When I describe mental illness as “an extreme, socially disapproved preference,” the most convincing counter-example people offer is depression.  Do I really think people “want to be depressed” or choose depression as a bizarre alternative lifestyle?

My quick answer: These objections confuse preferences with meta-preferences.

No one chooses to have the gene for cilantro aversion.  Yet people with the cilantro aversion gene are perfectly able to eat this vegetable.  They just strongly prefer not to.

Similarly, when I say that alcoholics are people who value heavy consumption of alcoholic beverages more than family harmony, this doesn’t mean that they like having these priorities.  If they could press a button which would eliminate their craving for alcohol, I bet many alcoholics would press it.  But given their actual cravings, they prefer to keep drinking heavily despite the suffering of their families.

The same holds even more strongly for the typical person diagnosed with clinical depression.  Most people with loving families and successful careers are happy.  Clinically depressed people, however, often have both loving families and successful careers, yet still want to kill themselves.  Their preference is so extreme that it confuses the rest of us.  They’d almost surely rather have a different preference.  But it is their preference nonetheless.

Not convinced?  Think back to the early 1970s, when psychiatrists still classified homosexuality as a mental disorder.  I object, “Mental disorder?  No, it’s just an extreme, socially disapproved preference.”  When critics incredulously respond, “Do you really think people choose to be gay?,” I say they’re confusing preferences with meta-preferences.  To be gay is to sexually prefer people of your own gender.  This doesn’t mean that gays want to feel this way.  If a gay-to-straight conversion button existed in the intolerant world of 1960, I bet that most gays would have gladly pushed it for themselves.  Even today, I think many gay teens would press the conversion button to fit in and avoid conflict.  But so what?  Hypothetical buttons can’t transform a preference into a disorder.

Is this all just a word game?  No.  The economic distinction between preferences and constraints that I’m drawing upon has three big substantive implications here.

First, people with extreme preferences could make different choices.  People with cilantro aversion are able to eat cilantro.  Alcoholics are able to stop drinking.  The depressed can refrain from suicide.  And so on.  This is fundamentally different from my inability to bench press 300 pounds – or live to be 150 years old.

Second, as a corollary, people with extreme preferences can – and routinely do – respond to incentives.  People with cilantro aversion are more likely to eat cilantro if other foods are expensive or inconvenient.  Alcoholics respond to alcohol taxes – and family pressure.  Depressed parents may delay suicide until their kids are grown.  Even in a tragic situation, incentives matter.*

Third, as a further corollary, people with extreme preferences can – and routinely do – find better ways to cope.  People reshape their own preferences all the time; perhaps you can do the same.  Failing that, perhaps you can discover more constructive ways to satisfy the preferences that you’re stuck with.  For example, if you’re extremely depressed despite great career success, you really should try some experiments in living.  Perhaps you’ll be miserable whatever you do.  But if you’ve only experienced one narrow lifestyle, how do you know?  Maybe you’d feel better if you tried putting friendship or hobbies above “achievement.”

It’s tempting to insist that there’s something pathological about having conflicting preferences and meta-preferences.  On reflection, however, these conflicts are a ubiquitous feature of human existence.  Almost everyone would like to feel differently in some important dimension.  Almost everyone reading this probably wishes they were less lazy, more patient, more outgoing, more loving, more ambitious, or more persistent.  But you still are the preferences you really have.  There’s plenty of room for improvement, but that doesn’t mean you’re sick.

* I’m well-aware that many physical symptoms also respond to incentives.  You can pressure a diabetic to lose weight, which in turn reverses his diabetes.  But all of these incentive effects require time to work.  The symptoms of mental illness, in contrast, can and often do respond to incentives instantly, because they are choices that are always within your grasp.  “I’m divorcing you unless you stop drinking right now” is a viable threat.  “I’m divorcing you unless you stop being diabetic right now” is silly one.