Conversable Economist

But correlation isn\’t causation, of course, as economists (and this study) are quick to note. For example, say that there is a strong correlation between families in poverty and a lower education level for the parents. Perhaps a substantial share of the problems for children in poverty are not caused by lower family income, but by the lower education level of parents. If the root cause is lower parental education levels, then raising these families above the poverty line in terms of income won\’t have much effect on the long-term problems faced by children from these families.  

Neuroscientists have produced striking evidence of the effect of early-life economic circumstances on brain development. Drawing from Hanson et al. (2013), Figure 3-3 illustrates differences in the total volume of gray matter between three groups of children: those whose family incomes were no more than twice the poverty line (labeled “Low SES” in the figure); those whose family incomes were between two and four times the poverty line (“Mid SES”); and those whose family incomes were more than four times the poverty line (“High SES”). Gray matter is particularly important for children’s information processing and ability to regulate their behavior. The figure shows no notable differences in gray matter during the first nine or so months of life, but differences favoring children raised in high-income families emerge soon after that. Notably, the study found no differences in the total brain sizes across these groups—only in the amount of gray matter.\”

This study is again a correlation, not a proof of causality. As the report notes: \”However, the existence of these emerging differences does not prove that poverty causes them. This study adjusted for age and birth weight, but not for other indicators of family socioeconomic status that might have been the actual cause of these observed differences in gray matter for children with different family incomes.\” But with all due caution rigorously observed, it seems to me a highly suggestive correlation. 

Other studies look at the long-term effects of existing government programs that have raised income levels for poor families. Here\’s another example:

In their 2016 study of possible long-term effects of Food Stamp coverage in early childhood on health outcomes in adulthood, Hoynes, Schanzenbach, and Almond focus on the presence or absence of a cluster of adverse health conditions known as metabolic syndrome. In the study, metabolic syndrome was measured by indicators for adult obesity, high blood pressure, diabetes, and heart disease. Scores on these indicators of emerging cardiovascular health problems increased (grew worse) as the timing of the introduction of Food Stamps shifted to later and later in childhood (Figure 3-4). The best adult health was observed among individuals in counties where Food Stamps were already available when these individuals were conceived. Scores on the index of metabolic syndrome increase steadily until around the age of five.

Add all these kinds of studies and factors up, and you can obtain a rough-and-ready estimate a total cost of child poverty. 

Holzer et al. (2008) base their cost estimates on the correlations between childhood poverty (or low family income) and outcomes across the life course, such as adult earnings, participation in crime, and poor health. … Their estimates represent the average decreases in earnings, costs associated with participation in crime (e.g. property loss, injuries, and the justice system), and costs associated with poor health (additional expenditures on health care and the value of lost quantity and quality of life associated with early mortality and morbidity) among adults who grew up in poverty. … Holzer et al. (2008) make a number of very conservative assumptions in their estimates of earnings and the costs of crime and poor health. … All of these analytic choices make it likely that these estimates are a lower bound that understates the true costs of child poverty to the U.S. economy.

The bottom line of the Holzer et al. (2008) estimates is that the aggregate cost of conditions related to child poverty in the United States amounts to $500 billion per year, or about 4 percent of the Gross Domestic Product (GDP). The authors estimate that childhood poverty reduces productivity and economic output in the United States by $170 billion per year, or by 1.3 percent of GDP; increases the victimization costs of crime by another $170 billion per year, or by 1.3 percent of the GDP; and increases health expenditures, while decreasing the economic value of health, by $163 billion per year, or by 1.2 percent …

McLaughlin and Rank (2018) build on the work of Holzer and colleagues by updating their estimates in 2015 dollars and adding other categories of the impact of childhood poverty on society. They include increased corrections and crime deterrence costs, increased social costs of incarceration, costs associated with child homelessness (such as the shelter system), and costs associated with increased childhood maltreatment in poor families (such as the costs of the foster care and child welfare systems). Their estimate of the total cost of childhood poverty to society is over $1 trillion, or about 5.4 percent of GDP. …  They do make it clear that there is considerable uncertainty about the exact size of the costs of childhood poverty. Nevertheless, whether these costs to the nation amount to 4.0 or 5.4 percent of GDP—roughly between $800 billion and $1.1 trillion annually in terms of the size of the U.S. economy in 2018—it is likely that significant investment in reducing child poverty will be very cost-effective  over time.

I don\’t propose spending $160 billion lightly. But I will point out that the expansion of the health insurance under the Patient Protection and Affordable Care Act of 2010 costs the US government over $100 billion per year.  Similarly, the costs of the Tax Cuts and Jobs Act passed in 2017 are projected to have an average cost of $100 billion per year (or more?) In short, our political system does seem fully capable of belching up expenditures of this size when the stars are properly aligned. 

As the report points out, some American cousins have taken the plunge to reducing child poverty by half.

The United States spends less to support low-income families with children than peer English-speaking countries do, and by most measures it has much higher rates of child poverty. Two decades ago, child poverty rates were similar in the United States and the United Kingdom. That began to change in March 1999, when Prime Minister Tony Blair pledged to end child poverty in a generation and to halve child poverty in 10 years. Emphasizing increased financial support for families, direct investments in children, and measures to promote work and increase take-home pay, the United Kingdom enacted a range of measures that made it possible to meet the 50 percent poverty reduction goal by 2008—a year earlier than anticipated. More recently, the Canadian government introduced the Canada Child Benefit in its 2016 budget. According to that government’s projections, the benefit will reduce the number of Canadian children living in poverty by nearly half.

Personally, I would be a lot more comfortable with the extent of US inequality if the child poverty rate was considerably lower, and thus the starting points for American children were closer together. 

Binder and Bound set the stage: 

Here\’s a figure from their paper showing labor force participation by level of education for \”prime-age\” males in the 25-54 age group. In the late 1960s, prime-age men of all education levels had very high labor force participation. But it has sagged over time for all education levels, and sagged the most for those with lower education levels. 

This drop-off in labor force participation has been accompanied by a wave of other symptoms, as discussed in the paper by Coile and Duggan. As one example, consider mortality rates for prime-age men, using their table. 

The overall mortality rate for men (bottom row) dropped dramatically from 1980 to 2000, but barely budged from 2000-2016. In particular, from 2000-2016, the mortality rate rose for men age 25-34 and for white men in the 25-age group as a whole. Looking at cause of death, there are big falls in death rates for prime-age men from heart disease and cancer in the 1980s and 1990s, but much smaller falls since then. Meanwhile, death rates for this group from accidents, suicides, and homicides went up from 2000-2016. Data on cause of death doesn\’t include education level, but the authors go on to show that in areas with lower education levels, these rises in death rates were more pronounced. 

When Coile and Duggan look instead at reporting of health problems, they find:

\”There is a steep health gradient with respect to education—within each age group, the share in fair or poor health is roughly 2.5 times as large for men with a high school education or less than for men with some college or more. Men with less education are similarly more likely to report having a work-limiting disability, limitations in physical activity or ADLs/IADLs [Activities of Daily Living or Instrumental Activities of Daily Living], and obesity …  Men’s health … is getting worse over time. … [T]he fraction of men reporting a health problem is higher in 2015 than in 2000 in nearly every case.\”

Coile and Duggan look at a variety of other patterns for prime-age men, focusing on lower skill levels where the data makes it possible. For example, they note the sharp rise in incarceration rates for men from 1980 to 2000. The pattern that emerges is that the incarceration rate for men in the 45-54 age group is higher in 2016 than in 2000, reflecting large numbers of younger men sentenced to prison in the 1980s and 1990s. However, the incarceration rate for men in the 25-34 and 35-44 age group is generally down in 2016 compared to 2000. As one example, the incarceration rate of black men ages 25-34 was 5.5% in 1980, 12.8% in 2000, and 7.4% in 2016. 

Marriage rates have been on a generally downward trend as well, although the drop-off from 2000 to 2016 is a lot smaller than the fall for the 1980-2000 period. Here\’s an illustrative table from Coile and Duggan: 

In some general way, all of these factors seems to combine into a shadowy picture. Low-skill men are working less, reporting worse health, were for a time more likely to be locked up, and seem less likely to form family ties. How do these factors connect? 

Binder and Bound focus on the task of explaining the drop in labor force participation. They argue that the reduced demand for labor of low-skill but prime-age men (perhaps because of shifts in technology or international trade) isn\’t nearly enough to explain the drop in their labor force participation.  They also offer back-of-the envelope estimates that while higher disability rates may affect men in the 45-54 age bracket, they aren\’t likely to explain less labor force participation for the younger prime-age men. They write:

On its own, falling labor demand does not sufficiently explain the secular decline in less-educated male labor-force participation—at least, not without allowing for substantial adjustment frictions in the long run as well as the short run. Rising access to Disability Insurance is at most a partial explanation for the 45–54 year-old group and matters quite little for younger men and for high school dropouts. Rising exposure to prison may be a significant factor for dropouts and for blacks without college education, but labor-force participation for these groups began declining decades before prison populations skyrocketed. Certainly no single explanation can sufficiently explain the decline, and even in combination, the explanations appear insufficient.

We suspect that there is another factor at play. We will argue that the prospect of forming and providing for a new family constitutes an important male labor supply incentive; and thus, that developments within the marriage market can influence male labor-force participation. A decline in the formation of stable families produces a situation in which fewer men are actively involved in family provision or can expect to be involved in the future. This removes a labor supply incentive; and the possibility of drawing support from one’s existing family … creates a feasible labor-force exit.

The paper by Edin, Nelson, Cherlin and Francis is by a group of sociologists, based on in-depth interviews with working-class men who have children but are not married to the mothers and do not live with them. They argue that low-skilled men are often trying to renegotiate their relationship to jobs, family and religion–but that many of them are in a social setting where these attempts lead to \”haphazard lives.\” They write (citations omitted):

[W]e show that working-class men are not simply reacting to changes in the economy, family norms, or religious organizations. Rather, they are attempting to renegotiate their relationships to these institutions by attempting to construct autonomous, generative selves. For example, these men’s desire for autonomy in jobs seems rooted in their rejection of the monotony and limited autonomy that their fathers and grandfathers experienced in the workplace, along with a new ethos of self-expression. Similarly, these working-class men focus on their ties to their children even when they have little relationship with the children’s mothers, and they seek spiritual fulfillment even though they disdain organized religion. … In sum, these working-class men show both a detachment from institutions and an engagement with more autonomous forms of work, childrearing, and spirituality … . Autonomy refers to independent action in pursuit of personal growth and development. Personal growth has come to be highly valued among middle class Americans but until recently has not been associated with the working class. … [P]ast scholarship typically assumed that such forms of action would usually only be found among those so materially comfortable that they needn’t spend time worrying about their economic circumstances …

Our interviews strongly suggest that the autonomous, generative self that many men described is also a haphazard self. For example, vocational aspirations usually remain nebulous and tentative, rarely taking the form of an explicit strategy. In the meantime, career trajectories are often replaced by a string of random jobs. These men typically transitioned to parenthood more by accident than design, and in the context of tenuous romantic relationships. … Religious community and a systemic belief system have been replaced by a patched-together religious identity that holds little sway over behavior, especially as it is divorced from the communal aspects of faith that have adhered working-class men to a set of behavioral norms. …

The optimistic reading of the developments we have described is that workingclass men are now sharing in the autonomy and generativity that was largely the province of middle- and upper-class men in previous generations. Moreover, the interest they show in being involved as fathers and in helping others could represent a widening of the boundaries of masculinity in ways that are more consistent with contemporary family and work life. The pessimistic reading is that these men are pursuing goals that they are unlikely to achieve due to their lack of social integration. They must find their way without ties to steady work, stable families, and organized religion. Without social support, their chances of success diminish. Those who fail to achieve the autonomous, generative selves they crave will have little to fall back on and few people to prevent them from sinking into despair.

In other words, the problems of low-skilled men in US society are certainly not just a matter of income, and not just a matter of having a job, either. Instead, they are related to a more wide-ranging disconnectness, which shows up across many domains of behavior and outcomes.  

Arguments over taxes often make fairness claims about the share of taxes paid by various income groups. Whatever one\’s ultimate conclusions about what should happen, it\’s useful to start from teh basis of what is actually happening.

It\’s common to hear a complaint that those with high incomes are evading federal taxes. Some do, of course. It\’s a big country. If a very rich person puts all their money into tax-exempt bonds, with the associated lower interest rates for being tax-free, they won\’t pay taxes on that income. But on average, those with higher incomes do pay a much larger share of taxes. Robert Bellefiore offers a couple of illustrative graphs. The first figure focuses only on federal income taxes.

The second figure includes the share of all federal taxes: that is, income, payroll, corporate (as attributed to individuals who benefit from corporate profits), excise taxes on gasoline, tobacco, and alcohol, and so on. Again, those with higher income levels pay a larger share of total federal taxes.

One can of course still argue that the share of taxes paid by those with high incomes should be larger. But again, arguing that those with high incomes don\’t already pay a larger share of federal taxes is not a true statement.

What about taxes paid at the very tip-top of the income distribution? Erica York offers this figure on the average tax rates paid by the top 0.1%. To be clear, the \”average\” rate rate is the actual share of income paid in taxes, which is different from the \”marginal\” tax rate charged on the highest $1 of income earned. Back in the 1950s, the highest marginal income tax rates sometimes reached 90%. The fact that the average tax rate is so much lower tells you that those very high marginal tax rates were largely for show, in the sense that they didn\’t actually apply to very much income. York writes: \”The graph below illustrates the average tax rates that the top 0.1 percent of Americans faced over the last century, based on research from Thomas Piketty, Emmanuel Saez, and Gabriel Zucman. The blue line includes the impact of all federal, state, and local taxes on individual income, payroll taxes, estates, corporate profits, properties, and sales. The purple line shows income taxes only, including federal, state, and local.\” The overall pattern is while effective tax rates on the top 0.1% were higher in the 1950s, they haven\’t shown much long-term trend one way or the other in the last half-century or so.

When listening to arguments over tax policy, it\’s common to hear complaints about whether deductions should be limited for purposes like mortgage interest, state and local taxes, or charitable contribution. It\’s useful to remember that those deductions don\’t apply to most taxpayers. Erica York explains: \”In 2016, barely a quarter of households with adjusted gross income (AGI) between $40,000 and $50,000 claimed itemized deductions when filing their taxes. In contrast, more than 90 percent of households making $200,000 and above itemized their deductions.\” One effect of the 2017 tax reform law is that the number of taxpayers who find it useful to itemize deductions will drop by as much as 60%.

The share of total federal taxes paid by those with high incomes has been rising over time. Part of that change is because the share of those who owe zero in federal income tax has been rising over time. Robert Bellefiore provides a graph. One main reason for the rise share of taxpayers who owe zero is the expansion of refundable tax credits aimed at those with lower income, including the Earned Income Tax Credit and the Child Credit. You can also see the share of those with zero income taxes owed rises in the Great Recession.

In a different post, Robert Bellefiore offers a chart showing the overall effects of federal tax and transfer policy on the share of income received by different groups. He writes: \”The lowest quintile’s income nearly doubles, while the second and middle quintiles experience relatively smaller increases in income. The fourth quintile’s income share remains constant, and only the highest quintile has a lower share of income after taxes and transfers. The top 1 percent’s share of income, for example, falls from 16.6 percent to 13.2 percent.\”

Again, one can argue that the amount of redistribution should be larger. But it would be untrue to argue that a significant amount of redistribution–like doubling the after-taxes-and-transfers share of the lowest quintile–doesn\’t already happen.

The authors suggest that one way in which modern technology is different from the 1890s is that it does not reward or encourage this kind of extreme division of labor. They write: 

The massive division of labor documented front and center in the Hand and Machine Labor study dramatically affected the nature of the human capital investment decision facing successive cohorts of American workers contemplating whether to enter the manufacturing sector. Earlier in the nineteenth century, the human capital investment problem such workers faced was mastering the diverse set of skills associated with most or all of the tasks involved in making a product, along with managing the affairs of a (very) small business, an artisan shop. The human capital investment problem facing the prospective manufacturing worker in the 1890s was quite different. There was little or no need to learn how to fashion a product from start to finish; mastery of one or two tasks
would do, and such mastery might be gained quickly on the job. The more able or ambitious might gravitate to learning new skills, such as designing, maintaining, or repairing steam engines, or clerical/managerial tasks, the demand for which had grown sharply as average establishment size increased over the century.

For many decades in the twentieth century, specialization was economically beneficial to workers—the costs of learning skills were relatively modest and the return on the investment—a relatively secure, highly paid job in manufacturing—made that investment worthwhile. The prospect of widespread automation has arguably changed this calculus. No single “job” is safe and the optimal investment strategy may be very different—a suite of diverse, relatively uncorrelated skills as insurance against displacement by robotics and artificial intelligence. This is perhaps the sense in which the history of how technology affects jobs is not repeating itself, and “this time” really is different.

In watching the cohort that includes my own children move from high school into young adulthood, this observation seems to me to contain a lot of truth. When it comes to training for a future job, many of us are still mentally in the 1890s, looking for one or a few particular focused skills that will guarantee a \”good job.\” But modern technologies are likely to disrupt what tasks are actually done in a very wide array of jobs, which will put a premium on workers with the ability to shift flexibly as the job situation is reshaped. 

____________________

\”Automation and New Tasks: How Technology Displaces and Reinstates Labor,\” by Daron Acemoglu and Pascual Restrepo
We present a framework for understanding the effects of automation and other types of technological changes on labor demand, and use it to interpret changes in US employment over the recent past. At the center of our framework is the allocation of tasks to capital and labor—the task content of production. Automation, which enables capital to replace labor in tasks it was previously engaged in, shifts the task content of production against labor because of a displacement effect. As a result, automation always reduces the labor share in value added and may reduce labor demand even as it raises productivity. The effects of automation are counterbalanced by the creation of new tasks in which labor has a comparative advantage. The introduction of new tasks changes the task content of production in favor of labor because of a reinstatement effect, and always raises the labor share and labor demand. We show how the role of changes in the task content of production—due to automation and new tasks—can be inferred from industry-level data. Our empirical decomposition suggests that the slower growth of employment over the last three decades is accounted for by an acceleration in the displacement effect, especially in manufacturing, a weaker reinstatement effect, and slower growth of productivity than in previous decades.
Full-Text Access | Supplementary Materials

\”Artificial Intelligence: The Ambiguous Labor Market Impact of Automating Prediction,\” by Ajay Agrawal, Joshua S. Gans and Avi Goldfarb
Recent advances in artificial intelligence are primarily driven by machine learning, a prediction technology. Prediction is useful because it is an input into decision-making. In order to appreciate the impact of artificial intelligence on jobs, it is important to understand the relative roles of prediction and decision tasks. We describe and provide examples of how artificial intelligence will affect labor, emphasizing differences between when the automation of prediction leads to automating decisions versus enhancing decision-making by humans.
Full-Text Access | Supplementary Materials

\”`Automation\’ of Manufacturing in the Late Nineteenth Century: The Hand and Machine Labor Study,\” by Jeremy Atack, Robert A. Margo and Paul W. Rhode
Recent advances in artificial intelligence and robotics have generated a robust debate about the future of work. An analogous debate occurred in the late nineteenth century when mechanization first transformed manufacturing. We analyze an extraordinary dataset from the late nineteenth century, the Hand and Machine Labor study carried out by the US Department of Labor in the mid-1890s. We focus on transitions at the task level from hand to machine production, and on the impact of inanimate power, especially of steam power, on labor productivity. Our analysis sheds light on the ability of modern task-based models to account for the effects of historical mechanization.
Full-Text Access | Supplementary Materials

\”The Rise of Robots in China,\” by Hong Cheng, Ruixue Jia, Dandan Li and Hongbin Li
China is the world\’s largest user of industrial robots. In 2016, sales of industrial robots in China reached 87,000 units, accounting for around 30 percent of the global market. To put this number in perspective, robot sales in all of Europe and the Americas in 2016 reached 97,300 units (according to data from the International Federation of Robotics). Between 2005 and 2016, the operational stock of industrial robots in China increased at an annual average rate of 38 percent. In this paper, we describe the adoption of robots by China\’s manufacturers using both aggregate industry-level and firm-level data, and we provide possible explanations from both the supply and demand sides for why robot use has risen so quickly in China. A key contribution of this paper is that we have collected some of the world\’s first data on firms\’ robot adoption behaviors with our China Employer-Employee Survey (CEES), which contains the first firm-level data that is representative of the entire Chinese manufacturing sector.
Full-Text Access | Supplementary Materials

Symposium on Fiscal Policy

\”Ten Years after the Financial Crisis: What Have We Learned from the Renaissance in Fiscal Research?\” by Valerie A. Ramey
This paper takes stock of what we have learned from the \”Renaissance\” in fiscal research in the ten years since the financial crisis. I first discuss the new innovations in methodology and various strengths and weaknesses of the main approaches to estimating fiscal multipliers. Reviewing the estimates, I come to the surprising conclusion that the bulk of the estimates for average spending and tax change multipliers lie in a fairly narrow range, 0.6 to 1 for spending multipliers and -2 to -3 for tax change multipliers. However, I identify economic circumstances in which multipliers lie outside those ranges. Finally, I review the debate on whether multipliers were higher for the 2009 Obama stimulus spending in the United States or for fiscal consolidations in Europe.
Full-Text Access | Supplementary Materials

\”Rising Government Debt: Causes and Solutions for a Decades-Old Trend,\” by Pierre Yared
Over the past four decades, government debt as a fraction of GDP has been on an upward trajectory in advanced economies, approaching levels not reached since World War II. While normative macroeconomic theories can explain the increase in the level of debt in certain periods as a response to macroeconomic shocks, they cannot explain the broad-based long-run trend in debt accumulation. In contrast, political economy theories can explain the long-run trend as resulting from an aging population, rising political polarization, and rising electoral uncertainty across advanced economies. These theories emphasize the time-inconsistency in government policymaking, and thus the need for fiscal rules that restrict policymakers. Fiscal rules trade off commitment to not overspend and flexibility to react to shocks. This tradeoff guides design features of optimal rules, such as information dependence, enforcement, cross-country coordination, escape clauses, and instrument versus target criteria.
Full-Text Access | Supplementary Materials

\”Effects of Austerity: Expenditure- and Tax-Based Approaches,\” by Alberto Alesina, Carlo Favero and Francesco Giavazzi
We review the debate surrounding the macroeconomic effects of deficit reduction policies (austerity). The discussion about \”austerity\” in general has distracted commentators and policymakers from a very important result, namely the enormous difference, on average, between expenditure- and tax-based austerity plans. Spending-based austerity plans are remarkably less costly than tax-based plans. The former have on average a close to zero effect on output and lead to a reduction of the debt/GDP ratio. Tax-based plans have the opposite effect and cause large and long-lasting recessions. These results also apply to the recent episodes of European austerity, which in this respect were not especially different from previous cases.
Full-Text Access | Supplementary Materials

Symposium on the Problems of Men

\”The Declining Labor Market Prospects of Less-Educated Men,\” by Ariel J. Binder and John Bound
Over the last half century, US wage growth stagnated, wage inequality rose, and the labor-force participation rate of prime-age men steadily declined. In this article, we examine these worrying labor market trends, focusing on outcomes for males without a college education. Though wages and participation have fallen in tandem for this population, we argue that the canonical neoclassical framework, which postulates a labor demand curve shifting inward across a stable labor supply curve, does not reasonably explain the data. Alternatives we discuss include adjustment frictions associated with labor demand shocks and effects of the changing marriage market—that is, the fact that fewer less-educated men are forming their own stable families—on male labor supply incentives. In the synthesis that emerges, the phenomenon of declining prime-age male labor-force participation is not coherently explained by a series of causal factors acting separately. A more reasonable interpretation, we argue, involves complex feedbacks between labor demand, family structure, and other factors that have disproportionately affected less-educated men.
Full-Text Access | Supplementary Materials

\”When Labor\’s Lost: Health, Family Life, Incarceration, and Education in a Time of Declining Economic Opportunity for Low-Skilled Men,\” by Courtney C. Coile and Mark G. Duggan
The economic progress of US men has stagnated in recent decades. The labor force participation rate of men ages 25-54 peaked in the mid-1960s and has declined since then (according to the Bureau of Labor Statistics), while men\’s real median earnings have been flat since the early 1970s. These population averages mask considerably larger declines in participation among less-educated and non-white men as well as substantial increases in wage inequality. In this paper, we seek to illuminate the broader context in which prime-age men are experiencing economic stagnation. We explore changes for prime-age men over time in education, mortality, morbidity, disability program receipt, family structure, and incarceration rates. We focus on prime-age men, namely those ages 25-54, and on the years 1980-2016 (or 2017 when possible), encompassing much of the period of reduced economic progress for low-skilled men.
Full-Text Access | Supplementary Materials

\”The Tenuous Attachments of Working-Class Men,\” by Kathryn Edin, Timothy Nelson, Andrew Cherlin and Robert Francis
In this essay, we explore how working-class men describe their attachments to work, family, and religion. We draw upon in-depth, life history interviews conducted in four metropolitan areas with racially and ethnically diverse groups of working-class men with a high school diploma but no four-year college degree. Between 2000 and 2013, we deployed heterogeneous sampling techniques in the black and white working-class neighborhoods of Boston, Massachusetts; Charleston, South Carolina; Chicago, Illinois; and the Philadelphia/Camden area of Pennsylvania and New Jersey. We screened to ensure that each respondent had at least one minor child, making sure to include a subset potentially subject to a child support order (because they were not married to, or living with, their child\’s mother). We interviewed roughly even numbers of black and white men in each site for a total of 107 respondents. Our approach allows us to explore complex questions in a rich and granular way that allows unanticipated results to emerge. These working-class men showed both a detachment from institutions and an engagement with more autonomous forms of work, childrearing, and spirituality, often with an emphasis on generativity, by which we mean a desire to guide and nurture the next generation. We also discuss the extent to which this autonomous and generative self is also a haphazard self, which may be aligned with counterproductive behaviors. And we look at racial and ethnic difference in perceptions of social standing.
Full-Text Access | Supplementary Materials

Features
\”Retrospectives: Ricardo on Machinery,\” by Samuel Hollander
We are currently experiencing an outpouring of concern both popular and professional regarding technological unemployment. I shall be discussing an apparent about-turn on the subject by David Ricardo (1772-1823), who at different times, even in different chapters of the same book, and, indeed, even at different places in the same chapter, seemed to be on both sides of the argument as to whether technological unemployment should be a matter for concern. In a chapter entitled \”On Machinery,\” added to the third edition of his Principles of Political Economy (1821), which comprises volume 1 of his Collected Works (1951-73), Ricardo announced that he had become concerned about the possibility, even likelihood, of technical change detrimental to labour\’s interests. However, in the very same \”On Machinery\” chapter, Ricardo also outlined qualifications to show that there was little need for concern. Ricardo\’s opposing messages are reflected in contrasting reactions to the chapter \”On Machinery.\” Some readers—including Thomas Robert Malthus and J. R. McCulloch—understood it as supporting working-class opposition to machinery. Others—including John Stuart Mill and Sir John Hicks—find therein the answer to such opposition.
Full-Text Access | Supplementary Materials

\”Recommendations for Further Reading,\” by Timothy Taylor
Full-Text Access | Supplementary Materials

\”Surprisingly, one commonality between globalization’s supporters and its critics is that both tend to believe the world is already far more globalized than it really is. …  The world is both more globalized than ever before and less globalized than most people perceive it to be. The intriguing possibility embodied in that conclusion is that companies and countries have far larger opportunities to benefit from global connectedness and more tools to manage its challenges than many decision-makers recognize.\”

The authors point to survey data on what people believe about globalization, compared to the actual data. Here are the results of a survey of business managers (footnotes omitted):

Figure 1.1 also highlights how managers tend to greatly overestimate measures of the depth of globalization. The actual levels are juxtaposed on the graph against perceived levels from a survey of 6,035 managers across three advanced economies (Germany, the UK, and the US) and three emerging economies (Brazil, China, and India) that we conducted in 2017. On average, the managers guessed that the world was five times more deeply globalized than it really is! In fact, their perceptions were no more accurate than those of students surveyed across 138 countries or members of the general public in the United States. And CEOs and other senior executives had even more exaggerated perceptions than did junior and middle managers—perhaps because their own lives tend to be far more global than those of their employees and customers.

Thus, some of the discussion in the report emphasizes that most economic activity is domestic, even for multinational firms.

The combined output of all multinational firms outside of their home countries added up to only 9% of global economic output in 2017, and just 2% of all employees around the world worked in the international operations of multinational firms. In part, those statistics reflect the fact that most companies are still domestic. Less than 0.1% of all firms have foreign operations and about 1% export. Small firms are, on average, much less international than large ones, and most companies are small. But even among the Fortune Global 500, the world’s largest firms by revenue, domestic sales still exceed international sales. … The same pattern of limited breadth prevails at the firm level as well. Among the world’s 100 largest corporations ranked by foreign assets, the average firm earns roughly 60% of its revenue in just four countries (home plus three international markets).

Although the buzzword is \”globalization,\” what this typically means in practice is trade with a few neighbors who are geographically close, or share the same language, or both.

Most countries’ international flows are so highly concentrated with key partner countries (usually neighbors) that it hardly makes sense to think of them as global at all. In fact, flows between countries and their single largest partners (e.g. export destinations for trade) make up nearly one-quarter of all merchandise exports and more than onequarter of all of the other flows …  Expanding the same analysis beyond only countries and their single largest partners, more than half of all flows except merchandise exports and inbound students take place between countries and their top three partners, and 75% or more are between countries and their top 10 partners. Even in the case of merchandise trade, more than half takes place between countries and their top five export destinations. Most countries simply do not maintain strong connections to a large number of other countries. 

Geographic distance, along with cultural, administrative/ political, and economic differences go a long way toward explaining the distributions of countries’ flows across locations. For example, if one pair of countries is half as distant as another otherwise similar pair of countries, greater physical proximity alone would be expected to increase the merchandise trade between the closer pair by more than three times and to more than double the stock of foreign direct investment (FDI) between them. And to highlight a cultural commonality, sharing a common official language roughly doubles both trade and foreign direct investment. Thus, despite the widespread perception that advances in transportation and telecommunications technologies are rendering distance irrelevant, international activity continues to be more intense among proximate countries. 

The chapter also has some intriguing hints about future directions for globalization. One possibility is an expansion of international flows beyond the closest neighbors and those with common languages, and so that more small companies become involved. Globalization is not close to any theoretical ceiling. 

Another set of possibilities is that the international connections of the future will tend to put less emphasis on movements of goods, and instead more emphasis on flows of information.  Comparing from 2000 up to the present, flows of goods, finances, and services are all up about 20-25%. But flows of information across international borders–measured by international flows of internet traffic, phone calls, printed publications– have nearly tripled in that time. The authors write: 

On the other hand, trade growth might be slowed by developments that would reduce the attraction of trade motivated by labor cost arbitrage. Automation and 3-D printing could potentially reduce the attraction of offshoring to access low labor costs. And macroeconomic trends imply some narrowing of the scope for such trade as well. One very rough measure of the potential for labor cost arbitrage across countries is the GDP-weighted average of the ratios of countries’ per capita incomes (higher over lower). As large emerging economies (especially China) have become richer, this ratio has already fallen from 8 in 2001 to 5.6 in 2017, and projections from Oxford Economics suggest it will continue falling (more slowly) to about 4 by 2050. While wage arbitrage will continue to motivate trade, exports of labor-intensive products from emerging economies may become a smaller driver of trade growth than in the recent past.

Flows of international tourists and of international students have been up. But on a global basis, migration has risen less than many people seem to believe. 

On a global basis, migration is on a rising trend, but a very modest one. … The proportion of people living outside of the countries where they were born has risen from 2.8% in 2001 to 3.4% in 2017. Both of those values, however, still round to 3%—the same level that global migration depth has approximated for more than a century!The modest global increase in international migration,  however, masks significant increases that have taken place in some countries. In advanced economies, the share of immigrants in the population increased from 9% in 2001 to 13% in 2017. In the United States, 2017 was noteworthy as the year that the proportion of immigrants in the population first surpassed its 1910 peak level.

Thus, the globalization of the future could follow a path where the ties across information, ideas, finance, and people are rising briskly, but the share of actual goods moving across national borders rises much more slowly. Altman, Ghemawat, and Bastian write: 

To summarize, the depth and breadth of trade, capital, information, and people flows—as well as the international business activity of multinational firms—fall far short of levels that are commonly presumed. National borders and the distances and differences between countries still have large dampening effects on international activity.