Exceptional Proff 612

600

FAILING CITIES AND THE RED QUEEN PHENOMENON

SAMIR D. PARIKH* ZHAOCHEN HE**

Abstract: Cities and counties are failing. Unfunded liabilities for retirees’ health- care benefits aggregate to more than $1 trillion. Pension systems are underfunded by as much as $4.4 trillion. Many local government capital structures ensure ris- ing costs and declining revenues, the precursors to service-delivery insolvency. These governments are experiencing the Red Queen phenomenon. They have tried a dizzying number of remedies, but their dire situation persists unchanged. State legislatures have failed to respond. More specifically, many states have re- fused to implement meaningful debt restructuring mechanisms for local govern- ments. They argue that giving cities and counties the power to potentially impair bond obligations will lead to a doomsday scenario: credit markets will respond by dramatically raising interest rates on new municipal and state bond issuances. This argument—which we term the “paralysis justification”—has been employed widely to support state inaction. The paralysis justification, however, is anecdotal and untested. This Article attempts to fill a significant gap in the literature by re- porting the results of an unprecedented empirical study. Our study aggregated da- ta for every fixed-rate general obligation, municipal bond issued in the United States from January 1, 2004 to December 31, 2014. It included over eight hun- dred thousand issuances in total. By employing multivariate regression analysis, we conclude that the paralysis justification is a false narrative. Municipalities lo- cated in states that offer meaningful debt restructuring options enjoy the lowest borrowing costs, all other things equal. This Article removes one of the largest obstacles to financial relief for many cities and counties. We hope to encourage recalcitrant state legislatures to enact the structural changes their local govern- ments need desperately.

© 2017, Samir D. Parikh & Zhaochen He. All rights reserved. * Associate Professor of Law, Lewis & Clark Law School. ** Assistant Professor of Economics at Christopher Newport University. We thank Douglas Baird, Jay Westbrook, JJ Prescott, Kellen Zale, Rafael Pardo Charles Tabb, Charles Silver, David Skeel, Colin Marks, Brian Blum, Katherine Porter, and Jim Grant for their input and guidance. We thank Oren Haker for helping formulate our research design. We would like to acknowledge Brian Adams and the University of Portland for granting us access to the Bloomberg Terminal. We thank Peter Farrell, Josh Goldberg, and Aaron Johnson for their research assistance. Finally, we would like to acknowledge our families for their unwavering support.

2017] Failing Cities 601

INTRODUCTION

The atomic bomb had its genesis in the city of Chicago. In 1942, the ex- periment producing the first controlled, self-sustaining nuclear chain reaction occurred at Stagg Field.1 The primitive reactor was too weak to power even a single light bulb but proved that the power of the atom could be unleashed and controlled.2 This nascent event was emblematic of the type of innovation for which the city was known during the early twentieth century. Chicago was “the pace–maker of the world.”3 The city’s trajectory since these halcyon days, however, has been disappointing.

In May of 2015, Moody’s Investor’s Service dropped Chicago’s bond rat- ing to junk status, the only major city in the United States to enjoy this igno- miny.4 The downgrade was a result of the city’s almost $9 billion in bond debt coupled with the fact that its four primary pension plans had a combined un- derfunding of nearly $27 billion.5 The aggregate, unfunded public-worker pen- sion liability per Chicago resident is almost $20,000.6 As a point of compari- son, the same liability per Detroit resident shortly before that city filed for bankruptcy was less than $5000.7 Without some sort of radical intervention, the city has no chance of servicing this debt. Chicago is home to a financial atomic bomb, and it is not alone.

So how does one dismantle an atomic bomb? Well, as the proverb in- structs: do not build it in the first place. Unfortunately, path dependence is ush- 1 See MICHAEL F. L’ANNUNZIATA, RADIOACTIVITY: INTRODUCTION AND HISTORY, FROM THE QUANTUM TO QUARKS 151 (2d ed. 2016). 2 See id. Some of the scientists who conducted this revolutionary experiment would go on to join the Manhattan Project and develop the atomic bomb. Id. 3 Newton Dent, The Romance of Chicago, MUNSEY’S MAG., Apr. 1907, at 2, 20. 4 See Moody’s Downgrades Chicago, IL to Ba1, Affecting $8.9B of GO, Sales, and Motor Fuel Tax Debt; Outlook Negative, MOODY’S (May 12, 2015), https://www.moodys.com/research/Moodys- downgrades-Chicago-IL-to-Ba1-affecting-89B-of-GO--PR_325213 [https://perma.cc/Q9SM-ZW3P] [hereinafter Moody’s Downgrades Chicago]. Further, in March 2016, Fitch Ratings downgraded Chi- cago’s bond rating to ostensibly junk status. Elizabeth Campbell, Chicago’s Rating Cut by Fitch After Pension Overhaul Dashed, BLOOMBERG (Mar. 28, 2016), http://www.bloomberg.com/news/ articles/2016-03-28/chicago-s-bond-rating-lowered-to-bbb-by-fitch-after-court-loss [https://perma. cc/T3AL-XH5F]. 5 See Jessica Corso, Chicago’s Fiscal Future Bleak but Not Hopeless, Experts Say, LAW360 (Oct. 26, 2015), http://www.law360.com/articles/7190099 [https://perma.cc/H6BV-MBRZ]; Moody’s Downgrades Chicago, supra note 4; see also KRISTEN DEJONG, CHICAGO’S FISCAL STRESS: NEW TERM, SAME PROBLEMS 1–4 (2015). Note that the underfunding is most likely far greater than $20 billion due to the high discount rate that the city has used in calculating its future liabilities. See Rob- ert Novy-Marx & Joshua D. Rauh, The Liabilities and Risks of State-Sponsored Pension Plans, 23 J. ECON. PERSP. 191, 191–92 (2009). 6 Rahmbo’s Toughest Mission: Can Rahm Emanuel Save Chicago from Financial Calamity?, THE ECONOMIST (Jun. 14, 2014), http://www.economist.com/news/united-states/21604165-can-rahm- emanuel-save-chicago-financial-calamity-rahmbos-toughest-mission [https://perma.cc/79EL-QXCE]. 7 See id.

602 Boston College Law Review [Vol. 58:599

ering municipalities8 towards financial Armageddon.9 States could provide municipalities the means to address their problems and potentially avoid ca- lamity, but an unfathomable paralysis has set in. States have been disengaged from the municipal restructuring process.10 Thirty-nine states fail to offer a meaningful restructuring process for their distressed municipalities.11 One of the primary excuses for this high level of disengagement is the fear that bor- rowing costs on municipal bond issuances will increase materially if a state begins offering debt adjustment options to its cities and counties.12

State law has a profound impact on a municipal borrower’s interaction with lenders and investors. State law is essentially a product that affects a municipali-

8 For purposes of this Article, the terms “municipality” and “municipalities” describe counties, cities, and certain other local governments that enjoy taxing authority. 9 In making this statement, this Article considers the financial condition of subnational govern- ments based predominately on a municipality’s ability to fulfill its “financial obligations to creditors, consumers, employees, taxpayers, suppliers, constituents, and others as they become due” and to satis- fy its service-delivery obligations to its current and future residents. See Martin Ives, Financial Man- agement Implications of “Service Delivery Insolvency,” MUN. FIN. J., Summer 2015, at 45, 45–47 (quoting Robert Berne from the Governmental Accounting Standards Board). 10 See Samir D. Parikh, A New Fulcrum Point for City Survival, 57 WM. & MARY L. REV. 221, 238 (2015); see also James K. Conant et al., State Budgeting in the Aftermath of the Great Recession: A Comparative Perspective, MUN. FIN. J., Summer 2012, at 1, 31–33 (concluding after a study of six states during the post-Great Recession period that none of the states took meaningful action to address municipal turmoil); Gregory Lipitz, Stephen Fehr, Thomas Neff & William Kannel, State Oversight, Bankruptcies, and Recovery, MUN. FIN. J., Winter 2015, at 55, 55–57 (2015) (concluding that only a few states are “aggressively involved” in their municipalities’ fiscal health). 11 See infra notes 90–109 and accompanying text (creating a taxonomy for all fifty states and the District of Columbia based off of the level of support and the options they provide to their municipali- ties that are under financial stress). The thirty-nine states that fail to provide a meaningful restructur- ing process for their municipalities are Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, Georgia, Hawaii, Idaho, Illinois, Iowa, Indiana, Kansas, Louisiana, Maryland, Massachusetts, Minnesota, Mississippi, Missouri, Montana, Nebraska, New Hampshire, New Mexico, North Dakota, Oklahoma, Oregon, South Carolina, South Dakota, Tennessee, Texas, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, and Wyoming. 12 See Wenli Yan, Revenue Diversification and State Credit Risk, MUN. FIN. J., Winter 2011, at 41, 45–48 (2011); Martin Bricketto, Atlantic City Bankruptcy Would Hurt Other Towns, Mayor Says, LAW360 (May 15, 2015), http://www.law360.com/articles/656664 [https://perma.cc/5WZC-L4J3]; Peter Coy, The Case for Allowing U.S. States to Declare Bankruptcy, BLOOMBERG (Jan. 25, 2016), https://www.bloomberg.com/news/articles/2016-01-21/the-case-for-allowing-u-s-states-to-declare- bankruptcy [https://perma.cc/AV3U-59PP]; Kelly Nolan, Detroit Bankruptcy Plan Called Unfair to Bondholders, WALL STREET J. (Jul. 18, 2013), https://www.wsj.com/articles/SB1000142412788732 4448104578614501316630488 [https://perma.cc/2A2A-S9U4]; Mary Williams Walsh, Woes of De- troit Hurt Borrowing by Its Neighbors, N.Y. TIMES (Aug. 8, 2013), https://dealbook.nytimes.com/ 2013/08/08/detroit-blocks-other-cities-from-bond-market/?_r=0 [https://perma.cc/M8HC-2QG5]; State of Pay: What Do the Woes of Detroit Mean for Muni Bonds?, THE ECONOMIST (Jun 22, 2013), http://www.economist.com/news/finance-and-economics/21579861-what-do-woes-detroit-mean- muni-bonds-state-pay [https://perma.cc/A9QD-T4F5] [hereinafter State of Pay].

2017] Failing Cities 603

ty’s risk profile, credit rating, and borrowing costs.13 Indeed, credit markets have historically punished debt issuances by municipalities subject to state laws that fail to protect bondholders in a manner commensurate with market expecta- tions.14 Subnational governments are dependent on borrowing to fill budgetary gaps and fund capital projects.15 Consequently, they seek to avoid legislation that they believe will diminish their credit rating or displease credit markets.16 Many states have refused to implement meaningful debt adjustment options for local governments because these options could provide a municipality the ability to impair17 bond obligations.18 State legislators and policymakers are concerned that increased impairment risk would prompt municipal credit markets to de- mand above-market interest rates on new issuances.19 Increased borrowing costs could cripple municipalities and negatively affect states.20 This argument— which we refer to as the “paralysis justification”—has been advanced to stifle the implementation of meaningful debt adjustment mechanisms, including the one recently proposed by Professor Samir D. Parikh.21

The paralysis justification, however, is purely anecdotal. State legislators and policymakers seem content to be the wardens of municipal demise based

13 See generally Roberta Romano, Law as a Product: Some Pieces of the Incorporation Puzzle, 1 J.L. ECON & ORG. 225 (1985) (outlining the theory that state laws can effectively be viewed as prod- ucts in certain situations and examining this theory in relation to state incorporation laws). 14 This punishment invariably comes in the form of a demand for an above-market interest rate on issuances. 15 See Clayton P. Gillette, Fiscal Federalism, Political Will, and Strategic Use of Municipal Bankruptcy, 79 U. CHI. L. REV. 281, 312 (2012). 16 PEW CHARITABLE TRS., THE STATE ROLE IN LOCAL GOVERNMENT FINANCIAL DISTRESS 12 (2013) [hereinafter STATE ROLE IN LOCAL GOVERNMENT]; Gillette, supra note 15, at 291–92, 305 (“[C]entralized governments that intervene in the face of municipal fiscal distress are motivated large- ly by a perception of contagion risk.”); Steven L. Schwarcz, A Minimalist Approach to State “Bank- ruptcy,” 59 UCLA L. REV. 322, 333 (2011). 17 The impairment would typically come in the form of deferred interest payments or an interest rate adjustment. 18 See Yan, supra note 12, at 45–48; Bricketto, supra note 12; Coy, supra note 12; Nolan, supra note 12; Walsh, supra note 12; State of Pay, supra note 12. 19 See Coy, supra note 12 (“State governments said they didn’t want to be eligible for bankruptcy, fearing that the very possibility would spook investors in municipal bonds and drive up their borrow- ing costs.”); see also Yan, supra note 12, at 45–48; Bricketto, supra note 12; Nolan, supra note 12; Walsh, supra note 12; State of Pay, supra note 12. But see David A. Skeel, Jr., Is Bankruptcy the Answer for Troubled Cities and States?, 50 HOUS. L. REV. 1063, 1069 (2013) (concluding that the states that haven chosen to enact municipal bankruptcy laws have not subsequently experienced crip- pling increases in bond costs); David A. Skeel, Jr., States of Bankruptcy, 79 U. CHI. L. REV. 677, 717– 22 (2012). 20 See Parikh, supra note 10, at 258. States have historically acted as implicit guarantors of mu- nicipal debts and service delivery. See id. at 259. 21 See id. at 221 (proposing a comprehensive, fiscal monitoring system that identifies and then directs distressed municipalities into a dynamic negotiation model supported by the option to file for federal bankruptcy).

604 Boston College Law Review [Vol. 58:599

on an untested argument. Indeed, as far as the authors of this Article are aware, no scholarly study has examined, in a statistically meaningful manner, how a municipality’s borrowing costs are affected by its state and federal debt re- structuring options.

This Article addresses this significant gap in municipal literature by re- porting the results of an unprecedented empirical study designed to test the paralysis justification. The study aggregated data for every fixed-rate general obligation, municipal bond issued in the United States from January 1, 2004 to December 31, 2014 (the “Observation Period”).22 It does not rely on a sample of bond issuances or survey results to estimate key features of our dataset. Ra- ther, the study meticulously reviews the entire universe of actual issuances dur- ing the Observation Period, over eight hundred thousand issuances in total. Overall, the study’s breadth and depth are unparalleled.

The data show that there is no empirical basis for the paralysis justifica- tion. Other things being equal, a municipality’s borrowing costs do not in- crease based on the fact that its home state provides meaningful debt restruc- turing options. In fact, after controlling for eighteen independent variables— including economic conditions and political affiliations—and drawing data from myriad sources, we conclude that that municipalities with the lowest bor- rowing costs are those located in states that offer meaningful out-of-court debt adjustment options. Our results establish that the paralysis justification is a false narrative and challenge conventional wisdom in this area.23 Further, by extrapolating the data, we posit that municipalities located in states that adopt Professor Parikh’s proposed debt adjustment mechanism would have the low- est borrowing costs of any group.24

In delineating these exciting results, this Article proceeds as follows: Part I addresses the challenges facing subnational governments and explores the contours of the Red Queen phenomenon25 and the paralysis justification.26

22 See infra notes 90–126 and accompanying text (performing an empirical analysis of these bond issuances and how differences in state bankruptcy law affected these issuances). 23 See infra notes 90–126 and accompanying text. 24 See infra notes 117–125 and accompanying text (extrapolating the results of this empirical analysis to evaluate a hypothetical municipality that chose to adopt Professor Parikh’s proposal). Keep in mind that a local government financial manager’s primary goal in debt issuance is to minimize the cost of capital. See Justin Marlowe, GASB 34’s Information Relevance: Evidence from New Issue Local Government Debt 6 (Apr. 14, 2010) (unpublished manuscript), http://ssrn.com/ abstract=1589343 [https://perma.cc/FW43-DFKN]. As one scholar has pointed out, because local government debt issuances are invariably long-term fixed coupon securities, “a few basis point differ- ence in interest costs on those securities can have multi-million dollar implications for overall cost of capital.” Id. 25 Definition of Red Queen, FIN. TIMES LEXICON, http://lexicon.ft.com/Term?term=Red%20 Queen [https://perma.cc/E69C-BZDM]. The Red Queen phenomenon is named after the eponymous

2017] Failing Cities 605

Municipal distress is a well-known phenomenon, but the overlooked facet is the reluctance of state officials and policymakers to address the problem. This recalcitrance leaves many municipalities with few options to avoid service- delivery insolvency. Part II presents our empirical study and methodology.27 Part II also details our dataset’s composition and high quality and explains the regression analysis used as well as the chosen dependent and independent vari- ables. Conventional wisdom posits that municipalities located in states that do not offer meaningful debt adjustment options and restrict access to Chapter 9—a group we refer to as no-option states—should enjoy the lowest borrowing costs.28 The rationale for this argument is that noteholders would not want their municipal borrowers to have any means to modify debt payments. However, we observed that municipalities in no-option states had some of the highest borrowing costs. Part III explains how the phenomenon we observed can be reconciled with accepted financial restructuring principles.29

Part IV explores the dramatic implications of this Article’s findings.30 Many states have been content to allow their municipalities to deteriorate based on a false narrative. Our findings transform the debate in this area by demonstrating that key constituencies benefit when a municipality is afforded meaningful debt restructuring options. Scholars and policymakers have disa- greed about whether the optimal approach for distressed municipalities can be found at the state or federal level, or if inaction is the best course. Our results unify the diverse arguments by demonstrating empirically that states should offer their municipalities meaningful debt restructuring options at the state and federal level.31

Ultimately, our study reveals that credit markets do not punish municipal- ities that possess meaningful debt restructuring options under state or federal laws. In fact, the opposite is true. We believe that the reason for this is that an orderly process for debt adjustment extracts proportional concessions from both employee unions and noteholders. This shared-pain dynamic increases the character from Lewis Carol’s Through the Looking-Glass, who tells Alice, “[I]t takes all the running you can do, to keep in the same place.” Id. 26 See infra notes 81–89 and accompanying text. 27 See infra notes 90–126 and accompanying text. 28 The borrowing cost assessments found in this article are generally subject to the “all other things being equal” qualifier. 29 See infra notes 127–147 and accompanying text. 30 See infra notes 148–163 and accompanying text. 31 At the state law level, states should monitor their municipalities and offer a debt adjustment mechanism that facilitates negotiation with creditor constituencies and meaningful debt restructuring. See Parikh, supra note 10, at 284–94. At the federal law level, states should authorize their municipal- ities to seek relief under Chapter 9 of the Federal Bankruptcy Code in the event these negotiations fail. See id. at 257–58. The Federal Bankruptcy Code requires state authorization for a municipality to file. 11 U.S.C. § 109(c)(2) (2012).

606 Boston College Law Review [Vol. 58:599

odds of sustainable viability for the municipality and is instrumental in avoid- ing a wholesale payment default. A delineated mechanism engenders certainty, which reduces the risk of holdouts and free riding. Not surprisingly, creditors have responded well to the implementation of debt restructuring processes in other contexts.32 Many state officials have overlooked this fact. State inaction based on unfounded fears will continue to decimate the interests of all local government constituencies, including employees, noteholders, residents, and even the home state itself. At a macro level, systemic municipal failure has ripple effects, imposing significant economic costs on state and national econ- omies.33

This Article seeks to remove one of the largest obstacles to implementa- tion of meaningful municipal debt restructuring options. Our findings support the conclusion that reluctant states should begin the process of enacting legis- lation that will allow municipalities to restructure their obligations and address their fiscal deficiencies at an earlier stage of deterioration; in essence, disman- tling a financial atomic bomb before it is built.

I. THE DESOLATE MUNICIPAL LANDSCAPE

A. An Overview of Decline

The national economy is recovering from the Great Recession, but subna- tional governments are unable to participate in this rebirth. 34 In fact, the dark-

32 See Michael Bradley, James D. Cox & Mitu Gulati, The Market Reaction to Legal Shocks and Their Antidotes: Lessons from the Sovereign Debt Market, 39 J. LEGAL STUD. 289, 295–98 (2010) (finding that the addition of collective action clauses—which enable supermajority voting to change payment terms—in sovereign bond indentures did not increase sovereign borrowing costs); Anne O. Krueger, First Deputy Managing Director, Int’l Monetary Fund, Sovereign Debt Restructuring Mech- anism—One Year Later (Dec. 10, 2002) (explaining that the creation of a sovereign debt restructuring mechanism could in fact reduce country borrowing costs). 33 See Robert D. Ebel, John E. Petersen & Ha T.T. Vu, The Great Recession: Impacts and Out- look for U.S. State and Local Finance, MUN. FIN. J., Spring 2013, at 33, 35 (“There is a high national interest in what happens to the state and local sector. The 50 states and their nearly 90,000 local gov- ernments represent 11.9% of the national product. That is one-and-a-half times the size of the federal sector with national defense spending included, and four-and-a-half times the federal government when one considers only federal non-defense spending. Furthermore, taken together, the states and localities employ about one out of every eight workers and provide the bulk of all basic governmental services consumed by individuals and firms, such as education, public safety, and public works. State and local governments also act as agents in the delivery of many federal services, including the pass- ing on of many federal transfer payments from places to people.”); see also STATE BUDGET CRISIS TASK FORCE, REPORT OF THE STATE BUDGET CRISIS TASK FORCE 2 (2012). 34 Municipal deterioration has been thoroughly chronicled, and a variety of articles ably handle the articulation of this phenomenon. See, e.g., Michelle Wilde Anderson, The New Minimal Cities, 123 YALE L.J. 1118, 1130 (2014); Christine Sgarlata Chung, Zombieland / The Detroit Bankruptcy: Why Debts Associated with Pensions, Benefits, and Municipal Securities Never Die . . . and How They Are Killing Cities Like Detroit, 41 FORDHAM URB. L.J. 771, 778 (2014); Parikh, supra note 10, at

2017] Failing Cities 607

est days lay ahead. A critical mass of municipalities is suffering from capital structures that ensure rising costs and declining revenues, undermining basic service delivery.35

Costs are fueled by a torrent of employee salaries, wages, pensions, health care, and other related expenses.36 Municipalities have attempted to curtail sal- aries and other benefits, but savings have been nominal.37 Further, mandatory pension contributions and health care costs are the most significant debt driv- ers, and these costs will continue to rise for the foreseeable future.38 In fact, the amounts that municipalities are currently setting aside to satisfy these long- term obligations are woefully insufficient. Unfunded liabilities for state and municipal retirees’ healthcare benefits exceed $1 trillion.39 Depending on the discount rate, pension systems are underfunded by as much as $4.4 trillion.40

238. A full exploration of this issue is beyond this Article’s scope but a general overview is instruc- tive. 35 See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 2–4 (summarizing the myriad of problems that many municipalities have in their financial capital structures); Ives, supra note 9, at 46 (describing how the financial degeneration of municipalities undermines basic service delivery). 36 See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 2–4. States have also contributed to this cost torrent by imposing unfunded mandates. For example, a state will pass a law requiring local governments to install signs that provide that cellphone use while driving is prohibited in school zones but fail to provide the funding for the signs. See J.J. Smith, New Texas Law Prohibits Drivers from Using Cell Phones While Driving in School Zones, ROCKWALL NEWS (Tex.) (Aug. 26, 2013), http:// myrockwallnews.com/new-texas-law-prohibits-drivers-from-using-cell-phones-while-driving-in- school-zones/ [https://perma.cc/56FV-ME94] (noting that the responsibility for posting state- mandated traffic signs fell on local governments in Texas). 37 See Parikh, supra note 10, at 235. Municipalities spend more than 35% of their budget on sala- ries and wages. See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 12. Collective bargaining agreements severely limit adjustments to employee headcount and benefits. See Parikh, supra note 10, at 235. Municipalities invariably wind up making minor reductions in headcount that produce minimal cost savings. See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 36. Hiring freezes are most frequently invoked but similarly futile. See Michael A. Pagano & Christiana McFarland, City Fiscal Conditions in 2013, MO. MUN. REV., Nov. 2013, at 12, 13–14. Through a combination of layoffs, attribution, hiring freezes, and furloughs, municipalities are able to reduce their workforce. These cuts, however, offer only marginal relief, and have accounted for only a 3.4% reduction in workforce be- tween September 2008 and December 2011. See LIZ GROSS ET AL., PEW CHARITABLE TRS., THE LOCAL SQUEEZE: FALLING REVENUES AND GROWING DEMAND FOR SERVICES CHALLENGE CITIES, COUNTIES, AND SCHOOL DISTRICTS 13 (2012). Collective bargaining agreements also restrict attempts to reduce health care and pension benefits. 38 See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 14 (illustrating graphically how the rapid growth of health care costs poses the greatest threat to state and municipal budgets). The report from the State Budget Crisis Task Force also details how pension costs for state and local govern- ments are underfunded by almost $3 trillion. See id. at 2. 39 See id. at 3. 40 See Stuart Buck, The Legal Ramifications of Public Pension Reform, 17 TEX. REV. L. & POL. 25, 27 (2012); Jean Burson et al., Do Public Pension Obligations Affect State Funding Costs?, MUN. FIN. J., Summer 2014, at 17, 17; see also Robert Novy-Marx & Joshua Rauh, The Crisis in Local Government Pensions in the United States, in GROWING OLD: PAYING FOR RETIREMENT AND INSTI- TUTIONAL MONEY MANAGEMENT AFTER THE FINANCIAL CRISIS 47, 48 (Yasuyuki Fuchita et al. eds.,

608 Boston College Law Review [Vol. 58:599

For example, Chicago’s retiree pension plans are only 39% funded. The state and its municipalities face a $165 billion liability, but have funded less than $65 billion.41 Additionally, the underfunding is swelling. The funded ratio has declined every year since 2011.42

Debt service further undermines this situation. Municipalities have over- grazed at the debt commons, relying on bond debt to fund choices and fill budgetary gaps.43 States have attempted to limit their municipalities’ debt load, but municipalities have successfully circumvented these obstacles.44 Large municipalities have embraced a model dependent on borrowing. Approximate- ly 44,000 subnational governments issue debt.45 The municipal bond market approaches four trillion dollars in principal and is comprised of over one mil- lion different municipal bonds.46 As a point of comparison, there are fewer than fifty thousand different corporate bonds.47 Debt service is a fixed cost that can undermine optimal resource allocation.

Exacerbating this indefinite rising-cost phenomenon is the fact that mu- nicipalities have been subject to a perfect confluence of declining revenues. Intergovernmental aid has evaporated.48 The loss of intergovernmental aid has 2011). The PEW Charitable Trusts has found that there is a $968 billion funding gap across all state governments. This determination, however, relies on state reports, which use an entirely unrealistic expected rate of return for discounting purposes. See PEW CHARITABLE TRS., THE STATE PENSIONS FUNDING GAP: CHALLENGES PERSIST 1 (2015) [hereinafter FUNDING GAP]. Finance scholars agree that the underfunding is much larger than reported by state agencies. See Joshua Rauh, Professor of Fin., Stanford Univ. Graduate Sch. of Bus. & Senior Fellow, Hoover Inst., Stanford Univ., Financial Economics for Public Policy, Presentation at the George Mason University School of Law Law & Economics Center Workshop for Law Professors on the Economics of Public Pension Reform (Sept. 17, 2015). 41 See FUNDING GAP, supra note 40, at 3. 42 See id. 43 See Gillette, supra note 15, at 287–88. 44 See, e.g., Ives, supra note 9, at 52 (“Although virtually all municipalities are required to pre- pare balanced budgets, many municipalities prepare budgets that are balanced in form, but not in sub- stance.”); see also James E. Spiotto, The Role of the State in Supervising and Assisting Municipalities, Especially in Times of Financial Distress, MUN. FIN. J., Spring 2013, at 1, 8. 45 See Amicus Curiae Brief by the Sec. Indus. & Fin. Mkts Ass’n at 6, In re City of Detroit, No. 13-53846 (Bankr. E.D. Mich. May 12, 2014). 46 See id. at 7. 47 See id. For example, Chicago has over $9 billion in bond debt. See Moody’s Downgrades Chi- cago, supra note 4. 48 Intergovernmental aid includes grants, transfers, and other funds a municipality receives from federal, state, county, or other local governments through ongoing revenue-sharing agreements and one-time infusions. See PEW CHARITABLE TRS., AMERICA’S BIG CITIES IN VOLATILE TIMES 10 (2013) [hereinafter AMERICA’S BIG CITIES]. The American Recovery and Reinvestment Act (“AR- RA”) was signed in 2009 as a short-term stimulus bill seeking to infuse $787 billion into the econo- my. A significant portion of these funds represented direct aid to states, funds that often times went to municipalities. See VICE PRESIDENT JOSEPH BIDEN, ANNUAL REPORT TO THE PRESIDENT ON PRO- GRESS IMPLEMENTING THE AMERICAN RECOVERY AND REINVESTMENT ACT OF 2009, at 9 (2010). In fact, states have historically funded on average close to one-third of local budgets. See LIZ GROSS ET

2017] Failing Cities 609

been compounded by declining tax collections.49 Primarily, property tax reve- nue has historically been a stalwart for municipalities during economic down- turns.50 In previous downturns, residential home prices were stable.51 This sta- bility—coupled with the fact that approximately 97% of property taxes go to local governments—made property tax revenue vital for municipal rehabilita- tion.52 An imploding housing market, however, precipitated the Great Reces- sion.53 The unprecedented fall in home prices decimated county coffers.54 Be- tween 2007 and 2011 home prices fell almost 20% nationally.55 Even after re- cent appreciation, home prices and attendant property tax revenue are still well below 2007 levels.56 For example, housing prices in Las Vegas are still 40% below their peak.57 This shift in the housing market continues to suppress local revenues.58

Municipalities have employed a staggering variety of traditional and non- traditional means to achieve fiscal stability, but most have fallen victim to the Red Queen phenomenon.59 Distressed municipalities have shifted revenue,60 swept fund balances,61 and raided rainy day funds to simply maintain their cur- AL., supra note 37, at 5–6 (“Many states provide grants [to localities] for general operations; in other cases, money is set aside for certain uses, such as road repair. States also sometimes share a portion of tax revenues with cities, counties, and school districts based on factors like population, need, and the community’s existing tax burden.”). Funding through the ARRA helped stabilize localities for a brief period of time, but the ARRA and other measures have merely served to delay the day of reckoning. By 2010, state aid to municipalities decreased by $12.6 billion from the previous year and decreased again in 2011, 2012, and 2013. See id. at 4, 7. 49 See GROSS ET AL., supra note 37, at 4 (highlighting graphically the drop in tax revenue as a result of the Great Recession). 50 See id. at 1. 51 See id. 52 See Ebel, Peterson & Vu, supra note 33, at 45. 53 See id. at 46. 54 See GROSS ET AL., supra note 37, at 4. 55 See id. 9–10. Home prices in metropolitan areas fell 24.7% from 2007 to 2010. See Ebel, Pe- tersen & Vu, supra note 33, at 47 n.21. 56 See Daily Chart: American Housing Prices, THE ECONOMIST (Aug. 24, 2016), http://www. economist.com/blogs/graphicdetail/2016/08/daily-chart-20 [https://perma.cc/67HQ-L27W]. 57 See id. 58 See Ebel, Petersen & Vu, supra note 33, at 47. General sales tax and individual income tax revenue also declined significantly during the Great Recession, further eroding municipal finances. See id. at 47–53. 59 See supra note 25 (providing background for the “Red Queen phenomenon” as a term used to describe situations where a great deal of effort is spent to essentially maintain the same position). 60 See James K. Conant et al., supra note 10, at 6 (defining revenue shifting as a method by which municipalities “accelerate the collection of tax revenue or fees by moving revenue ‘backward’ into the current fiscal year”). An obvious downside of revenue shifting is that any revenue “shifted” is re- moved from the accounting of the following fiscal year. See id. 61 See id. “Sweeping fund balances” describes the method by which municipalities transfer cash balances “from dedicated (or earmarked) funds or reserves into the general fund for the current fiscal year.” Id. As some scholars have pointed out, “[a] cost of [sweeping fund balances] is that the capacity

610 Boston College Law Review [Vol. 58:599

rent downward trajectory.62 Meaningful increases in tax revenue have been elusive because forty-six states severely limit their municipalities’ ability to increase taxes.63 Even without this obstacle, municipalities face significant political deterrence.64 Further, raising taxes does not necessarily lead to in- creased revenue. In many cases, tax increases accelerate migration trends and can decimate a municipality’s tax base.65

Truly desperate municipalities have attempted to sell government assets and privatize functions, including parking enforcement, park maintenance, graffiti removal, collection of delinquent taxes, and operation of public ven- ues.66 Unfortunately, all of these stabilization methods are characterized by short-term cash infusions that produce disproportionate future expenses or lost future revenue.67 For example, in late 2010, Newark opted to sell and then lease back sixteen of the city’s buildings, including its police and fire head- quarters and symphony hall.68 The sale yielded $74 million for the city, but leasing the buildings back will cost the city $125 million during the lease term.69 Similarly, the city of Chicago leased its parking meter system to a con-

to deliver products or services supported by such funds will be diminished in the upcoming fiscal year or years.” Id. 62 See id. at 6–7. See generally Carolyn Boudreaux & W. Bartley Hildreth, Pullback Manage- ment: State Budgeting Under Fiscal Stress, in THE OXFORD HANDBOOK OF STATE AND LOCAL GOV- ERNMENT FINANCE 816 (Robert D. Ebel & John E. Petersen eds., 2012) (analyzing how states’ budg- eting decisions are impacted by periods of significant declines in revenue). For example, Sacramento tapped cash reserves beginning in 2007. See AMERICA’S BIG CITIES, supra note 48, at 16. At that time, the reserve balance was 31% of general revenue. By 2011, the reserve was down to just 6% of general revenue. See id. Sacramento, as well as the vast majority of other distressed municipalities, cannot rely on its cash reserve to address future fiscal challenges. 63 LIZ GROSS ET AL., supra note 37, at 11. Tax and expenditure limitations (“TELs”) are the pri- mary means for limiting local financial autonomy. Critics have argued that TELs compounded the budgetary dystopia triggered by the Great Recession. See generally James M. Poterba & Kim Rueben, State Fiscal Institutions and the U.S. Municipal Bond Market, in FISCAL INSTITUTIONS AND FISCAL PERFORMANCE 181 (James M. Poterba & Jürgen von Hagen eds., 1999) (focusing on the effects of fiscal policy on budget deficits within subnational governments, specifically the states within the United States). 64 See PAGANO & MCFARLAND, supra note 37, at 6–7. Elected officials are prone to eschew tax increases in favor of less controversial revenue-generating measures, such as raising fees that are applied to city services. See id. These fee increases, however, often fail to generate significant funds. See Parikh, supra note 10, at 235. 65 See Ebel, Petersen & Vu, supra note 33, at 52–53; Sally Wallace, The Evolving Financial Ar- chitecture of State and Local Governments, in THE OXFORD HANDBOOK OF STATE AND LOCAL GOV- ERNMENT FINANCE 156–75 (Robert D. Ebel & John E. Petersen eds., 2012). 66 Ianthe Jeanne Dugan, Facing Budget Gaps, Cities Sell Parking, Airports, Zoo, WALL STREET J. (Aug. 23, 2010), http:///www.wsj.com/articles/SB10001424052748703960004575427150960867176 [https://perma.cc/GVA3-ECCU]. 67 See Parikh, supra note 10, at 235–36. 68 See Anderson, supra note 34, at 1167–68. 69 See id.

2017] Failing Cities 611

sortium led by Morgan Stanley in order to balance its budget.70 Chicago will ultimately receive $1.16 billion from its parking meter lease, but the consorti- um is now expected to make more than ten times that amount over the course of the deal.71 Further, these one-time sales temporarily fill budgetary gaps but fail to produce structural reform that improves the municipality’s viability.

Municipalities are desperate for meaningful debt restructuring options, but many states have blindly embraced the paralysis justification.

B. The Paralysis Justification

The acceleration of municipal demise should have prompted state legisla- tors and policymakers to undertake aggressive structural changes. A crippling paralysis, however, has predominated. Thirty-nine states fail to offer meaning- ful debt restructuring options to their municipalities.72 Federal bankruptcy is the only option available to municipalities in many of these states, but federal bankruptcy is a deeply flawed process73 that is designed to provide a venue for nearly terminal municipalities, like the city of Detroit in 2013. Municipalities that seek to address their eroding capital structures at an earlier stage of deteri- oration have few options. In light of the turmoil afflicting municipalities, the failure by state legislators and policymakers to offer meaningful debt adjust- ment options is bizarre. The justification for this failure is no less bizarre.

As noted in this Article, subnational governments are dependent on the credit markets to fund choices and fill budgetary gaps.74 Borrowing costs— and, more specifically, interest rates—fluctuate based on a borrower’s credit rating.75 Consequently, states and municipalities are enormously concerned with their credit rating.76 Even minor fluctuations in borrowing costs can be debilitating. Subnational government debt issuances are invariably long-term fixed coupon securities. Therefore, “a few basis point77 difference in interest

70 See Darrell Preston, Morgan Stanley Group’s $11 Billion Makes Chicago Taxpayers Cry, BLOOMBERG (Aug 8, 2010), http://www.bloomberg.com/news/2010-08-09/morgan-stanley-group-s- 11-billion-from-chicago-meters-makes-taxpayers-cry.html [https://perma.cc/P6ZZ-FWH2]. 71 See id. 72 See infra notes 101–102 and accompanying text (creating a taxonomy for all fifty states and the District of Columbia based off of the level of financial support the state provides to its municipalities). 73 See Parikh, supra note 10, at 242–48 (explaining the systemic deficiencies of Chapter 9 of the Bankruptcy Code). 74 See Gillette, supra note 15, at 287–88. 75 See, e.g., Yan, supra note 12, at 45–48. 76 See STATE ROLE IN LOCAL GOVERNMENT, supra note 16, at 12; Gillette, supra note 15, at 291–92 (“[C]entralized governments that intervene in the face of municipal fiscal distress are motivat- ed largely by a perception of contagion risk.”); Schwarcz, supra note 16, at 333. 77 A basis point is a common unit of measure for interest rates. One hundred basis points equates to one percent.

612 Boston College Law Review [Vol. 58:599

costs on those securities can have multi-million dollar implications for overall cost of capital.”78

Subnational governments must operate against this backdrop and the un- derstanding that their laws and legal structures affect their risk profile. State and municipal action can have a profound impact on how lenders and investors perceive a municipal borrower. State law is essentially a product that affects a municipality’s risk profile, credit rating, and borrowing costs.79 Municipalities will often enjoy reduced borrowing costs if the state undertakes action that is perceived as protecting bondholder rights or otherwise strengthening a munici- pal issuer’s financial position.80 Conversely, credit markets have historically punished debt issuances by municipalities subject to state laws that fail to pro- tect bondholders in a manner commensurate with market expectations.81

States acknowledge that their municipalities need meaningful debt ad- justment options.82 As noted above, however, subnational governments are dependent on the credit markets. States believe that enacting such options could negatively impact borrowing costs.83 The fear is that a meaningful debt adjustment mechanism will increase the likelihood that a distressed munici- pality will seek to restructure their debts. This option could lead to a height- ened impairment risk for bondholders.84 State legislators and policymakers have accepted the argument that this risk would compel bondholders to de- mand higher interest rates or additional protections.85 A contagion risk is in-

78 Marlowe, supra note 24, at 6. 79 See generally Romano, supra note 13 (discussing law as a “product”). 80 See generally Cynthia Sneed, An Examination of the Effects of Balanced Budget Laws on State Borrowing Costs, 14 J. PUB. BUDGETING, ACCT. & FIN. MGMT. 159 (2002) (concluding that states with balanced budget laws have lower borrowing costs). 81 This punishment invariably comes in the form of a demand for an above-market interest rate on issuances. 82 See, e.g., Jeannie O’Sullivan, Atlantic City Chamber Pushes for Casino Stabilization Laws, LAW360 (Nov. 5, 2015), http:///www.law360.com/articles/724003 [https://perma.cc/4SG6-CD8B] (discussing a financial stabilization bill that the Chamber of Commerce of Atlantic City, New Jersey, advocated for the New Jersey Governor to sign into law). 83 See Yan, supra note 12, at 45–48; Bricketto, supra note 12; Coy, supra note 12 (“State gov- ernments said they didn’t want to be eligible for bankruptcy, fearing that the very possibility would spook investors in municipal bonds and drive up their borrowing costs.”); Nolan, supra note 12; Walsh, supra note 12; State of Pay, supra note 12; see also Barrie Tabin Berger, Telling the Truth About State and Local Finance, GOV. FIN. REV., Apr. 2011, at 79, 80 (explaining that Ray Scheppach, the executive director of the National Governors Association, told the Senate Budget Committee that the enactment of bankruptcy legislation for states would “likely increase interest rates, raise the cost of state government, and create more volatility in financial markets”). 84 See Coy, supra note 12. 85 See id.; see also Yan, supra note 12, at 45–48; Berger, supra note 83, at 80; Bricketto, supra note 12; Nolan, supra note 12; State of Pay, supra note 12.

2017] Failing Cities 613

herent in this fear.86 More specifically, states fear that bondholders will de- mand higher interest rates not just from municipalities that enjoy these options but also the states that offer them.87 Consumed by the prospect of above- market interest rates or even complete denial of access to credit, state legisla- tors and policymakers have refused to act.

This is the paralysis justification; a seemingly plausible argument that supports crippling inaction. As explained above, a critical mass of municipali- ties is eroding as their costs rise, revenues plummet, unfunded pension liabili- ties escalate, and debt service requires the elimination of basic service delivery. State inaction is decimating the interests of all local government constituen- cies, including employees, noteholders, residents, and even the home state it- self. At a macro level, systemic municipal failure has ripple effects that impose significant economic costs on state and national economies.88 Nevertheless, states argue that the alternative is worse.

The paralysis justification, however, is purely anecdotal. This high-risk approach is based on conjecture. We are unaware of any attempt to empirically analyze bond issuances to evaluate these fears. In fact, to the best of our knowledge, no scholarly study has examined, in a statistically meaningful manner, how a municipality’s borrowing costs are affected by its state and fed- eral debt restructuring options.

This Article attempts to addresses the significant gap in municipal litera- ture. Our study used detailed data for every fixed-rate general obligation, mu- nicipal bond89 issued in the United States from January 1, 2004 to December 1, 2014. The dataset’s high quality and depth allowed for a comprehensive evalu-

86 See Gillette, supra note 15, at 302–08; see also Alex Wolf, Christie Blasts AC Takeover Foes, Says Other Cities at Risk, LAW360 (April 6, 2016), http://www.law360.com/articles/781310 [https:// perma.cc/NK67-ZPZD]. 87 See Gillette, supra note 15, at 302–08; Coy, supra note 12; see also Yan, supra note 12, at 45– 48; Bricketto, supra note 12; Nolan, supra note 12; Walsh, supra note 12; State of Pay, supra note 12. 88 See STATE BUDGET CRISIS TASK FORCE, supra note 33, at 2; Ebel, Petersen & Vu, supra note 33, at 35 (“There is a high national interest in what happens to the state and local sector. The 50 states and their nearly 90,000 local governments represent 11.9% of the national product. That is one-and-a- half times the size of the federal sector with national defense spending included, and four-and-a-half times the federal government when one considers only federal non-defense spending. Furthermore, taken together, the states and localities employ about one out of every eight workers and provide the bulk of all basic governmental services consumed by individuals and firms, such as education, public safety, and public works. State and local governments also act as agents in the delivery of many feder- al services, including the passing on of many federal transfer payments from places to people. Moreo- ver, due to the discipline of the municipal credit markets, state and local governments typically bal- ance their operating budgets.” (footnotes omitted)). 89 Fixed-rate, general obligation bonds are invariably unsecured debt and represent the primary means of borrowing used by municipalities. General obligation bond interest rates are more likely to shift based on changes to the issuers’ general risk profile, as compared to the revenue bond interest rates.

614 Boston College Law Review [Vol. 58:599

ation of the paralysis justification. As explained in the following Part, the find- ings were revelatory.

II. OUR METHODOLOGY AND FINDINGS

Part I describes the current municipal landscape and the dire need for meaningful debt adjustment options at the state and federal level. It explains that the vast majority of state policymakers have refused to provide these options based in large part on the paralysis justification, a purely anecdotal argument. Part II explains the dataset used and the use of multivariate regression analysis to evaluate over eight hundred thousand actual bond issuances. Our comprehensive analysis allowed us to determine how a municipality’s borrowing costs are af- fected by the debt restructuring options offered by its home state.

A. The Bond Issuances Studied

Extensive bond-level data were collected from the Bloomberg Terminal’s municipal bond database to determine whether borrowing costs rise based up- on a municipality’s debt adjustment options. Specifically, data were collected for all fixed-rate general obligation municipal90 bonds issued in the United States between January 1, 2004 and December 31, 2014 (the “Observation Pe- riod”).91 Over 800,000 issuances populated our dataset (the “Parikh-He Munic- ipal Bond Project”). We did not rely on a sample of bond issuances during this period or survey results to estimate key features of the dataset. Bond character- istics observed in the dataset include (1) the interest rate,92 (2) the maturity date, (3) duration until maturity, (4) whether or not the bond was callable, and, if available, (5) market prices and yields for the bond in question. Qualitative data about the purpose of the bond and its source of funding was also recorded.

90 As noted, for purposes of this Article, the terms “municipal,” “municipality,” and “municipali- ties” describe counties, cities, and certain other local governments that enjoy taxing authority. Note that 99% of all general obligation bonds issued during the Observation Period had a fixed-rate. 91 We would like to note a few aspects of the bonds that populated the dataset. Primarily, almost all bonds were fixed-rate. Consequently, we decided to eliminate variable-rate bonds because they are of a different nature and were usually the result of unorthodox borrowing circumstances. Further, we did not include revenue bonds because the yield on revenue bonds is driven primarily by the estimated revenue stream from the discrete service attached to that debt. For example, a county that wishes to fund a new sewer system may issue revenue bonds, payments on which will come from fees paid by system users. The yield for this bond issuance will be driven primarily by the stability of the revenue stream from the service provided, not by the overall health or credit worthiness of the county. Inclu- sion of revenue bonds would have distorted our results. Note that we only considered initial bond issuances. We did not consider sales in the secondary market because reliable information on second- ary market trading is unavailable. 92 Interest rate is sometimes referred to as the “coupon rate” or “bond yield.” The interest rate attached to a municipal bond is the best measure of the borrowing costs associated with that bond.

2017] Failing Cities 615

B. Dependent and Independent Variables

To contextualize our data, additional information was collected about each issuer’s fiscal, socioeconomic, and political environment.93 Data from the Census of Governments was used to establish the fiscal health of each issuer and its home state. In particular, we observe total debt outstanding, annual def- icits, and cash holding for each state and its municipalities.94 County-level data on income, education, and demographic composition was then matched to each bond according to location of the issuer and issuance date. This information allowed us to control for variances in municipal borrowers’ socioeconomic status, which is considered in determining credit-worthiness. We also included state-level political affiliations as a variable.95 The Appendix contains tables that summarize this dataset.

The variables noted above were selected because they capture the infor- mation on which credit rating agencies rely in evaluating municipal bonds. Sophisticated bond investors rely on a host of credit rating agencies96 to evalu- ate the risk profile of bond issuances.97 These agencies have formulated an evaluation process that seeks to include all variables that (1) can be measured and (2) have a material effect on an issuer’s risk profile.98 There is significant overlap across the prominent rating agencies. We believe that these variables constitute a viable starting point for our analysis.99 93 For each municipal bond, the county and state of the issuing municipality was established, as well as the quarter and year in which that bond was written. 94 We selected these variables because we believe they capture the information on which credit rating agencies rely in evaluating municipal bonds. See generally STANDARD & POOR’S, TOP 10 MANAGEMENT CHARACTERISTICS OF HIGHLY RATED CREDITS IN U.S. PUBLIC FINANCE (July 2010) (providing an overview of the characteristics that Standard & Poor’s uses to evaluate creditworthi- ness). 95 We include this variable because it could affect state policy and borrowing costs. The variable was based on the average democratic margin of victory in the 2000, 2004, and 2008 presidential elec- tions. For example, assume the results of a given state were as follows: in 2000, 60% of voters voted for Al Gore and 40% for George W. Bush; in 2004, 50% voted for George W. Bush and 45% for John Kerry; in 2008, 65% voted for Barack Obama and 30% for John McCain. For that state, the politics variable would be computed as the average of 20%, –5%, 35%, or 15%. Note that this variable would be negative for a state that had voted for a Republican candidate. 96 Moody’s, Standard & Poor’s (“S&P”), and Fitch are the most notable agencies. 97 See Yan, supra note 12, at 50–54. 98 See generally MOODY’S, U.S. LOCAL GOVERNMENT GENERAL OBLIGATION DEBT (2014) (providing an overview of the rating criteria used by Moody’s Investor Services); STANDARD & POOR’S RATINGS SERVS., U.S. LOCAL GOVERNMENTS GENERAL OBLIGATION RATINGS: METHOD- OLOGY AND ASSUMPTIONS (2013) (providing an overview of the rating criteria used by Standard & Poor’s Ratings Services). 99 Our study’s scope and transformative conclusions have naturally attracted scrutiny. The prima- ry critique we have received is that our study omits a measurable variable (the “Omitted Variable”) that in fact causes the phenomenon we observe in Part II.B. We believe that the probability of an Omitted Variable is extremely low. Our study focuses on the response of the municipal credit markets

616 Boston College Law Review [Vol. 58:599

C. Two Policy Dimensions: Available Debt Adjustment Options and Access to Federal Bankruptcy Court

In addition to the variables noted in Section B of this Part, we considered two policy dimensions in developing our analysis. Primarily, we categorized states based on the debt adjustment options that were offered to their munici- palities.100 As has been detailed by Professor Parikh, states have failed to offer municipalities truly effective debt adjustment options.101 In surveying state law, however, there exists a spectrum of options that supports our comparative analysis.

As captured by Table 1 below, at one end of the spectrum there are thirty- one states that have no formal municipal debt restructuring mechanism or in-

to state laws, and, more specifically, state restructuring options vis-à-vis municipal bonds. Credit market response is premised primarily on the risk-profile of the bond issuer, and the probability of a payment default. See Yan, supra note 12, at 45. Investors have demonstrated that they believe that these two factors are evaluated accurately by credit ratings agencies, including Moody’s, S&P, and Fitch. See id. at 45–48. Borrowing costs are largely driven by the rating given to a bond issue. See id. at 45. Therefore, in compiling our list of variables, we began by capturing the variables that these credit rating agencies consider in rating bond issuances. Sophisticated investors rely on these variables in pricing and evaluating bond issuances. See id. We then considered additional variables that we believe further enhanced our analysis, including resident education and racial composition (the “Addi- tional Variables”). Ultimately, the bond market is a highly efficient market. We find it extremely unlikely that there is a measurable variable that could cause the phenomenon we observe—and essen- tially disrupt pricing in the bond market—but is unidentified by credit rating agencies or not captured by our Additional Variables. 100 As noted throughout this Article, approximately thirty-one states lack statutes that have any material effect on restructuring or payment of debt at the municipal level (the “Debt Restructuring Statutes”). For the other nineteen states and the District of Columbia, amendments to, or enactment of, Debt Restructuring Statutes could arguably have some effect on borrowing costs. To the extent that such an amendment occurred during our Observation Period, we would likely have to account for this shift. Luckily, unlike many other statutes that affect local governments, Debt Restructuring Statutes change infrequently. The vast majority of states enacted their Debt Restructuring Statutes well before the Observation Period; a few enacted them after. For those that did make amendments to their Debt Restructuring Statutes during the Observation Period, we do not believe any have impacted borrowing costs in a material way. For example, in August 2011 the California legislature passed a bill that at- tempted to encourage negotiation by restricting a distressed municipality’s access to Chapter 9. See CAL. GOV. CODE § 53760 (West 2012). The bill provided that under California state law, a bankrupt- cy filing is conditioned on the municipality satisfying one of two prerequisites: participating in a non- binding negotiation for up to ninety days with interested parties holding claims of at least $5,000,000, or passing a resolution declaring a fiscal emergency. See id. Due in part to the ease with which munic- ipalities can avoid having to comply with the first requirement, architects of the legislation acknowl- edged that the new process did not increase bondholder exposure or protection in any material way and should not impact California debt pricing. See Karol K. Denniston, Neutral Evaluation in Chapter 9 Bankruptcies: Mitigating Municipal Distress, 32 CAL. BANKR. J. 261, 286–87 (2012); see also STATE ROLE IN LOCAL GOVERNMENT, supra note 16, at 4. 101 See Parikh, supra note 10, at 240–41.

2017] Failing Cities 617

tervention policy at all (the “No Option States”).102 These states do not monitor meaningfully their municipalities’ fiscal health or offer mechanisms for munic- ipalities to attempt to restructure their debts. Further, nineteen of the No Op- tion States do not allow their municipalities to seek bankruptcy protection un- der Chapter 9.103 This presumably provides bondholders and other creditors significant bargaining leverage when dealing with municipalities facing a payment default.

Table 1: Spectrum of Debt Restructuring Options

Meaningful Debt Ad- justment Options

Limited Debt Ad- justment Options

No Debt Ad- justment Options

Total States

Access to Chapter 9

0 2 9 11

Conditional Access to Chapter 9

10 1 3 14

Access to Chapter 9 Prohibited

2* 5 19 26

Total States 12 8 31

* Includes the District of Columbia

Moving further along the spectrum, we discovered eight states that pro- vide limited debt adjustment options to their respective municipalities (the “Limited Option States”).104 Most of these states—including Illinois and Indi- ana—offer distressed municipalities general financial planning and supervision

102 See JAMES E. SPIOTTO ET AL., MUNICIPALITIES IN DISTRESS?: HOW STATES AND INVESTORS DEAL WITH LOCAL GOVERNMENT FINANCIAL EMERGENCIES, at app. B (2012); see also STATE ROLE IN LOCAL GOVERNMENT, supra note 16, at 9–10. The thirty-one No-Option States are Alabama, Alas- ka, Arizona, Arkansas, Colorado, Connecticut, Delaware, Georgia, Hawaii, Iowa, Kansas, Louisiana, Maryland, Mississippi, Missouri, Montana, Nebraska, New Hampshire, New Mexico, North Dakota, Oklahoma, South Carolina, South Dakota, Texas, Utah, Vermont, Virginia, Washington, West Virgin- ia, Wisconsin, and Wyoming. 103 See SPIOTTO ET AL., supra note 102, at app. B. States that do not allow their municipalities to seek bankruptcy protection under Chapter 9 are Alaska, Colorado, Delaware, Georgia, Iowa, Kansas, Maryland, Mississippi, New Hampshire, New Mexico, North Dakota, South Dakota, Utah, Vermont, Virginia, Wisconsin, Hawaii, West Virginia, and Wyoming. Georgia is the only state that has legisla- tion that specifically forbids its municipalities from filing a Chapter 9 petition. 104 The eight Limited Option States are California, Idaho, Illinois, Indiana, Massachusetts, Min- nesota, Oregon, and Tennessee.

618 Boston College Law Review [Vol. 58:599

as well as limited financial support in the form of small cash infusions or guar- antees on debt. Some Limited Option States—including California and Ore- gon—attempt to provide a forum in which distressed municipalities can nego- tiate with their creditors. These states, however, do not offer municipalities any (1) state resources or (2) additional powers under state law to adjust debts.105

Finally, at the other end of the spectrum, there are eleven states and the District of Columbia that offer debt restructuring mechanisms coupled with fiscal monitoring (the “Meaningful Option States”).106 None of these mecha- nisms are particularly sophisticated, but they provide municipalities a forum to seek adjustment of obligations owed to bondholders and unions.107 For exam- ple, Florida requires municipalities to submit to the state a detailed financial report each year.108 Reports that indicate significant unresolved financial issues are submitted to a state oversight committee that has the capacity to intervene or otherwise facilitate some sort of debt adjustment.109

Access to federal bankruptcy court was the second dimension. Conse- quently, within the primary dimension noted above, we further divided states based on whether they (1) unconditionally allow municipalities to file for pro- tection under Chapter 9 of the federal bankruptcy code, (2) conditionally allow a Chapter 9 filing, or (3) preclude municipalities from filing for Chapter 9.110

These two dimensions allowed us to test the paralysis justification. If the justification is accurate, debt issued by municipalities located in states that of- fer absolutely no debt adjustment mechanism and preclude Chapter 9 filings would have the lowest yields, all other things equal; borrowing costs would rise as one progresses along the spectrum with the highest borrowing costs as- sessed for debt issued by municipalities that had debt adjustment options and access to Chapter 9.

Our data analysis identified a strikingly different phenomenon within the municipal credit market.

105 See Parikh, supra note 10, at 239–41. 106 The eleven Meaningful Option States are Florida, Kentucky, Maine, Michigan, New Jersey, New York, Nevada, North Carolina, Ohio, Pennsylvania, and Rhode Island. 107 See, e.g., SPIOTTO ET AL., supra note 102, at 103–07. For example, in Florida, state agencies monitor municipalities. See id. The state has delineated a list of events that indicate financial distress (the “Trigger Events”). See id. If any agency determines that a Trigger Event has occurred and an emergency is identified, the governor has the authority to take measures to stabilize the municipality, including creating a financial control board that could negotiate with municipal creditors. See id. 108 See FLA. STAT. § 218.39 (2016). 109 See id. Though the Meaningful Option States offer the most impactful debt adjustment mecha- nisms available, these mechanisms are still woefully insufficient and fail to create a structure that offers municipalities viable debt restructuring options. See Parikh, supra note 10, at 237–41. 110 Due to federalism and the separation of powers, a municipality can file for Chapter 9 only if the filing is explicitly permitted under state law. See 11 U.S.C. § 109(c) (2012).

2017] Failing Cities 619

D. Results

Figure 1 below presents average rates of interest for the municipal bonds in our dataset, broken down by state policy. As has been noted, two policy di- mensions were considered in this study: (1) the debt adjustment options of- fered to a state’s municipalities and (2) a municipality’s right to access federal bankruptcy court.

Figure 1: Raw Averages of Municipal Borrowing Costs for Fixed-Rate General Obligation Bonds (2004–2014)

Source: Parikh-He Municipal Bond Project

This preliminary analysis indicates that municipalities with meaningful debt restructuring options have lower borrowing costs. A preliminary tallying of interest rates by state policy, however, is insufficient to establish a causal relationship between policy and borrowing costs because other factors simul- taneously affect a state’s attitude towards municipal debt and the interest rate on such debt. For example, imagine that XYZ state has struggled economically in the past few decades. Those woes have increased borrowing costs to XYZ’s cities, and incentivized XYZ to enact legislation that provides for municipal debt restructuring. In contrast, imagine ABC state has been at the forefront of economic development. ABC maintains high levels of per capita income and education—factors correlated with lower municipal interest rates. These fac- tors have suppressed legislative incentives to install debt adjustment options for local governments. Observing ABC without understanding its economic

620 Boston College Law Review [Vol. 58:599

status might lead one to reach the conclusion that the lack of debt adjustment options is the cause of ABC’s low borrowing costs. In reality, both of these facts are attributable to the state’s economic vitality.

To address this deficiency, we used multivariate regression analysis to control for a state’s fiscal, economic, and political disposition, as well as a bat- tery of other demographic and socioeconomic variables. Multivariate regres- sion analysis allows one to measure the effects of a number of explanatory var- iables on an outcome variable—in this case, borrowing costs—independently from one another.111 This process allowed us to distill the effects of confound- ing factors and reveal the true relationship between debt adjustment and bor- rowing costs.

Table 2 below presents the results of our regression analysis. For compar- ison, three specifications are included. The first is a naive ordinary least squares regression which includes only the dependent and independent varia- bles of interest: (1) the coupon rate, (2) the type of debt restructuring options available, and (3) whether the municipality has access to federal bankruptcy. The coefficients in this table represent the effect on borrowing costs for the corresponding policy scenario, compared to a baseline case. For example, the coefficient in the first line implies that a bond subject to a sophisticated state- level debt adjustment policy commands an interest rate that is .342 percentage points lower than a bond issued by a municipality located in a no-option state. In contrast, the presence of a limited debt adjustment policy raises rates by 0.09 percentage points, compared to a scenario in which such policy is absent.

111 This methodology has been used previously by other scholars. See generally, e.g., Jun Qian & Philip E. Strahan, How Laws and Institutions Shape Financial Contracts: The Case of Bank Loans, 62 J. FIN. 2803 (2007) (using a regression analysis to evaluate how financial contracts respond to legal environments).

2017] Failing Cities 621

Table 2: Regression of Debt Restructuring Policies’ Effect on Municipal Borrowing Costs (2004–2014)112

(1) (2) (3)

Naive OLS Bond Lvl Ctrls Other Ctrls Debt Adjustment Options Meaningful -0.342** -0.228*** -0.139*

(-2.43) (-3.14) (-1.79) Limited 0.0917 0.0780 0.0311

(0.64) (0.90) (0.38) Chapter 9 Status Ch.9 Allowed 0.0708 -0.0121 -0.0361

(0.47) (-0.14) (-0.49) Ch.9 Conditional 0.406*** 0.238*** 0.0465

(3.04) (2.85) (0.75) Bond Level & Other Controls Callable -0.333

*** -0.178***

(-5.03) (-3.59) Maturity Size 0.0106 0.00664

**

(1.46) (2.18) Maturity Length (Yrs) 0.110

*** 0.0876***

(6.99) (4.49) Total Local Debt 0.00193

(0.29) Total State Debt 0.00284

(0.63) Total Local Deficit 0.0155

(0.41) Total State Deficit -0.00669

(-0.42) Political Affiliation 0.00109

(1.42) Education 0.407

(1.47) Pct. African American 1.322

***

(3.92) Median HH Income -0.000992

(-0.35) Total Population 0.0371

112 Probability values are indicated as follows: * p < 0.10, ** p < 0.05, *** p < 0.01. t statistics are in parentheses. Standard errors are clustered at state by year level. See generally A. Colin Cameron, Jonah B. Gelbach & Douglas L. Miller, Robust Inference with Multiway Clustering, 29 J. BUS. & ECON. STAT. 238 (2012) (explaining a method of variance estimation for data sets with “cluster- robust” standard errors). Education was measured as per-capita attainment of a bachelor’s degree. Political attitude was measured as mean difference in Democratic and Republican votes in the 2000, 2004, and 2008 presidential elections. Median household income is in thousands USD per year. Ma- turity Size is in millions USD. Column three omits years 2012–2014 due to data availability.

622 Boston College Law Review [Vol. 58:599

(1.47) Pop. Density 0.00358

(1.57) Observations 874478 872186 635594

R2 0.011 0.243 0.392

The regression in the second column controls for bond-level characteris- tics, including duration until maturity and whether or not the bond is callable. Two important facts should be noted in this table. First, the estimated coeffi- cients corresponding to maturity size, maturity length, and callability are all consistent with classical bond pricing theory. In particular, factors that lead to lower risk for the borrower should also lead to a lower coupon rate. This ex- plains why callability yields a negative point estimate, while estimates for ma- turity size and length are both positive. Second, the inclusion of additional controls changes the estimated effects of the policy variables. More specifical- ly, all of the policy coefficients have diminished in absolute value. In simple terms, this phenomenon implies that some of the net effects on borrowing costs previously attributed to policy differences are explained by differences in bond characteristics.113

The final specification adds controls for (1) the debt load, cash holdings, and yearly deficit/surplus for all municipalities located within a given state, (2) each individual state’s debt load, cash holdings, yearly deficit/surplus, and po- litical affiliation, and (3) resident demographics, education, and income. As mentioned, these factors could affect jointly state policy and borrowing costs, necessitating their inclusion in the regression. For example, median household income is a proxy for a state’s economic vitality. It is a factor that lowers bor- rowing costs while also encouraging states to ignore calls to implement mean- ingful municipal debt adjustment options. If income is neglected, one might infer incorrectly that the lack of a debt adjustment policy causes lower interest rates. This is why the negative estimates in the first line of Table 2—the inter- est reducing effects of a sophisticated debt adjustment system—become milder with the inclusion of demographic controls. Despite the addition of these con- trols, however, the estimate is still negative in sign and statistically significant at the ninety percent level.114

Figure 2 below uses the regression results above to construct counterfac- tual estimates of borrowing costs for the average city under varying policy en- vironments. More specifically, multivariate regression analysis allows the val-

113 See WILLIAM H. GREENE, ECONOMETRIC ANALYSIS 339–78 (5th ed. 2003) (explaining how multivariate regression can be used to control for the effects of observable factors, and the assump- tions implicit in this methodology). 114 The reduction in sample size in the third specification is due to the fact that data on municipal finances was not available for all years in our observation period.

2017] Failing Cities 623

ue of a given dependent variable to be expressed as a weighted sum of numer- ous independent variables. Based on these weights, we were able to predict the value of that dependent variable in the event the independent variables as- sumed values different from what they are in the sample. In this case, the de- pendent variable in question is the coupon rate of each bond, while the inde- pendent variables include state policies regarding debt adjustment, characteris- tics of the bond in question, and numerous demographic controls.115

Our estimates were constructed by considering a hypothetical city located in a county whose characteristics are equal to their national averages. The val- ues in the figure reflect differences in borrowing costs by state policy for a county with an average population, average level of education, and average income, among other factors.

Figure 2: Municipal Borrowing Costs For Fixed-Rate General Obligation Bonds, All Other Things Being Equal (2004–2014)

Source: Parikh-He Municipal Bond Project Figure 2 captures this Article’s primary contribution and provides insight

into a number of issues. First, borrowing costs are uniformly lower in the pres- ence of meaningful debt adjustment options. Further, note the ninety-five per- cent confidence intervals on each bar. The lack of overlap between the confi-

115 See GREENE, supra note 113, at 7–19 (providing definitions for dependent and independent variables in the context of regression models).

624 Boston College Law Review [Vol. 58:599

dence intervals116 indicates that the variance in costs is statistically significant regardless of Chapter 9 filing status. The Chapter 9 filing dimension, however, clarifies market response. Indeed, allowing municipalities to file uncondition- ally for Chapter 9 further suppresses borrowing costs. For both Limited Option and No Option States, municipalities with relatively unrestricted access to Chapter 9 enjoyed lower borrowing costs.

In addition, we observe that municipalities with limited debt adjustment options and conditional access to Chapter 9 had the highest borrowing costs. As discussed further in Part III, this is not surprising. In states where munici- palities have poorly formed restructuring options and their right to file for Chapter 9 is subject to the whims of the state legislature or governor, creditors may fear rent-seeking and collective action. More specifically, creditors price the risk that, in a distress scenario, a small subset of stakeholders will be able to influence the state legislature or governor to intervene and facilitate action that is most advantageous to the subset but detrimental to the creditor body as a whole. Consequently, the bond market is more receptive when a municipality has full access to Chapter 9—a process that attempts to solicit equitable con- cessions from all stakeholders—or no access at all.

E. Extrapolating the Data to Determine the Efficacy of the Parikh Proposal

Professor Parikh has proposed a normative debt adjustment mechanism that can be enacted at the state-law level.117 This mechanism has five key fac- ets. Primarily, Professor Parikh proposed that states engage in soft monitoring of municipal financial conditions within a framework that is built on identify- ing at-risk municipalities.118 Verified at-risk municipalities graduate to a hard monitoring system. Thereafter, a host of financial criteria are reviewed and triggers are used to determine fiscal distress.119 Devolution120 is reversed for

116 For example, there is no overlap among the confidence intervals for municipalities with mean- ingful debt restructuring options and conditional access to Chapter 9 and (1) municipalities with lim- ited debt restructuring options and conditional access to Chapter 9; and (2) municipalities with no debt restructuring options and conditional access to Chapter 9. Similarly, there is no overlap among the confidence intervals for municipalities with meaningful debt restructuring options and no access to Chapter 9 and (1) municipalities with limited debt restructuring options and no access to Chapter 9; and (2) municipalities with no debt restructuring options and no access to Chapter 9. 117 See Parikh, supra note 10, at 277–96. 118 Only fifteen states monitor their local government’s financial condition. See Philip Kloha et al., Someone to Watch Over Me: State Monitoring of Local Fiscal Conditions, 35 AM. REV. PUB. ADMIN. 236, 240, 252 (2005). 119 In his article A New Fulcrum Point for City Survival, published in the William and Mary Law Review, Professor Parikh stated:

Because we are attempting to identify municipalities at an earlier stage of financial de- terioration, relying exclusively on customary red flags—including payment defaults—is

2017] Failing Cities 625

fiscally distressed municipalities that face payment defaults or service delivery insolvency. A restructuring control board is appointed to manage all operation- al and financial matters for the local government.121 As Professor Parikh has stated, “[t]he board’s ultimate goal [is] to develop a recovery plan and new operating budget that restructures the municipality’s expenses and obligations

ineffective; they represent symptoms of a disease that has already proliferated. States would need to formulate their own lists of financial triggers, but they should incorpo- rate a collection from the following criteria: (1) the municipality’s credit rating has been downgraded; (2) the municipality executed an intra-fund transfer that suggests a deficit somewhere in the budget; (3) the municipality failed to file timely financial reports; (4) a major employer located within the municipality has closed an office or significantly downsized its operations; (5) the municipality failed to transfer taxes withheld on em- ployee income or employer and employee contributions for a pension, retirement, or benefit plan; (6) the municipality’s mandated audit report shows funds with deficit fund balances; (7) local officials have failed to correct problems—including internal control problems—after being notified by state officials; (8) the municipality has insufficient cash to meet required payroll payments in a timely manner; (9) the municipality has vi- olated a covenant in its credit agreements; (10) the municipality has recognized sizeable losses as a result of unnecessarily aggressive investment practices; (11) the municipali- ty has expended restricted funds in violation of applicable terms and provisions; (12) the ending balance in the municipality’s general fund has declined for two consecutive years; (13) the municipality is experiencing significant service delivery interruptions; (14) the municipality faces the likelihood of a default on debt payments or an inability to pay vendors, employees, or creditors with uncontested claims; and (15) the munici- pality has experienced high levels of revenue inefficiency.

Parikh, supra note 10, at 279–80. 120 The term “devolution” describes a state’s delegation of power and management of local affairs to municipalities and their residents. See Devolution, BLACK’S LAW DICTIONARY (10th ed. 2014). This practice is the foundation of the “home rule” movement, which seeks to “ensure[] that govern- mental power is exercised closest to the people.” David J. Barron, Reclaiming Home Rule, 116 HARV. L. REV. 2255, 2259 (2003). For an analysis of the complex bargaining and the legal rules that charac- terize the power-sharing relations between all levels of government in the United States, see generally Erin Ryan, Negotiating Federalism, 52 B.C. L. REV. 1 (2011). 121 Professor Parikh described the operations and goals of the control board by stating:

[T]he board attempts to preserve local democracy by giving residents a voice through elected state and local officials who either directly participate in the board’s decision- making processes or select board members who will be directing board action. A dif- fused decision-making structure makes the board less susceptible to capture and self- dealing . . . . [T]he board should be allowed to conduct all aspects of municipal opera- tions through majority vote. These options include effectuating debt service, making necessary contributions to pension funds, managing tax policy, hiring and removing municipal employees, seeking additional financing, exercising the authority of local of- ficials, and restructuring municipal offices, departments, and agencies. The board will have the discretion to formulate an optimal delegation structure. The board would be bound by state law but could request that the state legislature waive certain restrictions if such a waiver process exists under state law.

Parikh, supra note 10, at 282–83.

626 Boston College Law Review [Vol. 58:599

to ensure sustainable viability.”122 He has also proposed a clear negotiation structure that enables the board to seek material concessions from bondholders and current and former employees. Finally, his proposal set limited negotiation periods, after which the board must either propose a recovery plan or authorize the municipality to file a petition under Chapter 9 of the Federal Bankruptcy Code.123

The three key aspects of Professor Parikh’s proposal are meaningful debt restructuring options, comprehensive fiscal monitoring, and unconditional ac- cess to Chapter 9.124 We wondered how the bond market would respond to a municipal bond issuer with this profile. Consequently, we sought to estimate the borrowing costs for a hypothetical municipality located in a state that had enacted Professor Parikh’s proposal (the “Parikh Proposal”) at some point be- fore January 1, 2004. This was done by extrapolating the data from the Parikh- He Municipal Bond Project. The predictions with respect to borrowing costs for a municipality with meaningful debt restructuring options, fiscal monitor- ing, and unconditional access to Chapter 9 were made by employing the same methodology underlying the conclusions in Figure 2.

The cross-hatched bar in Figure 3 below captures the assessment. Extrap- olating the data leads to the conclusion that a hypothetical municipality located in a state that had adopted the Parikh Proposal would have the lowest possible borrowing costs, all other things being equal. We assert that our data analysis supports the Parikh Proposal’s efficacy. Indeed, borrowing costs for this hypo- thetical municipality would be approximately twenty-four basis points lower than a municipality with limited debt adjustment options and conditional ac- cess to Chapter 9 and approximately twenty-two basis points lower than a mu- nicipality with no debt-adjustment options and conditional access to Chapter 9.

122 Id. at 283. 123 See id. at 294–95. Professor Parikh proposed unconditional access to Chapter 9 after an exten- sive negotiation period. The benefit of unconditional access to Chapter 9 after such discussions is to incentivize recalcitrant creditors to come to the negotiating table and make concessions. If a state legislature or governor is empowered to block access to Chapter 9, the holdout problem would pre- clude any possibility of a meaningful restructuring. See id. at 257 (“Under my proposal, restructuring officials have autonomy to authorize a Chapter 9 filing if the negotiations required under the system’s parameters prove fruitless. This option is subject to satisfying a variety of negotiation prerequisites, but, once authorized, cannot be altered by state officials or legislatures. This aspect, coupled with the elimination of bailouts, brings holdouts to the negotiating table.”). 124 In determining whether a state had given its municipalities an unconditional right to file for Chapter 9, we considered whether the municipality had, at any point during their restructuring en- deavors, an unconditional option to file for Chapter 9. For example, a state could give its municipali- ties a blanket right to file for Chapter 9, as Texas does. A state would also qualify for this categoriza- tion if a municipality had the right to file for Chapter 9 upon completing settlement discussions with its creditors.

2017] Failing Cities 627

Figure 3: Municipal Borrowing Costs For Municipality Located in a State That Had Enacted the Parikh Proposal (2004–2014)

Source: Parikh-He Municipal Bond Project These borrowing spreads are significant. A local government financial

manager’s primary goal in debt issuance is to minimize the cost of capital.125 As one scholar has pointed out, because local government debt issuances are invariably long-term fixed coupon securities, “a few basis point difference in interest costs on those securities can have multi-million dollar implications for overall cost of capital.”126 As detailed in Part IV, suppressed borrowing costs can have myriad benefits for a municipality and its stakeholders.

There remains, however, the question whether these results can be recon- ciled with inveterate principles of municipal borrowing and restructuring. We believe they can.

III. UNDERSTANDING THE PHENOMENON

Part II explains the transformative results of the study and data analysis used in this Article. More specifically, based on our dataset’s depth and high quality, we are able to conclude that municipalities located in states that offer

125 Marlowe, supra note 24, at 6. 126 Id. Part IV presents the actual savings a municipality could realize.

628 Boston College Law Review [Vol. 58:599

meaningful debt restructuring options—through both state and federal law127— will enjoy lower borrowing costs on their bond issuances. Further, we extrapo- lated our data to conclude that municipalities located in states that adopt the Parikh Proposal128 would realize the lowest possible borrowing costs. Our re- search reveals a distinct phenomenon in the municipal credit market. The idea that a lender would favor municipalities that have heightened debt impairment options may appear counterintuitive. It is not. Our conclusion aligns with ac- cepted financial restructuring principles.129 Recent changes to sovereign bond issuances present an insightful analogy.

Almost all developed, sovereign governments issue unsecured bonds. At- tendant debt obligations are repaid pursuant to a delineated schedule. Before the 2000s, in most cases, the amount and timing of this repayment could be impaired only if all holders of a particular bond agreed to the modification.130 Historically, this fact did not preclude debt restructurings because of the small pool of bondholders.131 The repeat-player model encouraged cooperation and discouraged holdouts and free riding.132 In the 1980s, however, the pool of bondholders became much larger and far less homogenous.133 Consequently, sovereigns with unsustainable debt burdens faced especially pernicious collec- tive action problems.134 In response, officials from “systemically important

127 At the state law level, states should monitor their municipalities and offer a debt adjustment mechanism that facilitates negotiation with creditor constituencies and meaningful debt restructuring. At the federal law level, states should authorize their municipalities to seek relief under Chapter 9 of the Federal Bankruptcy Code in the event these negotiations fail. 128 See Parikh, supra note 10, at 284–94 (proposing a comprehensive, fiscal monitoring system that identifies and then directs distressed municipalities into a dynamic negotiation model supported by the option to file for federal bankruptcy). 129 See, e.g., THOMAS H. JACKSON, THE LOGIC AND LIMITS OF BANKRUPTCY LAW 24 (1986); Thomas H. Jackson, Bankruptcy, Non-Bankruptcy Entitlements, and the Creditors’ Bargain, 91 YALE L.J. 857, 860–65 (1982); Elizabeth Warren, Bankruptcy Policymaking in an Imperfect World, 92 MICH. L. REV. 336, 346 (1993). 130 ANNE O. KRUEGER, INT’L MONETARY FUND, A NEW APPROACH TO SOVEREIGN DEBT RE- STRUCTURING 6–7 (2002). This statement refers to bonds subject to New York state law. Bonds not subject to New York state law occasionally offer issuers more flexibility. These types of bonds are a fraction of the overall market, however. Bonds subject to New York law dominate the sovereign debt market. See Bradley, Cox & Gulati, supra note 32, at 295–96; Anna Gelpern & Mitu Gulati, Public Symbol in Private Contract: A Case Study, 84 WASH. U.L. REV. 1627, 1640 (2006) (noting that New York law bonds dominated the sovereign debt market and they accounted for over 80% of all emerg- ing market debt outstanding in 2002). 131 See KRUEGER, supra note 130, at 6–7. 132 See id. 133 See id. 134 See Gelpern & Gulati, supra note 130, at 1693. Typical collective action problems that arise when a sovereign experiences financial distress include (1) panicked selling of bonds (also known as the “rush to the exits”); (2) preemptive litigation (the “rush to the courthouse”); (3) holding out and hoping for an oversized settlement; and (4) hoping to free ride on a restructuring agreement. See Barry

2017] Failing Cities 629

economies” began formulating debt-restructuring provisions that could be in- serted into bond agreements.135

Among many competing options, the collective action clause (the “CAC”) was the most compelling. The CAC involved altering the unanimity requirement for debt modification. More specifically, the bond agreement would allow a ma- jority or supermajority of bondholders to authorize an impairment of the debt that would affect all bondholders.136 This change was intended to address hold- out and collective action problems. In 2003, Mexico tested the market by issuing twelve-year global notes that allowed the holders of seventy-five percent of the outstanding principal to amend financial terms, which marked a departure from market convention.137

Not surprisingly, Mexico’s CAC received scrutiny and criticism. Debt impairment is extremely difficult without a provision like the CAC. Mexico’s CAC diminished a significant safeguard for bondholders. As a result, many observers were certain that credit markets would demand higher interest rates on issuances.138 That result did not materialize, however. In fact, by 2006, CACs had become virtual boilerplate in new sovereign debt agreements.139 Further, recent empirical research established that CACs did not increase sov- ereign borrowing costs, all other things being equal.140 There are a variety of reasons that explain this phenomenon, and those very same reasons explain the phenomenon we observe in the municipal credit market.

Primarily, delineated debt restructuring options offer certainty. During the underwriting process, certainty regarding a possible restructuring allows credi- tors to evaluate risk exposure.141 Further, certainty and an actual restructuring mechanism can minimalize holdout risk.142 During financial distress, informed

Eichengreen, Restructuring Sovereign Debt, 17 J. ECON. PERSP. 75, 81–82 (2003); see also JACKSON, supra note 129, at 11–14. 135 See GROUP OF TEN, THE RESOLUTION OF SOVEREIGN LIQUIDITY CRISES: A REPORT TO THE MINISTERS AND GOVERNORS PREPARED UNDER THE AUSPICES OF THE DEPUTIES 16 (1996); INT’L MONETARY FUND, REPORT OF THE WORKING GROUP ON INTERNATIONAL FINANCIAL CRISES 28–30 (1998). 136 See GROUP OF TEN, supra note 135, at 16. This type of CAC was already a fixture in sover- eign bonds issued in the United Kingdom, but are still only a fraction of the overall market. As noted in this Part, New York-law bonds dominate the sovereign debt market. See Bradley, Cox & Gulati, supra note 32, at 296. 137 Gelpern & Gulati, supra note 130, at 1641. 138 See KRUEGER, supra note 130, at 6–7. 139 Gelpern & Gulati, supra note 130, at 1641. 140 See Bradley, Cox & Gulati, supra note 32, at 295–98. 141 See KRUEGER, supra note 130, at 6–7. 142 See Parikh, supra note 10, at 254 (“A process where parameters and procedures are defined ex ante incentivizes municipality and creditor constituencies to engage fully. Clarity engenders certainty. Under this premise, all key constituencies have a meaningful understanding of available options. This

630 Boston College Law Review [Vol. 58:599

constituents and clear parameters increase resolution speed because debt im- pairment is implemented uniformly and pro rata. That principle explains this Article’s finding that municipalities located in states with limited restructuring options and conditional access to Chapter 9 had the highest borrowing costs.143 Indeed, states with delineated restructuring options offer creditors an apprecia- ble level of certainty, as do states with no municipal restructuring options at all—though this latter type of certainty is of a vastly different flavor. States with limited restructuring options, however,—which oftentimes involve ad hoc provisions fashioned at the whim of the state legislature144—offer marginal certainty.145 The bond market’s response to this marginal certainty is decidedly negative.

Finally, a delineated system allows distressed borrowers to address cor- rupt capital structures at an earlier stage of deterioration and thereby, as one economist put it, “avoid[] the exhaustion of official reserves and unnecessarily severe economic dislocation.”146 Debt restructuring options provide a borrower with means to address its financial distress and hopefully avoid a full-scale payment default.147

dynamic minimizes posturing and irrational threats. Without this certainty, the prospect of a state or federal bailout emboldens holdouts.”); see also Skeel, supra note 19, at 694–701. 143 See infra notes 144–156 and accompanying text. 144 See Parikh, supra note 10, at 239 (“For example, Massachusetts offers an example of an ad hoc, reactive system. Under Massachusetts state law, legislation is passed to address municipal dis- tress on a case-by-case basis.”). 145 See id. Professor Parikh has written previously on ad hoc solutions:

Primarily, as with all ad hoc systems, intervention is at the whim of the state legislature. Procedures are not codified, which leads to crippling uncertainty and disparate treat- ment among municipalities. This approach undermines bargaining. Counter-parties, in- cluding bondholders and unions, believe that the state will come to the rescue, which emboldens holdouts. Local officials believe that the state will come to the rescue, which creates moral hazard. Access to credit and borrowing costs are also distorted. Reactive approaches are similarly deficient because they greatly increase the odds that the mu- nicipality will have experienced irreparable deterioration by the time the state inter- venes.

Id. at 240 (footnotes omitted). 146 KRUEGER, supra note 130, at 5; see also Parikh, supra note 10, at 285 (“My system is prem- ised on a shared burden among all creditor constituencies and seeks to capture distressed municipali- ties at a time where less sweeping concessions will be sufficient.”). 147 To understand the severity of full-scale debt defaults see J.F. HORNBECK, CONG. RESEARCH SERV., ARGENTINA’S DEFAULTED SOVEREIGN DEBT: DEALING WITH THE “HOLDOUTS” 3 (2013) (“On December 20, 2001, President de la Rua resigned and six days later, an interim government defaulted on Argentina’s sovereign debt . . . . Total public debt mushroomed . . . and the default left the Argentine government in arrears with a number of international creditors. At the time, Argentina owed private investors bonds with a face value of $81.8 billion, the Paris Club countries $6.3 billion, and the IMF $9.5 billion, among other domestic and multilateral obligations.”).

2017] Failing Cities 631

The confluence of these factors harmonizes the phenomenon we observe in the municipal credit markets with financial restructuring literature. The next Part explores the implications of these findings.

IV. EXPLORING THE IMPLICATIONS

The previous Part detailed how our empirical results can be reconciled with accepted financial restructuring principles. This Part explores the implica- tions of those results.

Empirical researchers frequently fail to understand that statistical signifi- cance does not equate to policy significance. Our results are statistically signif- icant and support the conclusion that the paralysis justification is a false narra- tive. This conclusion also has profound policy significance, cutting across an array of government and debt issuance dimensions. Indeed, disengaged states should begin implementing meaningful debt adjustment mechanisms for their failing municipalities, and they should arguably consider one modeled on the Parikh Proposal.148 The benefits could be transformative.

A. Reduced Borrowing Costs

Municipalities that gain access to a meaningful debt restructuring mecha- nism characterized by fiscal monitoring, creditor negotiation, and full access to Chapter 9 could enjoy a significant reduction in borrowing costs. Reduced cost of capital is the primary benefit that a municipality will enjoy. The impact of this benefit is best understood in absolute dollars. For example, from January 1, 2004 to December 31, 2014, the city of Hartford, Connecticut issued 4,355 fixed-rate, general obligation municipal bonds. If Connecticut had adopted Professor Parikh’s municipal debt restructuring proposal prior to these issuanc- es, we posit that Hartford would have enjoyed a twenty-two basis point reduc- tion in its interest rate.149 In any given year during the Observation Period, our research suggests that Hartford would have saved $32.93 million in interest payments. This amount represents approximately 6.2% of Hartford’s yearly general fund total expenditures.150 Over the life of the debt, Hartford would have paid approximately $336 million less in interest.

148 See Parikh, supra note 10, at 284–94. 149 Note that Connecticut is a No Option State that offers its municipalities a conditional grant to file for Chapter 9. 150 From 2004 to 2014, Hartford’s general fund total expenditures averaged approximately $535 million per year. See Budgets and Proposals, HARTFORD.GOV, http://www.hartford.gov/management- and-budget/budgets-and-proposals [https://perma.cc/386R-22XT] (containing a collection of Hart- ford’s budget documents).

632 Boston College Law Review [Vol. 58:599

This benefit is not limited to relatively small cities.151 Large counties would benefit similarly. For example, from December 1, 2004 to December 31, 2014, Fulton County, Georgia - where Atlanta is located - issued 14,296 fixed- rate general obligation municipal bonds. If Georgia had adopted the Parikh Proposal, we posit that Fulton County would have enjoyed an 18.5 basis point reduction in its interest rate.152 In any given year during the Observation Peri- od, our research suggests that Fulton County would have saved $19.28 million in interest payments. This amount represents approximately 3.4% of the coun- ty’s yearly general fund total revenues.153 Over the life of the debt issued dur- ing the Observation Period, Fulton County would have paid approximately $208 million less in interest.

Hartford and Fulton County are not outliers. Table 3 below lists thirty- five municipalities and the estimated interest savings they would have enjoyed if their home state had enacted the Parikh Proposal.

Expanding this analysis further, Table 4 below lists the aggregate savings for all municipalities located within the top thirty states with the largest bor- rowing cost savings.

151 See Welcome to Hartford, HARTFORD.GOV, http//www.hartford.gov [https://perma.cc/7KXD- ZZS6] (noting that Hartford has a population of approximately 125,000). 152 This is especially noteworthy because Georgia is one of the few states that explicitly prohibits its municipalities from seeking protection under federal bankruptcy law. 153 From 2004 to 2014, Fulton County’s general fund total revenues averaged approximately $587 million per year (the yearly totals are as follows: $575M in 2004; $591M in 2005; $650M in 2006; $600M in 2007; $597M in 2008; $612M in 2009; $637M in 2010; $534M in 2011; $550M in 2012; $529M in 2013; $585M in 2014). See FULTON CTY. GA., FISCAL YEAR 2014 ADOPTED BUDGET; FULTON CTY. GA., FISCAL YEAR 2012 ADOPTED BUDGET; FULTON CTY. GA., 2008 BUDGET BOOK.

2017] Failing Cities 633

Table 3: Municipalities With the Largest Interest Savings If Their Home State Had Enacted the Parikh Proposal (2004–2014)

County

State

Savings Per Year

(Millions)

Total Savings (Millions)

Bonds Out- standing (Millions)

Sacramento California 188.0 2923 86975 Cook Illinois 84.91 1133 54993

Los Angeles California 54.13 863.7 27689 New York New York 38.15 511.2 58075

Suffolk Massachusetts 35.68 491.5 23491 Harris Texas 34.26 526.0 32811

Hartford Connecticut 32.93 336.0 21210 Ramsey Minnesota 19.54 185.3 16126 Dallas Texas 19.49 276.5 18368 Fulton Georgia 19.28 208.4 14296

Anne Arundel Maryland 17.08 158.6 14836 Dane Wisconsin 16.77 181.0 12824

Honolulu Hawaii 16.27 196.1 12951 Santa Clara California 13.92 226.3 8401 San Diego California 13.56 264.9 7588

Bexar Texas 13.13 225.8 12638 Maricopa Arizona 12.97 141.9 12485

King Washington 11.45 142.2 11223 Tarrant Texas 10.95 162.2 10465 Franklin Ohio 10.87 109.9 20085 Collin Texas 10.81 163.3 10724

Davidson Tennessee 10.66 124.2 7035 Dauphin Pennsylvania 10.53 119.8 16875 Alameda California 10.49 169.9 6392

Leon Florida 10.01 152.0 15612 Clark Nevada 9.324 133.7 12854

Salt Lake Utah 8.586 77.41 5657 Hennepin Minnesota 8.208 74.69 6653 Dupage Illinois 8.007 103.1 4793 Fairfield Connecticut 7.950 75.55 4921 San Fran. California 7.758 84.11 4727 Riverside California 7.365 125.9 4006

Wake N. Carolina 7.293 76.17 11471 Middlesex Massachusetts 6.892 61.95 4807

E. Bat. Rouge Louisiana 6.614 69.82 5409

TOTAL $794 $10,875

634 Boston College Law Review [Vol. 58:599

Table 4: Aggregate Savings for All Municipalities Within a Given State If That State Had Enacted the Parikh Proposal (2004–2014)

State

Savings Per Year

(Millions)

Total Savings (Millions)

Bonds Outstanding (Millions)

Arizona 17.24 183.4 16196 California 338.1 5354 172240 Colorado 23.16 351.2 17316

Connecticut 50.11 509.2 31904 Florida 13.40 203.8 20850 Georgia 34.10 338.2 25233 Hawaii 17.70 213.1 14044 Illinois 129.0 1654 82479 Kansas 18.38 186.9 15134

Louisiana 13.75 149.8 10361 Maryland 40.52 406.7 34586

Massachusetts 58.19 695.3 38968 Michigan 20.02 252.6 31664 Minnesota 49.94 488.0 41053 Mississippi 10.25 121.7 7574 Missouri 12.69 145.7 12938 Nevada 12.59 173.3 17706

New Jersey 18.52 192.0 31636 New York 62.52 739.6 100298

North Carolina 13.68 138.5 21143 Ohio 23.91 296.6 41909

Oregon 18.17 234.3 12517 Pennsylvania 41.62 520.7 66186

South Carolina 14.06 143.0 13524 Tennessee 29.87 340.3 19831

Texas 171.5 2613 170229 Utah 13.97 132.8 9942

Virginia 28.96 309.7 23065 Washington 25.35 311.7 25723 Wisconsin 40.81 379.8 32819

Total 1,362 17,087

2017] Failing Cities 635

Our research suggests that municipalities in Limited Option States154 could have saved an aggregate of approximately $8.8 billion in borrowing costs over the life of all fixed-rate general obligation, municipal bonds issued during the Observation Period, and that municipalities in No Option States155 could have saved an aggregate of $7.2 billion during that same period.156 If used judiciously, these savings could stabilize distressed municipalities within a given state.

B. The Benefits of Stabilized Municipalities

Stabilized municipalities provide significant benefits to all constituencies. Primarily, there is an archetypical multiplier effect. As noted in this Article, credit rating agencies place significant emphasis on the fiscal health of the state’s municipalities.157 Strengthened municipalities bolster a state’s credit rating, which, in turn, reduces the state’s borrowing costs.158 Stabilized munic- ipalities are less likely to default on debt obligations or service delivery. His- torically, states have acted as implicit guarantor of municipal debts and obliga- tions.159 Naturally, a guarantor is absolved of its burden if the primary obligor can fulfill its obligations.

Meaningful debt restructuring options assist a municipality—and its con- stituencies—by offering a means to address serious fiscal challenges. As de- tailed in Part III, delineated debt restructuring mechanisms minimize free rid- ing, holdout risk, and collective action problems, which ultimately engenders debt restructuring and accelerates resolution speed.160 Municipalities will also hold the option of addressing fiscal challenges at an earlier stage of financial deterioration. By doing so, they avoid depleting resources and can invariably seek less dramatic concessions from creditor constituencies.161 This one facet

154 The Limited Option States are California, Idaho, Illinois, Indiana, Massachusetts, Minnesota, Oregon, and Tennessee. 155 The No Option States are Alabama, Alaska, Arizona, Arkansas, Colorado, Connecticut, Dela- ware, Georgia, Hawaii, Iowa, Kansas, Louisiana, Maryland, Mississippi, Missouri, Montana, Nebras- ka, New Hampshire, New Mexico, North Dakota, Oklahoma, South Carolina, South Dakota, Texas, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, and Wyoming. 155 The 8 Limited Option States are California, Idaho, Illinois, Indiana, Massachusetts, Minneso- ta, Oregon, and Tennessee. 156 These numbers do not capture any expected interest savings on bonds issued by the applicable state. 157 See STANDARDS & POOR’S, supra at note 94, at 4 (“At the state level, we believe that local government fiscal difficulties can increase and become a funding challenge for the state.”); see also Parikh, supra note 11, at 258–59. 158 See STANDARDS & POOR’S, supra at note 94, at 4; see also Parikh, supra note 10, at 259. 159 See Parikh, supra note 10, at 237. 160 See Jackson, supra note 129, at 860–65; see also Warren, supra note 129, at 346. 161 See KRUEGER, supra note 130, at 6–8; Parikh, supra note 10, at 285, 288.

636 Boston College Law Review [Vol. 58:599

increases materially the likelihood of a successful restructuring that establishes sustainable viability.162 Further, states can enhance the attractiveness of munic- ipal bonds as an asset class by injecting more certainty into the municipal bond market. This is a compelling proposition because borrowing costs oftentimes move inversely to the supply of funds in the market.163

Finally, and perhaps most importantly, the confluence of these factors in- creases the probability that a municipality would be able to provide essential services and resources to its residents.

CONCLUSION

In this Article, we employed empirical analysis to support the conclusion that municipal credit markets do not demand higher borrowing costs from mu- nicipalities that have meaningful debt restructuring options. In fact, our re- search establishes that these municipalities have lower borrowing costs, all other things being equal. Professor Samir D. Parikh has theorized that a munic- ipality enjoying meaningful debt restructuring options, fiscal monitoring, and unconditional access to Chapter 9 would have the absolute lowest borrowing costs. Our results support that theory. This Article transforms the discussion in the municipal restructuring arena. States have misguidedly used the paralysis justification to block implementation of meaningful debt restructuring options vital to avoid service-delivery insolvency.164 State legislators and policymakers should now begin the process of proposing legislation that will hopefully lead to structural solutions. Naturally, there remain obstacles to implementation, but the paralysis justification should not be one of them. Ultimately, municipal literature has produced disparate views on how distressed municipalities can achieve sustainable viability. This Article attempts to unify the diverse argu- ments and illuminate a path. State inaction will serve only to amplify financial distress, scaling a manageable problem to an apocalypse.

162 See Parikh, supra note 10, at 285, 288. 163 See Aaron Kuriloff, Muni Bonds Headed for a Rough Patch, WALL STREET J. (Mar. 6, 2015), http://www.wsj.com/articles/muni-bond-headed-for-a-rough-patch-1425838695 [https://perma.cc/ 28D3-GPTB]. 164 See, e.g., Coy, supra note 12.

2017] Failing Cities 637

APPENDIX

SUMMARY STATISTICS FROM REGRESSION

Source: Parikh-He Municipal Bond Project

TOP 10 USES OF FUNDS RECEIVED FROM ISSUANCE OF FIXED-RATE GEN- ERAL OBLIGATION MUNICIPAL BONDS (2004 – 2014)

Number of Bonds Percent

Refunding Bonds 260,659 36.06

Public Imps. 179,711 24.86

School Imps. 148,460 20.54

Refunding Notes 41,219 5.7

Water Utility Imps. 26,034 3.6

Sewer Imps. 10,988 1.52

Univ. & College Imps. 10,361 1.43

Recreational Fac. Imps. 10,192 1.41

Crossover Refunding 5,950 0.82

Current Refunding 4,312 0.6 Source: Parikh-He Municipal Bond Project

638 Boston College Law Review [Vol. 58:599

TOP 10 FUNDING SOURCES FOR REPAYMENT OF BOND INTEREST AND PRINCIPAL (2004 – 2014)

Number of Bonds Percent Ad Valorem Property Tax 699,147 96.56 Special Assessment 5,793 0.8 Miscellaneous Revenue 3,832 0.53 General Fund 2,905 0.4 Miscellaneous Taxes 2,690 0.37 Water Revenue 2,292 0.32 Sales Tax Revenue 1,618 0.22 Tax Incrmt./Allctn.Rev. 1,215 0.17 Sewer Revenue 1,190 0.16 New Jobs Training 537 0.07 Source: Parikh-He Municipal Bond Project

Reproduced with permission of copyright owner. Further

reproduction prohibited without permission.