The 54th Massachusetts Regiment was famously depicted in the movie Glory. The regiment’s main claim to fame is for being one of the first Black units to engage in combat during the Civil War. But, aside for their own freedom, the 54th also fought for something else too — the equal pay that they were promised. When the 54th Massachusetts was offered $7 a month instead of the $13 paid White soldiers, they simply refused to be paid at all. Call it one of the original racial income gaps. Eventually, after an entire year of not being paid and a lot of lobbying, the regiment was granted its rightful pay.
The story of the racial income gap in the U.S. does not have such a happy ending. In 1975, the median Black full-time worker made 26 percent less than the median White worker. Today, that number is pretty much unchanged. Lot’s of reasons exist for the persistence of this gap. Researchers at Harvard have identified one that any real estate agent would understand: location. It turns out, Black individuals are more likely than White people to grow up in areas with low economic opportunity. And, this lack of opportunity ultimately has a big impact on the persistence of the income gap.
Location Matters
Saying that location has an impact on the racial income gap is a statement of causality. You know how everyone is always saying: “correlation isn’t causation”? Well, sometimes a correlation is causation, and economists spend a lot of time trying to figure out when. Indeed, I put an entire classroom of 20-year old students to sleep twice a week teaching them about identifying so-called “causal effects”.
Lots of data show that place is correlated with kids’ later life economic outcomes. The map below illustrates this correlation clearly. It shows how far children from low-income families climb the economic ladder based on where they grow up. In the red areas, children from low-income families end up as lower-income adults. But, in the green areas, these children are able to further climb the ladder.
Figure. Household Income at Age 35 for Children of Low-Income Parents
But, is this correlation causal? Two possibilities exist. The first is just that some places are home to people who would struggle anywhere. Perhaps kids who grow up in certain areas just aren’t the type to work hard in school. In this case, no matter where these kids live, they wouldn’t make lots of money as adults. So, place would not be causal.
The second possibility is that place itself drives outcomes. For example, if some neighborhoods have low quality schools, then the kids who grow up there might not learn valuable skills and then struggle as adults. Had they grown up somewhere else, they would have learned those skills. Then, place would be causal. So, which is it?
(Note: for a visceral example of differences in school quality, listen to the first 5-6 minutes of this podcast.)
Why do We Think that Location is Causal?
Proving the causality of location on adult outcomes is hard. It requires controlling for all the stuff that that isn’t location, but that might vary across location, and that could affect adult outcomes. Like the tendency to work hard in school from the example above. In other words, you have to try and compare otherwise identical people, who just happen to live in different places. How does one do that?
A group led by Raj Chetty had a clever approach: use siblings. The approach is actually pretty simple. Look at families with multiple children that move from one neighborhood to another. Then, see which sibling does better in the long run. In a green area from the map above, if the younger sibling does better, it would suggest causality. After all, the younger sibling has more time for the “good” area to have an impact. In red areas, if the older sibling does better it would suggest causality. After all, they have less time for the bad effect of the place to “rub off”.
Using siblings is a great idea. After all, these individuals grow up in the same household. So, their parents have similar parenting styles, income, wealth, professional networks, and anything else that might affect adult earnings. And, the two people have pretty similar genes, which might account for things like the tendency to work hard. So, whatever is left over is presumably due to the actual effect of being exposed to a given location for differing amounts of time. Here in Boston, we call that a “Wicked Smaht” approach. (Cue native Bostonians rolling their eyes at the transplant from Maryland trying to be all Boston.)
The findings in this study suggest that place does indeed have a causal effect on adult outcomes. On average, young siblings outshine their older ones when the family moves to a green area. And, the older sibling does better when moving to a red area. So, place is probably causal. So, what the heck does this have to do with the racial income gap?
Location and the Racial Income Gap
Have you ever heard the term “reversion to the mean”? What it means (no pun intended) is that if you start out below the mean, you have a tendency to come up to it, if you start out above, you fall back down. The term is often used in sports. If a typically good player is missing all of her free throws, we expect her to start making them soon. Or, if someone is on a world record pace in home runs or goals scored, reversion to the mean would suggest the rest of his season won’t be nearly as strong.
Reversion to the mean would suggest conversion between racial groups over time with respect to their income. After all, Black households start out below the mean income, so they would have a tendency to rise. And, White households a tendency to fall from their position above the mean. Indeed, Raj Chetty and his research group estimate that if opportunity were equal across racial groups, then it would take just two generations for the racial gap to shrink to about a tenth its current size.
That hasn’t happened…at all. Why not? Because, Black people and White people do not face the same opportunity, even if they come from households where parents make the same income. For example, a Black child whose parent’s are at the 25th percentile of the income distribution would make it to roughly the 30th percentile as an adult. A White child with the same parental income would make it to the 42nd percentile on average. Indeed, this opportunity gap is such that from generation to generation to generation, the wage gap stays completely stuck.
And, a big reason for this opportunity gap is: location. According to that Harvard team, a full 20 percent of the racial wage gap is due solely to where Black children grow up. In other words, if you just moved Black children to where White children lived, the causality of location would reduce the racial income gap by a fifth.
Reckoning with the Past
So, why do Black children live in places so different in terms of opportunity than White children? Well, at least part of the reason goes way back. You’ve undoubtedly heard of redlining. The practice determined where the government felt it was safe for home loans to be made, and had a huge resultant impact on how much was invested in those neighborhoods. Areas that were “risky” for loans to be made had a red line drawn around them. Hence the name. Those areas often had one thing in common with each other: they were disproportionately Black.
But, despite being outlawed since 1968, the practice still has a huge impact on housing in America. It particular, areas that were redlined back in the day remain disproportionately Black to this day. And, areas that were given “green” for being not risky (i.e., areas that were white), remain mostly white. You can explore all the evidence here.
The takeaway of all of this is pretty simple. Location has a causal impact on adult outcomes. Black people tend to live in areas that are especially low on opportunity. And, the places Black people live were largely determined by government policy from over fifty years ago. If we don’t reckon with this past by changing housing policy to facilitate racial and economic desegregation of neighborhoods, the racial gap may be stuck at 1975 levels forever. Luckily, you can find lots of such solutions here.