The American Dream Is Not a Coin Flip, and Wages Have Not Stagnated

In my last column, I showed that Americans’ assessments of the economy have tracked the official unemployment rate well over the long run. That is important because it suggests that both public opinion and objective measures indicate that the labor market is historically strong (though accelerating inflation during and after the COVID-19 pandemic has caused these trends to diverge). Is public opinion also consistent with long-run trends in living standards and opportunity?

According to populists on the right and left, Americans have seen meager income growth at best. National conservatives like Oren Cass claim that workers have endured “decades of stagnant wages.” Progressives typically agree, as when New York Times journalist David Leonhardt asserted in his recent book that, “For almost half a century now, incomes and wealth have grown at a sluggish pace for most Americans.” Leonhardt’s book was inspired by research led by Harvard economist Raj Chetty, who he quotes as finding that “achieving the American dream is a 50-50 proposition.”

These are claims about the evolution of “real” income—its purchasing power after accounting for the rise in the cost of living. Debates over real income trends are contentious in part because they require good measures not only of “nominal” income—how much people actually receive—but of prices (which affect what those incomes can afford). If a worker gets a raise of 3 percent but prices go up by 4 percent, real income declines.

Sadly, debates about living standards and opportunity often bog down in technical arguments about measuring income and prices. And the measurement issues make a big difference, as we will see. However, there are good reasons to prefer some methods over others, and those methods often show that living standards have increased more than is appreciated.

Unfortunately, these debates are technical, and it is easy for doomers to dismiss the methods they do not like. Rather than get into the weeds, it is easier to claim the mantle of populism and tell Americans they “are right to believe their lyin’ eyes.”

The problem for the populists is that subjective measures strengthen the case for the methods that disprove the doomer narrative. We can see this by digging into Chetty’s view that achieving the American dream is tantamount to a coin flip.

His claim summarized the headline finding from a paper published with his colleagues at Opportunity Insight (OI): only 50 percent of children born in 1984 ended up with higher family income at age 30 in 2014 than their parents had at the same age. That was down from over 90 percent of children born in 1940. Less conspicuously, however, the paper included other estimates, using different methods to measure income and prices, suggesting that this “absolute mobility” remained above 50 percent. As we’ll see, his data are consistent with mobility being well above that level.

Fortunately, a number of polling organizations have asked Americans directly over the years whether they believe they are better off financially than their parents were at the same age. We can compare those responses to objective measures of absolute mobility using different methods and see which are most consistent with each other. That can tell us whether the methods that technical arguments favor are also most consistent with subjective measures of absolute mobility. That, in turn, gives us information about how to measure real income and wage trends accurately.

Most People Think They’re Doing Better Financially than Their Parents Were at the Same Age

Let’s start with what people tell pollsters. I searched polling questions that (1) referred explicitly to adults’ economic circumstances relative to their parents at the same age and (2) were asked of a nationally representative group of American adults. That ruled out a few questions asked only of voters or certain demographic groups, and it excluded questions asking how one generation as a whole compared with a past one.

I found 78 times that one of 37 surveys included a question fitting these criteria between September 1978 and December 2023. These questions differed in their wording and allowed different responses. For example, many offered responses that were “about the same” as their parents’ rather than requiring “better” or “worse” answers. In those cases, I confined responses only to the “better” or “worse” options and computed the “better” share. (The assumption is that if the “about the same” respondents were forced to choose better or worse, they’d break the same way as everyone else.)

Figure 1 presents the results. (Ignore the thick blue line for now.) I connect the data points in cases where the same organization asked the same question in the same way multiple times; the rest are shown as points that may or may not be comparable to each other.

Figure 1. Percent of Adults Saying They Are Better Off Financially than Their Parents at the Same Age, 1978-2023

Source: Tabulation of data obtained from Roper Center’s iPoll archive of public opinion questions, https://ropercenter.cornell.edu/ipoll/; Karlyn Bowman and Samantha Goldstein, “The Nostalgia Impulse,” https://www.aei.org/wp-content/uploads/2023/01/The-Nostalgia-Impulse.pdf?x85095; Barry Sussman, “Inflation Pessimism is Rising,” Washington Post, https://www.washingtonpost.com/archive/politics/1981/04/09/inflation-pessimism-is-rising/32ffa935-5445-458e-9691-7cfdb74c2952/; Deseret News and Center for the Study of Elections and Democracy, “The American Family Survey, 2022 Report,” https://media.deseret.com/media/misc/pdf/afs/2022-american-family-survey.pdf; NORC at the University of Chicago, General Social Survey, https://sda.berkeley.edu/archive.htm; CBS News/YouGov poll, December 6-8, 2023, https://d3nkl3psvxxpe9.cloudfront.net/documents/cbsnews_20231210_1.pdf.

‍

The first finding is that 76 out of 78 times Americans were asked over 55 years, more people said they were doing better than their parents than said they were doing worse. The two exceptions involved questions allowing respondents to say their circumstances were about the same as their parents’. But even in those instances, only a minority of adults said they were doing worse than their parents.

On average, 76 percent of adults said they were doing better (out of those saying better or worse), with the range running from 42 percent to 93 percent. The average is 74 percent using only the subset of questions that offered only a better-or-worse choice.

Fewer People Than in the Past Think They’re Doing Better than Their Parents

A second observation is that the trend is downward. This conclusion is somewhat more ambiguous than it might seem, given that the points are not necessarily comparable. For example, we don’t have similar 1978 estimates for those stand-alone points below the 70 percent line after 2010; if we did, they might indicate levels below 70 percent even in 1978. The eye is also drawn to those outlier points even though the average of all the data points since 2010 is 71 percent.

However, even the comparable points connected by trend lines tend to show declines. For instance, the dark green trend from 1994 to 2022 is from the respected General Social Survey (GSS). The 2022 value from that survey indicates that 76 percent of adults thought they were better off than their parents, down from 82 percent in 1994.

To summarize the overall trend, I computed the average of all data points from each year or the two years before and after it. (All points had to be within five consecutive years of each other.) The result is the thick blue line in Figure 1.

Again, the trend is downward. The average for 2023 is 65 percent—down from 85 percent in 1978. The average fluctuates with the business cycle, though it does not turn up in our current post-COVID expansion. There are only a handful of data points during these recent years, which suggests caution. In particular, the GSS trend indicates less of a decline than my public opinion average (6 points instead of 10 between 1994 and 2022).

Several Ways of Tracking Income Indicate that Intergenerational Mobility Is in Worse Shape than People Think

In a forthcoming paper with Thomas O’Rourke, I reanalyze Opportunity Insights’ absolute mobility results. I use the OI data to replicate their findings, extend the estimates for ten additional years, and then make different measurement choices.

Figure 2 carries over my public opinion average as the thick blue line and displays three ways of measuring absolute mobility. The lowest line looks at 30-year-olds and compares their family income to that of their parents when they were thirty. When the OI team published its paper, the 2013 point was the most recent estimate, indicating that just 50 percent of adults experienced absolute mobility. (For reasons discussed in my forthcoming paper, it’s more accurate to say this data point reflects the absolute mobility of the 1983 birth cohort as of 2013 than the 1984 cohort as of 2014, as the OI researchers do.) The absolute mobility rate temporarily fell to 49 percent in 2019, but it has been rising since. In 2023, it was 57 percent. That was down from 80 percent in 1978.

Figure 2. Percent of Adults Saying They Are Better Off than Their Parents vs. Estimates of Actual Absolute Mobility, 1978-2023

Source: See Figure 1 for public opinion average. Absolute mobility estimates are from reanalyses of data from Raj Chetty, Nathaniel Hendren, David Grusky, Maximilian Hell, Robert Manduca, Jimmy Narang, “The Fading American Dream: Trends in Absolute Income Mobility Since 1940.” Replication files available at https://opportunityinsights.org/data/?geographic_level=0&topic=0&paper_id=546#resource-listing.

These results were the basis for OI's claim that “the American Dream is fading” and Chetty’s evocation of it as a “50-50 proposition.” However, note how much worse those estimates look than subjective beliefs about absolute mobility, as reflected in the public opinion average. There are two possibilities here. My public opinion average could be too rosy, perhaps because people want to look good to pollsters and exaggerate how well they’re doing or because they have an overly negative bias about the past. (Of course, they could also have an overly positive bias about the past.) Alternatively, the OI results could be too dour. While objectively measuring how people are doing relative to their parents might seem straightforward, it is anything but that.

Fortunately, we can improve the OI results using the information they provide other researchers. Before making these adjustments, let’s switch to comparing parents and their adult children at age 40 rather than looking at 30-year-olds. This is probably more appropriate for comparing absolute mobility to my public opinion average since the latter includes adults of all ages. The average adult was 43 years old in 1978 and 48 years old in 2023.

At least based on the difference between the lines for 30- and 40-year-olds, it looks like long-term trends and levels aren’t impacted much by the age when intergenerational comparisons are made. The absolute mobility of 40-year-olds fell from 86 percent in 1979 to 56 percent in 2023. (See the red line in Figure 2.) Like the line for 30-year-olds, it’s not a good match with the public opinion average.

The yellow line uses a different methodology to adjust 40-year-old incomes for the rising cost of living over time. (Technically, it uses the “personal consumption expenditures price index,” or PCEPI, preferred by the Federal Reserve Board and the Congressional Budget Office—and by me—over the “consumer price index research series” used by OI.) This switch increases absolute mobility to 60 percent in 2023 and slightly reduces the decline over time. But the rates still don’t align well with the public opinion trend.

Better Ways of Tracking Income Align Better with People’s Impressions of Intergenerational Mobility

In a recent working paper, I argued that even the Fed- and CBO-preferred cost-of-living adjustment from Figure 2 strongly overstates inflation over the long run. I developed an alternative called the “more accurate consumer price index,” or MACPI. (The MACPI goes back only to 1973. Here, I extend it back in time by assuming it grows one percent less per year than the consumer price index.)

The bottom line in Figure 3 shows what happens if you use the MACPI to adjust incomes before assessing absolute mobility. Doing so moves the absolute mobility trend closer to the public opinion average. As of 2023, 63 percent of Americans had exceeded their parents’ income. In 1979, the rate was 89 percent.

Figure 3. Percent of Adults Saying They Are Better Off than Their Parents vs. Better Estimates of Actual Absolute Mobility, 1979-2023

Source: See Figure 1 for public opinion average. Absolute mobility estimates are from reanalyses of data from Raj Chetty, Nathaniel Hendren, David Grusky, Maximilian Hell, Robert Manduca, Jimmy Narang, “The Fading American Dream: Trends in Absolute Income Mobility Since 1940.” Replication files available at https://opportunityinsights.org/data/?geographic_level=0&topic=0&paper_id=546#resource-listing. MACPI is from Scott Winship, “Introducing the More Accurate Consumer Price Index,” https://www.aei.org/wp-content/uploads/2024/11/price-index-working-paperFINAL.pdf?x85095.‍

‍

Many researchers prefer to adjust incomes for family size when looking at long-term trends. Supporting a family of two at a given income level is easier than supporting a family of four. Moreover, families have become smaller over time as marriage and fertility have declined. That suggests that looking at income trends without accounting for this “fewer mouths to feed” factor will understate improvement in living standards over time. Put another way, poll respondents are likely to think they are better off than their parents if they have $80,000 to split among three people while their parents had $80,000 to split among five.

The next line up in Figure 3 returns to the PCEPI but adjusts incomes for family size in a standard way (technically, dividing incomes by the square root of the number of family members). Whether to size-adjust income is a more complicated question than it might seem, however, this trend line is even closer to the public opinion average than the previous absolute mobility estimates. It indicates that 69 percent of 40-year-olds were better off in 2023 than their parents were at the same age. That’s down 20 points since 1979, but nearly 20 points higher than OI’s 50 percent headline finding.

Finally, the red line in Figure 3 includes the family size adjustment and uses the MACPI. This trend line, too, fits the public opinion average relatively well, particularly from 1999 to 2020. It indicates that while absolute mobility is still down 20 percentage points since 1979, in 2023 the rate was 72 percent.

The Best Subjective and Objective Measures Agree: 70 to 75 Percent of Americans Are Better Off than Their Parents

As a final check, we can leverage the accessibility of the high-quality General Social Survey data, running from 1994 to 2022. I used that data to produce a trend in subjective assessment of absolute mobility confined to people ages 35 to 45. That drops younger and older adults and makes the public opinion trend more consistent with the objective absolute mobility trends. It also has the virtue of using a single survey asking the same question over 28 years in the same way, rather than having to average disparate survey results.

Figure 4. Percent of 40-Year-Old Adults Saying They Are Better Off than Their Parents vs. Estimates of Actual Absolute Mobility, 1994-2022

Source: General Social Survey, https://sda.berkeley.edu/archive.htm. Absolute mobility estimates are from reanalyses of data from Raj Chetty, Nathaniel Hendren, David Grusky, Maximilian Hell, Robert Manduca, Jimmy Narang, “The Fading American Dream: Trends in Absolute Income Mobility Since 1940.” Replication files available at https://opportunityinsights.org/data/?geographic_level=0&topic=0&paper_id=546#resource-listing. MACPI is from Scott Winship, “Introducing the More Accurate Consumer Price Index,” https://www.aei.org/wp-content/uploads/2024/11/price-index-working-paperFINAL.pdf?x85095.

‍

As above, I look at the share saying they are better off than their parents among those who either say they are better or worse off. Figure 4 compares the subjective measure to the objective measures shown in the previous chart and my extension of the Chetty OI trend for 40-year-olds. Clearly, the best-fitting objective measure is the one that both uses the MACPI and adjusts for family size. According to the best subjective and objective measures, then, in 2022, 75 percent of 40-year-old adults said they were doing better than their parents at the same age, and 70 percent were. In 1994, the figures were 78 percent and 81 percent. Pretty consistent.

Middle-Class Incomes Have Risen Impressively Using the Measures Favored by Technical Arguments and that Align Best with Subjective Assessments of Well-Being

What do these results tell us about income and wage stagnation claims? We’ve seen that adjusting income for family size and measuring price change using the MACPI produces absolute mobility estimates that better align with what Americans say about their mobility. The OI data allow for annual estimates of the typical (“median”) 40-year-old’s family income. Using the methods that produced OI’s headline results, the income typical of a 40-year-old rose by just 15 percent between 1979 and 2019 (both business cycle peaks). However, using the MACPI to adjust for the rise in the cost of living, the increase was 52 percent. If we also adjust for changing family size, the increase was 73 percent. These estimates all involve pre-tax income and exclude government benefits.

We can also revisit Cass’s complaint about wage stagnation using the lessons learned here. His claim comes from his data analysis using a price measure known to have major flaws for assessing long-run trends. In recent analyses, I showed that when using that price index, the hourly wages of the typical worker rose by just 2 percent from 1973 to 2023. Using the OI-preferred price index, the increase is a little better—14 percent—but not great. However, using the MACPI, the increase was 60 percent. (Wage analyses typically don’t adjust for family size.)

The doomers are left in a pickle. They want to argue that the methods with the best technical arguments behind them present too rosy a view of how Americans are doing. They want to appeal to the incongruence of these estimates with how Americans supposedly say they are doing. But it turns out that the best methods do better matching Americans’ actual views. There’s a great irony here: to maintain their position that social science is wrong about how the economy is doing, doomerists must also accuse everyday Americans of having lyin’ eyes.

Equating the American Dream with Absolute Mobility Is a Mistake

A final note: it's true that no matter what the measure, absolute mobility has fallen over the long run. While doomers may be tempted to take a win (a loss?) where they can get it, equating absolute mobility with the American Dream is a mistake. The OI research shows that Americans who grew up poor have the highest rates of absolute mobility—over 80 percent by even their preferred measure, compared with less than 20 percent for adults who grew up with the richest parents. It is, after all, easier to exceed a low parental income than to exceed a high one. It would be difficult, however, to find adults raised poor who would lament that their children are unfortunate for having a higher living standard than they did but a lower likelihood of moving up even further when they grow up. The odds remain solid that their kids will move up again; to the extent those odds diminish, it’s because the parents rose so high that their children could unlikely repeat such a leap.

Scott Winship is a senior fellow and the director of the Center on Opportunity and Social Mobility at the American Enterprise Institute. He is a contributing editor to Civitas Outlook.