Conversable Economist

About 10 years ago, the big topics that had economists all atwitter included the “secular stagnation” hypothesis put forward by Lawrence Summers and the rising-inequality hypothesis put forward by Thomas Piketty. Put the two theories together, and you have forecasts for a slow-growth, high-inequality future. The Review of Political Economy (38:3) has put together an eight-paper symposium on the topic: “What Have We Learned from Summers and Piketty Ten Years On?” There’s lots of thought-provoking stuff here, but I found myself especially drawn to an article by Steven Pressman, “Secular Stagnation After Piketty and Keynes” (pp. 961-979).

Pressman points out that the themes of secular stagnation and wealth inequality have been linked for a long time. The basic idea is that the wealthy have a lot of money that is being saved, rather than spent, and this lack of spending makes the economy grow more slowly. Phrased this way, there is an obvious two-birds-with-one-stone policy choice to fix both problems: tax the wealthy, and thus reduce inequality, and spend the money, thus pushing up aggregate demand and jolting the economy out of secular stagnation. As Pressman points out, advice along these lines goes back at least to the Francois Quesnay and the French physiocrats in the 18th century:

As far as I know, the first presentation of secular stagnation appears in static versions of the Tableau Économique (see Pressman 1994a). Quesnay was interested in economic reproduction and explained why the French economy languished while the British economy was growing rapidly and living standards were improving there substantially. Quesnay wanted to reverse the long-term stagnation of the French economy and to have France become more like Britain. His solution was for France to adopt a number of Physiocratic policy prescriptions and to end feudal restrictions on agricultural production. Some of his more noteworthy proposals were that taxes should fall on the class that consists (mainly) of wealthy landowners or proprietors (Pressman 1994b), encouraging consumption of goods produced by the productive agricultural sector, and discouraging savings and waste (Pressman 1994a).

Another linkage between the two ideas, going back in time, was that if a secular slowdown in growth was more-or-less inevitable–two centuries ago, a common idea was that rising output was going to run into diminishing marginal returns fairly soon–then the options were either mass starvation (Thomas Malthus) or widespread redistribution of income (John Stuart Mill).

The Great Depression brought the idea of long-term secular stagnation back to prominence; indeed, the term was coined in 1934 by Alvin Hansen. Paul Samuelson developed the idea of a “balanced budget multiplier” in 1948: tax the rich, who save a lot, have the government direct the money to the middle class and the poor who will spend more, and the result would be stronger economic growth without a need for budget deficits.

In short, there’s a long-standing current in economic and policy thought, manifesting itself in different ways at different times, that redistribution from those with high wealth or income to those with lower wealth or income would speed up the economy. A decade ago, Summers and Piketty were playing a role in resuscitating this long-standing theme. In this spirit, Pressman suggests that the modern policy along these lines would include higher marginal tax rates for those with the highest incomes and higher corporate tax rates. He’s not a fan of an annual wealth tax, pointing out that such taxes have been extraordinarily hard to implement in the past, but notes that an estate tax could function better as a form of a wealth tax.

Personally, I wouldn’t have a problem with moving the estate tax back to, say, the rates and rules tha applied in early 2000s. I’m less certain about higher marginal tax rates, but in 2026, the highest marginal tax rate kicks in at $768,700 for a married couple filing jointly. I’d be willing to try out a higher income tax bracket set well above that level. But I also think that the linkage from tax-the-rich, to more spending, to a permanently improved rate of economic growth makes the economic challenge facing us much too neat.

For example, the economy of China has for a half-century has famously high savings rates, more than double US levels. It has also had rising inequality. But the combination has been consistent with rapid economic growth. It used to be a standard recommendation that the US needed a higher rate of national savings, either from higher personal saving or lower government borrowing. Was that advice completely without merit?

As Pressman duly notes, the causes of a productivity slowdown in modern times are many and often self-reinforcing: slower population growth and an aging population, lower investment, the possibility that “most of the good ideas have already been discovered,” high US government debt, the shift to a service economy and intangible investment (areas where productivity growth may be slower), and others. I would add to this list other factor like the poor performance of the US education system for many children, problems of the US labor market in matching potential workers with jobs, underinvestment in building the nation’s reserach and development capacity, and other factors. Collecting higher taxes from those with high incomes and high wealth levels isn’t going to address those most of those issues in any direct way.

I’d also add that a widespread current concern seems to be that the US economy is about to enter a period of disruptively rapid economic growth, driven by new AI techologies. I think those worries about overly rapid AI-driven growth in the short- and medium-term are overstated. But if it happens, it would invert the earlier arguments that wealth inequality is associated with too-slow growth. Instead of redistribution to speed up the economy, we would be talking about redistribution in response to surging and unequal economic growth. I am congenitally wary about advice that remains the same, even when the situation shifts dramatically: for example, is it plausible that redistribution the answer to both slow growth and fast growth? Perhaps some level of redistribution is just a good thing for its own sake, rather than the answer to every question.

Looking at the benchmark interest rate for 30-year US Treasury debt, long-term interest rates have been rising. Any price change might happen for a number of different reasons: 1) expectations of future inflation are causing investors in Treasury debt to demand a higher rate; 2) as US government debt continues to climb, the perceived risk of this debt is rising; 3) the surge in productivity that will come from recent developments in information technology is pushing up demand for capital, and thus pushing up interest rates; 4) The higher intereset rates aren’t a US phenomonon, but rather a global one, and thus need a global explanation; and more. Hanno Lustig makes the case for a risk-based explanation in “America’s Risky Debt: What Markets See That Policymakers Don’t (Aspen Economic Strategy Group, August 2026, forthcoming The American Economy in a New Era, edited by Melissa S. Kearney and Luke Pardue).

Lustig lays out several pieces of evidence for a risk-based explanation. For example

If US Treasury bonds are perceived as safer than other assets, then the US government should be able to pay a lower interest rate than other borrowers. Up to about 2020, this pattern held true. But in recent years, investors seem to be viewing US Treasuries as very in risk to alternatives; indeed, the Wall Street Journal reported last fall that some big US companies like Microsoft and Johnson & Johnson were able to borrow long-term at lower rates than the US government.

Another risk-related pattern is called “flight to safety”: basically, when stock market prices fall or risk rises in some other way, investors head for the safe asset of US Treasury bonds. To put this another way, returns on stocks and bonds have a negative correlation. Again, this pattern held up to to about 2020. Now the correlation has flipped, and returns on stocks and US Treasury bonds are moving together — which is what you expect of two assets exposed to similar risks.

Managers of reserves at central banks aroun the world used to treat US Treasury debt as the primary saf asset: “Foreign reserve managers treated US Treasurys as the dominant safe asset, allocating more than 70 percent of allocated world foreign-exchange reserves to dollar-denominated assets. … At longer maturities, global investors now seem to prefer the safety of foreign G10 bonds.” (The G10 is shorthand for a group of 11 high-income countries–they decided not to change the name after Switzerland joined many years ago.)

Behind the scenes, the effects are apparent. For example, as the US Treasury has had to pay higher interest rates on long-term borrowing, it’s been moving to shorter-term borrowing. A problem with shorter-term borrowing, of course, it that you are planning to roll it over–that is, borrow the money again and again. In doing that, the US government as a borrower is not locking in long-term rates as often, but instead is more exposed to fluctuations in short-term interest rates. An increasing share of US Treasury debt is being purchased by hedge funds, who in turn are using that debt as basis for various strategies to make a buck, often with fairly short-term time horizons. These investors are not like, say, life insurance companies that have traditionally purchased long-term Treasury bonds as a way of making sure they could pay off long-term commitments to policyholders. If the hedge funds stop demanding as much Treasury debt, the interest rates on the debt will rise further.

As Lustig writes: “The marginal foreign holder of US Treasurys is no longer a central-bank reserve manager but a private, yield-sensitive investor: predominantly foreign banks, asset managers, and hedge funds … [T]hose Treasury investors with less price-sensitive demand have pulled back. The shortfall has been picked up by private leveraged investors whose demand is more elastic to yield and more fragile in stress.”

Lustig argues that the Federal Reserve has gradually, step by step, become entangled in keeping interest rates on Treasury borrowing lower than they would otherwise be. He argues that it’s time for the Fed to take a step back from the bond markets. If the US government is going to keep running enormous deficits, which means riskier borrowing, than the US government also face the higher interest rates that result. None of Lustig’s evidence suggest that a catastrophe or crash in US Treasury debt is just around the corner. But it does suggest that some yellow warning lights are flashing in world financial markets about the rapid build-up of US government debt.

The number of grains of sand on a beach is a classic example of an “infinite” number, in the sense that it would be impossible to count. But sand itself isn’t infinite. Demand for sand is way up, around the world, because it’s a key ingredient for some boom industries like concrete for construction and fracking for oil and gas drilling. But even lowly sand plays a role in many ecosystems. I’ve been commenting about the global market for sand for more than a decade nowg (for example, here, here, and here), but for those who would like an introduction or an update to the topic, the UN Environmental Program has published “Sand and Sustainability: An Essential Resource for Nature and Development” (April 2026). From the “Introduction”:

Sand, gravel, crushed rock and aggregates (hereinafter ‘sand resources’) are the most consumed solid materials on Earth (UNEP 2022). As fundamental ingredients in contemporary concrete and construction materials more broadly, sand has played — and continues to play — a central role in the built environment worldwide. Global demand for sand tripled between 2000 and 2020, fuelled by rapid population growth, urbanisation, and economic expansion (UNEP 2022). Since then, demand has remained stable but is projected to increase again. The global sand market was valued at USD 569.4 billion in 2024 and is projected to grow at around 3% annually, driven by urbanisation and infrastructure development (IMARC 2024) and the material demands of climate change adaptation and mitigation efforts.

While recent debates on raw materials have focused on critical minerals for the energy transition, sand plays an equally strategic, yet far less recognised role. In fact, the scale of sand use vastly exceeds that of critical minerals: around 50 billion tonnes of sand are extracted every year, compared to the projected annual extraction of 30 million tonnes of critical minerals by 2030 under a Net Zero Emissions by 2050 scenario (IEA 2023). Even for renewable energy infrastructure, sand and gravel constitute the largest volume of materials. For instance, they represent about 70% of the total volume needed for wind farms (Aska et al. 2025).

Despite the significant scale of demand and extraction, there are no exact global figures on how much sand exists, how much is extracted, where, or for which uses. … Even less discussed or understood is how sand underpins both human development and natural ecosystems. As highlighted above, sand, once extracted, is indispensable for housing and infrastructure, which form the foundation of economic activity and growth. Yet in nature, sand creates habitats for numerous species, shapes rivers and coasts, supports food and water security, and contributes to climate resilience (Torres et al. 2017). These ecological functions make sand essential for conserving biodiversity and maintaining the ecosystem services upon which humanity depends.

For an economically-minded reader, like me, the report suffers from a lack of even localized data on prices. Within the US economy, however, the US Geological Survey reports that prices for “Sand and Gravel (Construction)” rose from $10.52 per metric ton in 2021 to $14.50 per metric ton in 2025, as total production dropped modestly. The lack of data makes it hard to draw broad conclusions, but there are certainly localized examples where extensive sand extraction harmed ecosystems, including in ways that affected industries like tourism and fishing. The report concludes:

The sand crisis is no longer hypothetical. Globally, shortages are already halting major infrastructure projects. Demand for sand in the building sector alone could rise by 45 per cent by 2060 (Zhong et al. 2025). Yet unlike many environmental challenges where delayed action has led to escalating and irreversible costs, an opportunity for timely, coordinated intervention remains. Sand governance is ultimately a development choice.

The good folks at “Our World in Data” have been putting together a data series which shows monthly spending on the “on-site work to build data centers each month, including materials and construction labor. It excludes the cost of IT hardware like servers and storage, which can account for a large share of total investment.” Here’s the just-updated figure:

There’s an upward trend from about 2014, but the big rise kicks in around early 2023. At least as of the most recent data on this graph, for June 2026, there is no sign of the rise slowing down. Again, these are monthly figures, so on current trends, it’s plausible that just on-site construction spending for data centers will total $40 billion or more in 2026.

I’m of course aware that there is a degree of political pushback against building data centers, which the figure suggests isn’t having a dramatic effect, and which I think is largely misguided. Sure, there should be limits on where such centers can be built. But when I go on road trips around the country, I often find myself smiling when I drive into a modest-sized town, out in the countryside somewhere, and there is a sign and a turnoff for a hopefully named “industrial park”–without any actual firms or buildings visible. If the local community has already set the rules for industrial zoning, then data centers would seem to qualify. The on-site construction spending, as well as the longer-term running and maintenance of these sites, is largely blue-collar work. Yes, there are issues to work out concerning energy, water, noise and traffic. But some version of those issues will be true for pretty much any firm that wants to build in an industrially-zoned area.

Some of the opposition to data centers seems to arise because of a specific opposition to the new AI technologies., and yes, the new AI technologies are another reason for the recent surge of data center construction. But trying to limit AI in particular by hindering construction of data centers seems poorly targeted; for example, why not also try to limit AI by hindering US production of semiconductor chips, computers, smartphones, or other information technology products? Of course, if you just oppose all uses of information technology, it makes some sense. Bit If you have a smart-phone, or shop online, or work from home, or stream on your computer or television, or basically use the internet in any way, then you are using data centers. As the makers of the figure note, “Data centers power a wide range of online services beyond AI, such as streaming services and cloud storage.”

I’ll also add that AI as an industry may have the worst public relation campaign I’ve ever seen. As one of many examples, Genspark ran an ad during the Super Bowl that celebrated the idea that AI could do all your work–so that you could “take the day off.” If AI about taking everyone’s job, while producing endless online “AI slop” and deepfakes in text and video, while enabling hacking and simultanously making a few tech bros rich beyond the dreams of avarice, then there isn’t much to celebrate. But maybe, just maybe, the worst-possible scenarios are not all there is? Maybe, just maybe, there can be positive effects as well?

As one example (among many) of potential gains, Saloni Dattani recently published an essay titled “Every disease is a policy failure” (Works in Progress, August 11, 2026). The subtitle reads: “Few people know how much medicine has progressed in their lifetimes. Fewer know how much faster it could be progressing.” You can read about the gains over time in treating heart disease, diabetes, cancer, low-birthweight babies, and many others. If AI-assisted research has the potential to speed up health and medical innovation, so that it adds a decade or two to the average healthy human life, that possibility seems to me worth being open to a building boom in data centers.

When writing feels hard, perhaps it is because you are on the cusp of saying something significant or interesting? The thought is oddly comforting to me. Donald Murray originally described this connection in his essay “Making Meaning Clear: The Logic of Revision,” published in the Journal of Basic Writing in 1984. Murray wrote:

[I]n fact, syntax often breaks down just at the point where a new or significant meaning is beginning to break out of its shell. That meaning has an awkward and clumsy time of it, but if the writer listens carefully and nurtures the meaning, it may grow into significance. Or it may not. It may have to be put aside. But first it has to be understood before it can be rejected.

One concern about teaching writing in the age of AI is that generating a first draft that has surface-level plausibility and smoothness has now become very low-cost. As a result, for many practical real-world purposes, the task of collaborator with an AI-written first draft is to figure out how to edit the draft. But the AI tools are already good enough that they avoid obvious breakdowns in syntax. So how does the human AI-collaborator figure what kind of additional editing is needed?

My own (limited) experience in teaching writing is that when I offer copious comments, many students will respond to those comments in the most minimal way possible: that is, they alter the specific word or sentence I’ve pointed out, but otherwise leave the first draft untouched. They have not yet developed a sense that meaning can often be “awkward and clumsy,” as Murray puts it, and that you don’t fix awkward and clumsy by swapping out a few words. They don’t yet have a feel for the idea that good writing is more than avoiding errors in syntax.

Murray argues in his essay that teachers of writing should actually write in front of a class, to show students the series of erratic and continual adjustments that happen. In an AI age, it seems to me perhaps more useful for a teacher of writing to start from a text that has already been generated, whether by AI or a previous writer, and then to raise possible concerns and issues and discuss how one might address them. The challenge is to equip students with a set of questions that an editor would ask, so that they can see text as fluid.

In the past, teachers of writing have always encouraged students to revise their own work, and to say things like “writing is rewriting.” But in many classrooms, a draft that didn’t have errors in syntax was already treated as worth a B+ grade, maybe better. In an AI world–and especially the world outside the classroom–the AI-generated draft will often come first (or very early) in the process with the syntax already correct. In that setting, teaching editing is becoming more important than teaching writing, and one might even reverse the previous slogan and say that “rewriting is writing.”

There’s a standard problem in the social sciences in moving from the results of a smaller-scale study to larger-scale applicability. Consider a program in a certain school district that seems to increase academic performance and high school graduation rates. Would the program work in another school district? Will it work statewide? Would it work nationwide? Would it work in other countries? How “generalizable” is the result? The question doesn’t have a simple answer, but John A. List offers his own angle in “Make science more reliable: study people as they go about their lives” (Nature, June 25, 2026, 654, pp. 863-866). List offers various memorable examples of the generalizability problem.

Take Scared Straight, a programme run by more than 30 US states between 1978 and 2015 that aimed to dissuade at-risk teenagers from becoming hardened criminals by bringing them face-to-face with people incarcerated in maximum-security prisons. The programme was extended after a pilot project, the subject of a 1978 documentary, found that 80–90% of teenage participants stayed out of trouble. But the intervention did not work when scaled up. In some places, criminal behaviour among teenagers even rose.

A core problem is that (brace yourself for the insight here) people are different, and they live in different societies, economies, cultures and subcultures. Moreover, when people know that they are participating in a study, and they may even get attention from people involved with running the study, they may react differently than if they have similar incentives but are not gettign the attention from being watched while participating in a project.

List is a prominent supporter of “natural field experiments,” which are the idea that a research designs a study in a way that the participants don’t know they are in a study. A classic example here are “audit studies,” in which the reseachers who are studying discrimination train a bunch of assistants who have very similar paper credentials, but different races or ethnicities, and then send out the assistance to rent an apartment, get a bank loan, buy a car, or apply for a job (for examples of such studies, see here, here, and here). In these studies, the potential landlords, bank lending officers, car sellers, and employers don’t know that they are part of the study.

List has carried out roughly a zillion of these studies. He is careful to note that such studies should be pre-approved by an ethics board, and should basically only ask the unwitting participants to carry out the same range of choices and activities that they would in their normal lives–that is, the unwitting participants should not be put under stress or threat. But this leaves a wide field of possibilities. For example, List has carried out natural field experiments in areas like how charitable donors respond to different kinds of announcements or matching rates in a fund-raising drive, or might drop wallets in certain places around campus or neighborhoods and see how many are returned. List is currently chief economist at Walmart, where “my team is running natural field experiments with more than 6,000 suppliers to test which factors will most effectively incentivize suppliers to reduce their carbon emissions.”

Once you start thinking along these lines, lots of possibilities come up. A governemnt authority can experiment with different kinds of letters reminding people of various obligations. A retailer can experiment with different labels on a product. An online firm can experiment with different information formats to gauge the response. A researcher can also study whether some groups react more than others–which in turn offers insights about likely generalizability.

The natural field experiment approach doesn’t work easily for all topics. But for both policymakers and the private sector, it embodies a certain view about the world. We know that there is lots of variation in programs with similar goals (timing, subtance, presentation) and we know that private firms can offer their products in a variety of ways (information, advertising, price discounts). Instead of just letting such variation happen, as it does, a policymaker or private firm can choose the variation systematically, and learn from it. From this perspective, the world has far too few natural field experiments.

Is it really possible that pretty much every single college and university around the country will, this fall, be welcoming its best-ever, most qualified, most curious, most diverse set of first-year students? Seems unlikely. In an interview with Evan Goldstein and Len Gutkin in the Chronicle of Higher Education, Andrew Delbanco offers an alternative. The interview is titled “Is Academic ‘Smugness’ on the Rise? Andrew Delbanco on humility, complacency, and the value of self-criticism” (August 7, 2026). Delbanco is Pofessor of American studies at Columbia University, In 2012, he was awarded the National Humanities Medal.  He comments:

But to do serious academic work, research or teaching, self-doubt and curiosity — implying that there are things out there that you don’t already know and that you want to learn about — are central. Did there used to be more humility and less smugness? Having been in the academy now for more than 50 years, I’m inclined to say yes.

Some of the negative reaction to elite institutions is of their own making — a sense of superiority that we’ve got the answers figured out. If you’re inside the gates, you deserve something that people outside the gates don’t. If you think back to the very early days of the preeminent, Northeastern institutions, which all arose from one version of Protestantism or another, the president would have said to incoming students, “You are all worthless sinners. You’re here entirely by the grace of God. You haven’t earned it. What you should hope to achieve is a true sight of your own sinfulness and your own imperfection.” The speech today is: “Welcome to our great university. You are the best and brightest who have ever walked through the gate. We’re privileged to have you here.” That message should be mitigated or modulated.

There’s a common script about artists who live hard and die young. It says that the combination of a self-destructive lifestyle and art means that they were in some way fully alive, in a way that would be impossible, say, for some happily married husband and father, living in the suburbs and working as an editor. Indeed, the script reaches beyond the young and the dead to those who become obsessive about the pursuit of something. In this case, it says that the purity and single-mindedness of their obsession, even as they sacrifice other values and people along the way, shows that they are living fully.

I am extraordinarily hesitant to judge whether other people are living fully. The economist who inhabits one corner of my soul tells me that people make choices, and those choices generate costs and benefits. If a mountain-climber dies while trying to summit a lofty peak, I am willing to consider the possibiliy that the climber was willing to make that tradeoff. But many other people (including me) would not make that choice, and I do not think we are somehow living less fully as a result.

As always, George Orwell had a useful comment on this theme. It’s in a book review, “Review of ‘The Rock Pool’ by Cyril Connolly, ‘Almayer’s Folly’ by Joseph Conrad,” which appeared in New English Weekly, July 23, 1936. It appears in The collected essays, journalism and letters of George Orwell (1920-1940), by George Orwell, Sonia Orwell, and Ian Angus (1970, pp. 225-227). In paricular, he is reacting to the book by Connolly. Orwell wrote:

Obviously, modern mechanised life becomes dreary if you let it. The awful thraldom of money is upon everyone and there are only three immediately obvious escapes. One is religion, another is unending work, the third is the kind of sluttish antinomianism – lying in bed till four in the afternoon, drinking Pernod – that Mr. Connolly seems to admire. The third is certainly the worst, but in any case the essential evil is to think in terms of escape. The fact to which we have got to cling, as to a lifebelt, is that it is possible to be a normal decent person and yet be fully alive.

In our hypersensitive age, even using words like “normal” and “decent”–or at least using them as descriptors rather than ironically–can trigger a negative response. It is perhaps useful to note that Orwell is not making an exclusionary claim that only normal decent people can be fully alive. He is leaving open the possibility that a life of, say, risk-taking, intoxication, promiscuity, disloyalty, and friendlessness can also be fully alive. Instead, Orwell is making a more limited claim that normal decent people can be fully alive; indeed, that the idea of being fully alive in this way should be something to cling to, like a lifebelt. It’s interesting to me how many 21st century people have a nearly allergic reaction to that idea.

The great historian David McCullough (1933-2022) won the Pulitzer prize for biography twice, for his books on Truman and John Adams, but he also wrote best-selling books on the Panama Canal, the Johnstown Flood, the Brooklyn Bridge, the Wright brothers, and others. Here are his comments in a 2002 interview with Bruce Cole on the importance of writing as an act of thinking and learning (Humanities, “The Danger of Historical Amnesia,” July/August 2002, 23:4, my emphasis):

 I am adamant on the subject that we must not cut back on funding of the teaching of the arts in the schools: music, painting, theater, dance, all of it. The great thing about the arts is that the only way you learn how to do it is by doing it. If a child learns nothing but that as a guide to life, that’s invaluable. You can’t learn to play the piano without playing the piano, you can’t learn to write without writing, and, in many ways, you can’t learn to think without thinking. Writing is thinking. To write well is to think clearly. That’s why it’s so hard. …

We all know the old expression, “I’ll work my thoughts out on paper.” That’s exactly right. There’s something about the pen that focuses the brain in a way that nothing else does. That is why we must have more writing in the schools, more writing in all subjects, not just in English classes. And the teacher who teaches history should be grading the writing, too.

Is new technology that replaces a worker a good thing? We have a tendency, it seems to me, to believe that most previous inventions turned out to have overall net social benefits, but to be worried that all future inventions will instead have overall net social costs. Philipp Ager and Davide Coluccia tackle the question of how the arrival of the sewing machine affected women in the second half of the 19th century in “Liberation Technology? The Impact of the Sewing Machine on Women” (CEPR Discussion Paper 21496, May 2026). Tim Phillips interview Philipp Ager about the research in “Did the Sewing Machine Liberate Women?” (VOXEU Talk, June 12, 2026). I’ll quote here from their discussion.

The sewing machine was expensive enough that it was only used by industry up to about 1860. At that point, Isaac Singer introduces a much less expensive machine for household use. Sales to households rose into the hundreds-of-thousands per year, and by the end of the 1870s, about half of all US households had their own sewing machine.

To set the stage for what happens, Phillips offers some context from the time when the sewing machine arrives. The New York Times apparently wrote in 1860: “No one invention has brought with it so great a relief for our mothers and daughters as these iron needle women. Indeed, it is the only invention that can be claimed chiefly for women’s benefit.” Also, Phillips notes: “In 1860, the American magazine Godey’s Lady’s Book — the most widely read women’s magazine in the US at the time — did the arithmetic for us. A man’s shirt, it reckoned, took 20,620 stitches and 14 hours 26 minutes.”

How did all those new sewing machines affect work for women? Ager and Collucia mostly focus on detailed US census data for women in Massachesetts, which they can supplement with some Massachusetts-based data. More specifically, they have individual-level data on “occupation, fertility, marital status, place of residence,” as well as household wealth. They don’t have individual-level data on who owned a sewing machine, so Ager explains how the tackle the question of “exposure” to the new sewing technology:

[T]here are two sort of exposure to sewing machines — one is like an industrial exposure to sewing machine, and the other one is an exposure to sewing machines as a household tool. So, like basically we have information on industries where we know that they widely adopted the sewing machine during our sample period. And so, like basically we expect that in these towns where these industries were initially located that there is a higher impact on female employment, for example, in these industries. So, like you know, we expect a stronger effect in these locations where more of these industries were initially located before the sewing machine was widely adopted. And for the sewing machine as a household tool, we actually digitized information on retailers from business directories. So, we have information about sewing machine retailers that sold this appliance to households, and we use as a primary measure whether, like, you lived in a town that had access to the first sewing machine retailer in that town. And that’s our primary way how we capture exposure to sewing machine as a household appliance.

Their analysis suggests that the sewing machine increased overall employment for women, and in turn led to a lower overall marriage and fertility rate. However, these effects were strongest for women from poor and working-class households, and look quite different for woment from middle-class and high-income households.

[T]he sewing machine overall increased the employment opportunities of working-class women in these industries that widely adopted the sewing machines. Right, this was not only the garment industry, but you can also think about, for example, the boot and the shoe production — where many of these women were pulled into the factories — but we also find some impact on, for example, general merchandise stores that started to sell ready-made clothes, and they needed seamstresses and dressmakers in their departments. … [W]e find that for these working class or poorer women that they started to reduce fertility. Why is that the case? This is consistent with the argument that there were better employment opportunities, and this basically increased the opportunity costs of having children. … We find for poor women that — if they live before the rollout of the sewing machine really happened at scale — if they lived in towns with high industrial exposure, they were less likely to be married when they were young adults. And we can even follow them a bit longer when they completed fertility, and we also find that they had fewer children, and many of them remained unmarried when we look at them, say, as 50- or 55-year-olds. … [M]iddle class women that had access to a sewing machine were less likely to work as an adult because they used their free time for earlier family formation. They married earlier, and they also started to have children earlier, and that’s consistent with the general notion at that point in time that there was a strong stigma associated with women working outside the home once they were married.

The new sewing machine technology thus changed the time use of women, but certainly didn’t lead to “free” time. Instead, the technology interacted with class and gender norms. For poor and working-class women it opened employment opportunities to satisfy the demand for low-cost machine-sewn clothing, although these jobs often involve long hours, low pay, and sweatshop working condidtions. For middle-class women, it led to earlier marriage and more children. Like so many technology stories, there isn’t simple good-or-bad bottom line here, but a broader economic and social evolution. However, I find it impossible tor regret the “loss” of the earlier seamstress jobs that took 20,000 stitched and 14 hours of work for a single shirt.