Unfree Choices: Hanania's Dubious Defense of Fat Shaming

Richard Hanania recently argued that being fat is a blameworthy “choice” that should be shamed.  This begs the question, “what is a choice?”  Hanania’s answer is interesting though ultimately unpersuasive.

He defines choice as “a human condition or behavior that responds to incentives.”  Accordingly, things like height, race, and biological sex are not choices because they do not respond to incentives.[1]  Whether to go on a walk or finish reading this essay are “choices” because observed behavior will strongly correlate with incentives.  How much people eat and how much they exercise respond to incentives.  Being fat is the result of eating too much and/or exercising too little.  Accordingly, by the Hananian definition, being overweight is a choice.

Hanania’s definition makes it easy to explain why Americans have become fatter over the last 60 years.  Buying a calorie requires less labor than it used to.  Physical work has become rarer.  Norms against bullying, including fat-shaming, have become stronger.  Because fat shaming is unpleasant, people endeavor to avoid it, decreasing obesity.  Less fat shaming means fewer incentives towards self-control and greater obesity.  Hanania does not mention this, but declines in the number of marriageable men and increases in female earnings give today’s women less reason than their grandmothers to diet and exercise.

Hanania’s strange definition of “choice” lets it flourish in the absence of free will:

I’m willing to go all the way and reject free will, accepting that you are a combination of your genes and your environment, and you didn’t “choose” either one, whatever that means. But then you have to accept this view for everything. You can’t say that murder, rape, going to the opera, and racism are choices, but absolve the overweight of any responsibility for their situation.

Most people conceptualize “choice” in ways that are either inconsistent or incoherent.  If humans lack free will, then murderers and rapists lack free will, and those who want to justify punishing murderers and rapists should ditch the language of choice if they care about being truthful.

Indeed, jettisoning the language of choice is not so difficult.  Suppose Xerxes is trying to murder Sam with an AK-47.  Sam has no good reason to care whether Xerxes’ actions are the product of structural racism, bipolar disorder, or the application of the categorical imperative by a noumenal subject imbued with free will.  Sam’s urgent and overriding interest is in staying alive, and few would pause to query Xerxes’ volitional faculties or blameworthiness.  The best reason to examine the inner workings of Xerxes’ mind is to understand and thwart his tactics.

Hanania’s core claim is solid: rejecting free will does not render incentives impotent.  Determinism means that when a situation is specified down to the tiniest subatomic detail, there is only one possible outcome.  The catch is the world is so physically complex that the exact same material arrangement will never occur twice.  My brain will never again have the exact same constellation of neural connections it has right now.  Indeed, my brain subtly changes with every evanescent thought and every keystroke.  I experience different moods and feelings – contentment, anger, love--  but these words are clumsy tools for describing the imponderable complexity of my brain.  I loved French fries 34 years ago and I love them still, but my brain has formed and discarded trillions of neural connections in the interim.  I can never again love deep fried potatoes exactly the way I did 34 years ago, and, though I can tantalize myself through mild hunger, I can never relive my youthful joy of eating a huge order of fries after I made football weigh in.

Psychology, at its best, lets us predict human actions by drawing upon a vast reservoir of prior human experience.  Any psychological category is squishy because no psychologist has ever come close to understanding every neural connection in any human’s brain.  No two brains are identical.  Bill and Fred might both truly have bipolar disorder, but that hardly means their brains are wired the same way.  All psychologists can really do is compare patients to different archetypes, knowing that if a patient and an archetype have acted similarly, they will probably respond similarly to treatment.  The strength of these correlations is an empirical question.

Hanania is correct to focus on incentives because creating beneficent incentives -- beneficent meaning no more than “desireable” — correlates with beneficent outcomes.  However, retaining the word “choice” while rejecting free will is a dubious rhetorical trick.  Hanania’s definition of choice is flatly inconsistent with the dictionary definition.  In normal English, choice implies free will.  The Oxford English dictionary I looked at on google defines choice as “an act of selecting or making a decision when faced with two or more possibilities.”  If humans lack free will, there is only one possible future and no human has ever been “faced with two or more possibilities.”  Many things may seem possible, but this is only because we lack the knowledge and computational power to model how every subatomic particle will behave.  In any event, choice implies the ability to have acted otherwise and, therefore, requires free will.

Choice, blame and punishment are an ancient triad.  Many western humans are more comfortable punishing those who have “chosen” to act wickedly than those who were simply born that way or coerced.  Ask a prosecutor for clemency, and their go to reply is that the defendant should have acted differently.  Hanania knows deep down that the psychological link between “choice” and punishment will survive his redefinition.  Through clever rhetoric, he recruits Christian/Kantian intuitions about punishing bad choices while rejecting the possibility that evildoers could have acted otherwise.

That a behavior responds to incentives does not imply that any human being would or could have acted differently under the actual incentives he faced.  Fat people would probably weigh less had they been relentlessly shamed since birth.  That does not prove fat people could have responded to their actual incentives differently.  At instant X, every human being has the only BMI he or she could possibly have at instant X.  Brains and incentives will change, and waistlines oscillate with them, yet only one present and only one procession of future states is possible.

Hanania’s defense of fat shaming is deeply incomplete.  Even accepting his core claim that fat shaming reduces obesity, it does not follow that the benefits outweigh the cost.  The real question is whether fat shaming produces more mental anguish than it averts.  This is a deep, empirical question which Hanania never engages.  Nor is all fat shaming created equal.  Saying that big sodas should be taxed is different than exalting thin women in bikinis.  Lavishing praise upon the athletic is gentler than refusing to hire a fat woman for an office job, which is douchey but far more defensible than publicly flogging anyone with a BMI over 30.  Nor is shaming the only possible disincentive to obesity.  Building awesome public parks and trimming hours spent at sedentary jobs would surely decrease obesity.

I wish that, rather than defending fat shaming, Hanania had thought about humane nudges towards losing weight.

---

[1] This is a little squishy.  Height could respond to incentives.  While maximum height is genetically determined, actual height can be stunted by malnutrition.  If parents were offered $10,000,000 cash prizes to keep their children a few inches shorter than the modern averages, some would manage to starve their kids enough to win the prize, and average height would decrease somewhat.

It’s better to view Hananian “choice” as a spectrum of elasticity to incentives rather than as a binary descriptor.  An act or condition is a choice to the extent it responds to incentives.  It would take huge incentives to reduce average height by a couple inches, so height is inelastic to incentives.  Conversely, the number of people who attend a concert will vary materially with the price of tickets and quality of the concert.  (Good music is an incentive to go, steep ticket prices are a disincentive).  Accordingly, concert attendance would be incentive elastic.