As time grows short and it looks more and more likely that the government will not increase the debt limit by August 2, the question of whom the Treasury Department should pay if it has to make due with current revenues becomes more interesting. There have been a fair number of off-handed comments about this issue in the press, but I have yet to see much written on how to think about the problem conceptually.
One approach is to approach the problem like a business or individual that has a serious cash flow problem but isn’t insolvent. This would suggest ordering payments based on how much the government is likely to need particular creditors on an ongoing basis going forward. If an individual can’t pay all the bills, he is likely to first pay the mortgage (or at least this was true before mortgage defaults became so common that lenders lack the capability to foreclose quickly) and the utilities. If the bank takes your house or the power company refuses to provide heat next month, you have a serious problem. You can stiff the phone company and the cable company because it is less of a problem if they refuse to continue doing business with you. The plumber who already repaired your leaky faucet is even easier to stiff because you don’t need him in the future at all. (Yes, you might have another plumbing problem but unless the plumber gossip network is extremely efficient you will be able to find another one).
This view implicitly underlies the common assumption that bondholders will be paid first. It isn’t that they have some superior moral claim to be paid compared to, say, social security recipients or federal employees. The problem is that, if we don’t pay our bondholders, they (and others like them) are likely to refuse to loan us more money in the future, or at least demand more compensation to do so. (The bondholder gossip network is extremely efficient thanks to the news media and rating agencies).
Based on the principle of paying the people whose cooperation the government most needs in the future, who comes after the bondholders? If the government thinks like a business, it should be contractors who provide the most essential services and who are most likely to refuse to continue to work if they aren’t paid. Social security recipients and soldiers would come last. Social security recipients are done paying into the system and current workers don’t have a choice, so it is of relatively little concern if people don’t fully trust the government’s promise to make social security payments in the future. Soldiers aren’t allowed to walk off the battlefield. Federal employees are also relatively easy to stiff. In the long run they might look for new jobs, but the transaction costs are too high for most to do so if the cash-flow crunch is relatively short lived.
If government officials think in terms of which creditors are most important to them personally rather than to the government as an entity (think managers who serve their own interests rather than those of the corporation’s shareholders), the calculus changes drastically, and the question is not how keep the loyalty of the nation’s most important business partners but how to keep the loyalty of the greatest number of voters. This perspective points toward making sure that the social security checks go out. There are a lot of social security recipients, and they vote.
A very different approach would be to allocate incoming revenues based on the principle of equality. Rather than paying some in full and others not at all, this approach would counsel toward paying everyone to whom the government has a moral obligation — as a result of contract, promise, or reasonable reliance on the expectation of payment — the same amount. This approach would lead the government to pay bondholders, social security recipients, Medicare and Medicaid providers, etc. etc. 60% of what they have been promised (they get an IOU for the rest).
A third approach is to pay in order of the recipients’ need. Those who want to pay social security recipients first are implicitly adopting the need principle, based on the terrible thought of old people being evicted and eating dog food. The need principle has a lot of appeal, and especially so if we assume everyone will ultimately be paid and the only issue is timing, but it would be very tricky to implement given that decisions would have to be made on a class-by-class basis, not individual-by-individual. Many seniors could make do without social security better than many government contractors or federal employees could live without their checks. And while it might be tempting to think that doctors who care for Medicaid patients are better positioned to wait for their reimbursement than poorer creditors, if those doctors stop caring for Medicaid patients, some of the most vulnerable in the country will suffer substantially as a result.
Yet another approach is to try to pay those with the strongest moral claim to receive money from Uncle Sam first. One might argue that this principle would dictate that soldiers serving in Afghanistan or Iraq get paid first, since they are risking their lives to protect the rest of us. This would be even more difficult to implement than the need principle, however, because it arguably requires the government to try to figure out either with which of our creditors we have negotiated the best “deal” or which of our creditors provide services to the nation based in whole or in part on altruistic motives rather than due to naked self interest.
Ultimately, of course, the payment order will be determined based on a messy combination of some or all of these principles, but debates about (a) which principle(s) should take precedence and (b) what the order of payment should be under each of the various approaches would help to rationalize the decision making process.