Post #719: Why Latinos?

Source:  Commonwealth of Virginia Department of Health, COVID-19 dashboard, 6-16-2020.

This is just a quick note on the racial/ethnic mix of persons diagnosed with COVID-19 in Virginia, and in particular, in Fairfax County.  It is heavily skewed toward Latinos.  And I can’t quite put my finger on why, except to say that this dovetails with an age mix that suggests that working-age people are the ones at risk and being infected.


If we take the COVID-19 case counts, above, and match to the population divided the same way, here’s the relative prevalence of disease in Fairfax county.   The incidence falls heavily onto Latinos.

As far as I can tell, this is slightly (but only slightly) influenced by differences in age mix across these race/ethnicity categories.  This disease has tended to hit working-age people (as opposed to retirees), and Latinos are somewhat over-represented in that group relative to the population as a whole.  (I.e., under-represented among the elderly.)

What’s most startling is if you do the long division, and compare disease prevalence among Latinos to that of non-Latino Whites.  Roughly speaking, you are twelve times more likely to catch COVID-19 in Fairfax County if you are Latino than if you are a White, non-Latino person.

I have a hard time getting my mind around that large a difference.  I don’t believe that any genetic component to susceptibility could produce anything close to that large a difference.  But I don’t know.  I doubt that there are huge ethnicity-related differences in tendency toward large social gatherings, at least not at present.

Surely there’s an average income difference, as White non-Latino families have higher incomes that other ethnicities in Fairfax County.  But, interestingly, there’s little average difference between household income of Latino and Black non-Latino families — yet there’s a large differential in prevalence of COVID-19.

Source:  Racial Inequalities in Fairfax County, Urban Institute, December 2017, authors Leah Hendey and Lily Posey

One alternative is that this is a reflection of occupational mix.  Given that COVID-19 appears to be most prevelvant in the working-aged population, possibly the disproportionate impact on Latino population somehow is a reflection of occcupational mix?

From the same source as above, it does appear that a disproportionate share of the Fairfax Latino population works at low-wage jobs (defined here as full-time employment providing less than $35,000 per year income).

Source:  Same as above.

Geographic distribution may play some modest role in this.  In Fairfax County, the Latino population tends to be located in the more urbanized eastern portion of the county.  Still, prevalence differences across (say) ZIP codes are nowhere near large enough to account for the vastly higher disease prevalence among Latinos.

So this simply remains an odd puzzle.  Latinos in Fairfax County are vastly more likely to contract COVID-19 than any of the other identified race/ethnic groups.  COVID-19 largely hits working-age people.  And Latinos fill a disproportionate share of relatively low-wage jobs.

That said, what we are seeing here in Fairfax is just part of a broad national trend.  Based on this New York Times article, physicians in many locations in New York have noted the high prevalence among Latinos.   Ditto UtahNashville.  (Nashville?)  Charlottesville, VANorth Carolina.   MarylandDCCaliforniaNew OrleansAustinNew Bedford, MA?

The DC article just cited explicitly speculates that this is due to occupational mix.

"“Latinos are highly represented among health care workers and emergency providers, and among those who are still working in the actual workplace rather than at home,” he says, citing cleaning companies and restaurant and hospitality workers as among those at highest risk. “All of that increases their exposure to the virus.”

So this is hardly an isolated phenomenon.  The question is, does this really signal that certain occupations are at high risk, and that Latinos are so heavily concentrated among those workers that it skews the prevalence data by ethnicity this strongly?  If so, that has some pretty important implications for, at a minimum, what sort of protective equipment ought to be required for some common occupations.

Surely the Commonwealth has the information that would allow it to tabulate cases by occupation.  In Fairfax, is COVID-19 largely a disease hitting low-wage workers?  If that is the case, then for me, at least, that changes how I think about this, how I think about various forms of economic relief, and how I think about my own odds of contracting the disease.

As we progress with re-opening the economy, it’s high time for the Commonwealth to start providing more information.   It should take the data it has, on persons infected with COVID-19, and start providing the public with more detail on what affects the odds of infection.  How can anyone look at that 12-fold difference in prevalence of disease and not ask why that is occurring?