Massachusetts Cities and Coronavirus

Last week, I put together a few graphs of Massachusetts covid-19 county cases and correlates. Unfortunately, working at the county level didn’t create many data points, so there wasn’t much insight to be gained. I ended the blog post with a wishy-washy pledge to maybe try to compile a city-level data set.

Well, I didn’t do it. But the Massachusetts Department of Public Health did! Starting April 15, the DPH began tracking coronavirus cases in Massachusetts’s 351 towns and cities. I took those case counts and paired them with demographic and economic data from the 2018 American Community Survey to see if any patterns emerged.

Due to some quirks of the Census, it would have been really tedious to get the data I wanted for every town and city. I ended up opting for a shortcut of sorts — using cities and towns that were their own Census Designated Place. (Don’t ask.) After all was said and done, I was left with 54 cities and towns that I could pair with data from the 2018 American Community Survey.

This data set is biased toward larger areas, which probably also means younger and more diverse areas too. And, of course, all cities and towns in the set are from Massachusetts, which probably introduces other biases.

Anyway, here’s what I’ve found so far:

(Note: I’m using log-scale to condense case counts and, in the case below, population. Boston, for example, has over 4,000 cases [log(4,000) = 3.602], while Springfield has about 600 [log(600) = 2.778].)

Log cases are best predicted by log population. This is pretty much common sense: more people means more vectors for disease, and usually denser population. A regression summary is in the caption below the graph.

log(Cases) = 1.3472 * log(population) – 4.0471
R^2 = .7526
p-value < 2.2e-16

After population, race is the next-best indicator of case counts. I’d expected — based on media reports and the word of local officials — to find a relationship between the percentage of black or Hispanic residents and log cases. But that didn’t really show up. Instead, the proportion of residents that are non-Hispanic whites has the best linear relationship to log cases — and the only one with a negative slope. The regression summary statistics in the caption are only for the white-log case relationship.

log(cases) = -1.8623 * white + 3.5474
R^2 = .545
p-value: 1.887e-10

This does actually fit the narrative pretty nicely if we lump all non-white ethnic groups together mentally. But as I noted in the last post, race is collinear with so many other variables that it’s hard to know what we’re seeing.

There is, however, at least some indication that what we’re observing might really be about race. Other variables we would imagine to correlate to race and population have much weaker relationships to case counts. (“Public transport” is the percentage of people who take public transport to work, and “Poverty line” is the percentage of people below the poverty line.) I grabbed a bunch of variables like these, but so far, none of them are very helpful in explaining what’s going on.

Combining log population and the proportion of residents who are white gives us an adjusted R^2 of .8209, which is nice. But when I tried to use that model to predict case counts for four municipalities that weren’t among the 54 I’d defaulted into working with, it only did okay.

The problem, I think, is that the selection of cities in the data set I’m working with is biased toward larger areas. It’s also possible (in fact, likely) that there’s both more to the story and that an element of randomness is at play, too. Aside from age, it’s been hard for professionals to isolate significant risk factors.

A few MA covid graphs

This is a low-stakes post.

Massachusetts has been releasing county-level coronavirus case counts, which I paired with data from the US Census to look for patterns. I actually didn’t end up finding anything particularly interesting, but some of the graphs are nice, so I thought I’d share.

On case growth

A few days ago, it looked like the growth of covid cases in Massachusetts might be flattening. But as of yesterday, it seems like that’s not quite the case across the board. Here are the total case counts per 1,000 residents of each county county since March 15:

Dukes and Nantucket counties omitted.

And here are cases per 1,000 residents on April 7, with the geometric growth rate of cases over the last week indicated by color:

Growth rate calculated as (x_1/x_0)^(1/7)-1

Berkshire, Barnstable, and Franklin counties have the lowest case growth rates, ranging from 6.2% to 7.3% on average per day over the last week. These counties have some common characteristics:

  • They’re geographically remote;
  • They are the only MA counties to have experienced population decline over the last decade;
  • They have the highest non-Hispanic white populations per capita and the least foreign residents;
  • They have the greatest proportion of residents over 65 (at least 22% in each case!);
  • Franklin and Berkshire counties have the lowest population densities, at 102 and 141 people per square mile, respectively.

To me, the above is consistent with the idea that economic activity is a vector for the spread of coronavirus (not literally, but it gets people in contact which causes person-person transmission).

Plymouth, Hampden, and Bristol are the counties with the fastest-growing case counts, each of them averaging an increase of over 11% daily over the last week. These counties don’t have much in common, so I’m having trouble putting together a potential unifying narrative.

Race and population density as correlates

It’s starting to look like black Americans might be more susceptible to coronavirus than other racial and ethnic groups. At first glance, that appears to show up in county-level data. But upon closer examination, that doesn’t appear to be the case — first because population density and the percentage of black residents are collinear, and population per square mile has a higher correlation coefficient; and second because Suffolk county (Boston) is influencing the linear relationship in both cases. Adjusted R-squared drops heavily if we exclude Suffolk county from the data set. (Race and population density were the best predictors I found of cases per 1,000 residents.)

This isn’t to say race and its many correlates aren’t good predictors. I think it speaks more to the (severe) limits of the data set I’m working with. If I have time, I may try to build a city-level data set. If anyone knows of one (or something better), link me!

Should we be surprised that “young people” make up a substantial amount of coronavirus cases?

I’ve noticed media outlets are reporting that young adults make up a significant share of coronavirus cases with an air of incredulity.

My local paper posted on Facebook that “More than 50% of coronavirus cases in Massachusetts are people under the age of 50.” Very similarly, the Pittsburgh Post Gazette writes that “more than half of Pennsylvania’s confirmed COVID-19 patients are under 50 years old.” The New York Times, for its part, reports that “nearly 40 percent of patients sick enough to be hospitalized were age 20 to 54.” *

I can’t decide if this is a psyop to get young people to take the epidemic more seriously (as numerous spring-break photos show they should) or genuine surprise. If it’s the latter, I’m not sure if that’s warranted.

In each case, the age ranges in question are massive and not very meaningful without comparison to the age distribution of the general population. For example, in Massachusetts, about 63% of the population is under the age of 50. So if the incidence of coronavirus were age-independent, we might actually expect more cases among people under 50.

I think the issue, then, is that people seem to be assuming prevalence of the virus should be age-dependent to a higher degree than we’re observing. Maybe there’s a good Bayesian case to be made for this null hypothesis; I don’t know. But I feel like laypeople — local papers included — ought to be preceding with the assumption of age-independence, especially because we still don’t have much information.

Also, what’s going on with the under-20 crowd, which makes up 23% of the population but only 2.2% of MA coronavirus cases? Is Gen Z+ holding out on us?

The typo in the y-axis comes from the MA DPH data set, which has decided there are no 19-year-olds in the Bay State.

* This isn’t as egregious as the other two examples. It’s still a huge age range: about 48% of America is between the ages of 20 and 54. But since we’re talking about the severity of symptoms and hospitalizations, it seems much more noteworthy.

Smoking and the Hispanic Paradox

In the course of writing last month’s post about U-haul’s no-nicotine policy, I created the following graph:

This visualization didn’t make the final cut, but it’s nonetheless cool. It demonstrates that smoking rates among Hispanics are far less responsive to income than those of other ethnic groups (though even for Hispanics, the relationship between income and smoking rates is statistically significant). I was surprised to find this relationship, but apparently it’s a known factor of the phenomenon called “the Hispanic Paradox” (alternatively known as the “Latino Paradox”).

The paradox is that, on average, American Hispanics live longer than their non-Hispanic white counterparts, even though the former tend to have lower incomes and less education. The causes aren’t entirely understood, but Hispanics’ low smoking rates are thought to be a major contributor.

Source: Center for Disease Control, 2013

Some of the difference in smoking rates can be explained by immigration. Latin American countries tend to have lower smoking rates than the United States. Among those born in the United States, only Mexican-Ameicans seem to retain lower smoking rates and the attendant mortality advantage over non-Hispanic whites. It will be interesting to see if the Paradox ebbs as native-born Hispanics begin to account for more of the Hispanic population.

The Hispanic Paradox illustrates the capricious power of cultural influence on real-world outcomes — and conversely forces us to confront our limited ability to re-engineer the world.

We tend to think of (the physical, policy, social, or economic) environment and choice as the chief determinants of human behavior and outcomes. But we are just as much a product of the commingling of genetics and culture. The paths before us are well-worn by our predecessors, and we would be arrogant to think we can wholly resist their inclinations.

Kicking Away the Ladder?

A report from the Pew Center is the latest to document America’s rapidly declining religiosity. Pew’s numbers show a 12-point decrease in the percentage of Americans calling themselves Christians between 2009 and 2019, while those describing themselves as atheists, agnostics, or “nothing in particular” have risen from 17% to 26%.

These findings are mirrored in the General Social Survey, which has been asking respondents about their religious affiliation since 1972.

religious nones
Shaded area represents 95% confidence interval

The demographic surge of the religiously unaffiliated is a story of treatment effects triumphing over selection effects. A little context will help explain.

Natural selection seems to decidedly favor the religious. As the social psychologist Ara Norenzayan details in Big Gods (an excellent if at times slow book), religious societies, specifically those that follow omnipotent, moralizing “Big Gods,” have historically been able to outcompete others. To summarize Norenzayan’s findings, this is due to three factors, the first two of which are increased trust and greater social stability, both made possible by supernatural monitors (gods) that have allowed societies to scale by “building moral communities of strangers.”

In recent history, some societies (think Scandinavia and Japan) have been able to “kick away the ladder” of religion, replicating the monitoring effects of Big Gods through trusted civic institutions—police, courts, and others that allow for anonymous actors to cooperate in the same way religion used to. But there’s one other important advantage of religious societies that secular societies haven‘t been able to engineer: above-replacement fertility rates.

Fertility by Religion
I’ve restricted my search to 30 – 45-year-olds to try to account for the fact that older Americans are on average more religious and have more children. Huge confidence interval for Jewish women reflect low sample sizes.

The religiously unaffiliated reproduce at notably lower rates—so much so that, even as I’m writing a blog post about their astronomical demographic growth in America, the Pew Center projects they will decline as a share of the global population, from 16% to 13% between 2010 and 2050.

shrinking religiously unaffiliated.png

Followers of Big Gods, especially those of fundamentalist branches of religion, tend to have more children. Note: this is mirrored by the liberal-conservative fertility gap, and both are fueled in part by divergent views on reproduction.

To give you a sloppy illustration, consider the graph below, which I constructed using national-level data from 55 countries that appear in the 2012 Win-Gallup International Global Index of Religiosity and Atheism and Gapminder total fertility rate data from the same year. Nearly every country with an above-replacement fertility rate has a majority-religious population by quite a large margin.

more religion more babies
Horizontal line represents “replacement rate” of 2.1 children per woman. Vertical line represents median religious society in the data set.

In a personal email quoted by Norenzayan in Big Gods, a colleague confides that despite reviewing all available data and case studies back to early Greece and India, he was unable to find a single example of a secular society maintaining a birth rate higher than two children per women for even a century. France, Germany, Japan, and quite a few other countries are trying—and failing—to address this problem with a variety of subsidies. Thus far, there is no secular substitute for religion’s fertility premium.

So with forces of nature solidly in support of religion, why is it rapidly losing ground across the rich world? It turns out there are countervailing secularizing forces that, it feels safe to say, have grown powerful enough to chip away at the natural demographic advantage of the religious. Unlike the selection effects that propel religiosity, these are treatment effects, meaning they’re driven by exposure to certain conditions. In addition to the aforementioned creation of secular civic institutions, those conditions include education, rising incomes, and the general removal of existential threat (of the “where is my food coming from” variety, anyway).

Notably, somewhere around 78% of all “religious nones” are “converts,” if you will, meaning they were born into a religion and ceased to identify with it over time.

While there is some observable increase of religious unaffiliation within generations, the flight from religiosity is largely driven by generational replacement. In other words, it’s not like longtime worshipers have suddenly lost faith en masse—it’s that their grandchildren aren’t interested, and older generations are losing ground demographically. This fits the pattern of other paradigmatic shifts in public opinion, and to me, suggests there’s an element of timing involved, that conditional secularization may be contingent upon one’s formative environment.

The question that remains to be seen is whether or not the secularizing rich world can support itself. Our economies, infrastructure, and social welfare systems are reliant on people, and a large population that doesn’t reproduce will age with dramatic consequences (see Baby Boomers). In the long run, this is probably one of the most consequential political issues out there.

Birds, Bees, and (Abortion) Bans

Last Tuesday, abortion-rights advocates around the country held rallies in response to restrictive abortion laws or bills passed or introduced in several states. At the events, legislators and protesters decried the bills as an attack on women’s rights, an attempt by men to control women’s bodies.

This refrain, that abortion is an issue that divides the sexes, is a common narrative — at least in my social and professional circles. But it’s discordant with data that shows that men and women within a given society, including the United States, often have very similar views on abortion. A graphic from the Pew Center illustrates this nicely:

FT_18.12.14_femaleMaleAbortion420px

Broadly speaking, with respect to the above graph, the differences between nations are much greater than the sex-based differences within them. This suggests to me that cultural factors play a larger role than sex in determining one’s position on abortion and that men and women seem roughly equally sensitive to these forces.

I wanted to check if a similar phenomenon could be observed within the United States’ population. To get a sense of this, I pulled data from the 2018 General Social Survey, which asks respondents “whether or not [they] think it should be possible for a pregnant woman to obtain a legal abortion if the woman wants it for any reason” — a question similar but not identical to the one the Pew Center asks above.

Unfortunately, the GSS doesn’t release data about which state respondents live in unless you pay for it, which I’m not going to do (since I have certainly not made any money with this blog). So that means we can’t examine the opinions of residents where legislators have moved toward a more restrictive stance on abortion. We can, however, get data at the regional level, which seems like an OK proxy.

The regions used by the GSS aren’t entirely conventional — for example, Montana and New Mexico are both in the “Mountain” region — so here’s a chart for reference:

Gss Map

Keeping with the format, if not pleasing aesthetics, of the Pew graph, here are the results by region, arranged by the percentage of women who think a pregnant woman should be able to get an abortion for any reason.

abortion any reason
My chart title game is weak. Forgive me. Again, the text of the prompt is: “Please tell me whether or not you think it should be possible for a pregnant woman to obtain a legal abortion if the woman wants it for any reason.”

Our regional chart resembles Pew’s international chart in that it shows larger variances between regions than within them. Notably, the four regions below the national average — South Atlantic, West South Central, West North Central, and East South Central — contain states where restrictive abortion bills have been introduced. Even more notably, women in three of those four regions are less likely than men to have responded affirmatively.

Let’s take a look at another question from the GSS that asks respondents for their views on the morality of abortion. The question is, “Leaving aside whether or not you think abortion should be legal, are you morally opposed to abortion or not, or would you say it depends?”

Again, results are sorted by the percentage of female respondents — this time those stating a moral opposition to abortion.

Moral opposition abortion
The question as written is: “Leaving aside whether or not you think abortion should be legal, are you morally opposed to abortion or not, or would you say it depends?”

In nearly every region in the United States, the percentage of women morally opposed to abortion is greater than the share of men reporting the same. I was so surprised by these results I checked my code three times, but there you have it. It may be explained in part by greater religiosity among women. Nonetheless, it’s out of sync with the narrative that the push for a more restrictive stance on abortion is a manifestation of men’s desires to enforce their beliefs on women since by and large the women in question share these beliefs. (That said, the bills and laws, from what I can tell, are wildly out of step with the way Americans broadly think about abortion.)

I’ll wrap up by saying that I don’t believe the people decrying a war on women’s rights are being disingenuous. I know enough people, men and women, who hold this belief to know that’s sincerely the way they see things. I don’t know very many, possibly any, strictly anti-abortion people — New Englander here — but I think it’s necessary to take their claims on good faith, too. Presumably 30% of women in the South East are not hostile to women’s rights as they see them.

The Kids Are All Right: Follow-up

My post on the relationship between ideology and fertility rates generated some great feedback and critiques (albeit mostly on a Facebook thread). Sadly, none of this was related to the awesome pun in the title of the piece. (Seriously, no love for “The Kids Are All Right”?)

Well, life goes on.

In light of the interest in the subject, I’ve decided to do a quick follow-up piece to address some readers’ questions and adding a bit of information, particularly as relates to this graph from the original post:

Conservative have more kids

1. Are there more people on the political left?

A couple people asked about the ideological composition of the nation and the sample I used. This is an important question, because if the political right makes up a small enough minority of the population or sample, then my graph, which shows the average number of children per respondents of different ideologies but doesn’t convey sample sizes, is a bit misleading—or at least less compelling. So my fault for not going into it in the first place.

Per the most recent polling by Gallup, The American electorate identifies as roughly 26% liberal, 35% moderate, and 35% conservative. This is after two decades of a slow, steady increase in the percentage of Americans calling themselves “liberal.” More on that later.

The sample I pulled from the General Social Survey (GSS) reflects Gallup’s national numbers pretty well: out of the total 8,539 respondents sampled, 2,346 (27.47%) identified as some degree of liberal, 3,285 (38.47%) as moderate, and 2,908 (34.50%) as some degree of conservative.

sample distribution

2. Are there more women on the political left?

I believe this question is getting at the same idea: if the majority of women are left of moderate, then the higher fertility of women on the right is less consequential for the electorate. According to Gallup’s national numbers, 30% of women identify as liberals—the same percentage as call themselves conservatives. For men, those numbers are notably different: 40% and 21%, respectively.

The sample I used showed more gender parity in ideologies, but it’s not hugely off. At any rate, the important thing is that the elevated fertility rates of conservative women can’t be written off as the effect of a small sample size.

men and women political ideologies

3. But the population has been getting more liberal. Doesn’t that kind of throw a wrench in this narrative?

Only time will tell, I suppose! To be clear, this is how many people read the tea leaves, and the story I’m telling is a bit of heterodoxy. While I can’t offer a firm answer to this question now, I have a few remarks:

  • The past is no guarantee of the future. (Ask GE shareholders, amirite?) Just because the electorate has been getting more liberal doesn’t mean it will continue to.
  • I suspect the secular trend toward liberalization is as influenced by macroeconomics and sociological factors as it is individual characteristics and experiences. The question is, what will be the effects of today’s macroeconomic and sociological upheaval on future voters—or their children?
  • Relatedly, I think time horizon matters a great deal when evaluating whether or not the future looks liberal or conservative. This is theory on my part, but maybe populations naturally move to the right over the long term (because conservatives reproduce more) unless cultural forces pulling leftward—economic globalization?—are sufficiently strong and sustained.
  • Finally, I think it’s worth noting that America’s liberal ranks are mostly swelling at the expense of its moderate contingent, perhaps due to increasing political polarization.

americans becoming more liberal
America has been getting more liberal at the expense of its moderate contingent

*

There were a few questions brought up to which I don’t have good answers at the moment. Without promising when, if at all, to address them—I’ve learned not to make firm commitments relating to this fine blog—here are two excellent threads I should follow:

  • Does completing various life milestones (having children, buying a house, getting married) make people more conservative? There must be some longitudinal studies on this somewhere…
  • To what extent do children’s political views match their parents’, and is there symmetry between liberals and conservatives in this regard? (I linked to one study by Gallup in the last post that suggested a 70% match between parents and their children, but there’s probably a lot more work on this out there.)

The Kids Are All Right

A notion that’s become somewhat common among the left in recent years is that American conservatives are demographically doomed. Very often, this is discussed in terms of race and age—an allusion to the declining share of the non-Hispanic white American population and the accompanying erosion of its political sway, as well as the fact that conservatives tend to be older.

On its face, this seems like a tidy theory. But under the surface, I think it’s a great deal more complicated than the (sometimes wishful) theorizing of liberal pundits allows.

First, as I wrote a couple of weeks ago, the American concept of race, whiteness in particular, is a moving target. I think there’s a solid chance it will look quite a bit different in a couple of decades. But a more interesting foil, I think, to the liberals-own-tomorrow theory manifests itself in fertility rates.

People are having fewer children the rich world over, causing consternation among governments of countries whose economic futures depend on population growth. The political implications of this alone are fascinating, but the general trend obscures another interesting story: the intranational ideological disparity in fertility.

Liberals, it appears, are having fewer children than their conservative counterparts. Combing through General Social Survey (GSS) results from the last 10 years provides a clear view of this phenomenon. Among heterosexual men between the ages of 35 and 50, those who identify as “extremely liberal” had on average 1.79 children, whereas those describing themselves as “extremely conservative” had 2.43. For women, the difference was even starker: “extremely liberal” women had 1.69 on average to “extremely conservative” women’s 2.63.

Conservative have more kids

Many factors contribute to this disparity—most of which concern liberal women’s increased preferences for family planning. Anecdotally, liberals are later to marry, more likely to pursue advanced degrees, less religious, and, among women, more likely to pursue careers—the confluence of which makes for fewer babies. This trend seems to have gotten stronger over the past few decades.

Widening Child Gap
Note: Data for this graph are inclusive of all respondents 18 and older, meaning there’s probably some bias toward conservatives having more children, as they tend to be older. It’s also possible that causality flows both ways: having families might make people more conservative.

All this is significant because there’s at least some evidence that, most of the time, children end up inheriting their parents’ political views. This makes sense whether you view political ideology as a product of nature—differences in brain structure that give some a proclivity for novelty and others an aversion to risk—or nurture. Either way, if liberals are bearing and raising fewer children, it could mean fewer liberal adults down the line.

There are a few signs this might already be happening. One study found that after decades of decline, high school students’ support for traditional gender roles in the family has been rising steadily since 1994. Goldman Sachs pegs the ascendant Generation Z as especially fiscally conservative. Finally, a survey by the Hispanic Heritage Foundation of 50,000 14 – 18-year-olds found—shockingly, in my view—that the majority identified as Republicans and would support Donald Trump in the 2016 election. (An important caveat to this survey was that nearly a third of those polled would have declined to vote, had they been able.)

The policy preferences of an increasingly conservative nation are one thing—and obviously ideology colors one’s assessment of how good or bad that would be. But what really worries me is the thought of an even more politically segmented society; one in which an increasingly liberal minority of elites maintains control of the nation’s cultural power centers and an increasingly conservative majority grows frustrated with its obsolescence in the new economy, which, intentionally or not, places a premium on educational attainment, city-living, and delayed entry into family life.

Whether it comes to pass or not, we shall see.

Demographics Aren’t Destined

Last June, the Census reported the white-alone population to have declined by .2% in absolute terms between July 2016 and July 2017. Though it may seem trivial, this factoid has immense significance to those on opposing sides of the culture wars, both of which have taken it to herald the decline of white political significance and the rise of a more diverse, and therefore liberal, electorate.

Frankly, there’s too much there to talk about in one blog post. Instead, I’d like to address an issue that’s bugged me for a long time:

In projection after projection showing a minority-white America, Hispanic members of each racial category are separated and lumped into their own group, despite the racial diversity of Latin America. This is significant because the rising tide of American diversity is mainly the result of a four-decade wave of immigration from Latin America and the high fertility rate of their descendants (though both forces have recently calmed). In 1960, 3.5% of the country identified as Hispanic or Latino. Nearly 60 years later, that figure has risen to 18%, with expectations that a quarter of the country will identify as Hispanic by 2065.

But disaggregating Hispanics from racial categories is inconsistent, not just with official Census convention—which designates Hispanic/Latino an ethnicity, a variable mutually independent from race—but also with evidence that suggests many Hispanics are beginning to assimilate more wholly into the white population.

This isn’t a (just) pedantic rant about Census data. I think there’s a solid argument to be made that we’re actually in the middle of an expansion, rather than contraction, of American whiteness.

Take, for starters, that a slim majority of American Hispanics already identify as white, at least when asked about their race on the Census. This doesn’t seem like a vestige of a more racially animose time. Per the 2010 Census, 53% of US Hispanics describe themselves as “white alone,” up from 48% in 2000.

US Hisp Race
Between 2000 and 2010, the share of US Hispanics identifying as white alone increased, while the proportion selecting “some other race” when asked to identify decreased.

Secondly, Hispanic identity seems to fade the further removed from immigration a one is. According the Pew Center’s 2015 National Survey of Latinos, all but 3% of foreign-born Americans with Hispanic ancestry identify as Hispanic or Latino. In the second generation, that share increases only slightly, to 8%. But by the third and fourth generations, it climbs rapidly, to 23% and 50%, respectively. This is truer among younger cohorts.

All told, 11% of US adults with Hispanic ancestry do not identify as such. Because immigration has been replaced by native births as the main driver of US Hispanic population growth in the last few decades, it’s not unreasonable to expect this fraction of “non-Hispanics” to grow.

Hisp growth
Source: Based on Pew Research Center tabulations, Pew Research Center historical projections (Passel and Cohn, 2008).

Also worth considering are Hispanics’ growing geographical dispersion and high rate of intermarriage, especially among younger generations. Twenty-eight percent of 18- to 35-year-old US Hispanics are married to non-Hispanics. Again, this trend grows stronger the longer one’s family has been in the United States: nearly 60% of third-generation US Hispanics ages 18 to 35 are married to someone who isn’t Hispanic. Moving out of the city and marrying extra-ethnically seem, admittedly conjecturally, indicative of cultural assimilation.

It seems like Hispanics are following the arc of other (European and Levantine) immigrant groups who were once, and in some cases still are, considered outside the bounds of conventional whiteness. All of this is to say, I’m skeptical that the way Hispanics view themselves in 2018 is the way they will in 2050—especially as they become more enmeshed in mainstream American society.

Of course, this is just a prediction I’m making in my living room. I don’t have a crystal ball or any special insight into the minds of the American public. I’m going to end with some reasons things might not go as I imagine:

The 101 reason would probably be “politics,” with which race seems to have a bicausal relationship in America. It’s not hard to imagine the Republican party alienating Hispanics with nativism while selling themselves, intentionally or otherwise, as the party of White America. Similarly, Democrats’ ability to court Hispanics relies to some degree on the extent to which they feel shut out from the cultural and political mainstream. Both could push Hispanics to think of themselves as non-white more frequently.

Relatedly, the Office of Management and Budget could affect Hispanics’ racial identities through bureaucratic means. A few years ago, there was talk of combining the race and ethnicity questions, with “Hispanic” offered as a choice alongside Asian, black, white, etc. As I noted at the time, this might bring the Census questions more in line with the way Americans think about race today—but it would also be putting a thumb on the scales. There’s really no neutral position for the Census to take in this matter.

Anecdotally and finally, it also seems like the psychic benefits of whiteness have waned a lot over the last few decades—especially as regards low-status whites. Part of this owes to good news: cultural progress on matters of race, which has begun to erode the relatively elevated status enjoyed by whites at the expense of minorities. Other explanations are more sinister and reflect anomic decay in the white population: rising rates of suicide, drug overdoses, and voluntary unemployment. For one reason or another, whiteness no longer feels as enviable a club as it probably did in the 20th century when Italians, Jews, and other so-called “white ethnics” made the conscious effort to join its ranks.

A Glance at College Attainment in NYC

I’m taking a couple days off from work this week. Unfortunately, it’s raining, so I’m inside playing around with some county-level education data in R, and I thought I’d throw up a quick blog post. The data set, which goes back to the 1970s, comes courtesy of the USDA and can be found here.

A quirk of New York City is that each of the five boroughs is also its own county. I took the opportunity to make some graphs illustrating how educational attainment has changed in the city and its boroughs over the past 50 years. There are a few interesting insights to be had here:

New York City’s college attainment rate closely mimics the country’s. Since 1970, the percentage of residents over the age of 25 in both groups who have attended at least some college has risen from just over 20% to just under 60%.

US NYC

But between the boroughs, there’s quite a bit of diversity. Manhattan is the only borough to have ever had a higher than city-average rate of college attainment. What’s more, the gap between the Manhattan and the New York City average has only grown over time. That said, because it started out with a higher rate of college-educated residents, Manhattan has experienced the lowest rate of growth in this area. Here’s each borough compared with the city’s average over the last 50 years:

Boroughs Some College

Perhaps unsurprisingly, given its reputation as a hotbed of gentrification, Brooklyn leads the way in terms of educational growth among its residents. In 1970, Brooklyn and the Bronx had similar rates of residents with at least some college (12 and 13 percent, respectively), compared with a US average of 21.3%. By 2016, however, the gap between the former had widened to 10.5%. Brooklyn has surpassed Queens and is closing in on the NYC average.

A more detailed look at the change in educational attainment in Brooklyn, which now has slightly more residents with than without some college experience:

Rise of Educated Brooklyn