Hypothetical Mean

Commentary from an Actuarial and Economic Perspective

Adverse Selection is Not the Problem

leave a comment »

I hate to say it, but I think the excellent bloggers/economists Andrew Samwick and Donald Marron are entirely wrong when they say that “The problems of adverse selection and moral hazard in insurance markets are well known — they are what stands in the way of extending the benefits of competition to health care.  Addressing them should be the central features of the reform, with a risk-adjustment mechanism to address the former and high-deductible plans to address the latter.”

Adverse selection does not exist when the insurance company knows more about your expected costs than you do.  Here’s a quick question to illustrate:

You’ve had a simple fracture in the past three years.  What is your additional expected healthcare cost next year?  Do you know?  I do.  Change the question to a “wait-and-see” diagnosis for a pre-cancerous lesion and I’d still say I know more about that expected cost than you do.  We’ll rate for that expected cost, load for profit, and sign the contract.  Cool?

Insurance companies can issue surcharges for significant extra risk.  Profits don’t come from only insuring the healthy.  Insurance companies actually earn more profit insuring the <i>identifiably</i> sick <i>who also choose to be covered</i>.  Those contracts have more risk and more associated profit, almost always (if nothing else, profits are generally priced as a percent of premium, and the sick will be charged more premium because insurance companies are able to identify them with reasonable accuracy).

The problem is that there comes a cost level where the expected cost plus needed risk charges will exceed the person’s willingness / ability to pay.  And/or the needed price will be so high that it will make a demure health insurance company blush, making them worry about PR.  And rates are also frequently capped by regulatory limits on surcharges (I’m not arguing against those regs, I am simply pointing out that their existence leads people to be denied coverage).

This leads naturally into a second error I believe that Andrew and Donald are making when they trumpet potentially imperfect and administratively costly risk-adjustment mechanisms (actuaries are a top 3 profession and someone has to pay us for something).  Risk adjusters aren’t really a solution for adverse selection; risk-adjusters are really a way to transfer resources between individuals.  Again, an illustration:

A 20 year old male with a risk-score of 0.5 walks into a bar that sells insurance (yeah, yeah).  The bartender says “what’ll you have?”  The male says a $1,000 deductible plan.  The bartender says that he used to be able to sell that for $100 … but now he has to compensate for the 20-year old’s low risk score, so that he has to charge $300.

A 55 year old woman walks into that same bar.  She has a risk score of 1.5 because she had a zit on her nose when she was a teenager.  Yadda, yadda, yadda, and instead of being charged $600, the bartender can sell a policy to her for $400.  The bartender can (has to) do that because he has $200 shifted from the 20-year old.

Has any adverse selection problem been solved here?  Nope.  Either could have gotten the service for the initial price (and notice that the initial prices are independent of the quantity of competitors since they are simply a function of their true cost, which insurance companies can guess at better than you).  The difference is that after risk-adjustment, the healthier pay more and the sick pay less.

In fact, risk-adjustment may <i>cause</i> adverse selection, because, suddenly, the bartender can’t charge the 20-year old the low rate his health deserves.  The 20-year may just tell him to kiss off.  In which case everyone except the actuary will lose since you’ll have to pay the actuarial salaries to offer the exact same contract to the 55 year old that you would have without risk-adjustment.  Because even risk-adjusted markets can cause adverse selection, actuaries are (misguidedly) demanding that an insurance coverage mandate accompany this push to deform healthcare.

One of the fundamental questions of healthcare reform is “Who has the right to benefit from or pay for your health?”  If you answer that you do, then you don’t support risk-adjusters.  That has very little, if anything, to do with adverse selection.  Conversely, if you answer that no single person should benefit from or pay for their health, then by all means, risk-adjust away. They are great make-work for actuaries with tons of cool intellectual problems I’d like to try to be clever and solve.  And if that isn’t enough of a reason to support them, they can also be very successful at redistributing the financial gains and losses that would otherwise be caused by your health.

Just don’t pretend that you did it to solve a perceived adverse selection problem.  You did it to solve a perceived maldistribution of resources.


Written by Victor

October 8, 2009 at 3:33 am

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: