The Case for State Borrowing as a Response to the Current Crises

Introduction

The coronavirus pandemic is a national emergency that requires a national response. Asking states to absorb the budgetary losses caused by the pandemic while they are tasked with providing essential frontline services is comparable to asking states during World War II to pay for the landing in Normandy.

This article is a contribution to Project SAFE: State Action in Fiscal Emergencies.¹ We have already argued, more than once², that the federal government should borrow to prevent steep state and local budget cuts. But because the federal government will apparently not take sufficient action,³ we offer these ideas to states for how to proceed with borrowing absent sufficient federal aid.

Additional Clarification

We are only encouraging states to borrow against the backdrop of federal failure. Some have argued that having states borrow to cover revenue shortfalls caused by the pandemic would be a better option than direct federal grants,⁴ but we are emphatically not making that argument here. At least sometimes, the contention that state borrowing would be better is based on a notion of limited federal borrowing capacity. Yet we think this notion is both false and often hypocritical. It is false because the federal government does have substantial borrowing capacity,⁵ certainly more than states do, and it is also hypocritical for anyone who supported the Tax Cuts and Jobs Act to make this claim. This is because the TCJA incurred over $1.5 trillion in federal debt during an economic expansion in order to mostly benefit the already well-off. One obvious problem with pursuing this kind of reckless policy is precisely that there might not be enough borrowing capacity in an emergency.⁶

Sometimes this borrowing-is-better argument is instead couched in terms of fairness between states.⁷ This claim is also off base. For one, it seems to assume that only profligate high-tax states are suffering.⁸ This is not true; almost all states will suffer collapsing revenue, and many of the so-called high-tax states did not enter the coronavirus crisis in poor fiscal shape.⁹ Further, there is no reason that federal aid cannot be limited to help with the current crisis (that is, no help with accrued pension debt), perhaps even with some kind of per capita limit so as not to reward supposedly overly generous states. A per capita grant that makes Texas or Florida whole would be a huge relief for California, too.

Raising Revenue Is Possible and Preferable

Federal failure has prompted us to encourage states to raise revenue so as not to just make deep cuts to services.¹⁰ However, we are not against states also pursuing smaller cost-cutting measures like hiring freezes.

Many revenue-raising tools are available, and we previously urged state legislatures to consider a number of promising reforms to cope with the current crisis. Indeed, because state and local revenue systems were in such need of reform even before the beginning of this crisis, there are lots of sensible reforms to raise significant revenue fairly and efficiently.¹¹

Because states face balanced budget constraints, the fact that we are in a recession is not an argument against raising taxes.¹² States and localities should prioritize raising tax revenues from those who can best pay over slashing services needed by those who cannot afford to lose them (or their jobs), especially because many of those services play important roles in preventing further harm to state economies.¹³

The Case for State Borrowing

Another objection to the strategy of raising revenue is that it might not be enough. The cost of the pandemic is so large that even a state that takes us up on our complete menu of revenue-raising tools might still come up short. Current estimates of the need range from $500 billion to $1 trillion.¹⁴ Revenue tools might not be enough,¹⁵ but that does not mean states should not try to prevent unnecessary cuts to the extent that they can.

The deeper counter to this objection to revenue raising is that states can and should borrow, with the borrowing secured by the new revenue streams. It seems quite reasonable to surmise that in three years state finances will be looking a lot better, even without new taxes. Indeed, state revenues will likely look better still in three years if they do not eat their seed corn now by, for example, reducing funding for vital services. Consider that if states worked on five-year budget cycles, a budget planner would be well advised to borrow from later in that cycle to pay for necessary expenses now, at the nadir.

Thus, the basic idea is that states should smooth their taxing and spending by borrowing from a better future to sustain the challenging present. To ensure that there will be more revenue in the future, states should improve their revenue systems now — and there is a lot to improve. Because there is so much to upgrade, putting into place a suite of revenue enhancements now might substantially help states and localities through the ongoing emergency.

This is a different proposal than simply borrowing without reform, because borrowing without putting into place new revenue can be hamstrung by one of two problems. First, the borrowing could be too small. In that case, there is little harm done to the state’s future revenue, but it has done little to address the enormous current crisis. On the other hand, a state could look to borrow on a scale sufficient to the crisis, but borrowing like that could place a large burden on the state down the road without new revenues.¹⁶

To be specific about securitizing revenue, suppose California put into place our proposal to broaden its corporate income tax base by shifting to mandatory worldwide combination.¹⁷ In better times, this was estimated to raise $2 billion per year, a roughly 20 percent increase in state corporate income tax revenues. Suppose California securitized 10 percent of its corporate income tax base — an estimated $1 billion per year — for 20 years at a 5 percent rate. That would be worth about $12 billion to California right now, which is about 85 percent of the $14 billion that the state is hoping the federal government will provide to head off deep cuts.¹⁸ The cost of federal funds is closer to 1 percent for a borrowing for 20 years — so if the Fed were to purchase these bonds, as we have argued it should,¹⁹ then California could raise $17 billion through this expedient.

Note that before the onset of the coronavirus, California was running a surplus without this reform. So even if it were to raise 50 percent less than expected, the state’s net fiscal position when the recession ends would still be the same as it had been expected to be before the crisis and this borrowing.

But we do not mean to be overly prescriptive about the need for pairing new revenue with borrowing. First, the possible crisis that states are heading into is so severe that it would be worth it to cannibalize the future at least a little bit. Put another way, the future crisis in revenue might be even worse if states allow their economies to contract too much now. Second, states like California that had been running a surplus really can roughly borrow from future surpluses. Third, states like New Jersey have recently implemented many smart corporate income tax reforms with favorable results.²⁰ It would be odd that only those states that have still not made common-sense reforms should borrow. Fourth, states can always raise taxes later, and the politics of doing so might be different during a recovery.

In the end, however, it would be better for new borrowings to be paired with new revenue. One compromise would put in place tax increases that would only kick in as the economy expands or if revenues come in too far below what is needed.

What About Borrowing Limitation Rules?

Special debt rules have been part of American public finance for almost two centuries, and for a reason: It is tempting for current generations, especially politicians, to saddle the future with debt. States incurred huge debts in the 1830s building infrastructure (such as canals), and many of these projects ended up in the red when the economy turned. This experience of states facing default and near-default led to special constitutional rules as to incurring debt.²¹

However, to review that history and the associated rationales for borrowing limits is to understand why the current situation is different. The current generation is not trying to saddle the future with a speculative white elephant, but rather must ensure that the future generation is given schooling and other services at least roughly comparable to what was being provided just a few months ago. States could hardly have socked away $500 billion in case of pandemic. And in fact, as we already noted, seared by the Great Recession, many states entered into this crisis with a reasonable level of budgetary reserves.²²

So how can this sort of borrowing work? Can a stream of future tax revenue actually be sold in this way (that is, used to back borrowing in advance of when those future tax revenues are raised)? Yes, and indeed, there is a long history of financing new development projects through, in effect, selling the speculative increase in tax revenue that the development is expected to generate. This is called tax increment financing.²³

And there are even more interesting models. For instance, states and localities sold their right to revenue from the giant settlement with the tobacco industry even though it was unclear how much revenue that settlement would bring.²⁴ Those bonds contained features to manage the uncertainty in the future revenues,²⁵ and any issued coronavirus deficit bonds could do likewise.²⁶

But Doesn’t State Constitutional Law Forbid General Deficit Spending?

Even if a good idea, we still must consider obstacles to states borrowing in these ways because of balanced budget rules in state constitutions. Our first response is that, in some states, either amending the constitution or putting a constitutionally authorized borrowing proposal on the ballot for voter approval is not that difficult. Each state should consider holding these votes or elections to authorize borrowing under its rules. The rationale is compelling and should carry the day in those votes or elections.

Of course, we appreciate that (for various reasons) holding votes or elections might not be appealing or possible in some states. But all is not necessarily lost. State rules prohibiting borrowing without an election have long been found by the courts to have important exceptions. We think our proposed borrowing can fit under several of these exceptions, although any state pursuing our approach would have to carefully consider its own specific case law. Here, we can only speak in broad strokes.

As to available exceptions, we will review two because we think the rationales for the legal rules track our policy arguments reasonably well. Some other exceptions might be available as well and could be justified on similar policy grounds.²⁷

First, there is the cash flow borrowing exception. Here is a classic example: Property taxes are typically collected in one or two lump sums by local governments. Should a county be able to borrow at the nadir of its cash flow until it gets its second installment of property tax receipts? Courts in every state we know of say that it can.²⁸ As we explained, the borrowing we are envisioning is also a cash flow borrowing, albeit over a longer cycle. Thus, so long as the overall structure is reasonable and does not put future taxpayers at any greater risk, we think the sort of borrowing we propose should also fit within this cash flow borrowing exception.

Second, there is the special fund exception.²⁹ Under this exception, if all that investors are promised is a specific revenue stream — and no more — then those borrowings are not considered to trigger the election requirement or other state constitutional bars against deficit spending because the government’s general taxing power is not being promised. This exception is common, but state courts also have varied on whether they interpret it broadly or narrowly. The narrow version of the exception potentially requires that the borrowing be secured only by funds that would not have existed if not for the borrowing — say, by using only toll revenues to pay for a bridge.³⁰ The broader version of the exception focuses on the separateness of the revenue stream rather than the nexus with the project. Generally, prominent commentators have considered the narrow interpretation of the exception more favorably than have the courts.³¹

Though we would not endorse the broader view in every case, we think the narrow view is unduly restrictive. To go back to the beginning, the classic debt problems that ended in trouble in the 1830s involved states backing up the debt of private entities — such as a railroad — with the taxing power of the state. If the project succeeded, disproportionate benefits would flow to some interest groups. Failure, however, would be borne by all.

This blending of public and private was why the debt limitation rules were typically accompanied by constitutional provisions regarding public purpose, gift of public funds, subscription of stock, etc.³² This is also why circumventing debt limitations through privatization is problematic in ways that resonate with the historical problem these provisions were to solve.³³

A revenue bond using securitized general tax revenue — even new tax revenue, which is what we propose as ideal — is also a burden on taxpayers, but it is a cabined liability in time and amount. Importantly, our proposals would not involve taxpayers taking on unknown risks, much less unknown risks based on speculative projects particularly likely to benefit a powerful few. Indeed, to the extent that the taxpayers get money now, while bondholders have to wait to see how much revenue that tax increases yield, taxpayers have shifted risk from themselves to the investors, effectively the opposite of the classic problematic financings of the 19th century.³⁴

It is true that the financings we propose could balkanize the general tax base, and there are good reasons to be concerned with this practice. But this was not the primary target of the debt limitation provisions, and we think courts should be wary of constitutionalizing this concern. This is because there is also an argument that earmarking taxes does not destroy a tax base, but preserves it because voters or interest groups will protect a tax with clear benefits or beneficiaries.³⁵ We are not taking sides on the earmarking debate; our primary concern is to argue that courts should also not take sides absent a clear mandate. To fully fit within our argument and the law in some jurisdictions, it might be necessary for states to securitize new revenue in this way.

Do State Constitutional Limits Even Apply In an ‘Emergency’?

Beyond our earlier discussions, some state debt limitation rules apply “except in case of war to repel invasion or suppress insurrection.”³⁶ Those exceptions support the legal case for borrowing under any exception. At the broadest level, the emergency exception indicates an understanding that borrowing in response to an emergency is not subject to the same political process concerns as the speculative financings of the 19th century. Thus, a state court that has an exception of this kind in its borrowing limitation has a textual warrant for the proposition that debt limitation rules were not meant to prevent borrowing in case of emergency. This could be a useful complement to using some other exception.

In at least one state, there is no need to use the emergency language as a complement. In New Jersey, this debt limitation provision grants an exception for “acts of god,” and so borrowing in response to the pandemic would seem to be plainly permitted.³⁷ Appropriately, the New Jersey Legislature approved an almost $10 billion borrowing authorization. The statute was challenged, but the New Jersey Supreme Court unanimously upheld the law.³⁸

In states with the narrower insurrection/invasion-type language, we think that the legislature could still argue that no other exception is needed by appealing to the canon of ejusdem generis — these are the types of emergencies that suspend the usual rule — rather than that of expressio unius, which would limit the emergencies to the ones named in the constitution. And this makes sense: Why would the authors of the state constitution except emergency spending to treat the victims of a war, but not a pandemic?

At least a few state courts have interpreted their provisions broadly in roughly this manner. For example, the Washington Supreme Court upheld a bond issue for unemployment relief in the midst of the Great Depression.³⁹ The court reasoned that mass unemployment had caused signs of insurrection, and that the Legislature’s determination that an “incipient” insurrection existed was conclusive “unless, giving effect to every presumption in its favor, the court can say that such legislative declaration, on its face, is obviously false and a palpable attempt at dissimulation.”⁴⁰ The California Supreme Court found the issue whether an insurrection existed to be a political question and permitted borrowing in connection with bonuses to Civil War soldiers⁴¹ and the financing of a railroad.⁴² Note that the California bonus provision was made during the Civil War. Several state supreme courts upheld borrowing for bonus provisions for soldiers made after World War I.⁴³

If a state does plan to use any of these exceptions, then it would be prudent to also create a fast-tracked procedure so that the courts can hear any possible legal objections quickly.

Conclusion

To go back to the beginning, we reemphasize that what we are proposing here is inferior to the federal government stepping in to do its job adequately.⁴⁴ However, especially if the only other feasible choice is savage cuts to needed spending programs, we consider borrowing of the sort that we have proposed to be the far superior option.