This is a post we publish each Friday with links to articles that touch on economic inequality and growth. The first section is a round-up of what Equitable Growth published this week and the second is relevant and interesting articles we’re highlighting from elsewhere. We won’t be the first to share these articles, but we hope by taking a look back at the whole week, we can put them in context.
Equitable Growth round-up
Earlier this year, the U.S. Consumer Price Index registered inflation at 5 percent, leading some observers to panic about prolonged inflation and stagflation. Yet, write Francesco D’Acunto and Michael Weber, while long periods of inflation do have direct and immediate impacts on the economy and can exacerbate inequality, policymakers must assess whether the threat of inflation is real or if this is a short-term adjustment that will recalibrate as the economy begins to reopen. D’Acunto and Weber discuss four potential drivers of inflation in the medium to long term: demand pressures, supply chain disruptions, labor market pressures, and inflationary expectations. They detail what each of these drivers is, how it can affect inflation, and how it is relevant to the particular situation in which the U.S. economy currently finds itself. They conclude that while these factors may influence short-term inflation, they do not appear to imply that there will be sustained inflationary pressure in the coming 2–5 years. This means, they explain, that the Federal Reserve probably does not need to take any action to address inflation and that the Biden administration should continue to pursue its current policy agenda.
There has long been a gap in political participation along income lines in the United States, with wealthier Americans turning out to vote in higher numbers than their middle- and low-income peers. But in 2020, many states enacted new voting laws to ease access to the polls amid the coronavirus pandemic, such as expanding vote by mail and increasing the number of ballot drop boxes available, and new data released by the U.S. Census Bureau reveals the impact these laws had on election turnout. In a follow-up column to their February 2021 report on the relationship between voter suppression and economic inequality, Austin Clemens, Shanteal Lake, and David Mitchell analyze the new Census data to determine whether the income divide in voter turnout narrowed law year. They find that in states that made it easier to vote by mail, turnout was higher in the 2020 election across income groups, but that the effect was larger for lower-income individuals. The co-authors explain why the rash of new state laws restricting voting access, as well as the recent U.S. Supreme Court ruling that further defangs the Voting Rights Act, will have a disproportionate impact on low-income voters and voters of color—and what that means for U.S. economic policy. They conclude by urging the federal government to intervene with legislation that protects the right to vote and access to the polls for all Americans.
This week, the U.S. Bureau of Labor Statistics released data on hiring, firing, and other labor market flows from the Job Openings and Labor Turnover Survey, better known as JOLTS, for the month of May 2021. This report contains useful information about the state of the U.S. labor market, such as the rate at which workers are quitting their jobs and the ratio of unemployed workers-to-job openings. Kathryn Zickuhr and Clemens put together a series of graphics highlighting the trends in the data.
Check out Brad DeLong’s latest Worthy Reads column, where he provides summaries and his analysis of recent must-read content from Equitable Growth and around the web.
Links from around the web
Wondering what inflation means for you and how it is impacting the broader U.S. economy? Vox’s Rani Molla and Emily Stewart explain the inflation debate in the United States, what the recent increase in prices means, and the industries most affected right now and why. They detail the impact the coronavirus pandemic and the resulting recession have had on global supply chains and look at some specific examples of goods and services that have become more expensive as a result. Molla and Stewart then provide an update on what is happening in the lumber industry after perhaps the most talked-about price surge in the U.S. economy in the past year. They conclude by discussing how experts are evaluating these higher prices and what to look for in terms of how long this will last.
Last week’s Employment Situation Report from the U.S. Bureau of Labor statistics shows a healthy 850,000 jobs were added in June. Julia Coronado compiles seven charts that explain what is going on with the U.S. economy and recovery in The New York Times. She looks at how the “go early and go big” policy response to the coronavirus recession made a huge impact on the speed with which the economy is recovering. She also examines the impact of this recession and the responding policy on wealth in the United States, consumer confidence and loan delinquency, labor market trends such as job openings, the racial unemployment divide, and inflationary pressures both now and in the future. She concludes that while the policies enacted to counteract the coronavirus recession were not perfect, the bold action is paying off and will allow us to emerge from this downturn into a stronger, healthier economy.
As the West Coast prepares for another heat wave this weekend, Robinson Meyer explains in The Atlantic how unprepared U.S. infrastructure is to handle extreme weather. He details the rise in climate change-related weather events, including heat waves, and how the Biden administration’s plans to include climate policy in its infrastructure plan will bolster U.S. preparedness for the increasing prevalence of these events. Meyer also writes that engineering standards and new physical infrastructure construction have not incorporated the changing climate quickly enough to be innovative and well-equipped to handle the future—and that making these big shifts in resilience standards is more challenging and time-consuming than it may seem.
After a decade of below-target rates of inflation and the widespread fear of a Japanese-style prolonged economic stagnation with low inflation and zero growth, the incipient economic recovery in the United States from the coronavirus recession is suddenly turning the tables on the debate about inflation.
In May 2021, the Consumer Price Index registered inflation running at 5 percent, the highest reading in a decade for this broad measure of prices—so high that some observers worry about strong and sustained inflation and reference the stagflationary period of the 1970s. At the same time, the Federal Reserve is largely downplaying these concerns, so much so that in late June, Fed Chair Jerome Powell argued before Congress that “it is very, very unlikely” that the United States will face strong inflationary pressures going forward.
Prolonged periods of high inflation do have direct and immediate effects on the macroeconomy and also the potential for more subtle effects of further increasing economic inequality and crimping more sustainable and more equitable growth. Indeed, because many low-income and middle-class households, including many households of color, spend most of their income over the course of their everyday lives, prolonged price increases would affect them more than wealthier households, which save part of their income in real assets whose returns are shielded from inflation.
So, ultimately, what should policymakers really expect to happen to inflation over the next 2 to 5 years? The answer is important, as the Biden administration and the U.S. Congress debate the merits of the proposed infrastructure packages to address fragilities in the U.S. economy and move it toward more stable and sustainable economic growth.
Before focusing our discussion on the plausible and implausible drivers and effects of inflation going forward, we need to stress that the recent 5 percent inflation rate appears much less concerning once we account for what economists call the base rate effect—the price level 12 months ago, in the midst of a pandemic recession in which demand was artificially low due to lockdown measures and inflation was unusually depressed. As this base rate effect vanishes over the next few months—due to the sharp drop in the price level early in the recession partially reversing in the second half of 2020—then, all else being equal, policymakers should expect more moderate Consumer Price Index inflation readings going forward.
But there still remain four very relevant potential drivers of inflation over the next 2 years:
Demand pressures
Supply chain disruptions
Labor market pressures
Inflation expectations
We’ll examine in this column the extent to which these four factors might or might not be relevant in the current situation.
Demand pressures
Inflation can result when demand for goods and services outstrips suppliers’ ability to produce the desired goods and services to such an extent that permanent price increases are needed to bring the market to a new equilibrium. These dynamics, at first glance, seem to fit the economic recovery from the coronavirus pandemic, during which many suppliers were forced to shut down and needed time to bring production back to normal levels. But those supply chains are now recovering.
Demand pressure was also seemingly further fueled by fiscal and monetary policy interventions aimed at alleviating the most immediate negative impacts of the pandemic and resulting recession. Direct state and federal transfers in the form of direct payments to U.S. families, extended Unemployment Insurance, and emergency supplemental UI benefits, among other recession-fighting policies, ensured that households in which the breadwinners suddenly lost their jobs could maintain, at least in part, their purchasing power and repay their debts.
Yet these fiscal and monetary pressures were one-off interventions. A large majority of households might have received income transfers even when they were not facing large drops in their disposable income based on the high income thresholds used to target the program, such as a joint income of more than $160,000 for married couples. Paired with enhanced precautionary savings motives and limited opportunities to spend on leisure and travel during the pandemic, U.S. households’ savings did surge from a rate of 8.3 percent in February 2020 to 33.7 percent in April 2020 and 27.7 percent in June 2021.
This accumulation of higher savings, paired with new spending opportunities and the reopening of the U.S. economy, made the scope for demand pressure very concrete. Indeed, new car and home sales are at their highest levels since before the Great Recession of 2007–2009.
The ultimate question is whether these demand forces are a substantial concern in terms of inflation over the next 2 years. We find it unlikely. Whereas demand pressure will certainly impact prices in the short run, the stock of household savings accumulated over the past year and a half is limited, and suppliers will be able to adjust production to scale in a few months. What’s more, the delayed spending during the pandemic was primarily for services, rather than durable goods, which means the economy can only make up so much lost ground on delayed services spending.
Overall, then, there is no compelling reason to expect a long-lasting demand pressure on prices.
Supply chain disruptions
Moving to the supply side of the U.S. economy, inflation can also arise when firms experience increases in the costs of production—for example, of their intermediate inputs—or those costs that firms often pass onto their customers depending on pricing strategies and profit margin considerations. Global supply chains, whereby crucial components of finished goods are provided by suppliers around the world at greatly reduced costs, have been on the rise over the past several decades but were severely disrupted by pandemic-related transportation restrictions and the shutdown of production sites in response to the pandemic-induced public health crisis. This severely impacted the cost of outsourcing through global supply chains and hence put upward pressure on consumer prices.
For instance, car manufacturers currently face severe bottlenecks in supply chains due to the shortage of semiconductors, which increase their input costs and pressure them to increase the consumer prices of their finished goods. Another example is medical supplies that are largely produced in Southeast Asia, which was largely shut down during the early parts of the pandemic, resulting in stock shortages and price increases.
Similar to the demand pressures discussed above, though, there is little reason to believe that supply-side disruptions will last beyond the time when global supply chains will again be fully operative. Therefore, we do not see how supply chain disruptions might impact inflation in the medium to long term.
Take the price of lumber. Due to the increased demand for construction, timber prices almost doubled in the first half of 2021. Many observers argued that we would face sustained price pressure for lumber and other commodities. Yet, as of June 2021, the price of timber is pretty much back to its January 2021 level. Similarly, while the price of financial contracts for Crude Oil WTI futures—the most actively traded crude oil contract—is at levels previously seen in 2018, there is no reason to believe that prices will keep going up in the medium run.
Labor market pressures
Another way through which costs of production for firms can increase is through higher wages for workers. Of course, higher wages will ensure that consumers’ purchasing power is sustained, but waves of wage increases could be inflationary. So, let’s look at the details.
With demand picking up across the U.S. economy as it reopens, many businesses are starting to increase their hiring. Total unfilled vacancies have reached their highest level since 2000. At the same time, the overall U.S. labor market still has a substantial amount of slack—that is, a large amount of workers who have not yet returned to the workforce. The June 2021 employment-to-population ratio stands at 58 percent, recovering from a low of 51.3 percent in April 2020 but still substantially below the pre-pandemic level of 61.1 percent in February 2020.
Moreover, the pre-pandemic level was still below the employment-to-population ratio before the Great Recession of 2007–2009. Hence, the U.S. labor market is still a long way off from a full recovery and has ample slack to absorb additional labor demand, despite the unemployment rate decreasing from 14.8 percent in April 2020 to 5.8 percent in May 2021.
This latter point is important because macroeconomists typically argue that slack in the economy affects realized inflation due to the Phillips curve—the relationship between inflation and the extent to which the resources in the economy are utilized. Ample slack implies little price pressure.
Recent evidence indicates a flatter Phillips curve, which means that even if the economy were running at full capacity, there is little chance for sustained inflationary pressure to materialize. At the same time, we also observe that the pandemic resulted in a wave of early retirements alongside many workers who are not willing to work in contact-intense industries because of health concerns. This reduction in labor supply can result in temporary labor shortages and higher hiring bonuses and wages. These effects can result in higher price levels—for example, for restaurants—but they are unlikely to result in sustained inflationary pressures because of the overall amount of slack in the U.S. economy.
Inflation expectations
Expectations are a crucial determinant for virtually all decisions consumers make in their daily lives, such as savings-consumption choices, wage-bargaining decisions, and investment choices. Expectations matter for inflation because if consumers think prices are going to rise in the future, then they are more likely to take advantage of what seem like discounted prices today—behavior which itself bids up prices, leads to calls for higher wages, and creates a self-fulfilling feedback loop.
The Federal Reserve focuses on core inflation—overall inflation excluding volatile price series such as groceries and energy—because core inflation tends to better predict sustained inflationary pressures. The Fed also assumes that households have well-anchored inflationary expectations close to their annual target of 2 percent, on average. By contrast, recentresearchshows that households think about inflation in a very different way than what the Fed assumes.
Rather than focusing on the change in prices of an overall consumption basket and weighting those price changes by their expenditure shares, consumers tend to think about the prices of concrete goods and services they face in their daily lives. In particular, most consumers focus on the price changes of salient goods, such as gas, and the price changes of goods they purchase frequently, such as groceries.
What’s more, consumers tend to assign a larger weight to price increases, relative to similar-sized price decreases, when forming their inflation expectations. And these sources of divergence between perceived inflation and official inflation measures are more relevant for women, who tend to be more exposed to grocery prices than men due to traditional gender roles.
Now, with demand picking up, it is evident there is a substantial increase in gas prices at the pump and highly volatile grocery prices, both of which are transmitted into a large spike in household inflation expectations. The expected median inflation rate 1 year from now, as reported in the New York Fed Survey of Consumer Expectations, hit 4 percent in May, the highest number on record.
The median, however, masks a long right tail—statistical parlance for the existence of households with expectations substantially above the median—with average expected inflation across all U.S. households being substantially higher. The average number also masks substantial heterogeneity across demographics. The within-group average for the oldest age group is an inflation rate of 4.8 percent, 1.6 percentage points higher than for the lowest age group. Moreover, those with lower levels of education, on average, expect an inflation rate of 5.6 percent, 2.2 percentage points higher than the expected rate for the college-educated.
Then, there is recent research on what is known as the Euler equation, which explains why there might be some additional, but temporary, price pressures, according to a NBER working paper co-authored by one of us and also a forthcoming working paper in the Review of Financial Studies co-authored by both of us and Daniel Hoang at the Karlsruhe Institute of Technology. But to the extent that heighted inflation expectations result in increased demand pressure in the short run, more permanent price pressures might only occur if the U.S. household sector demanded higher wages due to heighted inflation expectations, which we haven’t seen materializing in recent decades of low and stable inflation and which would require sustained pressure to make up for mostly stagnant wages over the past several decades.
Conclusion
Taken together, the four highly debated factors that might influence inflation in the coming months do not appear to imply sustained inflationary pressure in the medium to long term. Hence, we do not see the need for the Federal Reserve to take policy actions in the short run because of sustained inflationary pressure. But more guidance might be needed by the Fed on how to interpret their new policy framework.
Moreover, given fading inflationary pressure in the medium run, we do not see any evidence for the Biden administration to change its policy agenda because of mounting inflation on the horizon. The pros and cons of those policies should be discussed on their merits rather than on their impact on inflation, which we believe will be temporary.
— Francesco D’Acunto is an assistant finance professor at the Carroll School of Management at Boston College. Michael Weber is an associate professor of finance at the University of Chicago’s Booth School of Business.
The 2020 election year was unprecedented in many ways. The coronavirus pandemic changed most Americans’ relationships and interactions with society and the economy, including their civic participation in the 2020 election, which saw numerous new voting rules and regulations implemented across states in response to the public health crisis.
So far, the evidence on how these new rules and regulations impacted voter turnout last year hasbeenmixed. But thanks to recently released data from the U.S. Census Bureau that measure turnout across race, income, and place, we can see these rules helped boost voter turnout, especially among low-income Americans, and reduced the income divide in political participation.
As we found in our previous report on political inequality and voter suppression, full participation in our democracy matters not just for fairness and government legitimacy, but also for combating economic inequality and spurring strong, stable, and broad-based economic growth. Yet structural racism and systemic flaws in the country’s voting apparatus—such as a lack of guaranteed time off for hourly workers to vote and underresourced polling places in densely populated areas—mean that White, high-income Americans vote at higher rates than low-income Americans of color.
The result? Richer White voters have more influence on which laws get passed, biasing economic policymaking in favor of the rich and powerful and against evidence-based proposals that would combat structural racism and deliver equitable growth, such as a higher minimum wage, increased taxes on the rich, and a stronger system of income supports.
Unfortunately, but not unpredictably, this income- and race-based voting bias was not reversed in the 2020 election. Our analysis of the new Census data shows that it continues to fester dangerously in the heart of our democracy.
Using the Current Population Survey’s Voting and Registration Supplement, we divide 2020 voters into three groups based on annual earnings: low-income individuals with incomes lower than $40,000; middle-income individuals with incomes between $40,000 and $75,000; and high-income individuals with incomes of more than $75,000. The 2020 election continued a long trend of stark divides among these groups.
According to the data, about 65 percent of low-income individuals voted, compared to 88 percent of those in the high-income category. In recent election cycles, turnout has been up in every income group, but this disparity between high- and low-income voter turnout persists. (See Figure 1.)
Figure 1
Within these top-line national numbers, though, there is much variation between states both in terms of changes to voting laws amid the pandemic and voter turnout. This variation can help researchers better understand the relationship between different voting reforms, expanded access to the ballot booth, and voter turnout.
There is already strong evidence that certain changes to voting regulations would provide more equitable access to voting across race and income levels in the United States. These changes include easing the voter registration process, reducing wait times, guaranteeing paid time off to vote, restoring voting rights to felons, and using federal authority to crack down on racial discrimination in state election administration. But the evidence has been more mixed on whether so-called convenience reforms—which make voting easier for those already registered by, for example, expanding early voting and vote by mail, or increasing the number of ballot drop boxes—also help close the racial and income gaps in political participation.
Because state election administrators responded in many different ways to the coronavirus pandemic, the 2020 election provides the means to test which convenience reforms most increase voter participation—keeping in mind, of course, that many extraordinary factors influenced voting in the 2020 election. As such, the statistics that follow are offered as descriptive evidence only. Recent research shows that voting increased at similar rates in most states in 2020, regardless of mail-in voting policies. But mail-in voting has been shown to increase voter turnout and may help decouple the income-voting link. We pick apart the data on income and race to show some suggestive evidence of how these policies play out across different groups.
Using the National Conference of State Legislatures’ guide to voting policies for the 2020 election, we divide states into three categories. First are those that either mailed applications for a mail-in ballot or simply mailed ballots to eligible voters without requiring voters to request anything from their state or county boards of election. These are the 23 states (including the District of Columbia) where voting in 2020 was easiest. Next are states that did not automatically mail anything to voters but did allow all voters or voters with COVID-19-related excuses to request a mail-in ballot. Some of these 24 states always allow voters to request a mail-in ballot, while others changed their rules for 2020 in response to the coronavirus crisis.
Finally, there are four states—Indiana, Louisiana, Tennessee, and Texas—that made no modifications to their voting policies amid the pandemic. These states kept in place rules that restrict access to mail-in ballots to only those voters with an excuse for not being able to vote in person, such as being out of the country or being physically unable to go to the polls. These were the hardest states to vote in during the 2020 election.
We calculated turnout rates in each state and then averaged state-level values in each of the three categories to avoid biasing statistics toward larger states. It should also be noted that there are relatively few states in the third, “no change from normal policy” group, so estimates in this category are less reliable. We dropped results when a race-by-income-by-state calculation had fewer than 10 unweighted survey respondents to draw on.
Voter turnout overall was very high in 2020—the highest presidential election turnout since 1900. Given the difficult circumstances surrounding the pandemic, this is a notable feat, and it may be at least partially attributable to the ease with which many states made it possible to vote by mail. Our analysis shows that states where vote by mail was easier saw higher turnout overall. (See Figure 2.)
Figure 2
While states that made it easier to vote by mail saw higher turnout across the three income groups we analyzed, these policies were more beneficial for low-income individuals. In 2020, turnout for high-income individuals was only 1.4 percentage points higher in states that automatically mailed ballots or applications to voters, compared to those that only moderately expanded accessibility, and just 3.6 percentage points higher than states that did not change their policies at all. But for low-income individuals, turnout in the most accessible states was 4.8 percentage points higher than in moderately accessible states and 7.5 percentage points higher than in states that made no policy changes. (See Figure 3.)
Figure 3
Accordingly, there is suggestive evidence that the reforms that made it easier to vote by mail did close the income divide in political participation. In the most permissive states, the gap between high-income and low-income voter turnout was about 17 percentage points. But in states that made fewer accommodations or made no policy changes at all, the participation gap between rich and poor voters was 21 percentage points.
Manyobservers of state election policy have suggested that less permissive voting regimes are intended to reduce participation by Black voters and other voters of color. States with less permissive voting rules in 2020 did have larger Black populations: States in the most permissive voting category are, on average, 9.3 percent Black, while those that somewhat expanded mail-in voting or made no changes at all are, on average, 15.3 percent and 14.5 percent Black, respectively.
Yet voting restrictions do not appear to have affected Black voters considerably more than White voters. Average White turnout in 2020 was 5 percentage points higher in the most permissive states, compared to the two less permissive categories. In the same comparison, the difference in Black turnout was 6 percentage points. One reason for this relative parity by race is that middle- and high-income Black voters showed incredible resilience to their states’ voting regimes. In fact, these individuals voted at about the same rate in all three of our policy regime categories. In contrast, high-income White individuals voted less in states with more restrictive voting regimes. (See Figure 4.)
Figure 4
Latinx citizens tend to vote at lower rates than both Black and White citizens in every income category. Their sensitivity to the ease of voting in the 2020 election was similar to that observed for White voters. Low-income Latinx citizens in the most permissive states were 4.5 percentage points more likely to vote than those in states that only somewhat expanded accessibility, compared to about 6 percentage points for White citizens. For middle-income Latinx voters, the discrepancy was about 6 percentage points, compared to about 3 percentage points for White citizens.
It is difficult, however, to make any inference about Latinx voting in the least permissive states as this category is composed of three states with relatively low Latinx populations and one state with a disproportionately large Latinx population—Texas—making the results mostly reflective of this one state. In addition, because of the limited sample sizes of Asian American, Pacific Islander, and Native American communities in the data, we were not able to conduct a reliable cross-state comparison of voting regimes’ impact on these communities.
Nonetheless, the evidence is clear that expanded access to vote-by-mail increases voter turnout. Yet many state legislators are working hard to reverse these reforms and even enact new laws that make it more difficult to vote. According to the Brennan Center for Justice, 28 laws restricting voting access have been enacted in 17 states, with these numbers expected to rise across those state legislatures still in session.
Florida’s new voting law, for example, includes several provisions that make it harder for voters to request and return absentee ballots. In addition to requiring proof of identity for absentee voting, the law also restricts the use of ballot drop boxes. Similarly, a provision in Georgia’s new voting law requires voters to have a state-issued driver’s license or ID on file to request an absentee ballot. This requirement could adversely affect voters of color, who are less likely to have a state-issued ID than their White peers.
Then, there’s the U.S. Supreme Court ruling from earlier this month, which will make it even more difficult to challenge state election regulations that have a disparate impact on voters of color under the already defanged Voting Rights Act.
One way to circumvent these attacks on voting rights and access is to enact federal legislation that overrides the Supreme Court’s limited reading of the Voting Rights Act and guarantees more permissive state rules. Congress is currently considering a major overhaul of federal voting law via the For the People Act (also known as H.R. 1 in the U.S. House of Representatives and S. 1 in the U.S. Senate) and the John Lewis Voting Rights Advancement Act (or H.R. 4 in the previous Congress, but not yet introduced in this Congress).
If passed, these bills would address major barriers to voting by requiring at least 2 weeks of early voting in all states, limiting voter purges and other discriminatory voter suppression tactics, and automating voter registration, among many other provisions. As former U.S. Treasury Secretary Robert Rubin recently argued, these bills are “key to our economic future,” with the potential to “pave the way for policies that achieve the interdependent objectives of strong growth, widespread economic well-being and reduced inequality.”
The data from the 2020 elections are clear: Making voting easier helps close the gap in political participation along income lines and consequently fosters greater consideration of the voices and economic policy priorities of more marginalized individuals when policymakers craft legislation. Indeed, it is a vital aspect of a flourishing and strong democracy. Yet if some states continue to enact restrictive voting legislation—and the federal government does not step in to stop them—then our democracy may reverse much of the progress made in 2020 and set the country back for generations to come.
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 May 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 declined to 2.5 percent as nearly 3.6 million workers quit their jobs in May, down from a series high in April.
The vacancy yield declined slightly, remaining extremely low as job openings and hires stayed relatively constant in May.
The ratio of unemployed-worker-per-job-opening was 1.01 in May, approaching its level of less than 1.0 prior to the coronavirus recession.
The Beveridge Curve continues to be in an atypical range compared to previous business cycles, as the unemployment rate declined slightly but job openings remained elevated.
1. The first reason that tax cuts on capital income and on income for the rich did not produce higher investment and faster economic growth is that an economy at full employment (or in a U.S. economy where the Federal Reserve has a very strong sense of the wedge it should maintain vis-à-vis full employment), any incentive effects of tax cuts on investment will be offset by the drag imposed by the larger government deficit on investment. Only when the economy is not at full employment, and the Fed either is or thinks it is out of ammunition, can tax cuts have any significant effect on aggregate demand. And, even then, it is not clear that the component of aggregate demand that will be boosted is the investment component. There is an income as well as a substitution effect associated with tax cuts for the elite and for capital income. And I have never seen any convincing evidence that the substitution effect dominates—and I have locked, hard. Read Corey Husak, “The relationship between taxation and U.S. economic growth,” in which he writes: “In the 1980s, policymakers followed these [neoclassical] models and drastically lowered the top individual marginal income tax rate, from about 70 percent down to 28 percent, and lowered the top corporate rate from 46 percent down to 34 percent. But instead of booming, income growth slowed. As the government reduced statutory tax rates, especially for those at the top, inequality exploded, and income growth rates went down. … Broad empirical data are the opposite of what neoclassical models predict. The U.S. national savings rate, for example, declined in the 1980s after taxes on capital dropped, and it again declined after capital tax cuts in the 1990s and 2000s.”
2. This I find depressing, but not surprising. I am enough of a materialist utilitarian to believe that ex-ante skepticism about the value of these tax refunds does not translate into ex-post lower utility. But it does mean that these tax refunds are not perceived to be as large of a benefit as they should be, and thus have a difficult time attaining political economy salience. Read Sydnee Caldwell, Scott Nelson, and Daniel Waldinger, “Tax Refund Uncertainty: Evidence and Welfare Implications,” in which they write: “Transfers paid through annual tax refunds are a large but uncertain source of income for poor households. We document that low-income tax-filers have substantial subjective uncertainty about these refunds. We investigate the determinants and consequences of refund uncertainty by linking survey, tax, and credit bureau data. On average, filers’ expectations track realized refunds. More uncertain filers have larger differences between expected and realized refunds. Filers borrow in anticipation of their refunds, but more uncertain filers borrow less, consistent with precautionary behavior. A simple consumption-savings model suggests that refund uncertainty reduces the welfare benefits of the EITC by about 10 percent.”
Worthy reads not from Equitable Growth:
1. it is so nice to see smart, intelligent, rational, things from the White House Council of Economic Advisers. Read Cecilia Rouse, Martha Gimbel, Ernie Tedeschi, Evan Soltas, and Ryan Cummings. “Distinguishing Between Signal and Noise in Recent Jobs Data,” in which they write: ‘The onset of the COVID–19 pandemic resulted in historically large job losses, and its retreat this spring has allowed millions of Americans to return to work. On average since January 2021, the U.S. has added about 500,000 new jobs per month. However, job growth has also been noticeably volatile month-to-month since January. This volatility in job growth during the pandemic reflects both real volatility—economic reverberations of the pandemic shock—as well as heightened measurement error due to the challenge of collecting statistical data amidst a pandemic. These considerations warn against placing too much weight on any single data point in assessing the current state of the economy, even as the worst of the pandemic in the U.S. fades away.”
2. Here is Phil Lubin, one of the managerial and entrepreneurial forces behind Evernote and the video presentation program mmhmm. He’s talking his book here. But it is a very good book to talk. We are still trying to figure out what our stuffing a decade of virtual and hybrid managerial business-model experimentation into a year will do to the pattern of future economic growth. It will certainly do something. And we need to find this out pronto. Read his” Lessons for CEOs,” in which he writes: “Imagine a parallel universe where we never had to commute before, and I came in and said, “Listen up guys, I need you to waste two hours of your day sitting in traffic. No, you won’t get any work done, no, it won’t be time spent with your friends or family, and yes, it’s terrible for the environment, but two hours a day, you have to sit in traffic.” Essentially that’s what companies are asking employees to do by telling them to come back to the office. So rather than saying that we’re only doing this because we have to, think about the amazing new powers we have, and how much better our lives are if we embrace them.”
This is a post we publish each Friday with links to articles that touch on economic inequality and growth. The first section is a round-up of what Equitable Growth published this week and the second is relevant and interesting articles we’re highlighting from elsewhere. We won’t be the first to share these articles, but we hope by taking a look back at the whole week, we can put them in context.
Equitable Growth round-up
A new issue brief measures the economic costs of racial, ethnic, and gender inequities in the United States, providing estimates of the benefits of eliminating these disparities and achieving equal opportunity for all. In the brief, author Robert Lynch quantifies the significant economic costs of failing to address not only overt racism and sexism, but also discrimination in such areas as access to jobs, education, credit and loans, high-quality pre-Kindergarten, and healthcare, disparities in the justice system, disproportionate exposure to pollutants, and uneven access to quality physical infrastructure. These disparities, he continues, place heavier burdens on women and people of color. Lynch calculates the impact that equal opportunity would have had in 2019 in terms of total average earnings, Gross Domestic Product, tax revenues, the poverty rate, and Social Security solvency. While he finds significant benefits for people of color and women in eliminating U.S. racial and gender disparities, Lynch also details the gains for the economy overall, which, he notes, are even conservative estimates considering that he did not assess the benefits for non-Hispanic White males in his calculations. It might be difficult to actually realize a society that is free from disparities, he concludes, but the results would be “well worth the effort.”
Economists and economic policymakers have long debated the relationship between taxation and economic growth. These arguments sometimes rely on abstract theoretical models to prove the value of a free market economy, but empirical studies using real world data typically don’t find that increasing taxes is bad for economic growth. In an issue brief, Corey Husak examines whether recent U.S. economic history can provide evidence linking economic growth and fluctuations in the tax code—namely, in top individual income tax rates, corporate tax rates, and capital taxation. He finds that tax changes can have large effects on the U.S. economy, but these effects do not include impacts on overall economic growth or corporate investment. Husak also finds a clear correlation between lower top marginal tax rates and growing income and wealth inequality in the United States. He concludes by recommending that policymakers focus on the real, meaningful, and measurable effects of tax policy—using revenue analysis and distribution tables—rather than on economic growth as an outcome.
As policymakers in Congress debate the merits of reforming or eliminating the filibuster, new research looks into the economic and policy consequences of retaining the Senate rule as it currently stands. Nathan J. Kelly explains that by making any kind of policy change more difficult, the filibuster contributes to growing income and wealth inequality in the United States and leads to a highly unequal economy. This, he continues, is because the filibuster enhances status quo bias, which entrenches divides between the haves and have-nots in the U.S. economy by making government intervention to reduce these divides less likely. Kelly details several recent economic policies that have run up against the filibuster, from increasing the minimum wage and indexing it to inflation to implementing financial market regulations and policies centered on worker power. He then dives into a recent study he and his co-authors released that examines the distributional effects of both partisan polarization and legislative inaction, finding that the latter has a bigger impact on distributional economic outcomes. This means, Kelly concludes, that the longer the filibuster remains part of the U.S. policymaking process, the longer it will take to close income and wealth divides in the United States.
Today, the Bureau of Labor Statistics released its monthly employment report, providing data on the labor market and unemployment in the U.S. economy in June 2021. The data show that, among other things, the economy added 850,000 jobs last month, a higher-than-expected amount. Kathryn Zickuhr and Austin Clemens put together five graphs that highlight the trends in the data, while Zickuhr and Carmen Sanchez Cumming go into more detail in a column about what this Jobs Day report means for foreign-born workers in the United States. These workers already face disparities in job losses, compared to U.S.-born workers, as well as disparities in access to important income support and worker protection programs such as Unemployment Insurance and the Supplemental Nutrition Assistance Program. This, coupled with the fact that foreign-born workers are more likely to be employed in hard-hit industries during the coronavirus pandemic and recession, means foreign-born workers are disproportionately vulnerable in this economic downturn amid the ongoing public health crisis. Zickuhr and Sanchez Cumming urge policymakers to bolster and expand safety net programs rather than cut them, and to ramp up labor standards enforcement to protect workers.
Links from around the web
While racial gaps in educational attainment have narrowed over the past four decades, the Black-White wage divide hasn’t changed. In 2020, average full-time Black workers earned approximately 20 percent less than their White peers and are far less likely to have a job. The New York Times’ Eduardo Porter asks if the underlying cause of these ongoing divides is either racial bias or a changing economy. After Black Americans made strong gains in closing racial gaps at work between the 1940s and 1970s, these gains have stagnated over the past 40 years—and scholars are looking into why. Considering that Black workers today earn less than their White counterparts even relative to educational attainment, many scholars argue that race is the decisive factor in why disparities in wages and work persist. Others say industrial change, globalization, and automation have affected the economy in ways that disproportionately impacted Black workers. These questions are important because the most urgent task at hand, Porter writes, is to find a way to close the divides—and what’s causing them is key to knowing how to address them.
This summer will likely be one of the strangest ever for the U.S. labor market, writes Bloomberg’s Katia Dmitrieva. As the economy overall seems to be growing at a steady pace, millions of workers remain unwilling or unable to return to their pre-pandemic jobs, particularly in hard-hit sectors such as the retail and food-service industries. There are a number of reasons why that’s the case, whether because they want to pursue higher-paying, more stable opportunities, because they are fearful of catching COVID-19, or because they don’t have good, affordable child care options available to them. Likewise, pauses in immigration and international travel have left many seasonal jobs unfilled. Dmitrieva dives into the responses of many corporations and businesses to these trends and explains why the recovery will likely continue to look very different from previous recoveries in the coming months.
In a recent interview with The Washington Post’s Joe Heim, Mark Rank discusses poverty and the safety net in the United States. Rank explains his research on the myths of poverty in the United States, and the denial that is rampant about how much poverty actually exists in the richest country in the world. Rank also specifically calls out the need for a stronger social safety net, citing the fact that, according to his analysis, 60 percent of Americans will spend at least a year of their life in poverty. He also touches on the media’s role in portraying and reporting on poverty in the United States and how the American mindset and the myth of the American Dream play into how the general public views poverty and how much income support goes to people who are struggling. The conversation closes with his insight into how the coronavirus pandemic and recovery may parallel the Depression and resulting New Deal effort to address poverty in the United States.
June was a month of strong employment gains. According to the U.S. Bureau of Labor Statistics’ latest Employment Situation Summary, the economy added 850,000 jobs between mid-May and mid-June, well above the average month-to-month growth of 540,000 jobs of the previous 3 months. The prime-age employment-to-population ratio, a measure that captures the share of U.S. adults between the ages of 25 and 54 who are employed, increased slightly from 77.1 percent to 77.2 percent.
Amid a strong report, however, the overall unemployment rate actually rose slightly from 5.8 percent in May to 5.9 percent in June, led by reentrants to the labor force and workers leaving their jobs. At 61.4 percent, the labor force participation rate is just 0.2 percentage points above where it was at this time in 2020. Adding 343,000 jobs, the leisure and hospitality sector continued to make big gains in June, but in other industries, net employment declined. Construction lost 7,000 jobs, and financial activities lost 1,000 jobs.
In addition, job gains continue to be greatly uneven along the lines of race, ethnicity, and gender. For Latina women, employment has shrunk 5.9 percent since the onset of the coronavirus recession. In June, 763,000 fewer Black women were employed than in February 2020—a 7.2 percent decline with respect to its pre-pandemic level. (See Figure 1.)
Figure 1
Another group that has been hit hard during the coronavirus recession is foreign-born workers in the United States. Between February 2020 and June 2021, the number of foreign-born workers employed declined 5.9 percent, compared to a 3.1 percent fall for U.S.-born workers. For both foreign- and U.S.-born workers, the decline in employment has been much deeper for women. (See Figure 2.)
Figure 2
Data from the U.S. Bureau of Labor Statistics also show that a substantial share of the decline in the U.S. labor force from 2019 to 2020 can be explained by the large number of foreign-born workers who dropped out of the labor force over that period. Strikingly, while workers born abroad only make up 17 percent of the employed or unemployed adults in the United States, they account for almost 40 percent of the overall drop in the U.S. labor force in that time period. In other words, foreign-born workers represented 1.1 million of the 2.8 million workers who stopped participating in the labor force from 2019 to 2020.
What is driving these disparate labor market outcomes? An analysis by Rakesh Kochhar and Jeffrey Passel at the Pew Research Center found that while 42 percent of U.S.-born workers held jobs which could be done remotely in February 2020, only 31 percent of foreign-born workers had the ability to telework.
Foreign-born workers are also more likely than their U.S.-born counterparts to work in some of the occupations that were most affected as the health and economic crises sent ripples through the U.S. economy in 2020. For instance, more than 20 percent of workers born outside of the United States were employed in service occupations in 2020—jobs in food preparation and serving, for example—compared with just more than 14.4 percent of native-born workers. (See Figure 3.)
Figure 3
In addition, foreign-born workers are also overrepresented among the essential workforce. Research by Donald Kerwin and Robert Warren at the Center for Migration Studies, for example, estimates that while 65 percent of native-born workers hold jobs categorized as essential by the Department of Homeland Security—jobs in sectors such as meatpacking and poultry processing, agriculture, construction, and healthcare—69 percent of all foreign-born workers do. For undocumented workers, that number climbs to 74 percent.
Foreign-born workers also make up a disproportionately large share of the workforce in sectors in which workers are most at-risk of contracting the coronavirus—food or agriculture has been the deadliest sector, according to a study by researchers at the University of California, San Francisco. Alarmingly, workers in food manufacturing, agriculture, and other essential occupations are also among the most likely to still be unvaccinated.
As such, the fact that foreign-born workers also face barriers to accessing many worker protections and income supports that are most needed during economic downturns—not to mention amid a global pandemic—leaves them particularly exposed to economic insecurity, as well as to the risk of getting sick. Under the current Unemployment Insurance system, for instance, immigrant workers not authorized to work in the United States are not eligible for unemployment benefits.
Immigrant workers are also less likely to apply for assistance programs even when they are eligible for them. A recent analysis by the Urban Institute finds that immigrant adults and families eligible for income support and social insurance programs, such as Medicaid and the Supplemental Nutrition Assistance Program, often decide not participate out of concern that doing so will hurt a green card application process or that information-sharing will lead to the deportation of an undocumented family member.
Even as the U.S. labor market is on track to a robust and relatively quick recovery from the coronavirus recession, there are still 6.8 million fewer jobs than prior to the pandemic. Workers who have been historically excluded from labor market opportunities and protections are at risk of being left behind in the recovery. This not only diminishes the economic security of vulnerable workers, but it translates into a less competitive labor market, lost productivity, and a drag on economic growth.
More than 9.5 million workers actively looking for a job in the United States do not have one. Despite the fact that many workers and families are still struggling, as of the release of this column, 26 states have decided to slash federally funded Unemployment Insurance benefits prior to their official expiration date in early September. Rather than cutting these vital benefits, policymakers should enact permanent reforms that expand UI benefit duration, level, and eligibility, including to freelance workers, seasonal and temporary workers, gig workers, undocumented workers, and workers who do not have a recent work history.
Policymakers should also ramp up the enforcement of labor standards. Research shows that violations to labor law such as wage theft are prevalent in sectors such as agriculture, and that workers who are not U.S. citizens—and Latina noncitizen women in particular—are especially vulnerable to these violations.
Perhaps most urgently, the Occupational Safety and Health Administration—which, at the onset of the pandemic, had the lowest number of workplace inspectors in more than four decades—should be strengthened to guarantee workers are protected during the ongoing pandemic. Ensuring that all workers are protected and have access to the income supports they need is essential for a strong and equitable recovery.
On July 2, the U.S. Bureau of Labor Statistics released new data on the U.S. labor market during the month of June. Below are five graphs compiled by Equitable Growth staff highlighting important trends in the data.
The prime-age employment-to-population ratio remained steady in June, while the labor market added 850,000 jobs.
Top-line unemployment, also known as U-3, and a broader measure of labor underutilization, known as U-6, remain low in June, falling below their levels prior to the Great Recession.
Unemployment rates by race and ethnicity saw little change and remain highest for Black workers at 9.2 percent and Latinx workers at 7.4 percent.
The unemployment rate remains highest for workers with less education and increased slightly to 10.2 percent for workers with less than a high school degree.
Employment grew for some of the industries hit hard by the coronavirus recession, including leisure and hospitality, but still remains below pre-pandemic levels.
The relationship between taxation and economic growth is hotly debated in economics. Free market economic ideology is based on the premise that constraining “the market” through policies such as increased taxes is bad for economic growth.1 Yet the economy is not perfectly represented by abstract theoretical models. Empirical studies based on real-world data are often unable to find these negative growth effects. Because neoclassical economic models consistently fail to accurately predict economic growth patterns, policymakers need to rethink using them to analyze tax changes.
Researchers have studied many different fluctuations in U.S. tax policy over the past several decades. In 1997, for example, the United States had much higher average tax rates and especially much higher rates of taxation on wealth and capital than it does today.2 Recent years have seen the passage of a major tax cut in the 2017 Tax Cuts and Jobs Act and new proposals from the Biden administration to raise revenue from the richest Americans.3
This issue brief examines if, in recent U.S. economic history, there is empirical evidence linking economic growth and:
The broad tax regime
Top individual income tax rates
Corporate tax rates
Capital taxation
U.S. income and wealth inequality
In light of the evidence surveyed, the brief closes by discussing the proper way for policymakers to judge and evaluate tax proposals. Because classical models lack strong explanatory power in practice, U.S. policymakers should not focus on economic growth as an outcome in tax policy formulation. Rather, they should focus on the actual, meaningful, and measurable effects of tax changes, such as their impact on government revenues and the distribution of income.
The broad relationship between tax rates and U.S. economic growth
Supply side and neoclassical models rose to prominence in the 1970s and 1980s, promising faster economic growth. Unlike the Keynesian model used in the mid-20th century, which had very little to say about taxes, supply-side economics and other neoclassical models champion a “free market”-centered view that frames taxes as the enemy of economic growth.4
In the 1980s, policymakers followed these models and drastically lowered the top individual marginal income tax rate, from about 70 percent down to 28 percent, and lowered the top corporate rate from 46 percent down to 34 percent.5 But instead of booming, income growth slowed. As the government reduced statutory tax rates, especially for those at the top, inequality exploded, and income growth rates went down. (See Figure 1.)
Figure 1
Drawing definitive conclusions from these types of correlations is not possible because there is no counterfactual. But it is clear that the U.S. economy grew more slowly after top rates were drastically lowered than it had grown previously.
The Congressional Research Service also finds that broad empirical data are the opposite of what neoclassical models predict. The U.S. national savings rate, for example, declined in the 1980s after taxes on capital dropped, and it again declined after capital tax cuts in the 1990s and 2000s; neoclassical models predict the opposite would happen.
Additionally, U.S. labor supply, measured by the number of hours worked, has broadly declined as top personal income tax rates have declined; neoclassical models would likewise predict the opposite. Again, there is no counterfactual, but even subcomponents of growth—in these cases, savings rates and labor supply—have behaved the opposite way free market, or supply-side, economists would predict.6
In addition, Chye-Ching Huang of New York University and Nathaniel Frentz of the Congressional Budget Office produce a review of dozens of peer-reviewed studies on the relationship between taxation and economic growth in 2014 and find that the academy was highly conflicted. “Taking all of these studies into account, there is simply no consensus that, as a general proposition, cutting taxes is a good strategy to boost economic growth,” they report.7 More recent evidence has not changed this conclusion, as detailed below.
Top individual income tax rate and U.S. economic growth
The current top marginal income tax rate is 37 percent, or up to 40.8 percent for some taxpayers when combined with two Medicare surtaxes. The Biden administration offered proposals to raise the top rate back to 39.6 percent, where it was for most of the 1990s and 2010s.8 Analysts who rely on neoclassical models argue that there are large trade-offs between having a strongly progressive tax system and economic growth.9 But this theoretical trade-off is not apparent in the economic literature.10
A tax and economic growth trade-off also is not present in the data. Over time, there has been no obvious relationship between U.S. economic growth and the top marginal rate applied to individuals’ regular income. (See Figure 2.)
Figure 2
In fact, over time, high top rates are correlated with higher economic growth for most Americans, according to research from University of California, Berkeley economist Emanuel Saez. His research on the effects of the Obama administration’s 2013 tax increases on individual taxpayers making more than $250,000 per year concludes that they were efficient at raising revenue and “the top tax rate increases of 1993 and 2013 do not seem to have hurt overall economic growth, quite the contrary.”11
Saez finds that “the best growth years for the bottom 99% incomes since 1990 have taken place in the mid to late 1990s and since 2013, shortly after increases in top tax rates.” On individual rates, the empirical pattern has been the opposite of what neoclassical models predict, casting doubt on these models’ ability to say anything meaningful about growth following tax increases on the rich.
Corporate rate cuts have not boosted U.S. economic growth
The biggest recent test case for the theory that tax rates have strong effects on economic growth was the 2017 Tax Cuts and Jobs Act. Among other changes, this law lowered the corporate tax rate from 35 percent to 21 percent. At the time, the Trump administration’s Council of Economic Advisers claimed that “reductions in effective corporate tax rates have substantial, positive short- and long-run effects on output,” primarily by increasing “firms’ investment, desired capital stock, and potential output,” which would lead to “wage increases for U.S. households of $4,000 or more” and “much of this boost to U.S. output may be apparent in the near term.”12
As manyanalysts confirm, these predictions did not occur. Steve Rosenthal at the Tax Policy Center reports that even a full 2 years after passage, the United States was “without resulting investment or wage growth, or even green shoots.”13
The 2017 tax cuts’ lack of effect on wages in its first 2 years surprised few analysts, but the law’s lack of effect on corporate investment was particularly striking. While investment can be volatile, the Congressional Research Service notes that the biggest bump in investment since its passage occurred in the first half of 2018, too early to be the plausible result of a tax change just months before because investment decisions take time to plan and execute.14
Moreover, the Congressional Research Service finds that investment increases (such as they were) were in subcategories that didn’t correspond to the provisions of the 2017 Tax Cuts and Jobs Act. For instance, intellectual property investment grew fastest in 2018, even though the law paradoxically raised the user cost of investing in intellectual property, with factors other than tax cuts driving those changes.15
Indeed, there was strong growth in fixed investment in the year before the 2017 tax cuts passed, which carried over into 2018 before stagnating in 2019. This is despite the long and sustained fall in tax revenue, which began dropping in 2017 as corporations used accounting techniques to ensure losses would start appearing the year before the new rates hit to take full advantage of the cut. (See Figure 3.)
Figure 3
So, what did corporations spend their large tax cut on, if not wages or investment? Analysts at the International Monetary Fund find that 80 percent of the corporate tax cuts were repurposed into stock buybacks and dividends, which overwhelmingly benefited wealthy shareholders.16 And Lenore Palladino of the University of Massachusetts Amherst documents that these corporate buybacks and dividends also widened the racial wealth divide, finding that White stock-owners hold $27 for every $1 in corporate equity and mutual fund value held by a Black or Hispanic stock-owner.17
The main effects of the Tax Cuts and Jobs Act were less government revenue and regressive tax cuts for corporations, wealthy shareholders, and executives who bear nearly all the burden of corporate taxes even as the rate changes had little effect on business investment or workers.18
Capital taxation and U.S. economic growth
In the late 1990s and early 2000s, the United States significantly reduced the taxation of capital. For instance, the top capital gains rate was lowered from 28 percent in 1997 to 15 percent in 2003, before being raised back to 20 percent, plus a 3.8 percent Medicare surcharge, in the early 2010s. Meanwhile, taxes on dividends were reduced from nearly 40 percent down to the capital gains rate of 15 percent in 2003.
There are often confounding factors that make it difficult for data to precisely show what effect a tax change has on the economy, but occasionally, natural experiments do arise. The 2003 dividend rate cut provides an ideal test case to measure the effect of a capital tax cut on growth. Equitable Growth grantee and UC Berkeley economist Danny Yagan (now at the White House Office of Management and Budget) compares the behavior of C-corporations, which experienced the dividend tax rate drop from nearly 40 percent to 15 percent, to other, similar businesses (S-corporations) that were not affected by the tax cut to demonstrate why only shareholders gained from those dividend cuts.19 (See Figure 4.)
Figure 4
Similar to the pattern after the 2017 tax cuts passed, Yagan finds that the main effect of the dividend tax cut was that C-corporations increased payments to their shareholders. C-corporations did not increase employee compensation or investment after they received their tax cut, when compared to unaffected S-corporations. Contrary to claims from the law’s proponents and neoclassical economic models, cutting dividend taxes more than in half did not boost economic growth but did reduce government revenue and increase inequality.
U.S. income and wealth inequality and growth
Cutting taxes for the owners of businesses and other forms of capital rarely had perceptible effects on investment or economic growth, but these cuts still have economic effects. Because capital income is so unequally distributed, lowering business and investment tax rates creates an upward income redistribution, enriching the already-wealthy. (See Figure 5.)
Figure 5
While there is no discernable positive correlation between upper-income tax cuts and U.S. economic growth, there is a clear correlation between these tax cuts and income inequality. When the United States began lowering top marginal income rates and taxes on capital income, such as investments, corporations, and other businesses, the income of the richest 1 percent of U.S. families grew. The 1980s were a period with many economic and policy changes besides tax cuts for the wealthy, so tax cuts cannot be said to be completely responsible for this trend, though they contributed. (See Figure 6.)
Figure 6
Low taxes on accumulated wealth also help the already-wealthy and powerful maintain and grow their privilege over other Americans. Wealth inequality is growing.20 Refusing to tax these gains helps the beneficiaries of past policy choices maintain their economic and social power, even when past wealth was gained in a context of racist and sexist economic structures.21 (See Figure 7.)
Figure 7
In general, neoclassical economic models have neglected economic inequality as a driver of slowing growth, decreasing dynamism, and stagnating well-being over the past several decades, argues Heather Boushey in her most recent book Unbound: How Inequality Constricts Our Economy and What We can Do About It. Boushey finds that lowering taxes at the top helped to fuel the rise of inequality, which has obstructed, subverted, and distorted the pathways to broadly shared growth.
Post-1980, economic policymaking focused more on allowing the already-wealthy to keep more and more gains, meaning this wealth has not “trickled down” or been reinvested in ways that build growth for middle- and lower-income Americans. This also starved the public sector of resources to build structures that lift up those who have been systemically disadvantaged historically. The result is that public investment has fallen, and the top 1 percent benefits disproportionately from the slower economic growth the United States is still experiencing.22
The economy is complex, and there are many factors that determine growth and well-being that short-run neoclassical models fail to capture. For instance, the returns from many investments in children and families are so high that they outweigh the plausible economic benefits claimed by tax cut boosters, says the former Vice Chair of the Federal Reserve Board Alan Blinder.23 This is the crucial part of analyzing a tax that many models undervalue: the benefits from the spending that taxes enable.
To the extent policymakers judge economic policy on its “growth” effects, they should consider who benefits from the growth that occurs.24
How policymakers should judge tax changes
So, how should policymakers judge proposed taxes if the data on tax changes rarely reflect classical analyses’ predictions? The answer, as former Equitable Growth Tax Policy Director Greg Leiserson writes, is “if U.S. tax reform delivers equitable growth, a distribution table will show it.”25 A revenue analysis explains how much money will be collected or lost, and a distribution analysis explains who will pay for or benefit from the tax. Numerous complex calculations and assumptions underlie the production of revenue and distribution tables, but when done well, they provide the key information that policymakers need to know about how tax changes will impact the populations they care about.
Analyses of taxes’ effects on economic growth, while often sparking interesting academic debates, do not tell policymakers anything they cannot learn by looking at a revenue analysis and a distribution table produced by rigorous, nonpartisan groups such as the U.S. Congress’ Joint Committee on Taxation and the Tax Policy Center.26
Conclusion
This issue brief shows that while tax changes can have large effects on the U.S. economy, they have not noticeably affected overall economic growth or corporate investment in recent decades. Instead, the research and data firmly establish that the main effects of tax changes are to increase or decrease inequality and government revenue.
Racial, ethnic, and gender disparities across the U.S. economy are a personal tragedy for the millions of individuals who suffer from the emotional, psychological, and social consequences. These disparities also are an economic calamity for communities of color, for women, and indeed for all U.S. workers, their families, and their communities, resulting in a United States that is far off from its promise of equal opportunity for all.
In this issue brief, the economic costs of racial, ethnic, and gender inequities are quantified and illuminated by providing estimates of the economic benefits of eliminating them. This analysis imagines an America free of racial, ethnic, and gender disparities, where one’s skin color, ethnic origin, or gender are no obstacle to worker productivity, labor force participation, and advancement—in short, a United States of America where everyone has the same opportunity to achieve their potential.
Specifically, this analysis examines the benefits that would have been realized in 2019 of closing these racial, ethnic, and gender divides by estimating the resulting:
Increases in the average total earnings of women and Black, Hispanic, and Asian workers
Growth in Gross Domestic Product
Rises in tax revenues
Reductions in poverty
Improvements in the solvency of Social Security
To achieve an equal-opportunity society where there are no barriers to economic success that are the consequence of race, ethnicity, or gender disparities would obviously require tackling overt racism and sexism.
But true equal opportunity requires more than that. It involves addressing many other disparities that undermine worker productivity, such as unequal access to jobs, education, credit in the form of home mortgages and business loans, affordable child care, high-quality pre-Kindergarten, and senior care; inconsistencies in family leave and workplace policies and in treatment in the justice system; imbalances in access to healthcare and insurance; unequal exposure to damaging levels of environmental pollutants; and inadequate and unequal access to high-quality physical infrastructure, which is essential to maximizing economic efficiency.
All of these disparities, and more, impose greater burdens on women and racial and ethnic workers and their families and communities. And all of these and other inequities undermine the productivity of women and people of color, contribute to inequalities in the distribution of income and wealth, and weaken the national economy.
This study clarifies the significant economic costs of failure to address these inequalities and of not creating equal opportunity. The specific findings are detailed below, but overall, if equal opportunity were the reality, then in 2019:
Total average earnings of all workers would have been 33.5 percent, or $18,134, higher ($72,263 versus $54,129).
Collectively, Black, Hispanic, and Asian workers would have earned 60.6 percent more ($71,472 versus $44,495).
U.S. Gross Domestic Product would have been $7.2 trillion higher and totaled $28.6 trillion instead of $21.4 trillion.
Federal, state, and local tax revenues would have been $1.82 trillion higher.
The overall U.S. poverty rate would have dropped from 10.5 percent to 6.6 percent, lifting 12.2 million people out of poverty.
There would have been a $429 billion improvement in the finances of the U.S. Social Security system.
Although the economic benefits of leveling the playing field identified in this study accrue disproportionately to women and people of color, many policies to achieve equal opportunity would benefit non-Hispanic White males as well. Improvements to the solvency of Social Security, for example, would benefit all people, and enhancements to physical infrastructure would increase the productivity and earnings of all workers.
Yet the magnitudes of these benefits to non-Hispanic White males are not assessed in this analysis. As such, the economic benefits identified are an underestimate of the total economic benefits. Despite the underestimate, these findings demonstrate that enacting economic and social policies that create a more equal-opportunity United States would deliver enormous benefits to all workers, not just to women and people of color, and would support more equitable and sustainable economic growth for decades to come.
U.S. earnings
If equal opportunity prevailed in the United States in 2019, then the average total money income—a reflection of the nation’s prosperity—would have been significantly greater, and racial, ethnic, and gender earnings differences would have largely disappeared. The average total money income of all workers would have been 33.5 percent, or $18,134, higher: $72,263 versus $54,129.
The breakdown by race, ethnicity, and gender is also telling. The average earnings of non-Hispanic White workers would have been 22.8 percent higher ($72,897 versus $59,376) because of the significant increase in the earnings of White women. Black workers would have earned 73.7 percent more ($71,611 versus $41,222). Hispanic workers would have earned 83.4 percent more ($70,578 versus $38,485). And Asian workers would have earned 10.2 percent more ($73,502 versus $66,679). (See Figure 1.)
Figure 1
Collectively, Black, Hispanic, and Asian workers would have earned 60.6 percent more ($71,472 versus $44,495), but the gains would have accrued especially to women of color. In particular, men of color would have earned 39.6 percent more ($71,122 versus $50,962), while women of color would have earned 88.2 percent more ($71,805 versus $38,150). Non-Hispanic White women would also have experienced large gains: Their earning would have been 58.2 percent higher, with average total money income of $72,929 instead of $46,100. The earnings of all these groups of workers would have closely matched the $72,863 in earnings of non-Hispanic White male workers, and earnings gaps would have largely disappeared.
The increase in earnings across all workers in the United States in 2019 would have totaled $4.3 trillion. Total money income would have risen from $12.7 trillion to $17 trillion in 2019. Put simply, earnings would have risen substantially if equal opportunity were the defining feature of the U.S. economy and society, with far more equitable earnings for women and workers of color.
U.S. Gross Domestic Product and tax revenues
If equal opportunity prevailed in the United States in 2019, then U.S. Gross Domestic Product—the most commonly used measure of the size of the economy—would have been $7.2 trillion higher and totaled $28.6 trillion instead of $21.4 trillion. (See Figure 2.) To get a sense of the magnitude of this $7.2 trillion increase, consider that at the actual 2.2 percent rate of real (inflation-adjusted) growth that the U.S. economy experienced in 2019, it would take more than 13 additional years to reach a GDP of $28.6 trillion.
Figure 2
The larger GDP would have generated a considerable increase in annual tax revenues in 2019. The elimination of racial, ethnic, and gender disparities would have increased federal, state, and local tax revenues by $1.82 trillion in 2019—$1.184 trillion for the federal government and $636 billion for state and local governments.27
The positive impact on government budgets would have been even greater than these tax revenue increases if reductions in means-tested public services and spending were taken into account. With the higher estimated earnings, many fewer people would have needed federal, state, and local food aid, housing assistance, cash assistance, and other support services.
U.S. poverty rates and number of poor people
If equal opportunity prevailed in the United States and disparities in income earnings were largely eliminated, then the overall poverty rate would have dropped from 10.5 percent to 6.6 percent in 2019. The poverty rate for non-Hispanic White people in the country would have fallen to 5.7 percent (from 7.3 percent) due to the earnings growth of White women. Across the country, the poverty rate for Black people would have dropped to 9.8 percent (from 18.8 percent); the poverty rate for Hispanic people would have fallen to 7.2 percent (from 15.7 percent); and the poverty rate for Asian people in the country would have declined to 6.7 percent (from 7.3 percent). (See Figure 3.)
Figure 3
Compare these poverty levels to those actually registered in 2019. That year, 34 million people were living in poverty, as defined by the U.S. Census Bureau.28 This translates into 14.2 million non-Hispanic White people and 19.1 million people of color who lived below the federal poverty threshold in 2019.29 The poverty rate for males was 9.4 percent and it was 11.5 percent for females, representing 15 million men and 19 million women.
If race, ethnicity, and gender-based income inequality were largely eliminated in 2019, then this would have had the effect of lifting 12.2 million people out of poverty across the country, including 3 million non-Hispanic White women, 3.9 million Black people, 5.2 million Hispanic people, and 119,000 Asian people. The largest beneficiaries would have been women in general, and women of color in particular. Of the 12.2 million reduction in the total number of people living in poverty, more than three-quarters (9.3 million) would have been women, and roughly half (6.3 million) would have been women of color.
The solvency of the U.S. Social Security system
The effects on the long-run solvency of the U.S. Social Security system from eliminating racial, ethnic, and gender disparities are multifaceted. The projected rise in earnings would have increased the wages subject to Social Security payroll taxes, and thereby raised revenues for the system. In addition, since only the earnings below a maximum ($132,900 in 2019) are subject to Social Security payroll taxes, wage growth that was concentrated among lower-earning communities of color and women would have generated higher tax receipts than wage growth that was more evenly distributed.
In contrast, the higher earnings would have also increased the future benefits paid out by the program because initial Social Security benefits are indexed to wages. This means that greater payroll tax revenues would have been partially offset by larger future benefit payments.
On balance, the impact on the financial solvency of Social Security would have been positive and boosted the solvency of the system for decades to come. For example, the productivity and GDP increases calculated in this analysis are of such a magnitude that they would have generated significant additional Social Security tax revenue. The $7.2 trillion increase in GDP in 2019 would have increased Social Security tax revenues in 2019 by $429 billion, adding to the small $2 billion surplus in total revenues relative to outlays that year.30 The extra $429 billion in tax revenues would, in turn, have contributed to the long-run solvency of the system.
Conclusion
The U.S. economic consequences of eliminating racial, ethnic, and gender inequalities are large, particularly for individuals of color and women. The total earnings in the nation would swell well beyond what they currently are, and the incomes of Black, Hispanic, and Asian workers would be substantially higher. The largest earnings gains would be experienced by non-Hispanic White, Black, Hispanic, and Asian women.
There would be a substantial concomitant rise in GDP and dramatic declines in poverty rates overall, within communities of color and for women in particular. The nation as a whole would benefit from an enormous increase in tax revenue and an improvement in the solvency of the Social Security system.
It is important to emphasize that it is not just people of color and women who would benefit from the creation of an equal-opportunity society. Addressing the inequalities that undermine a level playing field would entail policies that benefit all members of society. Investing in high-quality pre-Kindergarten programs and in the nation’s physical infrastructure, improving the K-12 educational system, providing adequate healthcare and insurance, removing lead from drinking water, strengthening the solvency of the Social Security system, and increasing access to high-quality senior care would raise the productivity, earnings, and economic well-being of people of all races, ethnicities, and genders, including non-Hispanic White males. Since these benefits to non-Hispanic White males were not estimated, the economic benefits of equality of opportunity identified in this study are understated.
It is also important to note that the proportion of people of color in the total U.S. population is anticipated to grow from less than 40 percent in 2019 to more than half by 2045.31 This means the future positive impact of eliminating inequalities on earnings, GDP, tax revenues, poverty rates, and the solvency of Social Security would very likely be substantially larger than that estimated for 2019 in this analysis. The sooner policymakers address these disparities, the sooner we will be able to enjoy the benefits of moderating them.
A large body of research documents disparities in earnings by race, ethnicity, and gender that are not a function of ability, effort, or aptitude.32 This means talent and skill are not being fully recognized and put to their greatest use, let alone fairly rewarded. This results in a misallocation of our nation’s most precious resource—its people—and undermines our long-run economic growth and competitiveness.33 The economic costs of these disparities in economic opportunity are enormous, as this issue brief illustrates. Thus, although it may be difficult to realize an equal-opportunity society, the results for all of us are well worth the effort.
Following the practice of the U.S. Census Bureau, this analysis uses non-Hispanic White males as the comparison group for other race groups, Hispanics, and women in the country.34 The analysis then calculates what earnings in the United States would have looked like in 2019 if all racial, ethnic, and gender groups earned the same average “total money income,” adjusted for age differences.35
Total money income is defined as all income from wages and other income sources, such as net income from self-employment and government income supports. Total money income is calculated by analyzing the responses to the Annual Social and Economic Supplement to the Current Population Survey carried out by the U.S. Census Bureau and the U.S. Bureau of Labor Statistics. The Current Population Survey is the most highly regarded source of information on earnings and labor force statistics for the population of the United States.
Accordingly, the hypothetical earnings in the United States in 2019 are estimated assuming that Black, Hispanic, and Asian men and women and White women earned the same average total money income as non-Hispanic White males did in 2019. This is done by assigning various racial and ethnic groups and women the same mean (average) total money income and percentage of income earners, adjusted for age, as that of non-Hispanic White males, as determined by the Current Population Survey of 2020, which estimated income in the United States for 2019.
This methodological approach implies that productivity and labor force participation are equalized across workers of all races, ethnicities, and genders. Average total money income across racial/ethnic groups and women in our hypothetical, disparity-free United States continues to vary somewhat across groups in this model because the age distribution varies by group. Hispanic workers for example, continue to have a lower average total money income than other groups because they are younger on average and not as advanced in their careers, and are thus lower down the salary ladder.
A final methodological note. As is standard in economic research that explores the consequences of narrowing racial, ethnic, and gender-based earnings divides, the model used in this analysis assumes that the relatively higher earnings of non-Hispanic White males in 2019 were a function of their higher productivity and labor force participation. If, instead, the higher earnings of non-Hispanic White males partly reflected their power to extract earnings from economic activity that exceeded the value of their current contributions to production—perhaps as a function of systemic racism and sexism or other factors—then the model overstated the economic benefits of leveling the playing field.
As race, ethnicity, and gender barriers to full participation in the economy are overcome, people of color and women would produce and earn more and grow the economy. Yet as obstacles to equal opportunity crumble, non-Hispanic White males may find that their capacity to extract earnings in excess of their productive contributions may evaporate, and they may earn somewhat less than at present, dampening the overall economic benefits identified in this analysis. This dampening effect would likely be most evident at the highest ends of the income distribution, where the racial, ethnic, and gender-based earnings divides are most pronounced. The resulting decline in the estimated total economic benefits would occur not because non-Hispanic White men would be less productive, but rather because the productivity of people of color and women would match a more accurate but somewhat lower level of actual non-Hispanic White male output than that assumed in this study.
Comparisons to other recent research on the effects of creating equal opportunity
The findings of this analysis are consistent with and similar to those of other recent studies. A 2021 working paper published by the Federal Reserve Bank of San Francisco concluded that closing racial and gender earnings divides would have increased total “earnings” in 2019 by 37.6 percent, or by $2.62 trillion from $6.97 trillion to $9.6 trillion.36 By comparison, the analysis conducted in this study concludes that earnings in 2019 would have increased 33.5 percent, or by $4.3 trillion from $12.7 trillion to $17 trillion.
The differences in these findings are mostly attributable to two factors. First, the Federal Reserve Bank of San Francisco working paper used a narrower definition of earnings—Average Hourly Earnings (those from wages and salaries)—than did this study, which analyzed changes in Total Money Income (which includes earnings from wages and salaries, as well as other forms of income such as self-employment income). Second, the San Francisco Fed’s working paper considered only earnings from people aged 25 to 64, while this study considered earnings from a broader group, people aged 15 and older.
That working paper also assumed that physical capital (man-made goods such as machinery, equipment, tools, and factories that assist in the production of products and services) would remain fixed even as earnings grew, whereas this study assumed that physical capital would grow in tandem with earnings growth. This difference in assumptions had a significant impact on estimated GDP growth. It caused the working paper to assume that GDP in 2019 would grow only by the same $2.62 trillion as would earnings in 2019. Thus, their estimated growth rate of GDP is implicitly only 12.2 percent (from $21.43 trillion to $24.05 trillion), compared to their estimated 37.6 percent growth in earnings. By contrast, this study assumed that GDP in 2019 would grow at the same rate as earnings, which is more consistent with the historical relationship between earnings and GDP. Thus, GDP in this study also grew 33.5 percent, or $7.2 trillion, from $21.43 trillion to $28.6 trillion.
A Citi GPS study in 2020 found that if racial divides between Black and White workers alone had been closed 20 years ago, then the U.S. economy would have generated an additional $16 trillion in output, or $800 billion in extra GDP per year.37 The study also projected that GDP would grow by an additional average of $950 billion per year between 2021 and 2025. This closely matches the $877 billion increase in earnings of Black workers in 2019 calculated in this study but is only a little more than half the estimated $1.48 trillion increase in overall GDP in this analysis that is attributable to creating equal opportunity for Black workers. This difference is largely because the Citi GPS report estimates the economic effects of closing only a subset of Black/White racial divides such as those in access to credit, education, healthcare, and housing.
A McKinsey & Company report in 2019 estimated that closing the Black/White racial wealth divide alone would add between $1 trillion and $1.5 trillion in inflation-adjusted GDP in the U.S. economy in 2028.38 This is similar to the $1.48 trillion addition to GDP in 2019 from creating equal opportunity for Black workers estimated in this study. But it is not clear how comparable the results are from these two studies because of the time difference (2028 versus 2019) and the differences in focus. The McKinsey & Company report calculated the GDP effects of closing only the Black/White wealth divide in housing, private consumption, and stock market investments, while this study analyzed the economic effects of creating equal opportunity for Black workers in all areas of the economy.
A W.K. Kellogg Foundation/Altarum study in 2018 calculated that the GDP of the United States would be 22 percent, or $8 trillion, larger in 2050 if the country eliminated racial earnings differences due to disparities in health, education, incarceration, and employment.39 The increase in GDP is smaller than the 33.5 percent calculated in this study largely because the W.K. Kellogg/Altarum study does not include the effect of closing gender-based earnings divides.
Acknowledgments
I would like to thank Marcus Corbin, Michael Ettlinger, Louisa Koch, Guy Molyneux, Charles Call, and Ed Paisley for their comments and suggestions, Ed Paisley and Emilie Openchowski for their excellent edits, and Dave Evans and Austin Clemens for their assistance with graphics.
