Skip to main content

Representative David Schweikert - Vice Chairman

The Wealth of Relations: Expanding Opportunity by Strengthening Families, Communities, and Civil Society

The Wealth of Relations: Expanding Opportunity by Strengthening Families, Communities, and Civil Society

Additional Material:

Throughout its history, economics has been centrally concerned with capital. In his magnum opus, The Wealth of Nations, Adam Smith defined it as the portion of someone’s possessions “which, he expects, is to afford him…revenue.”1 Until the second half of the 20th century, capital was analyzed primarily in its physical and financial forms: factories, machines, and equipment; stocks, bonds, and insurance contracts.

Over the past 60 years, Gary Becker and other economists have established the concept of “human capital”—personal attributes such as skills, knowledge, and personality traits.2 Like other forms of capital, human capital reflects investment and is valuable (revenue-affording) to its possessor. Most obviously, it pays off in the labor market. But more broadly, human capital helps us achieve whatever our aims might be—economic or otherwise. Empathy and the ability to cooperate, for instance, facilitate the forming of close friendships, an important source of happiness in life.  

In more recent decades, researchers and theorists have described another source of wealth: social capital. While not previously unknown to economists, social capital was first comprehensively analyzed by political scientist Robert Putnam.3 It refers to the aspects of human relationships that may be expected to afford value to their possessors. Relationships inhere in social networks as well as in the institutions that people create together for specific purposes and in which they participate. These institutions are ubiquitous, ranging from families to schools to book clubs to unions to churches to athletic leagues.4

The social capital literature has suffered from inconsistent and imprecise definitions, and like human capital, the social variety presents complex measurement challenges.5 Yet, in the same way that individual attributes are almost self-evidently of greater or lesser value, so, too, with relationships.6 Relationships provide individuals happiness, identity, self-worth, knowledge, skills, financial support, emotional support, values, preferences, habits, and more. Webs of relationships also produce collective goods such as norms, institutions, civil society, and culture, which then, in turn, become part of the stock of social capital available to individuals embedded in these webs.

Even from the narrowest of economic perspectives, who we know, the institutions to which we are connected, and what they can do for us are obviously related to economic outcomes. We find jobs through our contacts. We develop skills (human capital) that pay off in the labor market through the mentorship of teachers, parents, and neighbors (reflecting the social capital inhering in those relationships).

Just as obviously, economic precursors affect the value of individual and community social capital. Economic development has created more-extensive but less-personal relationships at least since ancient Greeks adopted coinage and institutionalized market exchange within open-air agoras.7 Today, a poor neighborhood with community members who have low attachment to the workforce will afford insufficiently valuable social capital to students living there who aspire to become a doctor.8 A town experiencing a factory closing may see local businesses fail and an exodus of the most resourceful residents, leaving behind places where community is all but absent.9

For two years, the Social Capital Project within the Joint Economic Committee (JEC) has documented trends in associational life—what we do together—and its distribution across the country. With this evidentiary base established, the Project, now situated in the JEC Chairman’s Office, turns to the development of a policy agenda rooted in social capital. Specifically, the focus of the Project will be to craft an agenda to expand opportunity by strengthening families, communities, and civil society.

The following sections of this paper offer a justification for our focus on opportunity and on social capital as a means to opportunity. Subsequently, we shift to discussing how public policy might affect the value and distribution of social capital. We introduce five policy goals around which the Social Capital Project’s research will be organized. The conclusion previews how we will do so.


Opportunity as the “Leading Object” of Public Policy

On the 85th anniversary of the Second Continental Congress’s ratification of the Declaration of Independence, President Abraham Lincoln addressed a special session of the US Congress. The state of the Union was decidedly precarious. Lincoln had assumed office just four months earlier. In between, four states had seceded from the Union in the aftermath of the Battle of Fort Sumter, joining the seven that had formed the Confederate States of America prior to Lincoln’s inauguration.

Lincoln asked Congress to authorize an increase in troops and funding sufficient to wage the Civil War, but he was also trying to win the hearts and minds of legislators—especially border-state legislators—in favor of the cause of the Union. His case was ultimately expressed in his assertion that “the leading object” of the federal government was “to elevate the condition of men; to lift artificial weights from all shoulders; to clear the paths of laudable pursuit for all; to afford all an unfettered start and a fair chance in the race of life.10

In other words, Lincoln’s pitch to nervous and wavering Americans staring a bloody war in the face—his best case for patriotism and loyalty—was an appeal to the federal government’s responsibility to remove barriers to opportunity.

Today’s policy and political debates often revolve around opportunity and the role of public policy in expanding it. These debates are rendered less productive than they could be because of a number of short-sighted views. More effectively promoting opportunity requires a broader and more subtle perspective that moves beyond the blind spots of the left and right.

For starters, there is a tendency to equate “opportunity” almost exclusively with economic outcomes—educational or financial success, for instance. Policymakers are too often blind to the reality that many people find happiness not in maximizing their years of formal schooling, standardized test scores, college ranking, seniority at work, or annual salary, but in enjoying their family life, being embedded in communities, feeling spiritually fulfilled, and having a sense of self-worth.

Many observers of the 2016 presidential primaries had a difficult time understanding the resonance of then-candidate Donald Trump’s assertion that, “Sadly, the American Dream is dead.” Tim Carney, in his important new book, Alienated America, argues that Trump’s pessimistic appeal was attractive not because economic times were especially bad that year. What had died, in the view of his core supporters, was “the American Dream of robust community life.”11

“The materialistic view of the American Dream,” according to Carney, “misses the point.” Instead, he speculates,

the things we think accompany the American Dream are the things that really are the American Dream. What if the T-ball game, the standing-room-only high school Christmas concert, the parish potluck, and decorating the community hall for a wedding—what if those activities are not the dressings around the American Dream, but what if they are the American Dream?12

Just as there is too often a narrow focus on economic outcomes in debates about opportunity, discussions too often emphasize economic or personal barriers. Among political liberals, in particular, “lifting artificial weights” and “clearing paths” mostly mean giving more money to poor, working-class, and (increasingly) middle-class people. Hence the calls on the left for guaranteed jobs, a $15 minimum wage, universal child care, universal college, and a universal basic income guarantee.

In contrast, conservatives have tended to point to personal barriers to opportunity. Different income levels in adulthood, for instance, may be due to unequal economic resources growing up, but they also may be the product of different orientations, preferences, values, and personal strengths and weaknesses. Equalizing incomes will not necessarily change these differences.

However, the conservative perspective is not without its own problems. Conservatives have tended to wield the concept of opportunity defensively, affirming their support for “equality of opportunity” as against the “equality of outcomes” that they accuse liberals of seeking. The distinction is rooted in conservatism’s view of people as mostly the captains of their own ships. Given that we have made great strides as a nation achieving formal political equality, the US is often thought to have realized actual equality of opportunity. If someone fails to realize her own definition of the good life—perhaps as a consequence of problematic orientations, preferences, values, and weaknesses—many conservatives view this failure as a personal shortcoming.

Most conservatives would agree with Martin Luther King Jr. that “a productive and happy life is not something you find, it is something you make.”13 But we do not navigate our lives in isolation, we make a productive and happy life with other people. Supportive relationships and institutions are instrumental for expanding opportunity. In part, that is because they are instrumental in forming our orientations, preferences, values, and personal strengths and weaknesses.

That is to say, opportunity depends on social capital—what is available to us from our relationships with family, friends, neighbors, congregants, coworkers, and others. In particular, the people to whom we are born and around whom we live are consequential for our opportunities. “Artificial weights” are not only economic, not only personal, but social.


Social Capital and Opportunity

Establishing the empirical importance of social capital is more difficult than it may seem at first glance. After all, we cannot see or touch social capital; we cannot measure it directly, and it has both quantitative and qualitative elements. Research on human capital has tended, until recently, to simply equate it with educational attainment or scores on tests of cognitive ability. As limiting as this convention is, the field of social capital research cannot even claim a conventional measure.

Moreover, it is difficult to establish causal relationships in social science even when looking at well-measured phenomena. Absent a randomized experiment—where half of participants are, for instance, given $50,000 while the other half are not—social scientists have limited options for understanding how the outcomes of two groups that differ on some attribute would change were there no such difference. The two groups may differ in any number of other ways, and those differences may be the real attributes that produce disparate outcomes.

Because we lack many experiments in which social capital is distributed randomly between people, we mostly are left with necessarily flawed studies that attempt to mimic an experiment using statistical methods.

One important exception is a paper by economists Raj Chetty, Nathaniel Hendren, and Lawrence Katz evaluating a policy experiment called “Moving to Opportunity.”14 Moving to Opportunity involved nearly 5,000 families in five cities and randomly gave public housing residents a rent subsidy to be used only in a low-poverty neighborhood, a subsidy with no such restriction, or no subsidy at all (beyond continuing to live in their public housing project). The study found that compared with remaining in place, moving to a low-poverty neighborhood in childhood (but not adolescence) increased college enrollment, lifted adult earnings by one-third, reduced single parenthood, and improved neighborhood quality in adulthood. Social capital’s power burns slowly; the longer children lived in a low-poverty neighborhood, the better their outcomes. Fundamentally, the results point to the importance of the local networks and institutions to which children are exposed.

This paper is part of a series of studies illustrating the importance of social capital for the intergenerational mobility of children by Chetty, Hendren, John Friedman, and a number of their colleagues at the Equality of Opportunity Project (EOP) and now at Opportunity Insights.

In its seminal work on the subject, “Where is the Land of Opportunity?”, the EOP found that intergenerational mobility varies widely across the country.15 To measure mobility, the authors of the study ranked more than 40 million children and their parents by income percentiles. Places where two grown children from different families tend to be ranked more closely together than were their parents were considered to have high relative mobility. The authors found that areas with high levels of relative mobility also tended to produce higher-ranked adults from their low-income children. (That is, they produced what EOP calls simply, “upward mobility.”)16

Many features of places seem to go along with high or low mobility, but the research of EOP and Opportunity Insights consistently has found measures of social capital to be among the most strongly related. Out of 34 features they examined in “Where is the Land of Opportunity?”, the one most strongly related to upward mobility and relative mobility was the share of children in an area that was raised by a single mother. In second place as a predictor of upward mobility was a fairly limited index of social capital. Community rates of mobility were also higher where there were more religious adherents, fewer divorced adults, and more married adults, and where there was less economic and racial segregation.17 All of these indicators bear on the value of social capital in an area.18

Research by the Social Capital Project confirms the EOP findings. Relative mobility is higher not only when states score better on our social capital index and when family unity is higher, but also the higher are their levels of family interaction, social support, community participation, and charitable giving.19

These relationships may or may not mean that social capital causes mobility. It may be that upwardly mobile families with high social capital simply tend to congregate together; if a community looks like it promotes mobility, it might just be because it is home to a lot of families who would be upwardly mobile anywhere. Chetty and Hendren have made a strong case that places do promote or impede mobility.20 In one paper, they found that children who moved to better areas—places with high income mobility among permanent residents—had better long-term outcomes in proportion to the amount of time they spent there.21

It is even more difficult to establish that specific features of places affect mobility. In another paper, Chetty and Hendren first estimated the causal effect on mobility of living in each of hundreds of communities, such as Memphis or Salt Lake City.22 To do so, they compared the adult outcomes of brothers who moved to an area but, because they differed in age, subsequently lived in it for different durations. This causal effect is distinct from the part of a place’s mobility rate that simply reflects the sorting of high- or low-mobility families into Memphis or Salt Lake City. Then, rather than looking at how community features were related to mobility, Chetty and Hendren looked at how they were related to the causal effect of living in Memphis, Salt Lake City, or some other community.

These analyses revealed that nearly two-thirds of the strong correlation between single parenthood and upward mobility was due to the fact that single parenthood was related to an area’s causal effect on mobility. That is, communities with high rates of single parenthood tended to be worse for mobility than communities with low rates (not simply home to families with lower mobility). The relationship between an area’s score on the social capital index and its upward mobility entirely reflected the causal effect of moving to the area rather than sorting. Social capital had the fourth-strongest relationship with area causal effects on mobility (out of 40 factors), while single parenthood ranked 10th. The share of adults married came in close behind at 15th place.

Chetty, Hendren, and their colleagues have also assessed the extent to which features of communities can explain unequal outcomes between men and women and between blacks and whites. One paper looked at why low-income girls have higher employment rates as adults than poor boys do.23 The association between this employment gap and a community’s rate of single parenthood when the boys and girls were children was the fifth strongest of 28 factors, and community marriage rates came in at sixth.

In yet another paper, the team looked at adults who, as children, were poor but lived in low-poverty neighborhoods. They found that black men had stronger upward mobility the more low-income black fathers there were in their childhood neighborhood.24 This was true regardless of whether someone’s own father was present, suggesting that even the family cohesion of other black children in the neighborhood affected them.25 Meanwhile, having more low-income white fathers in the neighborhood did not increase the upward mobility of poor black children; nor did having more low-income black men who were childless. More low-income black fathers in a neighborhood also corresponded with higher future employment among poor black boys. The relationship between the number of low-income black fathers and future employment was much smaller for poor white boys and negligible for poor black girls.

The Chetty-Hendren-Friedman research on upward mobility suggests that our intuitions about social capital are correct. The nature of our relationships provides more or less value to us as we pursue our aims in life.26


The Limits of Public Policy

If social capital matters for opportunity, there remains the question of whether government—or the federal government specifically—should (or can) do anything to reduce social inequalities.

The extent to which government policy can expand opportunity or reverse America’s deteriorating family and community life is, first and foremost, limited by our commitment to classical liberal principles of personal responsibility, individual choice, and freedom from state encroachment on private decision-making. One might argue that if free-willed individuals choose to act in ways that weaken community cohesion—by, for instance, relying more on markets to provide services such as child care and less on neighbors—so much the worse for communities.

What is more, practically speaking, public policies to change behavior have often proven ineffective or, worse, counter-productive. In the mid-1980s, after years of evaluating social policies, Peter Rossi declared with his “iron law of evaluation” that, “The expected value of any net impact assessment of any large scale social program is zero.”27 Further, his “brass law of evaluation” asserted that, “The more social programs are designed to change individuals, the more likely the net impact of the program will be zero.” Almost thirty years later, David Muhlhausen ratified Rossi’s observations in a comprehensive study of rigorously evaluated multi-site federal social programs. He reported, “the evidence overwhelmingly points to the conclusion that federal social programs are ineffective.”28

Even worse, policies not primarily intended to change behavior—aimed at, for instance, providing greater income security—often nevertheless create perverse incentives for beneficiaries to act in ways that impede opportunity. Many safety-net programs include features that discourage work, marriage, and saving. Policies to mitigate risks often counterproductively generate more risk-taking—the phenomenon known as “moral hazard.”

When the federal government expands the provision of goods and services on offer through social policy, it runs the risk of “crowding out” civil society—another potentially counterproductive effect of public policy. Increased reliance on government to address needs weakens the selfish rationale for community members to invest in social capital. Because individual investment in social capital often creates benefits for the entire community, such as norms of trust and reciprocity, the result may be the loss of communal benefits completely ancillary from the goods and services provided by government. Greater subsidy of child care might end up producing less volunteerism.

Any policy agenda to expand opportunity must confront these two constraints: the inadvisability of government intervention in many cases, and its frequent incompetence. However, even with these limitations, public policy can still influence our institutions of civil society in beneficial ways. Most obviously, government can remove or reform policies that weaken institutions or that promote less desirable alternatives to them. It can also promote the bases for a flourishing civil society: strong families, a stable and growing economy, effective safety nets, and local decision-making. Finally, public policy can facilitate the identification and replication of successful interventions that promote opportunity.


Expanding Opportunity through Social Capital Investment

The concept of social capital encompasses an enormous amount of social and economic life. A place’s social capital reflects the sum of all its relationships—between community members, between individuals and local institutions, and between members and those outside the community. To bound the types of policies under consideration, the Project has identified five broad goals related to opportunity, based on the past two years of its research.

These goals include making it more affordable to raise a family, increasing how many children are raised by happily married parents, connecting more people to work, improving the effectiveness of investments in youth and young adults, and rebuilding civil society.

Making It More Affordable to Raise a Family

The most important source of social capital for most people is the family in which they are raised. As Yuval Levin has eloquently expressed, the “middle layers” of society—between the individual and the state—“begin in loving family attachments.” From this base, the middle layers

spread outward to interpersonal relationships in neighborhoods, schools, workplaces, religious communities, fraternal bodies, civic associations, economic enterprises, activist groups, and the work of local governments. They reach further outward toward broader social, political, and professional affiliations, state institutions, and regional affinities. And they conclude in a national identity that among its foremost attributes is dedicated to the principle of the equality of the entire human race.29

When families are unhealthy or diminishing in number, the social capital effects ripple across our other relationships and civil society, reducing happiness, hurting opportunity, and exacerbating inequalities. In particular, if a large or growing number of men and women cannot afford to start or expand a family in line with their preferences, that represents a profound loss that merits national attention.


Figure 1. Percent of Women Married, 1880-2018

Source: Social Capital Project analyses.30


Both marriage and fertility have declined dramatically since the mid-20th century. The share of women married peaked around the mid-1950s (Figure 1), and marriage rates for single women have fallen steadily since 1950.31 Fertility began falling in the late 1950s (though from a relative peak, since fertility was much higher for most of the 19th century—see Figure 2).


Figure 2. Total Fertility Rate, 1800-2017

Source: Social Capital Project analyses.32


Do these trends reflect a long-term increase in the cost of raising children? Or perhaps a deterioration in personal finances that has made fewer people—in particular, fewer men—marriage material? Has adequately-sized housing become prohibitively expensive? Has student loan debt scared off would-be newlyweds? If these explanations are behind the fall in marriage and fertility, then our economy would be deeply implicated.

However, the timing of these changes suggests a different explanation. Marriage and fertility both fell during the 1960s, a period of robust income growth for Americans of all walks of life. The steepest declines in both came before 1980. While marriage rates have continued to fall, the total fertility rate was 1.79 in 1977, and 40 years later, it was 1.77.33 These trends are inconsistent with accounts focused on economic decline since the 1980s.

An alternative explanation that better fits the timing is that increased educational attainment and employment among women led to later and fewer marriages and births. Female labor force participation rose throughout the 20th century, but the sharpest rise came in the 1970s, when both marriage and fertility were rapidly declining.34 The share of women getting four-year college degrees began to accelerate in the mid-1960s and the share in graduate school sharply rose after 1970.35

This historical development has expanded opportunity for women greatly. However, because men did not choose to take on more of the burden of home- or community-making, the shift has also weakened our associational life.36

That, in turn, likely has made it more difficult to raise a family. Because there are fewer stay-at-home parents today, two-worker families have increasingly had to rely on more formal (and expensive) child care arrangements. Because the workplace has not better accommodated two-worker families, their increasing number has led many to feel they face an inescapable time crunch. Long work hours and commutes lead to less family time at home.

Meanwhile, families that might prefer the traditional breadwinner/homemaker model may find they must sacrifice other wants to do so, as the dual-income families bid up the price of housing, health care, higher education, and other expenses. Many traditionalists will find their preferences unattainable.

Falling marriage (paired with an increasingly generous federal safety net) has also contributed to an explosion in single-parent families, which tend to be poorer than their two-parent counterparts.

If this story about the declining affordability of raising a family is correct, then policy should address the highest costs of raising a family while accepting that we are unlikely to return to mid-20th-century levels of marriage and fertility. It should focus in particular on lower-middle class, working class, and poor families, and it should not marginalize or penalize families wishing to pursue the traditional sole-breadwinner model. Policy should also consider ways to increase family time. It should promote two-parent families and discourage single-parent families (see below). This story of changing family affordability, more so than the story of economic deterioration, also highlights the importance of choices and trade-offs. No one has a right to form their ideal family on their ideal schedule, and it is unreasonable to expect other taxpayers or employers to pay the cost of doing so.

A third explanation for falling marriage and fertility combines the first two, asserting that economic deterioration is behind the increase in work among women. This explanation sees rising female labor force participation and infers that wives have had to bail out their husbands due to their falling wages and employment. While the timing of the changes in educational attainment, work, marriage, and fertility—and the broad pervasiveness of these changes across rich countries around the world—offer little support for this story, it, too, would call for bold policy reforms to address economic problems that apparently have been plaguing the developed world for half a century.37

Has it become less affordable to raise a family, or is it just too expensive for too many people, regardless of the change over time? Who faces affordability problems? Which of the three stories about the decline in marriage and fertility is right? And above all, what should be done to help Americans who cannot afford the family that they desire? The Social Capital Project will pursue these questions over the coming months.

Increasing How Many Children are Raised by Happily Married Parents

Partly because marriage has declined—including shotgun marriages following unplanned pregnancies—the share of children in married-parent families has steadily declined over the past 50 years.38 In 1967, 88 percent of children were living with two parents, but that number had fallen to 69 percent by 2017. Half of children today will experience at least some time living without both parents.39 The share of births that were to unmarried women rose from 5 percent in 1960 to 40 percent in 2017 (Figure 3).40


Figure 3. Share of Births to Unmarried Women, 1940-2017

Source: Social Capital Project analyses.41


There are good theoretical reasons to think that growing up with a single parent is, on average, disadvantageous for children. And indeed, hundreds of studies find that on just about every outcome, children who grow up with single parents do worse than children who grow up with married parents.42

However, the problem is that we do not know how today’s grown children of single parents would have done if their parents had been induced, somehow, to marry or stay married. In fact, different forms of inducement would have affected different families. It is not reasonable to assume—as most social science research does—that the children of parents who would have been nudged to marry would later have the same outcomes as we see in the real world among the children of married parents.43

Consider identical twin studies, which compare the children of twin sisters, where one twin is married while the other is single. As in other studies, the children raised without a father present tend to do worse than those with two parents. This difference, by design, holds constant everything twin parents share in common (including all of their genes).

However, what we want to know is not how the children of the married twins compare with the children of the single twins, but how much better or worse the latter would have done with their father present. The married and single twins differ in some regards, as do their partners. Those differences might have affected child outcomes. We cannot simply assume that the single twins’ children would have done as well as their cousins if only their fathers had been around.

At the very least, it seems reasonable to believe that if the single twins and their partners would have had marriages as happy as those of the married twins and their partners—with all of the benefits that would have brought to their children—they would have chosen to get or stay married. Alternatively, if they would have experienced high or even elevated conflict as a consequence of a nudge-induced marriage, their children might have done worse than they did in the real world raised by a single parent.44

The key is having happily married parents. We don’t need to determine how well the academic literature estimates the typical effect on children of growing up with a single parent. We only need to stipulate that children generally are happier, healthier, and better prepared for life when they have two happily married parents rather than a single parent.

They get to see both parents every day, spend the holidays with both, and they don’t have to feel guilty about spending or enjoying more time in one household than the other. Nor do they have to question whether they caused their parents to break up. They have a single set of household rules, a single bedroom and wardrobe. Their schedule does not depend on which parent they are staying with. They get engagement from both parents and avoid hearing parents acidly complain about each other. Their parents are less exhausted by childrearing. They get the material benefits of economies of scale and of higher family income. They are witness to what a loving relationship looks like and have first-hand evidence that such relationships are secure and sustainable. And they avoid having to adjust to the changing romantic lives of their mother or father—changes which can include disruptive remarriages and family-blending.

It is easy to see how much more valuable family social capital is likely to be in such families than in disrupted families. And in terms of community social capital, the Chetty-Hendren-Friedman team has uncovered suggestive evidence that being surrounded by more single-parent families can also hurt upward mobility independently of whether they grow up with both parents themselves.45

From a policy perspective, then, we want more children in happy married-parent families. That is subtly different than simply minimizing the share of children in single-parent families, and it stands in contrast to being indifferent about the kinds of families in which children are raised.

How to achieve this goal depends on knowing how we got here, where and among whom the two-parent ideal has receded, and how much we should blame economic hardship versus cultural change or the unintended consequences of public policy. It could entail a variety of strategies: fostering the conditions that lead to more happy marriages; assisting fragile families so that differences and conflicts may be effectively cooled and ultimately resolved; preventing unintended pregnancies; improving men’s economic prospects to make them more “marriageable;” or removing penalties in the tax code and in safety net programs that discourage marriage among those who would otherwise wed. We will explore these and other policy approaches during the rest of this Congress.

Connecting More People to Work

Accounts of rising single parenthood that emphasize economic decline tend to focus on trends in men’s wages but also on the decline in the share of working-age men who have a job. This fraction has been falling since the late 1960s, driven by the decline in the share of men who are even looking for work. The top lines in Figure 4 show that the share of men ages 25 to 54 who were employed or actively looking for a job (“in the labor force”) fell from 97 percent in 1953 to 89 percent in 2018.


Figure 4. Employment, Labor Force Participation, and Weekly Hours among Working-Age Men, 1900-2018

Source: Social Capital Project analyses.46


Should we worry about the decline in male labor force participation? Many observers fear that it reflects a deteriorating national economy. However, it is notable that labor force participation among working-age women soared 35 percent in 1948 to 77 percent by 1999, and it was still at 75 percent in 2018.47

Moreover, a number of studies find that relatively little of the decline in labor force participation can be attributed to men who have given up on finding a job or who say they want a job.48 The biggest part of the story involves men who indicate they are disabled when asked in household surveys. Despite the attention given to the recent increase in mortality (driven by the opioids crisis)49, most evidence indicates that over the past 50 years, health has improved and work has become both less physically arduous and more accommodating of disabled employees.50 Meanwhile, it has become both easier to qualify for federal disability benefits and more financially attractive to try.51

Indeed, one possible read of Figure 4 is that the decline in labor force participation is simply the latest manifestation of rising national affluence. The two lowest lines in the chart display average weekly hours worked for employed adult men. In 1900, that average was over 58 hours, but since the Great Depression it has hovered around 41 hours (except for an upward blip during World War II). Without rising productivity, it is unclear that workers could have won a 40-hour work week by the 1930s. Since the early 1950s, rather than a further decline in the work week, the nation has seen more and more working-age men opting out of employment—either temporarily or permanently. That may reflect higher earnings among wives, more generous federal safety nets, or reduced expectations of support from nonresident fathers on the part of single mothers who have earnings, government benefits, or both.

But even if declining work might not primarily reflect problems with the economy, it still should concern us. Arguing that falling labor force participation “is of a piece with the broader turn away from community in America,” New York Times columnist Ross Douthat warns that

the decline of work carries social costs as well as an economic price tag. Even a grinding job tends to be an important source of social capital, providing everyday structure for people who live alone, a place to meet friends and kindle romances for people who lack other forms of community, a path away from crime and prison for young men, an example to children and a source of self-respect for parents.52

Concurring, social analyst and incoming Manhattan Institute president Reihan Salam notes that family ties are weakening, worrying that, “those who find themselves disconnected from the world of market work find themselves socially isolated along many other dimensions.”53

Social Capital Project research on “disconnected men” confirms these fears.54 Compared with employed men, men out of the labor force are more socially isolated and less happy.55 In our study, they believed more than employed men that they were left uninvited to do things by others, that they would have a hard time finding someone to help them with a move, that they had no one with whom to share their worries, and that they lacked anyone to turn to for advice related to personal problems. They were less likely to be married, to live with adults, or to live with children, and they were more likely to be divorced. Men out of the labor force were more likely to have ever been depressed and to have attempted suicide. The worse their score on an index of mental health, the less likely they were to have someone with whom they could confide in.

In addition to widespread reports of poor physical and mental health, our report turned up suggestive evidence that past incarceration and the threat of child support collection were additional employment challenges.

In future reports, the Social Capital Project will assess the causes of declining work and propose policies to reverse the decline. The goal should not be to increase employment rates indiscriminately, which would conflict with our interest in enabling more parents to afford to stay at home while their spouse works, if that is their preference. Nor should we be concerned about non-workers if they are enrolled in school or have enough resources to retire.

Rather, we will focus on idle able-bodied men and women. Our proposals will seek to reverse the deadening social disconnection that is subsidized by work-discouraging federal benefits, inflicted by government regulations that price many people out of employment, and exacerbated by economic policy that limits job creation and wage growth.

Improving the Effectiveness of Investments in Youth and Young Adults

If the families into which we are born made no difference to our adult outcomes, then poor children would be as likely as rich children to become well-off adults. For instance, one in five children raised in the poorest fifth would make it to the top fifth in adulthood, and one in five children raised in the top fifth would stay there. In reality, as few as one in 33 poor children rises to the top fifth, and two in five children starting at the top remain there as adults.56 Almost half of poor children (46 percent) end up in the poorest fifth of adults.

As children age into adults, we tend to assign more responsibility to them for their outcomes. But what are we to make of this fact: entering kindergartners who are in the bottom fifth of socioeconomic status typically lag their peers in the top fifth on math test scores by the equivalent of 19 points on an IQ test?57 And how are we then to feel about that gap still being the equivalent of 13 points between high school seniors who have a parent with a graduate degree and those whose parents did not graduate from high school?58

Such inequality of opportunity calls out for attention from policymakers of all ideological stripes.

Unfortunately, we have failed monumentally to narrow these gaps over time. In fact, they have remained essentially unchanged over at least 50 years.59 Rossi’s metallic laws of evaluation would predict nothing more; most of the social interventions we have tried have disappointed time and again.

For example, the Department of Education’s What Works Clearinghouse has evaluated 385 different interventions designed to affect skills acquisition and behavior.60 Of these, only 44 (11 percent) showed clear positive effects, meaning that the research was sufficiently strong and there was no contrary evidence worth considering.61 Another 54 (14 percent) were examined by multiple studies and showed “potentially positive effects,” meaning that it was supported by a preponderance of the evidence, though that evidence might be less rigorous than the best study designs and some of it might fail to find effects. Another 25 percent of interventions showed potentially positive effects but were only examined by a single study.

In contrast, 41 percent of the interventions were deemed to have “no discernable effects,” 3 percent had “potentially negative effects,” and 5 percent had “mixed effects.” This sampling of interventions is, unfortunately, likely to produce too optimistic a conclusion, since the vast majority of interventions are never studied, and those that are generally seem promising to begin with. Furthermore, studies finding positive effects of interventions routinely fail to replicate.62

One response would be to throw up our hands in defeat and live with a reality in which some children bear artificial weights on their shoulders as they attempt to traverse a path of laudable pursuit littered with debris.

We think Lincoln would throw a variety of strategies at the problem, and we think the strategies most likely to succeed will increase the effectiveness of investment in youth and young adults by parents and institutions outside the family. That is, we are most interested in strategies to increase the value of the social capital available to young people.

Over the course of this Congress, the Social Capital Project will explore ways to strengthen parents’ ability to invest in their children. The policies we consider related to family affordability and family cohesion will be relevant for this policy goal. Other possible reports might promote residential mobility (to elsewhere within a local community or to a different part of the country), assess policies to expand educational choice, or highlight promising interventions that seek to improve parenting skills. Some of these interventions—such as programs that deliver information via text messaging—might be considered “social capital hacks,” aimed at circumventing relationship-intensive strategies that are difficult to replicate or scale up.

The Project will also consider ways to improve the quality of investments provided by institutions outside the family. Many such institutions—schools being the most prominent—offer programs seeking to expand the opportunities of youth and young adults, to build their skills, and to alter their behavior. We are particularly interested in mentoring programs, which leverage relationships in an intensive way, as well as apprenticeships and other school-to-work programs. And since inequality of opportunity, as evidenced by test score gaps, appears early in life, we will pay close attention to early childhood interventions as well.

Often, our research will simply shine a light on private efforts that are succeeding, in order to encourage other private organizations to follow their lead. Other times, we may recommend local experimentation, potentially federally funded. Given the low rate of success that social programs have demonstrated, however, government must evolve to emphasize more evidence-based policymaking—requiring evaluation of publicly-funded programs and strict accountability. Programs that cannot be shown to be effective should, quite simply, be shut down.

A variety of public and private organizations are compiling data on “what works” (and what doesn’t).63 It is not far-fetched to imagine someday having the equivalent of the Congressional Budget Office or Joint Committee on Taxation “scoring” social policy legislation in terms of the likely effectiveness it will have in achieving its intended goals.

We will also consider ways that policy might encourage a greater role for the private sector in funding interventions and discovering what does and does not work. Social impact bonds provide one model, where private organizations fund an intervention with the understanding that they will be paid a dividend by a governmental partner if the intervention produces public savings. If the intervention fails, the taxpayer loses nothing. Income share agreements offer a similar model, whereby private investors partially fund the costs of college in exchange for an agreed-upon portion of a student’s future income. Income share agreements transfer risk from students to investors, and given the current problems with the student loan system, they could be attractive to many undergraduates.

There are any number of ways to increase the value of social capital accessible to youth and young adults through parental and institutional investment in them, and policymakers should be open to all of them.

Rebuilding Civil Society

In our flagship report, “What We Do Together,” the Social Capital Project documented long-term declines in the health of American associational life along a number of dimensions.64 Membership in a church or synagogue is down, as is attendance at religious services. Union membership has plummeted. Social interaction with neighbors and coworkers has declined. Rich, middle class, and poor are less likely to live alongside each other. Trust in our fellow man has eroded, along with trust in federal and state government, policymakers, and the media. Confidence in organized religion has fallen, as has confidence in banks, newspapers, big business, organized labor, and the medical system.

In short, our institutions of civil society have weakened and withered, and our relationships have become more circumscribed. Political polarization has deepened at the same time, increasingly taking a regional form that pits coastal cosmopolitans against heartland traditionalists. And, not unrelatedly, policymaking has become more concentrated in the federal government.

Rebuilding civil society will require a fundamental change in how we perceive policymaking. It will require policies that respect the strengths of localism while accounting for its weaknesses. That is to say, they must promote subsidiarity. Subsidiarity is a concept from Catholic social thought that recommends that if something can be more effectively done at a smaller, more local, and less centralized level than at a larger, more distant, and more centralized level, it should be.

Subsidiarity has a number of advantages. It leverages local expertise and relationships rather than relying on far-off and impersonal bureaucracies. It allows a diversity of solutions to respond to a diversity of situations across the country instead of relying on one-size-fits-all approaches handed down from the federal government. By giving more responsibility to local residents and institutions, it provides valuable roles to community members they might otherwise lack.

Further, by encouraging participation in local groups, subsidiarity provides us with firmly rooted identities, nurturing self-worth. That reduces the likelihood that people will cement their identities to non-local groups based on ideology or ethnicity that reinforce social and political polarization. By keeping decision-making and political authority at the local level, subsidiarity avoids the polarization that results when regional cultural polarization collides with federal politics. Finally, it forces local residents to interact to govern themselves, which then creates communal benefits. These benefits constitute social capital enjoyed by the whole community—strong institutions, dense and active social networks, and norms that encourage reciprocity and promote opportunity.

Unfortunately, there is a chicken-and-egg problem here for those who want to rebuild civil society through subsidiarity. Strong institutions, dense and active social networks, and norms of reciprocity themselves may be prerequisites for successful localism. But government (and markets) have crowded out civil society over the years, by serving specific needs that civil society used to fulfill (be they income support or personal services). By reducing social interaction, the number of groups to which we might belong, and what we get out of both, crowd-out has further debased our social capital beyond its taking over specific responsibilities. Members of “communities” have become less and less so. As a consequence, simply reintroducing responsibilities to civil society may fail. It may no longer be up to the task. Our social capital muscles may be severely atrophied.

It is not just—or even primarily—encroachment by the federal government that has reduced our need for civil society. Affluence has made social relationships less necessary, and as a result, we have chosen to invest less in them and more in our own happiness as individuals (happiness as we perceive it anyway). Affluence has allowed us to outsource the responsibilities we used to have toward one another to impersonal institutions, including the federal government, the personal service sector, private insurance and consumer credit companies, and the educational system. Affluence is also reflected in technological development, which has allowed us to maintain relationships with far-flung friends and family as we de-prioritize getting to know our neighbors better.

In particular, affluence made greater female labor force participation possible by improving the ability of women to control their fertility as well as by making paid child care affordable and increasing household productivity to the point where homes could be maintained even with both husband and wife working for pay. These developments resulted not only in less home-making, but less community-making, because husbands did not take on the roles that wives increasingly turned away from. Affluence also made single parenthood more viable, as a generous-enough safety net came to substitute for fathers.

As if the challenges of rebuilding civil society were not enough, localism has shortcomings to which policy should attend. Some places are poorer than others, which means residents will need more help even as their state or locality is the least able to provide it. Relatedly, the business cycle creates problems for states and localities, which often have limited options for financing their costs through deficit spending during recessions. Furthermore, in competing to attract business or in trying to avoid attracting too many needy residents, states and localities may face various kinds of “races to the bottom.” They may spend wastefully to bring in employment or stingily tighten the purse strings of safety-net spending.

Another problem for localism is that compared with a powerful federal government, states and localities may not be able to stand up to non-local business interests or dominant local businesses, which may exercise undue political power. Finally, sometimes federal policies can be superior to local ones—more effective (at running Social Security), more efficient (less inclined toward land use regulations or occupational licensing), or more just (in enforcing civil rights).

The Social Capital Project intends to think creatively about how policy—including federal policy—can rebuild civil society. First, federal policy can strengthen local institutions. It can leverage existing institutions and refrain from impeding their efforts. In particular, it can reduce barriers to church-based service provision using federal funds. It can also devolve more authority to local institutions in administering federal programs, and it can reduce its involvement in policies that could be better pursued locally. Policy can offer incentives for donations of time or money to local institutions. And it can create new institutions, such as worker co-ops that could provide portable benefits to employees.65

Policies at different levels could also promote trust in a community (restoring trust in police, for example, within minority communities), or they could promote pro-social norms, pro-opportunity norms, or information-sharing (about, for instance, job demand and availability).  They could seek to increase social interaction, through investment in “social infrastructure” that brings people together (including libraries, parks, and shopping malls).66 Policies could even attempt to influence patterns of social interaction by, for example, discouraging land use regulations and zoning that segregates by income (and inefficiently drives up the cost of living).

Finally, federal policies could counteract the problematic aspects of localism. The federal government could provide a limited degree of federal redistribution between rich and poor states. It could encourage individual savings so that states’ residents aren’t dependent on federal taxpayers during downturns. And it can ensure a baseline of civil rights protections to guard against the threat of discrimination.

By finding ways to rebuild local institutions of civil society and revive social interaction at the local level, the Social Capital Project hopes to catalyze a virtuous cycle of reinvestment in social capital to fulfill more and more needs—the mirror image of the vicious cycle that has incapacitated our ability to accomplish things with fellow members of our community.


Social capital is a form of wealth to which policymakers have devoted too little attention. The benefits we get from it are as diverse as information, identity, financial and emotional support, and culture. It is the stuff of which life is made.

An important product of social capital is opportunity—opportunity to pursue whatever individual priorities we may hold dear. Public policy should aim to expand opportunity and to temper inequalities of opportunity. An important way it can do that is by strengthening families, communities, and civil society.

In the coming months, the Social Capital Project will release a series of reports—one for each of the five policy goals outlined above—providing an overview of the goal, outlining the nature of the policy problem, and summarizing different approaches to achieving the goal. With these complete, the Project will then issue a series of analyses and recommendations related to specific policy approaches. These narrower reports will sometimes recommend federal policy proposals in greater or lesser detail. Other times, they will suggest appropriate state and local policies. Still other reports will suggest actions that private organizations might take or highlight successful private efforts to address a policy goal. The result will be a coherent policy agenda to expand opportunity by shoring up families, communities, and civil society.

This agenda will necessarily be incomplete, but it should offer a policy menu of sufficient variety to appeal to policymakers with a range of priorities. It will, hopefully, inspire others to think more creatively about investment in social capital and how public policy might make us wealthier not just financially, but in terms of our relationships with each other.


1 Adam Smith and Edwin Cannan (1994). The Wealth of Nations. New York: Random House, Modern Library Edition, 302.

2 Gary S. Becker (1994). Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. Chicago: University of Chicago Press.

3 Robert D. Putnam (2001). Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

4 Other examples of institutions: fraternal organizations, arts centers, community gardens, farmers’ markets, parades, festivals, block parties, museums, libraries, local businesses, and local governments.

5 For criticism of the concept see Steven N. Durlauf (2002). “On the Empirics of Social Capital.” The Economic Journal 112(483): F459–79.; Steven N. Durlauf and Marcel Fafchamps (2004). “Social Capital.” Working Paper No. 10485, National Bureau of Economic Research.; Michael W. Foley and Bob Edwards (1999). “Is It Time to Disinvest in Social Capital?” Journal of Public Policy 19(2): 141–73.; Alejandro Portes (1998). “Social Capital: Its Origins and Applications in Modern Sociology.” Annual Review of Sociology 24: 1–24.; Joel Sobel (2002). “Can We Trust Social Capital?” Journal of Economic Literature 40: 139–54.

6 The value of social capital depends on a variety of factors, including the amount and value of information and other resources that can be shared between people, the qualities of relationships that facilitate sharing of resources (trust, empathy, respect, affection, mutual understanding, shared culture), the degree and nature of collective action that social interaction produces (including civic engagement and the creation of institutions), and the degree and intensity of investment and reinvestment in social relationships.

7 Jack Weatherford (1997). The History of Money. New York: Crown Publishers.

8 William Julius Wilson (1993). The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press.

9 Timothy P. Carney (2019). Alienated America: Why Some Places Thrive While Others Collapse. New York: Harper Collins. U.S. Congress, Joint Economic Committee, Chairman’s Staff (2019). “Losing Our Minds: Brain Drain across US States.” Report prepared by the Chairman’s staff, 116th Cong., 1st Sess. (April 2019).

10 Abraham Lincoln (1861). “Message to Congress in Special Session.” July 4, 1861. Washington, DC.

11 Carney (2019), 90.

12 Ibid, xiii.

13 Martin Luther King, Jr. (2010). Where Do We Go from Here? Chaos or Community? Boston: Beacon Press, 28.

14 Raj Chetty, Nathaniel Hendren, and Lawrence Katz (2016). “The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment.” American Economic Review 106(4): 855-902.

15 Raj Chetty, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez (2014). “Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States.” Quarterly Journal of Economics 129(4): 1553-1623.

16 Chetty et al. (2014) referred to it as “absolute mobility,” but a subsequent paper redefined that term to mean “having a higher income in adulthood than one’s parents.”

17 These correlations are estimated across aggregations of counties called “commuting zones.” However, in one of their most recent papers, the Chetty-Hendren team finds strong correlations across census tracts—broadly, neighborhoods—between single parenthood and upward mobility. Raj Chetty, John N. Friedman, Nathaniel Hendren, Maggie R. Jones, and Sonya R. Porter (2018). “The Opportunity Atlas: Mapping the Childhood Roots of Social Mobility.” Working paper.

18 Scholars have generally viewed the study of family structure as distinct from social capital research, but examining the nature of intra-family relationships should be seen as an inquiry into social capital.

19 U.S. Congress, Joint Economic Committee, Vice Chairman’s Staff of the Joint Economic Committee (2018). “The Geography of Social Capital in America.” Report prepared by the Vice Chairman’s staff, 115th Cong., 2nd Sess. (April 2018). Social Capital Project Report No. 1-18.

20 Raj Chetty and Nathaniel Hendren (2018a). “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects.” Quarterly Journal of Economics 133(3): 1107-1162. Raj Chetty, Nathaniel Hendren, and Lawrence Katz (2016).

21 Chetty and Hendren (2018a).

22 Raj Chetty and Nathaniel Hendren (2018b). “The Impacts of Neighborhoods on Intergenerational Mobility II: County-Level Estimates.” Quarterly Journal of Economics 133(3): 1163-1228.

23 Raj Chetty, Nathaniel Hendren, Frina Lin, Jeremy Majerovitz, and Benjamin Scuderi (2016). “Childhood Environment and Gender Gaps in Adulthood.” American Economic Review Papers and Proceedings 106(5): 282-288.

24 Raj Chetty, Nathaniel Hendren, Maggie R. Jones, Sonya R. Porter (2018). “Race and Economic Opportunity in the United States: An Intergenerational Perspective.” Working paper.

25 Other social-capital-related factors correlating with higher black upward mobility were higher racial and economic segregation and higher racial and immigrant concentration.

26 Similar research looking at other outcomes would be invaluable, and it is increasingly possible thanks to the data that the Chetty-Hendren-Friedman team have assembled. Their team, for instance, found that communities with higher scores on the social capital index and fewer children with a single mother actually have lower life expectancy among the poor people living there. (This study looks at the community of residence in adulthood rather than in childhood, however.) See Augustin Bergeron, Raj Chetty, David Cutler, Benjamin Scuderi, Michael Stepner, Nicholas Turner (2016). “The Association Between Income and Life Expectancy in the United States, 2001-2014.” The Journal of the American Medical Association 315(4): 1750-1766. Correlations are population-weighted and estimated by the Social Capital Project using data from the Opportunity Insights website (

27 Peter Rossi (1987). “The Iron Law of Evaluation and Other Metallic Rules.” Research in Social Problems and Public Policy 4: 3-20.

28 David B. Muhlhausen (2013). Do Federal Social Programs Work? Santa Barbara, CA: Praeger.

29 Yuval Levin (2017). The Fractured Republic: Renewing America’s Social Contract in the Age of Individualism. New York: Basic Books, 4.

30 The 1880-2010 trend is for women ages 18-54 and comes from the 1% samples of the decennial censuses, analyzed via the Integrated Public Use Microdata Series Online Data Analysis System, See Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek (2019). IPUMS USA: Version 9.0 dataset (Minneapolis: University of Minnesota).

The series covering 1940, 1950, 1955, 1960, 1965, and 1970 is for women aged 18 and older and comes from various Census Bureau Current Population Reports, Series P-20. The estimates are included in Table 79 of US Bureau of the Census (1971). Statistical Abstract of the United States: 1971. (92nd edition.) (Washington, DC).

The 1962-2018 series is for women 18-54 and comes from the Annual Social and Economic Supplement to the Current Population Survey, analyzed via the Integrated Public Use Microdata Series Online Data Analysis System, See Sarah Flood, Miriam King, Renae Rodgers, Steven Ruggles, and J. Robert Warren (2018). Integrated Public Use Microdata Series, Current Population Survey: Version 6.0 dataset (Minneapolis: University of Minnesota).

The census and CPS series do not count “married, spouse absent” as married, because prior to 1950 “separated” was not an option given to respondents, who likely reported their status as “married, spouse absent,” creating a discontinuity in the trend.

31 No author (1973). “100 Years of Marriage and Divorce Statistics, United States 1867-1967.” Data from the National Vital Statistics System 21(24). (Rockville, MD: National Center for Health Statistics). Table 3. Amber Lapp and David Lapp (2016). “A Generation Conflicted about Marriage.” In 2016 Index of Culture and Opportunity: The Social and Economic Trends that Shape America. (Washington, DC: Heritage Foundation).

32 Estimates are rates per 1,000 women entering their childbearing years. Both the series for whites for 1800-1990 and the overall series from 1933-1990 are from Michael R. Haines (2006). “Total fertility rate and birth rate, by race and age: 1800–1998.” Tables Ab52 and Ab63 in Susan B. Carter, Scott Sigmund Gartner, Michael R. Haines, Alan L. Olmstead, Richard Sutch, and Gavin Wright eds. Historical Statistics of the United States, Earliest Times to the Present: Millennial Edition (New York: Cambridge University Press.)

1991-2009: Joyce A. Martin, Brady E. Hamilton, Michelle J.K. Osterman, Sally C. Curtin, and T.J. Mathews (2015). “Births: Final data for 2013.” National Vital Statistics Reports 64(1). (Hyattsville, MD: National Center for Health Statistics).

2010-2017: Joyce A. Martin, Brady E. Hamilton, Michelle J.K. Osterman, Anne K. Driscoll, and Patrick Drake (2018). “Births: Final data for 2017.” National Vital Statistics Reports 67(8). (Hyattsville, MD: National Center for Health Statistics).

33 Ibid.

34 Claudia Goldin (2018). “Gender Gap.” In The Library of Economics and Liberty (Carmel, IN: Liberty Fund).

35 Thomas D. Snyder (1993). “120 Years of American Education: A Statistical Portrait.” (Washington DC: National Center for Education Statistics). Claudia Goldin and Lawrence F. Katz (2002). “The Power of the Pill: Oral Contraceptives and Women’s Career and Marriage Decisions.” Journal of Political Economy 110(4): 730-770.

36 For a discussion, see U.S. Congress, Joint Economic Committee, Social Capital Project (2017). “What We Do Together: The State of Associational Life in America.” Report prepared by the Vice Chairman’s staff, 115th Cong., 1st Sess. (May 2017). SCP Report No. 1-17.

37 According to OECD data, births per 1,000 women were down 59 percent in Canada between 1960 and 2010, down 24 percent in Ireland, and down by an intermediate amount in 15 other developed countries (Australia, Austria, Belgium, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, New Zealand, Norway, Spain, Switzerland, and the United States). Sweden was an outlier, with only a 9 percent drop. Of the 17 of these countries with data from 1960 to 2000, 14 experienced the biggest declines in fertility in either the 1960s or (most often) the 1970s. Data for the United Kingdom is available only through 1994, but it experienced its biggest decline in fertility in the 1970s too. See Female labor force participation plateaued in the US in the late 1990s but continued rising in the rest of the OECD. See

38 See the Census Bureau’s Historical Living Arrangements of Children Tables, Table CH-1.

39 Larry L. Bumpass (1984). “Children and Marital Disruption: A Replication and Update.” Demography 21(1): 71-82; James A. Sweet and Larry L. Bumpass (1987). American Families and Households (New York: Russell Sage Foundation); Arthur J. Norton and Paul C. Glick (1986). “One Parent Families: A Social and Economic Profile.” Family Relations 35(1): 9-17.

40 For a discussion, see U.S. Congress, Joint Economic Committee, Social Capital Project (2017). “Love, Marriage, and the Baby Carriage: The Rise in Unwed Childbearing.” Report prepared by the Vice Chairman’s staff, 115th Cong., 1st Sess. (December 2017). SCP Report No. 3-17.; U.S. Congress, Joint Economic Committee, Social Capital Project (2018). “Rising Unwed Pregnancy and Childbearing across Educational and Racial Groups.” Report prepared by the Vice Chairman’s staff, 115th Cong., 2nd Sess. (February 2018).

41 1940-1999 estimates are from Stephanie J. Ventura and Christine A. Bacharach (2000). “Nonmarital Childbearing in the United States, 1940-99.” National Vital Statistics Reports 48(16). (Hyattsville, MD: National Center for Health Statistics).

2000-2014 estimates are from Brady E. Hamilton, Joyce A. Martin, Michelle J.K. Osterman, Sally C. Curtin, and T.J. Mathews (2015). “Births: Final Data for 2014.” National Vital Statistics Reports 64(12). (Hyattsville, MD: National Center for Health Statistics).

2015 estimate is from Joyce A. Martin, Brady E. Hamilton, Michelle J.K. Osterman, Anne K. Driscoll, and T.J. Mathews (2017). “Births: Final Data for 2015.” National Vital Statistics Reports 66(1). (Hyattsville, MD: National Center for Health Statistics).

2016 estimate is from Joyce A. Martin, Brady E. Hamilton, Michelle J.K. Osterman, Anne K. Driscoll, and T.J. Mathews (2018). “Births: Final Data for 2016.” National Vital Statistics Reports 67(1). (Hyattsville, MD: National Center for Health Statistics).

2017 estimates are from Joyce A. Martin, Brady E. Hamilton, Michelle J.K. Osterman, Anne K. Driscoll, and Patrick Drake (2018). “Births: Final data for 2017.” National Vital Statistics Reports 67(8). (Hyattsville, MD: National Center for Health Statistics).

42 For a review of the most sophisticated research, see Sara McLanahan, Laura Tach, and Daniel Schneider (2013). “The Causal Effects of Father Absence.” Annual Review of Sociology 39. For a critical review of the literature through 2004, see Scott Winship (2005). “The Effects of Parental Marital Status and Living Arrangements on Child Outcomes in the United States: An Overview of the Field.” Unpublished paper.

43 This is a problem even conditioning on fairly gross demographic and economic characteristics, as research tends to do, which presents other methodological problems at any rate.

The same problem recurs if we consider a different counterfactual and compare children born to unmarried parents to the different children who would have been born if the parents could have been nudged to delay fertility until they were married (potentially to other people).

There is a symmetric problem with establishing the outcomes of children who grow up with two-parents under the counterfactual that their parents were induced to not marry or to divorce. Different inducements would affect different families. Moreover, it is unlikely that the real-world outcomes of the children of single parents are a good guide to the outcomes we’d see among the children of happily-married parents if a father or mother were somehow removed from the picture. The closest example we have to a randomized experiment is when a parent dies unexpectedly, but the effect of having a parent die on child outcomes is unlikely to resemble the effect of going through a parental divorce, for example. (And parental death is generally not random, being at least somewhat predictable from various demographic and economic factors.)

44 This issue generalizes to all social science inquiries involving human choice. It is very likely that the human subjects in a study have a better sense of what is best for them (or in this case, their children) than does the researcher observing them as data points.

45 Chetty et al. (2014) report that the correlation between a community’s share of children with a single parent and its upward mobility was -0.76 and was still as strong as -0.66 if they estimated a community’s mobility only from the children of married parents. Chetty, Friedman, Hendren, Jones, and Porter (2018) report that the correlation across neighborhoods between the share of single parent households and upward mobility was -0.59, and it remained as strong as -0.52 if they only looked at the children of single parents. That is, even among the children of single parents, upward mobility is lower when there are more single parents in a tract. And as noted above, Chetty, Hendren, Jones, and Porter (2018) found that black men had stronger upward mobility the more low-income black fathers there were in their childhood neighborhood, regardless of whether or not their own father was present.

46 Percentage employed and labor force participation rate: From 1910 to 1960, estimates are for non-institutionalized men ages 25 to 54 who are not in the Armed Forces, and they come from the 1% samples of the decennial censuses, analyzed via the Integrated Public Use Microdata Series Online Data Analysis System, See Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek (2019). From 1948 to 2018, estimates are for the civilian non-institutionalized population of men ages 25 to 54, and they come from the Labor Force Statistics Database available at

Average weekly hours: From 1900 to 1977, estimates are for employed men not in school, from John Owen (1976). “Workweeks and Leisure: An Analysis of Trends, 1948-1975.” Monthly Labor Review 99(8): 3-8; John Owen (1988). “Work-time Reduction in the United States and Western Europe.” Monthly Labor Review 111(12): 41-45. Both cited in Robert Whaples (2001). “Hours of Work in U.S. History.” EH.Net Encyclopedia, edited by Robert Whaples. From 1980 to 2017, estimates are average usual weekly hours worked for employed men, and they come from the 1% samples of the decennial census and from the American Community Survey, analyzed via the Integrated Public Use Microdata Series Online Data Analysis System, See Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek (2019).

47 Ibid.

48 For a discussion, see U.S. Congress, Joint Economic Committee, Social Capital Project (2018). “Inactive, Disconnected, and Ailing: A Portrait of Prime-Age Men Out of the Labor Force.” Report prepared by the Vice Chairman’s staff, 115th Cong., 2nd Sess. (September 2018).

49 On the opioids crisis, see U.S. Congress, Joint Economic Committee, Social Capital Project (2018). “The Numbers Behind the Opioids Crisis: Revised Utah Edition.” Report prepared by the Vice Chairman’s staff, 115th Cong., 2nd Sess. (October 2018).; and U.S. Congress, Joint Economic Committee, Social Capital Project (2017). “The Rise in Opioids Overdose Deaths.”

50 Scott Winship (2017). “Declining Prime-Age Male Labor Force Participation: Why Demand- and Health-Based Explanations are Inadequate.” Mercatus Working Paper. (Arlington, VA: Mercatus Center at George Mason University).

51 Scott Winship (2015). “How to Fix Disability Insurance.” National Affairs 23.

52 Ross Douthat (2013). “A World Without Work.” New York Times (February 24, 2013).

53 Reihan Salam (2013). “Worklessness, Stratification, and Normative Diversity.” National Review Online. The Agenda blog (February 25, 2013).

54 U.S. Congress, Joint Economic Committee, Social Capital Project (2018). “Inactive, Disconnected, and Ailing.”

55 On loneliness, see U.S. Congress, Joint Economic Committee, Social Capital Project (2018). “All the Lonely Americans?” Report prepared by the Vice Chairman’s staff, 115th Cong., 2nd Sess. (August 2018).

56 Scott Winship (2017). “Economic Mobility: A State-of-the-Art Primer, Contemporary Levels of Mobility.” (Washington, DC: Archbridge Institute). Figure 3.

57 Assuming the standard deviation of the IQ test scores is 15 points. The math score gap between low and high socioeconomic status kindergartners entering school in 2010 was 1.25 standard deviations. See Emma Garcia and Elaine Weiss (2017). “Education Inequalities at the School Starting Gate: Gaps, Trends, and Strategies to Address Them.” (Washington, DC: Economic Policy Institute).

58 The math score gap between the children of parents with less than a high school education and the children of parents with graduate degrees is 0.882 standard deviations in 2015. See the NAEP Data Explorer tool at

59 Eric A. Hanushek, Paul E. Peterson, Laura M. Talpey, and Ludger Woessmann (2019). “The Unwavering SES Achievement Gap: Trends in U.S. Student Performance.” NBER Working Paper No. 25648 (Cambridge, MA: National Bureau of Economic Research).

60 These calculations are based on Social Capital Project analyses of Intervention Reports data from the What Works Clearinghouse, available at

61 No author (undated). “What Works Clearinghouse Procedures and Standards Handbook, Version 3.0.”

62 Ben Todd (2017). “Is It Fair to Say that Most Social Programmes Don’t Work?” 80,000 Hours.

63 For example: the What Works Clearinghouse, Arnold Ventures, Abdul Latif Jameel Poverty Action Lab at MIT, the Campbell Collaboration, Blueprints for Healthy Youth Development, Social Programs that Work, and the Washington State Institute for Public Policy.

64 U.S. Congress, Joint Economic Committee, Social Capital Project (2017). “What We Do Together: The State of Associational Life in America.”

65 Oren Cass (2018). The Once and Future Worker: A Vision for the Renewal of Work in America (New York: Encounter).

66 Eric Klinenberg (2018). Palaces for the People: How Social Infrastructure Can Help Fight Inequality, Polarization, and the Decline of Civic Life (New York: Crown).

Latest News