Apparently cancel culture is still alive and well in the Democratic Party as they throw Rep. Haley Stevens off the island for daring to call out El-Sayed's continuing antisemitism. Apparently you can be any kind of bigot so long as you're antiTrump.
R's like Vance are rightly criticized for refusing to disavow Tucker Carlson and the like. D's from Schumer to AOC have shown they are no more principled.
2
2
393
fractalsofchange.substack.co…
Ensuring Domestic Fuel Supplies Will Lower Prices Throughout the Economy
Higher US prices are unfair, economically destructive, and incent Iran to continue the war
The problem
Although the US exports more oil and oil products than it consumes, since the war with Iran began the average US retail price for gasoline has increased 45% and for diesel 75%. Since diesel is essential for farming and manufacturing as well as for transportation, expensive diesel is pushing up prices for almost everything else. The price of heating oil, which is chemically the same as diesel, will be much higher this winter than last. Americans are hurt by rising prices and angry because they are caused by an unpopular war and come amid record reported profits for oil companies. Iran, seeing American resentment of the war, is refusing to settle even though the US blockade of Iranian exports and sanctions are crushing their already damaged economy.
The causes
Even though only a small percentage of US fuel is imported, prices have gone up here because the parts of the world which were supplied through Hormuz have attracted rising US exports at the expense of US domestic supply. Although US production of crude oil has increased from 13.5 million barrels per day (mbpd) prewar to 13.8 mbpd now, exports have increased from 4.0 to 5.2 mbpd leaving us with a net 0.9 mbpd less crude oil for our refineries. Reflecting decreased supply and competition with foreign purchasers, the price of domestic crude oil has gone up from about $65/barrel to a range of between $95 and $102, a 45% to 50% increase. The price of gasoline, which is mainly determined by the cost of the crude oil used to make it, has gone up accordingly. Prices at the pump have gone up more on the East and West Coast than in the rest of the country, because these regions lack refineries and do import some of their gasoline.
The price of diesel has gone up even more dramatically than gasoline because that price depends on refinery capacity as well as the cost of crude. A lot of refined product used to come through Hormuz. Exacerbating the problem, refineries have been damaged and gone offline in both the Middle East and Russia because of separate wars in these regions. We have increased our production of diesel in the US from 5.0 to 5.1 mbpd but have increased exports from 1.1 to 1.6 mbpd cutting domestic supply by 0.4 mbpd and putting a crushing burden on our agricultural and manufacturing sectors as well as driving up the cost of transportation.
The raw numbers above come from U.S. Energy Information Administration (EIA), Petroleum Supply Monthly / Weekly Petroleum Status Report, “Distillate Fuel Oil Supply and Disposition.” [eia.gov/petroleum/supply/mon…]
The cure
The US should require oil exporters and refiners to maintain their prewar baseline of domestic sales for crude oil and diesel on a rolling basis, allowing unrestricted exports of any volume produced in excess of that obligation. Note that this will still leave these companies with more product to export than they had before the war and they will be able to sell that at lofty international prices. Restriction of exports to return the domestic supply to its prewar levels means that oil companies will compete with each other as they were doing before the war to sell those quantities domestically; and prices in the vast middle of the country, which is served with domestic fuel, will return to near prewar levels. Not only will the price of diesel go down for much of the country, so will the prices of heating oil, gasoline, transportation, food, and manufactured goods for the whole country.
Prices will still be somewhat elevated in the Northeast, which has resisted both new local refineries and increased pipeline capacity to deliver energy from other parts of the US and so actually imports some of its fuel supply. Prices will remain even more elevated on the West Coast, as they were before the war, because, since there aren’t pipelines over the Rockies and refineries there have been pushed into shutdown by hostile regulation, that region imports fuel, some of it even through Hormuz.
US agricultural and manufactured products will become more competitive internationally as the cost of the fuel they depend on is increasingly less than what competitors pay. Export restrictions are a much better tool for increasing competitiveness than tariffs because tariffs raise domestic prices and export controls decrease them. I’m not recommending perpetual commitment to controls; they should be regarded as an emergency measure. However, there has been significant damage to refineries worldwide and it’s not clear that the prewar Pax Americana, which served to protect shipping freedom, will be restored.
Are export controls practical?
Western Australia is a working example (see industry.gov.au/news/austral…) of a domestic gas reservation system like the one I’ve proposed here. Their system applies to natural gas and has been in effect since 2006. Since producers can export whatever they can produce in excess of the domestic reservation, they have economic incentives to increase production even though domestic prices are below world prices. The system has resulted both in low natural gas and electricity rates in Western Australia and the region being a major supplier of LNG to Asia.
The US already administers a scheme at least this complex to force refiners to blend a fixed amount of ethanol into their refined products. Part of what makes that workable is that refiners can buy and sell credits among themselves, and so those best able to do blending do more and others do less but the cost impact is shared. BTW, eliminating the forced blending of ethanol would lower overall costs for American consumers but that’s another topic for another day.
What’s wrong with export controls?
Oil refiners and exporters would prefer to sell all their product at high international prices, even the part consumed domestically. They are correct when they say a total export ban would be counter-productive. We don’t have enough demand domestically to use all the oil and refined products we produce. We’d have to shut down refineries and wouldn’t see much drilling either if we ban all exports – and we’d lose lots of revenue. They also argue that it is logistically impossible to replace all the product which was being imported on the Coasts from the Gulf prewar because of pipeline and shipping limitations. However, if we just restore domestic supply to what it was before the war, we will be doing logistically exactly what we were doing six months ago. No question of feasibility. Just lower (but still very high) profits for oil companies. There is no logistical argument against export restrictions which maintain domestic supply.
Export controls are contrary to unrestricted free trade and maximum globalization. I was an over-enthusiastic proponent of both globalization and totally unregulated markets. We’ve seen that globalization taken to its extreme led both to fragile supply chains and domestic polarization as the benefits of cheaper foreign supply went to the enrichment of some and the impoverishment of too many Americans. We should not make the same mistake during a global energy shock which does not need to have a major effect on US prices. We should not allow all the domestic profit from higher prices to flow to relatively few Americans while everyone else pays more. Oil companies and oil workers will still get a huge benefit by selling our surplus internationally at much higher prices, even if they cannot divert our baseline supply. They will still have an incentive to drill and build and run refineries since all increased production can be exported.
Government regulation in general has a checkered history and is often subverted to favor a politically well-placed few. That is a real danger with export controls. The temptation to swap loopholes and exemptions for “campaign contributions” will be enormous. For that reason alone, the controls must sunset within a couple of years and sooner if pressure on domestic supply ceases.
Foreign prices will go up if the US restricts exports but not as drastically as some “experts” predict. We do not reduce the amount of fuel available to the world by reducing exports; we just reallocate it. Some argue that the US will no longer be regarded as a reliable supplier. That’s irrelevant since we are clearly a more reliable supplier than the Mideast and will be for a long time to come. Hormuz may reopen but will still be vulnerable as is the entrance to the Red Sea.
Former British Prime Minister Keir Starmer famously said that the war in Iran is “not our war.” However, it’s Britain’s energy supply which comes through Hormuz, not ours. It’s the UK which decided to stop drilling in the North Sea as existing wells were exhausted. There is no reason American consumers should pay both for the fleet to keep Hormuz open and for supplying Europe with replacement energy.
Iran counts on domestic opposition to the war in the US based largely on high energy prices. An effective method for lowering the pain Americans suffer may result in Iran becoming more reasonable. We all win in that case, especially Iranian citizens.
The oil companies could spare themselves and us regulation if, instead of opposing all efforts to assure domestic supply, they proposed a voluntary plan to ensure that the supply stays at prewar baseline or better. They would need some antitrust immunity to put such a plan in place but could probably get it. Although they do have a responsibility to maximize return for their shareholders, they will be better able to do that without formal regulation. If they don’t act voluntarily, there should be both executive and legislative action with sunset provisions to protect domestic fuel supply and, indirectly, prices throughout the economy.
See also: fractalsofchange.substack.co…
203
Q: What’s Scarier Than a Swarm of AI Agents Trying to Take Over the Internet?
A: A Cartel of Tech CEOs Trying to Take Over AI.
Background
Anthropic CEO Dario Amodei wrote [darioamodei.com/post/we-must…]:
"I believe that AI could cure most major diseases in the next 5–10 years, greatly accelerate economic growth rates, create a world of abundance and empowerment, and usher in a renaissance of democracy and freedom…
"But like many technologies before it, AI brings risks, and because it is such a powerful technology, these risks are serious…
"We must slow the pace at which we improve the capabilities of AI models."
Amodei goes on to propose:
1. Embedded Evaluators who work in the AI firms and have employee-like access to what is being done there. He said Anthropic is committed to this step and Sam Altman quickly agreed OpenAI would do the same.
2. Democratic Coordination. “Frontier AI companies within democratic countries coordinate to establish common safety standards as well as limits on the rate of unchecked AI progress. Some forms of coordination that would be impactful for pacing are legally challenging, and will require government support.” By government support, he means loosening antitrust law which is specifically meant to exclude this type of anticompetitive behavior.
3. Global Coordination. “The US and other democratic governments attempt to coordinate with authoritarian governments, to the extent this is possible, while taking seriously the challenges of verifying compliance.”
IMO
As a geezer, I would prefer that we not slow the pace of curing most major diseases.
Even the seemingly harmless suggestion of neutral embedded evaluators (as if there were such a thing) seated in the AI labs is a step towards strangling competition. Costs the big guys nothing and might even shield them from some liability. An impossible burden for a startup which might someday challenge the current leaders. If I start a one-man AI company (I would if I were younger), do I need to pay a couple of “observers” to sit in my garage and use my precious computer time? But, if Anthropic, OpenAI, and xAI want observers in their labs or want any other kind of help with their own safety responsibilities, more power to them. Just don’t let them inflict observers on rivals.
The demand that antitrust law be waived so that the companies can observe each other’s product development directly defeats the purpose of that law which makes it a criminal offense for companies to share information to avoid competing. It will let the big guys slow down but, they hope, retain their cartelized lead over new entrants who won’t get the benefit of the information which is being shared. Elon Musk, whose company is NOT in the lead in AI, quickly agreed that they ought to look at each other’s models every couple of weeks for safety’s sake. He had not offered similar access to SpaceX or Tesla’s robot technology (a different form of AI) even though rockets, cars, and trucks loaded with fuel can be weaponized as planes were on 9/11. No way we should waive antitrust laws to let Musk, Amodei, and Altman form a cartel. Even if they don’t think so, they have plenty of market power already.
The proposal for global coordination on “safe” development of AI is the escape hatch for these companies. A nuclear test ban treaty could work for a while because nuclear testing is detectable with a seismograph. Even if China agreed to such an arrangement (they won’t), how would we verify or, from their PoV, how could they verify we aren’t cheating. Is Iran going to agree to a verifiable slowdown? North Korea? What about Russia? Or Ukraine which has proved pretty adept at weapon development but can’t afford to fall behind Russia? Amodei knows damned well this isn’t going to happen. He wrote:
"A full pacing, or even “pause”, in which participating governments agree to substantially limit the overall rate of AI development. I support floating this, but I think it is unlikely to actually happen any time soon: defecting from such an agreement by evading monitoring could radically shift the balance of global power, so I expect the incentives to do so to be enormous and the level of confidence we would need in verification to be very high."
The world isn’t going to be saved by unilateral AI disarmament. It will become an even more dangerous place as the bad guys move ahead with both offensive and defensive applications of AI, which often can’t be separated, while we pledge ourselves to a suicidally slower pace.
Why a Proposal Which Clearly Can’t Work?
I’m certainly not privy to the motivation or long run game plan of the tech titans, but here’s one guess. They are currently facing two dangers. First a Bernie Sanders – Steve Bannon ban on the work they are doing and/or nationalization of AI companies. Terrible ideas but there is some danger of them being at least partially implemented.
Second danger is purely economic. Until recently they were raising money at huge valuations based on lofty expectations and localities were tripping over each other in a rush to offer subsidies and tax breaks for data center construction. In other words, they were getting a lot of free money. But, as the rush to build data centers progressed, localities have begun to push back. Not only are they denying subsidies; they are demanding that builders of new data centers pay the full cost of infrastructure they will require and maybe then some (see The Art of the Data Center Deal [fractalsofchange.substack.co…]).
This increase in costs as well as the difficulty the large AI companies will have ever getting a return on the huge investments they must make to maintain a competitive lead are about to become very clear in the documentation Anthropic and OpenAI must issue in preparation for going public. In other words, the tinkle of free money is becoming a deafening silence. Making things even worse is competition from open-source AI models, which people can run on their own computers, and low-cost AI models from China. OpenAI, Anthropic, and SpaceX can’t charge enough for their AI models to make them profitable because of low-cost competition and they can’t stop spending because they are in competition with each other.
Short-term they want a truce so they won’t have to build out as quickly and they want to fend off proposals which would simply end their businesses. They also want to make it regulatorily impossible for domestic competitors to emerge. But, as they slow down, open source and Chinese models will start to overtake them. If the US tech giants have already established that fast AI development is bad, they will quickly ask for regulation against open source and foreign competitors and they will probably get that regulation. We already don’t allow cheap Chinese electric cars into the country to compete with Tesla. Same thing, right?
But then, as they know will happen, there won’t be verifiable global non-proliferation. The US will begin to fall behind. That can’t be allowed to happen so now they will need government help to advance the cartel’s work on AI domestically. Federal mandates that data centers be allowed. Subsidies. Grants. Bailouts.
This may be good strategy for the tech companies. It’s a terrible strategy for our country.
What Should be Done about AI Dangers?
AI is a tool which, like a knife or a crowbar, can be and is used for both good and evil purposes. The same algorithms which can find new vaccines can design new dangerous pathogens. The ability to detect vulnerabilities in systems will be used both to strengthen systems and to attack them. If we don’t build the world’s most capable AI, we will be at the mercy of whoever does. Musk, Amodei, and Altman like to talk about “alignment”, making sure that AI systems have the values that we want them to have so they don’t do things we don’t want them to do. Bull! Tools can have safety features but they don’t have values. Bad people will use tools to do bad things. Defunding AI development is as poor an idea as defunding the police because some police are abusive.
Making the world safer requires both that we build the best AI tools humans (now aided by AI) can build and that we do what we can to to ensure that good uses outweigh and sometimes counteract the bad. Even before AI, it was possible to make bioweapons. Regardless of the actual origin of the COVID virus, there is little disagreement that it could have been designed in a lab even in those days just before generally available AI. Computer systems and networks have been hacked from the time they were invented. Our power and other utility grids have been dangerously fragile for years. Our government systems are porous. One positive action that AI giants can take is to fund a White Hat Institute (WHI) and provide it with the latest models prerelease so that the WHI can use these models both to build stronger systems and to find where existing systems need to be patched. This would be something like an Underwriters Lab but its purpose is to strengthen defenses rather than certify products. WHI’s purpose is not to “pace” development but to speed benefits. It would be appropriate for government to participate in the WHI and to help with its security and benefit from its learnings. Workers in the WHI may well require security clearances.
Another step government can and should take is to outlaw the payment of ransom to hackers by corporations and institutions like hospitals. Much of the motivation for hacking and much of the funding for it even when it is done by governments comes from ransom payments. If such payments are impossible for corporate or public institutions to make, the motivation for many attacks and the funding to develop attack weapons is diminished.
The dangers of AI are intensifications of threats we faced before AI. The potential benefits of AI are huge and literally can’t wait. Unilateral and unverifiable “pacing” (aka slowing down) of AI development by allowing
2
1
286
Q: What’s Scarier Than a Swarm of AI Agents Trying to Take Over the Internet?
A: A Cartel of Tech CEOs Trying to Take Over AI.
1
1
44
There is no way Elon and the others should be free from antitrust constraints. That would be more dangerous than AGI. Just put resources into the use of AI against biologic and computer viruses. Use AI to make systems antifragile. The holes which AI can exploit in systems already exist and are already exploited by human wih and without AI.
1
79
hard to know whether to be more appalled by the arrogance of Amodei, Musk, and Altman thinking they're a collective Prometheus deciding when to give fire to mere humans or the nakedly self-serving proposals to allow them to conspire without antitrust constraints while strangling new competitors in red tape.
1
170
fractalsofchange.substack.co…
Just when it seemed that everyone was against data centers – even the governors of red states – trade unions have come out strongly in favor of building them and against politicians who stand in the way. The Wall Street Journal reports[wsj.com/economy/jobs/blue-co…]:
… the Steamfitters UA Local 602, whose members install mechanical piping systems in Virginia, Maryland and Washington, D.C., wrote that it is drawing a “clear line” and won’t support politicians who oppose the facilities. “This is an existential moment for Local 602,” it said.
… “You’re not a friend if you’re taking away great career opportunities,” said Don Slaiman, political coordinator of IBEW Local 26, a union representing electricians in Washington, Maryland and Virginia. Many of the jobs pay good wages with benefits, he said. “This is a once-in-a-generation opportunity to really get in the upper-middle class .”
The opportunity is for the largely non-college-educated tradespeople who actually make things. Management guru Peter Drucker divided [archive.org/details/practice… workforce into builders, sellers, and measurers. Measurers include middle managers, operations staff, finance analysts, and marketing coordinators. AI can do much of the work of this latter group, according to Cloudflare CEO Matthew Prince explaining[thenextweb.com/news/cloudfla…] a layoff of 1100 people in May of this year. The measurers often focused on reducing the cost of builders through automation, outsourcing, or “efficiencies”. White collar measurers received an increasing share of corporate revenue as builders were eliminated. Inside the buildings the carpenters, steel workers, and masons are erecting, computers powered by the electricity the electricians are installing will do many of the jobs white collar measurers used to do. The builders are reducing the number and cost of measurers; the builders are being paid more and getting a bigger share of the corporate revenue. Turnabout is fair play!
Outside of work we have shared our earnings with what “The 2028 Global Intelligence Crisis” [citriniresearch.com/p/2028gi… Citrini Research calls a “giant rent-extraction layer on top of human limitations”. In this layer are human agents whose pitch is “I will navigate complexity that you find tedious”. The layer includes insurance agents, real estate agents, tax preparers, many lawyers, investment advisers, travel agents, even personal shoppers and nutrition counselors etc. But the AI-powered agents which run in the new computer centers or even on your own computer can navigate much of this complexity at much lower cost. Whoops! There go a bunch more white-collar jobs.
This stunning reversal of the traditional view that the path to upward mobility includes a college degree and a desk job is a huge blow to those who financed an exorbitant education only to find that they should have and could have apprenticed as welders. Soon the workers will be paid more than the people whose offices they used to build. It’s simple supply and demand. There are more qualified office workers than there is demand for them. There are fewer qualified builders than needed. Electricians will buy fancier houses than insurance agents. The people on the factory floor (there still are many despite generations of automation) will be paid more, at least in the aggregate, than those remaining in the offices. Nurses (who are “builders” in Drucker’s definition) will earn more than hospital administrators.
The political response
Both major political parties trumpet their support of “working American families”. Democratic Socialists like Bernie Sanders, almost none of whom are from a builder background, say they are especially devoted to workers. Apparently by “workers” they all mean the college-educated measurers whose career paths are being truncated by blue- and pink-collared builders. Bernie Sanders was one of the first major politicians to come out against data centers despite the jobs they create. OK, that’s not surprising. Data centers are financed by “cawperations” and we know what Bernie thinks of them. They are also supposed to make a profit, anathema to Democratic Socialists. Data center building is also powering our current good economy.
Much of both the traditional press and the blabbering blogosphere picked up the leftwing characterization of data centers as taxpayer-subsidized, pollution-spewing, water-guzzling, electricity-hogging, landscape-desecrating, noisy engines of mass job destruction. This is something like the hatchet job on the reputation of nuclear power which left us with an unnecessary amount of fossil-fueled electricity generation. The doyens of AI didn’t do themselves or us any favor by seeking public subsidies, making secret deals with public bodies, and boasting of the jobs they would make unnecessary. Once the public perception had been sufficiently poisoned, the polling began in earnest “would you like to have a data center next door?” Guess what most people answered. In fact most of us don’t want anything built in our own backyard.
The polls on data centers have been grim. Politicians deserted in droves. Among the first was Governor Hochul (D) of New York with a moratorium which prevents the energy-rich and job-poor areas near her native Buffalo from permitting data centers. Next the more centrist Democratic governors like Pritzker of Illinois and Shapiro of Pennsylvania, which have been courting and subsidizing development of data centers for years, declared their own moratoria. Even some Republicans couldn’t stand against the polls: Abbot of Texas and DeSantis of Florida jumped on the bandwagon. Finally bipartisanship, even if in a bad cause!
Traditionally Democrats have had the important support of labor leaders. The members of government and teachers’ unions have been especially enthusiastic and effective parts of the Democratic base. These white-collar unions can be counted on both to continue to support Democrats and to oppose the automation of white-collar jobs, although actual classroom teachers should welcome an opportunity to divert funds from the until-now mushrooming cohort of administrators back into the classroom. The members of trade unions have long been abandoning their leaders’ devotion to Democrats for both social and economic reasons and have trended Republican. They’ve been the swing vote in much of the industrial Midwest. Looping back to the beginning of this post, it looks like the leaders of the trade unions now may be following their workers’ desire to take wealth share from the office workers. Can this trend slow the moratorium-madness? Is it the beginning of a realignment of our tired old parties? Don’t know but will speculate more.
A reasonable response
I’ll post more on data centers, our need for them, and how to address legitimate concerns about them. Meanwhile, as a teaser, here’s what Republican Governor DeWine of Ohio, who is term-limited so not running for reelection, has to say[wosu.org/2026-06-11/gov-dewi…]:
Look, data centers are important. I know it's been controversial. I think, here's my message to any community that is not [sic], is worried about data centers: cut a good deal. The deal is in your hands…
They’re making money. They gotta pay their own way. They should. But we can’t throw them completely out and say, ‘Oh no, we don’t – we want to close the walls of the state of Ohio, and we don’t want any data centers to come in,’ anymore than we can do that to any other business.
He also announced a pause in Ohio’s data center tax exemption and said that consumers should not bear the brunt of rising energy costs and that data centers are “going to be built somewhere” and that there's an economic advantage to them being built in Ohio.
Children Are Now a Poor Investment for Parents
But humanity still needs them.
fractalsofchange.substack.co…...
Used to be that people needed to provide for their old age. Even if they owned or had rights to a farm, it wouldn’t support them without someone to work it. Same for other productive assets like a forge. Those without assets hoped to be cared for as part of their children’s households and/or by their children’s earnings. The lower the survival rate of infants and children, the more pregnancies a woman needed to have for a good chance of support in her old age.
The rich didn’t need to have as many children as the poor. Their children had a better survival rate; and their assets, if necessary, could be used to pay for food and people to care for them. For most of history, as survival rates improved, women had fewer children but always enough children to keep the population as a whole slowly growing.
Even today very poor countries have higher fertility rates than better off ones. Until about 1977,as shown in the chart (teapawt.net/JesusFerna7026/status/…) above by Jesús Fernández-Villaverde, the OECD (most developed countries) tended to have lower fertility rates the richer they were. But after 1977 the correlation flipped: richer countries tend to be more fertile. Note that the total fertility rate (TFR, the average number of live births per woman over her lifetime) of OECD countries was 1.48 as of 2024, well below the 2.1 needed to maintain the population level without migration.
In his paper “Terra Incognita” (sas.upenn.edu/~jesusfv/Demog…) written with Patrick Norrick, Fernández-Villaverde says:
""As of 2026, humanity is likely below replacement fertility. That has never happened before, not in wars or pandemics. But the real surprise is that the fall has been concentrated in low- and middle-income countries and among poorer and less educated women…
"Falling below replacement follows from a fast and accelerating drop in fertility worldwide. Demographers and economists have long expected fertility to fall, but its decline over the last decade has been unusually steep, above all in countries where no one expected such a fast decline. Mexico’s total fertility rate (TFR) in 2025, 1.51, for instance, is likely to be below that of non-Hispanic whites in the U.S., 1.57, despite a GDP per capita at purchasing power parity (PPP) of only around 28% of the U.S. level. Even more striking is Tunisia, whose TFR in 2024 stood at 1.45 with a GDP per capita at PPP of 18% of the U.S. level.
"Mexico and Tunisia are not anomalies. In 2025, Thailand was at 0.87, Colombia at 1.01, Iran at 1.35, and Brazil at 1.52, all at a fraction of U.S. per capita income. . .
"This inversion of the historical relationship between income and fertility across countries also seems to be happening within countries. The negative relationship between household income and fertility, one of the oldest regularities in demography … flattened in the U.S. between 1980 and 2010, a phenomenon that seems to be spreading to other countries (a similar statement holds for the relation between fertility and education). Even more surprisingly… the cross-country correlation between the TFR and the female labor force participation rate has also become positive in the OECD data.”
Why, for the first time in known history, are the poor having even fewer children than the rich? Why are working women having more children than those who don’t work outside the home? After extensive statistical analysis, the authors can’t find any single factor to account for the observed data. They speculate that both the drop in TFR worldwide and the reversal of the traditional relationship between affluence and number of children are caused by a complex they call “modernity.”
"Our conjecture is that a modern society makes the third child expensive and childlessness cheap, even when GDP per capita is only a fraction of the frontier (i.e., the U.S. level). Children are expensive because a child must be equipped to function in a world of formal knowledge and credentials, which takes years of costly investment. . . At the same time, children are no longer productive in the household, not even for smaller chores, having been replaced by “engines of liberation” such as the electric appliances now available even to poor households . . . . Modernity also imposes tight career schedules, set by large-scale organizations, that collide with the biological clock and with the kin networks that used to share the work of raising children. And modernity makes childlessness cheap, because an impersonal society imposes next to no penalty on those who have no children: pensions and health care do the work that sons and daughters once did. . . and lineage counts for little."
In other words, children, who once were an essential investment for which there was no substitute, are now a very poor investment to which there are many alternatives. Instead of paying for the college education of someone who may or may not put that education to productive use and may or may not support me in my old age, I can simply put the money into annuity. Similarly children take a lot of time; if I put that time into work, I’ll have more assets for a secure old age. In fact with Social Security and Medicare and the like, society is doing this saving for me (partly with my tax dollars); I can live better now on the money children would have cost me. (It’s important to note that the birth control pill makes family planning much more effective than it ever was and, along with modern equality, gives women much more control over the number of children they will bear the burden of raising.)
OK. But that still doesn’t explain why the affluent are now more fertile than the poor. There is a simple (although conjectural) explanation: many people like children; they want to have them. The children are a financial burden rather than an asset, but so is a yacht. Children are often fun. I’m speaking as a father and not an economist; but, in economic terms, children are a luxury good, something you pay for because you want it, not because you need it.
Back when we all believed that a Malthusian nightmare of population explosion would doom humanity, we might have just said: “Great news. Less children means less overcrowding, less anthropogenic global warming, less poverty. Only those who can afford children are having them.”
But there’s a problem!
We may not need children family by family, but we still need enough of them collectively to take care of us as we get old. Maybe robots will change our geriatric diapers and feed us; but we still have a problem. The collective pension schemes like Social Security and Medicare are Ponzi schemes; they count on payments by future workers to make payouts to the retired. Works fine so long as the number of current workers is much greater than the number of retired they support. Breaks catastrophically if there aren’t enough current workers. Even our private pension savings may be insufficient if there are not enough workers to invest in or if a worker shortage leads to rampant inflation. That’s where we’re headed!
Something must change. In fact, some things are changing but we’re not dealing with that change very well (see my post “Where Are We Going to Recede From?”). The paper I’ve been quoting from is called “Terra Incognita” because we literally have no experience in human history with a general decline in population; we’re in unknown territory. Norrick and Fernández-Villaverde conclude:
"All of this could change with AI or shifts in population composition. We will see. But on the current evidence, deep sub-replacement fertility is the ‘new new normal,’ and unless we find a way to reorganize our societies, we had better learn to handle it as best we can. We do not have a map. What we have birth numbers, and those are sobering. In these waters, the pretense of knowledge is the most dangerous temptation."
AI Notes:
From the authors of “Terra Incognita” : Generative AI was used for improving the readability of text, for helping with coding the statistical model, and for checking the final paper for technical accurary [sic] (Refine.ink).
From me: I used my custom GPT Citation Checker (chatgpt.com/.../6a872d9d-3f8…, ChatGPT based) to check the citations in “Terra Incognita” and they held up very well. I removed some of the citations for readability when I quote from the paper but, of course, they are available in the original. I then had a long “discussion” with ChatGPT about the paper which helped my understanding and gave me food for further thought. That conversation is at chatgpt.com/.../6a874adf-ed1… but it’s long and repetitive.
One of the claims in “Terra Incognita” that readers have found most interesting is that income per capita and fertility are now positively correlated across OECD countries: the richer a country is, the higher its fertility.
To document this claim more carefully, I have completed the following exercise.
I take total fertility rates from the U.N. World Population Prospects 2024 and real GDP per capita from the new Penn World Table 11.0 (expenditure side, chained PPPs, released in October 2025 and now running to 2023). For each year from 1954 to 2022, I compute the rank correlation across countries between fertility and real GDP per capita and plot the results. I use 30 of the 38 countries that are OECD members today (regardless of when they joined). I drop the other eight because they have data gaps, as former socialist economies did not compute GDP using the standard methodology (Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia).
In the mid-1950s, the correlation among the 30 countries was about -0.32, and it deepened through the 1960s and most of the 1970s, until a minimum of -0.78 in 1977. This is the world most people have in mind: the opportunity cost of children rises with income and, the richer you are, the fewer children you have. Gary Becker and dozens of economists wrote papers trying to understand this pattern.
But then, in 1977, well before anyone had heard about smartphones, social media, or some of the other “usual suspects” behind the recent drop in fertility, something changed: the rank correlation started climbing in a steady way, becoming positive in the second half of the 2010s. Right now, being richer means having higher fertility rates. And, as far as I can tell, there is no indication this trend is slowing down.
So, if you have a favorite explanation for the evolution of fertility over the last few decades, ask yourself: how can my explanation account not only for the drop in fertility but also for the rank correlation switching directions in 1977?
A couple of warnings. First, I use the rank correlation because it is immune to outliers. Luxembourg and Ireland report GDP per capita figures that nobody should take at face value.
Second, I am using the U.N. World Population Prospects 2024, which I have criticized for overestimating the fertility rates of the poorest OECD members (Mexico, Colombia, Costa Rica, Türkiye). If you use the fertility rates reported by their own statistical agencies, the rank correlation is even more positive. But I don’t want anyone accusing me of fiddling with the data.
Finally, credit where it is due: I first learned about this reversal from Doepke, Hannusch, Kindermann, and Tertilt, “The Economics of Fertility: A New Era,” in Lundberg and Voena (eds.), Handbook of the Economics of the Family, vol. 1, North-Holland, 2023, pp. 151-254. Anyone who wants to go deeper should start there.
Comments always welcome!
1
459
In Terra Incognita (sas.upenn.edu/~jesusfv/Demog…) @JesusFerna7026 shows how children have gone from becoming an essential investment in parents' future to a poor use of time and money. Used to be the poor needed to invest in children (and needed to have more children because of low survival rates), more than the rich so were more fertile. Now children are a luxury good so those who can afford them have more of them than those who can't as his chart shows.
One of the claims in “Terra Incognita” that readers have found most interesting is that income per capita and fertility are now positively correlated across OECD countries: the richer a country is, the higher its fertility.
To document this claim more carefully, I have completed the following exercise.
I take total fertility rates from the U.N. World Population Prospects 2024 and real GDP per capita from the new Penn World Table 11.0 (expenditure side, chained PPPs, released in October 2025 and now running to 2023). For each year from 1954 to 2022, I compute the rank correlation across countries between fertility and real GDP per capita and plot the results. I use 30 of the 38 countries that are OECD members today (regardless of when they joined). I drop the other eight because they have data gaps, as former socialist economies did not compute GDP using the standard methodology (Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia).
In the mid-1950s, the correlation among the 30 countries was about -0.32, and it deepened through the 1960s and most of the 1970s, until a minimum of -0.78 in 1977. This is the world most people have in mind: the opportunity cost of children rises with income and, the richer you are, the fewer children you have. Gary Becker and dozens of economists wrote papers trying to understand this pattern.
But then, in 1977, well before anyone had heard about smartphones, social media, or some of the other “usual suspects” behind the recent drop in fertility, something changed: the rank correlation started climbing in a steady way, becoming positive in the second half of the 2010s. Right now, being richer means having higher fertility rates. And, as far as I can tell, there is no indication this trend is slowing down.
So, if you have a favorite explanation for the evolution of fertility over the last few decades, ask yourself: how can my explanation account not only for the drop in fertility but also for the rank correlation switching directions in 1977?
A couple of warnings. First, I use the rank correlation because it is immune to outliers. Luxembourg and Ireland report GDP per capita figures that nobody should take at face value.
Second, I am using the U.N. World Population Prospects 2024, which I have criticized for overestimating the fertility rates of the poorest OECD members (Mexico, Colombia, Costa Rica, Türkiye). If you use the fertility rates reported by their own statistical agencies, the rank correlation is even more positive. But I don’t want anyone accusing me of fiddling with the data.
Finally, credit where it is due: I first learned about this reversal from Doepke, Hannusch, Kindermann, and Tertilt, “The Economics of Fertility: A New Era,” in Lundberg and Voena (eds.), Handbook of the Economics of the Family, vol. 1, North-Holland, 2023, pp. 151-254. Anyone who wants to go deeper should start there.
Comments always welcome!
2
92
Where Are We Going to Recede From?
The question used to be “where will we grow?”
fractalsofchange.substack.co…
Nationwide our fertility rate is below replacement level. Without massive immigration, our population will shrink. The number of young children is already falling in much of the country. Aging Vermont gives us a preview of what a declining rather than expanding population means.
Almost 50 years ago Mary and I helped get a federal grant to build a new elementary school in Worcester, VT — a farm town evolving into a bedroom suburb of Montpelier. The town was growing with rapidly breeding young couples, including us. The old school was a wreck and too small. Today with the birth rate cratering and school-age population declining, the school is one of two in the district considered for closure. Of course, there is intense local opposition to closing it; but it probably should be shuttered for both educational and economic reasons. Closing it would definitely mean longer bus rides and make the town even less attractive to people with or planning children.
Similarly the rural hospitals in Vermont and elsewhere are under intense financial pressure. The first service to go is usually the birthing center for good demographic reasons. Copley Hospital in Morrisville, VT, closed theirs after 90 years of service because of financial losses. Not enough babies. That’s little consolation for those who are still having babies in the area and now have to drive considerably farther (Copley can still do emergency deliveries). The Board of Brattleboro (Vermont) Memorial Hospital approved a plan to wind down obstetric services to help alleviate the hospital’s huge operating losses. So far attempts to get extra funding from the state and/or town have failed. The nearest alternative services are out-of-state (although not very far away).
If one family moves out into the woods, are we obligated to build a school and a hospital near them? Obviously not. What if only a few families remain using a school and hospital that were built for local population high tide? Are we obligated to fund a tiny school and hospital forever? Can a tiny school or hospital be a good one? I’d say “no” to both – but we do have to think about both ways that towns can avoid terminal shrinkage and what happens to the families who are stranded when the town infrastructure recedes around them.
How does a town escape terminal shrinkage?
A shrinking population means fewer customers for stores, schools, hospitals, etc. in a town. As these organizations become uneconomic, they reduce services or disappear completely. Population declines because people move out and new people don’t move in. The spiral continues down unless somehow interrupted.
To stop the spiral —duh—a town needs to attract new people and retain most of the people it has. One way to become attractive is to host new facilities that no other town currently wants: a data center, for example, a quarry, a factory, or even a prison. It’s ironic that towns with prisons resist their closure but towns without them resist their opening. Yeah, there’ll be local opposition to the new facility from “stakeholders” who don’t want the facility next door and want to maintain the “character” of their town. Unless the stakeholders are willing to fund the school and the hospital and keep stores open at a loss, they should be ignored.
A town can become competitive by making it cheap and easy to build housing. Get rid of all exclusionary zoning. Don’t insist that farmland be preserved for cows that are never coming back. Make permitting cheap and easy and predictable. Get the damn “stakeholders” out of the permitting process as well. Encourage (but don’t subsidize) developers. If money can be found, don’t use it to prop up dying institutions; build town infrastructure like sewer, water, and roads which encourage private investment.
Excellent schools are attractive but very hard to achieve in small towns. Some places in Vermont work successfully around that limitation by using their school funds to pay tuition (with some limits) for whatever other public school or approved independent school parents choose. Teachers’ unions hate this. It works. State law should encourage it everywhere for better education and small-town survival.
What happens to the people when a town shrinks below viability?
As population recedes, some people will be left stranded in towns that can no longer offer adequate services. This may just be an inconvenience if the needed services are close enough and the town is otherwise attractive. Personal transportation is about to get both easier and cheaper because of self-driving cars and electrification. Hormuz aside, the long-term trend in fossil fuel prices is down, not up. Every hybrid or full electric on the road makes gasoline cheaper for everyone else. Lower battery prices mean cheaper electrics; we will either make them in the US or import them from places like China which already make them very inexpensively. Better batteries also mean better local grids and feasible use of renewables even in remote places with poor grid access. Whether or not the fiber-at-any-cost crowd wants to admit it, Starlink means good communication is available almost everywhere cheaply and immediately. Look for Starlink and competitors to make universal cell service practical soon.
But we’ve learned from the time of COVID that remote learning is educationally inferior. We’ve also learned that socializing in schools is a crucial part of growing up. Kids need to be able to get to some good school. People need to be able to get to hospitals while they’re still alive. People need a place to work. Some towns will not be viable except possibly for vacation homes and retreats for hermits. The people in those towns will have to move in order to give their families a good life.
There have been ghost towns almost since America was first settled. In some places the soil ran out quickly. Mines were exhausted. Towns built for railroad workers weren’t needed anymore. The dust bowl happened. Although these closures caused real hardship for real people, the safety valve was mobility. Americans moved not only because of overcrowding but also because the towns they lived in were no longer viable. We’ve lost our mobility! According to the Census Bureau, the number of people who move each year in the US has halved since 1980 and is still declining. We must have mobility so that people can escape towns with disappearing infrastructure and lack of opportunity.
One reason people are stuck where they are is two-income households. It is easier to get one new job than two in a new place. The other two major reasons for declining mobility are housing costs in places they might want to move to and welfare incentives to stay put. Both are addressable.
Buying and renting are both too expensive in many places largely because of exclusionary zoning and excessive permitting costs. Places like Austin, TX —which is a magnet for job seekers—have shown that rental costs actually decline with increasing population so long as housing can readily be built. This is a positive feedback loop. Reasonably priced housing attracts workers who attract employers who attract more workers who populate the schools and hospitals and shop in the stores and attract more housing developers. The villains are again the “stakeholders”, the people who believe they are protecting their own property values by making sure no one else can build.
Stakeholders cannot be allowed either to keep towns in a death spiral or block new housing for those seeking opportunity. To some extent, government subsidies and mandates for “affordable” housing help drive up housing costs for everyone.
An incredible web of welfare benefits, most of which are administered locally, has both subsidized the cost of remaining jobless in place and also made it very difficult to move. Some people are too sick or in too much need of family assistance to move. Others, however, should move to places where there are jobs, good schools, and adequate health care. The goal of welfare for them should be to get them back as productive members of society SOMEWHERE in the US. We need them in the workforce. They need to be in the workforce. One-time relocation assistance is far preferable to perpetual dependence. But there must be someplace to move to.
We live in post-Malthusian times.
The problems of population shrinkage are different than those of constant growth, which we are used to thinking about. Growth may come again; meanwhile we must help people in towns stranded by the outgoing population tide.
🤖 Made with AI
2
156
permitting data centers can save towns stranded by the shrinking population tide which makes their schools, hospitals, and businesses unsustainable fractalsofchange.substack.co…
🤖 Made with AI
Opinion: California reaps the rewards of the AI industry while largely decrying the data centers that sustain it. In the Dakotas, however, hyperscale data centers are not an abstract concept. sfchronicle.com/opinion/arti…
1
52
Tom Evslin retweeted
Opinion: California reaps the rewards of the AI industry while largely decrying the data centers that sustain it. In the Dakotas, however, hyperscale data centers are not an abstract concept. sfchronicle.com/opinion/arti…
permitting data centers can save towns stranded by the shrinking population tide which makes their schools, hospitals, and businesses unsustainable fractalsofchange.substack.co…
🤖 Made with AI
Everyone needs to hear this story. A data center came to a dying North Dakota town. The result? It can finally afford to repave streets, build a public-safety complex, create walking trails, and update the public school.
How is that bad? The data center haters should explain.
45
The End of the AI Bubble Will Be Just the Beginning
Bankruptcies are fertile ground for future growth
@tylercowen posted “Nothing Can Stop the AI Revolution” on The Free Press (substack.com/home/post/p-209…) . He acknowledges that there will be bankruptcies along the way but says “…bankruptcy will likely be little more than a bump in the road.”
What Tyler misses is that the AI revolution will not just survive bankruptcies, it will be turbocharged by them.
Happened with railroads and data communication. An exciting new technology attracts massive investment. Stock prices run up nearly vertically and almost everyone with money or the ability to borrow money rushes in so as not to miss a sure thing. More infrastructure is built than there is demand for at early prices. Crash. Bankruptcy. But the assets - track, fiber, or datacenters full of Nvidia chips - still exist. They’re snapped up at pennies on the dollar.
The new owners of the infrastructure can afford to offer the service – transport, communication, or AI – at much lower prices. At the new price point, new applications are possible using the assets. Cheap transport both expanded the market for local commodities and made possible both concentrated and dispersed manufacturing. All of which led to another rail boom. And bust. And even greater growth at even lower prices.
After the over-built fiber was marked down, data communication was so cheap that voice traffic moved from dedicated networks to the internet and the web and all its applications became possible. Services like Facebook, like it or not, could be offered “free” supported by advertising.
Once many of the datacenter owners go broke – and perhaps even some of the power generators frantically building power for them, the new owners of these assets will be able to offer artificial intelligence – profitably - at a much lower cost than is possible today. Incredible new applications of AI – for better and for worse – will be developed.
What will these new applications be? I certainly don’t know. Just know that they will surely happen in the aftermath of another bubble or two bursting.
2
66














