The rhetorical attraction of opportunity is irresistible. Every politician in the land sings its praises, laments its absence, or promises its restoration. Opportunity is a leitmotif not only of American political discourse but of American culture: Horatio Alger, the frontier, the land of opportunity, the American Dream … you know the drill.

Take these two quotes—one from President Obama, a Democrat: “Opportunity is who we are … but upward mobility has stalled”—and the other from U.S. Representative Paul Ryan , a Republican and now Speaker of the House: “Upward mobility is the central promise of life in America, but right now, America’s engines of upward mobility aren’t working the way they should.”

Rhetorical agreement that America ought to be a land of opportunity is, of course, hardly news. But it is significant that most senior political figures now agree that we are falling far short of this ideal. Mounting empirical evidence that rates of intergenerational social mobility in the U.S. are low and flat has finally penetrated the American political consciousness. A chance for some bipartisan work to address social mobility has presented itself, a precious moment that ought to be seized.

But while the U.S. is a world leader in opportunity rhetoric, it is something of a laggard for opportunity metrics. Indicators are necessary to guide policy, drive data collection strategies, and measure progress. There are clear summary statistics of economic growth, poverty, and productivity, Why not opportunity? We need clear concepts and credible indicators of opportunity to have an idea of whether we are even headed in the right direction. Right now, indicators are the poor relation of the social policy world, lacking either the immediacy of strong rhetoric or the tangibility of policies and programs. Indicators are the missing link in our attempts to promote opportunity.

Indicators can act as the point of contact between goals, initiatives, and data. First, of course, the overall goal has to be established and given a clear conceptual basis. Then indicators can be drawn together or developed to show long-run progress toward that goal. In addition, shorter-term “leading indicators” can also be defined. Initiatives—a deliberately broad term encompassing government policies and programs, but also work by nongovernmental organizations or even corporations—can then be judged against these indicators.

Evidence-based policy is obviously preferable to what we often get, which is policy-based evidence making. But evidence of what is the important question—to which indicators provide an answer. Last but not least, the generation of indicators can shape and promote new approaches to data collection.

In the remainder of this chapter, I will:

  1. (a)

    Position equal opportunity as the unavoidably American vision of fairness;

  2. (b)

    Argue for a definition of opportunity based on intergenerational relative mobility ;

  3. (c)

    Describe current levels of mobility, and the relationships between mobility patterns and family structure , education, and race;

  4. (d)

    Provide a brief history of the social indicators movement in the U.S.;

  5. (e)

    Outline some of the theoretical terrain of indicator development;

  6. (f)

    Describe two current examples of indicator frameworks—from the United Kingdom and Colorado —and;

  7. (g)

    Propose four specific reforms to elevate the role of indicators in the promotion of opportunity: setting a long-term Goal for Intergenerational Mobility; a “dashboard” of Annual Opportunity Indicators; an American Opportunity Survey; and a Federal Office of Opportunity.

All-American: Equal Opportunity as Egalitarian Individualism

The volume in your hands (or perhaps, more likely on your screen) is one of thousands with the word “opportunity” in its title. Especially in America, opportunity is a term redolent of optimism, progress, and freedom. It is, in short, impossible to be against. The danger is that opportunity becomes a protean term, meaning almost anything, or something different to different people in different contexts. Some specificity is therefore required in order to move beyond rhetoric and into action.

I will shortly argue for a specific concept of opportunity, namely relative intergenerational income mobility . But first I will attempt to define equal opportunity as a distinctly American kind of fairness. In his second inaugural address in 2013, Obama declared: “We are true to our creed when a little girl born into the bleakest poverty knows that she has the same chance to succeed as anybody else, because she is an American; she is free, and she is equal, not just in the eyes of God but also in our own.”

So: the “same chance to succeed,” even though “born into the bleakest poverty.” This is the utopian ideal of American fairness, in which the inequalities of birth do not dictate the inequalities of life . While Obama, like most politicians, focused on upward mobility out of poverty, the equal opportunity ideal reaches all the way up the distribution. It is about the chance for a middle class kid to join the elite, as well as for a poor kid to join the middle class. The ideal also goes deeper than political rhetoric. Equality of opportunity is in America’s DNA. The moral claim that each individual has the right to succeed is even implicit in the proclamation of Declaration of Independence that “All men are created equal.” In his first draft of that historic document, Thomas Jefferson in fact wrote that all were created “equal and independent.” This is the distinctly American formula—equality plus independence adds up to the promise of upward mobility. Equal opportunity reconciles individual liberty—the freedom to get ahead and “make something of yourself”—with societal equality. It is how the ideal of natural equality—“born equal” is fused with the ideal of individualism—“born independent.” It is a philosophy of egalitarian individualism.Footnote 1

Chris Hayes writes of social mobility in his book Twilight of the Elites: “Those on the bottom who make it to the top rise from their class rather than with it. It is a fundamentally individualistic model of achievement” (Hayes 2013, 23). Hayes wishes it could be different. But that is wishful thinking. Individualism is hard-wired into the very idea of America. The challenge is to ensure that it is genuinely combined with equality of opportunity. Hayes laments, “[T]he meritocratic creed finds purchase on both the left and the right because it draws from each…. It is ‘liberal’ in the classical sense.” Indeed it is—just like America.

Opportunity Equals Intergenerational Relative Mobility

Even the term “equality of opportunity” is, of course, very broad. The philosopher Bertrand Russell, asked what he actually did all day, replied: “[Y]ou clarify a few concepts, make a few distinctions. It’s a living.” Concepts and distinctions will be important, too, for the motivating project of this volume. We have to be crystal clear what we mean when we talk or write about “opportunity” and equally clear about the distinctions being made between different variants. Amartya Sen , the Nobel Prize-winning economist, famously argued that because everyone favors equality of one sort or another, the key question is: “Equality of what?” (Sen 1979, 1). So, in the spirit of Sen, what do we mean by “equality of opportunity?”

In particular, is our main concern with absolute mobility or relative mobility ? Relative mobility is, as Scott Winship puts it, “a measure of how the ranking of adults against their peers is (or is not) tied to the ranking of their parents against their peers. That is to say, ignoring dollar amounts, did adults who rank high or low in the income distribution also have parents who ranked high or low?” (Economic Policies for the 21st Century 2014). By contrast, absolute mobility rates are all about dollar amounts. In Winship’s terms: “absolute mobility ignores rankings and simply considers whether adults tend to have higher, size-adjusted incomes than their parents did at the same age, after taking into account increases in the cost of living.”

Most people are upwardly mobile in the absolute sense: 84 % of U.S. adults, according to the latest estimates (Economic Mobility Project 2012). People raised in families toward the bottom of the income distribution are the most likely to overtake their parents’ income status, as Fig. 13.1 shows. It is hard, then, from an absolute basis, to see that the “engines” of upward mobility have “stalled.”

Fig. 13.1
figure 1

Absolute mobility: share of Americans who exceed their parents’ family income (Copyright © July 2012 The Pew Charitable Trusts)

The two key drivers of absolute mobility are the rates of economic growth and the distribution of that growth. Policy should therefore attempt to maximize real income growth for as wide a swath of the population as possible. Relative mobility, which tracks movement up and down the income ladder, captures a different idea of fairness, closer to the ideal of meritocracy . Which kind of mobility to focus on—or rather, what balance to strike between the two—is a normative, rather than empirical, question. But relative mobility gets closer to the ideal of “equality of opportunity.” Even if everyone is richer than his or her parents, we would be a deeply unfair society if everyone was also stuck on exactly the same point on the income ladder. We want growth and more prosperity, but we also want fluidity and more fairness. A common way to present this intergenerational relative mobility is to examine the relationship between the income quintile (one-fifth of the income distribution) that people end up in as adults compared to the quintile they were born or raised in. Alternative approaches include a measure of the correlation between the income rank of parents and their child, used in particular by Raj Chetty , and rank direction mobility (RDM), which tracks an individual’s position on the whole income rank compared to their parents’ rank—developed in particular by Bhashkar Mazumder (Mazumder 2011, 2014).

Three more questions of definition should be briefly addressed. First, there is an important distinction to be made between intergenerational and intragenerational mobility , which is a measure of how far individuals will move up and down the income ladder during their own lifetime, especially during the prime working age years. While these kinds of mobility are related, my primary focus is on the former.

Second, the choice of outcome is important. Most studies of mobility focus on income. But there are, of course, many other possibilities, including wages, education, well-being, and occupational status. Many of these will provide important information about the capabilities and opportunities enjoyed by individuals, but I focus here on income. Income is important in itself and is strongly correlated with other goods. It is also a yardstick that is reasonably easy to measure and compare over time

Third, the presumption underlying this approach to measuring equal opportunity presumes that an outcome—in this case of income—is a good enough proxy for opportunity. They are not the same thing, of course, because there is a difference between an opportunity being available and somebody seizing it (Swift 2004). But for the moment, patterns of outcomes appear to suffice as an accurate reflection of patterns of opportunities.

Mobility: The Current Picture

The current picture in terms of relative intergenerational income mobility (RIIM) is not the main focus of this chapter (see Chap. 8). But a brief overview will provide a context for my broader argument on the need for strong indicators to guide data collection strategies, policy development, and evaluation.

The top line is: Rates of RIIM in the U.S. are low and flat and vary significantly by family structure, education, race, and geography. The U.S. suffers from a high degree of intergenerational income “stickiness,” especially at the top and bottom of the income distribution as Fig. 13.2, using the dataset constructed from the National Longitudinal Survey of Youth (NLSY) for the Social Genome Model , shows. There is more than a twofold difference in the odds of a child born in the top quintile remaining in the top income quintiles (the “comfortable middle class”), compared to one born in the bottom quintile (56 % versus 23 %).

Fig. 13.2
figure 2

Relative intergenerational income mobility

Has this picture worsened over time? It seems not. In a comprehensive series of recent studies, making innovative use of administrative records of income, Chetty et al. (2014, 10) investigate geographical variations in mobility (see below) and long-term trends. Their conclusion: “children entering the labor market today have the same chances of moving up in the income distribution relative to their parents as children born in the 1970s.”

There are, however, stark differences in mobility patterns at different levels of education. Children with a college degree are more likely to be upwardly mobile. A comparison of Figs. 13.3 and 13.4 shows that among children raised in the poorest quintile, those with a college degree are 20 times more likely than their high school dropout counterparts to make it to the top (20 % versus 1 %).

Fig. 13.3
figure 3

Social mobility matrix: college graduate

Fig. 13.4
figure 4

Social mobility matrix: less than high school education

Even top-income children receive a boost by receiving a college degree—37 % of them stay at the top, far more than their high school dropout and graduate peers, as seen in Fig. 13.3. So college degrees can be a double-edged sword in terms of relative mobility, helping improve the economic situation of poor children who go on to get a bachelor’s degree but also preserving the economic situation of the affluent.

At the other end of the spectrum, failing to receive a high school diploma damages upward mobility rates. Bottom-income children without a diploma have a 54 % probability of remaining on the bottom rung as adults, as seen in Fig. 13.4. Rates of downward mobility from the middle three quintiles are also very high for those without a diploma (42 % at the second quintile, 37 % at the third, and 48 % at the fourth).

There are striking differences in mobility by race, especially between Black Americans (Fig. 13.5) and White Americans (Fig. 13.6). One in two Black children born into the bottom quintile will remain there in adulthood, compared to just one in four Whites, and only 3 % of Black children rise to the top income quintile. Also, Black children are more likely to be downwardly mobile from the middle: of Black children born to parents in the middle-income quintile, 69 % move downward.

Fig. 13.5
figure 5

Social mobility matrix: Black Americans. Note: Sample size too small for those born in top quintile

Fig. 13.6
figure 6

Social mobility matrix: White Americans

There are also big differences in terms of the mobility patterns of children born in different kinds. As shown in Fig. 13.7, children with never-married mothers face a roughly 50–50 chance of remaining in the bottom quintile, while as Fig. 13.8 shows, children raised by continuously married parents have high upward mobility rates. The two biggest factors behind the “marriage effect ” appear to be higher income, even within income quintiles, and more engaged parenting (Reeves and Howard 2014).

Fig. 13.7
figure 7

Social mobility matrix: children of never-married mothers. Note: Sample size too small for those born in top two quintiles

Fig. 13.8
figure 8

Social mobility matrix: children of continuously married mothers

Last, there are variations in mobility patterns by geography. Chetty et al. (2014, 26) estimate, for example, that “the probability that a child from the lowest quintile of parental income rises to the top quintile is 10.8 % in Salt Lake City (Utah), compared with 4.4 % in Charlotte (North Carolina).” Five factors correlate strongly with intergenerational mobility by geography: racial and economic segregation, school quality, income inequality, social capital, and family structure, together “explaining” 76 % of the variation in upward mobility.

This brief discussion of the shape of U.S. intergenerational mobility is intended to motivate the remainder of this chapter, which focuses on the role of indicators to frame and focus strategies to promote greater opportunity. I begin with a brief history of social indicators and an even briefer theoretical overview.

A Very Brief History of Social Indicators

The U.S. has had an on-off relationship with social indicators. Interest began with the 19th century temperance movement, when campaigners began to collect data showing the deleterious social effects of alcohol. The alcohol industry responded with data on how much employment and revenue it generated: The result was a loosely fact-based debate about alcohol in the 1830s. But the measurements of trends began in earnest with the establishment of the Massachusetts Bureau of Statistics of Labor in 1869. But it was far from objective. First, it was run by pro-union officials, leading to biased reports; then it was taken over by pro-business staffing and swung the other way. But it was nonetheless an attempt to give data some official grounding and status.

The Community Indicators Movement was kick-started by the Pittsburgh Study funded by the Russell Sage Foundation in 1910, which inspired similar studies in towns around the U.S., with measures of health, income, jobs and so on. This was a time of great optimism about the potential of social indicators to effect change, as Cohen writes: “They relayed the findings of the technical experts to the public, who, enlightened by the facts, were expected to mobilize public opinion and press for appropriate reforms” (Cohen, quoted in Cobb and Rixford 1998, 7). The idea was that facts could change the world, through a process of enlightenment. In 1933, Recent Social Trends was published, under the Hoover administration. At 1,500 pages long, it was a compendium of every piece of social data the authors could get hold of. It also had no impact. The burst of interest in the 1930s did help to create the conditions for a significant widening in the collection of data on social trends. The U.S. Census Bureau , in particular, has captured increasingly rich data on demographic and social trends, especially through the Current Population Survey , which replaced the Monthly Report on the Labor Force in 1948.

Social indicators were out of political fashion until the late 1960s and early 1970s when a series of major studies were undertaken, including Indicators of Social Change (Sheldon and Moore 1968) and Towards a Social Report (HEW 1969). Across a range of policy areas, including defense, there was a renewed emphasis on the role of indicators in supporting cost-benefit analysis. This helps to explain why a good deal of funding was provided by NASA, which wanted to look at the impact of the space program on American society. (Many reviews of this work said that the links between the space program and the social indicators work were “somewhat tenuous,” which seems kind.)

The Office of Management and Budget (OMB) and Census Bureau picked up the baton, issuing a series of Social Indicators reports in the 1970s and into the 1980s (U.S. Office of Management and Budget 1974; U.S. Bureau of the Census 1977, 1981). A Center for the Coordination on Social Indicators was established in 1972. Between 1967 and 1973, Senator Walter Mondale submitted a number of bills to create a Council of Social Advisers (to mirror the Council of Economic Advisers) and institutionalize an annual social indicators report.

The movement was largely halted during the Nixon administration, as the role of social indicators lost any normative force. As Clifford Cobb and Craig Rixford write: “Some had envisioned these as the beginning of institutionalized social reporting, but their hopes were quickly dashed as political pressure within the Nixon Administration turned them into neutral chartbooks, replete with facts but void of interpretation … the social indicators movement in the United States was effectively over by the early 1980s” (Ibid., 11, my emphasis).

At the same time, many international organizations, such as the Organisation for Economic Co-operation and Development , the United Nations, and the European Union started to get very interested in social indicators, and in the 1990s, surveys and indices of well-being began to gain some traction, partly inspired by the environmental movement. In more recent years in the U.S., there has been a modest renaissance of community indicators, led by the Community Indicators Consortium , Healthy Cities movement , and so on. In one sense, this takes us back a century to where Russell Sage started in 1910, with metro-based approaches to community indicators rather than at a national or federal level. Efforts to improve the quality and increase the salience of indicators at the national level have been led by the National Academy of Science under its Key National Indicator Initiative, resulting in a series of publications, notably an important 2012 report, Using Science as Evidence in Public Policy (Prewitt et al. 2012). In 2010, President Obama signed legislation intended to create a Key National Indicators System , following advice from a commission of experts. A budget of $70 million was set aside. The commission was appointed in 2010 but never convened. The money—which was included in a provision of the Affordable Care Act—was never appropriated.

Theory: Conceptual Issues

The selection of indicators is not a straightforward matter. Indicators come in a wide variety of forms. Borrowing heavily from Cobb and Rixford (Ibid.), indicators can be distinguished and defined on a number of axes: inductive or deductive; “pseudo-objective” or “partisan”; descriptive or prescriptive; “local” or national; broad or narrow; and indirect or direct. The choice of indicator is inescapably connected to the purpose of the indicator—this is why they can only even be “pseudo” objective. Indicators of progress toward greater social mobility ought to be deductive (based on a clear theory about what promotes and predicts mobility); as objective as possible; prescriptive (intended to guide policy); narrow (provide as much focus as possible); and direct (getting as close as possible to the causal connection to mobility). But in terms of the choice between national and subnational indicators, the answer can legitimately be “both.” Many leading indicators may work in most localities. But especially in a nation as large and diverse as the U.S., there may be some localities in which a particular indicator is more powerful than elsewhere.

In their review of the role of indicators, Cobb and Rixford (Ibid.) offer a number of important lessons, of which I would highlight the following:

  1. (a)

    A clear conceptual basis is needed for indicators—otherwise you end up with a forest of numbers but no path;

  2. (b)

    A number is not necessarily a good indicator—just because a number is available does not mean it is “getting at” the trend or factor you are interested in;

  3. (c)

    There is no such thing as a “value-free” indicator—the simple selection of a particular indicator is a value judgment. It is better to be clear and upfront about the purpose of the indicator;

  4. (d)

    Comprehensiveness is the enemy of effectiveness—five strong indicators are better than 105 indicators in terms of focusing political energy; and

  5. (e)

    Indicators should attempt to reveal causes, not symptoms—especially in terms of promoting social mobility, indicators that get close to causal relationships are the most valuable.

Indicators and the U.K.’s Social Mobility Strategy

I served in the U.K. Coalition Government from 2010 to 2012, as Director of Strategy to the Deputy Prime Minister, who was leader of the junior party, the Liberal Democrats. At the time, Prime Minister David Cameron was in favor of what he had labeled the “big society”—a deliberate contrast to both the idea of the “big state” and Margaret Thatcher’s claim that there “is no such thing as society.” But Cameron and his team refused to define their term clearly or apply any metrics. So my questions to them were always along the following lines: “How will you know when society is bigger? How big is it now? What are your measures?” In the end they stopped inviting me to the meetings. But the truth is they had no way to answer the questions. The “big society” was just a rhetorical device.

Of course “opportunity” is at least as nebulous a term as “big society.” But when the U.K. government made a strong commitment to promoting social mobility as its overarching social policy goal, that commitment was buttressed by indicators and institutions. In April 2011, the U.K. government issued a social mobility strategy, declaring: “A fair society is an open society, one in which every individual is free to succeed. That is why improving social mobility is the principal goal of the Government’s social policy” (Cabinet Office, HM Government 2012, 5).

The definition of social mobility guiding the U.K. efforts is fairly tight, with a declared focus on intergenerational relative mobility by both income and occupation. Deciding on this definition was a vitally important step, laying the foundations for the selection of key “leading indicators” that are—based on the best available evidence—predictive of long-term trends in mobility. These indicators are shown in Table 13.1 and include income gaps in low birth weight , school readiness , educational attainment at ages 11, 16, and 19, postsecondary education , access to the professions, and early-career wage progression. An independent analysis of the indicators suggests that together they should capture more than half of the likely trends in intergenerational mobility (Gregg et al. 2014). The U.K. government also took steps to institutionalize the social mobility commitment with the creation of a Cabinet committee and a new, independent statutory Commission on Social Mobility and Child Poverty that reports annually to Parliament and the administration.

Table 13.1 Dashboard of opportunity “Leading Indicators” in United Kingdom

Indicators and the Colorado Opportunity Project

The State of Colorado has also created an evidence-based indicator framework for opportunity, based in part on the Social Genome Model (Winship and Owen 2013). The overall goal is to help as many Colorado residents as possible become “middle class by middle age” (i.e., a household income of 300 % of the federal poverty line by age 40). Following a yearlong project involving multiple state agencies and key stakeholders, a series of indicators at key life stages have been developed, as shown in Table 13.2. These indicators, making use of data available at a state level, will be used to help identify the most effective programs and initiatives. The project is still evolving, but speaking at a stakeholder summit on the project in March 2015, Gov. John Hickenlooper set the bar high: “The Colorado Opportunity Project is going to make history.”

Table 13.2 The Colorado opportunity framework

The U.K. and Colorado are just two examples of the operationalization of opportunity goals and indicators: They are offered here not as definitive or comprehensive but as illustrations of the potential for such an approach of which I have firsthand knowledge. Are there any lessons here for the U.S. more broadly?

Opportunity Indicators for the U.S.: Four Proposals

Indicators can provide a powerful infrastructure for policy making. This is an established fact in economics but has yet to become so for social policy. The current bipartisan interest in opportunity and mobility, however, could allow for operationalization of key indicators of progress, with potentially long-term benefits. In particular, four reforms should be considered.

Invest in Data for Opportunity

Data is gold, especially in the field of opportunity. Without data, policy decisions are arbitrary, claims are untested, and progress is virtually impossible. Indicators amount, in policy terms, to a weaponized data point. But the data they are based on has to be good.

This is an area where the U.S. can do much, much better, especially given the national commitment to opportunity. There are some hopeful signs of bipartisan activity here, too. House Speaker Paul Ryan (R-WI) and Sen. Patty Murray (D-WA) are together pushing for the creation of an independent Evidence-Based Policymaking Commission to “expand the use of data to evaluate the effectiveness of federal programs and tax expenditures.” In particular, the commission, if approved by Congress and the President, will:

  1. (a)

    study the federal government’s data inventory, data infrastructure, and statistical protocols in order to facilitate program evaluation and policy-relevant research;

  2. (b)

    make recommendations on how best to incorporate outcomes measurement, institutionalize randomized controlled trials, and rigorous impact analysis into program design; and

  3. (c)

    explore how to create a clearinghouse of program and survey data.

This may not sound very exciting to most people (it is intended not to, so as to avoid stoking unfounded fears about individual privacy). But it is thrilling for policy. The Obama administration has also led a renewed charge for evidence-based policy, as recounted by my colleague Ron Haskins (2015) in his book Show Me the Evidence: Obama’s Fight for Rigor and Results in Social Policy.

There is, however, a basic data issue too. Progress in terms of understanding trends in and prospects for intergenerational mobility is limited by what Kenneth (Prewitt (2015), 272), former director of t he Census Bureau, describes as “a serious gap in the nation’s statistics.” One promising proposal is the creation of an American Opportunity Survey by linking together various administrative datasets, including the Census, American Community Survey, Survey of Income and Program Participation, as well as data from the IRS and Social Security Administration. As Grusky et al. (2015) argue, this approach would “provide a high-quality infrastructure for monitoring mobility without the cost of mounting a new mobility survey.”

Right now, as they point out, the technical infrastructure for measuring mobility in the U.S. is in disrepair. This makes the formulation of policy difficult: It is rather like, as they put it, “formulating monetary and labor market policy without knowing whether unemployment is increasing or decreasing.”

Getting better data is not a huge undertaking. The key is to be clear what the data is for. As Isabel Sawhill put it in 1969: “The principal barrier to quantification, in the long run at least, is not a lack of meaningful data but a failure to define what is meaningful … to give operational content to our ideals.”

Set a Long-Term Goal for Intergenerational Mobility

Indicators are most valuable when an overall goal has been established: in other words, when it is clear what they are indicating toward. Goals can act as powerful policy commitment devices, helping to sustain a consistent focus on long-term objectives (Reeves 2015). In terms of promoting or restoring opportunity, a high-profile bipartisan commitment to a long-term goal could galvanize action on a number of important fronts. Such a goal would sit alongside existing goals for economic growth, monetary policy, employment, education, health, and so on. Because upward relative mobility is the primary concern for most policy makers, the goal should relate to progress on that front. For the purposes of illustration, I propose the following goal: increase the proportion of people born in the bottom income quintile who make it to the middle quintile or higher.

Right now, that number lies at around 40 % (or less, according to numbers generated by the Panel Study of Income Dynamics). In a perfectly mobile society, it would be 60 %. So, without further justification, 50 % seems like a reasonable goal. There are, of course, a host of other possibilities. A weakness of this goal is that it focuses attention on mobility from one specific part of the income distribution—the bottom—whereas equal opportunity ought to apply all the way up. I offer the goal principally in order to generate debate and illustrate the point. But this headline goal does have the advantage of being noncontroversial (at any rate it is hard to see why somebody would oppose it); simple (even if tracking it would be highly technical and controversial); and proximate to the goal of greater relative mobility. Operationalizing a goal like this would, needless to say, require a considerable number of technical specifications, including (but not restricted to): choice of dataset; household size equivalence; income definition; and inflation adjustments.

While the headline goal would apply to the whole population, it could also be used to track progress toward closing opportunity gaps and thereby help to focus policy attention. For example, the proportion of Black and White individuals could be compared in terms of the overall goal. Data from the NLSY suggests that the proportion of Black Americans making the journey is 22 %, compared to 58 % for Whites.

The key point is that the overall goal would act as a “north star,” guiding the direction of policy and other activities. We would at least be able to see, over the longer term, if we were making progress. A vitally important caveat, however: Setting such a goal should not precede the establishment of reliable data from which to measure it (see the first proposal above). Of course, there are other strong candidates for a “north star” summary goal, including an improvement in rank-rank mobility (the association between parents’ rank in earnings as compared to that of their children’s rank as adults), or in occupational mobility, or perhaps in relation to another nation, such as Canada. Each approach will have strengths and weaknesses; each will fail to capture some dimensions of opportunity. But these concerns apply to almost all summary statistics, including those for GDP growth, productivity, and poverty.

Develop a ‘Dashboard’ of Annual Opportunity Indicators

It takes a generation to track intergenerational mobility: an obvious point, but an important one. It will also be valuable to develop “leading indicators” that can be tracked over a much shorter time horizon but are empirically proven to predict progress against the long-term goal. This is the approach taken in the Social Genome Model, where progress toward the long-term goal—“middle class by middle age”—is measured and predicted by a series of success measures for each crucial life stage. It is also a central part of both the U.K. and Colorado examples described earlier. A dashboard should contain shorter-term data points and trends that—based on the best available evidence—will likely lead to more upward mobility in the long run. As in the U.K. and Colorado, these leading indicators would be best organized around key life stages. The indicators should also emphasize the relative picture, rather than the absolute one: in other words, not just overall rates for each indicator, but the gap between different groups. Increasing college graduation rates will not improve mobility rates if most of the increase is made up of students from affluent backgrounds. For relative mobility, then, the mantra is always: mind the gap. The particular gap ought to be determined in large part by the long-term goal. So if the agreed focus was indeed on movement from the bottom quintile, the most appropriate short-term indicators for the annual dashboard should compare, say, rates of low birth weight births, school readiness, test scores, or postsecondary education between those in the bottom income quintile and those in the top two quintiles.

The point here is not to argue for specific elements of a dashboard—that will require a good deal of investigation—but for its creation. It should also be stressed that many of the indicators become valuable over time, with repeated measurement and reporting, rather than as snapshots at a particular moment in time.

For the purposes of illustration, Table 13.3 combines the indicators used in the U.K., Colorado, the Social Genome Model, and my own paper on “five strong starts.” The overlaps are clear. The opportunity dashboard should have as many indicators as are useful but no more. In policy, parsimony is power. Continuous analysis of the predictive capability of the overall dashboard, and the contribution of each of the indicators, should be carried out. If after a period of time a specific indicator appears to be adding little value to the overall predictive power of the dashboard, it can be safely removed.

Table 13.3 Indicators used in the U.K., Colorado, the Social Genome Model, and Reeves’ paper on “Five Strong Starts”

Create a Federal Office of Opportunity

Better data, a clear long-term goal, and a near-term dashboard are all key elements of a new policy architecture for social mobility. But there is also a strong case for giving social mobility an institutional anchor, in the form of an Office of Opportunity. I’ve argued elsewhere for such an institution at a federal level, but there is just as strong a case for state or city versions (Reeves 2014). The office would be charged with producing regular reports on progress in terms of both the long-term goal and the shorter-term indicators; for overseeing and advising on data collection; and for generating independent advice on the mobility-enhancing potential of various policy proposals. The office could be established as an executive body, a congressional one, or a hybrid.

Scott Winship has made a more ambitious institutional proposal, an Opportunity, Evidence and Innovation Office (OEIO), based in the White House. His OEIO would bring together a number of existing agencies and fund and evaluate programs and initiatives that “seek to promote upward mobility” (Winship 2015, 36).

Note that none of these proposals are in themselves about policy: rather they are about the generation of reliable data and clear indicators and strong institutional grounding for a focus on intergenerational mobility. They amount to a policy architecture rather than a policy. Which policies or programs will work toward the goal—and by association the leading indicators—is a second-order question, and one that should be settled empirically. We should be evangelical about the ends but agnostic about the means.


The development of key indicators, collection of data, and establishment of technical bodies lack the glamour and immediacy of new policies or programs. But it is partly for that reason that they are more likely to gain crucial bipartisan political support. Even if both sides agree there is a problem, there is very little agreement in terms of specific solutions. Efforts to gain bipartisan support for specific policy programs are likely to be unsuccessful. But there is space for bipartisanship in the creation of an institutional framework designed to track the nation’s progress toward greater opportunity, keep the attention of policy-makers on this long-term task, drive the collection and dissemination of higher quality data, and dispassionately assess initiatives intended to improve rates of intergenerational mobility.

Right now, political discussions of opportunity are replete with anecdote and soaring speeches about American exceptionalism. But in the end, the restoration of opportunity is not a matter of opinion or rhetoric. It is a matter of fact. If we are serious about a project to restore opportunity, we need to know when we’ve arrived.