Wednesday, April 1, 2020

Urbanization: Glaeser's Presidential Address to the Eastern Economic Association

Urban areas have traditionally been engines of prosperity and social mobility. But the technologies driving changes in urban structure and the ways in which government responds to these changes has evolved  over time. Edward L. Glaeser prepared the Presidential Address for the Eastern Economic Association on "Urbanization and Its Discontents" (Eastern Economic Journal, April 2020, 46:191–21). It's also available as NBER Working Paper # 26839. (Both links require a subscription, which most academic libraries will have.)

Glaeser offers a brief reminder of past urban patterns:
Urban fortunes are shaped by technological change. During some periods, technological shifts are largely centripetal, meaning that they pull people toward cities. During other eras, technological trends are centrifugal, meaning that they push people away from dense urban cores.
The nineteenth century was predominantly a centripetal century, marked by series of innovations, including steam engines, streetcars and skyscrapers, that abetted urban growth. The first 60 years of the twentieth century was largely a centrifugal era, largely because technological change reduced the tyranny of distance. Cheaper shipping costs, from highways, cheaper railroads and containerization,
allowed far-flung people to participate more fully in the global economy (Glaeser and Kolhase 2004). Radio and television enabled the rural population to enjoy previously entertainment.
These lower costs reduced the need to locate production near the urban ports and railroads that once anchored all of America’s cities. The mass-produced automobile enabled low density mobility and the rise of car-oriented suburbs (Baum-Snow 2007). These centrifugal technologies first slowed the rise of American cities and then enabled a mass exodus from urban America. The air conditioner made America’s warmer places far more appealing than they had been before World War II, and a move to sun accompanied the move to sprawl. Urban social problems, especially weak schools and crime, were exacerbated by suburbanization and then further encouraged the move to the suburbs and to lower density Sunbelt cities.
However, the last four decades or so have seen a resurgence of many urban areas, based in large part on a rise in the economic importance of proximity. Glaeser explains:
The industrial jobs that had once been the backbone of urban economies did not return. Instead, human capital-intensive business services became the new export industries for urban areas. Financial services expanded enormously in urban America from 1980 to 2007. At its height in 2007, finance and insurance generated over forty percent of the total payroll on the island of Manhattan. The urban edge in transferring knowledge is particularly valuable in finance, because a bit of extra information can make millions for a trader in minutes.
Face-to-face contact is often part of the delivery mechanism for urban services. Clients like to meet their accountants, bankers, lawyers and management consultants in person. Face-to-face contact is even more imperative for barbers and manicurists. Urban interactions enable young workers to become more skilled. ... 
Why didn’t improvements in electronic communication make face-to-face contact obsolete? While e-mail is possible almost everywhere, face-to-face interactions generate a richer information flow that includes body language, intonation and facial expression. As the world became more complex, the value of intense communication also increases. Physical immersion in an informationally intense environment, such as trading floor or an academic seminar, generates a rush of information that is hard to duplicate online. Moreover, dense environments facilitate random personal interactions that can create serendipitous flows of knowledge and collaborative creativity. The knowledge-intensive nature of the urban resurgence helps to explain why educated cities have done much better than uneducated cities. ..
While highly educated workers moved into professional and business services, successful cities also generated employment for less skilled workers in other parts of the service economy. Many workers switched from manufacturing to wholesale and retail trade during the 1990s. Hospitality and food services also expanded dramatically after 1980. Employment in these service industries depends on the demand generated by the success of more export-oriented services, like finance. In areas that lack viable export industries, the dominant sector is typically healthcare and social assistance, where demand is maintained by Federal transfers.
Cities also came back as places of consumption as well as places of production (Glaeser et al. 2001), which partially reflects the rise in returns to skill. As Americans became better educated and as educated people came to earn more, they spent more on higher-end urban pleasures, such as fine dining, art galleries and expensive retail. Young people increasingly lived in cities, even as they worked in suburbs. Prices rose dramatically in urban cores and remained flat in the suburbs.
This description also helps to what is being lost by the pandemic-induced recession. Yes, some workers can do many basic parts of their jobs from home, and students can do some work with online courses. But the "richer information flow" of "face-to-face interactions" in both production and consumption is being lost for a time, and while the network of such interactions can certainly be rebuilt, it doesn't flip on and off like a light switch.

The resurgence of many cities has also brought with it a new group of problems, as Glaeser details.

One change is that cities do not seem to be functioning as ladders of opportunity. It may be that the extent of social and ethnic segregation--in terms of who you have significant interactions with on a typical day--can be higher in urban areas. Schools in urban core areas have often not recovered from their declines back in the 1970s. Glaeser writes: "It is a great paradox that cities appear to be forges of human capital for adults, but places where children seem to learn less productive knowledge."

Cities are also places of growing income inequality. Skilled workers in a large city or a downtown typically earn more than workers of similar skill outside those locations, but unskilled workers often do not have a similar pay boost from working in a city or downtown area.

Part of the issue may be government regulation of lower-skilled entrepreneurs. As Glaeser trenchantly notes:
Somewhat oddly, much of America appears to regulate low human capital entrepreneurship much more tightly than it regulates high human capital entrepreneurs. When Mark Zuckerberg started Facebook in his Harvard College dormitory, he faced few regulatory hurdles. If he had been trying to start a bodega that sold milk products three miles away, he would have needed more than ten permits. One question is whether the inequality that persists in America’s system is exacerbated by the legal and regulatory system.
And of course, the extremely high cost of housing in a number of economically strong urban areas makes it very hard for the middle-class, let alone those with lower skill levels, to pay the rent. Glaeser says:
For much of the post-war period, many urbanites could find housing that cost substantially less than construction costs even in successful cities (Glaeser and Gyourko 2005a). Housing depreciates, like cars and clothing, and so poorer urbanites could find older apartments in less fashionable neighborhoods that cost less. Filtering models predict that neighborhoods go through transitions, and that the rich would live in a newer, nicer areas but the poor occupy older, more dilapidated areas. The rich vacate areas as they depreciate and then move to a new area that had been built with higher-quality housing. Apparently, this model appears to have broken down after 1970, probably because of regulation and increased neighborhood opposition to redevelopment.
There is a persistent theme in American culture of moving to the big city, finding a low-level job, and working your way up. But US cities have become places where it's more costly to move in because the rent looks unthinkably high, and harder to find that low-skilled job, and then harder to move up unless you develop a high skill level. Add traffic congestion, and concerns about poor schools and crime, and moving to the big city doesn't look so attractive.

Glaeser argues that today's urban problems often reflect poor performance by local governments, who when it comes to housing markets, labor issues, schools, and other areas, have in recent decades often focused on blocking change or supporting insider groups. He writes:
Why has urban success been accompanied by so much discontent? The most natural explanation is that the success of private enterprise in cities has not been accompanied by sufficient development of public capacity. The public sector has often focused on limiting urban change, rather than working to improve the urban experience. In many cases, this focus reflects the political priorities of empowered insiders. ... There are many good things about citizen empowerment, but the most empowered citizens tend to be longer-term residents with more resources. Those citizens do not internalize the interests of people who live elsewhere and would want to come to the city. Consequently, their political actions are more likely to exclude than to embrace.

Monday, March 30, 2020

Aging, the Demographic Transition, and the Necessary Adjustments

David E. Bloom has been thinking a lot about aging. Last fall he edited Live Long and Prosper? The
Economics of Ageing Populations, a free ebook with 20 short essays summarizing a range of research on the topic (October 2019, VoxEU.org, registration required). Then Bloom contributed the lead article, "Population 2020: Demographics can be a potent driver of the pace and process of economic development," in the most recent issue of Finance & Development (March 2020, pp. 4-9).  At the moment, of course, a primary concern is that older people may be more vulnerable to the spread of COVID-19. But more broadly, a shift in the distribution of ages across society will have broad consequences for social institutions and government policies.

Here's a figure from Andrew Scott, in his F&D essay "The Long, Good Life," which gives as sense of the shift. The horizontal axis of the figure shows the expansion of population over time. The vertical axis shows the shift in aging. Thus, the shaded area for 1950 is narrower (fewer people) and more pointed near the top (fewer older people). The time periods that follow get wider (more people) and also develop "shoulders," representing a population where more people stay older for longer.



Bloom explains the underlying patterns, including the "demographic transition" and the graying population, in his F&D essay:

In many developing economies, population growth has been associated with a phenomenon known as the “demographic transition”—the movement from high to low death rates followed by a corresponding movement in birth rates.
For most of human history, the average person lived about 30 years. But between 1950 and 2020, life expectancy increased from 46 to 73 years, and it is projected to increase by another four years by 2050. Moreover, by 2050, life expectancy is projected to exceed 80 years in at least 91 countries and territories that will then be home to 39 percent of the world's population. ... Cross-country convergence in life expectancy continues to be strong. For example, the life expectancy gap between Africa and North America was 32 years in 1950 and 24 years in 2000; it is 16 years today. ...
In the 1950s and 1960s, the average woman had roughly five children over the course of her childbearing years. Today, the average woman has somewhat fewer than 2.5 children. This presumably reflects the growing cost of child-rearing (including opportunity cost, as reflected mainly in women’s wages), increased access to effective contraception, and perhaps also growing income insecurity. ... Between 1970 and 2020, the fertility rate declined in every country in the world. ... 
If the population’s age structure is sufficiently weighted toward those in prime childbearing years, even a fertility rate of 2.1 can translate into positive population growth in the short and medium term, because low fertility per woman is more than offset by the number of women having children. This feature of population dynamics is known as population momentum and helps explain (along with migration) why the populations of 69 countries and territories are currently growing even though their fertility rates are below 2.1.

The result of this demographic transition is a population with a rising number and share of elderly. Bloom writes:
Population aging is the dominant demographic trend of the twenty-first century—a reflection of increasing longevity, declining fertility, and the progression of large cohorts to older ages. Never before have such large numbers of people reached ages 65+ (the conventional old-age threshold). We expect to add 1 billion older individuals in the next three to four decades, atop the more than 700 million older people we have today. Among the older population, the group aged 85+ is growing especially fast and is projected to surpass half a billion in the next 80 years. This trend is significant because the needs and capacities of the 85+ crowd tend to differ significantly from those of 65-to-84-year-olds.
Although every country in the world will experience population aging, differences in the progression of this phenomenon will be considerable. Japan is currently the world leader, with 28 percent of its population 65 and over, triple the world average. By 2050, 29 countries and territories will have larger elder shares than Japan has today. In fact, the Republic of Korea’s elder share will eventually overtake Japan’s, reaching the historically unprecedented level of 38.1 percent. Japan’s median age (48.4) is also currently the highest of any country and more than twice that of Africa (19.7). But by 2050, Korea (median age 56.5 in 2050) is also expected to overtake Japan on that metric (54.7).
Three decades ago,the world was populated by more than three times as many adolescents and young adults (15- to 24-year-olds) as older people. Three decades from now, those age groups will be roughly on par.
I won't try here to summarize all the discussions in the F&D and the ebook. Instead, I'll just list the tables of contents below. But here are a few thoughts: 
  • Consider your mental picture of an extended family gathering. Maybe it's a holiday with grandparents, parents, and children. Maybe it's a family reunion with a larger group of aunts, uncles, and cousins. Now in thinking about that family reunion, think about it  being much more common to including five generations: that is, from great-great-grandparents down to children. In addition, think about there being fewer people in each generation, as a result of fewer children. The "family tree" is going to look taller and skinnier.  
  • As we shift from (in Bloom's calculation) a world where there are three times as many young people as elderly to a world where those populations are equal, public institutions will also shift to meet the needs of the elderly. The design of parks, libraries, public transit, city streets, shopping malls, and much more will evolve to reflect more emphasis on the needs and desires of the elderly. On the other side, schools and education will be a shrinking part of what government does. 
  • Caring for the elderly who need a range of support from an occasional in-home visit to living in a full-care institution is going to be a growth industry, needing both additional workers and technological innovations. This will be especially true as the number of extremely elderly people rises--often defined now as those over 85, but in the future perhaps defined as those over 100. In addition, the elderly will have had fewer children, and thus are less likely to have access to within-family support. 
  • It will be important for the workplace to shift in ways that provide jobs with the flexibility and interest to appeal to at least the "young elderly," who might otherwise just choose to retire completely.  
  • Government spending on programs to support the elderly, including pensions and health care, are going to rise dramatically in size.
  • All over the world, including the US, it's time to start phasing back the age of retirement at which people become eligible for pension plans. Exactly how that is done, and what kind of flexibility is available for retiring earlier or later, is open for discussion. But an expectation that retirement ages will in general be later is a fundamental step in making government-provided social security or pensions sustainable in the long run.
  • Older people tend to save less, as they draw down their retirement accounts, but also to look for less risky and volatile investments (more bonds, less stock market).
  • Keeping older people healthy and functioning later in life will be an urgent need, both for the people themselves and also to reduce the need for outside support. 
Following Bloom's lead-off essay, other essays on this topic in the March 2020 F&D include: 

In the e-book, the Table of Contents is:

1) "The what, so what, and now what of population ageing," by David E. Bloom

Part I: Implications of Population Ageing: The 'So What'

2) "Who will care for all the old people?" by Finn Kydland and Nick Pretnar
3) "Employment and the health burden on informal caregivers of the elderly," by Jan M. Bauer and Alfonso Sousa-Poza
4) "Ageing in global perspective," by Laurence J. Kotlikoff
5) "What do older workers want?" by Nicole Maestas and Michael Jetsupphasuk
6) "The flip side of "live long and prosper": Noncommunicable diseases in the OECD and their macroeconomic impact," by David E. Bloom, Simiao Chen, Michael Kuhn and Klaus Prettner
7) "Macroeconomic effects of ageing and healthcare policy in the United States," by Juan Carlos Conesa, Timothy J. Kehoe, Vegard M. Nygaard and Gajendran Raveendranathan
8) "Global demographic changes and international capital flows," by Weifeng Liu and Warwick J. McKibbin
9) "Ageing into risk aversion? Implications of population ageing for the willingness to take risks," by Margaret A. McConnell and Uwe Sunde
10) "Life cycle origins of pre-retirement financial status: Insights from birth cohort data," by
Mark McGovern
11) "A longevity dividend versus an ageing society," by Andrew Scott

Part II: Solutions and Policies: The 'Now What'

12) "Understanding 'value for money' in healthy ageing," by Karen Eggleston
13) "Healthy population ageing depends on investment in early childhood learning and development," by Elizabeth Geelhoed, Phoebe George, Kim Clark and Kenneth Strahan
14) "Financing health services for the Indian elderly: Aayushman Bharat and beyond," by Ajay Mahal and Sanjay K. Mohanty
15) "Cutting Medicare beneficiaries in on savings from managed healthcare in Medicare," by Thomas G. McGuire
16) "Macroeconomics and policies in ageing societies,"by Andrew Mason, Sang-Hyop Lee, Ronald Lee and Gretchen Donehower
17) "Population ageing and tax system efficiency," by John Laitner and Dan Silverman
18) "Means-tested public pensions: Designs and impact for an ageing demographic," by George Kudrna and John Piggott
19) "Pension reform in Europe," by Axel Börsch-Supan
20) "Happiness at old ages: How to promote health and reduce the societal costs of ageing," by Maddalena Ferranna


Saturday, March 28, 2020

An Economist's First Tryst with Benefit-Cost Analysis

Célestin Monga has had an eminent career as a research economist at the World Bank, as Managing Director of the UN Industrial Development Organization, as Chief Economist and vice-president at the the African Development Bank, and now as a Senior Economic Adviser at the World Bank. Here, he tells of that intimate special moment in the life of any economist--that first encounter with cost-benefit analysis. 

(I'm quoting here from Monga's "Comment" (pp. 77-94) in response to an essay by Amartya Sen from The State of Economics, The State of the World, published in 2019 by MIT Press.)  
I still remember vividly the strange mix of excitement and bewilderment that overwhelmed me in my high school years when our professor of accounting taught us the fundamentals of benefit-cost analysis. I immediately went to my dormitory and spent most of the evening trying to apply this powerful technique, not to assess whether the advantages of a hypothetical investment project were likely to outweigh its drawbacks, but to evaluate my own life prospects. Benefit-cost analysis seemed like a rigorous and revealing tool to examine whether my minuscule and uncertain existence was a "profitable" venture, or at least a worthwhile escapade that deserved to be continued. Of course, the few friends to whom I confided this found it a ludicrous idea. ... They were right: ... But so what? I kept running the numbers. ... 
I also had to decide how to imagine and estimate the prospective benefits and costs of my entire life to come. Using my own personal value scale, I calculated the costs as the amount of compensation required to exactly offset negative consequences of being alive for 50 or so years of life expectancy ahead. The compensation was the monetary amount required that would leave me just as well off as before engaging in this exercise. Benefits were measured by my willingness  to stay alive and enjoy all the things and emotions that I could reasonably expect for the decades ahead. Knowing that, in the end, life always results in death, typically following either an abrupt and tragic event like a car or airplane crash, or a long and painful illness, I could not find many benefits show present and expected value could match and compensate for the pains and disappointments of the costs. The results of my benefit-cost analysis were not very promising: Taking into consideration all current and expected streams of good and bad news, life did not appear to be a "profitable" investment. 
Shocked by the outcomes, I quick did some sensitivity analysis to check the robustness of my findings: No  matter what discount rates I chose, the calculations still yielded disappointing numbers to the question of whether life was a worthwhile venture. This was all the more puzzling because I actually loved many aspects of my life. Not knowing what to do with the analyses, I concluded one should either doubt the validity of certain measurement instruments or our ability to use them "objectively," or radically give more weight to whatever we define as "positive" outcomes for our actions or inactions, or accept the very probable hypothesis that happiness may be an illusion but those who choose to live should learn to ignore its downsides. I could only forget the outcomes of my own study by learning to radically change whatever assumptions I used in carrying it out. "Live is impossible without the ability to forget," philosopher Emil Cioran once said. But some memories are just to long-lasting to ever be erased. 
Monga's reminisce serves as a reminder of teenage feelings about the world. It also illustrates that although benefit-cost analysis has a useful place in comparing certain limited sets of choices, the method does not contain solutions to the mysteries of life. However, if you are a young person who finds yourself tempted to carry out a benefit-cost analysis of your own life, you may wish to consider seriously a career as an economist. 

Friday, March 27, 2020

Value of a Statistical Life: Where Does It Come From?

One of the (many) questions that causes economists to pull their hair out takes the general form: "How can you economists even possibly try to weigh economic costs against the value of a life saved?" Even worse, the question is often delivered in a triumphalist tone of a deeper moral truth being unveiled.

But in the real world, people and governments actually weigh economic costs against the value of a life saved all the time. Certain jobs that pose a greater risk to life and limb also tend to pay more than jobs for similarly qualified workers without such risks. Those who take such jobs, or don't take them, are in part placing an economic value on a greater risk of losing their life. Many government regulations, from setting speed limits on the roads to health and safety standards for food, could be tightened in a way that would save more lives but impose greater costs, or loosened in a way that would save fewer lives but impose lesser costs. Deciding where to set such regulations will necessarily involve a decision about how much it's worth paying to reduce the risk of someone losing their life.

Thus, the relevant question is not "how" to put a monetary value on life or "why would anyone ever want" to put a monetary value on life. The discussion starts from the fact that people and governments are already putting a monetary value on life, albeit often implicitly, by the actual real-world decisions they make.  When economists say that the "value of a statistical life" is about $10 million, they are not just pulling a number out of the air. Instead, they are only pointing out the monetary values that people are already using.

Thomas J. Kniesner and W. Kip Viscusi offer a readable overview of the evidence behind such decisions in "The Value of a Statistical Life," which was published in June 2019 in the Oxford Research Encyclopedia, Economics and Finance. (If for some reason you don't have access, a version of the paper is available on SSRN.)

As Kneiser and Viscusi point out, evidence about the economic value that people place on a higher or lower risk of losing their life can come from several sources: "revealed preference" studies that look at choices people make about jobs or products with different risks, or "stated preference" studies that involve survey data. To understand the intuition here, it's important to recognize that they studies are not asking a question like: "How much money would we need to pay you before we kill you?" The "value of a statistical life" is about changes in risk. They write:
Suppose further that ... the typical worker in the labor market of interest, say manufacturing, needs to be paid $1,000 more per year to accept a job where there is one more death per 10,000 workers. This means that a group of 10,000 workers would collect $10,000,000 more as a group if one more member of their group were to be killed in the next year. Note that workers do not know who will be fatally injured but rather that there will be an additional (statistical) death among them. Economists call the $10,000,000 of additional wage payments by employers the value of a statistical life. It is also the amount that the same group of workers would be willing to pay via wage reductions to have safer jobs where one fewer of their group would be fatally injured or ill. In that sense the VSL measures the willingness of workers to implicitly pay for safer workplaces and can be used to calculate the benefits of life-saving projects by private sector managers and government policymakers.
Studies of specific jobs that compare risks of death and pay will come up with a range of numbers; after all, jobs differ in many ways other than just their mortality risk. Thus, in a 2018 study, Viscusi looked at 1,025 estimates of the value of a statistical life drawn from 68 publications. He looked both at the total group, and then also at a "best-set" subgroup of the estimates that used what he viewed as more reliable methods. He found: "The all-set mean VSL is $12.0 million and the best-set sample mean is $12.2 million, where all estimates are in $2015. The median values are somewhat lower—$9.7 million for the all- set sample and $10.1 million for the best-set sample."

Of course, not everyone will put the same value on reducing mortality risk, and those of different ages and income levels, for example, will prefer different values. But for evaluating a broad government regulation that affects a broad cross-section of the population, using an overall number makes sense.

Another branch of the literature looks at purchases of certain goods or services. For example, how much is the price of a house affected by being in a high-crime area or near a large source of air pollution? How does the price that people pay for bike helmets or smoke detectors compare to the reduction in risk from such purchases? Again, different studies have a range of answers: again, an estimate of $10 million as the value of a statistical life seems plausible.

Other studies have taken an approach that uses detailed scenario-setting surveys. For example, the questionnaire may lay out a starting scenario, which includes the health risk expressed in various ways, like the chance of living to 100 years of age or the annual risk of being killed in the next year by cancer or in a car accident. Then the follow-up question offers other scenarios, with a range of costs expressed in terms like expected changes in prices or taxes paid, and different health risks. Naturally, the construction and interpretation of such surveys can be controversial, and sometimes the answers seem crazy-high or crazy-low. But an OECD study a few years ago suggested, based on an overview of these studies, that using $3.6 million as the value of a statistical life was plausible.

When it comes to public policy, Kneiser and Viscusi note: "Most U.S. government agencies have now adopted VSL estimates in a similar range consistent with the economics literature." The point out that the  U.S. Department of Transportation (2016) uses $9.4 million as the value of a statistical life, compared with $9.7 million for Environmental Protection Agency and $9.6 million for the U.S. Department of Health and Human Services.

It's easy enough to come up with questions about the value of a statistical life. But again, it is simply a fact that people and governments make decisions all the time about weighing health and safety against costs. Blaming the economists for doing the calculations to figure out what values are actually being places on a statistical life is like blaming the bathroom scale, or perhaps the laws of gravity, when it tells you that you could stand to lose a few pounds.

In the midst of the coronavirus pandemic, an obvious question is what a value of $10 million for the value of a statistical life means about the ongoing strategy of causing a recession for the sake of protecting public health. The multiplication is straightforward. Imagine that the steps being taken to contain the virus save 500,000 US lives. With those lives valued at $10 million, a social cost of up to  $5 trillion in lost output would be justified. For comparison, US GDP is about $21 trillion. If steps taken to contain the virus save 50,000 lives, then a social cost of up to $500 billion in lost output would be justified. This calculation is so quick-and-dirty, and leaves out so much, that I hesitate even to include it  here. It does suggest to me that in these benefit-cost terms, it's plausibly worth a recession to contain the virus, even a deep-but-short recession. It also suggests that if looking at how health risks  have been valued by actual people and governments in the past, a long-term recession or depression would not be a price worth paying to contain the virus.

For some previous posts and articles on the value of a statistical life, and its cousin the "quality-adjusted life-year," see:

Wednesday, March 25, 2020

Does the US Tax Code Favor Automation Over Jobs?

Imagine a company that is considering two possible ways to improve efficiency and productivity. One is to pay for many of its employees to go through a training program to learn new sets of useful skills. The other is to pay for new equipment that will replace many of the employees. Daron Acemoglu, Andrea Manera, and Pascual Restrepo argue that the US tax code tends to favor the second option. The technical version of their argument, "Does the U.S. Tax Code Favor Automation?" is published in most recent Brookings Papers on Economic Activity (Spring 2020, a short readable overview of the paper is also available at the link).  They write (citations and footnotes omitted):
The most common perspective among economists is that even if automation is contributing to declining labor share and stagnant wages, the adoption of these new technologies is likely to be beneficial, and any adverse consequences thereof should be dealt with appropriate redistributive policies (and education and training investments). But could it be that the extent of automation is excessive, meaning that US businesses are adopting automation technologies beyond the socially optimal level? If this were the case, the policy responses to these major labor market trends would need to be rethought.
There are several reasons why the level of automation may be excessive. Perhaps most
saliently, the US tax system is known to tax capital lightly and provide various subsidies
to the use of capital in businesses. In this paper, we systematically document the asymmetric taxation of capital and labor in the US economy in the US tax system labor is much more heavily taxed than capital. ...
Mapping the complex range of taxes in the US to effective capital and labor taxes is not trivial. Nevertheless, under plausible scenarios (for example, depending on how much of healthcare and pension expenditures are valued by workers and the effects of means-tested benefits), we find that labor taxes in the US are in the range of 25.5-33.5%. Effective capital taxes on software and equipment, on the other hand, are much lower, about 10% in the 2010s and even lower, about 5%, after the 2017 tax reforms. We also show that effective taxes on software and equipment have experienced a sizable decline from a peak value of 20% in the year 2000.3 A major reason explaining this trend in capital taxation is the increased generosity [of] depreciation allowances ...
 I should emphasize that this  paper is part of an ongoing research effort by these authors to think about interactions between automation and jobs. I have blogged about a previous entry in this line of research in "Is Something Different This Time About the Effect of Technology on the Labor Market?" (May 6, 2019). I discussed there a paper by Daron Acemoglu and Pascual Restrepo titled  "Automation and New Tasks: How Technology Displaces and Reinstates Labor."

In that paper, they suggest a framework in which automation can have three possible effects on the tasks that are involved in doing a job: a displacement effect, when automation replaces a task previously done by a worker; a productivity effect, in which the higher productivity from automation taking over certain tasks leads to more buying power in the economy, creating jobs in other sectors; and a reinstatement effect, when new technology reshuffles the production process in a way that leads to new tasks that will be done by labor. In this model, the effect of automation on labor is not predestined to be good, bad, or neutral. It depends on how these three factors interact.

In that context, the authors of the current paper suggest the theoretical possibility of an "automation tax," defined as "a higher tax on the use of capital in tasks where labor has a comparative advantage." They would combine this with a lower tax on other forms of capital, as well as on labor. In my own words, they are proposing that the tax code encourage the kind of automation that complements what workers do in a way that leads to sharp increases in productivity and output, but that the tax code not encourage the kind of automation that mostly just replaces workers with a real but only modest cost savings for the employer.

Of course, it's reasonable to note that a theoretical economic model can just create variables for these two kinds of automation, while a real world policy might face some difficult challenges in distinguishing between them. Still, the authors are trying to break out of a binary choice where automation is viewed as always good or always bad, and automation is instead being viewed as a range of choices that include automation that is more likely to be job-destroying or more likely to be job-creating. It feels to me like a potential distinction worth investigating.

Monday, March 23, 2020

James Madison on Why to Fill Out Your Census Form

I saw a news release from the US Bureau of the Census over the weekend that only about one-sixth of US households have responded to the 2020 Census so far. Here's what James Madison had to say, back in 1790, about why to fill out the form.

To set the stage, section 2 of the just-adopted US Constitution called for an enumeration of people to determine the number of members each state would have in the House of Representatives: "The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct." But when the bill to enact the first Census came up in 1790, James Madison (then a member of the House of Representatives) argued that although collecting additional information would add to the difficulties, both the legislators and citizens would proceed with "more light and satisfaction" when they "rest their arguments on facts, instead of assertions and conjectures." 

Our records of Congressional debates from that time do not quote exactly verbatim, but instead are paraphrased. The fuller comments attributed to Madison are below, but here's are some highlights of what he had to say on January 25 and then on February 2, 1790:
This kind of information, he [Madison] observed, all Legislatures had wished for; but this kind of information had never been obtained in any country. ... If the plan was pursued in taking every future census, it would give them an opportunity of marking the progress of the society, and distinguishing the growth of every interest. This would furnish ground for many useful calculations, and at the same time answer the purpose of a check on the officers who were employed to make the enumeration ... I take it, sir, that in order to accommodate our laws to the real situation of our constituents, we ought to be acquainted with that situation. ... If gentlemen have any doubts with respect to its utility, I cannot satisfy them in a better manner, than by referring them to the debates which took place upon the bills, intend, collaterally, to benefit the agricultural, commercial, and manufacturing parts of the community. Did they not wish then to know the relative proportion of each, and the exact number of every division, in order that they might rest their arguments on facts, instead of assertions and conjectures? ... We should have given less encouragement in some instances, and more in others; but in every instance, we should have proceeded with more light and satisfaction.
Afterword: Here is a fuller version of the comments attributed to Madison. Here is the paraphrase of Madison's comments for January 25, 1790:  
Mr. Madison observed, that they had now an opportunity of obtaining the most useful information for those who should hereafter be called upon to legislate for their country, if this bill was extended to as to embrace some other objects besides the bare enumeration of the inhabitants; it would enable them to adapt the public measures to the particular circumstances of the community. In order to know the various interests of the United States, it was necessary that the description of the several classes into which the community is divided should be accurately known. On this knowledge the Legislature might proceed to make a proper provision for the agricultural, commercial, and manufacturing, interests, but without it they could never make their provisions in due proportion. This kind of information, he observed, all Legislatures had wished for; but this kind of information had never been obtained in any country. He wished, therefore to avail himself of the present opportunity of accomplishing so valuable a purpose. If the plan was pursued in taking every future census, it would give them an opportunity of marking the progress of the society, and distinguishing the growth of every interest. This would furnish ground for many useful calculations, and at the same time answer the purpose of a check on the officers who were employed to make the enumeration; forasmuch as the aggregate number is divided into parts, any imposition might be discovered with proportionable ease.
And here is a fuller paraphrase of Madison's comments on February 2, 1790:
And I am very sensible, Mr. Speaker, that there will be more difficulty attendant on the taking the census, in the way required by the constitution, and which we are obliged to perform, than there will be in the additional trouble of making all the distinctions contemplated in the bill. The classes of people most troublesome to enumerate, in this schedule, are happily those resident in large towns, the greatest number of artisans live in populous cities, and compact settlements, where distinctions are made with great ease.
I take it, sir, that in order to accommodate our laws to the real situation of our constituents, we ought to be acquainted with that situation. It may be impossible to ascertain it as far as I wish, but we may ascertain it so far as to be extremely useful, when we come to pass laws, affecting any particular description of people. If gentlemen have any doubts with respect to its utility, I cannot satisfy them in a better manner, than by referring them to the debates which took place upon the bills, intend, collaterally, to benefit the agricultural, commercial, and manufacturing parts of the community. Did they not wish then to know the relative proportion of each, and the exact number of every division, in order that they might rest their arguments on facts, instead of assertions and conjectures? Will any gentleman pretend to doubt, but our regulations would have been better accommodated to the real state of the society than they are? If our decisions had been influenced by actual returns, would they not have been varied, according as the one side or the other was more or less numerous? We should have given less encouragement in some instances, and more in others; but in every instance, we should have proceeded with more light and satisfaction.
Finally, this post repeats some material from a post back in March 2017 about the importance of government statistical agencies. But with the 2020 Census upon us, the message seemed to bear repeating.

Friday, March 20, 2020

Can the Economy Just Be Put on Hold for a Few Months?

The COVID-19 pandemic is an intertwined public health and economic event. Here are some comments from Richard Baldwin and Beatrice Weder di Mauro: "The social distancing policies are purposefully inducing an economic slowdown. ... The recession, so to speak, is a necessary public health measure. ... The recession is a medical necessity. That’s a given. But governments can and should try to flatten the economic recession curve. ... The key is to reduce the accumulation of ‘economic scar tissue’ – reduce the number of unnecessary personal and corporate bankruptcies, make sure people have money to keep spending even if they are not working."

The comments are from the "Introduction" to an e-book containing 24 short essays edited by Baldwin and Weder di Mauro, and just released by VoxEU: Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes. This is a follow-up to the ebook they produced less than two weeks ago, which I commented on in "Some Coronavirus Economics" (March 11, 2020). The collection has a lot of interesting analysis about fiscal and monetary policy in the current setting, along with specific discussions of the European Union, the European Central Bank, Italy, Germany,  China, Singapore, South Korea, and so on. 

Here, I want to focus a bit on a theme that comes up in a number of the essays: the idea that sensible economic policy can put the economy in the freezer for a few months, and the pull the economy out of the freezer, thaw it out, and restart it. I find myself in the awkward position here of largely being in agreement with this policy as a short-run approach, and at the same time also feeling that the ultimate consequences of the policy are going to be more difficult than a number of authors are envisaging. 

As one example of this theme in the book, Luis Garicano writes: "[A]t this point standard demand management is useless. Governments do not want to stimulate economic activity —they are doing all they can to stop it (they are asking people to stay at home!). Instead, economic policy is needed to ensure that the economy survives a ‘freeze’ of (hopefully, no more than) three to six months." 

As another, Pierre-Olivier Gourinchas writes: 
[I]n the short run, flattening the infection curve inevitably steepens the macroeconomic recession curve. Consider China, or Italy: increasing social distances has required closing schools, universities, most non-essential businesses, and asking ost of the working-age population to stay at home. While some people may be able to work from home, this remains a small fraction of the overall labour force. Even if working from home is an option, the short-term disruption to work and family routines is major and likely to affect productivity. In short, the – appropriate – public health policy plunges the economy into a sudden stop. ... A modern economy is a complex web of interconnected parties: employees, firms, suppliers, consumers, banks and financial intermediaries… Everyone is someone else’s employee, customer, lender, etc. A sudden stop like the one described above can easily trigger a cascading chain of events, fuelled by individually rational, but collectively catastrophic, decisions. ... To a first-order approximation, I would consider that governments may need to provide income support on a scale roughly comparable to the output lost.  
Rather than quote from a number of other authors, I'll try to sum up the economic policy goal in my own words. Sure, in normal times, it makes sense to have a dynamic economy where firms rise and fall depending on whether customers are willing to pay for what they produce, and it makes sense for a churning labor market to reallocate workers back and forth across these companies. 

But the novel coronavirus isn't a normal time. There is no reason to believe that it is a socially useful mechanism for sorting out which firms should expand or contract, or that it is a useful mechanism for reallocating labor across the economy. Instead, the goal of public policy should be to prevent firms from needing to fire workers permanently as a response to the pandemic. If firms need to use short-term layoffs, then government can help to replace lost income until workers get the call-back to return to work. The financial sector should be encouraged or subsidized to hold back on calling in loans or foreclosures. In some broad sense, the costs of an economic "freeze" due the coronavirus pandemic should be socialized. In the US, for example, perhaps the debt/GDP of the federal government rises by 5 percentage points of GDP, as the government provides support for the lost income across the economy.  

As I said earlier, I broadly share this vision of aggressive government action to ease the immediate economic burden of the pandemic. If the issue was as simple as whether to cut interest rates and raise government debt by 5 percentage points, then the policy choice would be simple for me. But ti's not that simple. 

1) As a matter of public health, there's no guarantee that COVID-19 is a one-time event. For example, the Great Influenza Epidemic of 1918-20 happened over three cycles. There is some reason for concern that in a globalized world, outbreaks of pandemics may be more common and spread faster than in the past. Even as we plunge into the immediate economic rescue, it's worth asking: Is this a pattern of public health and economic actions we are planning to repeat each year for the next 2-3 years? If such pandemics recur more frequently, is the the set of policies we are planning to follow once or twice a decade into the foreseeable future? For example, if very large fiscal policy actions to soften the economic blow of pandemics are going to be a recurrent and standard policy moving ahead, governments need to start planning ahead for them. 

2) The proposed plan as a matter of practical politics and legislation, will be able to turn on the rescue and then turn it off? For example, Olivier Blanchard offers this advice in the book: "[S]pend what you must on crisis containment and commit to wind down everything once the crisis is over. Full stop." The commitment to turning off the support is easy to say, but often hard to do, and the "winding down" process can extend a considerable time. 

3) As another matter of practical politics, support for business often tends to favor the big and the prominent. For example, in a US context there has been talk of assisting the airlines, which are big prominent companies, often with unionized employees, and thus lots of political clout. However, the tourism, entertainment, and restaurant industry is made up of much smaller firms, although as a group, they employ vastly more people.

4) It's easy to say that companies shouldn't be forced into bankruptcy by the coronavirus, and by and large, I agree. But we all know that some economic actors are more prudent than others. Some companies take on very high levels of debt, and some don't. Some people make sure to an amount equal to several months of income on hand, and others (who have similar levels of income) are rolling over large credit-card bills and paying interest from month to month. Yes, the pandemic was unexpected, but some firms and people build in a cushion deal with the unexpected, and some don't. Whenever government steps in to help those who are blindsided by an unexpected event, it benefits those who were not prudent over those who tried to be.

5) Finally, I'll add that any economy is not just a machine that can be switched off and on again. Helping households and firms to limp through the economic effects of this pandemic is worth doing, even at high costs to governments. But the economic disruption has costs of its own. Some number of the projects that companies were working on will be forever uncompleted. Relationships within and across organizations have been disrupted. It seems unlikely that customers and supply chains will simply restart their old patterns after a substantial disruption. Trust in providers will be shaken, sometimes for good reasons and sometimes for trivial ones.  There may be jump-starts toward companies doing more work online, or encouraging telecommuting. In health care, there may be shifts toward virtual consultations, or pharmacists providing more services, or how tests and treatments are approved and used. There may be surge to online education: after all, if online education is good enough for grades and credit at Harvard, Stanford, doesn't that show that it could be online at lots of other places after the pandemic passes? These pattern shifts and many more will outlast the pandemic itself, and will cause economic disruptions and costs of their own.