Skip to content
 

No-op: The case of Case and Deaton

In responding to some recent blog comments I noticed an overlap among our two most recent posts:

1. Mortality rate trends by age, ethnicity, sex, and state

2. When does research have active opposition?

The first post was all about the fascinating patterns you can find by analyzing and graphing data from the CDC Wonder website, which has information on all the deaths in the United States over a fifteen-year period. The post was motivated by the release of a new article by economists Anne Case and Angus Deaton, who pulled out a few things from these data and, from this, spun a media-friendly story of the struggle of white Americans. In that post, I emphasized that I think Case and Deaton’s work has positive value and that I hope journalists will use that work as a starting point to explore these questions more deeply by interviewing knowledgeable actuaries, demographers, and public health experts.

The second post explored how it was that now-disgraced eating-behavior researcher Brian Wansink managed to stay at the top of the heap, maintaining media exposure, government grants, and policy influence for something like 10 years even while his sloppy research practices were all in plain sight, as it were, in his published work. In that post I suggest one reason that Wansink stayed afloat all this time was that his claims were pretty much innocuous, and he was working in a noncompetitive field, so there was nobody out there with any motivation to examine his work with a critical eye.

And here’s the mash-up: Case and Deaton are writing about an important topic—mortality trends!—but their message is basically simpatico to all parts of the political spectrum. Struggling working-class white people, that’s a story that both left and right can get behind. There’s nobody on the other side!

Indeed, when the original Case/Deaton story came out a bit over a year ago, it was framed by many as an increase in the death rate of middle-aged white men, because that was what everyone was expecting to hear—even though the actual data (when correctly age-adjusted) showed a decrease in the death rate of middle-aged white men in recent years (the increase was only among women), and even though Case and Deaton themselves never claimed that anything was happening with men in particular.

The news media—left, right, and center—had a pre-existing narrative of middle-aged white malaise, and they slotted the Case and Deaton reports into that narrative.

Why did the media not interview any questioning voices? Why did we not hear from actuaries, demographers, and public health experts with other takes on the matter? Why no alternative perspectives? Because there was no natural opposition.

And it does seem that the news media need opposition, not just other perspectives. After the original Case and Deaton paper came out, I did some quick calculations, then some more careful calculations, and realized that their headline claim—an increase in mortality among middle-aged white Americans—was wrong. But when I wrote about it, and when I spoke with journalists, I made it clear that, although Case and Deaton made a mistake by not age adjusting (and another mistake by not disaggregating by sex), their key conclusion—their comparison with trends among other groups and in other countries—held up, so I was in agreement with Case and Deaton’s main point, even if I thought they were wrong about the direction of the trend and I was skeptical about their comparisons of different education levels.

Journalists’ take on this was, pretty much, that there was no controversy so everything Case and Deaton said should be taken at face value.

I don’t think this was the worst possible outcome: based on my read of the data, Case and Deaton are making reasonable points. I just wish there were a way for their story to motivate better news coverage. There are lots of experts in demographics and public health who could add a lot to this discussion.

As I wrote in my earlier post, Case and Deaton found some interesting patterns. They got the ball rolling. Read their paper, talk with them, get their perspective. Then talk with other experts: demographers, actuaries, public health experts. Talk with John Bound, Arline Geronimus, Javier Rodriguez, and Timothy Waidmann, who specifically raised concerns about comparisons of time series broken down by education. Talk with Chris Schmid, author of the paper, “Increased mortality for white middle-aged Americans not fully explained by causes suggested.” You don’t need to talk with me—I’m just a number cruncher and claim no expertise on causes of death. But click on the link, wait 20 minutes for it to download and take a look at our smoothed mortality rate trends by state. There’s lots and lots there, much more than can be captured in any simple story.

P.S. Just to emphasize: I’m making no equivalence between Wansink and Case/Deaton. Wansink’s published work is riddled with errors and his data quality appears to be approximately zero; Case and Deaton are serious scholars, and all I’ve said about them is that they’ve made a couple of subtle statistical errors which have not invalidated their key conclusions. But all of them, for different reasons, have made claims that have elicited little opposition, hence unfortunate gaps in media coverage.

17 Comments

  1. Paul says:

    “Why did the media not interview any questioning voices? … Why no alternative perspectives? Because there was no natural opposition.”

    Your point is unclear IMO — is it merely the standard lament that the media/journalists are lazy/biased & don’t do their job well… or that there just isn’t any “natural opposition” for journalists to earnestly consult ?

    And your term “natural opposition” seems quite vague, if that’s your point.

    • Andrew says:

      Paul:

      My point is that Case and Deaton know a bit about demography and public health, but these are important topics and there are other experts that journalists should talk with, rather than simply interviewing Case and Deaton and thinking that’s the last word.

      • Andrew says:

        Or, to put it another way, be wary when the existing narrative has no major opposition, as it’s easy to then misinterpret evidence to keep the story rolling forward.

        • Keith O'Rourke says:

          Not sure a boxing analogy will help many, but its like training with sparring partners that can’t hit back very well – its a lot of fun and makes you feel good – but you’ll pay dearly for it when you encounter a real opponent.

          Doing science requires real and constant (but manageable) opposition.

    • Steve Sailer says:

      Here’s Case and Deaton’s latest from last week:

      https://www.brookings.edu/bpea-articles/mortality-and-morbidity-in-the-21st-century/

      “Case and Deaton are writing about an important topic—mortality trends!—but their message is basically simpatico to all parts of the political spectrum. Struggling working-class white people, that’s a story that both left and right can get behind. There’s nobody on the other side!”

      Or you could look at this question from the opposite perspective: Why in the world didn’t anybody notice the White Death before November 2015? Would anybody have noticed it yet if Deaton hadn’t won the Nobel quasi-price in Economics in October 2015?

      Q. Why wasn’t this noticed years earlier?

      A. There’s nobody on the side of white people, qua white people!

      There are all sorts of government agencies and NGOs that scan statistics looking for unfortunate disparities involving blacks and, to a somewhat lesser extent, other peoples of color. But there are no respectable organizations whatsoever looking out for the interests of whites. To be in the business of caring about white welfare is a good way to get you on the Southern Poverty Law Center’s popular list of people to hate.

      • Steve Sailer says:

        Obviously, it’s an argument over whether the glass is part full or part empty …

        But nobody seemed to recognize the existence of the White Death before Deaton’s 2015 (quasi) Nobel Prize gave his (admittedly limited) paper News Interest. But then it turned out to be hugely import.

        Here’s my Taki’s Magazine column on Case and Deaton’s latest

        http://takimag.com/article/white_privilege_vs_white_death_steve_sailer/print#ixzz4chThjw5X

      • Andrew says:

        Steve:

        Fair enough. Demographers were aware that death rates of middle-aged whites were no longer declining, but this finding hadn’t made it into the general conversation until Case and Deaton wrote that paper. It helped that (a) the news media have lots of economics reporters who have the habit of deferring to econ professors, and (b) Deaton won the Nobel prize right around then. I remain frustrated that Case and Deaton said that the death rates for middle-aged whites were increasing, when this wasn’t actually true. (To not adjust death rates for age is the demographic equivalent of presenting prices in raw dollars and not adjusting for the consumer price index or something like it.) As I wrote, Case and Deaton’s main point holds up fine if you replace “mortality rates are increasing!” with “mortality rates are flat” or “mortality rates are increasing for women in the south and flat or slightly decreasing for other groups.” But then they lose the headline and they lose their connection to the popular struggling-white-men story. More recently, their analysis conditional on education has potential big problems due to selection bias. Had Case and Deaton done the appropriate analyses and given the appropriate cautions from the start, maybe their results wouldn’t have had the same publicity.

  2. Jordan Anaya says:

    I thought you were exaggerating the file size.

    Protip: do not try to open that in your browser.

  3. Jonathan says:

    People laugh at the concept of “alternative facts” but you’ve just given examples, meaning that how things exist and how they are viewed occurs in contexts and narrative storylines which can sometimes run roughshod over objective or non-context specific evaluation. An underlying theme to your blog might be that there are objective probabilities – which can at least be approached with care – and then there are abstractions of those probabilities, which tend from relatively accurate to wholly inaccurate, and which achieve life as abstractions, meaning as “alternative” versions of the underlying context-independent or context less specific rendering.

    • zbicyclist says:

      I heard this described as the philosophical difference between facts (objective) and truth (based on facts, but an inference).

      There can clearly be alternative truths (exhibit A: world religions). Alternative facts, though, are an abomination.

      • Martha (Smith) says:

        I would make a big distinction between an inference and a truth — inferences involve an often large degree of subjectivity.

        • Martha (Smith) says:

          Also, I would classify some inferences (in particular, religions) as “beliefs”. Statistical inferences involve a degree of uncertainty, which means one should not hold them as beliefs just because they are inferences from the data; holding an inference as a belief goes beyond an inference from the data.

  4. choncan says:

    two oversimplifications.

    in sports, I have often witnessed exchanges along the following lines:
    STUPID OLD MAN: player X is the best because of the fire in his eyes and the heart of a champion!
    YOUNG TURK: [argument from authority] you are an idiot! science shows player X is bad! player Y is better! because science! you are a science denier! you are an ignoramus!
    STUPID OLD MAN: [an ignoramus] [very stupid, easily-caricatured response]
    YOUNG TURK: [wind at his back, enjoying the power of murder] stupid idiot! science! stupid idiot! science! player Y!
    [time passes, further investigation reveals further measures of value/deeper complexity to past measurements — catcher framing, whatever — implying player Y is still better than player X, but player X is probably better than was thought, player Y was probably overrated somewhat, and in any case, player Z is the new hotness, player Z is the thing]
    YOUNG TURK: [thoughtful blog post, “rethinking player X, forget about Y, anyway Z!”]
    STUPID OLD MAN: …wtf?

    in political media, I have often witnessed exchanges along the following lines:
    TALKING HEAD/TELEVISION PRODUCER: [obeying “on the one hand, on the other, get both sides” instincts] some say climate change is a problem, others say it is a chinese conspiracy, we come back with experts bringing the debate to you, our treasured viewers.
    ENTIRE INTERNET: YOU MONSTERS!!!!!!! THAT’S NOT AN EXPERT THAT’S A DENIER!!!!!!!! CLIMATE CHANGE DENIAL!!!!!!!!! HOW DARE YOU MURDER THE EARTHMOTHER!!!!!!
    TALKING HEAD/TELEVISION PRODUCER: [various degrees of bafflement/defensiveness/general deep end of the pool panic and equivocation]

  5. Napata says:

    I think you are making a serious error. Yes they are scholars BUT their is clearly a sub text and that sub text few people are acknowledging. There was never any good reason to compare ‘white working class’ with non- whites. There was good reason to compare internationally. T his is clever propaganda because it legitimises a debate about a fictitious event… white workers suffering more than Black workers. Read in a scholarly way the paper does not quite say that but the message received by 90% of the population will be that and my contention is that is deliberate. By comparing with Black people it deflects attention and if it were not for the affront to Democratic sentiments might well be ‘universally acknowledged’. The authors asked why Black people in worse situation commit suicide less answered in effect that they were use d to it …despite the evidence being exactly the opposite i.e. Black people were dieing faster. No one challenged this because they all agreed BLACK LIVES DON’T MATTER … Further if they seriously bought their own argument they should have applied it to Black people in the past and identified a ‘sea of despair’ as the cause of high mortality and drug abuse…but they were not interested in consistency. Whatever the issue about white working class morbidity Black people should never have been brought into it ..they were deliberately and the motive was patently clear for anyone from the Black community to see. This was a political tract masquerading as a technical paper and most commentators evade the dark politics by focussing solely (as was intended by Deaton/Case) on the scholarly speech and not the political diatribe written into the subtext.

    • Kyle C says:

      I don’t endorse all of this, but I agree that C&D must have known they were feeding a narrative that Something Uniquely Bad happened to white working class men in the recent past (which most readers would assume correlated with the post-2008 economy), and now they realize this isn’t true, for the reasons Andrew and others have given, but they won’t say so, probably because the media attention that this particular narrative brings them in the wake of Trump’s election is just too gratifying.

  6. Jonathan says:

    Not sure if you were aware of this but Anne Case did a longform interview on the whole controversy and a lengthy discussion on the methodological issues you have covered:
    http://podcast.ft.com/2017/04/21/anne-case-on-mortality-and-morbidity-in-the-21st-century/

    She is really honest about a lot of the decisions made. Making for a good interview.

Leave a Reply