The Dubious Economics of Crop Insurance
Insurance is an essential part of the financial infrastructure of a market economy. By spreading losses among members of a group with similar exposure, insurance encourages people to take prudent risks while protecting individuals from ruin in case they are the unlucky ones. Not all risks are insurable, however. Attempts to insure the uninsurable create incentives to take excessive risks and burden the economy with costs to the many that exceed the gains to a few. So-called “crop insurance,” which has become a central feature of U.S. farm policy, is a case in point.
Why crop losses are not insurable
Over time, insurers have developed rules that identify which risks are insurable and which are not. Crop insurance violates at least three of them.
Not a pure loss. Insurance is normally limited to situations in which people face a pure loss. For example, if I insure my house against fire, I either experience a fire, in which case I suffer a loss, or I do not, in which case I have neither a loss nor a gain. In contrast, if I build a house for resale, I may suffer a loss if no one likes it or if the market declines, or make a profit if someone falls in love with it and pays me a premium price. The risk of fire is a pure loss, and is insurable; the risk of a business venture that carries the possibility of gain as well as of loss is not.
Insurance against crop risks, especially in the popular form of crop revenue insurance, departs from the pure loss principle. Crop revenue insurance does not just protect farmers against bad harvests due to natural causes like drought or floods. It also protects their profits against the economic risk of low prices, even when a good harvest is the cause of the low price. In fact, if the premium is low enough and the benchmark price is high enough, crop revenue insurance provides a guaranteed profit no matter what happens.
Congress is now considering changes to farm legislation that would introduce so-called shallow loss revenue insurance, which means that revenues would have to drop only 10 percent below the benchmark in order to trigger a payout, rather than the 25 or percent or more that has been common in the past. As an article in the High Plains/Midwest Ag Journal explains, shallow loss insurance means no-risk farming:
That’s a novel idea. I wouldn’t want to be the one to explain this concept to Wall Street investors, futures market speculators, Las Vegas gamblers and small business owners. All of these groups also face much uncertainty but with few exceptions receive no government subsidy.
The authors, writing in a farm journal, may not fully appreciate the degree to which Wall Street firms benefit from implicit government subsidies under the too-big-to-fail doctrine, but pointing out the similarity only strengthens their main point: Government policies that give private firms a sure profit while shielding them from the risk of loss are inadvisable, regardless of the field of business.
Losses not fortuitous. To be insurable, losses should be fortuitous, that is, they should be due to random events that are outside the control of insured party. Some crop losses, for example, damage from hail or tornados, fit this definition. They strike randomly, no one knows where. However, in many other cases, the risk of crop losses directly depends on the choice of farming practices.
A recent paper in the American Journal of Agricultural Economics by Barry Goodwin of North Carolina State University and Vincent Smith of Montana State University details some of the effects of crop insurance on farming practices (full text of the paper posted here.) Goodwin and Smith find that farmers with crop insurance choose to plant riskier crops, to extend planting to acreage that is marginally suitable for the chosen crops, and to substitute insurance for other farming practices that reduce risk of crop loss. Some of the effects they describe remind one of Nikita Khrushchev’s virgin lands fiasco of the 1950s.
The tendency of people to take greater risks when they are insured against loss is known as moral hazard. It is possible to control moral hazard, to some degree, by including deductible clauses in policies, so that the first part of any loss falls on the insured party. Unfortunately, the proposed introduction of shallow loss crop insurance reduces the amount of loss that is deductible. If enacted, it would make the perverse incentives noted by Goodwin and Smith significantly worse.
Premiums not affordable. A third requirement for insurability is that premiums must be affordable. In practice, that means two things. First, it means that insurance works best for risks that have low probabilities. Second, administrative costs of providing the insurance must be small.
Home fire insurance fits this pattern. If there is one chance in a thousand that my house will burn down, I might very well be willing to pay an insurance premium of, say, two-tenths of a percent of its value. Such a premium would give me security against an unlikely but catastrophic financial loss, and at the same time, since it is double the mathematical expectation of loss, it would provide enough of a margin to cover the insurance company’s reasonable costs plus a profit.
On the other hand, insurance is not practicable against events that are highly likely. For example, home insurance does not cover costs like repainting or snow removal. It is true that those expenses involve risk, to the extent that they do not occur on a fully predictable schedule, but sooner or later, they are highly likely to occur. Claims for painting or snow removal would be more frequent than for fire and administrative costs would be high relative to the value of claims paid. Even if some company would be willing to write insurance against such risks, homeowners would not be interested. It would be cheaper to pay for isuch items as they occur than to buy insurance at a premium high enough to cover claims and administrative expenses.
Insurance against some narrowly defined crop risks, like damage from hail, were historically available at affordable premiums before federal programs were largely preempted them. However, premiums on the broad form of crop revenue protection that is now popular would be unaffordable if they were set high enough to cover claims and administrative expenses in full.
The three features that make crop revenue risk uninsurable interact with one another. The fact that coverage extends to speculative gains as well as pure losses attracts more farmers to the program. The fact that losses are not independent of the behavior of insured parties serves to increase both risks and losses. More participants plus more losses means that actuarially sound premiums become unaffordable. When all three operate in combination, no unsubsidized private insurer would be able willing to offer the product. Enter the government.
Budgetary consequences of insuring the uninsurable
Because crop revenue insurance is not commercially viable, the government must provide heavy subsidies in order to offer it at all. As Goodwin and Smith explain, the subsidies take three forms.
The first subsidy goes to reduce premiums paid by farmers. In the 1980s, when the present program was first established, that subsidy averaged about 25 percent. Relatively few farmers found coverage attractive. Following new legislation in 1994, the subsidy rate increased to around 60 percent. As a result, the number of acres covered jumped sharply. Covered acreage is now some five times higher than it was in the early 1980s, when subsidies were lower. According to USDA data, premium subsidies have averaged 72 percent of the budgetary cost of the program over the past ten years.
A second subsidy goes to insurance companies to cover their administrative expenses. Without the reimbursement of administrative expenses, the private companies that offer the crop insurance would have to charge even higher premiums, and the premium subsidies would have to be even higher to make them affordable. Administrative expense reimbursements accounted for another 21 percent of the budgetary cost of the program.
The remaining 7 percent of the budgetary cost of the crop insurance consists of other administrative costs, including the costs of designing new programs for specialty crops, which the USDA must consider when farmers request them.
The actual cost of the insurance program each year depends not only on the level of subsidies, but also on crop yields and prices. In a bad year, payouts to farmers exceed premiums paid, even when premium subsidies are included. In good years, premiums plus subsidies exceed payouts. As the following chart shows, there have been ups and downs in program costs caused by changes in growing conditions, but the trend has clearly been up.
If Congress enacts current proposals to sweeten the program with shallow loss coverage, costs are certain to continue to increase, by an estimated $3 to $7 billion per year. The lower estimate assumes that crop prices will remain at or near their recent record highs; the higher estimate assumes that crop prices return to their long-term trends.
What are the alternatives?
The crop insurance program is sometimes justified as a correction for a “market failure” on the grounds that without subsidies, private companies would not write policies that guaranteed farm revenues. However, the fact that no one is willing to supply a service that customers are unwilling to buy at a price that covers its cost is not a market failure in any meaningful sense. Instead, it is a sign of market success, since the proper function of markets is not just to reward winners but also to winnow out losers.
Fortunately, there are alternatives. Critics have proposed several reforms that would reduce the budgetary costs of crop insurance without destroying the capacity of American farms to produce reliable supplies of food and fibers.
The simplest reform would be phase out federal crop insurance altogether. Doing so would not, as some claim, leave farmers without any tools of risk management. For many crops, futures markets allow farmers to lock in prices at the time of planting. For narrow risks such as hail damage, private insurance could fill the gap left by an end to federal programs. Changes in farming practices, including greater diversification and retreat from planting inappropriate crops on marginal land would further reduce risks.
A seemingly opposite reform would, instead of abolishing crop insurance, make it mandatory for all farmers. Mandatory insurance would reduce the required premium subsidy by mitigating the problem of adverse selection. Adverse selection occurs when farmers who know their risks are low, because of favorable local conditions or low-risk farming practices, do not purchase insurance. The result of nonparticipation by low-risk farmers is a higher rate of claims for those who buy into the insurance pool. Higher claims require higher subsidies, or else even more farmers will drop out of the program. Legislation passed in 1994 introduced a limited form of mandatory crop insurance, but Congress repealed those provisions in 1996.
Another possible reform would be to better integrate crop insurance with disaster relief. Under past practices, farmers have often been able to double-dip, collecting both insurance payouts and disaster relief for the same loss.
Still another suggested reform would be to move to a single-payer system under which the government provided crop revenue insurance at no charge to farmers. Doing so would eliminate the need to subsidize the administrative costs of the private firms that now provide such insurance—subsidies that not only help them cover costs but guarantee a 14 percent rate of return. Some analysts think that if free insurance included a reasonable deductible and if it were well integrated with existing disaster relief, it could provide substantial protection to farmers at a lower cost to the budget than the present public-private partnership.
Finally, any of the reforms that stop short of ending federal crop risk insurance altogether could include caps on payments to any one recipient. For example, a recent GAO study looked at the potential impact of a $40,000 annual cap on premium subsidies paid to any single operator. It found that such a cap would affect just 3.9 percent of all farms, yet, since those few large farms account for a third of all subsidies, it would shave $1 billion per year off the budgetary cost of the program.
The bottom line
Legislation now before Congress proposes to replace direct payments to farmers with expanded crop insurance. To some, this sounds like a good idea. Any fool can see that it makes little sense to pay billions of dollars to farmers whether they plant crops or not. The hope is that indirect payments, disguised as “insurance,” will not appear so transparently foolish. Yet, they are even worse. Replacing direct payments with shallow-loss crop revenue insurance would not only increase costs to the budget, it would further distort incentives and reduce farm efficiency.
The bottom line is that the proper framework for thinking about crop insurance is not market failure, but program failure. Crop revenue risks are inherently uninsurable. The time to end the charade of subsidized federal crop “insurance” is now.
3 Responses to “The Dubious Economics of Crop Insurance”
at least in theory i get "more crops." my beef is with federal flood insurance. we had just a few houses wiped out by floods up here…FEMA? nothing! wipe out 2nd homes that haven't been occupied in New Jersey etc for years? BILLIONS. absolutely ridiculous. we have so many subsidies for "housing" in this country i think we should start with that before we go after food supply. it's not like the demand isn't there given the food stamp program.
I suspect that the political reality is that the amount of money one way or another distributed to the farm industry will not be allowed to go down in practice. It may be called by different names, but in reality, well, there are 42 senators from farm states…
Which means calling this "federal ag PUT option" what it is won't actually help.
I agree flood insurance is also a wasteful program, probably in dollar value more wasteful even than crop insurance, but I don't see why not to go after both of them.