The Left's Worst Inequality Bogeyman Attempt Yet
This article originally appeared in Forbes.
Part 4 in the series, “Inequality and the Middle Class.” Also available, Parts 1, 2, and 3.
Over at Vox.com, Matt Yglesias summarizes what he characterizes as a breakthrough in messaging that will finally (finally!) make voters care about inequality. Elizabeth Warren has hit on solid gold! Check it out:
“When the top 10 percent gets 100 percent of the income growth over the course of a generation, then the America of opportunity is vanishing.”
Well, the idea that middle-class living standards haven’t risen much is pretty ubiquitous. The University of Chicago’s business school asks a panel of prominent economists to vote on a range of empirical questions, and a number of them indicated skepticism about rising living standards in the most recent poll.
This is just all sorts of badly wrong. Let’s back up though. Past attempts to beat voters over the head with the inequality cudgel have, Yglesias notes, leaned heavily on the Great Gatsby Curve, which purported to show a link between inequality and economic mobility across countries. As he says, linking to an essay I coauthored with Donald Schneider, it’s not clear that there’s even a correlation between the two, let alone a causal relationship. And while inequality has risen, mobility has remained steady.
But with the Warren framing, upward mobility means (in Yglesias’s words) “to grow up to have a better life than one’s parents.” In mobility research circles, this is referred to as “absolute mobility.” Traditionally, it is conservatives who emphasize its importance over “relative mobility,” which liberals tend to highlight. To understand the difference, imagine two scenarios. In the first one, kids of fast-food workers become fast-food workers and kids of CEOs become CEOs. But everyone ends up richer than their parents. That’s a world with absolute mobility but no relative mobility. Alternatively, we can envision a world in which it is no more difficult to become a CEO if your parents are fast-food workers than if they are CEOs, but where living standards have stagnated. In this scenario, there is relative mobility but no absolute mobility.
Liberals generally care more about relative mobility because it gets at the extent to which family origins limit opportunities. Conservatives typically care more about absolute mobility because they view unequal outcomes as not necessarily reflective of unequal opportunities and are mainly concerned that everyone is becoming better off over time. (For the record, I think both are important, but I’ve put special effort into persuading more conservatives to worry more about relative mobility.)
Warren’s quote suggests that because all income gains are going to the top, there is no upward absolute mobility for anyone else. As I’ve tried to show, it’s simply not true that anything like “all” of the income gains have gone to the top over the past generation (or even over the current recovery). But if Warren and Yglesias were familiar with the mobility literature, they’d also know that absolute mobility looks much stronger in the U.S. than relative mobility does.
There is much confusion–sadly, even among top economists–about the extent to which the middle class and the poor have or have not seen improvements in living standards over the past generation. Official Census Bureau figures, for instance, show that median household income—the income of the household in the middle when everyone is lined up, from poorest to richest—rose by just 15 percent between 1979 and 2007, after taking into account changes in the cost of living.* But a brighter picture emerges after using the best cost-of-living adjustment available, accounting for declines in household size (which make the amount of income per household member larger), including the value of employer-provided health insurance as income, deducting taxes, and focusing on non-Hispanics, to account for immigration. After those adjustments, median household income rose 39 percent between these two business-cycle peaks.**
Even these trends understate the extent to which adults are better off than their parents. Parents have lower size-adjusted incomes than other adults.*** And because it doesn’t follow that the median adult had parents at the earlier generation’s median, the intergenerational change in household income at the median is not the same as the median intergenerational change in household income. When adults’ incomes are compared with their own parents’ incomes at the same age, the median change between 1978 and 2005 was a 93 percent increase, after accounting for declining household size and the rising cost of living. That is, today’s adults typically have incomes twice as large as their parents had at the same age, once their smaller households are taken into account (or $26,000 more, without adjusting for household size). Fully 83 percent of today’s adults are better off than their parents were.****
Don’t believe me? Pew computed their own estimates using a different data set than me and a different span of years. They found 84 percent of today’s adults are better off than their parents were.
Now, progressives can respond like Robert Putnam and assert that regardless of what today’s adults experienced, today’s children will be worse off than their parents. Maybe. But Warren was making a claim about the past generation, which, after all, has experienced rising income inequality. I hope they’ll forgive me if I want to wait and see—and I hope they’ll forgive voters for failing to believe inequality-firstism over their own lying eyes.
* Census Bureau, Historical Income Tables: Households, Table H-5, https://www.census.gov/hhes/www/income/data/historical/household/2013/h05.xls.
** Author’s computations from the Current Population Survey Annual Demographic Supplement. I use the Bureau of Economic Analysis’s Personal Consumption Expenditures (PCE) deflator to adjust for inflation, and I divide incomes by the square root of household size, as is convention. These estimates count cash and non-cash transfers as income, but they do not include the value of Medicare or Medicaid, and non-cash transfers are relatively unimportant for the median household.
*** Author’s computations from the 1979 Current Population Survey Annual Demographic Supplement.
**** Author’s computations from the National Longitudinal Survey of Youth 1979 cohorts, based on a sample of 384 adults. Parental income is measured in 1978, when sample members were between the ages of 13 and 15 and their parents—either the father for male sample members with a father present or the mother for female sample members and male members with no father present—were between the ages of 40 and 42. The income of sample members in adulthood is measured in 2005, when they were between the ages of 40 and 42. Incomes were bottom-coded at $0 and top-coded at the 97th percentile of all family incomes in each year. Incomes are adjusted to constant 2014 dollars using the PCE deflator. The size adjustment divides incomes by the square root of family size.
Both 1978 and 2005 were near-peak years in the business cycle. If sample members’ 2009 incomes are used instead—so that a bad year for them is compared with a good year for their parents, 68 percent of adults were better off than their parents without adjusting for household size, 83 percent were better off after size adjustment, and the median adult had an income 120 percent higher than that of his parents. Sample members, however, were aged 44–46 in 2009—older than their parents were in 1979 by four years, so with more potential work experience. The sample also shrinks by 9 percent, so it could be that poorer adults dropped out of the survey.
Scott Winship is the Walter B. Wriston Fellow at the Manhattan Institute for Policy Research. You can follow him on Twitter here.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, e21 delivers a short email that includes e21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the e21 Morning eBrief.