by Allan Joseph
Welcome to Project Millennial’s 2013 summer journal club. For the next 10 weeks, we’ll be going through some of the most important concepts in health policy and health economics, especially regarding health insurance, by studying some classic research papers from the last 50 years. This week’s post focuses why people buy health insurance, using Kenneth Arrow’s classic 1963 paper “Uncertainty and the Welfare Economics of Medical Care,” available for free online here. Note: This paper covers a lot more ground than the topic at hand, but we’re specifically focusing on why people want to buy health insurance.
Imagine walking into a grocery store. You know you’re going to buy eggs, milk, and Double Stuf Oreos (because they’re the best kind). You grab your eggs, and see that the price of milk has gone up, so you buy the half-gallon instead of the gallon. When you get to the cookie aisle, you see a happy sight: Double Stuf Oreos are on sale, so you buy two packages instead of one (American obesity epidemic notwithstanding). You get to the register, and when the clerk rings up your total, he or she asks you for your food insurance card.
No, that last part isn’t right. No one buys food insurance — or, for that matter, many other types of insurance. So then why is healthcare different? What makes health insurance so important? That’s one of the questions Kenneth Arrow aims to answer in this paper, and it’s the one we’ll focus on here.
To start, let’s back up to that grocery-store story we told. If it made sense to you, you’re familiar with the mechanics of supply and demand. (If you want more of a tutorial, click here.) In most markets, supply and demand are very straightforward. If a natural disaster wipes out cotton crops, then the supply of cotton goes down, the price of cotton goes up, and the demand for cotton then goes down. In those markets, consumers tend not to buy insurance. There’s no need to. If the price of milk goes up, you just buy less milk.
The problem for “medical care” (Arrow’s term of choice) is that the market is very, very different from most. How? Let’s count the ways:
- Unlike other markets, demand for medical care is irregular and unpredictable. Food, clothing, housing, and other essential goods are all regular, predictable, and fairly manageable expenses. It’s not just hard to predict when you’ll need medical care, it’s also hard to predict how much medical care you’ll need.
- People who need medical care are sick, and sick people are at a real risk of losing their earning potential, whether through forced time off, disability, or death. Unlike in the market for food, this problem isn’t solved by merely having a large enough income.
- Physicians and care providers act under ethical rules that limit their ability to act like producers in normal markets. While later research would show Arrow was a little optimistic on whether physicians shape their treatment due to financial concerns, the widespread existence of charity care for the poor is a good example of this idea.
- Even if you go to a good doctor, there’s still significant uncertainty regarding the product’s quality, because very few things in medicine are guaranteed to work — in Arrow’s words, “recovery from disease is as unpredictable as is its incidence.” That means that even if you know you’re going to a top doctor at a top hospital, you can’t ever be sure of the quality of product you’re buying, because you don’t know if it’s going to work.
- In a normal market with lots of demand, more producers will enter the market to sell the product, because they see an opportunity to make money. Medical licensing and the long training period make this impossible, though, because they limit how many people can provide medical care.
- Price competition is rare in the medical-care field — often, it’s impossible to even comparison-shop by price.
So medical care is, in short, an unpredictable expense — and one of the central truths of economics is that most people don’t like unpredictability: they’re what we call “risk-averse.” Yet the market has come up with a way to transfer the risk of that unpredictability: insurance. If a company agrees to take a fixed payment from you and 1,000 other people in exchange for paying your bills, everyone wins. You get to pay a predictable amount and get your unpredictable medical expenses covered. The insurance company gets to “pool” the risk: it’s hard to predict one person’s chance of getting sick (and thus, their cost) based on statistics about the whole country, but it’s a lot easier when you have 1,000 people (or even better, 10,000, or more) — as your risk pool increases, your expenses should more closely match those of the average. It’s probability at work. (If you want a more detailed description, it’s in the postscript at the end of this post.)
That’s why we have health insurance, and that’s the most important takeaway from this paper. But there’s more!
Arrow goes on to detail some important problems in the medical care insurance market that become the source of some really important problems. We’ll run through a few important ones:
- When people know someone else is on the hook for paying their medical bills, they tend to do riskier things, but the price they pay for insurance doesn’t reflect the increased risk of needing care. Economists call this problem “moral hazard,” and it can seriously increase the price of insurance.
- Insurance companies aren’t costless. There are a lot of administrative costs involved in providing health insurance, and those also raise the price.
- Very importantly, some people are much likelier to get sick and need care than others. This leaves insurance companies competing to sign up healthy people (who will probably pay more than they claim in insurance costs) and makes insurance prohibitively expensive for those who are sick. As you might imagine, this is one of the problems the Affordable Care Act tried to address.
These problems leave the health-insurance market somewhat hobbled, and they’re big reasons why government has played such a big role in health insurance around the world. In fact, they’re some of the core problems in health economics, and if you have questions about all of this, feel free to shoot me an email.
But we’ve answered an important question already: Why do people want insurance? That, I think, is enough for this week.
Next week: Why do American employers provide health insurance?
Postscript: Risk pooling
Risk pooling isn’t necessarily an intuitive concept, but an example might help:
Imagine rolling a die once. You know the probability of rolling a 1 is 1/6, right? Would you be totally shocked if you rolled it six times and didn’t get a 1? Probably not. What if you rolled it 60 times and didn’t roll a 1 about 10 times? That would be more surprising. What if you rolled it 6,000 times and didn’t get almost exactly 1,000 rolls of a 1? That would be really shocking.
The same concept is at work here in insurance risk-pooling. As the number of rolls/customers increases, it’s easier to predict what happens — the risk gets “pooled” and therefore more manageable. By pooling risk, insurance companies have a very good idea of what they’re going to pay for medical expenses as a whole.
Allan Joseph is a senior at the University of Notre Dame studying economics and pre-medical studies. You can follow him on Twitter @allanmjoseph.