Government benefit programs are sometimes called "social insurance," but what exactly is being insured against? Look at the prescription drug benefit that died—for this year—in the Senate on Wednesday. Differences between this proposal and the one that passed the House in June do not loom large to the naked eye. Both parties claim to favor drug coverage for the elderly, and what they are quarreling about is as unclear as the philosophical basis for the plans they have come up with.
Under the House-passed plan, seniors would pay about $33 a month and then all of the first $250 of drug costs each year, then pay a fifth of the cost up to $1,000, then half of costs up to $2,000, then pay all of the costs again until they were out-of-pocket by $3,700 (which, if my math is correct, would happen when total costs hit $4,400), then pay nothing on drug expenses beyond that. Except that people with incomes less than 150 percent of the poverty line (about $9,000 for an individual and $12,000 for a couple) would pay nothing on costs below $2,000, and those between 150 percent and 175 percent of poverty would pay part of the monthly premium—on a sliding scale, of course.
Under the plan defeated in the Senate, seniors would pay $25 dollars a year and then 95 percent of drug costs up to an out-of-pocket $4,000 (which means total drug costs of around $4,185), and then nothing on annual costs after that. But seniors under 150 percent of the poverty line would pay only a few dollars for each prescription—the exact amount depending on whether it was a generic or preferred brand. The Congressional Budget Office estimated the 10-year cost to taxpayers of both plans at around $400 billion.
Thought of as insurance, the crazy-quilt set of benefit levels and cutoffs in the House bill and the slightly-less-crazy quilt rejected by the Senate were both attempts to insure against two different risks. One is the risk of being poor. The other is the risk of large drug costs. The principles underlying these efforts are that nobody should have to go without needed medicines because of a lack of money, and that no one—whatever his or her financial situation—should have to pay more than about $4,000 a year for needed medicines.
Put like that, the second principle seems odd. What is the difference, in terms of need for a subsidy from other taxpayers, between a senior with a yearly income of $30,000 but drug expenses of $5,000, and a senior with income of $26,000 but drug expenses of only $1,000? Both have incomes after pills of $25,000. Why is the first one more deserving?
There is a traditional argument—and not a bad one—that important social insurance benefits should be universal because this creates a sense of common citizenship. Although taking money from people in taxes and giving it back to them in benefit checks seems pointless, in the accounting sense, unless there is some kind of redistribution and/or behavioral incentive going on, the take-and-give-back does bond people to one another and also to the program itself. Social Security is the prized example.
That is why one of the disagreements between Democrats and Republicans that is preventing a prescription-drug benefit is whether it should be funded through Medicare or through private insurers. But the argument seems silly in this case. A benefit so scattershot is not going to generate any warm national fuzzies just because it is administered by a government agency. On the other hand, because it is actually a subsidy and not insurance in the business sense, turning it over to the private sector is a symbolic exercise that is unlikely to produce many efficiencies.
The idea of insurance is to share the risks of bad outcomes. You don't know whether your house is going to burn down next year, and neither does the insurance company. But the company knows the odds, and with enough customers the odds turn into a certainty that some percentage will suffer a fire. That cost, plus some vigorish, then gets spread around. But insurance depends on uncertainty. If your house is certain to burn down (never mind why …), you either have to keep that fact from the insurance company or be prepared to pay the insurance company just as much as it will ultimately pay you, plus the vigorish—which is pointless.
If social insurance worked the same way, poor people would have to pay more, not less, for income-based benefits because the chance of your being poor next year is a lot higher if you're poor already. Social insurance, however, does not depend on the future being unknowable. It insures against certainty or near-certainty. That is why it requires a subsidy. In fact if it doesn't require a subsidy, it is like fire insurance for arsonists: possible, but pointless.
When Congress takes up a drug benefit again, it should keep things simple and concentrate on the risk, approaching a certainty, that it wishes to prevent: people doing without drugs—or without food—because of the cost. That means concentrating on poor people. The risk that drug prices will move you down a notch in the middle class is not something an entire society can insure itself against anyway.