Issue Brief: Volume 2, Number 1 

The Education of Detroit’s Pension and Bond Creditors

Author: David Skeel

The unprecedented bankruptcy of Detroit was a shock not only to the residents, but to pension and bond holders in the city.  Although there are protections in place for pension beneficiaries and bondholders, they do not necessarily cover loss when a municipality files for bankruptcy.  Although the uncertainty about the pensions’ and bonds’ status is inconsistent with basic rule of law expectations, it also is temporary; and permitting pensions and bonds to be restructured in bankruptcy would bring considerable benefits.


  • The example of Detroit raises the question of whether state lawmakers should look for ways to more fully protect pension beneficiaries and bondholders. Several states that previously did not have special protections in place have recently taken steps in this direction.
  • Although these protections would spare future pension beneficiaries or bondholders the fate of their compatriots in Detroit, there are two very serious problems with them.
  • First, if these obligations—pensions especially—cannot be restructured under any circumstances, some U.S. municipalities will be incapable of addressing their financial distress.
  • Moreover, these cities may be forced to cut back dramatically on services, with substantial consequences.
  • The future of municipal governance actually would be brighter if pensions and government obligation bonds can be restructured, at least a little, in bankruptcy.

The Detroit residents who must wait fifty-eight minutes for a response to their 911 calls and whose streetlights do not work are not the only ones who are dazed and confused. Many of Detroit’s bond and pension holders thought they were nearly certain to be paid in full, even after Detroit filed for bankruptcy. But both are almost certainly mistaken. Bankruptcy judge Steven Rhodes explicitly stated in his recent ruling uphold- ing Detroit’s eligibility to file for bankruptcy that the pensions can be restructured, and this ruling is likely to be upheld on appeal. The ruling did not directly address the status of Detroit’s general obligation bonds, but these creditors too are likely to be forced to accept considerably less than one hundred cents on the dollar.

Why are both constituencies likely to receive so much less than they initially imagined?

One reason is simply that municipal bankruptcies are uncommon, and for major cities unprecedented. Even Rick Snyder, Michigan’s governor, proclaimed a bankruptcy for Detroit unthinkable only three years ago. Because creditors did not expect even so financially distressed a city as Detroit to file for municipal bankruptcy, they paid less attention than they should have to the possible implications of a bankruptcy filing.

But there is another, more significant reason for Detroit pension beneficiaries’ and creditors’ confusion as well. It isn’t just that bankruptcy is uncommon, and thus wasn’t on their radar screen. Both pensions and general obligation bonds have a radically different status, and much more protection, outside of bankruptcy than they do when a municipality files for bankruptcy. Both constituencies seemed to assume that the protection was, like a diamond, forever. But both constituencies were wrong.

In this Issue Brief, I will describe the basis for pension beneficiaries’ and bond- holders’ beliefs that they are fully protected, and explain why both are almost certainly mistaken about the efficacy of these protections in bankruptcy. I then will describe recent efforts to protect pensions in Illinois and bonds in Rhode Island, and argue that extending these kinds of protections would be a mistake. Although the uncertainty about the pensions’ and bonds’ status is inconsistent with basic rule of law expectations, it also is temporary; and permitting pensions and bonds to be restructured in bankruptcy would bring considerable benefits.

The Pension Bombshell

Last June, shortly before Detroit filed for bankruptcy, Emergency Manager Kevyn Orr filed a report that served as a rude awaken- ing for many of Detroit’s creditors, including its pension beneficiaries. Orr had been appointed at the end of March, pursuant to controversial state provisions that permit the governor to select an emergency man- ager who would largely displace the mayor and city council if a city is in financial dis- tress. Contrary to the trustees of Detroit’s two major pension funds, who claimed that the pensions were adequately funded, Orr’s report estimated that they were underfunded by $3.5 billion. Even more radically, Orr also insisted that the pensions would need to be reduced, a step that pension beneficiaries insisted is impossible under Michigan law.[1]

The key to the standoff between Detroit’s emergency manager and the pen- sion beneficiaries is a provision that was added to Michigan’s state constitution in 1963. This provision states that accrued pension benefits “shall not be diminished or impaired.”[2] To pension beneficiaries, the provision sounds like a trump card assuring that their pensions can’t be touched. Outside of bankruptcy, they may well be right.[3] And properly funded pensions are probably protected even in bankruptcy. But if a city has failed to fully fund its pensions, the unfunded portion of the pension may be subject to restructuring in bankruptcy. How can this conclusion be reconciled with the Michigan state constitution, with its very clear promise that pension benefits “shall not be diminished or impaired?” The simple answer is that federal law takes precedence over state law—even state constitutional law—under the Supremacy Clause of the U.S. Constitution.[4] Because the U.S. Constitution authorizes Congress to enact bankruptcy laws, and the municipal bankruptcy law that Congress has enacted permits a city to restructure its ordinary debts, the unfunded portion of a pension can be restructured even if state law seems to say pensions are sacrosanct.

Bankruptcy’s overriding of the Michigan constitution may seem problematic at first glance, but it makes a great deal more sense once we add two key details. The first is that it is an exaggeration to say, as I have just done, that bankruptcy over-rides Michigan’s constitution. Michigan’s prohibition against impairing pensions was not added with bankruptcy in mind at all. Prior to 1963, a pension promise was simply a “gratuity” in Michigan, as in many other states. A city like Detroit could withdraw the promise at any time, even after a school teacher or fireman had worked for the city for decades and was about to retire. Michigan lawmakers wanted to put pension promises on sounder footing, by making them enforceable contractual obligations. This does not mean that the obligation could never be restructured, even in bankruptcy; it means that a city like Detroit cannot simply decide to withdraw its promise. Michigan lawmakers could have gone further, and forbidden even the most financially distressed city from filing for bankruptcy if they had wished to do so. Cities can only file for bankruptcy if their state consents to municipal bankruptcy filings. This gave Michigan the power to just say no. But Michigan has permitted cities to file for bankruptcy since 1939, shortly after the first permanent municipal bankruptcy law was enacted. Indeed, not only did Michigan authorize municipal bankruptcy, but Michigan lawmakers such as Frank Murphy, the mayor of Detroit and later governor and then a U.S. Supreme Court justice, were among the most vigorous advocates for municipal bankruptcy in the 1930s.

The second key fact is that the Detroit pensions are likely to be protected to the extent they are adequately funded. The funds that Detroit and its employees contributed to its two major pension funds belong to the beneficiaries. A court is likely to conclude that Detroit and Detroit’s other creditors do not have any interest in the funds. Notice the implication of this: if Detroit had properly funded its pensions, the pension beneficiaries would not have anything to worry about. In fact, the same Michigan constitutional provision that prohibits the impairment of pensions requires that the pensions be fully funded each year as they accrue. Detroit seems to have simply ignored this constitutional obligation.

I should point out that my comments about the status of the Detroit pensions are somewhat speculative. Prior to the recent decision by Detroit’s bankruptcy judge holding that Detroit is eligible to file for bankruptcy, no judge had directly ruled on the question of whether a pension can be restructured in bankruptcy. Judge Steven Rhodes ruled that pensions can indeed be restructured, and seemed to suggest that even the funded portion may not be protected.[5]
I suspect that he will not go this far if and when Detroit does in fact propose a restructuring. But he has made clear that pensions are not immune from restructuring in bankruptcy.
The importance of this conclusion cannot be overstated. The Detroit decision is currently on appeal, and it could quite easily make its way to the Supreme Court. If the conclusion that pensions can be restructured is upheld, other major cities may think seriously about the bankruptcy option. Chicago, Philadelphia and other cities face severe structural deficits, and in each case by far the most serious problem is potentially unsustainable pension obligations. If pensions can be restructured, municipal bankruptcy may be a solution for an egregiously stressed city. If pensions cannot be restructured, municipal bankruptcy will offer little relief, and we are not likely to see any additional bankruptcy filings by major cities.

General Obligation Bonds: The Meaning of Full Faith and Credit

Detroit’s general obligation (GO) bond- holders are in an oddly analogous predicament. They too seem to have assumed that repayment was assured, and that their apparently fortified status would be recognized in bankruptcy. Like Detroit’s pension beneficiaries, its bond holders are almost certainly mistaken.
Why were the holders of GO bonds so sure they were fully protected? Part of the answer has nothing to do with the bond- holders’ legal status, and everything to do with the municipal bond market’s assumption that a major city would never file for bankruptcy, as I mentioned earlier. But the main reason for GO bondholders’ optimism about their status is that they are different than general creditors in several important respects. Unlike with other debt, such as the certificates of participation issued when Detroit issued debt to fund its pensions,the GO bonds were explicitly approved by Detroit’s voters. In addition to authorizing the borrowing, the voters pledged that the GO debt would be backed by Detroit’s “full faith and credit.” Shortly before and after Detroit’s bankruptcy filing, I had conversations with bond market participants who assured me that the full faith and credit provision gave GO bonds more protection than other debt, such as revenue bonds that are secured by water, sewer or casino revenues. Unlike revenue bonds, which have only a limited source of funding, the reasoning went, GO bonds are backed by all of Detroit’s revenues.

Although there is a certain logic to these assumptions, the belief that GO bonds are protected and revenue bonds are vulnerable is upside-down from a bankruptcy perspective: in bankruptcy, revenue bonds are protected and GO bonds aren’t.  With revenue bonds, the bankruptcy laws treat the specified revenue source as truly belonging to the bondholders, and as securing their repayment. Unlike most other creditors, revenue bondholders are explicitly permitted to continue collecting the payments even during the bankruptcy case, thanks to a 1988 reform that sought to make sure that bankruptcy does not interfere with the payment of revenue bonds.[6]GO bonds are not entitled to any such protection. The “full faith and credit” protection is not treated as giving GO bondholders a right to any particular source of revenue. They are simply general creditors, and Detroit’s obligations to them can be restructured just as its obligations to other creditors can be.

Not surprisingly, GO bondholders have crafted arguments that they too are entitled to protection. A number of Detroit’s GO bonds have contractual provisions saying that the bonds are entitled to payment from all funds legally available to the City, including ad valorem property taxes. Insurers of some of these bonds are now arguing that their bonds are actually revenue bonds, and that they are entitled to all the protections of revenue bonds.[7]The insurers have now filed a motion in the bankruptcy case asking the judge to set aside Detroit’s ad valorem taxes for the benefit of the bonds.

Notice that this argument does not claim that GO bonds are entitled to special treatment; rather, the insurers are arguing that the bonds actually are revenue bonds, not GO bonds. Although the argument is plausible, at least for the bonds involved in the suit, the bonds are rather different than ordinary revenue bonds. They are linked to particular revenues only in the sense that Detroit was required to increase its ad valorem taxes if it did not have enough funds for payment. Not until bankruptcy was imminent did the bond insurers suddenly ask for revenues to be segregated on their behalf. They seem to have hoped that if they started acting like holders of a revenue bond, the court might treat them as such. But it seems more likely that the bankruptcy court will treat even these GO bondholders and insurers as general creditors.

Although Detroit’s bondholders and pension beneficiaries have mixed their legal arguments with a heavy dose of posturing, both constituencies seem to have genuinely believed that the law was on their side, and that they would be protected even if Detroit filed for bankruptcy. This is the worst possible kind of misunderstanding. While, theoretically, it could have been beneficial for creditors to believe that they have less protection than they actually have,[8] it is hard to imagine anything good coming of their exaggerated belief in their own protection. Certainly policymakers did not have any incentive to signal to bondholders or pension beneficiaries that their status was more precarious than they thought.

Should States Enact Special Protections?

While virtually all large-scale state and local pensions are significantly underfunded, the magnitude of the funding gap varies substantially across geographic areas. In Figure 1, for example, I display the annual contribution necessary per household per year over a 30-year period to bring state and local (e.g. Detroit or Los Angeles) pensions into balance. At the high end are states like New York, Ohio, and California. Interestingly, Michigan is in the middle category.

Figure 1: Required Increases for Full Pension Funding in 30 Years

Given the magnitude of the problem and the likelihood that pensions and bonds will be restructured in the Detroit bankruptcy,[9] state lawmakers may look for ways to more fully protect pension beneficiaries or bondholders. Several states that previously did not have special protections in place have recently taken steps in this direction and there is every reason to believe that others will think seriously about following their lead.

The most important recent state actions thus far have taken place in Rhode Island and Illinois. Two years before Detroit filed for bankruptcy, the Rhode Island legislature enacted a new law purporting to give general obligation bondholders a lien on all ad valorem tax and general fund revenues,[10] Thus, Rhode Island has created as a matter of formal state law the same kind of protection some of Detroit’s bondholders claim their bond contract gives them. The new law was enacted shortly before Central Falls, Rhode Island filed for bankruptcy, and was used to ensure that bondholders were paid in full.

The first question to ask about this provision is whether it is enforceable in bankruptcy. From a legal perspective, there are several potential issues. First, although genuine liens are honored in bankruptcy, liens that apply only in bankruptcy or are not legitimate liens are not enforceable. The Rhode Island law is not limited to bankruptcy but it is debatable whether it creates a genuine lien. Ordinary liens apply to particular assets, such as a type of revenue or property. A lien on “all revenues” doesn’t have this quality. Given that state law now permits a creditor to obtain a lien on nearly any and everything,[11] the Rhode Island law might be upheld. But is a very unusual “lien,” to say the least.

The second issue arises from one of bankruptcy’s clawback rules. In effect, the lien statute diverted revenues from Central Falls to a particular creditor, without giving anything to the town in return. As a result it may run afoul of bankruptcy’s prohibition against “fraudulent conveyances,” at least for the first few years after enactment. (For readers who are interested, I provide the technical legal analysis of this point in the footnote below.)[12]

To my knowledge, no state has purported to create a similar protection for pensions, but Illinois may recently have achieved a very similar effect. In early December, Illinois passed legislation designed to reform its future pension obligations. The legislation includes provisions authorizing the state’s pensions to sue the state if their pensions are not adequately funded, and waving the state’s sovereign immunity for the purposes of the litigation. If Michigan had a provision like this, Detroit’s pensions would have the right to insist on payment from the state, even if Detroit restructured its pensions in bankruptcy.

A simpler way to protect pensions or bonds would be for the state to prohibit its municipalities from filing for bankruptcy altogether. If the state does not permit its cities to file for bankruptcy, the status of pension or bond obligations will be subject solely to state law. At least one state—Georgia—has explicitly forbidden its municipalities from making use of municipal bankruptcy. Prohibiting municipal bankruptcy would remove all of the benefits of bankruptcy, and thus might be seen as too draconian a corrective. The state also could change its mind, and remove the prohibition, more easily than it could withdraw a pension or bond specific protection. But taking bankruptcy off the table is another way to beef up protection for pensions or bonds.

Although these protections would spare future pension beneficiaries or bondholders the fate of their compatriots in Detroit, I believe there are two very serious problems with them. First and most important, if these obligations—pensions especially—cannot be restructured under any circumstances, some municipalities will be incapable of addressing their financial distress. Cities can limit their pension promises to future employees and in some cases adjust their obligations for not-yet-accrued benefits to current employees. They also can cut salaries or even lay off employees to reduce expenses. But for cities like Detroit that have unsustainable accrued obligations to retirees and current employees, none of these measures is sufficient.

Second, protecting one group of creditors will inevitably inflict pain on other constituencies. If a city is unable to restructure large portions of its obligations, it may be forced to cut back dramatically on services or may be unable to hire new teachers or police officers. The consequences may be considerable, and for many constituencies may be far more severe than if restructuring were possible. Although I am not aware of empirical evidence on the cost of lingering financial distress outside of bankruptcy, commentators have speculated that they may be substantial.[13]

Odd as this may sound, doing nothing is likely to be far better than stepping in to protect a particular group of creditors. One of the chief benefits of bankruptcy is that it distributes the sacrifice of financial distress more broadly and equitably than is the case if a city is left to its own devices. Municipal bankruptcy prohibits “unfair discrimination” against any group of general creditors, which ensures that each group will receive roughly comparable treatment.[14]

A second benefit of bankruptcy is that it may alter some of the perverse political incentives that have contributed to many cities’ financial distress. This is especially true with pensions. Although there are a variety of reasons for the current pension crisis, one of the biggest problems in many cities has been the absence of genuine bargaining over the terms of pension promises.  

Public employees obviously would prefer a giant pension to a modest one, but so too would the politicians who ostensibly bargain with them. In some cases, politicians are part of the same pension system; and even if they aren’t, they often depend on the votes of employees who are. If an unsustainably generous pension can be restructured in bankruptcy, this gives employees much more of an interest in making sure that the pension promises are realistic, and that pensions are properly funded. 

With GO bonds, the principal objection to bankruptcy is the risk of contagion. If bonds can be restructured, the reasoning goes, a bankruptcy filing by one city will have devastating effects throughout the municipal bond market. The Detroit bankruptcy suggests that contagion may be an issue in the short run for cities in the same state, but it is unlikely to extend beyond the state borders. After Detroit filed, other Michigan municipalities faced a more unforgiving bond market, but there was very little impact outside of Michigan. This is as one would expect. Detroit’s bankruptcy filing confirmed that the current Michigan administration is reluctant to bail out troubled Michigan cities, information that has relevance throughout Michigan. But it is much less relevant beyond Michigan’s borders. And even within Michigan, the largest effects were very short term, with several municipalities delaying bond offerings or accepting larger than expected interest costs. 

Reporters often speculate about whether a ruling or a case will “set a precedent” for future cases. Make no mistake: Judge Rhodes’ holding that Detroit’s pensions can be restructured, and the likelihood that both pensions and general obligation bonds can be adjusted in the Detroit bankruptcy, are hugely important developments that are being closely watched by struggling municipalities throughout the country. There will be a strong temptation for these constituencies to seek bankruptcy-proof protection from their state governments. This would be unfortunate. The future of municipal governance would be brighter if both sets of obligations can be restructured, at least a little, in bankruptcy.

About the Author

David Arthur Skeel, JD

Professor of Corporate Law, Penn Law School

David Skeel is the S. Samuel Arsht Professor of Corporate Law at the University of Pennsylvania School of Law. He is the author of The New Financial Deal: Understanding the Dodd-Frank Act and Its (Unintended) Consequences (Wiley, 2011), Icarus in the Boardroom (Oxford, 2005) and Debt’s Dominion: A History of Bankruptcy Law in America (Princeton, 2001), as well as numerous articles and other publications. He has been interviewed on The News Hour, Nightline, Chris Matthews’ Hardball (MSNBC), National Public Radio, and Marketplace, among others, and has been quoted in the New York Times, Wall Street Journal, Washington Post and other newspapers and magazines. Skeel has received the Harvey Levin award three times for outstanding teaching, as selected by a vote of the graduating class, the Robert A. Gorman award for excellence in upper level course teaching, and the University’s Lindback Award for distinguished teaching. 


1 I discuss many of the issues in this section more fully in a white paper prepared a few weeks before the Detroit bankruptcy judge’s ruling recent eligibility ruling.  David Skeel, Can Pensions be Restructured in (Detroit’s) Municipal Bankruptcy, available at

2 MICH. Const. ART. XXIV, sec. 9.

3 Indeed, shortly after Detroit filed for bankruptcy, Detroit’s unions persuaded a Michigan state court judge to issue injunctions designed to prevent Michigan state officials from pursuing a bankruptcy case.  The bankruptcy judge held that the state court orders were unavailing, however, because state court actions are halted by the filing of a bankruptcy case. 

4 The Supremacy Clause states that the “Constitution, and the Laws of the United States … shall be the supreme Law of the Land.” U.S. Const, Art. VI, sec. 2.

5 See In re City of Detroit, Michigan, Case No. 13-53846, Opinion Regarding Eligibility, at 80 (Bankr. E.D. Mich. Dec 5, 2013)(noting that the Michigan legislature could have created a property interest, security interest, or state guarantee when lawmakers added pension protection to the Michigan constitution in 1963, but did not).

6 The legislation was codified as part of the 1988 amendments to the bankruptcy laws.  Public Law No. 100-597 (1988).

7 See, e.g, Christian S. Herzeca, Detroit Chapter 9 Pension Impairment and Special Revenues Pledged to Secure General Obligation Bonds, MBS/MONOLINE LITIGATION COMMENTARY FOR SPECULATORS, available at (“T]he insured GOs are paid from the levy of ad valorem taxes levied by Detroit in connection with the issuance of the insured GOs.  … “[As with] revenue bonds, these tax receipts were specifically pledged to repay the insured GOs … and no authority was granted to apply these tax receipts other than towards the repayment of the insured GOs”).

8 I say “theoretically” because I have my doubts. Prior to the recent crisis, some commentators argued that “constructive ambiguity” can be beneficial—even desirable— in some contexts.  But the behavior of Lehman Brothers and AIG during their downward spirals cast considerable doubt on the thesis.

9 Efforts recently have been made to raise hundreds of millions of dollars to protect Detroit’s pensions and art, but even these funds will not ensure full payment of the pensions.

10 R.I. GEN. LAWS § 45-12-1(a) (2011) (“[A]ll general obligation bonds … shall constitute a first lien on … ad valorem taxes and general fund revenues.”) A number of other states also purport to protect bondholders in various ways, but Michigan is one of the states that does not.

11 The most recent revision of the Uniform Commercial Code, which every state has adopted, significantly expanded the kinds of property that can be used as collateral. Interestingly, however, “supergeneric” descriptions of the collateral, such as “all the debtor’s assets” are deemed to be too broad to qualify as creating a lien. U.C.C. section 9-108(c).

12 Under both state law and bankruptcy law, a transfer of the debtor’s property can be reversed if the debtor was insolvent or nearly so at the time of the transaction, and the debtor did not receive “reasonably equivalent value” in connection with the transfer. See, e.g., 11 U.S.C. § 548(a)(1)(B). Although it may sound a little odd to say that the enactment by Rhode Island of a bond protection statute was a “transfer” of the property of the debtor, (Central Falls), the statute purported to set aside Central Falls property for bondholders who would not otherwise have been entitled to it, and fraudulent conveyance law applies even if the debtor did not itself voluntarily transfer the property in question.

13 See, for example, Stephen D. Eide, Quantifying Crowd-Out, CIVIC REPORT, No. 81 (Oct. 2013).

14 11 U.S.C. sec. 901(a)(incorporating 1129(b)(1), which prohibits “unfair discrimination”).