August 22, 2007

The Moral Hazard


(This is Part 2-- click to read Part 1

If you behave badly because you know you'll get away with it, that's being "bad."

A Moral Hazard is different.  If you behave "worse" than you would have otherwise, solely because you know that you won't have to bear the consequences, then you have a Moral Hazard.

I'll emphasize: the key is that your behavior is in itself not necessarily "bad."  It is simply worse than your behavior otherwise would have been, because you know there won't be consequences. 

Here's why it's called a Moral Hazard: if there are no external consequences, the only thing that would prevent you from behaving worse is an internal set of rules.

Where do these internal rules come from? 

Let's run some examples, starting with the current credit problem:

  • Too much risk (credit) got the hedge funds (and us) in trouble.  If Bernanke lowers rates to save the economy, and thus inadvertently bails them out, the hedge funds have learned they can take bigger risks next time, because they know Bernanke will (inadvertently) bail them out again.  It’s not that Bernanke wants to bail them out, he doesn’t.  He wants to save the economy—but the hedge funds now know he will repeatedly do this and can act “worse.”
  • Insurance: If you have no car insurance, you drive very, very, very carefully.   The Moral Hazard describes when a person, upon getting insurance, drives “worse” (read: less slowly, rolling stops; doesn’t avoid talking on cell phone.)  NB: he is not driving badly; he is actually driving normally.  But he's driving worse than "very carefully."  The term “worse” is relative to other behavior.  The point is that his behavior has comparatively worsened because he doesn’t have to worry as much about the consequences. 
  • Moral Hazard is cited everywhere in the universal healthcare debates: if healthcare is free, people may  be less attentive to preventative measures, or overuse medical care.  Before you scoff, imagine behavior if you had to bear all the cost personally.  To counter the Moral Hazard, we have limits, deductibles, co-pays, etc.
  • A subtype of the Moral Hazard is the principal-agent problem, in which one party has more information/knowledge they can use to maximize their own benefit.   A doctor, having considerably more information/knowledge than a patient, can use it to maximize his benefit, not even necessarily at the expense of the patient.  For example, he could bill for an extra service.  He can ask the patient to return more often than necessary to get the extra visit fee, perform unnecessary tests, etc.  So, too, lawyers, and politicians, etc. 
  • A more interesting example is the doctors who recommend a treatment modality-- for example, Depakote for maintenance bipolar--because it furthers their career, though it may not be as effiacious as another modality or more toxic.
  • People invoke the Moral Hazard to complain about easy bankruptcy laws: if there are no consequences to declaring bankruptcy, people could borrow excessively and then simply walk away.

But what causes the Moral Hazard? Why do all people, to varying degrees, fall prey to different forms of the Moral Hazard?  Why, and to what extent, does behavior worsen when there are no consequences?

Causes of the Moral Hazard

Succumbing to the Moral Hazard means that you have priced your gain above the loss of the other.  Let me be clearly emphatic: your gain is priced not above the gain of the other, but specifically above the loss of the other.  That's what makes it a Moral Hazard, and not an economic or strategic hazard. They know their actions cause a harm to the other, but do it anyway. 

So the real question is, what makes people not care?  Let's be clear that I am not talking about evil; I'm talking about the context dependent decisions of gain vs. loss. 

There are two cases:

  1. The loss is determined to comparatively trivial.  The inattentive insured driver knows that insurance rates will be raised, but he figures he himself only caused a tiny part of that raise, which is considerably outweighed by his gain of getting to use a cell phone in the car. Similarly, each individual hedge fund figures it contributes very little to the overall risk cycle.
  2. Unknown sufferer.  While it is necessary that someone suffer some loss, it is not necessary that the specific identity of that someone be known.  In fact, not knowing the specific identity makes it that much easier to find in favor of your gain, that much easier to be seduced by the Moral Hazard. 

It's case 2 that is the psychiatrically interesting one.  What about when the person harmed is clearly identified?  How do you facilitate, psychologically, the harming of that person?  The answer, of course, is to strip that person of their individuality, and see them only as an extension of you, your situation.

And so we return to it. 

The Moral Hazard and Narcissism

The Moral Hazard is narcissism.   The degree to which one is susceptible to the Moral Hazard is the measure of the individual's narcissism.

Remember, this isn't about evil; it's about what psychic variables make "your gain vs. their loss" calculations easier or harder to solve. 

In an extreme case where a clearly identifiable person is greatly injured—say, a doctor who prescribes chemotherapy to a non-cancerous patient just to collect fees—the person can act precisely because he fails (or chooses not) to see the person as a separate entity—he only sees the person as it impacts himself.

The Simpsons' Mr. Burns used to refer to his employees as "organ banks."  That's narcissism.

The Moral Hazard is, in essence, the opposite of Kant's categorical imperative. 

So the causes of Moral Hazard are  the degree to which a person

  1. is able to see others as individuals and not only as they impact himself;
  2. has an internalized set of rules that do not change regardless of how it impacts himself.
Everyone has a level of narcissism, thus, everyone faces Moral Hazards. 

It's not that only narcissists fall prey to Moral Hazards; it's that narcissism is what allows you to see a Moral Hazard, where others don't.  Consider the principal-agent problem: Say you're a lawyer.  You could easily bill for an extra hour, no one would know. Ok: did that ever occur to you? 

So the variables all impact your response to a Moral Hazard, and are covariant: your level of narcissism; your gain vs. their loss; the identity of the person harmed, etc. 


The Application of the Moral Hazard to Others 

But an even better description of the relation of narcissism to Moral Hazard is how your behavior  creates a Moral Hazard for others.

A good example is parenting.  If a three year old throws a tantrum, you can either appropriately deal with it, or give in to it.  But if you give in, you know the kid will throw a bigger tantrum next time, because he has learned there won't be any repercussions. There are two things you want: you want to teach him that these rules will be consistently applied, "so don't try to get away with it;" but, more importantly, that rules in general are consistently applied, so learn the rules and internalize them, with no consideration of your own gain-- don't try to calculate if you can get away with it, even if you can get away with it.  That second lesson is how you prevent (reduce) narcissism.

So good parenting requires that you always address tantrums in the same way-- consistently. Eventually, he internalizes the rules, they become part of his personality.   You have thus averted his facing Moral Hazard; you have saved him from narcissism.

But now you're in a store and he throws a tantrum.  The Moral Hazard still applies, but there's a new variable: your embarrassment.  Do you deal with the tantrum the same way as at home, or do you give in-- stick a cookie in his mouth, buy the toy, whatever-- just to shut him up, hoping to deal with it back home?  Clearly, the latter likely reinforces bad behavior.  You know this.  What do you do? 

Do you teach him rules are rules, no matter where and what-- save him from narcissism; or do you let him slide this time and reinforce that rules are context dependent-- in essence, reinforce his narcissism? 

How you teach your kid about the Moral Hazard is, in fact, your own personal Moral Hazard: do you gain a little at the store, at a loss to his development? 

Now you can see how your creating Moral Hazards for other people reflect your degree of narcissism.  Let's say it's a fact that free auto insurance leads to worse driving. If you are a politician who votes for free auto insurance because you will gain in popularity, knowing that it causes a Moral Hazard for others which is to the loss of still others, then you have double the narcissism.

(Read Part 3 .  The markets are stable today, and GOOG is up, thank God.)