I seem to run into comments on a semi-regular basis about how the policies of either President Biden or President Trump either are bringing back US manufacturing jobs or have already done so. Thus, a quick reality check seemed in order.
This figure shows total US manufacturing jobs back to 1939. After the dramatic fall in manufacturing jobs in the first decade of the 21st century (for reasons including the dramatic rise in Chinese exports after China joined the World Trade Organization in 2001 and the effects of the Great Recession from 2007-09), There has indeed been a modest bounceback. It started around 2010, and after an interruption from the pandemic recession has continued since.
It’s important to keep the size of this bounceback in perspective. As you can see from the vertical axis, US manufacturing jobs had declined to 11.5 million in early 2010, and have now risen to 13 million. This increase of about 13% in manufacturing jobs is meaningful–although it’s perhaps worth noting that the rise was underway and much of the gain had happened before either the Trump or Biden presidency.
Another useful perspective is to look at US manufacturing employees as a share of all employees. Here’s the figure. On the far left, you can see the jump in US manufacturing jobs during World War II, and the fall immediately after the war. Since then, the long-term story of US manufacturing employees as a share of total US employees has been one of gradual decline, from about one-third of the US labor force back in the 1950s to about 8% of the labor force at present.
In the case of US manufacturing, two insights can both hold true. One is that it’s important to have an active and robust US-based manufacturing sector across a variety of industries, because it’s important for the US economy to be close to the frontier of innovation in a wide range of areas, and being physically close to actual manufacturing is useful in developing and applying new technologies and production techniques.
The other and complementary insight is that a substantial renaissance in US manufacturing jobs, in a way that the become a steadily and substantially rising share of US employment over time, is extremely unlikely. Manufacturing as a share of employment has been declining for the US for a long time, as well as for high-income countries across the world. Jobs in services (including services related to design, management, marketing, transportation, installation, maintenance, and repair of manufactured output) have been a rising share of employment for decades. For low- and medium-income countries, a key question is whether or how they can grow their economies based on a rise service-oriented jobs. Manufacturing across many industries has become increasingly capital-intensive, often making use of very expensive facilities (think semiconductors, for example), along with automation and robots.
So yes, the US needs a healthy manufacturing sector, and that employment option will work well for millions of people. But the other 90+% of the US workforce will need other opportunities for good jobs and prosperity.
It is remarkable but true that per capita US economic growth has hewed close to a trendline of 2% per year for the last 150 years. Here’s a recent figure showing this pattern from Charles Jones in his paper “The Outlook for Long-Term Economic Growth,” as prepared for the annual symposium at Jackson Hole hosted by the Federal Reserve Bank of Kansas City (August 2023, full symposium proceedings including Jones’s paper available here, Jones’s paper also available as NBER Working Paper #31648).
Notice that the left-hand axis of the figure is a logarithmic (that is, a proportional) scale. On such a scale, a 2% growth rate appears as the straight dashed red line. The path of the US economy has sometimes been a little above this line, and sometimes a little below below. But the fact that the 2% annual growth rate fits the long-term pattern so well is quite remarkable. This pattern isn’t a new discovery: for example, here’s a previous post from 2012 discussing it.
Jones presents a succinct summary of possible reasons why future growth rates might rise or fall. He tends to be a pessimist about future growth rates, for two main reasons for pessimism. 1) Growth comes from new ideas that can spread freely, and more people means more possibilities that some of them will develop these new ideas, so slower population growth will result in slower technological growth. 2) Some of the forces that contributed to US growth in the past, like an overall rise in US education levels and a rise rate of investment in R&D and new ideas, seem to have stopped rising or at least slowed dramatically.
Conversely, he also offers three potential reasons for optimism. First, economic development in previously poor countries like China and India means a larger pool of people who can participate in the search for new ideas, which can then spread to the US economy. Second, the US has been experiencing improvements in the allocation of talent as members of groups that have been historically underrepresented in science and technology have been playing a growing role. Third, artificial intelligence and increased automation may together improve the growth rate.
For myself, I think I have a pretty good grasp on the US economy with about a 2-3 year lag: that is, I’m currently feeling as if I have a reasonably good grip on what happened in the US economy up through mid-2021. Coming to an understanding the past is hard enough for me; I’m not in the business of future predictions. But for those who are predicting either a resurgence or collapse in long-term growth, I would suggest that they have some respect for the fact that a 2% annual growth trend for the US economy has persisted through some very dramatic changes in the last 150 years.
The World Development Report is one of the flagship annual reports for the World Bank: for 2023, the theme is “Migrants, Refugees, and Societies.” The report defies a simple summary, but here are a few key points.
The World Bank defines “migrants” as those who are living outside the country of their nationality: thus, someone who emigrates to the United States would be counted as a migrant unless or until they became a citizen, and then wouldn’t count as a migrant any longer. With this definition:
As defined in this Report, there are globally about 184 million migrants (about 2.3 percent of the world’s population)—37 million of them refugees:
About 40 percent (64 million economic migrants and 10 million refugees) live in high-income countries that belong to the Organisation for Economic Co-operation and Development (OECD). These are high- and low-skilled workers and their families, people with an intent to settle, temporary migrants, students, as well as undocumented migrants and people seeking international protection. This number includes 11 million European Union (EU) citizens living in other EU countries with extensive residency rights.
About 17 percent (31 million economic migrants) live in Gulf Cooperation Council (GCC) countries. Nearly all of them are temporary workers with renewable work visas. They represent, on average, about half of the population across GCC countries.
About 43 percent (52 million economic migrants and 27 million refugees) live in low- and middle-income countries. They moved primarily for jobs or family reunification or to seek international protection.
The share of migrants in the global population has remained relatively stable since 1960. However, this apparent stability is misleading because demographic growth has been uneven across the world. Global migration increased more than three times faster than population growth in high-income countries and only half as fast as population growth in low-income countries.
That final sentence, of course, is part of the reason for much of the political turmoil over immigration in high-income countries.
At the most basic level, the economics of migration starts from the evidence that the migrants themselves benefit. They may benefit in some cases because they are moving to places of greater opportunity, or in other cases because they are escaping repression or persecution, but their gains are real and should appear in the social calculus. But migration also affects non-migrants in both the sending and receiving countries in a variety of potentially positive and negative ways, and so political controversy is inevitable.
The World Bank suggests that one framework for thinking about these issues from the standpoint of the receiving country is the “match-motive matrix.” When immigrants are a good match for pre-existing needs in the receiving country and, the receiving country as a whole will gain. When the match is poor, as when the migrants do not have skills needed in the receiving country and perhaps are refugees seeking only to escape persecution, then the costs for the receiving country may exceed the benefits.
From the standpoint of those remaining in the origin country, migrants to other locations can also bring benefits by sending remittances and providing knowledge transfers. In some cases (nurses from the Philippines are a commonly cited example), the fact that migrants will go to work in other countries provides a strong reason to build up domestic institutions that can provide skills to those remaining in the country of origin as well.
The report works its way through the policy choices that need to be made in sending countries, receiving countries, and multilateral agreements to increase the benefits of migration and, in the case of refugees, to find ways to share the costs. I also expect that the ongoing spasm of immigration across the southern US border will lead to policy changes. Here, I’ll just close with a broader theme: rules about migration are in many ways similar to rules that affect society as a whole. For example, how can we support children from families with low incomes and limited educational background to realize their potential as adults contributing to society? How do we support people of diverse backgrounds in living side by side? As the report notes:
This debate plays out in a context in which societies and cultures are neither homogeneous nor static. There is no “pre-migration” harmony to return to. In every society, tensions, competition, and cooperation have always existed across a variety of groups that are partly overlapping and constantly changing. Some of these tensions reflect socioeconomic divides: they are not about migration but about poverty and economic opportunity—and large numbers of migrants happen to be poor. Because many of those who moved or their descendants have been naturalized, some of the cultural issues attributed to migration are, in fact, about the inclusion of national minorities. Migration is also just one of many forces transforming societies in an age of rapid change, alongside modernization, secularization, technological progress, shifts in gender roles and family structures, and the emergence of new norms and values, among other trends.
For some previous posts digging into the economics of immigration, one useful starting point is a symposium in the Fall 2016 issue of the Journal of Economic Perspectives (where I work as Managing Editor):
The “electrification” agenda is the idea that the move away from fossil fuels and carbon emissions can happen via electricity: that is, generate carbon-free electricity and then use it to replace fossil fuels.
But in-between the massive increase in electricity generation that would be needed and the uses of that electricity in homes, cars, and businesses, there will also need to be a dramatic expansion (maybe a tripling or quadrupling) in electricity lines to distribute that electricity where needed. In the US electrical grid, almost all of the distribution of electricity is done by regulated public utilities. These companies are owned by shareholders–that is, you can buy stock in them. But their pricing and investment plans are regulated by government. And these companies are the ones that will ultimately be making the decisions about whether and in what ways to expand the electrical grid.
Rate-regulated public utilities own and operate one-third of U.S generators and nearly all the transmission and distribution system. These firms receive special regulatory treatment because they are protected from competition and subject to rate caps. In the past decade, they also have been at the center of high-profile corporate scandals. They have bribed regulators to secure subsidies for coal-fired generators and nuclear reactors. They have caused wildfires and coal-ash spills that resulted in hundreds of deaths and billions of dollars in liability. Their failure to maintain reliable electric service has contributed to catastrophic blackouts. Perhaps most consequentially, they have emerged as powerful opponents of state and federal climate action. This Article describes the unique corporate governance challenges public utilities face and argues that these governance challenges contribute to the pervasive inefficiencies and the frequency of corporate misconduct that characterize utility industries.
As the authors point out, the shareholders that own rate-regulated public utilities are different than shareholders in other companies. Because of the regulations, the shareholders are never going to get a super-high rate of return, but because the regulators don’t want want the electric company to go out of business, shareholders are also somewhat protected from losing money. If the company performs quite well–providing reliable and green energy at low cost–it’s the users of that electricity, the ratepayers, who will actually reap the greatest benefit. Conversely, if the company performs badly, the shareholders are likely to be somewhat protected by the regulators, but those depending on the electricity supply will suffer.
In a meaningful sense, the shareholders of this kind of company are more like creditors, getting a steady return unless there is a bankruptcy. The ratepayers are more like shareholders, because they are the residual claimants who experience most of the benefit or harm when the company performs well or badly.
The proposed solutions of the authors to this situation are only mildly persuasive, in my view. For example, they want “ratepayers” to have a place on the board of directors of these regulated public utilities, which I suspect would have little effect on the underlying incentives.
Kovvali and Macey are focused on the overall governance problem for these companies, not really on the need to have a dramatic expansion of the electrical grid. But as they point out, existing public utilities that both generate and transmit electricity may have mixed incentives about building additional transmission lines. After all, additional transmission is likely to mean more outside competition from electricity produced elsewhere–which can be a good thing for rate-payers, but harder to justify to shareholders. In general, shareholders of publicly-regulated utilities are not likely to lobby the company or the regulators for a dramatic expansion of transmission lines, because such a step would involve taking on a lot of debt and probably (given regulated prices) will not improve shareholder returns. But unless regulators provide them with incentives to do so, public utilities aren’t likely to expand transmission lines. Along with the various practical hands-on issues of dealing with the costs and permissions needed to expand transmission lines, these institutional constraints are likely to be another complicating issue.
Traditionally, the basic US data on issues like employment/unemployment and family income are derived from government surveys: that is, either people are called on the phone or sent forms, or both, and their responses are tallied. There are obviously problems with this approach: for example, if you are receiving government benefits, but also have a side job to bring in some extra cash, are you likely to be honest when someone from the government asks if you are employed? However, one could at least argue that if the surveys are carried out in (roughly) the same way over time, then these kinds of biases would be (roughly) the same over time.
Here’s a table showing the problem of declining response rates for households n the last decade–especially since the pandemic hit in 2020. The Current Population Survey (CPS) provides the core data on employment, earnings, and the workforce. It had a 90% response rate a decade ago, now down to 73%. The Consumer Price Index (CPI) Housing survey collects data on prices for rental housing and also estimates the “owner equivalent of rent.” These numbers feed into the estimates of overall inflation. This survey had a 70% response rate in 2014, but was down to 52% by 2022. The Current Expenditure Survey (CE) collects data on how households are spending their money, and among other purposes, it is used to weight observed price changes and develop the inflation rate. It’s down to a 43% response rates. The Medical Expenditure Panel Survey collects data from households and health care providers. families and individuals, their medical providers. It’s was down to a 46% response rate before the pandemic. The American Community Survey (ACS) collects annual data on a wide array of social, economic, housing, and demographic factors down to the community level. Its response rate was 97% a decade ago, and it was down to 71% before the pandemic hit.
When the survey response rates drop this dramatically, the reliability of the information drops, too. The survey response problems are severe enough that a number of specialized surveys have been ended, or in some cases suspended for a time. The NAS report notes (citations omitted):
The increasing costs of obtaining participation and flat or declining budgets have led to the elimination, or threat of elimination, of multiple important programs and surveys. … For example, in 1996, the National Vital Statistics System, part of the National Center for Health Statistics, suspended the collection of detailed national records-based data on marriages and divorces. In 2008, after publishing fourth-quarter 2007 estimates, the U.S. Census Bureau terminated its quarterly survey measuring residential alterations, improvements, and repairs. In the absence of official statistics, private sector estimates of the size of the home-improvement marketplace vary widely. For 2020, private sector estimates ranged from $150 billion (Statista, 2022) to $325–333 billion. The elimination of the U.S. Bureau of Labor Statistics (BLS) Mass Layoff Statistics program, a BLS-state cooperative program, resulted in the loss of a standardized approach across states to identify, describe, and track the effects of major job losses. With the loss of the Information & Communication Technology Survey, there are no longer official annual estimates of information, communication, and technology equipment or software purchases—a huge and growing market. According to a report commissioned by the Census Project, a nonpartisan advocacy group, the future of the ACS is threatened. Experts argued that the ACS, a survey central to the nation’s data infrastructure, needs an additional $100–300 million in funding to address current limitations and introduce much-needed enhancements.
So you want to know about estimate of IT and software spending in the US economy? Might be important! But the survey ended. Want to study how people are spending money on fixing up their houses, perhaps rather than moving, in this era of work-from-home and higher interest rates? Might be important. But the survey ended. Want to study patterns of marriage and divorce, both the effects on those involved and also the effects on care-giving for children and elderly parents? Might be important. But the data isn’t being collected by the federal government.
Part of the issue here is just that the government needs to spend more on collecting statistics. As I’ve noted in the past, total federal spending on collecting statistics is 0.18% of the federal budget–that’s not 18%, but less than one-fifth of 1%. It would be a sensible social investment to spend more here: say, an additional 0.1% of federal spending. But as the traditional survey-based methods of collecting data become increasingly unreliable, the statistical base also need to shift to alternative sources of data.
This shift is already happening. Researchers are making much wider use of “administrative” data that is collected for other purposes: for example, being able to look at tax return data to measure income, or Social Security data to measure wages, may be more accurate than relying on household surveys. The obvious questions here are how to limit the use of administrative data so that it shows overall patterns but doesn’t invade the privacy of individuals. Also, because administrative data was not designed to be used for research purposes, it needs to be handled with care. But these hurdles are surmountable.
The second NAS volume launches a discussion of expanding the sources of publicly-available government data in other ways. In some cases, this can mean being open to finding ways of linking existing data. For example, the NAS report gives the example that there was data from the US Department of Housing and Urban Development about what housing had greater or lower risks of lead exposure for children. There was data from the National Health and Nutrition Examination Survey (NHANES) on levels of lead in children’s blood. But there was no link between the two: that is, it wasn’t possible to determine how living in a place with less exposure to lead affected the level of lead in the bloodstreams of children. However, it was possible to link records of actual households, both where they lived and the levels of lead in children’s blood–and to do so with an anonymous numbering system so that the information of any specific household was not available to the researchers.
But the report also considers a wide array of other data. Along of asking households about what they buy, for example, perhaps actual sales records from grocery stores or retail outlets could help. In learning about, say, health or car or home insurance, perhaps actual data from insurance companies could help. For issues like environmental measurement and agriculture, satellite images may help. Location data from cellphones and health data from fitness trackers might be useful. Data might be scraped from the web, or from social media, or crowd-sourced.
It’s easy to think of ways in which these alternative sources of data could go astray, either in accuracy or in revealing personal information. Also, government data needs to be available at regular intervals, broadly representative, and comparable over time, not just a one-time data dump. But it’s also obvious that the publicly-available information is widely useful, and that current methods of collecting public data are in the process of going astray themselves.
It’s a standard story at this point: work-from-home seems to be establishing itself as commonplace in many companies–at least in the “hybrid” form where many or most workers come into the office 3-4 days each week, but work from home the remaining time. For many workers, part of the benefit is avoiding a commute to work. But when fewer people are commuting to work, the company needs less office space. Moreover, the market for meals and for retail in general that had been supported by commuters is reduced in size. Demand for office space falls, and demand for residential real estate near that office space falls as well (because the benefits of being close to the office are smaller if you don’t need to be there physically every work-day).
McKinsey Global Institute reviews these patterns in its report Empty spaces and hybrid places: The pandemic’s lasting impact on real estate (July 2023). The report also offers some preliminary thoughts on how these shifts may alter the use of urban real estate. A theme that stuck with me was that if employers, retailers, and cities want people to commute, they need to act in a way that will “earn the commute.”
As background, the McKinsey report is focused on “nine superstar cities: Beijing, Houston, London, New York City, Paris, Munich, San Francisco, Shanghai, and Tokyo”–although a broader group of cities is also surveyed and discussed. They find that rates of work-from-home seems to be settling at about 30% of work-days. This will shift markets including office real estate, urban retail, construction, and residential real estate.
Employers may both reduce their office space but also shift to newer and higher-quality office space, with better meeting rooms and greater capacity for sophisticated audio-visual equipment and teleconferencing. The firms can’t just tell employees to show up five days each week. If employees don’t see actual benefits from commuting to work in person, they will push back.
Like employers, retailers need to provide a reason for an in-person trip to their real estate, rather than an online connection. “Retailers too may have to `earn the commute by designing spaces that cater to many different uses. A prime example is stores that easily accommodate omnichannel retail—a single, seamless experience for customers, whether they shop online or in person. Similarly, stores can provide experiential retail. For example, one department store brand is launching smaller stores where customers can pick up products bought online, get clothes altered, find style advice, and patronize a beauty salon.”
Those who are building or refurbishing urban real estate need to think more about how use of the space is likely to vary over time. “To adapt to declining demand for traditional office and retail space, developers could create hybrid buildings. The most ambitious vision is a universal, `neutral-use’ building whose design, infrastructure, and technology could be easily modified to serve different uses. Imagine a medical building that could be easily converted into, say, a hotel or an apartment building if customers’ preferences changed.”
Finally, residential real estate will shift as well: “Residential vacancy rates increased from 2019 to 2022 in every superstar urban core that we studied, from a 0.8-percentage-point increase in Tokyo to a 9.9-percentage-point increase in London; meanwhile, in the suburbs, vacancy rates grew much less or even declined. Prices followed suit, rising eight percentage points more slowly in US superstar urban cores than in their suburbs and 13 percentage points more slowly than in non-superstar urban cores. In San Francisco, nominal prices in some neighborhoods fell by 12 percent from the end of 2019 to 2022. Residences in San Francisco’s urban core are now worth $750 billion less than they would have been if prices there had risen at the national average rate. The effect seems to be a global phenomenon.”
I sometimes say that the work-from-home movement quite suddenly converted a lot of what had been residential real estate into commercial real estate–that is, a part of many houses became a place where people worked. The McKinsey report offers estimates of how that shift is likely to reverberate in the medium- and the long-term.
David A. Price carries out an “Interview” with Melissa Kearney, subtitled “On the American fertility decline, the role of social norms, and the link between single-parent households and economic gaps” (Econ Focus: Federal Reserve Bank of Richmond, Third Quarter 2023, pp. 16-20). Here are a few comments that caught my eye:
What is “two-parent privilege”?
More consequentially for children’s outcomes and socioeconomic gaps, children born to college-educated parents are now much more likely to live in a household with married parents and have the associated benefits of that. To be specific, 84 percent of children whose mothers have a college degree live with married parents, compared to less than 60 percent of children whose mothers don’t have a college degree.
This means that the kids born to college-educated mothers live in a household with much higher levels of income, not just because their mother has the potential to make more income, but because she’s much more likely to have a working spouse in the home or to have a spouse in the home at all. But also, there are many more parental resources in general when there are two parents in the home — more parenting time for supervision, nurturing, and so on. To the extent that parenting inputs shape children’s outcomes, this widens the gap in kids’ behavioral and educational outcomes and exacerbates class gaps. This is why I referred to this phenomenon as the “two-parent privilege,” because the two-parent home has now become another advantage of the college-educated class and their children.
Discomfort among academics about discussing the role of family structure
I was at a conference on poverty earlier this summer where someone in attendance, not me, brought up the subject of family structure. The panelists, who were not all economists — they included sociologists and a social worker — were visibly uncomfortable, even annoyed, and promptly dismissed the person’s question.
I think this discomfort stems from a well-intentioned instinct to not want to come across as sounding judgmental or shaming certain types of families. And I’m very sympathetic to that instinct. The problem, though, is that avoiding this topic is counterproductive. Denying the importance of family structure and the role of families to children’s outcomes and economic mobility is just dishonest, based on the preponderance of evidence. Another reason why I think it makes economists, in particular, uncomfortable is that we don’t have a ready solution for the challenge. It would be much easier for us to talk about this issue if there were an obvious policy lever to address it. ..
That’s something that highly educated Americans often speculate about this topic: “Oh, we’re just becoming more northern European in our attitudes about this.” But that’s not what’s happening. The decline in marriage among parents in the U.S. has not been replaced with a corresponding rise in unmarried parents stably living together for the long haul and essentially being married in all but name. In the U.S., cohabitation is a very fragile arrangement.
To what extent can government transfers address this issue?
I am in favor of increasing income support to low-income families with children, given all we know about the benefits of alleviating material deprivation for children. … I do want to acknowledge the concern that insofar as transfer payments increase the economic viability of single-parent households, that might lead to some small increase in these types of households. My read of the evidence is that the behavioral effects there are likely to be small. But either way, I’ll take the trade-off. I am firmly of the view that we cannot just allow children to continue suffering the consequences in the hope that entices some more parents to get married.
Having said that, even an increase in transfer payments isn’t going to fully make up for the absence of the second parent in the home. Parents do more than just pay the bills. They invest their time and energy into their children. They provide supervision and guidance. They read to them and play with them. We should be clear that a government check is never going to be able to fully make up for the absence of a second committed parent in the home.
Furthermore, the reality is that in this country, we couldn’t even muster the political support last year to maintain an annual child tax credit of $3,000 per year. That indicates to me politically how far we are from a situation where we might conceivably have income transfers to make up for the absence of a second earning parent in the home. What are the chances we’re going to have a child allowance equal to, say, the median earnings of a high school graduate, around $40,000 a year, year after year until the child is 18? The idea that we’re going to solve this problem with a government check is just not plausible. And it doesn’t account for all the many things that a second parent brings to the home beyond income.
Here’s the overall pattern. In 2000, the level of remittances was quite similar to “official development assistance” (ODA), while both of those were substantially lower than foreign direct investment (FDI). Now, remittances are roughly triple official development assistance. And if you take foreign direct investment going to China out of the picture, as shown in the bottom panel, remittances are roughly double the level of FDI going to all other low- and middle-income countries.
In short, remittances are a huge part of the capital flows between countries. However, they often don’t make the news. Development assistance from nation to nation or a big overseas investment by a US company can make headlines. But payments of remittances, often within families, are individually quite small–it’s just that they add up to big stuff.
The report provides lots of detail about remittances in regions and countries. Here’s one more chart. The left-hand panel shows remittances received by country. The right-hand panel shows remittances as a share of GDP, which is more than one-fifth of GDP for the countries listed there.
These flows also help to explain the interest, in many low- and middle-income countries, in finding ways to transfer money across international borders at lower cost, even if it involves using cryptocurrencies or other new financial technologies.
I’ll leave it for another day to write about whether remittances are “good” or “bad” (although personally, I lean heavily to the “good” side). On one side, there are concerns that countries sending migrants may suffer from “brain drain,” as those who are more educated or highly motivated may go other places. On the other side, when people of your country are earning and sending home tens of billions of dollars, or in some cases more than one-fifth of GDP, that’s not obviously a bad thing! Moreover, the skills learned by migrants and the economic connections they make can often be of considerable use to those who remained in the country of origin.
In the United States, we naturally celebrate the Labor Day holiday with a holiday from work. While this may be counterintuitive for others, economists have long see the price of labor and leisure as intimately intertwined. Although prices are often stated in terms of money, which is a useful convenience, the true price of anything is determined by its “opportunity cost,” which is what you give up or trade off. A standard example is that attending college has two costs: what you pay for tuition, room, and board, and also the opportunity cost of what you could have earned if you had been working at a job during that time. In a similar spirit, leisure has a opportunity cost–it’s the income you could have earned if you had been working. This insight isn’t new, of course: any gig worker in the modern economy knows it at a visceral level.
But the idea that leisure has an opportunity cost has some perhaps surprising implications. It means, for example, that as an economy becomes better off over a long period of time and the standard of living rises, leisure becomes more expensive in the opportunity cost sense. Moreover, as any good or service becomes more expensive, it will tend to be used differently. Alex Tabarrok explains some of the implications in a 2022 essay (“William Baumol and the Cost Disease,” in Research in the History of Economic Thought and Methodology: Including a Symposium on the Work of William J. Baumol: Heterodox Inspirations and Neoclassical Models ungated version here). Tabarrok writes:
Travelers to developing countries notice that haircuts and restaurant meals are cheap and often something else. Life in developing countries seems slower, more tranquil, and less harried. Meals are longer, conversations deeper, time seems to move less quickly. Development seems to be accompanied by fast food, fast talking and even fast walking (Levine & Norenzayan, 1999). In The Harried Leisure Class, Staffan Linder (1970) proved a theorem … As productivity increases, as measured by wages, so does the opportunity cost of leisure or more generally, time. A higher price of time encourages us to economize on time, so as wages rise, we schedule our time more carefully with time planners, “to do” lists, calendaring and incessant notifications. We consume more quickly and we consume more goods that are quick to consume so “fast food” becomes the norm and we choose to watch television or movies “on demand” rather than read books or go to plays or live music performances. We consume multiple goods at the same time as when we eat and watch, talk and drive, and exercise and listen. …
Linder notes that goods and services take time to consume, and time is limited so the price of good x should include not just its money price but also its time price, the wage rate multiplied by the time it takes to consume the good … Or, as Baumol (1973, p. 630) put it, “rising productivity decreases the demand for commodities whose consumption is expensive in time.” …
[T]here are qualifications and additional considerations. Most notably, since the … Linder theorem [is] driven by an increase in productivity, consumers are not made worse off by the respective increase in prices. The choice to consume fewer time-intensive goods as wages increase is welfare-maximizing. It is also true that time is not entirely fixed. Time-saving devices such as faster internet or supersonic aircraft are possible. We can also increase life expectancy giving us more time to enjoy our greater productivity of goods and services (Hall & Jones, 2007). Indeed, we are consuming more leisure over time. Nevertheless, the power of the Linder theorem should be evident. One reason we consume more leisure than in the past is that we manage our time at higher levels of intensity than in the past. A search at Amazon for “time management,” for example, leads to over 10,000 hits. It is also the case that leisure, especially passive leisure such as watching television, has become more concentrated in people with lower incomes (Aguiar & Hurst, 2009).
An obvious question here is the extent to which you wish to push back against more intensive management of your leisure time. Are you planning a day that involves travelling multiple places, preparing and eating multiple foods, while paying attention to multiple music channels and social media feeds? Do you have a desire, or perhaps the internal fortitude, to commit a substantial block of time to a single activity: perhaps sitting down for an uninterrupted conversation, or a swim or a hike, or a couple of hours with a single book. There’s no best answer here, of course. Preferences will differ. But if when you reach the end of your holiday you are exhausted in such a way that you feel as if you need another day off, maybe it’s time to take closer look at how you are being influenced by the rising time costs of leisure.
Considerable evidence has been suggesting for some time now that children suffered learning loss from the K-12 during the pandemic. From a research perspective, it’s trickier to demonstrate exactly why children suffered such losses. School closures are an obvious possibility, and probably also part of the answer. But the other disruptions of the pandemic to family health, economic prospects, and social lives surely played a complementary role. But for the affected children, the arguments over causes are also arguments about the past: for this group, the policy question is whether remediation efforts can help them catch up to where they would have been.
Santiago Pinto pulls together the existing research on this issue in “The Pandemic’s Effects on Children’s Education” (Economic Brief: Federal Reserve Bank of Richmond, August 2023, #23-29).
It’s worth putting that learning loss of about a half-year in perspective. In some ways, it doesn’t seem like a lot. Shake it off! Catch up! But US schools have had a substantial share of students lagging behind grade-level for decades now, and they have not shown much ability to help large groups of students catch up. Indeed, if there was a school reform that would boost average student learning by a half-year, it would be thought of as a near-miracle. Conversely, losing a half-year of schooling is a near-disaster.
Moreover, US schools do not seem to be mobilizing in a way that will help students catch up. For example, students have shifted out of K-12 schools to alternatives and absenteeism is high. Pinto writes:
The 2023 report “Where the Kids Went: Nonpublic Schooling and Demographic Change During the Pandemic Exodus From Public Schools” uses data from 21 states plus D.C. (covering the 2020-21 and 2021-22 school years) to explain what happened with students leaving public schools. The report shows that an increase in home-school enrollment and a decrease in school-age population each explain 26 percent of the decline in public-school enrollment, and 14 percent was explained by an increase in private-school enrollment. About one-third of the decline, however, remains unexplained.
We don’t yet know if those who shifted to alternative schools will be more able to catch up. We do know that school districts with less in-person learning had the biggest drops in test scores, which doesn’t suggest that the alternatives are performing especially well. In general, student learning is tied to the amount and the continuity of instructional time received, and neither the amount nor the continuity seems to be rising in a way that will help students as a group to catch up.
It’s of course dicey to estimate what learning loss means in the long-run. Pinto describes one effort this way:
[A]n additional year of schooling increases income by about 11 percent on average in the U.S. If it is assumed that the pandemic and the associated changes in schooling generated a loss of one-third of a school year, then this would translate into a loss of income for the affected students of about 3.5 percent over their entire working life.
The losses in lifetime earnings are of course only part of the issue: “Empirical evidence has linked school closures to several factors, including rising mental health concerns, lower levels of engagement, reports of violence against children, rising obesity, increases in teenage pregnancy, rising levels of chronic absenteeism and dropouts, and overall deficits in the development of socioemotional skills due to social isolation from networks and peers.”
These effects tend to be larger for student who are from economically disadvantaged families, and those who are already falling behind academically. Schools tend to be a “great equalizer,” making up to some extent for the unequal distribution of other educational resources across families. Those children who depended most on public schools also lost the most when the schools shut down.
What would an emergency-level response to these issues look like? One possible step would be a dramatic rise in programs of in-person tutoring. “The 2022 working paper “The Challenges of Implementing Academic COVID Recovery Interventions: Evidence From the Road to Recovery Project” claims that making up the gap would require approximately 40 to 100 hours of high-quality tutoring for the average student (slightly less for reading than math).” Think of this as maybe two hours of tutoring per student every week. Making this happen would require an immediate and dramatic expansion of tutoring programs, drawing on parents, retirees, college students , and others.
Another step would be to expand the school year by, say, six weeks: “The 2023 paper “The Fiscal and Welfare Effects of Policy Responses to the COVID-19 School Closures” shows that extending schools by three months (or six weeks over the next two summers) generates significant welfare gains for the children and raises future taxes to pay for the cost of this schooling expansion.”
Frankly, a dramatic rise in tutoring and expanded school years should have been much higher on the K-12 policy agenda before the pandemic. During the pandemic, schools should already have been planning for the implementation of these steps back in 2021 and 2022. A decrepit bridge can sometimes wait a few years for necessary repairs. A current fourth-grader or an eight-grader or a high school senior, lagging behind academically, doesn’t have the luxury of waiting.
