Skip to content

Why the death of the insurance mandate doesn’t matter that much

December 14, 2010

The idea for this post came, from of all places, an interesting and lengthy Facebook discussion. I thought it would be better to expound upon my views on my blog for posterity instead of studying for my impending Copyrights exam.

My argument is that the insurance mandate in its current form would not have been effective at stopping those who want to game the system. It should be obvious that many people who currently do not have insurance don’t buy it because of a lack of means, not will. In other words, they simply can’t afford it. Hence the Affordable Care Act.

The primary impetus for a mandate arises out of the general fear that people will not buy insurance because they a) are too lazy to bother filling out forms, or b) will wait until the moment before they need insurance to purchase it, passing on the costs to everyone else in the form of higher premiums. Optimistic technocrats like Cass Sunstein favor the mandate for the first reason—it nudges them into acting in their self-interest. More pessimistic (though arguably less paternalistic) folks favor it for the second—it deters them from acting in their self-interest. I am not sure how many folks fall into category a, and for those I am willing to concede that some would sign up for insurance.  It is a bigger hassle to deal with the IRS than it is to deal with an insurance company. This is the brave new world of libertarian paternalism, and it frightens me. I digress.

I am wholly unconvinced that the mandate has any effect on category b. First, I doubt the fear is justified. Certainly people who think that they can call up an insurance company on the way to the ambulance vastly over-estimate the efficiency and service of America’s private bureaucracies. Insurance companies have no incentive to speed up their client sign-up process to facilitate “on-the-way” purchases. They still need to run their actuarial models to determine premium prices and such. A larger problem results from those who would purchase insurance a few weeks ahead of elective surgery, such as to fix a bum knee. Employer-based health insurance is typically integrated into compensation, so presumably only those who are not afforded these benefits through their job would be candidates to try something like this.

However, the mandate in its current form does not have a harsh enough punishment to actually deter anyone who would try to take advantage of the Act’s prohibition on discriminating against those who have preexisting conditions. Right now, the individual penalty is a minimum of $95 per annum—up to 1% of income—and a maximum penalty of roughly $2000 for families. The current penalty simply would not provide the needed deterrence.

Any punishment is geared towards those who must be deterred, not those who are naturally law abiding, whether out of self-interest or risk aversion. Let’s take the example of bank robbery. There are three types of people: a) those who wouldn’t rob a bank regardless of illegality, b) those who wouldn’t rob one because of the fact of illegality, but not fear of punishment, and c) those who are only deterred by the punishment. The penalty for bank robbery is aimed only at category c, otherwise we wouldn’t punish at all. So among category c folks, any effective punishment must exceed the expected value of the crime. If a bank robbery nets you, say, $11,000 on average, and the punishment is $700, the expected value of robbery is positive. The punishment thus does not deter me from robbing banks.

Similarly, let’s say I derive no or minimal benefit from the security of year-round health insurance, and would just pay premiums at the moment I need coverage if I could. Insurance imposes a cost of $11,000 a year to me on average, while the penalty is somewhere between $95 and $2000. The expected value of not purchasing insurance is positive, hence it cannot deter me from failing to buy it.

The mandate in its current form is clearly effective as a deterrent only to the degree that it affects those who refrain from doing illegal activities merely because they are illegal, and not from fear of the punishment. If you think this number is substantial, fair enough. But I point to marijuana possession, speeding, copyright infringement and jaywalking as pretty strong counter-examples. The mandate imposes no criminal sanctions, and you can’t go to jail for willfully flaunting it. Thus, unless we substantially raise the penalties for noncompliance well above the cost of insurance to account for probability of detection, a more effective strategy is not to make it illegal, but to establish social norms. Norms like “continuous insurance is in each person’s self-interest,” or “purchasing it is a public duty akin to refraining from burglary” would work.  I’m sure Cass Sunstein is working on some method of mass inception to plant these ideas in our collective subconscious as we speak.

While it is regrettable that the “nudge” effect will be lost, there are plenty of other constitutional ways to get lazy folks to buy insurance, such as opt-out policies, tax incentives for the chronically part-time employed (above and beyond the substantial subsidies already provided by the Act), and requiring purchase as a pre-condition to receiving other federal benefits such as unemployment insurance.

And finally, even if the mandate in its current form is essential to the cost-cutting function of the bill, it does nothing to affect its primary moral purpose—to end rescission and expand coverage for those who currently can’t afford it. The budget concern was mostly a conservative one. So long as the Senate remains at least 40% Democratic and Obama has veto powers, the law will at worst be more expensive but roughly intact come 2014.  In fact, President Obama himself opposed mandates while he was a candidate.  Justice Kennedy may do him a favor if he ends up siding with the other conservative justices who would likely strike down the mandate.

6 Comments leave one →
  1. Justin Michael permalink
    December 14, 2010 10:49 pm

    I think this leaves out a more pressing problem– that mandate or no, this bill has the potential to bankrupt the insurance model. Pre-existing conditions clauses have gotten a bad name lately (mostly because of gross overuse), but there is a reason they’re there. Insurance is preventative medicine, as it were, not an emergency medical expense. The model works a lot like a reverse loan– you pay a little money in for long periods of time and every once in a while you get a huge amount of money back. The amount you pay in is figured by your risk for injuries or “payouts” over the term of your contract with the insurance company.

    While BCBS calls in the middle of an ambulance ride are rare and pretty much a non-issue, most of the really expensive first-aid care is not covered by insurance anyway (think trauma stuff), and all of the things that are, are planned surgeries that give you plenty of time to set up an account. What this means is that over a lifetime, person A pays the premiums for like 6 months total and gets payouts for every emergency. That system is a net negative for the insurance company, and as it grows it will bankrupt them (with intermittent price gouges doing the job of driving away business).

    Further, like you point out, it is a net positive for person A, and so it is in their best interests to do it. Once this exploit catches on (like technicalities on medicaid and unemployment benefits), the insurance companies will be next in line for the bailout. Onerous though they are, the idea was to align their interests, not remove them from the game.

    If the insurance companies wrote this bill, they need new actuaries.

    • December 15, 2010 12:38 am

      Yeah, agreed. I don’t think you are correct that insurance doesn’t pay for trauma care. But your point remains that with or without the mandate, people could start doing this. The question is whether they will. Massachusetts’ experience suggests not: 99.8% of children are insured, 97.1% of the state. Many of the remaining 2.9% are Hispanic, which suggests either access or immigration concerns rather than cheating. This is further bolstered by the fact that more people who report fair or bad health are uninsured than those who are in good or great health (we’d expect cheaters to be people who don’t anticipate going to the doctor).

  2. Nick Brubaker permalink
    December 14, 2010 11:33 pm

    Good point. Out of curiosity: These limitations of deterrence, a fairly basic economic concept, in the now contested mandate must surely have been recognized by the drafters if this legislation. Why do you think the penalties in this mandate were so toothless to begin with? Do you think it would have been substantially more difficult to get this component of the healthcare package through congress if the included penalties were more severe?

  3. Matthew Scott Rozen permalink
    December 15, 2010 3:02 pm

    I think you’re making the wrong comparison here. When people are deciding whether to purchase health care, they are not weighing the cost of health care against the tax. They are weighing the cost of purchasing health care against the cost of not purchasing health care. One way or another, the value of health care needs to be fit into the equation (either by considering the net cost of purchasing health care, or by treating the lost benefit as a cost of not purchasing health care). Although the possibility of purchasing health care right before elective surgery reduces the value of health care, it doesn’t reduce it to zero.

    There are two reasons for this. The first is that the probability of requiring at least a small amount of medical care during a given period of time is fairly high. Most people I know have visited the doctor in the past year. Most people I know are also on at least one prescription medication. Of course, most people I know are insured, so they might not be representative, but all this means is that there is a subset of medical expenses that they might not have incurred had they been insured, the value of which is therefore somewhat less than the actual cost of the expenses to the insurance company, though still greater than the cost of a copayment. In any case, a portion of the price of health care goes to the ordinary annual health care expenses of a normal, mostly healthy individual, and this portion generates value that weighs against the cost of insurance.

    The second reason is uncertainty. The probability of being able to purchase health care before needing it varies directly with how quickly medical care is needed. Thus, for emergency care, as you acknowledge, the probability of being able to purchase health care on the way to the hospital approaches zero. For doctor’s visits one week out, the probability of being able to obtain health care in time is still pretty low. In fact, one possible consequence of the preexisting condition requirement might be that insurers impose delays in peoples’ abilities to obtain coverage. I don’t know if the health care bill allows this, but it should. Health care companies can say that previously uninsured consumers cannot obtain health insurance for sixth months, or that certain expenses incurred during the first sixth months will not be covered. Even a one month delay could have significant effects. Of course, a portion of the value of insurance derives from elective surgery or other medical care that can be delayed indefinitely, but my guess is that this is a fairly small portion. Consumers who fail to purchase insurance lose the value equal to the sum of the expected additional expenses incurred during the delay period. In other words (probability of not being able to obtain insurance within x days) x (probability of injury within x days) x (cost of care for injury within x days with insurance – cost without insurance) . . . or something like that, I think.

    A portion of the cost of insurance is composed of these values. The rest is made up of savings that could be captured by gaming the system, and profit for insurance companies. I was going to make an argument that the profits portion of the cost of insurance is irrelevant, but I can’t express it concisely, so I’ll just leave that out for now. The point is, a consumer deciding whether to purchase insurance will weigh the value of insurance (in terms of average ordinary annual expenses + expected additional expenses that will not be covered if the consumer tries to game the system) plus the cost of the tax penalty against the cost of insurance. While the tax penalty alone is clearly less than the cost of insurance, when added to the value of insurance, it might bridge the gap to the cost of insurance.

    This is also where the nudge comes in. Consumers are probably bad a estimating the expected value of insurance. If they’re just comparing the cost of the tax penalty against the cost of insurance, they can do this math fairly easily. But failing to purchase insurance means that a consumer risks incurring health care expenses before he can get insured. If the decision is close, consumers might be nudged not necessarily in their own interest, but in the direction we want them to be nudged, by a default rule in favor of insurance.

    Another way to look at this: your post skews the problem by separating the two rationales for the individual mandate (paternalism and preventing people from gaming the system). When you put the individual consumer’s self-interest in purchasing insurance back in the equation, the tax penalty may be enough to discourage gaming the system.


  1. Mo’ Mandates, Mo’ Problems « Generalissimo

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: