Parts and Wholes: Unpacking Reports of White Working-Class Death Rates
Chicago Working-Class Studies
The white working class has been getting a lot of attention lately — not just for how they’re voting in primary elections, but also for dying at increasingly high rates. As we might expect, a lot of this attention is classist, especially when politics and death rates are discussed together, but even thoughtful and probing commentaries too often confuse parts and wholes, leading to loose generalizations that couldn’t possibly be true.
The discussion of death rates was initiated late last year by economists Anne Case and Angus Deaton who found “rising morbidity and mortality in midlife among white non-Hispanic Americans.” I suspect one reason their study attracted so many political commentators, including some who tried to link it to support for Trump, was because Case and Deaton define the white working class the same way political analysts do – whites with less than a bachelor’s degree. The man-bites-dog news value of the study, however, was that it showed death rates of a large group of Americans increasing rather than decreasing. Since 1900 life expectancy at birth has risen from 47 to 79 years, nearly doubling the average American lifespan. So it was big news that the death rates of U.S. whites aged 45 to 54 had increased by 8% from 1999 to 2013.
When Case and Deaton divided this white cohort by educational attainment, however, they found that all of the increase was accounted for by increased deaths among whites with high school educations or less. What’s more, this group of whites had substantially higher death rates than blacks or Hispanics in that age group, including much higher rates of deaths from drug overdoses, suicides, and chronic liver cirrhosis.
Mortality rates in 2013 for persons aged 45-54, organized from low to high (deaths per 100,000 population)
| “Racial”/Ed. Group | All-cause mortality | Poisonings
(drug OD) |
Intentional self-harm
(suicide) |
Chronic liver cirrhosis |
| White non-Hispanic w/BA or more |
178.1 |
8.1 |
16.2 |
6.98 |
|
Hispanics (all races) |
269.6 |
14.4 |
7.3 |
23.1 |
| White non-Hispanic w/some college |
287.8 |
20.6 |
19.6 |
14.9 |
|
Black non-Hispanic |
581.9 |
21.8 |
6.6 |
13.5 |
| White non-Hispanic w/high school or less |
735.8 |
58.0 |
38.8 |
38.9 |
SOURCE: Compiled from Case & Deaton, Table 1, p. 3.
This was startling news because we are so used to seeing blacks and Hispanics at the bottom of these kinds of lists. These “racial” minority groups consistently have much higher rates of unemployment and poverty – often double and triple white rates — and lower average incomes and much lower accumulated wealth. Why would any group of whites be dying at higher rates than minorities, let alone killing themselves or poisoning themselves with drugs and alcohol?
Speculations about causes range from broad-based economic factors to the psychological impact of crushed expectations. A few have even suggested that folks with white-skin privilege are not as resilient in dealing with hard times as blacks and Latinos, and others have claimed that personal morality has “collapsed” in the white working class. These speculations move very carelessly from one white age cohort – those who would have graduated high school between 1977 and 1986 – to the white working-class as a whole.
A confusion of parts and wholes, however, began with Case and Deaton themselves, as we can see in the table above. Whites are by far the largest group, so it makes sense to break them into three parts by educational attainment, but that leads Case and Deaton to compare all Hispanics and all blacks with three separate segments of the white population. It could well be that blacks and/or Hispanics with only high school or less have even higher death rates than comparable whites. We can’t tell because parts are being compared to wholes.
But the reverse is just as important: comparing patterns between white and black, with no recognition of class differences, erases substantial differences in life conditions and life chances among whites. Dividing the white population by education reveals that white-skin privilege may not be all it’s cracked up to be among the largest group of American whites – those with only high school educations or less.
Janell Ross recently provided a thorough rundown of black-white disparities in The Washington Post: “On just about every measure of social or economic well-being, white Americans fare better than any other group. That’s true of housing and neighborhood quality and homeownership. That’s true of overall health, health insurance coverage rates, quality of health care received, life expectancy and infant mortality. That’s true when it comes to median household earnings, wealth (assets minus debt), retirement savings and even who has a bank account.” Ross’s bouquet of links, based on very solid sources, documents an appalling degree of racial injustice, especially toward blacks. But, unlike Case and Deaton, these sources all compare the entire white population with the entire black and Hispanic populations, with no internal differentiation. As with death rates, all these disparities might look very different in a five-category comparison like Case and Deaton use. I’m betting, for example, that whites with only high school educations or less have nowhere near the “typical” white family’s wealth of $131,000. Routinely differentiating the white population by educational attainment would not show that we overestimate racial injustice, but it would almost certainly show that we grossly underestimate class injustice.
Differentiating the white part of the population by three levels of educational attainment provides a somewhat surprising profile of the American population, I think, even if it downplays class differences within other racial groupings:
“Racial” composition of U.S. population, 18 & over, in 2014, with “class” by educational attainment for non-Hispanic whites
| “Racial”/ed. Group | % of U.S. pop. | # in millions |
| White non-Hispanic total | 65.5 | 156.8 |
| WNH w/BA or more | 21.8 | 52.2 |
| WNH w/some college | 19.4 | 46.4 |
| WNH w/ high school or less | 24.3 | 58.2 |
| Hispanic | 15.2 | 36.4 |
| Black non-Hispanic | 12.3 | 29.4 |
| Asian | 5.6 | 13.4 |
| Others | 1.4 | 3.4 |
SOURCE: Compiled from U.S. Census Bureau, Educational Attainment in the United States: 2014
One surprise, I suspect, is that whites with only high school educations or less are the single largest group – nearly a quarter of all Americans 18 and older, some 58 million adults. This is not some small leftover group atypical of whiteness in the 21st century, and if they have high and increasing death rates, that’s not a “pocket of poverty” problem. Combined with much larger percentages of blacks and Hispanics with only high school educations or less, who also likely have higher death rates than what Case and Deaton report for blacks and Hispanics as a whole, they constitute more than 40% of our adult population. These are not canaries in a coal mine – they’re the miners, and the mine walls are collapsing.
It should be obvious that the entire group of whites without bachelor’s degrees, nearly 105 million people (adding those with “some college”), is too large to possibly share a single personality type, a uniform social psychology, or any one political ideology – as so many commentators are wont to assume. Why do critics insist on making judgments about working-class resilience and morality based on a handful of misunderstood facts?
Rising aggregate death rates, and especially those related to self-harm, drugs, and alcohol, are indicators of increasing stresses being experienced by a population. But only about 1/10th of one percent of whites without bachelor’s degrees are killing or poisoning themselves. They tell you nothing about how most people in that population are dealing with those stresses. Within my own extended white working-class family, a small sample to be sure, the addicted and de-moralized are a decided minority, with sometimes dramatic changes across their life stages. The vast majority, given the destabilizing challenges they’ve faced, demonstrate near-heroic levels of personal morality and resilience – and they rightly feel considerable pride in their capacity for “taking it.” If anything, their sturdy commitment to these admirable qualities may undermine their capacity for the kind of broader collective action that could change their fates – fates that could and should require a lot less resilience.
***
This has been reposted from the Center for Working-Class Studies.
