Harvard Researchers Also Know What Is Best For Your Children. Sort Of. Maybe.

Yesterday I talked about how the Colorado Department of Public Health knows, in creepy-abusive-spouse detail, what is best for your kids. Sometimes the government realizes all by itself that it is a better parent than you are. But other times the government has this epiphany as a result from prodding from its partners in academia.

Take, for instance, Harvard. Harvard has long been a source of Deep Thinkers with Important Ideas about government's role in society, leading to examples of well-thought-out government intervention like the Vietnam War. We've discussed the sort of insight that comes out of Harvard before.

Today, courtesy of the redoubtable Walter Olson, I saw a new instance inflaming the news media and blogosphere today: two Harvard researchers have written a commentary in the Journal of the American Medical Association called "State Intervention in Life-Threatening Obesity" suggesting that obese children should, in some instances, be removed from their parents by the state for the good of the children. The Harvard folk are David S. Ludwig and Lindsey Murtagh (hat tip to Future of Capitalism for those links).

My first instinct, like that of Walter, is outrage and contempt and all the other pulse-elevating feelings on the I-must-blog-about-this spectrum. My second instinct, though, is to ask this: exactly what are these Harvard researchers advocating?

I knew in advance not to expect to learn much from the news media. Like legal reporting, science reporting is so inadequate that it is the butt of jokes in comics — at least the sort of comics read by the sort of people who care about the quality of science reporting.

My expectation was not thwarted. The reporting on the article was vague and ambiguous and failed either (1) to educate me about the exact scope of the custody-meddling that Ludwig and Murtagh are advocating, and (2) to convince me that this wasn't the sort of OMG MAD SCIENCE! reporting that the comics linked above lampoon.

JAMA does not provide full text of its articles online. I was therefore thwarted in my attempt to figure out whether Ludwig and Murtagh are respectable clinicians or nasty totalitarians or something in between. I was ready to write about this anyway. Then I thought how can I complain about lousy science reporting while perpetrating it myself? So I bit the bullet and paid $30 to download the JAMA article. I comfort myself that (1) JAMA is, in general, a good cause, (2) this isn't the stupidest thing I will have spent money on this week, and (3) the Popehat promise is that we read things like JAMA so you don't have to.

I was concerned that the JAMA article might be beyond my little political-science-major brain. I should not have worried. "State Intervention in Life-Threatening Obesity" is a puff piece suitable for a popular magazine or newspaper and largely free of substance. With footnotes, it's two two-column magazine pages long. It does little but set up the existence of childhood obesity in America, discuss in brief some of the health risks of such obesity, note that existing law generally allows some state intervention in parenting choices when a child's health is at risk, and opine quite generally that in some extreme cases custodial interference by the state may be warranted. This is as specific as it gets:

In severe instances of childhood obesity, removal from the home may be justifiable from a legal standpoint because of imminent health risks and the parents’ chronic failure to address medical problems. Indeed, it may be unethical to subject such children to an invasive and irreversible procedure without first considering foster care. Nevertheless, state intervention would clearly not be desirable or practical, and probably not be legally justifiable,
for most of the approximately 2 million children in the United States with a BMI at or beyond the 99th percentile. Moreover, the quality of foster care varies
greatly; removal from the home does not guarantee improved physical health, and substantial psychosocial morbidity may ensue. Thus, the decision to pursue this option must be guided by carefully defined criteria such as those proposed by Varness et al, with less intrusive methods used whenever possible.

In short, Ludwig and Murtagh seem to be doing little more than attempting to "start a conversation" about when obese children should be removed from their parents by the state. For specific diagnostic criteria, they punt to an article in Pediatrics called Childhood Obesity and Medical Neglect; that article is available fully online through the link. That article is fairly cautious:

In our opinion, 3 conditions must be met to justify state intervention, that is, a high likelihood of serious imminent harm, a reasonable likelihood that coercive state intervention will result in effective treatment, and the absence of alternative options to address the problem. In the case of childhood obesity, it is not the mere presence or degree of obesity but rather the presence of comorbid conditions that is critical for the determination of serious
imminent harm. All 3 criteria are met only in very limited cases, that is, the subset of obese children who have very serious comorbid conditions and for whom all alternative options have been exhausted. In these limited cases, a trial of removal from the home to protect the child may be indicated.

The linked Pediatrics article gives some examples of types of health risks.

Supporters of this research could reasonably argue that this standard is not really different from what courts and child authorities would apply if children were facing a risk other than obesity. In short, to the extent that media coverage of this article suggests that Harvard researchers say that the government should take your kids away if they are fat, that coverage is woefully deficient.

But that doesn't let Ludwig and Murtagh off the hook.

The proof of the pudding is in the eating; the proof of the regulatory scheme is in its drafting and implementation. There are good reasons to be highly suspicious of any social movement — driven by medical professionals or not — to encourage interference with parental custody.

The medical community is an increasingly popular vector for social agendas and state control of the individual. Norms about government control start narrow in theory — like "in the most extreme cases, some morbidly obese children facing imminent serious health consequences might be taken from their parents' homes if those parents refuse to address the health problems." In practice, though, the norms get broadened and twisted to justify social, political, and religious agendas, to increase government fiefdoms, and to wage cultural wars. The norms are implemented not by the knowledgeable professionals who envisioned them, but by bureaucrats of mixed capacity and motives under regrettable conditions. The result, amongst actual people, is inevitable: norms permitting limited state intervention into the lives of citizens are capriciously and maliciously applied, usually in a disproportionate fashion against the unpopular or the powerless. Moreover, when the state trains and equips to exercise power upon some justification, it tends to see that justification whether or not it exists. In our society, the natural and probable result of normalizing state custodial interference of extremely obese kids is not sparing application; the natural and probable result is abuse and misapplication.

Ludwig and Murtagh may well take the Beckian stance that they are "just asking questions" — simply raising the topic for greater discussion, not advocating any hasty broadening of state power over citizens' children. I interpret their brief column as exactly that — an attempt to start the discussion and lend JAMA's credibility to the proposition that forcible removal of obese children is supported by medical academia at the highest level. But the state always thirsts for such power, and is always too eager to get academic support for seizing it. Sooner or later, Ludwig and Murtagh will be cited in some local bureaucrat's application to take a child from his or her parents. Will that child meet the narrow diagnostic criteria that Ludwig and Murtagh have in mind? Perhaps — and perhaps not. The strong possibility that those footnote-referenced diagnostic criteria will be neither understood nor met is, I submit, a factor that Ludwig and Murtagh have not weighed adequately. They have started the dialogue, but distant, barely-supervised and poorly-restrained state officials will finish it for them. Ludwig and Murtagh might also have suggested that parents could buy vicious attack dogs to scare their fat children away from the refrigerator; though the dogs could be trained, in theory, to do so in a safe and appropriate manner, the natural and probable result is blood on the floor.

Last 5 posts by Ken White


  1. delurking says

    "Nevertheless, state intervention would clearly not be desirable or practical, and probably not be legally justifiable, for most of the approximately 2 million children in the United States with a BMI at or beyond the 99th percentile. "

    There are not 2 million children in the US with a BMI at or beyond the 99th percentile. That would imply there are 200 million children in the US. Then again, they are Harvard researchers. Maybe they think of the vast majority of Americans as children.

  2. says

    I can't afford a platinum membership (yet) but I feel kind of guilty that you came $30 out of pocket for this. Does Popehat have a Paypal account for sums smaller than $750,000?

  3. Xenocles says

    Whenever I hear the phrase "researchers recommend that policymakers…" I reach for my gun.

    (What if they recommend that a policy be loosened? Let me know when it happens and I'll tell you how I feel.)

  4. C. S. P. Schofield says

    Both Doctors and Harvard graduates and/or employees have a tendency to confuse themselves with God. A Doctor who IS connected to Harvard is probably so delusional that he thinks he pisses rosewater.

  5. says

    CSP, I read that on my iPhone as "piss rottweilers," and I was thinking that would be an AWESOME superpower. Before you attacked your enemy's secret lair you would drink, like, three Big Gulps. But I sure hope they come out tiny and get big after.

  6. C. S. P. Schofield says


    That would certainly make passing a kidney stone feel trivial by comparison.

  7. says

    The problem, as always, is the creeping incrementalism. First, it's okay to take kids who are morbidly obese and suffering health problems. Next, it's just the fat kids at risk of health problems. Finally, CPS is at your house because somebody thinks your kid is too pudgy or eats poorly at school. For all the talk of 'slippery slope' being a logical fallacy, absent any limiting principle it's usually a prediction.

  8. says

    A Harvard story:

    My partner, who is extremely into Harvard, holds Harvard Club meetings here at the firm. I do not attend. Last night, the partner was away, but the Club held its meeting anyway.

    An older gentlemen arrived about half an hour early and was not-quite-shouting at our clerks, who were sitting at reception, about how some black man had done him wrong. He became dissatisfied with their response, and started hollering at a male clerk instead.

    I've managed to piece together what got him upset. Apparently he pulled up to the parking garage of our skyscraper — which is 50 stories high and contains dozens and dozens of businesses — and told the parking lot attendant that he was here for the Harvard Club meeting and demanded to know where it was. The parking lot attendant, not surprisingly, is not kept in the loop about what tenants of the building are hosting meetings for which clubs. Not even the Harvard Club. But the old guy claims the attendant told him he was in the wrong building and waved him down the street. This might have been a miscommunication — the old guy is hard of hearing. On the other hand, the attendant may have been confused — the Jonathan Club and the California Club, two clubs widely associated with old, white, poorly socialized rich men, are indeed just down the street. Anyway, this led the old guy to be only a half hour early instead of an hour early, and led to his rant about the black guy downstairs, apparently.

    What kind of building are they running here if the parking lot attendant doesn't know where the Harvard Club is meeting?

  9. Mike says

    Delurking – that was my first thought as well – its impossible for 2 million kids to be in the 99th percentile.

    I think the answer is that they use a chart prepared in the 1950s that tells them for a 5 foot tall kid, median weight is 100 pounds, 75th percentile is 120 pounds, 99th percentile is 140 pounds…

    So they say that any 5 foot tall kid that comes in today weighing more than 140 pounds is above the 99th percentile – even if that's 50% of kids.

    But you're right – it does scream innumeracy.

  10. Mike says

    But more troubling is the use of percentiles to determine risk. Leaving aside the innumeracy, the researchers are suggesting that they're justified in taking 1% of children into foster care solely based on the fact they're at the end of a normal distribution curve. No messy proof of harm or anything, just height and weight.

    And the dirty thing about statistics is that no matter how healthy the whole population is, 1% of kids are always on the wrong side of the 99th percentile. This program goes on forever, no matter how much broccoli we make them eat.

  11. says

    > What kind of building are they running here if the parking lot attendant doesn’t know where the Harvard Club is meeting?

    I could certainly understand the parking attendant not knowing where the Cornell Club meeting was…or perhaps even the Yale Club meeting… but the HARVARD meeting?

    Looks like you've not had the right kind of people explain things to you yet.

  12. says

    Mike: Actually, they're not advocating (at least openly) doing it based on percentage alone. They incorporated certain diagnostic criteria by reference, which I quoted and linked above.

  13. Xenocles says

    T, slippery slope is a logical fallacy. As that name implies, it has little relation to human behavior.

  14. Xenocles says

    I worded that awkwardly. The gist is that human behavior tends to have little to do with logic.