The payback period is itself a function of the prevailing wages. If you do a discounted cash flow analysis with the machines plus the labor still required, versus without the machines and a lot more labor, then the labor wages will greatly influence of outcome.
Hi Oren, I just now finished reading your article in the Feb. 6 New York Times, "The Finance Industry Is a Grift. Let’s Start Treating It That Way." I read it twice and will read it again this week. It's a masterpiece, so much to digest and consider. As you say, we need to invest in businesses and industries that build things of use, that employ people who get paid well and can raise families, who shop and dine at businesses near them.
This is how it used to be and we can make it that way again. One thing I never quite understood is how a product can be designed in one area, and then manufactured in another low-cost area. As a young adult, I worked in a factory here that made armored personnel carriers. I was later promoted to the front office as a Quality Control Analyst. The front office was right in front of the factory, not thousands of miles away.
If a problem came up, we received Quality Deficiency Reports from government personnel that formally informed us. It was my duty to investigate the problem in the factory or elsewhere (design, maybe, or vendors...), find a suitable corrective action, and report back to the government. I can't imagine doing this if the production site was in China. (I later got a BA and a MA, but those years at FMC were formative).
I also never understood why stock buybacks were needed, and were kind of a grift. They took money that could go to so many other things, like wages, research, building...I was so happy to see that you believe stock buybacks should be illegal, as they once were.
What the dairy farmer did is what dairy farms have been doing for decades. We’ve seen those machines at state fairs for over 50 years. Now, do something like that on a berry farm.
You're buying into a false argument. There was a strawberry farmer in south Florida that found himself in a bind. Most of the other local farms had sold to developers and he could no longer reliably get the cheap seasonal labor he had in the past. He switched over to raised bed greenhouse farming. As such his profits went up a bit while he was now able to offer year round stable employment with benefits, to a smaller number of workers.
The reality is that we have fallen very far behind where we could have been, while recreating segregation and the corruption it came with, though our lack of immigration enforcement.
In the Northwest, strawberry farms run 20-40 acres. Many are larger. Converting that to greenhouse is an unreasonable expectation. Profit per unit might go up due to year round growth, but the total goes down due to lower quantities. Plus, they flood once or twice a year, so structures are less than ideal.
Enough with the one datapoints. Look at whole industries.
We might havre to move to hydroponics at some point, but I don’t see square miles of farms converting to greenhouses. Where’s the capital to make that happen? If it was a good investment, someone would have done it at scale already.
That farmer in Florida went from 22 acres under cultivation to 4 acres of greenhouses yet was able to increase his total yield. Better, he was more able to manipulate the inner climate so that he could better time his harvest to avoid having to sell lots of strawberries at the same time everybody else's strawberries were hitting the market and driving down prices.
This is the point where you try to point out that it's cold up north. Well, there has been some real innovation in using underground heat pumps to warm up greenhouses. We're back to corruption holding back the needed investment into innovation.
By the way, I'm actually in construction and have been watching the same dynamic play out as more and more contractors and developers have been un-intentionally rewarded for using illegal immigrants rather then innovated to reduce the labor, material, or energy that goes into building stuff.
The same is happening across our industries as companies are allowed to exploit the much lower environmental and labor standards in foreign countries, even while relying on the increasingly bankrupt allies, who have hollowed out their industrial bases to protect the foreign supply chains feeding those foreign factories.
It's all cracking up. Mostly because, that's what you get when you run around trying to justify and make excuses for corruption. By the way, the next really big thing that is going to revolutionize the world is coming in biology and agriculture. Between genetically modified crops, agricultural drones, greenhouse technologies, it's going to be big. But as usual, corruption is holding it back. Why change what you're doing when you can pay off politicians and economists to let you keep exploiting people instead.
Corruption…you use that word a lot and as a reason for everything. Define it in the way you use it. What you say is corruption, which requires deliberate action on the part of one or more groups of people, may be simple economics and convenience.
The economics for that farmer work only because he is able to expand the growing season. We may well come to doing that on a massive scale. With renewable energy, vertical farming (multistory hydroponics buildings) becomes more feasible. We already have the tools and the costs are coming into line to make it work. Maybe five years.
Construction has long resisted advancements to reduce labor. It has less to do with corruption as it does a myriad of other factors, logistics being a big one. The time may come for that to change, too, but it isn’t because people aren’t trying.
I use the word a lot because it has a much bigger effect than most people realize. Why are Canada and America so wealthy while the rest of the Americas are fairly poor? We started off with jury trials which are much harder to corrupt over time than the systems the our southern neighbors inherited from the Spanish.
Of course it's worth pointing out that not all corruption is illegal. Slavery and segregation were legal for a long time, and what parts of the nation lagged behind? Similarly, free trade has been the name of the game these days, all the while ignoring how this has allowed companies to bypass the stricter labor and environmental protections, and hollowed out our industrial base while leaving us increasingly reliant on hostile powers for essential materials and goods.
Construction isn't special in avoiding change. That's the normal state of most industries. We have also had the tools to switch over towards greenhouse farming and other less labor intensive methods for most of my life. More to the point, slavery held off the shift to domestic migratory labor that wound up being more cost efficient than slavery had been.
By the way, I also grew up and live in the south, which has enter a period a massive growth over the last fifty years. It's not a coincidence that this happened after segregation ended. Segregation had been a legalized form of corruption that created millions of people that couldn't use the courts or unionize to fight back against exploitive labor practices.
So pause a minute and really consider all those economic arguments being thrown around. You can't really claim that having millions of people in the country illegally doesn't have a very real effect on the economic decisions being made on a daily basis.
Whether you're talking about a farmer living in a jungle in Columbia, a farmer growing berries in Oregon, or a contractor roofing houses in the suburbs of Dallas, they are all stuck with the reality of living in a world where certain laws aren't being enforced. That means that their competitors will get ahead and put them out of business if they don't also break the laws that aren't being enforced anyway.
Your last paragraph is true, as are your comments about resistance to change and corporations’ tendency to work very hard to skirt labor and environmental laws. Your use of “corruption”, while understandable given your explanation, is still misleading because it doesn’t fit the general use of the term. There are other words that more clearly communicate your point.
That said, many innovations in farming and construction have failed to catch on for both economic and non-economic reasons. You blame the use of immigrant labor as the cause of these failures. More likely, it’s the result of them, although there are probably aspects of both. In the Skagit Valley of Washington, migrants showed up, they weren’t actively imported. After strawberry season in June, they moved east to pick cherries, then apples and other fruits. Many migrants came back year after year (I know this from personal experience). The pay was bad, but a lot better than they could get at home, so they came. Further, many of those migrants are here legally.
Doesn’t this analysis repeat the exact mistake Frédéric Bastiat warned about in the broken window example—focusing on the visible gains from forced investment in automation while ignoring the unseen counterfactual? If a farmer is pushed into buying expensive milking robots because labor supply tightens, isn’t the real economic question what that capital would have done otherwise—expanding the herd, improving land, reducing debt, investing in a different line of production, or flowing to a higher-return sector altogether? In a functioning market, firms already compare labor and capital costs; if mechanization were clearly superior absent distortion, wouldn’t it have been adopted more broadly without enforcement shocks? How can we call this a net productivity gain without accounting for the opportunity costs of redirecting capital under constraint?
The Bastiat analogy is backwards. The broken window parable warns against creating destruction to stimulate activity. My argument is that we already have a distortion—a labor market shaped by H-2A expansion and decades of lax enforcement of immigration law—and that removing it would allow the market to function properly. Under Bastiat’s framework, the “unseen” is not the opportunity cost of buying a milking robot but the decades of foregone mechanization, productivity growth, and wage improvement that never materialized because artificially cheap labor suppressed the incentive to invest. Your question, “Wouldn’t mechanization have been adopted if it were clearly superior?” assumes the status quo reflects an undistorted market. Again, it doesn’t. When a key input is priced below its true cost because of public policy, firms rationally respond to the subsidy by underinvesting in substitutes.
I’m struggling with the claim that labor is being pushed below its “true cost” by public policy. What exactly is the mechanism that drives wages below competitive equilibrium?
A distortion is a tax, subsidy, price control, or legal constraint that wedges price away from marginal cost. A visa program that allows additional workers to enter the labor market does not, by itself, create such a wedge. It increases supply, right? That lowers wages relative to a smaller labor pool. That’s not price suppression below “true cost”; that is the market-clearing price given the expanded supply.
To say labor is priced below its “true cost” assumes there is some baseline supply that reflects the authentic market and that additional workers represent an artificial deviation. But labor markets are always defined by law — citizenship rules, work authorization, contract enforcement. Expanding eligibility is not the same thing as subsidizing an input. Unless H-2A involves an actual financial transfer to employers, or legally constrains workers in ways that generate monopsony power and prevent wage competition, it’s difficult to see where the price wedge is.
Doesn't this matter for the mechanization argument? Firms compare relative prices- right?.
If labor is cheaper than capital, they hire labor. If capital is cheaper, they automate. That’s not underinvestment; it’s rational allocation at prevailing prices. To argue that automation was “suppressed” presumes we know the socially optimal capital–labor ratio better than the market revealed at those relative prices. How do you know better than the market?
If the logic is that reducing labor supply is good because it raises wages and forces capital deepening, then the principle isn’t about correcting a distortion — it’s about intentionally tightening labor markets to engineer a different capital–labor mix. That logic doesn’t stop at immigration. By the same reasoning, any reduction in labor supply — lower labor-force participation, demographic contraction, even deliberate exclusion of workers — would be “productivity enhancing” because it forces substitution toward capital. That’s a very different argument than restoring a properly functioning market. It’s an argument for constraining supply to produce a preferred outcome.
If the position is normative — that we want fewer workers in order to drive investment and wages higher — that’s a coherent policy choice. But calling labor abundance a distortion requires demonstrating an actual price wedge below competitive equilibrium, not simply pointing to a larger labor supply and preferring a smaller one.
What you're missing is the corruption that is created through illegal immigration and temporary workers programs that are tied to employment. Both create, a more or less, segregated group that can be exploited with out the ability to turn to the state, the courts, or unionize. Add in that this distorts all adjacent economics as well.
Landlords can exploit the illegals. Citizens and legal immigrants are forced to compete against the segregated class that lacks the same rights and privileges which can, in and of itself, effectively mean that legal immigrants and citizens also can't turn to the state, the courts, or unionize.
In other words, we're recreating the bad old days of the south. There were a number of reasons the south largely stagnated while the rest of the nation leaped ahead. Investing in labor saving technologies pays off much better but that takes time, while using things like debt peonage makes more money upfront and that can often be used to buy off the political class to write the rules that favor the corruption of things like debt peonage.
So the farmer that uses illegal immigrants have more profits to spend buying off politicians today, than the farmers investing their profits in innovation. It's the same thing we've been battling since the beginning of the enlightenment.
I don’t disagree that corruption and exploitation can exist in sectors that rely on unauthorized labor or tied visa programs. But the source of that corruption isn’t “immigrants.” It’s the legal framework that creates vulnerability.
If workers lack portability, meaningful legal recourse, or the practical ability to assert rights without risking removal, that asymmetry is created by law. The distortion, then, isn’t labor supply — it’s institutional design. The remedy for that kind of distortion is fixing portability, enforcement symmetry, and access to courts — not shrinking the number of workers.
If the concern is monopsony power created by tied visas or status vulnerability, that’s an argument about labor rights and rule-of-law consistency. It’s not an argument that additional willing workers inherently corrupt markets. We shouldn’t conflate exploitation enabled by legal precarity with the simple fact of labor abundance.
There’s also a deeper principle here. Liberal markets function best under universalism — equal protection, equal recourse, freedom of contract, freedom of exit. When we create a legally segregated class with fewer practical protections, we undermine those Enlightenment commitments. But the answer to unequal liberty is not fewer participants; it’s equal liberty under neutral rules.
If we care about human rights, rule of law, and the dignity of labor, then the solution isn’t to remove workers from the system. It’s to ensure that anyone participating in the labor market can rely on the same legal protections and institutional safeguards. Markets disciplined by universal rights and freedom of movement are consistent with Enlightenment values. Markets structured around selective vulnerability are not.
So if the argument is that current immigration policy creates coercive asymmetries, that’s a serious institutional critique. But that’s fundamentally different from claiming that increasing labor supply is itself a market distortion.
You're fundamentally right about this being a political or legal problem. What you're overlooking is that this corruption has been going for so long that it has led to an overall distrust in our governments willingness or ability to enforce the law. Which in turn means we're not having even a remotely honest debate over the actual needs of the labor markets or what metrics we should honestly be using to determine immigration rates in the first place. Much less which people we should be bringing in.
I assume you're not actually trying to propose that we allow the 6 billion people on the planet that would happily contract at the lowest legally allowed wage to be able to immigrate just because they wanted to. Which brings us back to the need to first establish that our government will largely enforce the laws as written. There is no rule of law if the laws are selectively enforced.
"Markets disciplined by universal rights and freedom of movement are consistent with Enlightenment values", There are universal rights and freedom of movement within the nation state, those don't exist in nature. Therefor markets can largely only be disciplined with our nation's boundaries. I think that could be extended to other democratic, allied nations that can agree on enforceable minimum on defenses, environmental and worker protections. But then, that why the western world order is starting to fall apart. Too many people trying to live in theoretical worlds rather then the real one.
Seriously, it's all falling apart and will only continue to do so until we stop cherry picking what arguments we want to use, and start looking at the actual results of our policies and/or lack of enforcement of the law.
I think this response is mixing several distinct issues — rule of law, visa design, undocumented labor, and basic price theory — and they need to be separated for clarity.
First, on Bastiat. I don’t think anything has challenged my original point. The broken window fallacy warns against focusing on the visible gains from a forced shift while ignoring the unseen counterfactual. If enforcement pressure pushes capital toward automation, we can observe the mechanization and higher wages for retained workers. What we don’t observe is what that capital would have done otherwise — alternative investments, expansion, debt reduction, innovation in adjacent sectors. If we’re going to argue that tightening labor supply improves productivity, we have to account for opportunity cost, not just the visible substitution.
Second, on “corruption.” There’s a fundamental difference between undocumented workers operating in a shadow market and workers participating in a lawful visa program. Those are not interchangeable categories. A legal temporary worker program may have design flaws — especially if portability is weak — but there is nothing inherently corrupt about participating in a lawful work arrangement. Conflating legal visa holders with illegal entry muddies the analysis.
If the concern is that certain visa structures create dependency or monopsony conditions, then that’s a design critique. Fine. That’s not the same as saying labor supply itself is corrupting the market.
Third, the “6 billion people willing to work for the lowest wage” framing is a straw man. No serious person is arguing that anyone who wants a contract should automatically receive one. Contracts occur because both parties expect mutual gain. Firms hire when the marginal product of labor exceeds the wage. Workers accept when the wage exceeds their next-best alternative. That’s not charity; that’s exchange.
More broadly, the global integration of labor and goods markets — imperfect as it is — has been historically transformative. Expanding the set of people who can trade with one another across borders dramatically increased output, reduced extreme poverty, lowered consumer prices, and improved living standards worldwide. That wasn’t theoretical; it’s measurable. The fact that globalization produced distributional tradeoffs doesn’t negate the aggregate gains.
So we absolutely can debate optimal immigration levels, enforcement credibility, visa portability, and labor-market effects. Those are political questions. But we shouldn’t conflate (1) labor supply expansion, (2) unlawful status vulnerability, and (3) actual economic distortions (like those that come through regulation, taxation (tariffs), subsidies, and other government actions).
We shouldn’t assume that restricting labor supply automatically produces net productivity gains without reckoning with the unseen opportunity costs Bastiat warned us about.
If the argument is that we prefer a tighter labor market for distributional reasons — higher wages for incumbents, faster capital deepening — that’s a legitimate policy preference. But that’s different from claiming that lawful labor market participation is itself corruption or that increased labor supply is inherently a distortion.
I agree with you that rule of law matters. If laws are selectively enforced, institutional trust erodes. That’s real. I’m not arguing against national borders or against having immigration limits. Every country sets entry rules, and there are real benefits to expanding those (eg, how we have immigration freely between all 50 states). My concern is narrower: are we being precise about what counts as economic distortion versus what counts as a political preference about enforcement levels?
When we talk about corruption, are we distinguishing clearly between undocumented labor operating in a shadow market and workers participating in a lawful visa program? If a visa program has weak portability or creates vulnerability, that’s a design flaw worth fixing. But is participation in a legal work program itself corrupt? Or is the corruption located in how we structure and enforce it?
On the economic side, I keep coming back to the same question. If tightening labor supply forces investment in automation, how are we accounting for the unseen tradeoffs? As Frédéric Bastiat warned, it’s easy to see the visible shift — robots installed, wages rising for remaining workers — but harder to see what that capital would have done otherwise. Is it obvious that every dollar redirected into mechanization represents net productivity gain, rather than reallocation under constraint?
When we say things are “falling apart,” I wonder what specific causal chain we’re identifying. Are housing shortages primarily caused by immigration — or by zoning rules and permitting constraints that suppress supply? Are low labor-force participation rates driven by migrant competition — or by benefit structures and regulatory barriers that discourage work? Is crime a function of immigration levels — or of enforcement choices and prosecutorial priorities? These are empirical questions, and it seems risky to collapse them into a single narrative about labor supply.
You’re right that we shouldn’t cherry-pick arguments. But that cuts both ways. If we credit immigration with every downstream institutional frustration, aren’t we potentially overlooking the policy choices that more directly explain those outcomes? Expanding labor supply changes wage equilibria. Restricting it changes them in the opposite direction. That’s a tradeoff- nothing more or less.
I would suggest a more honest framing is this: what level and structure of immigration best balances wage growth, capital formation, institutional integrity, and rule-of-law credibility? That’s a difficult policy question, and one the article provides precious little empirical support. I just don’t think it helps to describe labor abundance itself as a distortion or to assume that broader social strain necessarily originates there.
It's lovely that Oren thinks DonOrenomics is unspooling so successfully. Joepa used to assure us of the same.
Perhaps someone should alert his fellow Americans, who rate DonOrenomics as exceedingly unpopular. Their stewardship ranks the same as Joepa's. Or worse. I love irony.
I guess life's pretty good for the Epstein class. For the working class stiffs, not so much.
I’m always impressed by posters like Karl whose profiles show no subscriptions, no reads and no likes. Just an empty profile. Like they are bots or trolls or something.
Hey Art, good to hear from you. No bot:) Just an ex-Republican trying to learn what the "new" right stands for beyond grievance and devotion to one politician. It's a riddle.
Of course, Borjas's comparable workers exist. But they are all employed in similar jobs and therefore not available for the comparable vacancies filled by lower paid H1B workers. Oren got out a little over his skis here and was too quick to criticize. Besides, the wage differential cited helps make his point. US employers need to pay the going wage and stop gaming the system.
ChatGPT reports the payback period for automated milking robots for 2,000 cow herd is 10 to 20 years, at a cost of $6mm plus which is a huge financial consideration beyond prevailing wages. I’m not an expert on the finances of dairy farms and I suspect Kishi isn’t either. PS I bet those milking machines are built in a country that is not USA.
And why have those machines been developed and are now being manufactured elsewhere? Might it be that by ignoring and even encouraging corruption here in America, we created perverse incentives? That's the point of the article.
You get what you reward. We've been rewarding farmers and an ever increasing number of businesses for exploiting cheap and often illegal foreign labor, both at home and abroad. Which means we have been unintentionally punishing those that try to develop more productive labor saving technologies.
Automated milkers are not that much - starting around $400-800K.
First fresheners still have to be trained to use the machine. But they learn fast.
ChatGPT stated 25 machines at $250,000 each would be needed for a herd of 2,000 plus other outfitting costs to bring the total to $6M.
I went to whatever AI Google uses and it shows that net revenue would increase (driven by increasedproduction and decreased labor costs) by about $1.6 million annually or a break even point of 4.6 years. As others point out, if the US market moved towards stronger use of the equipment prices would drop and technology improve (admittedly after a temproary increase until production catches up). PS - same size herd.
If the numbers were compelling I assume more dairy farmers would move faster to robots. If things go south you can fire manual labor and keep the farm, but you can’t fire the debt on robots, and the lender gets the farm.
Hmmm, why were the numbers compelling enough for Europe to move so much faster to robots? Could it be that illegal aliens were not readily available there to work far below market rates and. in turn, drive down the market rate?
BTW, a dairy herd of 2000 requires capital of about $20 million depending on location and farming method. A business this size will usually see gross revenue of $10 million plus and net of $2 million or so (without robotics). With the increased revenues paying for the cost of converting to robotics, the net would remain the same for about five years and then jump significantly thereafter.
The claim that illegal workers are needed for agriculture is simply not true. It is simply that too many farmers are not willing to adapt and/or not good enough business people to make better decisions.
No one has claimed illegal workers are needed, but ones that will accept the wages that the employer can afford to pay, based upon what the market will bear for their product. A lot of things affect milk production and pricing. All I'm saying it takes a HUGE investment in a business with razor thin profit margins. Too many farmers aren't willing to take the risk.
1) The entire point of the article is to refute claims that illegal aliens are needed.
2) If legal applicants won't accept the pay, then the business model is wrong and must be changed (or the business can just go under and someone else will fix it).
3) As with many industries, profit margins vary upon size and efficiency of the business. Profit margins for mid to large dairy herd actually run higher than average for most US businesses. Across all industries, the average is roughly 7.7% to 8.9%.while mid size dairies are 5-15% and large 15-25%.
What we are looking at in the case (and in a lot of agribusiness) is reality catching up on an outdated tradition of farms run on small scale to accomodate a lifestyle choice. Its been declining for a while but - when we eliminate the exploitation of illegal aliens with below market wages - those low margin farms will be taken over by the big ones (and, BTW, a lot of those farmers will go to work for the large dairies for a better income and working conditions).
You're missing the distorting effect of what is a combination of both legalized and illegal corruption that reduces the economies of scale that would have otherwise brought the cost of milking machines down while improving the efficiency of those machines. Just imagine the economies of scale if rather than 5% of dairy farmers in America were using those machines, but more like 75% of dairy farmers?
This is yet another example of basic economics being blatantly ignored when it's convenient to the existing order.
The price probably reflects the capitalized value of the future payments for the manual labor replaced. As adoption becomes more widespread competition among suppliers inevitably drives prices down and performance up.
For sure Frankie. However, the author vastly simplifies switching manual labor for robots. There’s a ton of risk for first movers. As most milk robots are deployed in Europe I’m assuming the robots are designed and built there too, that itself is a risk these days.
The arguments of slaveholders never go away. Slavers always argued that free men would never do the work of slaves. In his Treatise on the Patriarchal, or Co-operative System of Society, Zephaniah Kingsley argued that « the labor of the negro, under the wholesome restraint of an intelligent direction, is like a constant stream; that of the white man is economically measured out by his urgent necessities, or dissipated by his expenses. » Here we are two hundred years later and the same arguments are repeated.
Cheap labor is not only necessary, but advantageous. Kingsley argued that slavery was good for slaves too. Sound familiar? https://www.floridamemory.com/items/show/345402
You have a very stupid definition of slavery.
So the solution for a farmer is to take on the fixed expense of single function capital equipment and jettison the variable expense of multifunction labor. Someone has never had to run a business where the product selling price is highly variable.
Cryus McCormick has joined the chat. Also some unknown Babylonian who invited the seed drill.
If the argument is that innovation is systematically under-invested because innovators can't capture all the gains from what they invent, and therefore immigration restriction is good because it shifts the amount of resources spent on innovation closer to optimum, why not just subsidize innovation directly?
The payback period is itself a function of the prevailing wages. If you do a discounted cash flow analysis with the machines plus the labor still required, versus without the machines and a lot more labor, then the labor wages will greatly influence of outcome.
Hi Oren, I just now finished reading your article in the Feb. 6 New York Times, "The Finance Industry Is a Grift. Let’s Start Treating It That Way." I read it twice and will read it again this week. It's a masterpiece, so much to digest and consider. As you say, we need to invest in businesses and industries that build things of use, that employ people who get paid well and can raise families, who shop and dine at businesses near them.
This is how it used to be and we can make it that way again. One thing I never quite understood is how a product can be designed in one area, and then manufactured in another low-cost area. As a young adult, I worked in a factory here that made armored personnel carriers. I was later promoted to the front office as a Quality Control Analyst. The front office was right in front of the factory, not thousands of miles away.
If a problem came up, we received Quality Deficiency Reports from government personnel that formally informed us. It was my duty to investigate the problem in the factory or elsewhere (design, maybe, or vendors...), find a suitable corrective action, and report back to the government. I can't imagine doing this if the production site was in China. (I later got a BA and a MA, but those years at FMC were formative).
I also never understood why stock buybacks were needed, and were kind of a grift. They took money that could go to so many other things, like wages, research, building...I was so happy to see that you believe stock buybacks should be illegal, as they once were.
What the dairy farmer did is what dairy farms have been doing for decades. We’ve seen those machines at state fairs for over 50 years. Now, do something like that on a berry farm.
You're buying into a false argument. There was a strawberry farmer in south Florida that found himself in a bind. Most of the other local farms had sold to developers and he could no longer reliably get the cheap seasonal labor he had in the past. He switched over to raised bed greenhouse farming. As such his profits went up a bit while he was now able to offer year round stable employment with benefits, to a smaller number of workers.
The reality is that we have fallen very far behind where we could have been, while recreating segregation and the corruption it came with, though our lack of immigration enforcement.
In the Northwest, strawberry farms run 20-40 acres. Many are larger. Converting that to greenhouse is an unreasonable expectation. Profit per unit might go up due to year round growth, but the total goes down due to lower quantities. Plus, they flood once or twice a year, so structures are less than ideal.
Enough with the one datapoints. Look at whole industries.
We might havre to move to hydroponics at some point, but I don’t see square miles of farms converting to greenhouses. Where’s the capital to make that happen? If it was a good investment, someone would have done it at scale already.
That farmer in Florida went from 22 acres under cultivation to 4 acres of greenhouses yet was able to increase his total yield. Better, he was more able to manipulate the inner climate so that he could better time his harvest to avoid having to sell lots of strawberries at the same time everybody else's strawberries were hitting the market and driving down prices.
This is the point where you try to point out that it's cold up north. Well, there has been some real innovation in using underground heat pumps to warm up greenhouses. We're back to corruption holding back the needed investment into innovation.
By the way, I'm actually in construction and have been watching the same dynamic play out as more and more contractors and developers have been un-intentionally rewarded for using illegal immigrants rather then innovated to reduce the labor, material, or energy that goes into building stuff.
The same is happening across our industries as companies are allowed to exploit the much lower environmental and labor standards in foreign countries, even while relying on the increasingly bankrupt allies, who have hollowed out their industrial bases to protect the foreign supply chains feeding those foreign factories.
It's all cracking up. Mostly because, that's what you get when you run around trying to justify and make excuses for corruption. By the way, the next really big thing that is going to revolutionize the world is coming in biology and agriculture. Between genetically modified crops, agricultural drones, greenhouse technologies, it's going to be big. But as usual, corruption is holding it back. Why change what you're doing when you can pay off politicians and economists to let you keep exploiting people instead.
Corruption…you use that word a lot and as a reason for everything. Define it in the way you use it. What you say is corruption, which requires deliberate action on the part of one or more groups of people, may be simple economics and convenience.
The economics for that farmer work only because he is able to expand the growing season. We may well come to doing that on a massive scale. With renewable energy, vertical farming (multistory hydroponics buildings) becomes more feasible. We already have the tools and the costs are coming into line to make it work. Maybe five years.
Construction has long resisted advancements to reduce labor. It has less to do with corruption as it does a myriad of other factors, logistics being a big one. The time may come for that to change, too, but it isn’t because people aren’t trying.
I use the word a lot because it has a much bigger effect than most people realize. Why are Canada and America so wealthy while the rest of the Americas are fairly poor? We started off with jury trials which are much harder to corrupt over time than the systems the our southern neighbors inherited from the Spanish.
Of course it's worth pointing out that not all corruption is illegal. Slavery and segregation were legal for a long time, and what parts of the nation lagged behind? Similarly, free trade has been the name of the game these days, all the while ignoring how this has allowed companies to bypass the stricter labor and environmental protections, and hollowed out our industrial base while leaving us increasingly reliant on hostile powers for essential materials and goods.
Construction isn't special in avoiding change. That's the normal state of most industries. We have also had the tools to switch over towards greenhouse farming and other less labor intensive methods for most of my life. More to the point, slavery held off the shift to domestic migratory labor that wound up being more cost efficient than slavery had been.
By the way, I also grew up and live in the south, which has enter a period a massive growth over the last fifty years. It's not a coincidence that this happened after segregation ended. Segregation had been a legalized form of corruption that created millions of people that couldn't use the courts or unionize to fight back against exploitive labor practices.
So pause a minute and really consider all those economic arguments being thrown around. You can't really claim that having millions of people in the country illegally doesn't have a very real effect on the economic decisions being made on a daily basis.
Whether you're talking about a farmer living in a jungle in Columbia, a farmer growing berries in Oregon, or a contractor roofing houses in the suburbs of Dallas, they are all stuck with the reality of living in a world where certain laws aren't being enforced. That means that their competitors will get ahead and put them out of business if they don't also break the laws that aren't being enforced anyway.
Your last paragraph is true, as are your comments about resistance to change and corporations’ tendency to work very hard to skirt labor and environmental laws. Your use of “corruption”, while understandable given your explanation, is still misleading because it doesn’t fit the general use of the term. There are other words that more clearly communicate your point.
That said, many innovations in farming and construction have failed to catch on for both economic and non-economic reasons. You blame the use of immigrant labor as the cause of these failures. More likely, it’s the result of them, although there are probably aspects of both. In the Skagit Valley of Washington, migrants showed up, they weren’t actively imported. After strawberry season in June, they moved east to pick cherries, then apples and other fruits. Many migrants came back year after year (I know this from personal experience). The pay was bad, but a lot better than they could get at home, so they came. Further, many of those migrants are here legally.
Doesn’t this analysis repeat the exact mistake Frédéric Bastiat warned about in the broken window example—focusing on the visible gains from forced investment in automation while ignoring the unseen counterfactual? If a farmer is pushed into buying expensive milking robots because labor supply tightens, isn’t the real economic question what that capital would have done otherwise—expanding the herd, improving land, reducing debt, investing in a different line of production, or flowing to a higher-return sector altogether? In a functioning market, firms already compare labor and capital costs; if mechanization were clearly superior absent distortion, wouldn’t it have been adopted more broadly without enforcement shocks? How can we call this a net productivity gain without accounting for the opportunity costs of redirecting capital under constraint?
The Bastiat analogy is backwards. The broken window parable warns against creating destruction to stimulate activity. My argument is that we already have a distortion—a labor market shaped by H-2A expansion and decades of lax enforcement of immigration law—and that removing it would allow the market to function properly. Under Bastiat’s framework, the “unseen” is not the opportunity cost of buying a milking robot but the decades of foregone mechanization, productivity growth, and wage improvement that never materialized because artificially cheap labor suppressed the incentive to invest. Your question, “Wouldn’t mechanization have been adopted if it were clearly superior?” assumes the status quo reflects an undistorted market. Again, it doesn’t. When a key input is priced below its true cost because of public policy, firms rationally respond to the subsidy by underinvesting in substitutes.
I’m struggling with the claim that labor is being pushed below its “true cost” by public policy. What exactly is the mechanism that drives wages below competitive equilibrium?
A distortion is a tax, subsidy, price control, or legal constraint that wedges price away from marginal cost. A visa program that allows additional workers to enter the labor market does not, by itself, create such a wedge. It increases supply, right? That lowers wages relative to a smaller labor pool. That’s not price suppression below “true cost”; that is the market-clearing price given the expanded supply.
To say labor is priced below its “true cost” assumes there is some baseline supply that reflects the authentic market and that additional workers represent an artificial deviation. But labor markets are always defined by law — citizenship rules, work authorization, contract enforcement. Expanding eligibility is not the same thing as subsidizing an input. Unless H-2A involves an actual financial transfer to employers, or legally constrains workers in ways that generate monopsony power and prevent wage competition, it’s difficult to see where the price wedge is.
Doesn't this matter for the mechanization argument? Firms compare relative prices- right?.
If labor is cheaper than capital, they hire labor. If capital is cheaper, they automate. That’s not underinvestment; it’s rational allocation at prevailing prices. To argue that automation was “suppressed” presumes we know the socially optimal capital–labor ratio better than the market revealed at those relative prices. How do you know better than the market?
If the logic is that reducing labor supply is good because it raises wages and forces capital deepening, then the principle isn’t about correcting a distortion — it’s about intentionally tightening labor markets to engineer a different capital–labor mix. That logic doesn’t stop at immigration. By the same reasoning, any reduction in labor supply — lower labor-force participation, demographic contraction, even deliberate exclusion of workers — would be “productivity enhancing” because it forces substitution toward capital. That’s a very different argument than restoring a properly functioning market. It’s an argument for constraining supply to produce a preferred outcome.
If the position is normative — that we want fewer workers in order to drive investment and wages higher — that’s a coherent policy choice. But calling labor abundance a distortion requires demonstrating an actual price wedge below competitive equilibrium, not simply pointing to a larger labor supply and preferring a smaller one.
What you're missing is the corruption that is created through illegal immigration and temporary workers programs that are tied to employment. Both create, a more or less, segregated group that can be exploited with out the ability to turn to the state, the courts, or unionize. Add in that this distorts all adjacent economics as well.
Landlords can exploit the illegals. Citizens and legal immigrants are forced to compete against the segregated class that lacks the same rights and privileges which can, in and of itself, effectively mean that legal immigrants and citizens also can't turn to the state, the courts, or unionize.
In other words, we're recreating the bad old days of the south. There were a number of reasons the south largely stagnated while the rest of the nation leaped ahead. Investing in labor saving technologies pays off much better but that takes time, while using things like debt peonage makes more money upfront and that can often be used to buy off the political class to write the rules that favor the corruption of things like debt peonage.
So the farmer that uses illegal immigrants have more profits to spend buying off politicians today, than the farmers investing their profits in innovation. It's the same thing we've been battling since the beginning of the enlightenment.
I don’t disagree that corruption and exploitation can exist in sectors that rely on unauthorized labor or tied visa programs. But the source of that corruption isn’t “immigrants.” It’s the legal framework that creates vulnerability.
If workers lack portability, meaningful legal recourse, or the practical ability to assert rights without risking removal, that asymmetry is created by law. The distortion, then, isn’t labor supply — it’s institutional design. The remedy for that kind of distortion is fixing portability, enforcement symmetry, and access to courts — not shrinking the number of workers.
If the concern is monopsony power created by tied visas or status vulnerability, that’s an argument about labor rights and rule-of-law consistency. It’s not an argument that additional willing workers inherently corrupt markets. We shouldn’t conflate exploitation enabled by legal precarity with the simple fact of labor abundance.
There’s also a deeper principle here. Liberal markets function best under universalism — equal protection, equal recourse, freedom of contract, freedom of exit. When we create a legally segregated class with fewer practical protections, we undermine those Enlightenment commitments. But the answer to unequal liberty is not fewer participants; it’s equal liberty under neutral rules.
If we care about human rights, rule of law, and the dignity of labor, then the solution isn’t to remove workers from the system. It’s to ensure that anyone participating in the labor market can rely on the same legal protections and institutional safeguards. Markets disciplined by universal rights and freedom of movement are consistent with Enlightenment values. Markets structured around selective vulnerability are not.
So if the argument is that current immigration policy creates coercive asymmetries, that’s a serious institutional critique. But that’s fundamentally different from claiming that increasing labor supply is itself a market distortion.
You're fundamentally right about this being a political or legal problem. What you're overlooking is that this corruption has been going for so long that it has led to an overall distrust in our governments willingness or ability to enforce the law. Which in turn means we're not having even a remotely honest debate over the actual needs of the labor markets or what metrics we should honestly be using to determine immigration rates in the first place. Much less which people we should be bringing in.
I assume you're not actually trying to propose that we allow the 6 billion people on the planet that would happily contract at the lowest legally allowed wage to be able to immigrate just because they wanted to. Which brings us back to the need to first establish that our government will largely enforce the laws as written. There is no rule of law if the laws are selectively enforced.
"Markets disciplined by universal rights and freedom of movement are consistent with Enlightenment values", There are universal rights and freedom of movement within the nation state, those don't exist in nature. Therefor markets can largely only be disciplined with our nation's boundaries. I think that could be extended to other democratic, allied nations that can agree on enforceable minimum on defenses, environmental and worker protections. But then, that why the western world order is starting to fall apart. Too many people trying to live in theoretical worlds rather then the real one.
Seriously, it's all falling apart and will only continue to do so until we stop cherry picking what arguments we want to use, and start looking at the actual results of our policies and/or lack of enforcement of the law.
I think this response is mixing several distinct issues — rule of law, visa design, undocumented labor, and basic price theory — and they need to be separated for clarity.
First, on Bastiat. I don’t think anything has challenged my original point. The broken window fallacy warns against focusing on the visible gains from a forced shift while ignoring the unseen counterfactual. If enforcement pressure pushes capital toward automation, we can observe the mechanization and higher wages for retained workers. What we don’t observe is what that capital would have done otherwise — alternative investments, expansion, debt reduction, innovation in adjacent sectors. If we’re going to argue that tightening labor supply improves productivity, we have to account for opportunity cost, not just the visible substitution.
Second, on “corruption.” There’s a fundamental difference between undocumented workers operating in a shadow market and workers participating in a lawful visa program. Those are not interchangeable categories. A legal temporary worker program may have design flaws — especially if portability is weak — but there is nothing inherently corrupt about participating in a lawful work arrangement. Conflating legal visa holders with illegal entry muddies the analysis.
If the concern is that certain visa structures create dependency or monopsony conditions, then that’s a design critique. Fine. That’s not the same as saying labor supply itself is corrupting the market.
Third, the “6 billion people willing to work for the lowest wage” framing is a straw man. No serious person is arguing that anyone who wants a contract should automatically receive one. Contracts occur because both parties expect mutual gain. Firms hire when the marginal product of labor exceeds the wage. Workers accept when the wage exceeds their next-best alternative. That’s not charity; that’s exchange.
More broadly, the global integration of labor and goods markets — imperfect as it is — has been historically transformative. Expanding the set of people who can trade with one another across borders dramatically increased output, reduced extreme poverty, lowered consumer prices, and improved living standards worldwide. That wasn’t theoretical; it’s measurable. The fact that globalization produced distributional tradeoffs doesn’t negate the aggregate gains.
So we absolutely can debate optimal immigration levels, enforcement credibility, visa portability, and labor-market effects. Those are political questions. But we shouldn’t conflate (1) labor supply expansion, (2) unlawful status vulnerability, and (3) actual economic distortions (like those that come through regulation, taxation (tariffs), subsidies, and other government actions).
We shouldn’t assume that restricting labor supply automatically produces net productivity gains without reckoning with the unseen opportunity costs Bastiat warned us about.
If the argument is that we prefer a tighter labor market for distributional reasons — higher wages for incumbents, faster capital deepening — that’s a legitimate policy preference. But that’s different from claiming that lawful labor market participation is itself corruption or that increased labor supply is inherently a distortion.
I agree with you that rule of law matters. If laws are selectively enforced, institutional trust erodes. That’s real. I’m not arguing against national borders or against having immigration limits. Every country sets entry rules, and there are real benefits to expanding those (eg, how we have immigration freely between all 50 states). My concern is narrower: are we being precise about what counts as economic distortion versus what counts as a political preference about enforcement levels?
When we talk about corruption, are we distinguishing clearly between undocumented labor operating in a shadow market and workers participating in a lawful visa program? If a visa program has weak portability or creates vulnerability, that’s a design flaw worth fixing. But is participation in a legal work program itself corrupt? Or is the corruption located in how we structure and enforce it?
On the economic side, I keep coming back to the same question. If tightening labor supply forces investment in automation, how are we accounting for the unseen tradeoffs? As Frédéric Bastiat warned, it’s easy to see the visible shift — robots installed, wages rising for remaining workers — but harder to see what that capital would have done otherwise. Is it obvious that every dollar redirected into mechanization represents net productivity gain, rather than reallocation under constraint?
When we say things are “falling apart,” I wonder what specific causal chain we’re identifying. Are housing shortages primarily caused by immigration — or by zoning rules and permitting constraints that suppress supply? Are low labor-force participation rates driven by migrant competition — or by benefit structures and regulatory barriers that discourage work? Is crime a function of immigration levels — or of enforcement choices and prosecutorial priorities? These are empirical questions, and it seems risky to collapse them into a single narrative about labor supply.
You’re right that we shouldn’t cherry-pick arguments. But that cuts both ways. If we credit immigration with every downstream institutional frustration, aren’t we potentially overlooking the policy choices that more directly explain those outcomes? Expanding labor supply changes wage equilibria. Restricting it changes them in the opposite direction. That’s a tradeoff- nothing more or less.
I would suggest a more honest framing is this: what level and structure of immigration best balances wage growth, capital formation, institutional integrity, and rule-of-law credibility? That’s a difficult policy question, and one the article provides precious little empirical support. I just don’t think it helps to describe labor abundance itself as a distortion or to assume that broader social strain necessarily originates there.
Did Mr. Cass write a book with a title like "The American Worker"?
It's lovely that Oren thinks DonOrenomics is unspooling so successfully. Joepa used to assure us of the same.
Perhaps someone should alert his fellow Americans, who rate DonOrenomics as exceedingly unpopular. Their stewardship ranks the same as Joepa's. Or worse. I love irony.
I guess life's pretty good for the Epstein class. For the working class stiffs, not so much.
I’m always impressed by posters like Karl whose profiles show no subscriptions, no reads and no likes. Just an empty profile. Like they are bots or trolls or something.
Hey Art, good to hear from you. No bot:) Just an ex-Republican trying to learn what the "new" right stands for beyond grievance and devotion to one politician. It's a riddle.