The United States is bumping up against the debt ceiling once again. The Treasury department has begun taking extraordinary measures to ensure it can continue to service its debts on time. Secretary Yellen has said these measures should be sufficient until June. Some project they might work until August.
Many pundits act as if there is no good reason to have a debt ceiling. They say Congress implicitly approves an increase in debt when it passes revenue and spending bills. They think requiring Congress to explicitly raise the limit on issuing debt is unnecessary and—since it increases the odds of default—unnecessarily risky. Some go so far as to describe it as lunacy.
David Beckworth discussed the debt ceiling with Michael Strain on a recent episode of Macro Musings. Strain, who is the Director of Economic Policy Studies and the Arthur F. Burns Scholar in Political Economy at the American Enterprise Institute, provides a clear statement of the view described above:
[…] the gap between spending and revenues is implicitly determined by Congress. It doesn’t make any sense to require Congress to explicitly authorize these periodic debt ceiling increases.
The debt ceiling puts the President and the Executive Branch in a weird position of kind of picking which laws it will break and will not break. You know, it would be illegal for the Treasury department to issue additional debt if Congress didn’t raise the debt ceiling. It is also illegal for the Treasury department not to spend the money that it is required by law to spend. So, which of those two laws is Treasury going to break?
Given Congress’s implicit consent to raise the debt ceiling, Strain argues, there is no reason to require explicit consent.
I am sympathetic to Strain’s broader argument. We are a nation of laws. Our laws should be consistent. The government should not make promises and then fail to keep them. We should keep our promises and honor our debts.
That said, I don’t buy the implicit consent argument. And I don’t think the debt ceiling is lunacy. To the contrary, I think the debt ceiling might serve as a useful constraint.
On Implicit Consent
To be honest, I have never quite understood the implicit consent argument. If passing a spending bill that exceeds the revenues authorized by Congress implies consent for raising the debt ceiling, why is raising the debt ceiling so difficult? That it is so difficult suggests that a majority in the House or Senate has not already consented to a higher debt level.
We have known that majority voting systems can produce inconsistencies since the 18th century. One should not assume Congress prefers A to B just because they have indicated they prefer B to C and C to A. They might also prefer B to A. Therefore, it is inappropriate to assume that passing a spending bill that exceeds authorized revenues necessarily means Congress is willing to raise the debt ceiling. And, if Congress is not willing to raise the debt ceiling, it is not obvious why its decisions on revenues and spending should force it to change its decision on the debt ceiling. One might just as easily argue that its decision not to raise the debt ceiling should force it to change its decision on revenues or spending.
On Constraints
A binding constraint prevents one from doing what he or she would like to do. Relaxing such constraints, therefore, generally makes us better off. But it is naive to think all constraints are bad. Sometimes constraints make us better off.
Think about dieting. In principle, anyone can lose weight: just burn more calories than you take in. And, yet, many of us struggle to do so. You might say, “Just don’t eat the chips. It’s inconsistent with your goal of losing weight.” But, for some of us, having chips in the house is a temptation we can’t resist. If we are serious about losing weight, we adopt hard rules like “Never buy chips.”
Can one lose weight while still enjoying the occasional handful of wavy Lay’s? Sure. But, if we are likely to err in the eat-too-many direction (and likely to do so frequently), going cold turkey might be a better option.
The Potential for Strategic interaction
Richard Thaler and H. M. Shefin describe self-control problems, like the dieting decision discussed above, as a problem of two selves. My future self wants to be thinner. My current self wants to eat chips. The hard constraint aligns the decisions of my current self with the goals of my future self.
The need for hard constraints is perhaps even more clear when we are dealing with truly separate selves: different people with different objectives. With separate selves, we must think seriously about strategic interactions: individuals make decisions today based, in part, on the decisions they expect others will make in the future. If there is scope for strategic interactions, hard constraints might make us all better off.
There is certainly scope for strategic interaction in Congress. Congress is not a single acting entity: the decisions of Congress are actually the decisions of a majority of its members. Moreover, the majority specifying how much to spend might not be the same as the majority specifying how much can be raised in taxes. And the decisions individual members of Congress make with respect to how much to tax and how much to spend depend, in part, on the decisions they expect others to make in the future. By treating Congress as a single acting entity, those advancing the implicit consent argument obscure the potential for strategic interactions.
A simple example serves to illustrate. Suppose a Senator supports higher spending and higher taxes. A spending bill comes up, which he would happily support if he knew taxes would be increased to cover the additional spending. His colleagues assure him that they will support a tax increase in the near future.
What‘s a budget-balancing Senator to do?
If he doesn’t support the spending bill today, he risks seeing that bill fail. That’s bad, in his view, because he thinks the additional spending is warranted.
If he supports the spending bill today, he faces the risk that his colleagues will renege when a tax bill comes up in the future. That’s bad, in his view, because he thinks the additional spending should be paid for with additional tax revenues.
The potential for parties to engage in ex-post negotiation makes it harder to reach an agreement ex-ante.
Enter the Debt Ceiling
The debt ceiling provides a limit on the potential for ex-post renegotiations. Our hypothetical budget-balancing Senator can vote for the spending bill today, knowing that in the future his colleagues will be forced to grapple with the imbalance. They must either raise taxes (as he’d prefer), forgo the additional spending (his second best policy outcome), or—if he’s in the minority—raise the debt ceiling.
Obviously, this is a simple example. But it illustrates an important point: sometimes constraints make us better off.
Conclusion
It’s tempting to think that all constraints are bad. In fact, we often impose constraints on ourselves and our counterparties to achieve better outcomes than would be possible in the absence of constraints.
Left unconstrained, many individuals will eat more calories than they should. That’s why they impose constraints on themselves! Likewise, unconstrained politicians are likely to authorize more borrowing than they should. The debt ceiling might provide a useful—if somewhat limited—constraint against excessive borrowing.