angle-up angle-right angle-down angle-left close user menu open menu closed search globe bars phone store

Choosing Our Future: A Story of Opportunity in America
Irwin Kirsch, Henry Braun, Mary Louise Lennon, and Anita Sands

The Role of Social Capital

If income inequality is the main economic problem, it could be solved tomorrow, through confiscation and redistribution. If the main problem is the unequal generation of social capital in institutions such as families, schools and communities, the solutions get more difficult. One task can be accomplished by a tax collector; the other is the work of a civilization.

Michael Gerson, columnist for the Washington Post and senior policy advisor in the George W. Bush administration 55

The family structure and social networks into which children are born and raised, the behavioral norms they develop, and the trust that connects members of their communities – factors often referred to as social capital, impact opportunity throughout an individual's lifetime. Parents may regularly take a young child to their local library and help with homework as the child grows, fostering academic achievement as well as community and school connections. One teenager may alert another to a local business that has after-school jobs available, presenting the possibility of a paycheck along with the opportunity to develop work-related skills, including learning the importance of showing up on time, appropriately dressed, and ready to work. The leader of a religious community may help young members through the process of applying to college and seeking financial aid so that they can be the first members of their families to pursue higher education. In each of these instances, social capital acts as a catalyst to foster development and provide much-needed support.

Social capital includes both informal connections to family, friends, and acquaintances and more formal connections through participation in religious communities, sports teams, volunteer groups, political organizations, professional organizations, and unions. These connections foster norms that guide values, personal decisions, and social interactions. The domain of social capital is defined by some social scientists as having a number of different dimensions. 56 Two of the most important are:

  • Bonding capital, which consists of strong relationships within groups, connects individuals with similar backgrounds or characteristics, often family and close friends. Bonding capital builds social cohesion and a sense of solidarity. When members are sick, out of work, or having other difficulties, it can provide various types of help, acting as an informal social safety net.
  • Bridging capital, which consists of weaker relationships across groups, brings together individuals who may be of different races, ethnicities, educational backgrounds, classes, religions, or ages. Bridging capital fosters the exchange of information and ideas and can help build consensus, taking into account the diverse nature of its members.
thumbnail for social capital video

Social Capital

Source: Educational Testing Service

Strong social capital can have positive impacts on both individuals and communities. Networks of parents and community members focusing on the well-being of children, as well as individuals focusing on the vitality of their neighborhoods and larger communities, can lead to improved school performance, lower crime rates, better public health, and reduced political corruption.57 With strong social capital, community members are more likely to join together to clean up a neighborhood park, meet with political leaders to influence local policy, set up a childcare cooperative to support working families, or organize a food pantry to help neighbors in need.

A lack of social capital can hinder opportunity, even for those with strong human capital. For example, high-achieving students may not reach their full potential if they are not encouraged – or do not have access to – individuals and resources that can help them navigate the system to enroll in advanced coursework or apply to, and succeed at, competitive colleges.58

Although fostering positive social capital is one component of expanding opportunity, it is important to recognize that social capital can have negative impacts as well.59 Social networks can be exclusionary, denying access to resources and assistance to those outside of the group. They can endanger rather than build up communities by enabling violence or crime. Like all social endeavors, social capital reflects the perspectives and intentions of the individuals involved.

People have always had networks; their actions have always been influenced by values and social norms. However, just as changes in the economic landscape have increased the importance of human capital, as well as the differences in accumulated capital between those who have had opportunities to develop it and those who have not, changes in the social landscape have affected the ability of individuals to develop positive social capital. Indeed, changes over time in families, neighborhoods, and participation in social, civic, and religious organizations have impacted social connections and often narrowed the range of people with whom individuals interact.

In addition, there is growing evidence that social capital is becoming more strongly related to human capital and that the ensuing divergence in outcomes that results from differences in both types of capital has increased.

The Relationship between Human and Social Capital

In combination, human and social capital help to determine the kind of work you do, who your coworkers and friends are, where you live, and your choice of spouse or partner. There have, however, been significant changes in the nature of their relationship over the past several decades – changes that have contributed to the growing divergence in outcomes that we see today. Whereas, in the past, strong social networks and norms of civic engagement transcended socioeconomic status and levels of educational attainment, social capital is now more tightly tied to human capital. Individuals with strong human capital, and the higher economic status typically associated with greater skills, also tend to have the networks, norms, and values that provide a greater benefit in today's economy.60

Whereas, in the past, strong social networks and norms of civic engagement transcended socioeconomic status and levels of educational attainment, social capital is now more tightly tied to human capital.

In contrast, those with weaker skills not only tend to be much worse off economically than they were two generations ago but also are disadvantaged by more limited social capital. This is due, in part, to the changing characteristics of their neighborhoods and communities, along with the stresses associated with working multiple jobs and not earning a livable wage. Such circumstances make it likely that they will rely on the kinds of social connections that are essential for getting by, but that they will not develop the broader networks and norms that will help them, or their children, get ahead.

The Growing Divide

Evidence that the development of beneficial social capital has become more stratified by education and skills can be found in a variety of indices, including the extent to which people are engaged in civic activities, their trust in social institutions and in others, and, perhaps most important, their patterns of family formation.61 Each is described in more detail below.

Civic engagement

Civic participation, including activities such as voting and volunteering, has declined overall. Some 93 million eligible voters did not vote in the presidential election of 2012. Voter turnout was 57 percent, down from 62 percent in 2008 and 60 percent in 2004.62 Like many other indices of social capital, voting patterns are not uniform throughout the country but vary widely across states and localities. In 2012, for example, the percentage of eligible voters who voted ranged from a high of 76 percent in the District of Columbia to a low of 48 percent in West Virginia.63 Turnout for local elections is even lower and, based on data for 144 of the largest cities in the U.S., has been declining over the last several decades, with an average of about 21 percent of eligible voters participating in local elections.64

In addition, voting patterns have become increasingly stratified by age, education, and income. Citizens over the age of 35 vote at higher rates than do younger people. The highest voting rates are seen among the most educated. And those with higher household incomes are more likely to vote than those in poorer households.65

Volunteering, which both reflects and helps build social connections, is also increasingly associated with specific demographic characteristics. Those who are not in the labor force volunteer less than those who work full or part time. College graduates tend to volunteer at rates that are four times higher than those with less than a high school diploma. Robert Putnam and his colleagues report similar differences among Americans under the age of 18, with children of college-educated parents volunteering more, while rates have not changed for children whose parents have not gone beyond high school.66 Those young people who do not volunteer are less likely to benefit from the social connections and experience that volunteering affords – not to mention the satisfaction that comes from helping their communities.


Declines in civic participation may reflect a more fundamental issue. Across age groups, educational levels, and economic status, Americans are losing trust – in political leaders, corporations, social institutions, and in each other. A 2014 study based on two nationally representative surveys that have been conducted since the 1970s reported that, "Trust in others and confidence in institutions, two key indicators of social capital, reached historic lows among Americans in 2012."67 Across generations, Americans expressed a lack of confidence in large institutions including business, Congress, the presidency, the news media, religious organizations, and the medical establishment.

Generalized trust in others has also declined overall over the past 40 years. As we have seen with other indices of social capital, trust varies by levels of education. For example, Figure 3 shows that those with a bachelor's degree are more likely to say that people can be trusted than those without high school diplomas. Those with advanced degrees are even more likely to express trust in others and are the only group for whom trust has been on the rise over the past several years.68

Figure 3: Percentage of the U.S. Population Who Believes That "People Can Be Trusted" by Highest Level of Educational Attainment, 1972 – 2014

Source: General Social Survey, 1972-2014. Data have been smoothed.

In his book, Our Kids, Robert Putnam also shows strong associations between trust and economic status, citing evidence that residents in affluent neighborhoods trust their neighbors more than residents in poor communities do. This lack of trust may be rooted in decades-long declines in community bonds. Putnam documents this decline, focusing specifically on differences by educational attainment. Better educated Americans tend to report having more close friends and broader social networks (i.e., more bridging capital). In contrast, less educated adults tend to have fewer and more homogeneous networks and more familial than non-kin ties. The connections for those with wider social networks can provide expertise and support, both for themselves and their children, that are simply not available to those operating within a narrower social context.

Thumbmail: Increasing Gap in Single-Parent Families by Parents' Education and Percent of Children Living in a Single-Parent Family, United States

Increasing Gap in Single-Parent Families by Parents' Education

Chart showing Increasing Gap in Single-Parent Families by Parents' Education and Percent of Children Living in a Single-Parent Family, United States

Source: Details: Percent of children in single-parent families by parents' highest level of education. Source: Carsey School of Public Policy, University of New Hampshire. See: Accessed July 10, 2015. Adapted from IPUMS-USA, University of Minnesota, Data: 1960-2000 U.S. Decennial Census; 2008-2012 ACS Five-Year Estimates (2010).

Line graph showing the increasing gap in single-parent families in the U.S. by parents' education. Two lines show those with high school education or less and those with a bachelor's degree or more. The horizontal axis shows the years 1960 to 2010 in 10-year intervals. The vertical axis shows percentages of children by 10 percent intervals. The percent of single-parent families where parents have a high school education or less has increased from around 5 percent to just over 50 percent in 2010. The percent for those with a bachelor's degree or more has risen slowly from just over 0 percent in 1960 to slightly over 10 percent in 2010.


While family structure has become increasingly varied across social and economic lines over the past two generations, differences by education and skills can be found here as well. Changes in marriage rates provide one indicator. Census data from 2012 show that marriage rates increased by 3 percent over those in 2011. However, almost all of that increase (87 percent) was the result of increasing marriage rates among those with a college degree or higher. Today, poorly educated adults are less likely to ever get married than in the past.69

Any discussion of family structure must also take into account couples who choose to live together without marrying, as cohabitation rates have increased steadily over the past few decades. Cohabitation rates for adults between the ages of 30 and 44 have more than doubled over the past 20 years.70 However, adults without a college degree are almost twice as likely to cohabit as those who are college educated. This difference by educational attainment is important due to the associated economic consequences and, by extension, the impact on opportunity.

Figure 4: Median Adjusted Household Income by Educational and Partnership Status, 2009

Source: Paul Taylor & Richard Fry, "Living Together: The Economics of Cohabitation." Pew Research Center, 2012 Graph title: Median Adjusted Household Income by Education and Partnership Status, 2009. Based on 30- to 44-year-olds. "No partner" includes those living without an opposite-sex partner or spouse. Income adjusted for household size and scaled to a household size of three. Adapted from 2009 American Community Survey (ACS) Integrated Public Use Micro Sample.

As Figure 4 shows, there is little difference in household income for a college graduate who cohabits versus one who marries. However, an individual without a college degree who cohabits is typically worse off financially than one who is married.71 The fact that cohabiters without college degrees are more likely to have children than their college-educated counterparts contributes to this income difference. In general, cohabitating couples with children tend to be younger, less educated, and have less secure employment and lower incomes, all characteristics that can affect both their opportunities and those that they are able to provide for their children.72

In contrast, couples who begin their lives together with shared high educational attainment and skills are in the advantageous situation where, just like the "magic" of compound interest, they stand to benefit from compounding incomes, wealth, and social capital over their lifetimes and are likely to pass those benefits along to their children.

Compounding Advantage

The human and social capital accumulated by individuals is hugely important, as it impacts the transmission of opportunity from one generation to the next. In part, that transmission is influenced by the tendency for people to marry or partner with those like themselves, a phenomenon referred to as assortative mating. In the past, partners tended to share similar cultural, religious, or ethnic backgrounds. Today, educational attainment and economic status are more likely to be the characteristics that partners have in common. Those who attend and complete college or postgraduate and professional programs become part of a social network that includes other highly educated individuals. And, because that network includes growing numbers of women, this has changed the likelihood of assortative mating. That, in turn, leads to compounding economic and social advantages for some families and greater opportunities for their children.

The human and social capital accumulated by individuals is hugely important, as it impacts the transmission of opportunity from one generation to the next.

Women are increasingly better educated and more likely to be in the workforce. Today, women's graduation rates are more than five times greater than those in the early 1960s. In 1960, 6 percent of women ages 25 or older had completed four or more years of college. By 2014, that had increased to 32 percent.73 The numbers of women earning postgraduate degrees has also dramatically increased. In the 2011-2012 academic year, 60 percent of the students receiving master's degrees were women.74 The number of women receiving a professional degree in 2010 was almost 20 times the number in 1970.75

Similarly, women are working outside the home more than in the past. Although less than one-third of women were in the workforce immediately following World War II, by 2013 close to 60 percent of women were in the workforce, with a large share of them working full time and year round.76 For some women, this change in labor market status reflects a realization of aspirations. For others, it has been an economic necessity.

One consequence of these trends is that highly educated men and women are much more likely to meet and ultimately marry or partner with each other than in the past. In 1960, only 3 percent of married couples were both college educated; by 2012 that figure rose to 22 percent.77 This phenomenon, when combined with the earnings gap favoring workers with higher levels of education, results in an even greater stretching of the income distribution.

Figure 5 expands on the data presented in Figure 1, showing weekly earnings for both men and women by level of education over the past 50 years. This figure illustrates the substantial economic advantages available to a two-income couple where both partners have postgraduate degrees versus a couple with lower levels of educational attainment or, even more starkly, a single parent who has less than a college education.

Figure 5: Changes in Real Wage Levels of Full-Time U.S. Workers by Sex and Education, 1963 – 2012

Source: David H. Autor, Science 2014; 344: 843-851.

Analyses of U.S. Census data show that, in 1960, a family with two working postgraduate degree holders earned an income that was 176 percent above the national household average; by 2005, that increased to a 219 percent advantage. In contrast, couples with a high school education went from earning 103 percent of the national average in 1960 to 83 percent in 2005.78 For those with children, the contrast in economic resources between two working parents with postgraduate degrees versus a single parent with a high school diploma or less sets the stage for massive differences between these families in terms of human and social capital.

Changes in social capital that have impacted assortative mating are among the factors that have contributed to the growing disparities in income in this country. Figure 6 illustrates these differences by looking at trends in income inequality for children based on family incomes since 1967. The gap between children with family incomes at the top of the income distribution (those at the 90th percentile) and those at the bottom (the 10th percentile) grew steadily from 1970 to 2010, more than doubling in 25 years. From 1967 to about 1987, this was driven principally by the growing gap between those at the middle of the income distribution (those at the 50th percentile) and those at the bottom. The 50/10 income ratio grew by 64 percent over that time. After that period, the gaps were driven largely by increasing inequality between those at the 90th and 50th percentiles.79 This growing divide over the last 25 years or so is driven by a number of factors, one of which is the rise in assortative mating.

Figure 6: Trends in Family-Income Inequality among School-Age Children, 1967 to 2008 (Weighted by Number of School-Age Children)

Source: Whither Opportunity?: Rising Inequality, Schools, and Children's Life Chances by Duncan, Greg J.; Murnane, Richard J. Reproduced with permission of Russell Sage Foundation in the format Republish in other published product via Copyright Clearance Center. Author's calculations, based on U.S. Bureau of Census. Note: Each line shows the trends in the ratio of household incomes at two percentiles of the income distribution trends. All are divided by their value in 1967 in order to put the trends on a common scale.

Another way to understand these disparities is to look at the relative share of household income held by those at different points in the income distribution. Table 1 divides the population into five equal groups, or quintiles, and shows the share of household income held by each group. Note that within quintiles there have been changes in household income, with those in the bottom two quintiles seeing their share of total household income decrease by 20 and 22 percent, respectively. This table shows that the story is not just about the gap between the very rich and the very poor. For every group below the top 20 percent, the relative share of household income has declined since 1967.

Table 1: Shares of Household Income by Quintiles, 1967 – 2013

Source: Selected Measures of Household Income Dispersion: 1967 to 2013, Table A-2, U.S. Census, Income and Poverty in the United States, 2013. Income figures in 2013 dollars.

Real Hourly Wages by Select Percentiles, 1973 - 2012, All Workers (2012 dollars)

Real Hourly Wages by Select Percentiles, 1973 - 2012, All Workers (2012 dollars)

Source: Economic Policy Institute's The State of Working America, 12th Edition. Data from Current Population Survey Outgoing Rotation Group microdata. Note: The xth-percentile wage is the wage at which x% of the wage earners earn less and (100-x)% earn more.

Persistent gaps in human and social capital both reflect, and contribute to, the polarization of life outcomes and outlooks. In particular, gaps are associated with differences, ranging from economic outcomes such as wages and benefits to health and longevity, as well as to social outcomes such as trust and civic engagement. Equally important is the impact these gaps can have on life outlooks, including whether people have hope for the future and a belief that, with hard work, the American Dream is still within reach.

Such disparities are significant enough for the individuals living with them. But the fact that these gaps also characterize disparities in the circumstances of birth for the next generation makes them all the more important to understand. In the following section, we explore in more depth how differences in parental human and social capital are transmitted to children at birth and beyond, as well as how these differences shape children's opportunities.


55 Michael Gerson, "Our Disconnected Working Class," Washington Post, May 15, 2014,

56 See, for example, "Dimensions of Social Capital," The Social Capital Community Benchmark Survey,, and Social Capital Glossary, Saguaro Seminar: Civic Engagement in America (Cambridge, MA: Harvard Kennedy School, accessed September 10, 2015),

57 David Halpern, The Hidden Wealth of Nations (Boston: Polity, 2009).

58 See the Expanding College Opportunities Project, a low-cost intervention for high-achieving, low-income students, in Caroline Hoxby and Sarah Turner, "Expanding College Opportunities," Education Next 13, no. 4 (2013), See also a College Board project designed to expand access to scholarship opportunities earlier in high school to change students' trajectories and help inform their decisions about pursuing college in, "College Board Announces Major Expansion in Access to Scholarships for the Millions of Students Who Take the PSAT/NMSQT®," press release, January 29, 2015,

59 Answers to Frequently Asked Questions about Social Capital, Saguaro Seminar: Civic Engagement in America (Cambridge, MA: Harvard Kennedy School, accessed September 10, 2015,

60 Robert Putnam, Bowling Alone: The Collapse and Revival of American Community (New York: Simon & Schuster, 2000); Putnam, Our Kids.

61 This section draws on the work of Robert Putnam and Charles Murray looking at the stratification of social capital by education and skills.

62 2012 Voter Turnout Report, Bipartisan Policy Center, November 8, 2012,

63 U.S. Census Bureau, Current Population Survey, November 2012,

64 Mike Maciag, "Voter Turnout Plummeting in Local Elections," Governing the States and Localities, October 2014,

65 Richard J. Coley and Andrew Sum, Fault Lines in Our Democracy  (Princeton, NJ: Educational Testing Service, 2012),

66 Robert D. Putnam, Carl B. Frederick, and Kaisa Snellman, Growing Class Gaps in Social Connectedness among American Youth, Harvard Kennedy School of Government, Saguaro Seminar: Civic Engagement in America (Cambridge, MA: Harvard Kennedy School, August 8, 2012),

67 Jean M. Twenge, W. Keith Campbell, and Nathan T. Carter, "Declines in Trust in Others and Confidence in Institutions among American Adults and Late Adolescents, 1972-2012," Psychological Science, doi:10.1177/0956797614545133.

68 Authors' calculations based on General Social Survey data on trust by educational attainment, 1972–2014.

69 Murray, Coming Apart. Statistics from the Pew Research Center show that in 1950, 92 percent of 35- to 39-year-olds without a college degree had ever married; in 2012 that was the case for only 73 percent of this group. (Note that because of the increase in college graduation rates, the characteristics of those without a college degree have changed.)  Richard Fry, New Census Data Show More Americans Are Tying the Knot, But Mostly It's the College-Educated, Fact Tank, Pew Research Center, February 6, 2014,

70 Richard Fry and D'Vera Cohn, "Living Together: The Economics of Cohabitation," Social & Demographic Trends, Pew Research Center, June 27, 2011,

71 Ibid.

72 Family Structure:  Indicators on Children and Youth, Child Trends Data Bank, March 2015,

73 Authors' calculations based on figures in United States Census Bureau, Years of School Completed by People 25 Years and Over, by Age and Sex: Selection Years 1940 to 2014, CPS Historical Time Series Tables, Table A-1,

74 Most Popular College Degrees for Men and Women,, June 1, 2015,

75 The Rise of Women: Seven Charts Showing Women's Rapid Gains in Educational Achievement  (New York: Russell Sage Foundation, February 21, 2013),

76 U.S. Bureau of Labor Statistics, Women in the Labor Force: A Databook, BLS Reports, Report 1052, May 2014,

77 Wendy Wang, Record Share of Wives Are More Educated than Their Husbands, Fact Tank, Pew Research Center, February 12, 2014,

78 Rich Morin, "New Academic Study Links Rising Income Inequality to ‘Assortative Mating,' " Fact Tank, Pew Research Center, January 29, 2014,

79 Reardon, "Widening Academic-Achievement Gap."