In 2015, Americans spend a lot of time listening to advice about America from high IQ immigrants from other countries, such as Hillary Clinton’s Idea Man Raj Chetty:
Chetty was offered tenure at the age of 28 and accepted at 29, becoming one of the youngest tenured faculty in the history of Harvard’s economics department. He is a recipient of the John Bates Clark Medal and a 2012 MacArthur Fellow.[3] Currently, he is also an editor of the Journal of Public Economics.[4]
One minor problem with this is that individuals like Chetty tend to be pretty clueless about America, as I’ve noticed the more I’ve dug into the immense pile of data Chetty has assembled about “social mobility” and “equality of opportunity.”
… “We think there are really large causal effects of local communities on children’s long-term success,” Chetty said during his presentation.
… This new research builds on two 2014 studies, which you can read here. In one, the researchers followed the earnings of a core sample of children born between 1980 and 1982. They found huge geographic variations in upward economic movement, as you can see in the map below. Lighter-shaded areas have better mobility.
This map shows the average percentile rank of children who grow up in below-median income families across areas of the U.S. (absolute upward mobility). Lighter colors represent areas where children from low-income families are more likely to move up in the income distribution.
Chetty’s FAQ explains what’s being mapped a little more precisely:
Statistically, we define absolute upward mobility as the average percentile in the national income distribution of a child who is born to parents at the 25th percentile in the national income distribution. In areas with higher absolute upward mobility, children from low-income parents earn higher incomes on average as adults.
Chetty’s team got their hands on a giant amount of IRS data allowing them to match up the incomes of individuals around 30 years old in 2011/2012 with their parents’ incomes in 1996. This map shows how far individuals who were in families at the 25th percentile nationally (not regionally) in income in 1996 regressed toward the mean in 2011/12.
Among the top 50 “Commuting Zones” (and even more sprawling version of metro areas to include in more rural regions), the young people in Charlotte, NC in 1996 had the lowest “absolute upward mobility” in 2011. Somebody whose parents in Charlotte were at the national 25th percentile in 1996 was, on average, in the 36th percentile nationally in 2011/12. (Note that this is regardless of where they live in 2011/12.)
In contrast, Salt Lake City had the most absolute upward mobility: the average 25th percentile resident in 1996 is now at the 46th percentile nationally in 2011.
Obviously, Regression Toward the Mean is playing a huge role here. Chetty’s most American Dream-crushing hellhole, Charlotte, still saw over half as much absolute upward mobility as his utopia, Salt Lake City.
Nationally, the typical kid who was at the 25th percentile in 1996 was at about the 40th percentile in 2011. What’s the reason? No particular reason, other than regression toward the mean. That’s just the way the universe works.
Note that Chetty is not interested in where to move to in this decade; instead, he’s trying to measure which ex-teenagers had the foresight to pick the right place to be from in the 1990s.
The goal is to find something about the Salt Lake Cities that makes it so much better to have been from than the Charlottes.
Bloomberg goes on:
Figuring out why economic mobility varies so greatly and what causes mobility to increase or decrease is important. It might give economists and policy makers clues to how to give kids a better chance at climbing the economic ladder.
Does Charlotte have too much sprawl compared to Salt Lake City? Is it … segregation?
So let me help Prof. Chetty figure out why “economic mobility varies so greatly and what causes mobility to increase or decrease.”
First, as I’ve pointed out before, the most obvious thing about his map above is that it’s basically a map of where the blacks are, plus giant Indian reservations like the Navajos’ reservation that’s mostly in the Arizona quadrant of the Four Corners. In other words, blacks and reservation Indians tend to regress toward lower means.
Chetty seems irritated by people pointing this out, like I did in 2013. He didn’t get a huge grant just to discover that, all else being equal, black people tend to be poorer than white people everywhere. There must be some sinister reason instead. Chetty wrote in 2014.
This correlation could be driven by two very different channels. One channel is an individual level race effect: black children may have lower incomes than white children conditional on parent income, and hence areas with a larger black population may have lower upward mobility. An alternative possibility is a place-level race effect: areas with large black populations might have lower rates of upward mobility for children of all races. …
Unfortunately, we do not observe each individual’s race in our data. As an alternative, we predict race based on the parent’s 5-digit ZIP code (in the year they first claim their child as a dependent). We use data from the 2000 Census to measure racial shares by ZIP code. Figure IXa replicates the map of absolute upward mobility by CZ, restricting the sample to ZIP codes within each CZ in which at least 80% of the residents are non hispanic whites. In this subsample, 91% of individuals are white. The spatial pattern in Figure IXa is very similar to that in the original map for the full sample in Figure VIa. Most notably, even in this predominantly white sample, rates of upward mobility remain low in the Southeast and are much higher in the West.
Actually, the West doesn’t look so hot in this map of the >80% white zip codes within Chetty’s Commuting Zones: look at the northwest Pacific Coast from far northern California to the Olympic Peninsula. Instead, it’s the northern Great Plains where it looks like white people did pretty well from 1996 to 2011. (Apparently, that’s the West to Chetty.)
In contrast, Northern lower Michigan looks bad, as does Maine, and the southern Appalachians got hammered.
What horrible ghost of Jim Crow infests western Oregon, northern Michigan, Maine, and western North Carolina? What do they have in common that, say, North Dakota doesn’t have that causes them to look bad from 1996 to 2011, while blue collar kids from the Dakotas and Nebraska are flying high in 2011?
Well, here are the top 10 places in Chetty’s database for absolute upper mobility among mostly white zip codes from 1996 to 2011:
Oh, now I get it! The reason early 30s blue collar white people who are used to working outside in the Great Plains winter were making so much more money in 2011 than their parents were making in 1996 has less to do with Chetty’s Deep Thoughts about Sprawl and Segregation and Social Capital than it has to do with the North Dakota fracking boom. From Wikipedia:
North Dakota oil boom is an ongoing period of rapidly expanding oil extraction from the Bakken formation in the state of North Dakota that followed the discovery of Parshall Oil Field in 2006, and is continuing as of 2015. Despite the Great Recession, the oil boom has resulted in enough jobs to give North Dakota the lowest unemployment rate in the United States. … North Dakota, which ranked 38th in per capita gross domestic product (GDP) in 2001, rose steadily with the Bakken boom, and now has per capita GDP 29% above the national average.[7]
There are three reasons for the oil boom, not just in North Dakota but nationwide:
- the recent discoveries of shale gas reserves in the United States — initiatives to seek independence from unstable energy sources, such as Venezuela and nations in the Middle East — the successful use of horizontal drilling and hydraulic fracturing, which have made energy deposits recoverable[8]
This energy boom has drawn in blue collar workers from other cold weather states, like South Dakota, Minnesota, Nebraska, and Iowa. As far as I can tell, it hasn’t drawn in that many illegal aliens from warm weather Latin American countries, allowing some blue collar Americans to, for once in their lives, get ahead for awhile.
Chetty is staring at the same maps I am. Sadly, the Magic Eye pattern never emerges from murk for him. He writes:
For example, many parts of Texas exhibit relatively high rates of upward mobility, unlike much of the rest of the South. Ohio exhibits much lower rates of upward mobility than nearby Pennsylvania. The statistics also pick up much more granular variation in upward mobility. For example, South Dakota generally exhibits very high levels of upward mobility, …
What could the connection be?
Well, South Dakota is next to North Dakota. Even Chetty should have heard of the connection between Texas and energy. (There’s an oil and gas boom going on in the Eagle Ford Shale Formation in South Texas.) What’s the difference between, say, Ohio and Pennsylvania in 2011? Pennsylvania is where oil drilling was invented in 1859, so there remain energy resources in the ground that were untappable until the new technology came along over the last decade. Ohio, in contrast, is mostly dirt without good geology for oil and natural gas.
In contrast to the treeless Great Plains, why were blue collar regions with a lot of trees — e.g., western Oregon, Maine, northern Michigan, and especially Appalachia south of the Pennsylvania-West Virginia coal and fracking belt — doing badly in 2011 relative to 1996? Here is a countdown of Chetty’s bottom ten commuting zones for white zip codes for 2011 v. 1996:
Morristown TN, Winder GA, Morganton NC, Wilmington NC, Fayetteville NC, Rome GA, Wayne IN, Dublin GA, Gastonia NC, Hickory NC.
Here’s the Wikipedia explanation of the economy of Hickory, NC, Chetty’s worst place for white people in 2011 relative to 1996:
… wagon-making know-how, proximity to expansive forests, and excellent transportation via two intersecting railroads provided fertile ground for the emergence of the furniture industry. …
The furniture industry in Hickory is not as strong as in the decades previous, but still a primary component in the area economy, and includes HSM (formerly Hickory Springs, founded 1944), a leading manufacturer of mattress coils.
Currently the area is home to many leading manufacturers of furniture, fiber optic cable, and pressure-sensitive tape. It is estimated 60% of the nation’s furniture used to be produced within a 200-mile (320 km) radius of Hickory.
Emphasis on used to be. Here’s an article on the collapse of the North Carolina-Virginia hardwood furniture factory belt.
The Housing Bubble of the 2000s drove wood-oriented industries, such at lumbering, furniture-making, and home-building, to prosperity. The ensuing Housing Bust meant that the incomes of blue-collar whites who came from wood-working regions like western North Carolina got hit very hard in 2011-12.
There’s a direct causal connection between the oil boom and the wood bust. When the price of gasoline spiked upwards in 2008, that killed off the last hopes that the mid-2000s building boom in the distant exurbs made economic sense because long commutes would be cheap. High oil prices (along with much else) helped kill off construction of wooden homes in the exurbs, along with major purchases of new furniture to fill the mcmansions.
In turn, high energy prices justified spending a lot on wages for workers to do horizontal drilling and fracking in North Dakota, Texas, Pennsylvania, and the like. Like the Lion King says, it’s all part of the great circle of life.
In other words, Chetty is getting distracted by economic and technological cycles that aren’t particularly closely tied to very long term reasons why some places are more upwardly mobile than other places. It’s too bad that his data gets batted around so much by local booms and busts, but that’s a serious problem that he needs to deal with, perhaps by getting more years of data.
There is of course something to be learned from his data: one of the most obvious points is that it’s better economically to be an American these days in places that are too cold for the comfort of Latin American immigrants. The Census Bureau found North Dakota was only 2.9% Hispanic in 2013 vs. 17.1% for the whole country. In contrast, Hispanics have grown from 1.2% of North Carolina’s population back in 1990 to 8.9% by 2013.
Now, if you look at big Commuting Zones, what has had the most absolute upward mobility from 1996 to 2011 for young people from mostly white zip codes?
1. New York, New York
2. Newark, New Jersey
Please note that Chetty defines Newark not as the black-ruled inner city, but as the huge suburban sprawl that’s home to many people who commute to Wall Street or Midtown Manhattan.
So, the lesson to be drawn from Chetty’s analysis would appear to be that you should try to make sure your hometown is located within convenient commuting distance of Wall Street during a gigantic financial industry boom.
Does that mean the standard of living has gone up more in New York and Newark than anywhere else? Maybe. It has if you own your own home since 1996. If you are thinking of buying today, maybe not so much.
When The New York Times first aired Chetty’s map back in 2013, I pointed out in the NYT that Chetty should adjust for cost of living differences using the ACCRA data. He actually went out and got the ACCRA data, but now he claims that he doesn’t need to adjust because he checked and it didn’t matter much and you should just trust him on it. Why would you want to look at a cost-of-living adjusted map?
In summary, there’s a lot of interesting data that Chetty has assembled; it’s just too bad that over the two years he’s been promoting it, he doesn’t seem to have hired anybody who knows much about America to analyze it for him.
