Over the weekend, I posted on the U.S. crop insurance program. In that post, I struggled to figure out what I could learn about environmental policy from the crop insurance program. As I sit here looking at the pictures coming out of Alabama, Mississippi and Louisianna I find myself asking: Why do we have a federal crop insurance program and not a federal hurricane insurance program?
Before all our libertarian readers get their panties in a bunch, I'm not about to argue for a federal hurricane insurance program. Instead, I want to compare the common explanations for the U.S. crop insurance program and ask: If a federal crop insurance program is viewed as necessary by some, why isn't a federal hurricane insurance program (and vice-versa)[1]? The economics seem to be the same, so there must be another explanation.
From a traditional economic perspective, there are two reasons why insurance markets fail, and why government intervention might be needed to supply insurance: adverse selection and moral hazard. Adverse selection happens when only the customers with the highest risk purchase insurance. Unfortunately, the insuring company can't always tell which customers are high risk. But the customer knows. This is known as asymmetric information. For example, suppose I know that I am a high risk driver, but the insurance company has no way of knowing it. Based on the average driver, the insurance company sets a premium. Who is likely to purchase the insurance? Those that view their own risk as higher than average. So the company is left insuring the high risk drivers (I know, car insurance is mandatory. That's partly to eliminate this problem). So insurance companies are at a disadvantage because they can't observe the risk level of the customer. Same thing for crop insurance. Who is likely to purchase crop insurance? Those farmers that are either bad at farming, or those in areas with a high risk of crop failure. The insurance company can identify farmers in high risk areas, but observing the individual farmers ability to farm is difficult. Same goes for hurricanes. Who is likely to purchase hurricane insurance? Those in high risk areas (southeast and gulf coast residents), and those living in structures more likely to suffer more damage from hurricanes. The high risk areas are easy to identify, but the potential damage to a structure is better known by the owner than the insurance company. So adverse selection looks like it exists in both crop insurance and hurricane insurance markets.
Now for moral hazard. Adverse selection deals with what the customer knows (and the insurer doesn't) before she buys insurance. Moral hazard is what happens after the insurance is purchased. If a customer is insured, she is more likely to do something risky than someone not insured. This is why there is evidence that the number of car crashes go up with mandatory seat belt laws. If I know I'm protected, I might drive more recklessly. A farmer that has crop insurance might be more likely to take risks with their crop: try a new fertilizer or irrigate less in anticipation of rain. What about hurricanes? If a coastal resident has hurricane insurance, she is more likely to build in a high risk area and use cheaper materials and construction techniques. The premium the insurer sets acts as an incentive to take risk for the customer.
So for both crops and hurricanes, moral hazard and adverse selection make it difficult for insurers to earn a profit. If only high risk customers are purchasing insurance and those customers are taking above average risks once they have insurance, the insurance payouts are going to exceed the premiums collected (that's bad). But insurance companies have found ways to alleviate some of these problems. For example, in crop insurance, companies will offered group (pooled) insurance contracts where payouts are based group average crop yield. This reduces the incentive for deviating from what others are doing. Companies have ways of dealing with the market failures often listed for federal insurance.
Why then do we have federal crop insurance? I know many of you are going to say it's just an income support program, and it is, but there is also an economic argument to be made. Insurance companies rely on premiums to cover payouts. If the likelihood of a payout to a customer is unrelated to the likelihood of a payout to another customer, the insurance company can pretty accurately predict what the average payout will be and set premiums accordingly: that's why there are actuaries. But if the likelihood of a payout to customer A increases with the likelihood of a payout to customer B, there is systemic risk. That is, there is a possibility of having to pay a bunch of customers all at once (again, that's bad for the insurer). A drought will cause crop failure for all farmers in a region at the same time. When catastrophic crop failures happen the insurer is unable to cover all of the losses. Who pays the losses? You guessed it, the insurer of last resort: the government. So instead of waiting for a catastrophe, we instead have government subsidized crop insurance (actually reinsurance, but that's a detail I won't get into right now because this is getting really long).
So is there systemic risk from hurricanes? Dumb question. The better question: What is the difference between hurricanes and crops? The economics look the same. Both have adverse selection, moral hazard and systemic risk. So why do we have federal crop insurance and no federal hurricane insurance? Do politicians just like farmers better than coastal residents. I didn't say it...but you might.
[1] Currently, hurricane insurance is provided through most home-owners policies for wind and rain damage. Separate flood insurance is available in most areas that is federally subsidized. Catastrophic hurricane insurance is rarely available.
