Repeating part of Raj Chetty’s classic study using the American Nations model revealed strong differences across U.S. regional cultures, with southern regions performing worst
By Colin Woodard
Historically, the United States has prided itself as being “the land of opportunity,” a place where a child with the humblest of beginnings has a real chance of achieving their dreams. But in this era of growing and staggering economic inequality is it still the case?
A decade ago, Harvard University economist Raj Chatty and his collaborators asked these same questions and turned to anonymized Internal Revenue Service tax data in search of answers. Their central conclusion: upward mobility varies substantially at the regional level, with the worst outcomes in the Southeastern U.S. and the highest in the Great Plains, West Coast, and Northeast. “Going forward,” they wrote, “a key question is why some areas of the United States generate higher rates of mobility than others. We hope that future research will be able to shed light on this question by using the mobility statistics constructed here.”
Late to the party, as they say, but we thought we’d see what would come of running Chetty et. al.’s data through the American Nations model, which defines cultural regions via rival colonization patterns and the ideological and cultural characteristics they imparted on different parts of the North American continent. The answer: despite the presumably substantial influence of state-level policies on inequality, the American Nations cultural boundaries had a substantial degree of county-level continuity with intergenerational economic mobility, even within states riven by major early colonization fault line, like Arkansas, Virginia, Maryland, Ohio and New Jersey.
Chetty’s group parsed inequality via several approaches and found similar spatial patterns. For efficiency’s sake, we focused on what they described as their primary measure, what academics call Absolute Upward Mobility. This takes children raised by parents who were in the 25th percentile of national parent income distribution and looks at how they ranked among their peers at some fixed point of adulthood. This allows researchers to ascertain what the odds of “making it” or “doing better than their parents” are for a poor kid raised in a given place. Their experiment focused on the best data comparisons they could make at the time, which for various reasons were for children born between 1980 and 1982. It identified families in the 25th percentile based on their mean incomes in the tax returns they filed between 1996 and 2000 – when the children were 15 to 20 years old – and then looked at those children’s mean total family incomes in 2011 and 2012, when they were approximately 30 years old. (This cohort is now approaching their mid-40s.)
With our partners at Motivf, we ran this same experiment using county-level data (rather than the multi-county “commuter zones” at the forefront of Chetty’s team’s analysis.) The county-level results are plotted in the map at the top of this article. Note that this is mapping the mean mobility of children who grew up in each of these counties, not the mean mobility of people living there now. (In other words, a place can have high mobility not because there are local economic opportunities, but because large numbers of children from those places move to seek opportunity in places that do.) Also, the grey counties — mostly in the high Plains — are those (sparsely populated) ones that lacked sufficient data for Chetty’s team to confidently conduct their analysis.
Notice that the regions that were slave-centered societies until the mid-1860s and had formal racial caste systems until the mid-1960s have the worst social mobility, save for First Nation, the Alaska Native culture region that also has alarmingly bad life expectancy, suicide, and various health indices statistics. This is consistent with Chetty et al.’s observations that informal, present-day racial segregation in housing, schools and the like are highly correlated with poor social mobility – including for white people growing up in those places.
To this I would add that the Deep South and (legacy) Tidewater were societies built on the premise that people are not equal and that inequality (and, in the past, slavery) are natural and even desirable aspects of society, as they were in Ancient Greece and Rome. Historically, these regions actively avoided social investments that level the playing field – in public schools, health care, social supports for the poor and the like – in service of maintaining oligarchic or aristocratic control. (And the Deep South still does so.)
For clarity, here’s a map of Absolute Upward Mobility in the American Nations as a whole:
New Netherland – the Dutch-colonized area around what is now New York City – was, from the 1640s onward, a society of “self-made men,” and spawned the “American Dream” narrative in American life. It is still the most upwardly mobile region today, or at least it was in 2012, with kids who grew up in families at the 25th percentile of income growing up to be in the 44.5th percentile at age thirty. That’s 6.1 percentile higher than the Deep South, and a whopping 15.1 percentile better than First Nation, where centuries of imperial neglect and exploitation have severely damaged communities and social fabric.
All three “western” nations also perform very well, with Left Coast second in the country with children making a mean ascent to the 43.3rd percentile and El Norte and the Far West both at the 43rd. These regions – especially Left Coast – are more inclined to invest in public goods than their southern counterparts, which my research collaborators and I have shown elsewhere to have far better health and wellness outcomes, creditworthiness, and life expectancy. Two other strongly communitarian-minded cultures, Yankeedom and the Midlands, are just behind at 42.6 and 42.5 respectively.
This individualistic vs communitarian orientation may also explain why Greater Appalachia – where individuals prize their self-worth and have historically resisted overlords of all sorts – doesn’t do much better than the oligarchic lowland Southern regions. Appalachian culture may prize the dignity and honor of ordinary people, but it’s also hostile to having strong institutions or governments of any form, resulting in poor and underfunded social supports and services and a weak regulatory environment. Since the early 20th century, the region’s people have left or engaged in cross-regional migrant work as avenues to social mobility. This is likely why children raised in counties across West Virginia perform as well as they do in absolute mobility: in 2010 it was the state with the greatest “brain drain” in the country, according to a 2019 analysis of net educated out-migration conducted by the Republican-controlled Senate Joint Economic Committee.
It also lines up Chetty’s more recent work probing the causes of the spatial disparities his team found. In a 2022 paper in the journal Nature, he and his colleagues found that places with a high level of intra-class social connectedness – where poor and successful people share friendship circles with one another – is one of the most important predictors of social mobility. (Rates of volunteering and the density of civic organizations were also correlated with social mobility, but had much less of an influence.)
I was also struck by the pattern paralleled the one we saw when we mapped the distribution of immigrants as of the 2020 census, even though Chetty’s data excludes the children who weren’t yet U.S. citizens in 2013 (because it’s impossible to determine their parent’s income levels if they weren’t in filing U.S. tax returns in the late 1990s.) Check this out:
Notice the regions with the very highest intergenerational mobility are also the ones with the largest share of immigrants. Regions with a smaller share of immigrants don’t perform as well, which could mean immigrants are attracted to places with high mobility or that immigrant children may be more motivated to be upwardly mobile, or most likely a combination of the two. (But note the pattern isn’t as tightly linked when you compare county-level mobility, meaning there are definitely other significant factors at play when it comes to non-immigrant mobility.)
These findings call for, yes, more research, but they suggest a legacy of slavery and apartheid has a lasting and very negative effect on upward mobility, and that some things must be working pretty well in New Netherland and the West.
Thanks, as always, to our Motivf friends Tova Perlman (for the data crunching) and John Liberty (for the maps you find herein.)
- Colin Woodard is the director of Nationhood Lab, a project at Salve Regina University’s Pell Center for International Relations and Public Policy.