Back in 2010, when the Affordable Care Act (aka Obamacare) was passed, I didn’t know much about health insurance. At the time, I was a first-year graduate student getting my Ph.D. in Economics, and the national debate over health care started to steer my interests. I focused my studies on two sub-fields of economics: Health Economics and Public Economics. I studied the economics of health care and the economics of public policy. Now with my Ph.D. in hand, I’m perpetually surprised about how unwilling the popular press, pundits (including economists!), and politicians are in making a full argument for these major policy proposals. So, I would like to try to fill the void, and summarize in one place everything I now know about health insurance and the policies related to it.
Why People Buy Insurance
Imagine a simplified world where we exist in one of two states: either we are healthy or we are sick. When we are healthy, we can work and we’ll make $100 each day doing so. When we are sick, we cannot work and we earn $0. To make our math simple, let’s assume that 50% of the time we are healthy, and 50% of the time we are sick. So how much money do we earn on an average day? Well, on an average day there is a 50% chance we are healthy and earn $100, and a 50% chance we are sick and earn $0, so we’d average $50 a day in the long run.
Now let’s think about an insurance company in this environment. They can offer a plan where, on days that people are sick, they pay out $100. But on days that people are healthy, they don’t have to pay anything. If they get a lot of people to sign up, they’ll start to benefit from the law of large numbers, and the number of customers they have to pay on a given day will almost always be right at 50%. That is, with a large pool of customers, the insurance company would be paying 50% of their customers $100, and the other 50% would be healthy and they wouldn’t have to pay out anything. So, on any given day, the average customer is receiving $50. How much would this insurance plan cost per day? Well, the insurance company would break even if they charged $50 a day, since they pay out an average of $50 a day to each customer. To cover administrative costs and make a profit, they’d have to charge a little more than $50 a day.
So why do people buy insurance? In essence, they’re paying $50 a day to receive $100 on days they’re sick. They’re sick half the time so on average they’re getting $50 from the insurer. They’re paying $50 (or more) to get $50. How is this rational!?
The reason is that the utility of a guaranteed $50 each day is greater the average utility of getting $100 half the time and $0 the other half. With the insurance plan, I earn $100 on healthy days but receive $100 on sick days. So I get $100 every day no matter what. But I also have to pay $50 every day to the insurer no matter what. So that leaves me with a guaranteed $50 every day.
Economists represent people with “utility functions“. These are mathematical equations that define a person’s well-being. The details aren’t too important, but what is important is that these functions exhibit a property we call diminishing marginal utility. On a given day, the first sandwich I eat is of enormous benefit. It keeps me alive and functioning. The second sandwich I eat is joyful, but less important. The third is pleasant, but now I’m pretty full. The fourth is now painful. The fifth tortuous. The sixth evil. As we keep consuming, the benefit of having one more declines. The same is true for money, since money is just stuff we haven’t bought yet. The first dollar I have to spend is critical to the my livelihood. The millionth is pleasant. But as Brewster’s Millions shows, the 30 millionth can be a burden.
This translates to a surprising property: earning $100 is not twice as good as earning $50. I know that sounds crazy, but it’s definitely true. If you’re not convinced it helps to make the numbers larger. If you had $100 billion you would not be twice as happy as if you had $50 billion. You’d be pretty much the same with either amount.
Because $100 is not twice as good as $50, we are actually happier with a guaranteed $50 than an average of $50. Because without insurance, half the time we get the joy of $100 and the other half we get nothing. And half the joy of $100 is less than the joy of $50. That’s why we buy insurance. That’s why we’d even pay more than $50 a day for the insurance. In fact there is some number (what we call the certainty equivalent) less than $50, maybe $47 or $48, that makes us exactly as happy as we are on average without insurance.
But wait, why would I pay $52 or $53 dollars for insurance when I could just save $50 on healthy days to cover my sick days? Because that’s not how randomness works. Flip a coin 10 times and the odds of getting exactly 5 heads and 5 tails is only 24.6%. And the odds of more tails than heads is 37.7%. Sure, over the long run your savings scheme might work. But it’s also sure to run into deficit sometimes, leaving you destitute. So it isn’t the sure thing that insurance is. It doesn’t solve the problem.
Insurance spreads the burden of randomness across enough people that randomness no longer has any power. It’s the same savings scheme detailed in the last paragraph, but where the people running in surplus are loaning money out to those running in deficit, to be repaid when our positions are reversed. That’s why insurance exists and that’s why we buy it.
Insurance is Already Broken
With that explanation alone, we’ve already broken health insurance. In my example from the last section, everyone got sick 50% of the time. But in reality, we all face different odds. It’s also hopefully easy to see that the people willing to pay the most for insurance will be ones with the highest odds of getting sick. This is a problem we call adverse selection.
If the average person is sick 50% of the time, our insurance company will still function fine charging $50 a day from everyone, and paying $100 when they’re sick. For the person who gets sick 70% of the time, this is a HUGE benefit. They average only $30 a day without insurance, but with it they get the full $50 a day. They’d even be willing to pay $70+ for insurance, but are getting it for only $50!
But what about the person who gets sick only 30% of the time? They average $70 a day without insurance, but would only get $50 a day with it. As we noted, people are willing to pay a little extra for the peace of mind, but not that much extra. They’d probably choose to forgo insurance.
That’s a problem, because without the low risk people, the pool of the insured becomes higher risk. The average person is getting sick 50% of the time, but without the 30% or lower people, the group of insured might be getting sick 60% of the time. To cover those costs, insurance companies have to charge $60 a day, and that might push out the 40% people, raising the rate for the insured even more to 70%, pushing out the 50% people, raising it to 80% and pushing out the 60% people and on and on we go until we have the permanently sick, who aren’t earning the income needed to pay for insurance, and so the whole thing goes kaput.
Insurance Companies Have to be Dicks
I love this headline from the satirical news paper The Onion. Insurance companies have to be evil in order to prevent the death spiral of adverse selection from taking hold. In order to survive, they have to successfully distinguish the high risk people from the low risk people.
In our example, if the insurance company could charge all the 70% people $70 a day, and all the 50% people $50 a day, and all the 30% people $30 a day, they could keep things running. No one wants to quit their insurance plan, because the payout keeps them all at their long-run average, and the company covers costs and stays in business. Easy right? But which people are more likely to get cancer and need care? Who is at risk for Alzheimer’s? Who will need very little care in their life and die of natural causes? If we had any idea, we’d be acting on it, but we live in an age of superstition and unconfirmed findings from medical journals making headlines only to be quietly debunked later. Just about the only thing we do know is that smoking is bad. And that’s just about the only thing other than age that an insurance company has to distinguish you from others.
But you know a lot more than that. You know your eating habits. You know what your grandparents died from, and what that means for you. You know your own propensity to seek care. Some people go to the doctor with a cough. Others only when they’re on death’s door. You know more than the insurer, and that creates a problem of asymmetric information.
Because they can’t tell us apart, insurance companies often try to get us to self-sort. At my job, I can choose between three employer-provided health insurance options. Some offer more care options and coverage but are more expensive. And if I get onto the lower-cost plan, but end up needing lots of care, the insurance company is going to deny me that coverage. We may hate them for it, but frankly, the problem starts with us.
The result of asymmetric information is that insurance companies deny people certain coverage because of pre-existing conditions, cancel coverage, and reject requests for treatment. They do it because without these sorting mechanisms, the entire industry would collapse. And so it is the conditions of this market that lead us to a sub-optimal solution. Insurance companies are dicks so that we self-sort into different tiers, but this sorting is still imperfect, and so there are still a lot of people who don’t buy insurance because they are low-risk, and cannot find a plan at a reasonable price.
A Social Argument
I am partial to a philosophy put forward by John Rawls in a book called A Theory of Justice. Rawls argues that we should construct society in the “original position” behind a “veil of ignorance”. Philosophers are better at naming things than economists. What he meant was, right and wrong should be determined through a thought experiment, where we imagine ourselves prior to our birth, deciding on how society will be constructed. And when we are done, we will be born into a random position within this society. Rawls argues that in this environment, we wouldn’t tolerate a lot of severe poverty, because there would be too great a chance that we end up being born into it. Think what you want about that, but the thought experiment is a good one.
I would argue that given the option, most people would choose to buy health insurance prior to being born. If we found ourselves in Rawls’ original position debating ethics, a clever entrepreneur might notice the opportunity. Since our placement in society is to be random, no one knows what their health risks will be. Our best guess would be whatever the average for humanity is. In other words, this fictional veil of ignorance levels the playing field between the insurer and the insured. Both can only rely on broad statistics, and so both would find it much easier to come to some agreement without worry of being scammed.
The conclusion of my entrepreneurial version of Rawls’ thought experiment is that charging different prices for health insurance for different people isn’t really socially optimal, even if it is economically neutral to charging all the same price. And I would further argue Rawls’ point, that if we extend this discussion to right and wrong, we’d be willing to accommodate those born into poverty, who might not be able to afford the insurance premiums we agreed upon behind the veil.
And thus we have:
The Patient Protection and Affordable Care Act (aka Obamacare) was passed in 2010 to address all of the problems I’ve outlined so far. At its core is an attempt to solve the market failure of adverse selection and asymmetric information. It does so by further subsidizing the purchase of health insurance.
Health insurance is already subsidized in America by being exempt from income taxation. Most people get their insurance through their employer, and when they do, that payment gets deducted from their paycheck prior to taxes. And please note that this includes not only what the person pays, but also what the employer pays, which is still part of your overall compensation. The subsidy for health insurance is bigger the higher your income. If your top marginal income tax is at the 15% bracket, then the tax break is equivalent to a 15% subsidy on your health insurance. If you make a lot more money and fall into a top rate of 33%, then the subsidy is 33% of the cost of your health insurance. If you don’t get insurance through your employer, then you pay the full taxes on your income and must buy insurance post-tax, missing out on that large subsidy.
Contrary to everything I have ever read in the news (especially from economists!) this subsidy is good! The whole problem of adverse selection and asymmetric information is that the low-risk people drop out of the market, raising the risk of the remaining pool. In our example, the people getting sick 30% of the time were not willing to pay $50 a day for insurance. But the subsidy helps them to make up the difference. If they are willing to pay $30 themselves, and then the subsidy gives them an extra $20 to buy insurance, then they’ll do so happily. Subsidizing this market makes it work.
But there are a lot of people who don’t get insurance from their employer, and a lot of people who choose to forgo it when it is offered because the subsidy is not enough. Obamacare tries to correct that problem by doing two things. First it creates a marketplace for people to buy health insurance essentially as a group, offering them some of the benefits employer-provided coverage gets. By lumping people together onto an employer-provided plan, insurance companies hope to avoid some of the problems of adverse selection. It is hard to tell how risky an individual is, and thus how much to charge them. But it is much easier to assess the risk of a large group of employees, when the law of large numbers starts to take root. People who don’t get insurance from employers lose out on the benefits of this risk pooling, and the Obamacare exchanges try to provide that to them.
Second, Obamacare counter-balances the imbalance created by the tax code subsidy on health insurance. It offers larger subsidies to low-income groups without employer-provided coverage, whereas the tax code gives larger breaks to high-income groups with employer coverage. This, hopefully, evens the playing field some, and realigns incentives in favor of an efficient economic outcome where the key benefit is that insurers no longer need to differentiate between us. They can charge the same price to all (or close to it), and all will be willing to pay. This is both economically beneficial (a boost is our overall welfare as a society) and as I argued before, a social benefit that better approaches what we would agree to behind the veil of ignorance.
Obamacare also includes some other elements with purely social benefits that economics is not deeply concerned with. Namely, disallowing insurers to deny coverage based on pre-existing conditions balanced by a mandate that all people must hold a certain level of insurance (so that we don’t just wait till we are sick to buy it, since we cannot be denied). Think of those what you will. My point here is that Obamacare is a reasonable solution to a serious economic problem.
Obamacare is Very Bad
That isn’t to say Obamacare is without issue. There is a different problem that is made much worse by Obamacare, and it is our best named term in economics: moral hazard. I like to describe moral hazard this way: How much ice cream do you plan on eating today? Okay, how much ice cream would you eat today if I paid for it all? I’m gonna take a guess and say the answer to the second question was more than the first. That’s moral hazard. When someone else is footing the bill, we happily consume more.
That is true for health care too. When I take my car in for an oil change and they bring me a list of other things they’d like to “fix” for me, I say no to most of them. Sometimes for better, sometimes for worse. But I weigh the costs and benefits and try to come to a logical solution. When I go to the doctor with a sprained ankle and they recommend an MRI, I say “Great, let’s do it!” Because I’m not paying for the MRI. I don’t need to concern myself with the costs or benefits. Better safe than sorry as they say.
Because we don’t have to pay for medical care when we have insurance, we consume way too much of it. This piece has been light on citations, but a massive study called the RAND Health Insurance Experiment found that 30% of our health care is unnecessary. The study has been around a while, so a lot of criticisms have mounted, but most people who study this sort of thing don’t think the problem is that the number is too high.
The problem with this is that the supply curve for medical care slopes up. That is, the 30% of unnecessary care was more expensive that the previous 30% of necessary care. As we consume more, each bit of it costs even more. We have to raise salaries to attract enough doctors, nurses, and equipment into the health care sector of the economy to supply all that unneeded care. This sucks resources out of other industries where they could be doing more good for society. It also raises the cost of health care, and therefore also raises the cost of health insurance.
Obamacare simply incentivizes people to get health insurance. It does little to correct the problem of moral hazard, and so it just makes this problem worse. The headlines noting the large increases in insurance premiums since Obamacare was implemented often note that their rate of increase has still slowed. This is because more low-risk people are on the rolls, spreading the costs around and keeping premiums down. But premiums are still rising, and by a lot, because all of these newly insured people are using more care than they did when they had to pay for it out-of-pocket.
Ok, so now what?
I’ve written this whole thing because I see too few of the right arguments being discussed. Free markets cannot provide an efficient health insurance system. We’ve known this since the 1960s. Obamacare is a serious and well-designed attempt at correcting that problem using simple market mechanisms. While it is often derided as a government take-over of health care, it is the exact opposite. It leaves health care largely untouched, and focuses on balancing what government intervention there already is in order to make the market system work. But Obamacare ignores the problem that other countries designed their government take-overs of health care to solve.
Single-payer health care, where the government is the sole insurer and pays for everything, is an effort to correct moral hazard. The government can simply not supply the health care people want when it is free. They can stop production by not paying for what is deemed wasteful. The problem is, who determines what is wasteful? You end up with the mythical death panels Obamacare actually avoided. Personally, I’d be more in favor of creating incentives for doctors and health care providers to not provide more care than is necessary. Changes to malpractice laws and other schemes like prospective payment systems would likely work. We don’t need a single-payer system. I just hope we don’t throw out a good solution just because it puts pressure on us to solve a bigger and harder problem. Obamacare is well-designed, and it could be combined with legislation aimed at fixing the larger problem of moral hazard, legislation Republicans in congress and President Trump could be rightly immortalized by.
Nevertheless, what I would wish for the United States is a discussion with less head-nodding at fiery rhetoric and a more willingness to learn more about how this all works. As President Trump has said, “Nobody knew health care could be so complicated.” It is complicated. Very complicated. More complicated than you or I know. But this is what I have learned about it so far.