In my testimony before the Joint Economic Committee of the U.S. Congress this past summer, I presented some of the findings from my 2020 report, “The “Double Gap” and the Bottom Line: African American Women’s Wage Gap and Corporate Profits.” My research and my testimony that day highlights one of the most persistently pernicious wage inequalities in the United States today: Black women workers earn the least in wages on average, compared to their working counterparts among White men, White women, and Black men.
This “double gap” is a result of the intersecting effects of the racial wage gap—Black workers overall earn on average less than White workers—and the gender wage gap—women overall on average earn less than men. How big is this double gap? Well, the Institute for Women’s Policy Research finds that in 2019, women’s median annual earnings were about 81 percent of men’s. And according to the findings of the National Women’s Law Center, Black women earn about 61 cents for every dollar that non-Hispanic White men earn.
But these analyses do not take into account the complex factors that play a role in wage disparities for women, and women of color in particular. Among them are occupational crowding based on sex, gender socialization, employer bias, historical exclusionary practices on the part of unions, and the “motherhood penalty.” And these and other analyses do not fully account for the many and varied consequences of centuries of racial discrimination against Black Americans that remain very evident today.
The U.S. Bureau of Labor Statistics detailed this double gap in June 2020, releasing data on average weekly and annual salaries broken out by race and gender:
White Men: $1,115 weekly, $58,000 annually
Black Men: $828 weekly, $43,000 annually
White Women: $929 weekly, $48,300 annually
Black Women: $779 weekly, $40,500 annually
This is wage inequality by race and gender in a nutshell. I estimate this double gender-and-race wage gap costs Black women workers approximately $50 billion in involuntarily forfeited earnings, a large and recurring annual loss to the Black community.
In my testimony before Congress in June, I presented five ways in which this double gap could be closed:
Pass state and/or federal laws that prohibit employers from requesting previous salary histories from job applicants
Pass state and/or federal laws requiring pay transparency in the private, for-profit sector
Revise the so-called EEO-1 Form that employers are required to regularly submit to include compensation data so that the U.S. Equal Employment Opportunity Commission can better “support civil rights enforcement” in the wage arena
Make tuition free at community colleges and public colleges and universities throughout the United States to help more Black women afford to complete college
Raise the federal minimum wage
Funding more academic research on the double gap and analyzing an array of policy solutions in addition to the ones I presented to Congress earlier this year will be one of my top priorities as the president and CEO of the Washington Center for Equitable Growth. Sustainable economic growth and stability alongside growing racial equity are inseparable.
The U.S. House of Representatives late last week passed a key plank of President Joe Biden’s economic agenda, the Build Back Better Act. Coming on the heels of the president signing the Infrastructure Investment and Jobs Act on November 15, House passage of the other half of the president’s Build Back Better agenda signals momentum toward long overdue—and complementary—improvements in both the country’s physical and social infrastructure.
The Build Back Better Act enjoys privileged parliamentary status as a “budget reconciliation” bill, which means it can be passed by the U.S. Senate by a simple majority. It now moves across the Capitol, where it will likely be amended. But there is growing consensus in Congress on the core of the bill, which is designed to create jobs, cut costs, and grow the U.S. economy.
Misconceptions about inflation and debt have dominated the debate over the Build Back Better Act even though nearly all of the proposed spending is offset by provisions that would raise new government revenue. Most importantly, the $1.68 trillion of investments in the nation’s social and physical infrastructure would help expand the supply of labor and goods, and thus help ease cost pressures facing consumers, while contributing to sustained growth.
Indeed, the evidence detailed below demonstrates that, overall, the Build Back Better Act would deliver strong, stable, and broad-based economic growth up and down the income ladder and across the broader economy. (See Figure 1.)
Figure 1
The Build Back Better Act includes 10 broad ways that would achieve these results. Specifically, the House-passed legislation would:
Establish a national paid family and medical leave program
Expand access to child care and establish a universal pre-Kindergarten program
Make the Child Tax Credit permanently refundable and extend its expansion
Extend the Earned Income Tax Credit expansion
Invest in home- and community-based services
Reduce costs of prescription drugs
Enhance labor law enforcement capabilities
Increase antitrust enforcement funding
Make investments to combat climate change
Improve tax enforcement and impose taxes on large corporations and high-income earners
Let’s now turn to each of these 10 consequential provisions in the bill and the economic evidence that underpins the reasoning for their anticipated success.
Establishes a national paid family and medical leave program
The Build Back Better Act authorizes a new national paid family and medical leave program of up to 4 weeks for U.S. workers. This paid leave program would allow individuals to take time to care for and bond with a new baby, care for a family member with a serious health issue, and care for themselves if struck by a personal health issue.
Caregiving responsibilities are a significant driver of women’s exit from the U.S. labor force, as witnessed during the coronavirus pandemic, which disrupted more than half of family caregiving arrangements. Yet multiple studies find that access to paid leave increases the labor force participation rate of mothers, which we know is correlated with increased Gross Domestic Product. In one study, new mothers with paid leave are 18 percentage points more likely to be working a year after the birth of their child. (See Figure 2.)
Figure 2
Similarly, a synthesis of research finds that paid medical leave protects families during a health shock, ensuring financial stability even when things go wrong. Moreover, researchers have seen paid leave improve child well-being, strengthening the human capital of the next generation.
Expands access to child care and establishes a universal pre-Kindergarten program
The Build Back Better Act invests in children and supports caregivers by delivering a universal free pre-Kindergarten program for 3- and 4-year-olds and allocating an estimated $100 billion to expand accessible, affordable, and high-quality child care, including by providing higher wages for child care workers.
Detailed in Equitable Growth’s child care report and factsheet, research shows that achieving affordable, accessible, high-quality care and early education has benefits for both parents and their children. Studies find that a 10 percent reduction in child care costs increases maternal employment between 0.5 percent to 2.5 percent, and for every additional 100 child care slots that open up, a 2007 study in Maryland finds that women’s labor force participation rate goes up by 0.3 percentage points. Similarly, evidence from a universal preschool program in Washington, DC shows that mothers of these children increased their labor force participation rate by 10 percentage points.
For children, high-quality early care and education can lead to long-term improvements in their human capital—better education, economic, developmental, and social outcomes—all of which is also good for the broader U.S. economy. In fact, research finds that President Biden’s proposed universal preschool program—similar to the one included in the Build Back Better Act—pays for itself within 10 years and generates $4.93 in economic return in 35 years for every dollar spent.
Makes the Child Tax Credit permanently refundable and extends its expansion
The legislation makes the Child Tax Credit permanently refundable, important for lower-income families, and extends the expansion of the Child Tax Credit of up to $3,600 enacted earlier in 2021 as part of the American Rescue Plan in advanced monthly payments.
According to two studies, the expanded Child Tax Credit is projected to reduce childhood poverty in the United States between 40 percent and 45 percent, with poverty rates falling particularly rapidly for Black and Hispanic children. Additional research shows that income support programs, such as the Child Tax Credit program, also allow families to make investments in their children’s human capital development, improving children’s chances of upward intergenerational mobility and increasing both their future earnings and corresponding tax revenue.
Extends the Earned Income Tax Credit expansion
The Build Back Better Act extends the expansion of the Earned Income Tax Credit in the American Rescue Plan for low-wage workers without qualifying children through 2022.
The Build Back Better Act invests $150 billion in home- and community-based services that will allow more seniors and individuals with disabilities to receive services in their homes.
The use of home- and community-based services, compared to institutional care, is good for the economy as well. One study finds that transitioning from nursing home facilities leads to an 18 percent to 24 percent decline in healthcare spending in a patient’s first year in home- and community-based care. Economists have also pointed out that investing in home- and community-based services will create jobs and boost growth.
The Build Back Better Act reduces out-of-pocket prescription costs by allowing Medicare to negotiate the price of prescription drugs that lack competition.
Limited competition keeps prescription drug costs, and resulting national healthcare spending, high. One study finds that increased competition via generic drug entrants in the marketplace helped drop the price of drugs on average by 51 percent in the first year and 57 percent in the second year. Another study found that the U.S. healthcare system saved more than $1 trillion with the use of generic competitor drugs between 1999 and 2010.
Enhances labor law enforcement capabilities
The Build Back Better Act makes significant investments in the enforcement of labor laws. It provides more than $2 billion to rebuild agencies focused on worker protection, including the National Labor Relations Board, which protects the right of workers to organize into unions, and the Wage and Hour Division of the U.S. Department of Labor, which enforces labor laws, such as the federal minimum wage. The proposed law passed by the House also increases financial penalties for violations of wage theft and authorizes new financial penalties for unfair labor practices.
These investments would better protect the rights of workers and maintain a level playing field for compliant employers. Recent research shows that firms would be incentivized to abide by labor laws if there were a high chance of violations being detected. That’s why increasing the capacity of agencies to detect and respond to unfair labor practices, with the possibility of substantial monetary penalties, would likely change the calculation for employers and serve as an effective deterrent.
Increases antitrust enforcement funding
The Build Back Better Act provides $500 million each to the Federal Trade Commission and the U.S. Department of Justice’s Antitrust Division—the two federal antitrust enforcers—to support their work cracking down on unfair competition. This additional funding is a much-needed increase, as resource constraints over the past several years have limited these agencies’ ability to take enforcement actions against anticompetitive behavior. (See Figure 3.)
Figure 3
In an Equitable Growth co-authored report, we recommend an increase of $600 million in appropriations for the Federal Trade Commission and the U.S. Department of Justice’s Antitrust Division to implement a deterrence strategy to antitrust activities, including increasing investigations into problematic anticompetitive behavior and taking 60 to 100 enforcement actions a year rather than the current rate of roughly 40 enforcement actions per year.
Makes investments to combat climate change
The Build Back Better Act invests $555 billion to mitigate climate change, including promoting the shift to clean energy technologies and addressing the disproportionate harms that climate change brings to low-income communities and communities of color.
Ensuring these investments are equitable in nature is particularly critical. Low- and middle-income communities and communities of color face a “double threat” of significant costs from climate-related hazards and fewer financial resources to pay for the costs associated with climate change mitigation measures. One example of how the proposed legislation would achieve this end is that it allows consumers to take advantage of the electric vehicle tax credit at the time of sale, so that low- and middle-income buyers do not have to be financially strained by the full cost of the vehicle until they can claim the tax credit at the end of the year.
Improves tax enforcement and imposes taxes on large corporations and high-income earners
The Build Back Better Act includes provisions that move the United States toward a more equitable tax system and aims to cover the costs of the bill, though the federal government has the capacity to deficit-finance these growth-enhancing investments. It invests $80 billion in the Internal Revenue Service so that it can combat tax evasion by high-income earners, adds a surtax to Americans making incomes higher than $10 million, establishes a new “minimum tax” on profitable U.S. corporations, and imposes a 15 percent country-by-country minimum tax on foreign profits consistent with the recent international agreement on a global minimum tax on multinational corporations.
Multinational corporations have also found ways to not pay taxes. Historically, the U.S. tax system has encouraged profit shifting by multinational corporations to tax havens in other countries, costing the United States more than $100 billion in revenues per year. With the Build Back Better Act, the global minimum tax of 15 percent on large multinational corporations’ profits earned overseas will begin to crack down on this tax evasion. (See Figure 4.)
Figure 4
In addition, the 15 percent global minimum tax on profitable corporations in general would help to counteract corporations’ relatively low effective tax rate—as low as 10 percent for some industries—especially as we know that the tax cuts in 2017 were economically wasteful.
Conclusion
Taken together, these policies make significant investments in U.S. workers and their families with the potential to have positive short- and long-term economic outcomes on individuals, households, and the broader U.S. economy. To be sure, the House missed out on opportunities to address certain structural issues in our economy, such as reforming our patchwork Unemployment Insurance system, taxing the unrealized capital gains of billionaires, and improving IRS tax reporting. And, certainly, there could be retrogressive changes to the proposed law in the Senate.
Nonetheless, these 10 policy provisions detailed above go a very long way toward addressing longstanding economic fragilities rooted in economic inequality and exacerbated across racial and gender lines. Should these broad policies become enacted into law, the evidence demonstrates they would deliver on the promise of structural economic change that would help U.S. workers and their families and the broader economy achieve strong, stable, and broad-based growth.
The three components of the U.S. social insurance system serve retirees and those recipients facing challenging health conditions. But all three programs have grown in recent decades, sparking various calls for reform. The basic Social Security retirement system is now more than 80 years old, while the Social Security Disability Income system turned 65 in August and the Supplemental Security Income program, the youngster of U.S. retirement benefits system, celebrated its 47th birthday back in January of this year.
Basic Social Security remains a wildly popular retirement benefit and is thus considered an untouchable “third rail” in U.S. politics. But calls to reform Social Security’s two younger siblings often get caught in political crossfires because of concerns about individuals who may not have severe health conditions receiving SSDI and SSI benefits.
These two programs together provide access to health insurance and $200 billion annually in cash benefits to nearly 13 million Americans. At the same time, criticism of who qualifies for Social Security Disability Income and Supplemental Security Income has become more common in the United States. The argument goes that only those who suffer from the most severe health problems deserve access to these benefits. And the expansion of these two programs in recent decades has been attributed, at leastin part, to nonhealth factors, such as stagnating wages, resulting in concern that providing benefits to individuals without severe health conditions is diluting the value of these programs overall.
The findings in a recent working paper by me and my co-author Lee M. Lockwood at the University of Virginia address the premise of this argument. Our data-driven analysis of Social Security Disability Insurance and Supplemental Security Income finds that the value of these two programs, including value from insuring nonhealth risks, exceeds that of cost-equivalent of tax cuts by 64 percent, creating a surplus worth $8,700 of government revenue per recipient per year.
Moreover, we find that the high value of the SSDI and SSI programs is, in part, because of—not despite—mismatches with respect to health status. We estimate that benefits to recipients with less-severe conditions create a value over cost-equivalent tax cuts of $7,700 per recipient per year, or about three-fourths that of benefits to those recipients with more-severe conditions of $9,900.
These are important findings amid the debate about SSDI and SSI benefits for less-severe recipients. Benefits to less-severe recipients do not decrease the value of the two programs. Instead, they increase it considerably, accounting for about half of the total value. We discuss the policy implications of our findings in this column after briefly presenting the outlines of our research.
The outlines of our research
We conducted our research using a combination of survey and administrative data to establish new facts about the targeting of disability benefits on the basis of nonhealth factors. Our working paper details the complete methodology, but, broadly, our research examined disability recipients and nonrecipients with different health statuses. We find that less-severe disability recipients are, on average, much worse off economically than less-severe nonrecipients and, by many nonhealth measures, are even worse off than more-severe recipients.
We find in administrative data, for example, that prior to receiving disability benefits, less-severe recipients are 40 percent more likely to have experienced a mass layoff, 19 percent more likely to have experienced a foreclosure, and 23 percent more likely to have experienced an eviction than more-severe recipients. And from survey data, we find that less-severe recipients have similarly low consumption levels to more-severe recipients just prior to receiving disability benefits. Conversely, more-severe nonrecipients are, on average, better off than either more-severe or less-severe recipients.
The strong associations between receiving disability insurance payments and nonhealth shocks conditional on recipients’ overall health increases the value of Social Security Disability Insurance and Supplemental Security Income by both increasing insurance of nonhealth risk and decreasing distortion costs from discouraging work. Providing disability benefits, for example, to a worker who is laid off and has limited earnings prospects but not a severe health condition can increase the value of the disability program through both high insurance value and low distortion of earnings.
Our findings help reconcile two strands of the debate over these disability programs. The first strand raises concerns about disability programs, including that some disability recipients may not have severe health conditions, that labor market shocks increase disability enrollment, and that disability programs overall reduce the labor supply of marginal recipients. The second strand quantifies the so called welfare effects of various reforms to the SSDI and SSI programs, taking into account the gains, less any losses of these programs. This strand typically finds that expanding disability programs would increase the welfare of recipients.
Our results help reconcile the tension between these two strands by demonstrating that the nonhealth drivers of enrollment in the two programs do not just drive up the costs of the programs, but also drive up the benefits in terms of insuring nonhealth risk.
Policy implications
The public debate over Social Security Disability Insurance and Supplemental Security Income centers on concerns about individuals with less-severe health conditions receiving benefits. But we find that these programs are highly valuable and that including benefits to individuals with less-severe health conditions actually increases their value. Together, our findings show that these programs target well on the key determinants of the value of receiving disability benefits.
Our results suggest that proposed reforms to decrease benefit levels, allowance rates, or awards to individuals with less-severe health conditions would harm recipients without improving the two programs. A related but distinct question that our paper is not equipped to answer is whether these two programs should be expanded.
Another relevant question is whether disability programs are the best way to insure nonhealth risk. For individuals with high earnings capacity, insuring them through other means—perhaps by expanding Unemployment Insurance or creating a new wage insurance program—could be valuable. But the available evidence suggests that the vast majority of disability recipients, including those with less-severe health conditions, would earn little even if not receiving disability benefits.
Moreover, expansions of alternative programs would have efficiency costs of their own and may lack the advantageous targeting properties that we find for Social Security Disability Insurance and Supplemental Security Income.
Of course, the importance of nonhealth risk for the value of U.S. disability programs may be just one example of a broader phenomenon. No program exists in a vacuum—its effects reflect the diversity of risks in the U.S. economy, how well-insured those risks are by other programs and institutions, and how its tags and screens select on those risks.
Still, we find that Social Security Disability Insurance and Supplemental Security Income insure risks well beyond health, and that this “incidental” role is central to their overall value. Other programs might be similar in having their costs and benefits driven in large part by factors outside of their core aims. The extent to which they do is an empirical question, one that future research could investigate using similar methods.
—Manasi Deshpande is an assistant professor of economics at the University of Chicago, faculty research fellow at the National Bureau of Economic Research, visiting economist at the Social Security Administration, and a 2016 Equitable Growth grantee. Her co-author on the working paper featured in this column, Lee. M. Lockwood, is an associate professor of economics at the University of Virginia.
1. The extremely sharp Sam Abbott has what I think is an excellent column on how much the U.S. economy is held back by our collective failure to construct a reasonable and effective system for childcare. Read “The child care economy: How investments in early care and education can fuel U.S. economic growth immediately and over the long term,” in which he writes: “Insufficient child care options can prevent parents who wish to work from doing so, with mothers often bearing the brunt of this challenge. … High-quality early care and education provides critical socialization and learning opportunities when the brain is developing rapidly and is particularly responsive to the outside environment. … Adequate funding is necessary for human capital development. Fully funding the subsidy programs and devoting resources for state-level agencies to assist providers in qualifying for subsidies are two ways in which greater public investment could increase child care availability and quality. Supporting child care workers is crucial for promoting quality care and human capital development. Using public funds to support higher compensation would help stabilize the child care workforce, ensuring that these workers can afford to stay in their jobs. Investing in the nation’s children is one of the safest bets policymakers can make. Research on early care and education programs finds that $1 in spending generates $8.60 in economic activity.”
2. Equitable Growth’s Jonathan Fisher did very well before the House Select Committee on Economic Disparity and Fairness in Growth in detailing Equitable Growth’s push for improving the U.S. statistical system to be good at things other than simple national scale means. Read his “Testimony before the House Select Committee on Economic Disparity and Fairness in Growth,” in which he writes: “In which direction should the federal statistical system move?…. We describe four important steps. … GDP 2.0: Gross Domestic Product needs an upgrade. … Distributional National Accounts. … The Fiscal Analysis by Income and Race, or FAIR, Scoring Act, would direct the Congressional Budget Office to prepare distributional analyses by race and income for all legislation with substantial budgetary effects. … Congress should support the Census Bureau and Bureau of Labor Statistics working together to develop one survey that can: Measure income, consumption, and wealth. Having all three measures in one survey would decrease academic disagreement on the proper survey or definitions to use.”
3. A very nice paper with what I regard as very strong evidence that employers typically do not respond, at least in the short run, to employees’ outside offers. Wage bargaining between employers and employees in real time simply does not appear to be a thing. Instead, employers post wages and then overwhelmingly hold the line. Read the new Equitable Growth working paper by Marta Lachowska, Alexandre Mas, Raffaele Saggio, and Stephen Woodbury, “Wage Posting or Wage Bargaining? A Test Using Dual Jobholders,” in which they write: “This paper examines the behavior of dual jobholders to test a simple model of wage bargaining and wage posting. We estimate the sensitivity of wages and separation rates to wage shocks in a worker’s secondary job to assess the degree of bargaining versus wage posting in the labor market. We interpret the evidence within a model where workers facing hours constraints in their primary job may take a second, flexible-hours job for additional income. When a secondary job offers a sufficiently high wage, a worker either bargains with the primary employer for a wage increase or separates. The model provides a number of predictions that we test using matched employer-employee administrative data from Washington State. In the aggregate, wage bargaining appears to be a limited determinant of wage setting. The estimated wage response to improved outside options, which we interpret as bargaining, is precisely estimated, but qualitatively small. Wage posting appears to be more important than bargaining for wage determination overall, and especially in lower parts of the wage distribution. Observed wage bargaining takes place mainly among workers in the highest wage quartile. For this group, improved outside options translate to higher wages, but not higher separation rates. In contrast, for workers in the lowest wage quartile, wage increases in the secondary job lead to higher separation rates but no significant wage increase in the primary job, consistent with wage posting. We also find evidence in support of the hours-constraint model for dual jobholding. In particular, work hours in the primary job do not respond to wages in the secondary job, but hours and separations in the secondary job are sensitive to wages in the primary job due to income effects.”
Worthy reads not from Equitable Growth:
1. I continue to fail to understand claims that “team transitory” with respect to this year’s news about inflation needs to reevaluate its position. When I listen to Jason Furman, for example, I hear rhetorical alarm coupled with a suggestion for only a very small revision in the future of monetary policy—that the Federal Reserve’s first interest rate hike should probably be made next February rather than next October, leaving short-term interest rates about half a percentage point per year higher at the end of 2022 than the Federal Reserve is currently penciling in. That is not a huge difference. The inflation we have seen this year—that is just the rubber left on the road from the skid marks that always happens when you try to rapidly re-join highway traffic at speed. Nothing that has happened so far this year has surprised me, definitely not enough to make me revise my belief that the inflation we see is highly likely be transitory. And I am greatly encouraged by the fact that Wendy Edelberg of the Hamilton Project agrees. Read her “What does current inflation tell us about the future?,” in which she writes: “Should [we] expect continued extraordinary inflation for core goods—everything from automobiles to exercise mats—in the coming years? Three factors suggest no. First, the surge in spending on goods has put upward pressure on prices as suppliers have been unable to keep up with demand. Suppliers have strong incentives to iron out issues with the supply chain to get more product onto shelves; in addition, the problems with the supply chain that owe more directly to the pandemic will ebb as the pandemic is brought under control globally. Second, that surge in goods spending is no doubt temporary because households—as the pandemic recedes—will rebalance consumer spending toward services, which has been unusually depressed. … Third, the fiscal support to households that has helped to finance the surge in goods spending has largely waned. In contrast to spending on consumer goods, spending on services remains below its pre-pandemic peak. This pattern is a significant departure from previous business cycles where services were relatively unaffected.”
2. I concur 100 percent with Scott Sumner’s judgment that Paul Krugman’s 1998 Brookings paper both revolutionized thinking about monetary policy in the 21st-century and yet is also very often only partially understood and misunderstood. It is, I think, the most important macro paper of the last 40 years. Read “Scott Sumner on the Princeton School of Macroeconomics and Overcoming Inflationary Fears,” in which he writes: “In the early 2000s … five people … at Princeton … developed what we now view as the standard view of monetary policy in the 21st century, that is policy when interest rates are really low. … The key paper here is Paul Krugman’s 1998 paper where he looked at the liquidity trap in Japan. … I believe the paper is underestimated and misunderstood. … I view it as one of the most important macro papers of the last 40 years. … The problem in Krugman’s view was not just that money and bonds are equivalent at zero interest rates. The real problem is sort of bearish expectations of future monetary policy. … One interpretation is that monetary policy still works at zero interest rates. The central bank just needs to commit to future inflation. … The other interpretation … is that central bank promises to inflate are not likely to be credible … so you need Keynesian fiscal policy.”
3. Martin Ravallion says something that very much needs to be said: “degrowth” is deeply flawed both in its analysis of past economic development patterns and in its hopes for the future. Costa Rica is not a country that has remained poor and nevertheless upgraded its human development patterns. It is a country that has become much richer and has invested well in human development. Read his “The Degrowth Fallacy,” in which he writes: “In an example of the Degrowth Fallacy, in a radio interview … Jason Hickel (introduced as an economist and anthropologist at the London School of Economics) points to Costa Rica to support his claim that rich countries can maintain their social outcomes at lower mean income. Yes, Costa Rica has had good social and environmental policies over many decades, and other countries can learn from that experience. However, the country did this in combination with economic growth. Indeed, mean income has tripled in real terms since 1960, and the average growth rate has been above average for Latin America [according to the World Bank]. By combining social policies with policies that directly supported economic growth, Costa Rica was able to attain over time good social outcomes for a country at its level of mean income. Costa Rica is definitely not an example of how good social outcomes are possible without economic growth.”
Many neighborhoods and communities across the United States are shaped by pervasive economic inequality spanning multiple generations. This intergenerational economic inequality today limits the prosperity of many people who want to attain the American Dream of upward income mobility, and it stunts U.S. economic growth.
Research demonstrates a strong link between one’s neighborhood of residence and upward income mobility. Findings from economics research highlight that parental income and education, the quality of education, and employment opportunities are factors correlated with neighborhood of residence and economic mobility.
Intergenerational economic mobility is a critical economic issue. A person’s economic fate is heavily influenced by their family background and where they grew up. Yet the connections between individual outcomes and family backgrounds still aren’t resolved due to the complexity of these dynamics and how academics and policymakers weigh them, such as the environmental factors that a family faces and the impact of family income on investments in children.
Nonetheless, policymakers at the federal, state, and local levels alike understand that persistent intergenerational economic inequality not only hinders those in our society who begin lower on the income ladder but also inhibits the overall growth potential of the U.S. economy by blocking paths to success for those most in need of support.
In a recent working paper, “Family Background, Neighborhoods, and Intergenerational Mobility,” economists Magne Mogstad at the University of Chicago (also an EG grantee) and Gaute Torsvik at the University of Bergen attempt to understand these issues by reviewing the broad economic literature on location, family, and intergenerational mobility. Much of the existing literature uses the Becker-Tomes model, which looks at how family resources can be invested into a child’s education and how this, in turn, will affect a child’s future economic outcomes, mobility, and productivity.
This model—named after Nobel laureate Gary Becker at the University of Chicago and his co-author Nigel Tomes at the University of Western Ontario—uses parents’ income and spending, which encompasses a single parent’s consumption of everyday goods and necessities, such as rent, food, or other miscellaneous bills, to discern patterns of intergenerational mobility.
To further dig into how neighborhoods and families influence future outcomes, the authors review the Moving to Opportunity Experiment, a 1994 policy initiative that was implemented by the U.S. Department of Housing and Urban Development to evaluate how randomized home mobility for low-income households with children who moved from their neighborhoods to “private-market housing in much less distressed communities” affected future economic mobility. This experiment was an opportunity to delineate family effects and neighborhood in economic mobility by comparing three groups: those who received an experimental rental voucher to move within low-income areas within a city; those who received Section 8 vouchers that allowed them to move anywhere within the designated city; and those who received no additional assistance.
The results found that those who were able to use Section 8 vouchers to move to any location within a city had the most positive results, with declining poverty rates. This demonstrates that neighborhood has a significant effect on economic outcomes beyond what can be explained by family characteristics.
Mogstad and Torsvik’s survey of the research helps connect the links between neighborhood and intergenerational mobility. Yet the research generally looks at neighborhoods in a very broad sense, using commuting zones and counties. This means it is exceedingly difficult to pinpoint what neighborhood has the best or most opportunity for mobility.
Still, the sum of the research clearly indicates that location matters. The multiple barriers that multitudes of families face every single day in low-income neighborhoods affect intergenerational mobility. Furthermore, these barriers interact with family characteristics, such as the income, wealth, and the human capital allotments of parents.
The working paper by Mogstad and Torsvik supports the findings of other scholars in the field, among them research by Equitable Growth grantee and Opportunity Insights Research Scientist Matthew Staiger, which finds that children often have access to early employment at their parent’s employer, which puts them on a higher earnings trajectory, compared to children who do not work at their parents’ well-paying employer.
Children can’t help it if the schooling they rely upon isn’t as academically challenging as the schooling in neighborhoods with more funding. And this blame doesn’t fall on the parents who are providing their absolute best with what they were given from their own parents.
Research along these lines is equally compelling, including Harvard University economists Raj Chetty and Nathaniel Hendren in their study titled “The Impacts of Neighborhoods on Intergenerational Mobility II: County-Level Estimates.” Chetty and Hendren find that children’s future earnings, the rate at which they attend college, their fertility, and even their marriage patterns are determined by the neighborhood they grow up in. This was all found looking at looking more than 7 million families and their commuting zones and counties.
Potential policy solutions include making permanent the recent changes to the Child Tax Credit. The policy change, which makes the credit fully refundable and available on a monthly basis, swiftly helped alleviate some of the financial crunch faced by low-income families with children amid the coronavirus pandemic and ensuing recession.
Finding the financial resources to help their children advance in life is harder for parents who are low on the income ladder. The Child Tax Credit could be an economic vehicle that would help alleviate this problem, allowing these parents to capitalize on opportunities for the advancement of their children and improving intergenerational mobility.
But more extensive policy solutions are needed, given the complexity of the problem of intergenerational mobility highlighted in Morgstad and Torsvik’s survey of the research and the broader body of evidence on the importance of neighborhoods in propagating economic inequality over generations. To address place-based economic inequality, policymakers can build on the locally targeted Pathways to Work program, housed at the U.S. Department of Health and Human Services.
More broadly, federal policy with a “place-concious” lens has the potential to improve outcomes in specific underresourced places, argues University of Michigan economist and Equitable Growth grantee Robert Manduca. He makes a compelling argument for advancing universal policies that will reach the places most in need.
In the case of underresourced neighborhoods characterized by high levels of unemployment and poverty, structural reforms to the Unemployment Insurance system, strengthening the Supplemental Nutrition Assistance Program, and ensuring that the recent reforms to the Child Tax Credit are made permanent would efficiently bring economic resources into disadvantaged neighborhoods.
The guidance is clear for policymakers: They must find a way to ameliorate the disparity in economic mobility due to the effects of neighborhoods on intergenerational mobility.
—Vincient Whatley is an undergraduate Rogers Fellow at Kentucky Wesleyan College and was a participant in the AEA Summer Training Program at Howard University who was placed with the Washington Center for Equitable Growth as an experiential learning component of that program.
Statement for the Roundtable Meeting of the Committee on “Measuring Economic Disparity” House Select Committee on Economic Disparity and Fairness in Growth
November 17, 2021
I am Jonathan Fisher, the research advisor at the Washington Center for Equitable Growth, a nonprofit research and grantmaking organization dedicated to advancing evidence-backed ideas and policies that promote strong, stable, and broad-based economic growth. I have written this statement in partnership with Austin Clemens, Equitable Growth’s director of economic measurement policy.
This Select Committee’s charge is to “develop solutions to the key economic issue of our time: the yawning prosperity gap between wealthy Americans and everyone else.” I am here to discuss today the state of inequality measurement in the United States and steps the committee can support to help better define the problem and better understand the effects of potential policy interventions. What is the current state of inequality? How will you know if the solutions your committee develops and the government implements are successful at reducing disparity and increasing fairness? And how will future Congresses know the state of the U.S. economy in real time to craft well-calibrated policy?
It is important to measure and understand inequality because economic inequality harms the economy. Inequality today is high, close to its peak, using before-tax and before-transfer income, and increasing. When taking into account taxes and transfers, inequality is just below its 2006 peak, and only lower because of the health insurance provided through the Affordable Care Act.1This inequality constricts growth by:
Obstructing the supply of people and ideas into our economy and limiting opportunity for those not already at the top, which slows productivity growth over time
Subverting the institutions that manage the market, making our political system ineffective and our labor markets dysfunctional
Distorting demand through its effects on consumption and investment, which both drags down and destabilizes short- and long-term growth in economic output
Today, I describe four key shortcomings of the federal statistical system, or the federal agencies that provide essential statistical information when it comes to measuring disparities.2
The products of the Bureau of Economic Analysis, such as Gross Domestic Product, lack a distributional component, leaving policymakers, businesses, the media, and the American people in the dark about to whom economic growth is accruing.
A significant lag in the creation of distributional measures hinders Congress’ ability to know, in real time, the state of the economy and disparities in growth.
The system fails to measure multidimensional inequality and poverty, and it lacks a standardized definition of income, consumption, and wealth for the purposes of measuring poverty and inequality.
Federal statistics conceal substantial variation by race, gender, and ethnicity.
I will then discuss four concrete steps Congress should take to address these shortcomings and help improve measurement within the federal statistical system. It is important that these efforts reside within the federal statistical system. The federal statistical agencies act as the neutral arbiter and can adopt the best practices developed across researchers and other countries’ statistical agencies.
Critiques of the federal statistical system for inequality measurement
The federal statistical system today is insufficient for Congress, the business community, and for the American people. The economic stress during the coronavirus recession and amid the continuing pandemic makes clear the need for more consistent and more timely estimates regarding the health of the economy and growing inequalities.
Gross Domestic Product lacks a distributional component
One of the most-commonly referenced statistics from the federal statistical system is Gross Domestic Product, or GDP, and Personal Income. GDP measures the total final goods and services produced in the United States. Personal Income measures the income that people receive from wages and salaries, government benefits, dividends and interest, business ownership, and other sources. These measures are published quarterly.
This committee’s name references fairness in growth, but the federal statistical system generally only reports growth statistics in the aggregate. GDP growth was once a reasonable proxy for the economic fortunes of families because growth in the U.S. economy was broad-based. But growth over the past four decades has increasingly accrued to the top of the income distribution, creating a disconnect between headline GDP growth and the fortunes of families in the United States. (See Figure 1.)
Figure 1
It is helpful to understand first why current measures of economic growth can be so misleading. Let’s say 20 people live in a country, and everyone has $50,000 in income. National income is $1 million. Now, let’s say one person inherits $1 million from a relative living in a different country. National income is now $2 million. The Bureau of Economic Analysis would report that national income doubled. But only one person in the entire economy benefited. The headline growth number promises a massive increase in economic well-being that is illusory for 95 percent of the country.
Political scientists have found this effect can be significant. The tone of news coverage of the U.S. economy is closely correlated to GDP growth. Newspapers tend to see GDP growth as especially newsworthy, and this is reflected in reporting where GDP growth is often used as a proxy for the health of the entire economy. This has ill effects. Because GDP growth is no longer closely tied to income growth for most people, the overall tone of economic news is unrelated to income movement for all Americans outside the top 1 percent. Put another way, when GDP is used as a proxy, the economic news of today is largely just news about how the very richest Americans are doing.3 And because the public relies on the news to form their own assessment of the economy, headline GDP distorts economic perceptions across society.
A GDP measure that reflects the entire economy by reporting disaggregated statistics can better inform the media, policymakers, and businesses. Businesses can use disaggregated GDP to understand the income dynamics of the consumers they serve. A business that primarily sells to lower-income households, for example, will want to know the income dynamics of that population, which an improved measure of GDP and Personal Income will allow.
Lack of standardized definitions of income and long lag in measuring poverty
While the current GDP measure does not inform about inequality and poverty, the U.S. Census Bureau publishes poverty and inequality statistics annually in September, reporting these statistics for the previous calendar year. The Census Bureau uses an imperfect definition of family income—what it calls money income—to measure poverty and inequality. Money income includes earnings, unemployment compensation, Social Security, interest income, child support, and other money income sources.
But money income excludes taxes and tax credits, such as the Earned Income Tax Credit and the home mortgage interest deduction, and it misses economic stimulus payments such as the recent Economic Impact Payments. It also excludes noncash transfers, such as the Supplemental Nutrition Assistance Program, or SNAP, and housing assistance. Finally, money income also excludes capital gains. Furthermore, the Current Population Survey, or CPS, which the Census Bureau uses to estimate poverty and inequality, overestimates the extent of poverty through underreporting of income,4 and underestimates the concentration of income at the top of the distribution because of survey nonresponse.5
Because of the well-documented weaknesses of the Official Poverty Measure, the Census Bureau publishes a Supplemental Poverty Measure that improves upon the Official Poverty Measure in several ways. First, the Supplemental Poverty Measure includes noncash benefits, taxes, and tax credits. The poverty threshold is based on current spending patterns, while the Official Poverty Measure uses spending patterns from 1963, adjusted for inflation. The Supplemental Poverty Measure also adjusts the poverty threshold geographically, recognizing that the cost of living is different across the country. The Supplemental Poverty Measure can be used to indicate how much certain programs or policies affect poverty. The 2020 Supplemental Poverty Measure, for example, indicates that 11.7 million people were kept out of poverty because of the two Economic Impact Payments made in 2020.6
Poverty statistics also take significant time to be released. The poverty rate for 2020 was not released until September 2021. One recent innovation measures poverty in close to real time. The researchers use the monthly Current Population Survey poverty rates, with just a short lag in reporting.7 The real-time poverty measure shows how poverty fell in the months around the Economic Impact Payments in April 2020 and January 2021.8 Conversely, the official poverty measure and supplemental poverty measure summarize poverty in 2020 with a single number each. (See Figure 2.)
Figure 2
Lack of standard definitions of inequality
Various government agencies use different inequality measures. The Census Bureau uses its money income definition to publish inequality statistics, such as the Gini coefficient, a single number that measures income dispersion, and share of income received by different percentiles of the population. While not a statistical agency, the Congressional Budget Office, or CBO, publishes its own inequality statistics.9 CBO receives a sample of tax return data from the Internal Revenue Service about 2 years after the tax year ends, and CBO statistically matches the tax data and the Current Population Survey.
The Congressional Budget Office also adjusts for underreporting of benefit receipt and imputes the value of health insurance provided. For Medicaid, CBO assigns each household receiving Medicaid the average dollar cost to the government by subgroup. CBO uses administrative total Medicaid payments to a subgroup divided by the number of people in that subgroup.10
The measures of income used in estimating inequality by the Congressional Budget Office and the Census Bureau represent just two ways of doing so. Both ways show an increase in inequality over the past 40 years when using the Gini coefficient or when using the share of resources going to the top of the income distribution. Inequality researchers sometimes use other definitions of income, such as before-tax income from tax returns, which also shows increasing inequality.11 Almost universally, these different inequality measures show increasing inequality.
The only measures that won’t show increasing inequality over the past 40 years, or very little increase in inequality, are those that ignore the top of the distribution, such as the 90/10 ratio. The 90/10 ratio calculates the income of the household at the 90th percentile and divides it by the income of the household at the 10th percentile. Inequality over the past 40 years is driven by the top 10 percent, and even the top 1 percent. Using the 90/10 ratio as the measure of inequality excludes the portion of the distribution that is driving inequality.
Failure to capture multidimensional inequality
Another issue with official government statistics is that they fail to capture the multidimensional nature of poverty and inequality. Income, consumption, and wealth each convey important information.12 Income is how much money is coming into households. Consumption describes families’ standard of living. And wealth details the cushion that households have against hard times, which is especially for low-income households. Those who are poor measured by income may not be poor measured by consumption or wealth.
As with income, researchers use different definitions of consumption and wealth, and they use different inequality measures. Using consumption, one measurement disagreement regards the inclusion of education expenses. Some researchers exclude education spending on the argument that education spending is an investment.13 Others include education spending as part of consumption. Inequality is lower when excluding education expenses because of higher education spending by higher-income households, but inequality increased since 1984 using both consumption measures.14
Research using wealth to measure inequality typically uses either the Federal Reserve Board’s Survey of Consumer Finances, or it uses tax data. Regardless of data source or wealth definition, researchers find increasing inequality.15
Inequality in two and three dimensions is growing, and growing faster than inequality in any single dimension. In 1989, 1.7 percent of Americans were in the top 5 percent of the income, consumption, and wealth distributions. In 2007, 2.5 percent of Americans were in the top 5 percent of all three resource measures, indicating that the country is becoming more stratified across all three measures. While the percentages in the top 5 percent of all three have fallen since the Great Recession of 2007–2009, it still exceeds the 1989 levels.16 Those who are at the top of the income distribution are more likely to also be at the top of the wealth distribution and the top of the consumption distribution today than they were in 1989. (See Figure 3.)
Figure 3
This work was only feasible by imputing consumption to the Survey of Consumer Finances, or SCF. The SCF captures income and wealth but has very little on consumption. In addition, the SCF is collected by the Federal Reserve Board, which is not a statistical agency. It would be better for the data to be collected and distributed through the federal statistical system. The Survey of Income and Program Participation, or SIPP, captures income and some wealth elements. No federal survey collects all three measures.
Federal statistics conceal substantial variation by race, gender, and ethnicity
The federal statistical system cannot represent the full dimensions of inequality and poverty by race, gender, ethnicity, and other important characteristics. The American Community Survey, or ACS, surveys the largest number of households, including households on Indigenous lands and households in Puerto Rico. The ACS surveys more than 2 million households per year. But the ACS misses many relevant income sources and has little on consumption and wealth. The Current Population Survey, or CPS, is used for income poverty and inequality calculations and provides comprehensive income measurement, but it only surveys around 50,000 households. The Survey of Income and Program Participation, or SIPP, is also a fantastic data source that captures income and some elements of consumption and wealth, but the SIPP only interviews around 40,000 households as well. The CPS and SIPP lack the sample size to produce statistics about smaller populations.
Small sample sizes inhibit our ability to understand differences between groups in a timely manner. Statistical agencies typically combine groups such as Asian American, Native Hawaiian, and Pacific Islander people, or AANHPI, into one group because of small sample sizes, but this group is not homogenous. In aggregate, AANHPI men and women do have higher average earnings than men and women of the other major racial and ethnic groups. (See Table 1.)
Table 1
Yet these aggregate statistics mask important inequities within the AANHPI community.17 Research by Rakesh Kochhar and Anthony Cilluffo of the Pew Research Center, for example, shows that income is more unequally distributed among the Asian American and Pacific Islander population than among any other major racial or ethnic group, with Asian American and Pacific Islander families in the top 10 percent of the income distribution earning 10.7 times as much as those in the bottom 10 percent.18
In the case of AANHPI women specifically, disparities in U.S. labor market outcomes are evident in their overrepresentation in both high- and low-wage occupations. An analysis by the National Women’s Law Center shows, for example, that Asian American and Pacific Islander women make up about 3 percent of the U.S. workforce, yet they represent 4.2 percent of all workers in the highest-paying jobs and 4.3 percent of all workers in the lowest-paying jobs in the U.S. economy.19
Policy recommendations
In which direction should the federal statistical system move to help understand whether the important work of this committee will be successful? We describe four important steps that should be taken.
GDP 2.0
Gross Domestic Product needs an upgrade. Rather than ignore distributional differences, what if the federal statistical agencies pursued what the Washington Center for Equitable Growth calls GDP 2.0? GDP 2.0, which is referred to as Distributional National Accounts by academic economists, is the idea that we should add to our national accounts a breakout of growth for different populations. In our earlier hypothetical example of an economy with 20 individuals, one of whom inherited $1 million, using GDP 2.0, the Bureau of Economic Analysis would report that income increased 20-fold for the top of the distribution while it was unchanged for the bottom of the distribution. Everyone in the economy would then understand what happened much better.
Encouragingly, the Bureau of Economic Analysis has recently produced prototype statistics that distribute personal income growth. Personal income is a measure of total output, analogous to GDP, that requires fewer assumptions to distribute. Currently, this prototype data series distributes growth for each decile of the income distribution.
To see how a permanent, real-time version of this prototype could be useful in understanding and responding to recessions, we used the recently developed BEA prototype, produced with a 2-year lag, to chart growth in income during and after the Great Recession.20 The policy response to the Great Recession left millions of low- and middle-income families struggling while wealthy families saw significant additions to their incomes.
Consider the cumulative growth in disposable personal income of high- and low- or moderate-income Americans. High-income households in the top 10 percent initially suffered a steep drop in income at the onset of the Great Recession due primarily to the collapse of income from assets such as stocks and bonds, which cratered early in the economic downturn, alongside business income losses. Low- to middle-income households in the bottom half of the income distribution did not initially suffer as dramatic a fall. (See Figure 4.)
Figure 4
One reason for the less dramatic shock to income seen in Figure 1 among households in the bottom 50 percent was that disposable personal income incorporates transfers from the federal government to households, so losses in this group were partially compensated for by rising Unemployment Insurance payments, Supplemental Nutrition Assistance Program benefits, and other government benefits. But this group then became mired in years of stagnant, or even declining, income as these benefits ended amid a still-tepid economic recovery and did not experience substantial income gains until 2015. In fact, the bottom 50 percent did not recover all that was lost in wages in the Great Recession until 2017. (See Figure 5.)
Figure 5
By comparison, households in the top 10 percent of income recovered almost immediately after the end of the Great Recession and ended 2018 up 22 percent, compared to 2007. Importantly, the jump in top incomes in 2012 and steep decline in 2013 seen in Figure 3 does not represent a real decline. Rather, it is the result of households retiming their income to occur in 2012 so they could avoid rising top-income tax rates in 2013, as a result of the expiration of high-income tax cuts first enacted during the George W. Bush administration.
BEA’s dataset makes it clear that the labor market had not recovered by 2010 and that households in the bottom half of the distribution were in recession well into 2015. Although some U.S. labor market statistics hinted at this weakness, the relatively healthy headline GDP growth of 2010 convinced legislators in both parties to “pivot to austerity” in an attempt to reduce deficits. This pivot had a disastrous human cost, as many families saw their incomes dip as the government ratcheted down spending.
Had federal statistical agencies instead focused on the distribution of growth, then the public, the media, and policymakers alike would have been better informed. And policymakers might have greatly helped American families by passing stimulus targeted at this bottom 50 percent group instead of cutting programs.
In addition to reporting growth by level of income, GDP 2.0 also might eventually include reporting on growth for rural residents, Americans in marginalized groups, and so on.
Here are concrete steps Congress can take to improve the reporting of economic growth.
The BEA data series remains a prototype. Congress should give BEA a mandate and the necessary resources to expand and improve upon this product. Currently, it is released only annually and with a significant time lag. Congress should pass the Measuring Real Income Growth Act, which would require BEA to include distributional analysis with each GDP report.21 BEA also needs additional resources to create the necessary high-frequency and low-lag data series.
Congress should implement the administration’s proposal in the FY 2022 Treasury Green Book that would expand access to federal tax information to the Bureau of Labor Statistics, or BLS, and BEA.22 This expansion would not impact the privacy of tax data and would remedy a longstanding problem that has harmed the accuracy of BLS and BEA data series and prevented these agencies from pursuing promising areas of research.
FAIR Scoring Act
We want to highlight one related measure Congress should consider taking to achieve fair growth. The Fiscal Analysis by Income and Race, or FAIR, Scoring Act, would direct the Congressional Budget Office to prepare distributional analyses by race and income for all legislation with substantial budgetary effects.23 This bill is, in some ways, the other side of the GDP 2.0 coin. GDP 2.0 will help us assess, in retrospect, how policy is impacting the distribution of economic resources. The FAIR Scoring Act would help legislators understand the disparate effects of legislation in advance of passing it. Under the proposed bill, CBO would report not simply the expansion costs of the Child Tax Credit, for example, but would also note the benefits—a 50 percent decrease in child poverty that was previously only reported by independent analyses.
These scoring improvements would allow members of Congress to target particular economic disparities and fine-tune legislation to address them. Many members of Congress have expressed a desire to address the vast wealth divide between White and Black families. If the FAIR Scoring Act were implemented, then forecasts would show how much legislation would widen or narrow this divide.
Need for integrated and expanded data from the federal statistical system
To improve measurement of disparities for smaller subpopulations and disparities across income, consumption, and wealth, Congress should support the Census Bureau and Bureau of Labor Statistics working together to develop one survey that can:
Measure income, consumption, and wealth. Having all three measures in one survey would decrease academic disagreement on the proper survey or definitions to use.
Reconcile the reporting of these measures, so that they are close to agreement. Income and consumption are often underreported in surveys. Asking survey respondents to reconcile their income and consumption can help reduce underreporting.
Link to administrative records, such as tax data. Linkage to administrative data can also help with underreporting. In addition, linkage allows researchers to virtually follow respondents before and after the respondents actually participate in the survey. Linkage also would allow for more detailed study of intergenerational mobility.
Have a small longitudinal component. Income and consumption volatility harm the well-being of households, especially among low-income households.24 A survey that can track this volatility by surveying the same households over multiple years will allow for further study of volatility and the role fiscal policy plays in exacerbating or alleviating volatility.
Capture differences by race, ethnicity, and gender. This can be done by increasing the sample size and oversampling small subpopulations.
Next steps
The federal statistical system fails to provide the necessary information to Congress, the media, and the American people on the state of the economy and the extent of inequality. These four steps can improve the federal statistical system.
Congress must give the Bureau of Economic Analysis the resources to expand and improve upon its current Distribution of Personal Income product, what Equitable Growth calls GDP 2.0. Currently, it is released only annually and with a significant time lag. Congress should pass the Measuring Real Income Growth Act, which would require BEA to include distributional analysis with each GDP report. BEA also needs additional resources to investigate the creation of a high-frequency and low-lag data series.25
Congress should also implement the administration’s proposal in the FY 2022 Treasury Green Book that would expand BLS and BEA access to federal tax information. This expansion would, without impacting the privacy of tax data, remedy a longstanding problem that has harmed the accuracy of BLS and BEA data series and prevented these agencies from pursuing promising areas of research that will help policymakers, business, and the American people.
Congress should pass the FAIR Scoring Act. This bill would direct CBO to prepare distributional analyses by race and income for all legislation with substantial budgetary effects. This bill is, in some ways, the other side of the GDP 2.0 coin. GDP 2.0 will help us assess, in retrospect, how policy is impacting the distribution of economic resources. The FAIR Scoring Act would help legislators understand the disparate effects of legislation in advance of passing it.
Congress should support the Committee on National Statistics, which is part of the National Academies of Science, Engineering, and Medicine, and its push for an integrated statistical system that can measure income, consumption, and wealth poverty and inequality for the same households. This integrated statistical system should oversample smaller subpopulations, allowing for a more detailed understanding of which subpopulations are impacted or not impacted by fiscal policies.
Lastly, it is important that these efforts reside within the federal statistical system. BEA should continue to research and improve its GDP 2.0 efforts. BEA acts as the neutral arbiter among academics and can adopt the best practices developed among researchers and among other countries’ statistical agencies. The integrated statistical system should also live at a statistical agency, such as the Census Bureau or Bureau of Labor Statistics. These agencies contain the expertise to collect the surveys and design the questionnaires.
A new working paper sheds new light on the lack of bargaining power among low-wage workers in the United States by examining workers’ ability to bargain over wages when they are dual jobholders. The paper, “Wage Posting or Wage Bargaining? A Test Using Dual Jobholders,” examines whether improvements in the wages of workers’ second jobs increase their ability to bargain for higher wages in their primary jobs.
The new working paper, by economists Marta Lachowska at the W.E. Upjohn Institute for Employment Research, Alexandre Mas at Princeton University, Raffaele Saggio at the University of British Columbia, and Stephen Woodbury at Michigan State University, is an important contribution to economic research. It tests the extent to which two different models of wage determination—either wage posting or wage bargaining—come into play for workers at different wage levels.
The wage-posting model is a “take it or leave it” approach, where an employer posts a nonnegotiable wage at the point of making a job offer to a prospective employee. The wage-bargaining model, by contrast, proposes that wages are determined by a process of bargaining between workers and employers, which gives a worker the ability to seek a higher wage through the hiring process.
Using administrative linked employer-employee data on workers in Washington state, the researchers test which model is more prevalent by looking at how workers respond to wages at their second jobs, where they can use outside income to improve their bargaining position at their current employer or choose to leave their employer if they are not able to seek higher pay at their primary job. Their findings are telling when considered across the breadth of the U.S. labor market.
The four researchers find that the effects of wage bargaining are small overall and apply most to high-wage workers, who have a relatively greater ability to use their outside options to bargain for higher wages. But the wage-bargaining model does not appear to apply to low-wage workers. For them, the wage-posting model better explains their behavior in the job market because they generally can only raise their wages by quitting their primary job in exchange for a higher-paying job elsewhere.
Taken together, this means that bargaining over wages at a current job generally does not achieve a large impact on workers’ wages in general, but high-wage workers can more often bargain at their current jobs to improve their wages, while low-wage workers can do very little to improve their wages at their jobs other than leave for a new one.
These findings suggest that wages are posted at a given, nonnegotiable rate for low-wage jobs, leaving workers in these jobs with little ability to bargain over their wages. This helps explains why market-level solutions such as raising the minimum wage are so effective for low-wage jobs in the current legal landscape.
The new research also illuminates another possible way that the lack of bargaining power for low-income workers exacerbates income inequality, as other research suggests that wages tend to be higher when workers can bargain over them, compared to when the “take it or leave it” wage-posting model is employed at their workplaces.
These findings together can help economists and policymakers alike understand how wages are determined in the U.S. labor market and how this connects to wage inequality. A growing body of research shows that employers have outsized power in employment relationships in the United States, allowing them to undercut workers’ wages. As a result, workers are paid over one-third less than they would be in a truly competitive labor market.
The importance of unions is not lost on U.S. workers. Equitable Growth grantee Alexander Hertel-Fernandez notes that there is strong worker demand for unions, but federal and state labor laws generally do not allow for approaches that could help large numbers of workers, such as industrywide or statewide collective bargaining.
Conclusion
These findings are important for policymakers who want to improve wages for low-income workers and reduce income inequality. The working paper shows that low-wage U.S. workers don’t have appreciable bargaining power under current U.S. labor market dynamics and laws and institutions.
What’s more, other economic evidence demonstrates that policymakers can address market-level factors underpinning persistent low-wages by instituting statutory minimum wages that would increase posted wages and by addressing the structural conditions that increase low-wage workers’ bargaining power, such as income support programs and support for institutions that foster wage bargaining such as labor unions.
Because low-wage workers in the United States have so little bargaining power, the steady erosion of the federal minimum wage has contributed to rising economic inequality and stagnant and declining wages for workers in the lower part of the pay distribution. This means raising wage standards is one of the most important and effective approaches to improving wages for low-wage workers.
Higher minimum wages also are critical for addressing racial wage gaps. A 2019 analysis by economist Valerie Wilson at the Economic Policy Institute found that increasing the federal minimum wage to $15 an hour would increase earnings for 38.1 percent of Black workers and 23.2 percent of White workers.
Understanding the various ways in which wages are set in the U.S. economy is crucial to analyzing economic activity and worker well-being. Phenomena such as wage and earnings growth, worker mobility, income inequality, and the underlying causes of unemployment all relate back to wage setting and workers’ ability to push for higher pay or better working conditions.
Wages in the United States are typically set either through wage bargaining or wage posting. The former occurs when workers have the ability to negotiate their own compensation, and wages thus respond to workers’ outside options—their ability to find alternative employment. The latter, wage posting, arises when employers set wages that employees must either accept or reject, without holding influence over how much they are paid.
Little evidence exists on the extent to which workers have bargaining power and outside options that could affect their wages. But our new working paper seeks to expand the evidence by looking at dual jobholders—workers who take a second job because the available hours at their primary job (which typically has better wages) are limited by the employer. On average, dual jobholders rely on their secondary jobs for about 20 percent of their total earnings.
Using employer-employee linked administrative data on wages from Washington state’s Unemployment Insurance system, we analyze data on wage changes for a worker’s colleagues at their secondary jobs, or the job in which they work fewer hours. In particular, we estimate how workers’ wages and hours change at their primary jobs—as well as whether these workers quit their primary jobs—in response to wage increases at their secondary jobs.
We find that the relationship between changes in the wages of co-workers in a secondary job and changes in the worker’s wages at the primary job differ for worker in different parts of the wage distribution. Wage bargaining does play a role in setting workers’ wages but only in the top quartile of the wage distribution. For workers in the lowest quartile, wage posting appears to be the dominant method of wage setting.
The difference between wage setting in the top quartile of the wage distribution and lower in the distribution can be attributed to the greater economic surplus generated by high-wage workers. The gap between the value produced by high-wage workers and what they are paid—the surplus—is greater for high-wage than for low-wage workers largely because they are difficult to replace. The high costs of recruiting, hiring, and training high-wage workers create an incentive for employers to raise their wages rather than potentially losing them.
In contrast, workers in the lowest quartile of the wage distribution are more likely to quit their primary jobs in response to higher pay in their secondary jobs, consistent with a lack of bargaining power. This is likely because employers can more easily replace low-wage workers.
In short, high-wage workers are able to increase their earnings through wage bargaining at their primary jobs, whereas low-wage workers need to change jobs to increase their pay.
Our study also examines single jobholders, finding that these workers, on average, earn higher hourly wages than dual jobholders at either their primary or secondary jobs. The work hours of single jobholders tend to be more than the work hours of dual jobholders at their primary jobs, although the total work hours of dual jobholders exceed the total work hours of single jobholders.
We also find that dual jobholders more often work with other dual jobholders. Only 5 percent of single jobholders’ colleagues have secondary jobs, while 9 percent of dual jobholders’ colleagues in their primary jobs and 48 percent of their colleagues in their secondary jobs have two jobs.
In light of the rise in U.S. gig employment in recent years, it is important to note that the secondary jobs held by dual jobholders in our study are not informal, app-based, or contract-based occupations. Because we use data from Washington’s Unemployment Insurance system, and because gig workers are not eligible for UI benefits in most cases, we look at dual jobholders with second jobs that have more flexible hours, but who are still classified as employees by both of their employers.
The findings in this paper are closely linked to our related research on growing wage inequality in Washington since 2002. In “Do Firm Effects Drift? Evidence from Washington Administrative Data,” we find increasing wages in the top quartile of the wage distribution are responsible for most of that growing inequality. We also find that the skills possessed by high-wage workers have been rewarded with increasing wage premiums, and that high-wage workers have sorted increasingly to high-wage employers. The ability of high-wage workers to negotiate still-higher wages can be viewed as an outcome of the increased value employers have placed on workers’ skills in the U.S. labor market.
1. My impression is that Equitable Growth’s academic grantgiving is more short of good projects to fund than short of money with which to fund them, so it is definitely worthwhile to throw your hat into the ring if you have what you think is a good idea for how to discover important things about equitable growth. Go to Equitable Growth’s “2022 Request for Proposals,” which “seeks to deepen our understanding of how inequality affects economic growth and stability. To do so, we support research investigating the various channels through which economic inequality, in all its forms, may or may not impact economic growth and stability. Equitable Growth promotes efforts to increase diversity in the economics profession and across the social sciences. We recognize the importance of diverse perspectives in broadening and deepening research on the topics in this request for proposals.”
2. In an interview with NPR, Michelle Holder does a very good job explaining “how the economy has managed to pull off a disappearing act with people’s wages.” Listen to Stacey Vanek Smith and Julia Ritchey, “The Money Illusion: Have Americans really gotten a raise?”
Worthy reads not from Equitable Growth:
1. I confess I do not understand why anybody thinks that the 1970s are a better model for what is now going on with inflation than after World War II. This seems to me to be a likely mistake that nobody who has learned any history could make. Read Paul Krugman, “History Says Don’t Panic About Inflation,” in which he writes: “Back in July the White House’s Council of Economic Advisers posted a thoughtful article to its blog titled, ‘Historical Parallels to Today’s Inflationary Episode.’ The article looked at six surges in inflation since World War II and argued persuasively that current events don’t look anything like the 1970s. Instead, the closest parallel to 2021’s inflation is the first of these surges, the price spike from 1946 to 1948. … It was a one-time event, not the start of a protracted wage-price spiral. And the biggest mistake policymakers made in response to that inflation surge was failing to appreciate its transitory nature: They were still fighting inflation even as inflation was ceasing to be a problem, and in so doing helped bring on the recession of 1948–49. … Demand in the United States actually doesn’t look all that high; real gross domestic product … is still about 2 percent below what we would have expected the economy’s capacity to be if the pandemic hadn’t happened. But demand has been skewed, with consumers buying fewer services but more goods than before, putting a strain on ports, trucking, warehouses and more. These supply-chain issues have been exacerbated by the global shortage of semiconductor chips, together with the Great Resignation—the reluctance of many workers to return to their old jobs. So we’re having an inflation spurt. … So what can 1946–48 teach us about inflation in 2021? Then as now there was a surge in consumer spending, as families rushed to buy the goods that had been unavailable in wartime. Then as now it took time for the economy to adjust to a big shift in demand. … But the inflation didn’t last. It didn’t end immediately: Prices kept rising rapidly for well over a year. Over the course of 1948, however, inflation plunged, and by 1949 it had turned into brief deflation. … An inflation spurt is no reason to cancel long-term investment plans. The inflation surge of the 1940s was followed by an epic period of public investment in America’s future, which included the construction of the Interstate Highway System. That investment didn’t reignite inflation — if anything, by improving America’s logistics, it probably helped keep inflation down. The same can be said of the Biden administration’s spending proposals, which would do little to boost short-term demand and would help long-term supply…. People making knee-jerk comparisons with the 1970s and screaming about stagflation are looking at the wrong history. When you look at the right history, it tells you not to panic.”
2. Disagree with this assessment of the housing market bubble in the 2000s, but it is smart. After the start of the Great Recession, the United States was starved of housing construction for a decade, especially construction in places where people can get jobs of high productivity or where people can get very good value for their leisure. This shortage of supply means that today’s elevated housing prices are quite likely to be “fundamental.” But 2005 was half a generation ago, and then the supply demand balance in housing was different. So values in 2005 were a bubble, I think, even though the same inflation-adjusted values today are “fundamental.” Read Timothy B. Lee, “The 2000s housing bubble was greatly exaggerated,” in which he writes: “Housing prices are now above the supposedly unsustainable levels of 2006. And that’s after adjusting for inflation. And yet not very many people think we’re in the middle of a second housing bubble. Rather, most experts believe that today’s housing prices reflect ‘fundamental’ factors. Interest rates are at all-time lows, giving homebuyers more spending power. And regulatory restrictions have created housing shortages in many metropolitan areas. But that leads to a question that at first glance might seem crazy: What if those same explanations largely explain the housing boom that peaked in 2006? What if the big problem in the early 2000s wasn’t an excess of houses but a shortage of them? … That’s the thesis of ‘Shut Out,’ a 2019 book by Kevin Erdmann. … Erdmann argues that policymakers misdiagnosed the causes of the housing boom, and that led to catastrophic policy errors. In particular, because the Federal Reserve thought housing was overvalued in 2007, it didn’t cut rates fast enough in response to the housing crash. … If Erdmann and Schubert are right, we’re still living with the consequences of misdiagnosing the housing boom as a speculative bubble.”
Every month the U.S. Bureau of Labor Statistics releases data on hiring, firing, and other labor market flows from the Job Openings and Labor Turnover Survey, better known as JOLTS. Today, the BLS released the latest data for September 2021. This report doesn’t get as much attention as the monthly Employment Situation Report, but it contains useful information about the state of the U.S. labor market. Below are a few key graphs using data from the report.
The quits rate continued to rise in September to 3.0 percent as nearly 4.4 million workers quit their jobs, an increase of 164,000 since August.
The vacancy yield remains extremely low as job openings (10.4 million) and hires (6.5 million) stayed relatively constant in September.
Quits are elevated and continued to rise in industries such as manufacturing, leisure and hospitality, and education and health services.
The ratio of unemployed-worker-per-job-opening decreased from 0.79 in August to 0.74 in September. At the same time, 183,000 workers left the labor force in September.
The Beveridge Curve continues to be in an atypical range compared to previous business cycles, with the unemployment rate declining to 4.8 percent and the job openings rate still elevated at 6.6 percent.
post
post
