ÃÍãÏ ÃÈæ ÒäØ
08-17-2012, 01:12 PM
Freedom to Fail The Keystone of American Federalism
Paul E. Peterson & Daniel Nadler
INTRODUCTION
One can hardly imagine how much [the] division of sovereignty
contributes to the well-being of each of the States which compose
the Union. In these small communities . . . all public authority . . .
[is] turned towards internal improvements. . . . [T]he ambition of
power yields to the less refined and less dangerous desire for wellbeing.
–Alexis de Tocqueville1
It is one of the happy incidents of the federal system that a single
courageous state may, if its citizens choose, serve as a laboratory;
and try novel social and economic experiments without risk to the
rest of the country.
–Justice Louis Brandeis2
It is, of course, no longer politically correct to characterize
anything American as exceptional. In days gone by, descendants of
the Pilgrim faithful spoke easily of their country as a “city upon a
hill,” a “New Jerusalem”3 whose hallowed light shone as a beacon for
all nations to see. It was not difficult for nineteenth-century
Americans to imagine a national destiny that spread “from sea to
shining sea.” Even in the mid-twentieth century, school children
learned to sing of a “sweet land of liberty” made beautiful by its
“purple mountains,” “spacious skies,” and “amber waves of grain.”
Most felt that the United States had been called to end—or at least
† Henry Lee Shattuck Professor of Government, Harvard University.
†† PhD Candidate, Department of Government, Harvard University.
This paper has been prepared for the Understanding Education in the United States: Its
Legal and Social Implications Symposium held at the University of Chicago Law School on
June 17 and 18, 2011. The authors would like to thank Romain Zamour, Yale Law School, JD
Class of 2013, for his research assistance.
1 Alexis de Tocqueville, Democracy in America 141 (Barnes & Noble 2003) (Francis
Bowen, ed) (Henry Reeve, trans).
2 New State Ice Co v Liebmann, 285 US 262, 311 (1932) (Brandeis dissenting).
3 See Cecelia Tichi, The Puritan Historians and Their New Jerusalem, 6 Early Am Lit 143,
143–44 (Fall 1971) (discussing Puritan historians’ use of biblical metaphors).
254 The University of Chicago Law Review [79:253
contain—tyrannies of unimaginable villainy in Nazi Germany, the
Soviet Union, and Maoist China.4 When Americans looked at their
nation, they saw an exceptional land upon which God had shed his
grace.
In the aftermath of World War II, university scholars joined in a
secularized version of the hymn.5 They marveled at a pluralist
America able to hold its political leadership accountable while
avoiding mass uprising that could translate into totalitarian
tyrannies.6 Such talk now seems antiquated, even self-indulgent. For
many today, the United States is better understood as just another
society at the advanced stage of capitalism.7 American and European
problems and politics are converging. If any country is exceptional, it
is China, or one of the four Asian Tigers, or perhaps India or Brazil.
Or, to state the situation in its most undeniable terms: Every country
is exceptional. Each has its own distinct geographical location, origin,
history, social composition, and political institutions. The United
States is no more exceptional than Canada, or Mexico, or what have
you.
I. THE EXCEPTIONAL AMERICAN FEDERAL SYSTEM
Still, it is worth treating as exceptional the country’s federal
system, with its unique separation of authority between national and
state governments. According to a recent count, only 25 of the
world’s 193 countries are federal systems.8 And most of those 25 federal
nations circumscribe the authority exercised by lower tiers of
government in important ways. In some, the heads of lower offices
4 See, for example, Daniel Bell, Interpretations of American Politics, in Daniel Bell, ed,
The Radical Right 39, 50 (Doubleday 1963). See also Samuel P. Huntington, American Politics:
The Promise of Disharmony 240–45 (Belknap 1981).
5 See Louis Hartz, The Liberal Tradition in America 4–6 (Harcourt 1955); Huntington,
American Politics at 27–30 (cited in note 4); Sven H. Steinmo, American Exceptionalism
Reconsidered: Culture or Institutions?, in Lawrence C. Dodd and Calvin Jillson, eds, The
Dynamics of American Politics: Approaches and Interpretations 106, 117–24 (Westview 1994).
6 Bell, Interpretations of American Politics at 58–59 (cited in note 4); William
Kornhauser, The Politics of Mass Society 121–23 (Free Press 1959).
7 See, for example, Mancur Olson, The Rise and Decline of Nations: Economic Growth,
Stagflation, and Social Rigidities 92–94 (Yale 1982); Harold L. Wilensky, The Welfare State and
Equality: Structural and Ideological Roots of Public Expenditures 52 (California 1975); Phillips
Cutright, Political Structure, Economic Development, and National Social Security Programs,
70 Am J Sociology 537, 549 (1965). But see Harold L. Wilensky and Charles N. Lebeaux,
Industrial Society and Social Welfare: The Impact of Industrialization on the Supply and
Organization of Social Welfare Services in the United States 41 (Free Press 1958).
8 Forum of Federations, Federalism by Country (2007), online at www.forumfed.org/en
/federalism/by_country/index.php (visited Oct 6, 2011). The number is twenty-three, according
to one count given in Jonathan A. Rodden, Hamilton’s Paradox: The Promise and Peril of
Fiscal Federalism 39 (Cambridge 2006).
2012] Freedom to Fail 255
hold office at the pleasure of the central government.9 In others, the
lower tiers are heavily dependent on the central government for
revenue.10 Except in Canada and Switzerland, state debts in all
federal systems in the industrialized countries of the world are
implicitly or explicitly guaranteed by the federal government.11
The design of the US federal system owes as much or more to
historical circumstances as to theoretical intentions. When writing
the Constitution, those gathered in Philadelphia necessarily allowed
for autonomous action by state governments for the very practical
reason that no other form of government could have won ratification
by the supermajority of states required before the founding
document could take effect. Unless the national government’s
powers were limited and states continued to exercise considerable
power on their own, the citizenry, more fond of their former colonial
governments than the new national entity, would not have agreed to
important limits the Constitution did impose upon the states, such as
restrictions on their abilities to declare war, coin money, and
regulate interstate commerce. The cultural differences between the
slaveholding South and an increasingly antislavery North could be
contained only if each region was allowed to organize its own
domestic affairs. But if the US federal system was initiated to solve a
very practical problem, it gradually became an institutional form so
appropriate and effective that it persisted into the twenty-first
century even after the Civil War had been fought, slaves had been
freed, and a much more powerful federal government had been
established.
That exceptional federal system, best characterized as
competitive federalism, can be sustained only if the lower tiers of
government are held accountable to the marketplace—most
specifically, to the market for government bonds. Unless lower tiers
are subject to independent movements in the interest rates on their
bonds, and unless lower tiers remain at risk of default, or something
tantamount to default, the central government cannot afford to grant
9 See, for example, Republic of India Public Administration Country Profile 8 (United
Nations Division for Public Administration and Development Management Jan 2006), online
at http://unpan1.un.org/intradoc/groups/public/documents/un/unpan023311.pdf (visited Oct 6, 2011).
10 See, for example, Dele Olowu, Property Taxation and Democratic Decentralization in
Developing Countries *19 (Institute of Development Studies Seminar Paper, 2002), online at
http://www2.ids.ac.uk/gdr/cfs/pdfs/Olowu2.pdf (visited Nov 6, 2011).
11 Rodden, Hamilton’s Paradox at 9, 31, 93 (cited in note 8). For a discussion of default
risk within the Canadian federal system, see Stuart Landon and Constance E. Smith,
Government Debt Spillovers and Creditworthiness in a Federation, 33 Can J Econ 634, 636,
653–54 (2000). The argument of our paper does not turn on whether the United States is
exceptionally different from Canada and Switzerland.
256 The University of Chicago Law Review [79:253
wide discretion to state or local governments. For more than two
centuries, the US federal system has survived multiple economic and
political crises, but never has the autonomy of the lower tier of
government been circumscribed to such an extent that state and
municipal bonds do not have their own, independent standing in the
marketplace. Yet a striking new political development—the granting
of collective bargaining rights to those who work for state and local
governments—has posed a dramatically new challenge to the
viability of the American federal system as we have known it. Just
how that has occurred, as well as its potential consequences for the
country’s political institutions, is the topic this paper explores.
II. COMPETITIVE FEDERALISM
Historically, competitive federalism helped to generate the
extraordinary growth of the world’s largest economic power.12 Over
the decades, states and localities designed and maintained canals,
railroads, highways, sewage systems, schools, parks, and systems of
public safety. As Lord James Bryce wrote nearly a century ago,
[I]t is the business of a local authority to mend the roads, to
clean out the village well or provide a new pump, to see that
there is a place where straying beasts may be kept till the owner
reclaims them, to fix the number of cattle each villager may turn
out on the common pasture, to give each his share of timber cut
in the common woodland.13
As compared to the federal government, state and local
governments are more sensitive to political market forces, making
them better equipped to design and administer those types of
programs. Unless local government supplies public services to meet
the needs of local businesses and residents, citizens may “vote with
their feet” and migrate to a locality better attuned to their needs.
Since 12 percent or more of the population changes its residence
each year,14 the effects of policy choices on property values can be
quickly felt.
12 The discussion in this Section draws upon the theoretical statement developed in Paul
E. Peterson, City Limits 68–72 (Chicago 1981). See also Charles Tiebout, A Pure Theory of
Local Expenditures, 64 J Polit Econ 416, 422 (1956); Wallace E. Oates, Fiscal Federalism 240–41
(Harcourt Brace Jovanovich 1972); Paul E. Peterson, The Price of Federalism 18–19
(Brookings 1995). The contributions of local government to rapid economic growth in China
are explored in Gabriella Montinola, Yingyi Qian, and Barry R. Weingast, Federalism, Chinese
Style: The Political Basis for Economic Success in China, 48 World Polit 50, 67–78 (1995).
13 James Bryce, 1 Modern Democracies 132 (Macmillan 1921).
14 See US Census Bureau, Statistical Abstract of the United States: 2012 37 table 30, online
at http://www.census.gov/prod/2011pubs/12statab/pop.pdf (visited Oct 7, 2011).
2012] Freedom to Fail 257
Business and residential choices are influenced by factors other
than the quality of local public services, of course. Businesses want to
be close to both their sources of supply and the markets for their
products. Individual and family residential choices are influenced by
family ties, employment opportunities, and the quality of the natural
environment. But the quality of publicly provided infrastructure also
affects, on the margins, the choices businesses and households
make.15
Since small changes in supply or demand can have a significant
effect on price, residents of a community, eager to protect their
property values, can be expected to pressure government officials to
employ public resources efficiently in order to meet local
expectations and facilitate economic development. Poor policy
decisions can have rapid and lasting effects on a municipality’s
property values and corresponding tax income.16 Therefore, it is
reasonable to expect most state and local governments to be
relatively competent at designing and implementing developmental
policies. Admittedly, lower-tier officials in a system of competitive
federalism may exhibit “narrowness of mind and the spirit of
parsimony,” as Lord Bryce was the first to admit, but if it were
otherwise, “there would be less of that shrewdness which the practice
of local government forms.”17
State and local government can also facilitate the gathering of
information regarding the most efficient way of organizing public
services. Each state or city is a laboratory where experiments are
tried. If the experiment is successful, other governments will copy it.
If the experiment fails, the idea will soon be abandoned. In addition,
“states and localities pay close attention to the wages and salaries
paid to employees in adjacent communities” and will feel pressure to
bring them into line with those of their neighbors.18 So valuable is the
role played by lower tiers of government within the federal system
that, despite the growth in the role of the federal government, more
than 40 percent of all government spending for domestic purposes
was, as late as 2008, paid for out of revenues raised by state and local
governments from their own sources.19 The lower tiers are also the
predominant public-sector employer. No less than 87 percent of all
15 See Peterson, Price of Federalism at 18–19 (cited in note 12).
16 See id at 19.
17 James Bryce, 1 Modern Democracies at 132–33 (cited in note 13).
18 Peterson, Price of Federalism at 19 (cited in note 12). See also Tiebout, A Pure Theory
at 421–22 (cited in note 12).
19 See Figure 3.
258 The University of Chicago Law Review [79:253
nonmilitary public-sector employees work for either the state or
local government.20
In a system of competitive federalism, state and local
governments resist taking responsibility for large-scale redistributive
programs.21 If states and localities attempt in a serious way to tax the
rich and give to the poor, the rich will depart while the poor will be
attracted. If the rich leave and the poor migrate into the state, tax
revenues will plummet while expenditures escalate. Any debt
acquired by state and local governments must be borrowed from
investors; if a state borrows too much money, state bond ratings fall
and, unless the fiscal situation of the state is corrected, the state will
default on its debts.
III. SUPREME COURT JURISPRUDENCE
If a state defaults, it may not be sued without its consent. That
state sovereignty implies immunity from private lawsuits compelling
payment of debt was established in the early years of the Republic.
When the Supreme Court, in Chisholm v Georgia,22 ruled that the
State of Georgia had to pay a citizen of South Carolina a debt it had
incurred,23 Congress passed the Eleventh Amendment to the
Constitution, reversing that decision and “ma[king] it very difficult
[subsequently] for creditors to force states to repay debts.”24 Early
jurisprudence also established that a state’s own citizens could not
file a suit in federal court to secure repayment of debt25 and that a
foreign nation could not successfully compel a state to pay its debt.26
A state is not immune from a suit filed by a sister state or by the
federal government, but neither entity is likely to be a state
bondholder.27 Citizens within a state can file a suit within a state’s
own courts, but state courts have historically not had much success in
compelling other branches of government to honor their debts so
that, as a result, citizens have been “unable to collect on the bonds.”28
20 See Figure 2. See also US Census Bureau, Statistical Abstract: 2012 at 300 table 461
(cited in note 14).
21 See Peterson, Price of Federalism at 29–30, 70 table 3-3, 71 table 3-4 (cited in note 12).
22 2 US (2 Dall) 419 (1793).
23 Id at 453.
24 William B. English, Understanding the Costs of Sovereign Default: American State
Debts in the 1840’s, 86 Am Econ Rev 259, 260 (1996).
25 Hans v Louisiana, 134 US 1, 14–15, 21 (1890).
26 See Monaco v Mississippi, 292 US 313, 330 (1934).
27 Consider English, Understanding the Costs at 260–61 (cited in note 24), citing South Dakota
v North Carolina, 192 US 286 (1904) and United States v North Carolina, 136 US 211 (1980).
28 English, Understanding the Costs at 261 (cited in note 24).
2012] Freedom to Fail 259
One might think that ancient decisions dating back to the
earliest days of the Republic are no longer pertinent, but despite the
array of recent civil rights litigation against states in recent decades,
the original conception of the United States as a federal union in
which sovereignty is enjoyed by both the federal and state
governments has remained altogether relevant for contemporary
jurisprudence. In US Term Limits, Inc v Thornton,29 Justice Anthony
Kennedy, in a concurring opinion, characterized American federalism
in words little different from those James Madison might have used:
Federalism was our Nation’s own discovery. The Framers split
the atom of sovereignty. It was the genius of their idea that our
citizens would have two political capacities, one state and one
federal, each protected from incursion by the other. The
resulting Constitution created a legal system unprecedented in
form and design, establishing two orders of government, each
with its own direct relationship, its own privity, its own set of
mutual rights and obligations to the people who sustain it and
are governed by it.30
Nor can the federal government order a state to compensate its
creditors. The Rehnquist Court invalidated federal laws said to
violate state autonomy by “commandeering” the states. In New York
v United States,31 the majority held that Congress may not simply
“‘commandeer’ state governments into the service of federal
regulatory purposes.”32 Printz v United States33 applied this reasoning
to executive officers as well, holding invalid provisions of the Brady
Handgun Violence Prevention Act34 that required state and local law
enforcement officers to conduct background checks on prospective
handgun purchasers.35 Writing for the majority, Justice Antonin
Scalia concluded: “By forcing state governments to absorb the
financial burden of implementing a federal regulatory program,
Members of Congress can take credit for ‘solving’ problems without
having to ask their constituents to pay for the solutions with higher
federal taxes.”36
29 514 US 779 (1995).
30 Id at 838 (Kennedy concurring).
31 505 US 144 (1992).
32 Id at 175 (holding that Congress does not have the authority to force state
governments to take title to waste under the Tenth Amendment).
33 521 US 898 (1997).
34 Pub L No 103-159, 107 Stat 1536 (1993).
35 Printz, 521 US at 932–33.
36 Id at 930.
260 The University of Chicago Law Review [79:253
With the passage of the Fourteenth Amendment, and the
application of its Due Process and Equal Protection Clauses to the
states, state sovereignty was eroded by a wide variety of civil rights
lawsuits that were effectively prosecuted in both state and federal
courts.37 But Fourteenth Amendment suits generally have been
viewed as constituting exceptions to state sovereign immunity. In
Alden v Maine,38 the Court reaffirmed the states’ immunity to
lawsuits filed in state courts.39 Justice Kennedy rooted the decision in
“the Constitution’s structure, and its history,” saying that “sovereign
immunity derives not from the Eleventh Amendment but from the
structure of the original Constitution itself.”40 However, Kennedy
also said that state sovereign immunity does not extend to suits
brought by the federal government itself and those pursuant to
enforcement of the Equal Protection or Due Process Clauses of the
Fourteenth Amendment.41
Future attempts to limit state sovereignty can be expected to
exploit Fourteenth Amendment exemptions from the doctrine of
state sovereignty. Those who seek to compel states to honor state
pension and health care policies and collective bargaining
agreements can be expected to invoke equal protection and due
process arguments. Bondholders will argue that defaults deny them
property without due process of law. But it is doubtful that such suits
could be successfully pursued in the absence of federal legislation
requiring states to honor implied contracts with bondholders,
pensioners, or public employees.42 In other words, the jurisprudence
that allows states to claim a sovereign status within the federal
system seems as vibrant today as it has ever been. While the
individual constitutions of many states may be interpreted as
granting permission for lawsuits by bondholders, pensioners, or
those protected by collective bargaining agreements,43 states—as
sovereign entities—appear to enjoy today the same legal
prerogatives vis-à-vis bondholders and other creditors as states that
have defaulted in the past, if they so choose.44
37 See, for example, Fitzpatrick v Bitzer, 427 US 445, 456 (1976). Courts also got around
limitations on sovereign immunity by allowing suits to go forward when plaintiffs sued state
entities rather than the state itself. For examples of the courts applying this limitation, see
Brown v Board of Education of Topeka, 347 US 483, 493 (1954); Ex parte Young, 209 US 123,
165 (1908).
38 527 US 706 (1999).
39 Id at 712.
40 Id at 713, 728.
41 Id at 755–56.
42 See, for example, United States v Sherwood, 312 US 584, 590–92 (1941).
43 See, for example, Ill Const Art 13, § 5.
44 See notes 82–91 and accompanying text.
2012] Freedom to Fail 261
IV. COLLECTIVE BARGAINING IN THE PUBLIC SECTOR
Since the beginning of the Republic, states have managed their
fiscal affairs so well that only in a few instances have they defaulted
on their debts. But in the twenty-first century, the risk of default by
large and economically significant states has increased dramatically.
Among the most important contributors to the altered situation has
been the rise of collective bargaining within the public sector. In this
regard, events within the field of education are especially instructive,
as school personnel are the largest segment of the state and local
workforce and education costs constitute approximately one-third of
all state and local expenditures paid for out of locally generated
revenues.45
Public-sector collective bargaining was largely unknown prior to
the 1960s. Even Franklin D. Roosevelt, the most significant
presidential ally the labor movement has ever enjoyed, rejected
public-sector bargaining within the federal government: “All
Government employees should realize that the process of collective
bargaining, as usually understood, cannot be transplanted into the
public service. . . . The employer is the whole people, who speak by
means of laws enacted by their representatives in Congress.”46
George Meany, the head of the American Federation of Labor, did
not disagree. As late as the 1950s, he plainly stated, “It is impossible
to bargain collectively with the Government.”47 Other organizations
that represented government employees took the position that
collective bargaining “was demeaning for civil service professionals.”48
The National Association of Education (NEA), by far the largest of all
teacher organizations, was firmly opposed to the idea.49
Collective bargaining was introduced into the public sector in
part because the policy fit the political needs of the Democratic
Party, which had been closely affiliated with the labor movement
45 In 1990 the percentage expended by state and local governments from their own
resources (that is, excluding federal grants) for elementary, secondary, and higher education
was 34.1 percent. See US Census Bureau, Statistical Abstract: 2012 at 273 table 435, 300 table 462
(cited in note 14).
46 Letter from President Franklin D. Roosevelt to Luther C. Steward, President of the
National Federation of Federal Employees (Aug 16, 1937) (American Presidency Project),
online at http://www.presidency.ucsb.edu/ws/?pid=15445#axzz1bRZWSTlq (visited Oct 21, 2011).
47 George Meany, Meany Looks into Labor’s Future, NY Times Mag 11, 38 (Dec 4, 1955).
48 Paul E. Peterson, Saving Schools: From Horace Mann to Virtual Learning 106
(Belknap 2010).
49 Martin Raymond West IV, Politics, Public-Sector Unionism, and Education Policy:
Explanations and Evaluations *38 (unpublished PhD dissertation, Harvard University, 2006)
(on file with authors).
262 The University of Chicago Law Review [79:253
since the 1930s.50 Since the 1950s, private-sector unionization has
been on the decline, slipping from roughly one-third to less than
10 percent of the nongovernmental workforce.51 Mobilizing new
recruits to the Democratic coalition became critical, and nothing was
more appealing than reaching out to the growing segment of the
workforce employed by local, state, and federal governments.
Previously, public-sector workers had not shown any particular
partisan loyalty other than to the machine that hired them.52 The
many white-collar professionals working for government were, if
anything, more inclined to the Republican side of the aisle.53 But farsighted
union leaders and key Democratic members of Congress,
perceiving an opportunity, began to campaign for collective
bargaining rights for public-sector workers.54 Dwight Eisenhower, a
Republican, stoutly resisted all congressional efforts to pass such
legislation into federal law, but his Democratic successor, John
Kennedy, promised he would take action if elected President.55 Since
the close balance of power on Capitol Hill precluded passage of
collective bargaining legislation, the President signed an executive
order giving federal employees the right to bargain collectively.56
That executive order, in conjunction with the success of New
York City teacher unions in obtaining collective bargaining rights,
initiated a decisive transformation of the American public sector.
The number of teacher strikes increased from 9 to 107 in just three
years, between 1964 and 1967.57 Faced with the prospect of school
shutdowns and masses of teachers picketing outside once uneventful
classrooms, school boards gave in to public pressure to settle strikes
quickly and return children to school. Affiliates of the American
Federation of Teachers (AFT) won recognition rights in many large
cities, including Boston, Chicago, Cleveland, and Philadelphia.58 “For
the first time . . . since 1918, the AFT threatened to surpass the
50 For labor-friendly New Deal legislation, see National Industrial Recovery Act of 1933,
Pub L No 73-10, 48 Stat 31, codified at 15 USC § 701–10, terminated by Executive Orders 7252
(Dec 21, 1935) and 7323 (March 26, 1936); National Labor Relations Act, 29 USC § 151 et seq.
51 See US Bureau of Labor Statistics, Press Release, Union Members—2010 1 (Jan 21,
2011), online at http://www.bls.gov/news.release/union2.nr0.htm (visited Oct 21, 2011).
52 See West, Politics at *41 (cited in note 49).
53 See Clem Brooks and Jeff Manza, Class Politics and Political Change in the United
States, 1952–1992, 76 Social Forces 379, 393 figure 1 (1997).
54 See West, Politics at *55 (cited in note 49).
55 See id at *56–58.
56 Executive Order 10988, 27 Fed Reg 551 (1962). See also West, Politics at *64 (cited in
note 49).
57 Peterson, Saving Schools at 113 (cited in note 48).
58 See id.
2012] Freedom to Fail 263
NEA,” one historian has noted.59 That changed when the NEA
dropped its principled opposition to collective bargaining as it saw its
membership ranks rapidly defect to the AFT. Eventually both
organizations prospered, with “NEA membership climbing from
700,000 in 1960 to 3.2 million in 2007, while the smaller AFT grew
from under 60,000 to 1.3 million over the same period.”60 Scarcely
known in education before 1960, collective bargaining achieved
predominance in most states outside the South.61
In present times, collective bargaining is so pervasive within the
public sector that few remember Franklin Roosevelt’s objections to
such a practice. There remains a handful of critics who argue that
collective bargaining subverts the democratic relationship between
government and citizens by privileging a particular set of interests—
those of government employees. Notably, such critics draw a
distinction between collective bargaining in the public and private
sector. In the private sector, collective bargaining is often
appropriate, such critics argue, because workers may need to bargain
collectively in order to prevent a profit-maximizing management
from abusing its superior bargaining position vis-à-vis individual
employees. When resolute unions bargain with a management
indifferent to all but its bottom line, each protects its own vital
interests in the collective bargaining process. But within the public
sector, such countervailing power cannot be assumed. The
“management” in the public sector is made up of elected officials,
such as school board members, to whom unions contribute heavily
during the election process. Also, school employees participate more
frequently than others in “school elections, which are often lowvisibility,
non-partisan affairs that engage the attention of only the
most interested parties.”62 Campaign contributions and coordinated
voting blocs give employees special influence over the very school
board with which they negotiate. Though not quite self-dealing,
teachers’ unions are certainly not bargaining with hostile
59 Marjorie Murphy, Blackboard Unions: The AFT and the NEA, 1900–1980 220 (Cornell
1990).
60 See Peterson, Saving Schools at 113 (cited in note 48).
61 See Richard C. Kearney and David G. Carnevale, Labor Relations in the Public
Sector 38, 60–61 table 3.2 (Marcel Dekker 2001); Randall W. Eberts, Teachers Unions and
Student Performance: Help or Hindrance?, 17 Future Children 175, 178 (2007). For further
research on the growth of teachers’ unions since 1950, see Joseph E. Slater, Public Workers:
Government Employee Unions, the Law, and the State, 1900–1962 193 (Cornell 2004). Hanna
Skandera and Richard Sousa, School Figures: The Data behind the Debate 106–08 (Hoover
2003). AFT membership includes university faculty, paraprofessionals, and other school
employees.
62 Peterson, Saving Schools at 114 (cited in note 48).
264 The University of Chicago Law Review [79:253
management representing interests in opposition to those of the
employees.63
Union political power has been expanded by collective
bargaining agreements in other ways as well. Contracts in many
districts require an amount equivalent to union dues be deducted
from employee paychecks, the use of which is given to the discretion
of the employee’s union. These deductions include fees that, unless
specifically objected to by a member, may be used for political
purposes.64 With such resources, teacher unions have become among
the most influential groups in state politics: in 1985 “teachers’
organizations” were identified as the most influential interest group
in state politics; in 2002 they were found to be second only to
business groups. In both surveys, they outranked such powerful
groups as utility companies, insurance companies, hospitals, trial
lawyers, manufacturers, and those representing local governments
more generally.65
Assessing the consequences of collective bargaining within the
public sector is a controversial matter. But we do know that since the
mid-1960s, per-pupil expenditures on elementary and secondary
education have tripled in real-dollar terms—from less than $4,000
per pupil (in 2006 dollars) to nearly $12,000 in 2008.66 Much of the
increment is to be explained by the growth in the number of public
school employees both because the number of pupils per teacher fell
by one-third (from twenty-five to sixteen) and because many more
nonteaching professionals were hired to provide ancillary services
and to help manage an increasingly complex system.67 In 1960, school
districts employed 6 professionals for every 100 students; by 2005
they were employing more than 12 for that same number.
Nonprofessional hiring rose at a similar rate—from less than 2 per
100 students in 1960 to nearly 4 in 2005.68 Most of the rising cost of
education was borne by state and local governments, as the federal
contribution did not rise much above 10 percent of the total.69
63 Terry M. Moe, The Union Label on the Ballot Box: How School Employees Help
Choose Their Bosses, 6 Educ Next 58, 60 (Summer 2006).
64 Peterson, Saving Schools at 114–15 (cited in note 48).
65 Clive S. Thomas and Ronald J. Hrebenar, Interest Groups in the States, in Virginia
Gray and Russell L. Hanson, eds, Politics in the American States: A Comparative Analysis 100,
119 table 4-1 (CQ 8th ed 2004).
66 See Thomas D. Snyder, Sally A. Dillow, and Charlene M. Hoffman, Digest of
Education Statistics, 2009 100 table 64, 260 table 181 (Department of Education 2010).
67 See Thomas D. Snyder and Sally A. Dillow, Digest of Education Statistics, 2008 117
table 80 (Department of Education 2009).
68 See id.
69 Peterson, Saving Schools at 153 (cited in note 48).
2012] Freedom to Fail 265
During this period of time, teacher salaries have kept pace with
overall wage and salary increases nationwide.70 In addition, teachers
and other public-sector employees have been guaranteed steep
increases in pensions, health care, and other nonsalary benefits, as
elected officials choose to reach collective bargaining settlements by
rewarding workers with promises of future benefits rather than
immediate salary compensation.71 In most cases, the cost of these
benefits was postponed into a future well beyond current election
cycles.72 For years the growing imbalance between rising costs and
increasing liabilities, on the one side, and fiscal resources, on the
other, was ignored, except by the Cassandras of the policy world.73
But with the financial crisis of 2008, the possibility of state and
municipal defaults shifted from the theoretical to the plausible.74
V. STATE FISCAL CRISES
The lower tiers of the US government are facing a
contemporary fiscal crisis unprecedented since the days of the Great
Depression. While some resource-rich, less populous states—Alaska,
Montana and North Dakota, for example—continue to run balanced
budgets,75 most states are confronting large deficits. New York’s
deficit for the fiscal year 2012 was estimated in early 2011 to be
18 percent of the previous year’s budget, California’s to be 29 percent,
Texas’s to be 32 percent, and New Jersey’s to be no less than 38
percent.76 The gap in official state budgets was estimated to be at
$121 billion, or 19 percent of the budget in the forty-six states
running deficits.77 The size of these projected deficits may have
attenuated as state economies have recovered, but they would be
70 See id at 133.
71 See, for example, Letter from Daniel W. Hancock, Chairman of the Little Hoover
Commission, to Edmund G. Brown, Governor of California, and Members of the California
Legislature (Feb 24, 2011) (“Hancock Letter”), in Public Pensions for Retirement Security First
Page (Little Hoover Commission Feb 2011), online at http://www.lhc.ca.gov/studies/204
/report204.html (visited Oct 22, 2011).
72 See David Stella and Keith Bozarth, Pension Sustainability—The Wisconsin Example,
47 Benefits & Compensation Dig 32, 33 (Feb 2010).
73 See, for example, Hancock Letter at First Page (cited in note 71); Roger Lowenstein,
The End of Pensions, NY Times Mag (Oct 30, 2005), online at http://www.nytimes.com/2005/10/30
/magazine/30pensions.html?pagewanted=print (visited Oct 22, 2011).
74 See Kevin Hassett, California Leads Nation to Bond Default Abyss, Tulsa World A12
(June 2, 2009).
75 Daniel J. Nadler and Sounman Hong, Political and Institutional Determinants of Tax-
Exempt Bond Yields *3 (Harvard Kennedy School Report No 11-04), online at
http://www.hks.harvard.edu/pepg/PDF/Papers/PEPG_11-04_Nadler_Hong.pdf (visited Oct 10,
2011).
76 See id at *15 table 1.
77 Id at *3.
266 The University of Chicago Law Review [79:253
considerably larger if they were to include the revenues necessary to
fully fund state pension and health care obligations.78
With the onset of the financial crisis, the bond market
immediately took note of the increased risk of sovereign state
defaults. In late 2008 investors demanded a higher premium for state
and local bonds over safer US Treasury securities, despite the
exemption from federal taxation of interest received on most state
and local bonds. Although all states were affected by the crisis, the
perceived risks of default varied considerably among the states.
“Between September and December of 2008, the premium that
investors demanded to hold California debt over US treasuries
jumped from 24 basis points to 271 basis points, a ten-fold increase.”79
(100 basis points equals one percent.) Before the crisis, the
difference in the premium paid in California and Texas was only 15
basis points. But by 2011 the gap between the two states had
increased to 84 basis points.80 Similar jumps in the cost of borrowing
occurred in a number of other states as well.81 Clearly, investors had
become increasingly sensitive to the variation in the risk of defaults
among the sovereign states.
VI. IMPACT OF COLLECTIVE BARGAINING ON DEFAULT RISK
State and municipal defaults are not unknown to American
federalism.82 Eight states defaulted or repudiated debt between 1841
and 1843 when a severe economic depression restricted state ability
to pay interest on debt that had been assumed primarily for the
purpose of constructing canals and railroads.83 The federal
government refused to assume responsibility despite efforts by both
defaulting states and foreign banks to persuade the federal
government to intervene.84 While four states eventually repaid all of
their debt, three made only a partial repayment, and one,
Mississippi, never did.85
78 Alan J. Auerbach, Long-Term Objectives for Government Debt *10–13 (Swedish Fiscal
Policy Council Conference on Fiscal Policy and Labour Market Reforms, Feb 2008), online at
http://elsa.berkeley.edu/~auerbach/long_term_objectives_govt_dept.pdf (visited Nov 4, 2011).
79 Nadler and Hong, Political and Institutional Determinants at *4 (cited in note 75).
80 See id.
81 See Figure 4.
82 In this paper we ignore defaults by municipalities and other lower-tier governments.
These units of government do not have status in the US Constitution, and their status varies
from one state to the next, depending on state law.
83 See English, Understanding the Costs at 261–65 (cited in note 24).
84 See Rodden, Hamilton’s Paradox at 59–60 (cited in note 8).
85 English, Understanding the Costs at 265 table 3 (cited in note 24).
2012] Freedom to Fail 267
The boom-and-bust economy of the 1870s and 1880s provoked
another ten defaults, and Arkansas was unable to cover its debts
during the 1930s depression.86 Bondholders were not the only
creditors that states ignored during hard times. During the Great
Depression, Chicago teachers were on several occasions paid in
“scrip,” because the City could not find the cash in hand to
compensate them.87 Years later the “scrip” was made good, but few
teachers themselves ever received payment in full, as they had used
their highly discounted “scrip” to pay monthly bills.88
In none of these crises was there much hope that the federal
government would come to the rescue of states at risk of default.
During the 1840s, some political leaders invoked the precedent
established when Congress, prompted by Alexander Hamilton,
assumed the Revolutionary War debts incurred by some of the
states.89 But others argued that the precedent did not hold in that
revolutionary war debts had been incurred on behalf of a common
cause, while the state debts incurred prior to the 1840s were for the
purpose of setting up banking and transportation systems designed
mainly for the benefit of the state itself. Neither of the national
political parties saw any advantage in coming to the rescue of a few
states at the expense of the remainder.90
Nor has the US government guaranteed state debts in any
subsequent crisis. As a result, each state is held accountable by the
bond market in ways that lower-tier governments in most other
countries are not. Consider, for example, the differing response of
the bond market to the state bonds issued in the United States and in
the German Federal Republic. Even though the constitution of the
German Federal Republic, adopted in the aftermath of World War II,
was explicitly modeled on that of the United States and assigned
major responsibilities to state governments, the German federal
government has asserted control over state finances and guarantees
state debts.91 For that reason, the spread between German federal
and state securities is less than the spread between such securities in
the United States. As is shown in Figure 3, the 2008 financial crisis
86 Andrew Ang and Francis A. Longstaff, Systemic Sovereign Credit Risk: Lessons from
the U.S. and Europe *6 (NBER Working Paper No 16982, Apr 2011), online at http://www.nber.org
/papers/w16982 (visited Oct 8, 2011).
87 See Paul E. Peterson, The Politics of School Reform, 1870–1940 179 (Chicago 1985).
88 See William J. Grimshaw, Union Rule in the Schools: Big-City Politics in
Transformation 64 (Lexington 1979).
89 See Rodden, Hamilton’s Paradox at 57, 60 (cited in note 8).
90 See id at 62–63.
91 See id at 91. See also id at 83–87 table 4.1 (reporting that the variance in the credit
ratings of lower-tier governments in Germany is much smaller than in the United States).
268 The University of Chicago Law Review [79:253
had an impact on the perceived default risk of the average German
state relative to that of the German Federal Republic. The average
yield spread between the two types of securities from 36 basis points
to 112 basis points between June 2008, the eve of the financial crisis,
and December 2008, when the crisis was at its peak. But during that
same period, the average yield spread for US states relative to
federal securities increased from 19 basis points to 129 basis points.
Clearly, the bond market perceived that the risk of default by a state
in Germany was attenuated by the guarantee supplied by the
German Federal Republic.
In the United States, investors were willing to accept lower
interest rates on state debt securities relative to US Treasuries due to
their federal-tax-exempt status. After the financial crisis, however,
the yield on state bonds rose above that for comparable federal
securities, as any tax advantages were overwhelmed by perceived
increased risk.92 Rates of return on state bonds before the financial
shock trailed those for Treasury securities because federal taxes need
not be paid on the returns from most state and municipal bonds. But
after the financial crisis, the spread between state and federal bonds
turned from negative to positive, as the relative risk from state
investments outweighed any tax advantages. Moreover, the yield
spread between state and federal bonds varied significantly from state
to state, indicating that the market perceived greater default risk in
certain states.93
Notably, investors’ perceptions of the risk of default were
correlated with the unionization rate of the public-sector workforce.94
As shown in Figure 5, the relationship between union membership
and default risk was noticeably weaker in June 2008, prior to the
financial crisis, than it was over the next six months. Figure 5’s
vertical axis shows the spread for federal securities and state bonds
that will mature in one year, while the horizontal axis shows the
unionization rate for the state’s public sector. “The relationship
between the two variables, modest in June 2008, becomes
pronounced by June 2009, as bondholders became highly sensitive to
a state’s perceived political capacit[ies] to take actions needed to
bring budget deficits under control.”95 The differences in the
steepness of the slopes taken by the regression lines in Figure 5
92 Nadler and Hong, Political and Institutional Determinants at *8 (cited in note 75).
93 See Figure 4.
94 See Nadler and Hong, Political and Institutional Determinants at *8 (cited in note 75).
95 Id.
2012] Freedom to Fail 269
describe the strengthening of the simple relationship between the
union share of the public-sector workforce and default risk.96
This relationship persists when other factors are controlled, as
has been shown by Daniel Nadler and Sounman Hong.97 In this
study, unbiased estimates of the impact of political variables on state
default risks are estimated with models that solve for the
endogenous relationship between credit and yield, and that also take
into account the economic factors that James Poterba and Kim
Rueben have shown to be associated with a state’s default risk:
change in its unemployment rate, Gross Domestic Product (GDP),
and deficit-to-GDP ratio.98 Nadler and Hong present estimates of the
impact of a range of state-level political variables—such as the union
share of the public-sector workforce and partisan representation in
the legislature—on state municipal bond yield spreads in the context
of the unexpected deficit shocks seen following the 2008 financial
crisis. In particular, they evaluate the impact of the union share of
the public-sector workforce and partisan representation in the
legislature using “separate models, because unionism and partisan
balance are highly correlated with one another, making it difficult,
with the small number of observations available, to identify the
independent impact of each within a single model.”99
Their results are reproduced here in Table 1. According to
Nadler and Hong, unexpected deficit shocks of the size that took
place in 2008 especially affect state yield spreads when certain
political conditions are present. As can be seen in Table 1, a
1 percent difference in union membership in a state is associated
with an additional 2.02 basis point change in state borrowing costs, if
the state has experienced a billion-dollar change in its unexpected
deficit shock. In other words, a twenty percentage-point difference in
the share of the public-sector workforce that is unionized (one
standard deviation) is associated with an additional increase in the
96 Unlike the simple relationship between union share of the public workforce and bond
yield spreads, the simple relationship between yield spreads and the partisan composition of
the state legislature does not increase in the wake of the fiscal crisis. The impact of this political
factor is detectable only after estimating the model presented in Table 1.
97 Nadler and Hong, Political and Institutional Determinants at *6–7 (cited in note 75).
98 See id at *6, 8. Since the relationship between decisions to issue state bonds and yield
spreads are endogenous, an unbiased estimate of the factors affecting yield spreads cannot be
obtained by an ordinary least-squares regression. However, if a change in the size of the deficit
is unexpected, the demand for credit changes regardless of the price of the bond, permitting
unbiased estimates. The model thus estimates the interaction between deficit shocks and the
key political variables included in the models. The model follows those used by James M.
Poterba and Kim Rueben, Fiscal News, State Budget Rules, and Tax-Exempt Bond Yields, 50 J
Urban Econ 537, 539–44 (2001).
99 See Nadler and Hong, Political and Institutional Determinants at *9 (cited in note 75).
270 The University of Chicago Law Review [79:253
level of state bond spreads of 40.4 basis points, for every billiondollar
change in unexpected deficit shock that a state experiences.100
Similarly, Nadler and Hong found that a one-percentage-point
increase in the Democratic share of a state legislature is associated
with an additional 3.02 basis point increase in state borrowing costs
for every billion-dollar change in a state’s unexpected deficit shock.
This suggests that an increase in the Democratic legislative
representation by twenty percentage points is associated with a
60.4 basis point increase in the state-to-federal bond yield spread in
the context of a billion-dollar deficit shock. “The cost to the state
taxpayer of a standard deviation shift in either variable is, roughly
speaking, about one half of one percent on a five-year security
note.”101 As Nadler and Hong argue, “That amount is non-trivial. In
Illinois, an increase in the yield spread of that magnitude on its debt
of $145.5 billion amounts to $727 million dollars in additional
interest costs annually.”102
However, one should not reify these two indicators of a state’s
political situation. Union share of the public-sector workforce and
partisan representation in the legislature are actually indicators of a
broader set of factors affecting a state’s risk of default.103 As Nadler
and Hong’s work makes clear, the unionization rate of the publicsector
workforce is correlated with factors such as whether a state
has a right-to-work law and whether the legislature has permitted
public-sector collective bargaining, both of which are correlated with
the magnitude of bond yield spreads. In addition, the percentage of
the legislature that is Democratic is highly correlated with the
percentage that is Democratic in each house of the legislature, which
also are correlated with bond yield spreads.104 (However, as Nadler
and Hong note, that data shows that the partisan affiliation of the
governor is not correlated with yield spreads, “suggesting that
governors have broader constituencies than do members of the
legislature.”)105
The two interval variables emphasized by Nadler and Hong—
public-sector unionization and political orientation of the legislature—
100 Id.
101 Id.
102 Id.
103 Variations in state expenditures on Medicaid, a “highly redistributive program that
might be considered a default risk factor, were not correlated with yield” spreads. Bondholders
apparently think that expenditures incurred outside of collective bargaining agreements are
more easily managed. See Nadler and Hong, Political and Institutional Determinants at *9 n 16
(cited in note 75).
104 Id at *9.
105 Id.
2012] Freedom to Fail 271
should be understood as useful proxies for a broader set of collective
bargaining and partisan factors that affect bond yields. Economic
factors—growth, change in the deficit-to-GDP ratio, and change in
the unemployment rate—are also strongly associated with the
substantial interstate variation in yield spreads that occurred in the
wake of the financial crisis. But in addition to the impact of these
economic factors, political realities are clearly also taken into
account by bondholders: Nadler and Hong found that, when both are
entered into the same equation, the political variables seem to be at
least as important in explaining post-crisis interstate variation in
yield spreads as are core economic indicators such as state-level
growth in GDP and changes in state unemployment rates.106
VII. DEFAULT RISKS AND COMPETITIVE FEDERALISM
A system of competitive federalism has long been extolled as a
permanent feature of American government.107 Both early and
modern Supreme Court jurisprudence has recognized states as
sovereign entities that exercise autonomous power—and incur
concomitant risks—within the sphere allocated to them by the
Constitution.108 States and localities play a major role in the raising of
revenue, the delivery of services, and the servicing of public debt.
Nothing is as quintessentially American as the dual sovereignty
granted both to states and to the federal government.
Yet in Hamilton’s Paradox, a recent study of federal systems,
Jonathon Rodden argues that attempts to sustain systems of
competitive federalism usually fail, attributing to Alexander
Hamilton a similar appreciation of competitive federalism’s
fragility.109 When sovereignty is divided, lower-tier governments are
tempted to run debts that place themselves at grave risk of default in
times of financial crisis.110 And central governments, both to
safeguard their international credit rating and to respond to internal
political pressures, cannot resist providing the assistance necessary to
safeguard bondholders and other creditors from loss.111 Central
governments do not offer a helping hand without at the same time
asserting their authority, however. If they rescue states and localities
they will feel more than entitled to take preventative measures
106 See id at *10–12.
107 See note 12.
108 See, for example, Hans v Louisiana, 134 US 1, 21 (1890); New York, 505 US at 175.
109 Rodden, Hamilton’s Paradox at 2 (cited in note 8).
110 See id at 8.
111 See id at 78–79.
272 The University of Chicago Law Review [79:253
designed to preclude future defaults.112 Irresponsibility at the state
and local level thus undermines the dual sovereignty essential for the
survival of competitive federalism. Celebrated in theory as an
efficient government of Herculean proportions, competitive
federalism is but a ten-pound weakling in practice. To prove his
claims, Rodden inquires into the functioning of three large and
important federal systems—Australia, Brazil, and Germany. Based
on his findings, he recommends that new states be constructed either
as unitary systems or designed in such a way as to give the central
government undisputed fiscal authority.
So it is of considerable interest that Rodden does not apply his
argument to the United States. That exceptional nation, Rodden
suggests, can still enjoy the benefits of a system of competitive
federalism, because the arrangement has been woven so deeply into
the fabric of the society that it cannot be torn asunder. In the 1840s,
the national government stood aside when multiple states defaulted,
and it has never intervened to help them out in the decades since.113
Other factors have contributed to a stable system of competitive
federalism as well: the size of government has been relatively small,
the lower tiers of government have been responsible for a fairly large
share of all domestic spending, grants from the federal government
have remained only a moderate proportion of total state and local
spending, and debts have made up a small percentage of most states’
GDP.
Yet within five years of the publication date of Rodden’s
insightful study, competitive federalism in the United States seems
more fragile than it has ever been. Many of the stabilizing factors are
gradually being whittled away. In recent years, the size of federal,
state, and local government has grown from less than 30 percent to
over 35 percent of GDP, the federal share of overall domestic
expenditures has been on the increase,114 intergovernmental grants
are making up a greater share of total lower-tier expenditures, and
state and national debt is escalating at an astounding rate. In the
spring of 2009, Congress, as part of the stimulus package, transferred
hundreds of billions of dollars to states and localities, and tens of
billions in additional aid were appropriated the following year.115
Though presented as legislation that would protect public-sector
112 See id at 271.
113 See Rodden, Hamilton’s Paradox at 62, 67 (cited in note 8).
114 See Figure 1.
115 American Recovery and Reinvestment Act of 2009, Pub L No 111-5, 123 Stat 115.
2012] Freedom to Fail 273
jobs,116 the monies were most valued by governors and mayors as
mechanisms for reducing fiscal deficits.117 Though not a federal
guarantee against default, the stimulus packages provided a dramatic
example of the way in which federal aid can ameliorate state and
local distress when states find themselves at risk of default.
That sovereign entities may be at risk of default in the coming
decades is well understood. It is not just Greece, Ireland, Portugal,
Spain, and Italy whose debt situations have become a matter of
urgent concern. Even the US government is at risk, as the net foreign
debt of the US central government, in the absence of corrective
measures, is projected within the next twenty years to rise from
about $5 trillion dollars in 2011 (about 33 percent of GDP) to $50
trillion, or more than 140 percent of GDP,118 a level “far above any
levels that could be considered sustainable.”119 Those numbers do not
include the sovereign debts of the fifty states of the union, a debt
that is currently about 7 percent of GDP.120 Nor do they take into
account the value of the unfunded liabilities faced by public-sector
pension plans, officially estimated at $438 billion by states
themselves but which could in fact be as high as $3 trillion dollars,
about 20 percent of GDP.121
Within the United States the sovereign state default crisis is for
some states—Illinois, California, and New Jersey, for example—
serious enough that Washington policy makers are currently
debating the policy and constitutional implications of three
alternatives: federal loans that would bail out states at risk of
default,122 bankruptcy procedures,123 and simple defaults of the kind
116 See The White House, Press Release, State Governments Expected to Credit Recovery
Act with Creating, Saving at Least 250,000 Education Jobs Nationwide (Oct 19, 2009), online at
http://www.whitehouse.gov/the-press-office/state-governments-expected-credit-recovery-actwith-
creating-saving-least-250000-ed (visited Oct 9, 2011).
117 See William Murphy, Stimulus Helps Shrink Deficit, Newsday A20 (Aug 6, 2009).
118 William R. Cline, Long-Term Fiscal Imbalances, US External Liabilities, and Future
Living Standards, in C. Fred Bergsten, ed, The Long-Term International Economic Position of
the United States 11, 24 table 2.3 (Peterson Institute for International Economics 2009).
119 C. Fred Bergsten, The Global Crisis and the International Economic Position of the
United States, in Bergstern, ed, Long-Term Economic Position (cited in note 118). See also The
Current Account Deficit and US Foreign Debt Hearing before the Senate Budget Committee,
110th Cong, 1st Sess 3 (2007) (testimony of C. Fred Bergsten).
120 See Steven Maguire, State and Local Debt: An Analysis *5 (CRS Mar 31, 2011), online
at http://www.nasbo.org/LinkClick.aspx?fileticket=4sLYo0HTYI8%3D&tabid=81 (visited Oct
23, 2011).
121 Andrew G. Biggs, The Market Value of Public Sector Pension Deficits 5 (American
Enterprise Institute for Public Policy Research, Apr 2010), online at http://www.aei.org/files
/2010/04/06/2010RPOno1g.pdf (visited Nov 5, 2011).
122 See, for example, Douglas Turner, Serious Budget Troubles Brewing in Many States,
Buffalo News A8 (Jan 10, 2011).
274 The University of Chicago Law Review [79:253
that occurred during the 1840s.124 Representative Patrick McHenry,
chairman of a subcommittee of the Committee on Oversight and
Government Reform says that “already state and municipal
governments are coming to Washington, hat-in-hand, expecting a
federal bailout.”125 Berkeley School of Law Dean Christopher Edley
has proposed that the federal government bail out states by lending
them federal money at low interest in the expectation that it will be
paid in due course.126 The Obama administration’s proposal to loan
monies to states to help them cover deficiencies in their state
unemployment insurance accounts sets a precedent for larger and
more consequential federal actions in the future.
Bankruptcy protection has been proposed by University of
Pennsylvania law professor David Skeel. In his view, the country
needs a federal bankruptcy law designed specifically for sovereign
debts that would “enable a state to restructure [its] obligations.”127
Such a law, he argues, would be constitutional as long as state
sovereignty were protected by giving states the option to invoke
bankruptcy procedures rather than mandating them to enter
bankruptcy court if they would otherwise default.128 Voluntary
participation in bankruptcy procedures would give states the
opportunity to restructure their obligations to employees,
pensioners, and bondholders, much as bankrupt corporations may
continue to operate while under the protection of federal bankruptcy
law. Not only would bankruptcy give states the opportunity “to
restructure obligations that are [otherwise] extremely difficult to
restructure,” but it would “ensure[] that most or all of a state’s
constituencies make sacrifices, not just one or two.”129 Jeb Bush and
Newt Gingrich have proposed a similar plan that would give states
123 See, for example, Jeb Bush and Newt Gingrich, Let States Declare Bankruptcy:
Reorganization Allowing Breaking of Union Contracts May Be the Best Way for Some,
Baltimore Sun A13 (Jan 31, 2011).
124 See, for example, Jeff Segal, Martin Hutchinson, and Rob Cox, California’s Only
Option, NY Times B2 (June 10, 2009).
125 State and Municipal Debt: The Coming Crisis? Hearing before the Subcommittee on
TARP, Financial Services and Bailouts of Public and Private Programs of the House Committee
on Oversight and Government Reform, 112th Cong, 1st Sess 1 (2011) (“State and Municipal
Debt Hearings”) (Rep Patrick McHenry).
126 Christopher Edley, Let Treasury Rescue the States, NY Times A25 (July 8, 2010).
127 David A. Skeel Jr, States of Bankruptcy, 79 U Chi L Rev *18 (forthcoming 2012),
online at http://ssrn.com/abstract=1907774 (visited Oct 9, 2011).
128 Id at *23–25.
129 State and Municipal Debt Hearings at 5 (testimony of David A. Skeel Jr). See also
Skeel, States of Bankruptcy at *19–20 (cited in note 127).
2012] Freedom to Fail 275
the opportunity to seek bankruptcy protection in the event of a
deficit crisis.130
Nicole Gelinas of the Manhattan Institute argues that “state
bankruptcy would create more problems than it would solve.”131 Most
states do not owe their debt through a single entity, making it
difficult for any single bankruptcy court to handle the extraordinary
complexities involved. For example, pension obligations are typically
borne by local governments as well as by the state, adding to the
number of participants in bankruptcy procedures.132
None of the three proposed options are attractive, but if the
state fiscal crisis becomes increasingly severe, as could happen if
projected deficits in pension and health care accounts materialize,
then the federal loan option may prove to be the most politically
palatable. Multiple state and municipal defaults would likely
provoke a nationwide political crisis and could affect the credit of the
US government, especially if its debt-to-GDP ratio continues to rise.
Passage of bankruptcy legislation could allow for a more managed
imposition of costs on the full range of creditors, including
bondholders, pensioners, and beneficiaries of collective bargaining
agreements, but bankruptcy could also affect US credit in world
markets and would create a legal nightmare, given the complexity of
state contractual arrangements with its creditors. By comparison,
federal loans provide an attractive option to those elected officials
aligned with public-sector unions, a constituency at risk in any
bankruptcy proceeding. Even if power in Washington is divided
between the two political parties, the fear of international
consequences could induce compromises that require substantial
federal contributions to states along the lines of the stimulus package
passed in 2009.
The current contemporary flirtation with default, coupled with
demands for a federal rescue, poses a threat to the system of
competitive federalism. The threat comes not so much from the
accumulation of debt as the obligations that have been incurred as
part of the collective bargaining process, many of which may be
enforceable in court. So it is probably not surprising that a state’s
default risk, as judged by the contemporary bond market, is related
both to the share of the public-sector workforce that is unionized and
to the percentage of the members of the state legislature affiliated
130 See Jeb Bush and Newt Gingrich, Better Off Bankrupt: States Should Have the Option
of Bankruptcy Protection to Deal with Their Budget Crises, LA Times A19 (Jan 27, 2011).
131 State and Municipal Debt Hearings at 1 (testimony of Nicole Gelinas).
132 See id at 1–2.
276 The University of Chicago Law Review [79:253
with the Democratic Party.133 If that party is in control of the federal
government, it can be expected to look favorably on requests that it
rescue states in need. In the midst of the latest crisis, Warren Buffett,
a prominent investor with a large stake in the state and municipal
bond market, expressed the hope that such federal action would be
forthcoming, conceding that “[t]he bond insurers . . . have extraordinary
liabilities,” but doubting that “the federal government
[would] turn away a state that is having extreme financial difficulties
when in effect it honored” the debts of corporate entities, including
General Motors.134 Later, in an interview with the congressional
Financial Crisis Inquiry Commission, he qualified that assessment,
saying, “I don’t know how I would rate [state bond default risks]
myself. . . . It’s a bet on how the federal government will act over
time.”135
Making a bet on the federal response to a state sovereign debt
crisis is beyond the scope of this paper. We claim only that the
introduction of collective bargaining has magnified the risk of state
sovereign defaults, complicated the resolution of deficit problems
that provoke such crises, heightened the likelihood of a federal
intervention if such crises materialize, and set the conditions for a
transformation of the country’s federal system. The costs of such
actions are greater than just the dollar numbers explicitly on the
bargaining table. Within the past decade a system of competitive
federalism that once enjoyed an exalted, even Olympian, standing in
American political culture has now been placed at risk.
133 This Section draws upon the findings and analysis presented in Nadler and Hong,
Political and Institutional Determinants at 7 (cited in note 75). Their analysis is based on data
from the twenty states for which daily yield spreads are publicly available. It extends a previous
analysis of the economic and legal determinants of state bankruptcy risks by Poterba and
Rueben, 50 J Urban Econ at 537 (cited in note 98).
134 Svea Herbst-Bayliss and Jonathan Stempel, Buffett: US Can Bail Out States, Insurers
Pained, Reuters (May 1, 2010), online at http://www.reuters.com/article/2010/05/01/berkshirebuffett-
ratings-idUSN0118355720100501 (visited Oct 9, 2011).
135 Ianthe Jeanne Dugan, Investors Looking Past Red Flags in Muni Market, Wall St J C1
(June 14, 2010).
2012] Freedom to Fail 277
TABLE 1. THE EFFECT OF UNEXPECTED DEFICIT SHOCKS,
POLITICAL, AND LABOR INSTITUTIONS ON CHANGES IN
STATE BOND YIELDS (MODEL 2).
(1) (2)
Δ Defshock 24.17 -11.02
(14.25) (11.51)
Δ Defshock × Union membership 2.02***
(0.61)
Δ Defshock × Dem. share in State Legislature 3.02***
(0.96)
Δ Unemployment rate 37.41 57.65*
(23.11) (27.48)
Δ Real GDP 131.5 -128.3
(177.0) (139.2)
Δ Deficit to GDP 8.598** 13.23***
(3.335) (4.267)
Constant -100.3*** -131.9***
(31.96) (40.42)
N 20 20
R2 0.702 0.723
Source: Nadler and Hong, Political and Institutional Determinants at *16 (cited in note 75).
Paul E. Peterson & Daniel Nadler
INTRODUCTION
One can hardly imagine how much [the] division of sovereignty
contributes to the well-being of each of the States which compose
the Union. In these small communities . . . all public authority . . .
[is] turned towards internal improvements. . . . [T]he ambition of
power yields to the less refined and less dangerous desire for wellbeing.
–Alexis de Tocqueville1
It is one of the happy incidents of the federal system that a single
courageous state may, if its citizens choose, serve as a laboratory;
and try novel social and economic experiments without risk to the
rest of the country.
–Justice Louis Brandeis2
It is, of course, no longer politically correct to characterize
anything American as exceptional. In days gone by, descendants of
the Pilgrim faithful spoke easily of their country as a “city upon a
hill,” a “New Jerusalem”3 whose hallowed light shone as a beacon for
all nations to see. It was not difficult for nineteenth-century
Americans to imagine a national destiny that spread “from sea to
shining sea.” Even in the mid-twentieth century, school children
learned to sing of a “sweet land of liberty” made beautiful by its
“purple mountains,” “spacious skies,” and “amber waves of grain.”
Most felt that the United States had been called to end—or at least
† Henry Lee Shattuck Professor of Government, Harvard University.
†† PhD Candidate, Department of Government, Harvard University.
This paper has been prepared for the Understanding Education in the United States: Its
Legal and Social Implications Symposium held at the University of Chicago Law School on
June 17 and 18, 2011. The authors would like to thank Romain Zamour, Yale Law School, JD
Class of 2013, for his research assistance.
1 Alexis de Tocqueville, Democracy in America 141 (Barnes & Noble 2003) (Francis
Bowen, ed) (Henry Reeve, trans).
2 New State Ice Co v Liebmann, 285 US 262, 311 (1932) (Brandeis dissenting).
3 See Cecelia Tichi, The Puritan Historians and Their New Jerusalem, 6 Early Am Lit 143,
143–44 (Fall 1971) (discussing Puritan historians’ use of biblical metaphors).
254 The University of Chicago Law Review [79:253
contain—tyrannies of unimaginable villainy in Nazi Germany, the
Soviet Union, and Maoist China.4 When Americans looked at their
nation, they saw an exceptional land upon which God had shed his
grace.
In the aftermath of World War II, university scholars joined in a
secularized version of the hymn.5 They marveled at a pluralist
America able to hold its political leadership accountable while
avoiding mass uprising that could translate into totalitarian
tyrannies.6 Such talk now seems antiquated, even self-indulgent. For
many today, the United States is better understood as just another
society at the advanced stage of capitalism.7 American and European
problems and politics are converging. If any country is exceptional, it
is China, or one of the four Asian Tigers, or perhaps India or Brazil.
Or, to state the situation in its most undeniable terms: Every country
is exceptional. Each has its own distinct geographical location, origin,
history, social composition, and political institutions. The United
States is no more exceptional than Canada, or Mexico, or what have
you.
I. THE EXCEPTIONAL AMERICAN FEDERAL SYSTEM
Still, it is worth treating as exceptional the country’s federal
system, with its unique separation of authority between national and
state governments. According to a recent count, only 25 of the
world’s 193 countries are federal systems.8 And most of those 25 federal
nations circumscribe the authority exercised by lower tiers of
government in important ways. In some, the heads of lower offices
4 See, for example, Daniel Bell, Interpretations of American Politics, in Daniel Bell, ed,
The Radical Right 39, 50 (Doubleday 1963). See also Samuel P. Huntington, American Politics:
The Promise of Disharmony 240–45 (Belknap 1981).
5 See Louis Hartz, The Liberal Tradition in America 4–6 (Harcourt 1955); Huntington,
American Politics at 27–30 (cited in note 4); Sven H. Steinmo, American Exceptionalism
Reconsidered: Culture or Institutions?, in Lawrence C. Dodd and Calvin Jillson, eds, The
Dynamics of American Politics: Approaches and Interpretations 106, 117–24 (Westview 1994).
6 Bell, Interpretations of American Politics at 58–59 (cited in note 4); William
Kornhauser, The Politics of Mass Society 121–23 (Free Press 1959).
7 See, for example, Mancur Olson, The Rise and Decline of Nations: Economic Growth,
Stagflation, and Social Rigidities 92–94 (Yale 1982); Harold L. Wilensky, The Welfare State and
Equality: Structural and Ideological Roots of Public Expenditures 52 (California 1975); Phillips
Cutright, Political Structure, Economic Development, and National Social Security Programs,
70 Am J Sociology 537, 549 (1965). But see Harold L. Wilensky and Charles N. Lebeaux,
Industrial Society and Social Welfare: The Impact of Industrialization on the Supply and
Organization of Social Welfare Services in the United States 41 (Free Press 1958).
8 Forum of Federations, Federalism by Country (2007), online at www.forumfed.org/en
/federalism/by_country/index.php (visited Oct 6, 2011). The number is twenty-three, according
to one count given in Jonathan A. Rodden, Hamilton’s Paradox: The Promise and Peril of
Fiscal Federalism 39 (Cambridge 2006).
2012] Freedom to Fail 255
hold office at the pleasure of the central government.9 In others, the
lower tiers are heavily dependent on the central government for
revenue.10 Except in Canada and Switzerland, state debts in all
federal systems in the industrialized countries of the world are
implicitly or explicitly guaranteed by the federal government.11
The design of the US federal system owes as much or more to
historical circumstances as to theoretical intentions. When writing
the Constitution, those gathered in Philadelphia necessarily allowed
for autonomous action by state governments for the very practical
reason that no other form of government could have won ratification
by the supermajority of states required before the founding
document could take effect. Unless the national government’s
powers were limited and states continued to exercise considerable
power on their own, the citizenry, more fond of their former colonial
governments than the new national entity, would not have agreed to
important limits the Constitution did impose upon the states, such as
restrictions on their abilities to declare war, coin money, and
regulate interstate commerce. The cultural differences between the
slaveholding South and an increasingly antislavery North could be
contained only if each region was allowed to organize its own
domestic affairs. But if the US federal system was initiated to solve a
very practical problem, it gradually became an institutional form so
appropriate and effective that it persisted into the twenty-first
century even after the Civil War had been fought, slaves had been
freed, and a much more powerful federal government had been
established.
That exceptional federal system, best characterized as
competitive federalism, can be sustained only if the lower tiers of
government are held accountable to the marketplace—most
specifically, to the market for government bonds. Unless lower tiers
are subject to independent movements in the interest rates on their
bonds, and unless lower tiers remain at risk of default, or something
tantamount to default, the central government cannot afford to grant
9 See, for example, Republic of India Public Administration Country Profile 8 (United
Nations Division for Public Administration and Development Management Jan 2006), online
at http://unpan1.un.org/intradoc/groups/public/documents/un/unpan023311.pdf (visited Oct 6, 2011).
10 See, for example, Dele Olowu, Property Taxation and Democratic Decentralization in
Developing Countries *19 (Institute of Development Studies Seminar Paper, 2002), online at
http://www2.ids.ac.uk/gdr/cfs/pdfs/Olowu2.pdf (visited Nov 6, 2011).
11 Rodden, Hamilton’s Paradox at 9, 31, 93 (cited in note 8). For a discussion of default
risk within the Canadian federal system, see Stuart Landon and Constance E. Smith,
Government Debt Spillovers and Creditworthiness in a Federation, 33 Can J Econ 634, 636,
653–54 (2000). The argument of our paper does not turn on whether the United States is
exceptionally different from Canada and Switzerland.
256 The University of Chicago Law Review [79:253
wide discretion to state or local governments. For more than two
centuries, the US federal system has survived multiple economic and
political crises, but never has the autonomy of the lower tier of
government been circumscribed to such an extent that state and
municipal bonds do not have their own, independent standing in the
marketplace. Yet a striking new political development—the granting
of collective bargaining rights to those who work for state and local
governments—has posed a dramatically new challenge to the
viability of the American federal system as we have known it. Just
how that has occurred, as well as its potential consequences for the
country’s political institutions, is the topic this paper explores.
II. COMPETITIVE FEDERALISM
Historically, competitive federalism helped to generate the
extraordinary growth of the world’s largest economic power.12 Over
the decades, states and localities designed and maintained canals,
railroads, highways, sewage systems, schools, parks, and systems of
public safety. As Lord James Bryce wrote nearly a century ago,
[I]t is the business of a local authority to mend the roads, to
clean out the village well or provide a new pump, to see that
there is a place where straying beasts may be kept till the owner
reclaims them, to fix the number of cattle each villager may turn
out on the common pasture, to give each his share of timber cut
in the common woodland.13
As compared to the federal government, state and local
governments are more sensitive to political market forces, making
them better equipped to design and administer those types of
programs. Unless local government supplies public services to meet
the needs of local businesses and residents, citizens may “vote with
their feet” and migrate to a locality better attuned to their needs.
Since 12 percent or more of the population changes its residence
each year,14 the effects of policy choices on property values can be
quickly felt.
12 The discussion in this Section draws upon the theoretical statement developed in Paul
E. Peterson, City Limits 68–72 (Chicago 1981). See also Charles Tiebout, A Pure Theory of
Local Expenditures, 64 J Polit Econ 416, 422 (1956); Wallace E. Oates, Fiscal Federalism 240–41
(Harcourt Brace Jovanovich 1972); Paul E. Peterson, The Price of Federalism 18–19
(Brookings 1995). The contributions of local government to rapid economic growth in China
are explored in Gabriella Montinola, Yingyi Qian, and Barry R. Weingast, Federalism, Chinese
Style: The Political Basis for Economic Success in China, 48 World Polit 50, 67–78 (1995).
13 James Bryce, 1 Modern Democracies 132 (Macmillan 1921).
14 See US Census Bureau, Statistical Abstract of the United States: 2012 37 table 30, online
at http://www.census.gov/prod/2011pubs/12statab/pop.pdf (visited Oct 7, 2011).
2012] Freedom to Fail 257
Business and residential choices are influenced by factors other
than the quality of local public services, of course. Businesses want to
be close to both their sources of supply and the markets for their
products. Individual and family residential choices are influenced by
family ties, employment opportunities, and the quality of the natural
environment. But the quality of publicly provided infrastructure also
affects, on the margins, the choices businesses and households
make.15
Since small changes in supply or demand can have a significant
effect on price, residents of a community, eager to protect their
property values, can be expected to pressure government officials to
employ public resources efficiently in order to meet local
expectations and facilitate economic development. Poor policy
decisions can have rapid and lasting effects on a municipality’s
property values and corresponding tax income.16 Therefore, it is
reasonable to expect most state and local governments to be
relatively competent at designing and implementing developmental
policies. Admittedly, lower-tier officials in a system of competitive
federalism may exhibit “narrowness of mind and the spirit of
parsimony,” as Lord Bryce was the first to admit, but if it were
otherwise, “there would be less of that shrewdness which the practice
of local government forms.”17
State and local government can also facilitate the gathering of
information regarding the most efficient way of organizing public
services. Each state or city is a laboratory where experiments are
tried. If the experiment is successful, other governments will copy it.
If the experiment fails, the idea will soon be abandoned. In addition,
“states and localities pay close attention to the wages and salaries
paid to employees in adjacent communities” and will feel pressure to
bring them into line with those of their neighbors.18 So valuable is the
role played by lower tiers of government within the federal system
that, despite the growth in the role of the federal government, more
than 40 percent of all government spending for domestic purposes
was, as late as 2008, paid for out of revenues raised by state and local
governments from their own sources.19 The lower tiers are also the
predominant public-sector employer. No less than 87 percent of all
15 See Peterson, Price of Federalism at 18–19 (cited in note 12).
16 See id at 19.
17 James Bryce, 1 Modern Democracies at 132–33 (cited in note 13).
18 Peterson, Price of Federalism at 19 (cited in note 12). See also Tiebout, A Pure Theory
at 421–22 (cited in note 12).
19 See Figure 3.
258 The University of Chicago Law Review [79:253
nonmilitary public-sector employees work for either the state or
local government.20
In a system of competitive federalism, state and local
governments resist taking responsibility for large-scale redistributive
programs.21 If states and localities attempt in a serious way to tax the
rich and give to the poor, the rich will depart while the poor will be
attracted. If the rich leave and the poor migrate into the state, tax
revenues will plummet while expenditures escalate. Any debt
acquired by state and local governments must be borrowed from
investors; if a state borrows too much money, state bond ratings fall
and, unless the fiscal situation of the state is corrected, the state will
default on its debts.
III. SUPREME COURT JURISPRUDENCE
If a state defaults, it may not be sued without its consent. That
state sovereignty implies immunity from private lawsuits compelling
payment of debt was established in the early years of the Republic.
When the Supreme Court, in Chisholm v Georgia,22 ruled that the
State of Georgia had to pay a citizen of South Carolina a debt it had
incurred,23 Congress passed the Eleventh Amendment to the
Constitution, reversing that decision and “ma[king] it very difficult
[subsequently] for creditors to force states to repay debts.”24 Early
jurisprudence also established that a state’s own citizens could not
file a suit in federal court to secure repayment of debt25 and that a
foreign nation could not successfully compel a state to pay its debt.26
A state is not immune from a suit filed by a sister state or by the
federal government, but neither entity is likely to be a state
bondholder.27 Citizens within a state can file a suit within a state’s
own courts, but state courts have historically not had much success in
compelling other branches of government to honor their debts so
that, as a result, citizens have been “unable to collect on the bonds.”28
20 See Figure 2. See also US Census Bureau, Statistical Abstract: 2012 at 300 table 461
(cited in note 14).
21 See Peterson, Price of Federalism at 29–30, 70 table 3-3, 71 table 3-4 (cited in note 12).
22 2 US (2 Dall) 419 (1793).
23 Id at 453.
24 William B. English, Understanding the Costs of Sovereign Default: American State
Debts in the 1840’s, 86 Am Econ Rev 259, 260 (1996).
25 Hans v Louisiana, 134 US 1, 14–15, 21 (1890).
26 See Monaco v Mississippi, 292 US 313, 330 (1934).
27 Consider English, Understanding the Costs at 260–61 (cited in note 24), citing South Dakota
v North Carolina, 192 US 286 (1904) and United States v North Carolina, 136 US 211 (1980).
28 English, Understanding the Costs at 261 (cited in note 24).
2012] Freedom to Fail 259
One might think that ancient decisions dating back to the
earliest days of the Republic are no longer pertinent, but despite the
array of recent civil rights litigation against states in recent decades,
the original conception of the United States as a federal union in
which sovereignty is enjoyed by both the federal and state
governments has remained altogether relevant for contemporary
jurisprudence. In US Term Limits, Inc v Thornton,29 Justice Anthony
Kennedy, in a concurring opinion, characterized American federalism
in words little different from those James Madison might have used:
Federalism was our Nation’s own discovery. The Framers split
the atom of sovereignty. It was the genius of their idea that our
citizens would have two political capacities, one state and one
federal, each protected from incursion by the other. The
resulting Constitution created a legal system unprecedented in
form and design, establishing two orders of government, each
with its own direct relationship, its own privity, its own set of
mutual rights and obligations to the people who sustain it and
are governed by it.30
Nor can the federal government order a state to compensate its
creditors. The Rehnquist Court invalidated federal laws said to
violate state autonomy by “commandeering” the states. In New York
v United States,31 the majority held that Congress may not simply
“‘commandeer’ state governments into the service of federal
regulatory purposes.”32 Printz v United States33 applied this reasoning
to executive officers as well, holding invalid provisions of the Brady
Handgun Violence Prevention Act34 that required state and local law
enforcement officers to conduct background checks on prospective
handgun purchasers.35 Writing for the majority, Justice Antonin
Scalia concluded: “By forcing state governments to absorb the
financial burden of implementing a federal regulatory program,
Members of Congress can take credit for ‘solving’ problems without
having to ask their constituents to pay for the solutions with higher
federal taxes.”36
29 514 US 779 (1995).
30 Id at 838 (Kennedy concurring).
31 505 US 144 (1992).
32 Id at 175 (holding that Congress does not have the authority to force state
governments to take title to waste under the Tenth Amendment).
33 521 US 898 (1997).
34 Pub L No 103-159, 107 Stat 1536 (1993).
35 Printz, 521 US at 932–33.
36 Id at 930.
260 The University of Chicago Law Review [79:253
With the passage of the Fourteenth Amendment, and the
application of its Due Process and Equal Protection Clauses to the
states, state sovereignty was eroded by a wide variety of civil rights
lawsuits that were effectively prosecuted in both state and federal
courts.37 But Fourteenth Amendment suits generally have been
viewed as constituting exceptions to state sovereign immunity. In
Alden v Maine,38 the Court reaffirmed the states’ immunity to
lawsuits filed in state courts.39 Justice Kennedy rooted the decision in
“the Constitution’s structure, and its history,” saying that “sovereign
immunity derives not from the Eleventh Amendment but from the
structure of the original Constitution itself.”40 However, Kennedy
also said that state sovereign immunity does not extend to suits
brought by the federal government itself and those pursuant to
enforcement of the Equal Protection or Due Process Clauses of the
Fourteenth Amendment.41
Future attempts to limit state sovereignty can be expected to
exploit Fourteenth Amendment exemptions from the doctrine of
state sovereignty. Those who seek to compel states to honor state
pension and health care policies and collective bargaining
agreements can be expected to invoke equal protection and due
process arguments. Bondholders will argue that defaults deny them
property without due process of law. But it is doubtful that such suits
could be successfully pursued in the absence of federal legislation
requiring states to honor implied contracts with bondholders,
pensioners, or public employees.42 In other words, the jurisprudence
that allows states to claim a sovereign status within the federal
system seems as vibrant today as it has ever been. While the
individual constitutions of many states may be interpreted as
granting permission for lawsuits by bondholders, pensioners, or
those protected by collective bargaining agreements,43 states—as
sovereign entities—appear to enjoy today the same legal
prerogatives vis-à-vis bondholders and other creditors as states that
have defaulted in the past, if they so choose.44
37 See, for example, Fitzpatrick v Bitzer, 427 US 445, 456 (1976). Courts also got around
limitations on sovereign immunity by allowing suits to go forward when plaintiffs sued state
entities rather than the state itself. For examples of the courts applying this limitation, see
Brown v Board of Education of Topeka, 347 US 483, 493 (1954); Ex parte Young, 209 US 123,
165 (1908).
38 527 US 706 (1999).
39 Id at 712.
40 Id at 713, 728.
41 Id at 755–56.
42 See, for example, United States v Sherwood, 312 US 584, 590–92 (1941).
43 See, for example, Ill Const Art 13, § 5.
44 See notes 82–91 and accompanying text.
2012] Freedom to Fail 261
IV. COLLECTIVE BARGAINING IN THE PUBLIC SECTOR
Since the beginning of the Republic, states have managed their
fiscal affairs so well that only in a few instances have they defaulted
on their debts. But in the twenty-first century, the risk of default by
large and economically significant states has increased dramatically.
Among the most important contributors to the altered situation has
been the rise of collective bargaining within the public sector. In this
regard, events within the field of education are especially instructive,
as school personnel are the largest segment of the state and local
workforce and education costs constitute approximately one-third of
all state and local expenditures paid for out of locally generated
revenues.45
Public-sector collective bargaining was largely unknown prior to
the 1960s. Even Franklin D. Roosevelt, the most significant
presidential ally the labor movement has ever enjoyed, rejected
public-sector bargaining within the federal government: “All
Government employees should realize that the process of collective
bargaining, as usually understood, cannot be transplanted into the
public service. . . . The employer is the whole people, who speak by
means of laws enacted by their representatives in Congress.”46
George Meany, the head of the American Federation of Labor, did
not disagree. As late as the 1950s, he plainly stated, “It is impossible
to bargain collectively with the Government.”47 Other organizations
that represented government employees took the position that
collective bargaining “was demeaning for civil service professionals.”48
The National Association of Education (NEA), by far the largest of all
teacher organizations, was firmly opposed to the idea.49
Collective bargaining was introduced into the public sector in
part because the policy fit the political needs of the Democratic
Party, which had been closely affiliated with the labor movement
45 In 1990 the percentage expended by state and local governments from their own
resources (that is, excluding federal grants) for elementary, secondary, and higher education
was 34.1 percent. See US Census Bureau, Statistical Abstract: 2012 at 273 table 435, 300 table 462
(cited in note 14).
46 Letter from President Franklin D. Roosevelt to Luther C. Steward, President of the
National Federation of Federal Employees (Aug 16, 1937) (American Presidency Project),
online at http://www.presidency.ucsb.edu/ws/?pid=15445#axzz1bRZWSTlq (visited Oct 21, 2011).
47 George Meany, Meany Looks into Labor’s Future, NY Times Mag 11, 38 (Dec 4, 1955).
48 Paul E. Peterson, Saving Schools: From Horace Mann to Virtual Learning 106
(Belknap 2010).
49 Martin Raymond West IV, Politics, Public-Sector Unionism, and Education Policy:
Explanations and Evaluations *38 (unpublished PhD dissertation, Harvard University, 2006)
(on file with authors).
262 The University of Chicago Law Review [79:253
since the 1930s.50 Since the 1950s, private-sector unionization has
been on the decline, slipping from roughly one-third to less than
10 percent of the nongovernmental workforce.51 Mobilizing new
recruits to the Democratic coalition became critical, and nothing was
more appealing than reaching out to the growing segment of the
workforce employed by local, state, and federal governments.
Previously, public-sector workers had not shown any particular
partisan loyalty other than to the machine that hired them.52 The
many white-collar professionals working for government were, if
anything, more inclined to the Republican side of the aisle.53 But farsighted
union leaders and key Democratic members of Congress,
perceiving an opportunity, began to campaign for collective
bargaining rights for public-sector workers.54 Dwight Eisenhower, a
Republican, stoutly resisted all congressional efforts to pass such
legislation into federal law, but his Democratic successor, John
Kennedy, promised he would take action if elected President.55 Since
the close balance of power on Capitol Hill precluded passage of
collective bargaining legislation, the President signed an executive
order giving federal employees the right to bargain collectively.56
That executive order, in conjunction with the success of New
York City teacher unions in obtaining collective bargaining rights,
initiated a decisive transformation of the American public sector.
The number of teacher strikes increased from 9 to 107 in just three
years, between 1964 and 1967.57 Faced with the prospect of school
shutdowns and masses of teachers picketing outside once uneventful
classrooms, school boards gave in to public pressure to settle strikes
quickly and return children to school. Affiliates of the American
Federation of Teachers (AFT) won recognition rights in many large
cities, including Boston, Chicago, Cleveland, and Philadelphia.58 “For
the first time . . . since 1918, the AFT threatened to surpass the
50 For labor-friendly New Deal legislation, see National Industrial Recovery Act of 1933,
Pub L No 73-10, 48 Stat 31, codified at 15 USC § 701–10, terminated by Executive Orders 7252
(Dec 21, 1935) and 7323 (March 26, 1936); National Labor Relations Act, 29 USC § 151 et seq.
51 See US Bureau of Labor Statistics, Press Release, Union Members—2010 1 (Jan 21,
2011), online at http://www.bls.gov/news.release/union2.nr0.htm (visited Oct 21, 2011).
52 See West, Politics at *41 (cited in note 49).
53 See Clem Brooks and Jeff Manza, Class Politics and Political Change in the United
States, 1952–1992, 76 Social Forces 379, 393 figure 1 (1997).
54 See West, Politics at *55 (cited in note 49).
55 See id at *56–58.
56 Executive Order 10988, 27 Fed Reg 551 (1962). See also West, Politics at *64 (cited in
note 49).
57 Peterson, Saving Schools at 113 (cited in note 48).
58 See id.
2012] Freedom to Fail 263
NEA,” one historian has noted.59 That changed when the NEA
dropped its principled opposition to collective bargaining as it saw its
membership ranks rapidly defect to the AFT. Eventually both
organizations prospered, with “NEA membership climbing from
700,000 in 1960 to 3.2 million in 2007, while the smaller AFT grew
from under 60,000 to 1.3 million over the same period.”60 Scarcely
known in education before 1960, collective bargaining achieved
predominance in most states outside the South.61
In present times, collective bargaining is so pervasive within the
public sector that few remember Franklin Roosevelt’s objections to
such a practice. There remains a handful of critics who argue that
collective bargaining subverts the democratic relationship between
government and citizens by privileging a particular set of interests—
those of government employees. Notably, such critics draw a
distinction between collective bargaining in the public and private
sector. In the private sector, collective bargaining is often
appropriate, such critics argue, because workers may need to bargain
collectively in order to prevent a profit-maximizing management
from abusing its superior bargaining position vis-à-vis individual
employees. When resolute unions bargain with a management
indifferent to all but its bottom line, each protects its own vital
interests in the collective bargaining process. But within the public
sector, such countervailing power cannot be assumed. The
“management” in the public sector is made up of elected officials,
such as school board members, to whom unions contribute heavily
during the election process. Also, school employees participate more
frequently than others in “school elections, which are often lowvisibility,
non-partisan affairs that engage the attention of only the
most interested parties.”62 Campaign contributions and coordinated
voting blocs give employees special influence over the very school
board with which they negotiate. Though not quite self-dealing,
teachers’ unions are certainly not bargaining with hostile
59 Marjorie Murphy, Blackboard Unions: The AFT and the NEA, 1900–1980 220 (Cornell
1990).
60 See Peterson, Saving Schools at 113 (cited in note 48).
61 See Richard C. Kearney and David G. Carnevale, Labor Relations in the Public
Sector 38, 60–61 table 3.2 (Marcel Dekker 2001); Randall W. Eberts, Teachers Unions and
Student Performance: Help or Hindrance?, 17 Future Children 175, 178 (2007). For further
research on the growth of teachers’ unions since 1950, see Joseph E. Slater, Public Workers:
Government Employee Unions, the Law, and the State, 1900–1962 193 (Cornell 2004). Hanna
Skandera and Richard Sousa, School Figures: The Data behind the Debate 106–08 (Hoover
2003). AFT membership includes university faculty, paraprofessionals, and other school
employees.
62 Peterson, Saving Schools at 114 (cited in note 48).
264 The University of Chicago Law Review [79:253
management representing interests in opposition to those of the
employees.63
Union political power has been expanded by collective
bargaining agreements in other ways as well. Contracts in many
districts require an amount equivalent to union dues be deducted
from employee paychecks, the use of which is given to the discretion
of the employee’s union. These deductions include fees that, unless
specifically objected to by a member, may be used for political
purposes.64 With such resources, teacher unions have become among
the most influential groups in state politics: in 1985 “teachers’
organizations” were identified as the most influential interest group
in state politics; in 2002 they were found to be second only to
business groups. In both surveys, they outranked such powerful
groups as utility companies, insurance companies, hospitals, trial
lawyers, manufacturers, and those representing local governments
more generally.65
Assessing the consequences of collective bargaining within the
public sector is a controversial matter. But we do know that since the
mid-1960s, per-pupil expenditures on elementary and secondary
education have tripled in real-dollar terms—from less than $4,000
per pupil (in 2006 dollars) to nearly $12,000 in 2008.66 Much of the
increment is to be explained by the growth in the number of public
school employees both because the number of pupils per teacher fell
by one-third (from twenty-five to sixteen) and because many more
nonteaching professionals were hired to provide ancillary services
and to help manage an increasingly complex system.67 In 1960, school
districts employed 6 professionals for every 100 students; by 2005
they were employing more than 12 for that same number.
Nonprofessional hiring rose at a similar rate—from less than 2 per
100 students in 1960 to nearly 4 in 2005.68 Most of the rising cost of
education was borne by state and local governments, as the federal
contribution did not rise much above 10 percent of the total.69
63 Terry M. Moe, The Union Label on the Ballot Box: How School Employees Help
Choose Their Bosses, 6 Educ Next 58, 60 (Summer 2006).
64 Peterson, Saving Schools at 114–15 (cited in note 48).
65 Clive S. Thomas and Ronald J. Hrebenar, Interest Groups in the States, in Virginia
Gray and Russell L. Hanson, eds, Politics in the American States: A Comparative Analysis 100,
119 table 4-1 (CQ 8th ed 2004).
66 See Thomas D. Snyder, Sally A. Dillow, and Charlene M. Hoffman, Digest of
Education Statistics, 2009 100 table 64, 260 table 181 (Department of Education 2010).
67 See Thomas D. Snyder and Sally A. Dillow, Digest of Education Statistics, 2008 117
table 80 (Department of Education 2009).
68 See id.
69 Peterson, Saving Schools at 153 (cited in note 48).
2012] Freedom to Fail 265
During this period of time, teacher salaries have kept pace with
overall wage and salary increases nationwide.70 In addition, teachers
and other public-sector employees have been guaranteed steep
increases in pensions, health care, and other nonsalary benefits, as
elected officials choose to reach collective bargaining settlements by
rewarding workers with promises of future benefits rather than
immediate salary compensation.71 In most cases, the cost of these
benefits was postponed into a future well beyond current election
cycles.72 For years the growing imbalance between rising costs and
increasing liabilities, on the one side, and fiscal resources, on the
other, was ignored, except by the Cassandras of the policy world.73
But with the financial crisis of 2008, the possibility of state and
municipal defaults shifted from the theoretical to the plausible.74
V. STATE FISCAL CRISES
The lower tiers of the US government are facing a
contemporary fiscal crisis unprecedented since the days of the Great
Depression. While some resource-rich, less populous states—Alaska,
Montana and North Dakota, for example—continue to run balanced
budgets,75 most states are confronting large deficits. New York’s
deficit for the fiscal year 2012 was estimated in early 2011 to be
18 percent of the previous year’s budget, California’s to be 29 percent,
Texas’s to be 32 percent, and New Jersey’s to be no less than 38
percent.76 The gap in official state budgets was estimated to be at
$121 billion, or 19 percent of the budget in the forty-six states
running deficits.77 The size of these projected deficits may have
attenuated as state economies have recovered, but they would be
70 See id at 133.
71 See, for example, Letter from Daniel W. Hancock, Chairman of the Little Hoover
Commission, to Edmund G. Brown, Governor of California, and Members of the California
Legislature (Feb 24, 2011) (“Hancock Letter”), in Public Pensions for Retirement Security First
Page (Little Hoover Commission Feb 2011), online at http://www.lhc.ca.gov/studies/204
/report204.html (visited Oct 22, 2011).
72 See David Stella and Keith Bozarth, Pension Sustainability—The Wisconsin Example,
47 Benefits & Compensation Dig 32, 33 (Feb 2010).
73 See, for example, Hancock Letter at First Page (cited in note 71); Roger Lowenstein,
The End of Pensions, NY Times Mag (Oct 30, 2005), online at http://www.nytimes.com/2005/10/30
/magazine/30pensions.html?pagewanted=print (visited Oct 22, 2011).
74 See Kevin Hassett, California Leads Nation to Bond Default Abyss, Tulsa World A12
(June 2, 2009).
75 Daniel J. Nadler and Sounman Hong, Political and Institutional Determinants of Tax-
Exempt Bond Yields *3 (Harvard Kennedy School Report No 11-04), online at
http://www.hks.harvard.edu/pepg/PDF/Papers/PEPG_11-04_Nadler_Hong.pdf (visited Oct 10,
2011).
76 See id at *15 table 1.
77 Id at *3.
266 The University of Chicago Law Review [79:253
considerably larger if they were to include the revenues necessary to
fully fund state pension and health care obligations.78
With the onset of the financial crisis, the bond market
immediately took note of the increased risk of sovereign state
defaults. In late 2008 investors demanded a higher premium for state
and local bonds over safer US Treasury securities, despite the
exemption from federal taxation of interest received on most state
and local bonds. Although all states were affected by the crisis, the
perceived risks of default varied considerably among the states.
“Between September and December of 2008, the premium that
investors demanded to hold California debt over US treasuries
jumped from 24 basis points to 271 basis points, a ten-fold increase.”79
(100 basis points equals one percent.) Before the crisis, the
difference in the premium paid in California and Texas was only 15
basis points. But by 2011 the gap between the two states had
increased to 84 basis points.80 Similar jumps in the cost of borrowing
occurred in a number of other states as well.81 Clearly, investors had
become increasingly sensitive to the variation in the risk of defaults
among the sovereign states.
VI. IMPACT OF COLLECTIVE BARGAINING ON DEFAULT RISK
State and municipal defaults are not unknown to American
federalism.82 Eight states defaulted or repudiated debt between 1841
and 1843 when a severe economic depression restricted state ability
to pay interest on debt that had been assumed primarily for the
purpose of constructing canals and railroads.83 The federal
government refused to assume responsibility despite efforts by both
defaulting states and foreign banks to persuade the federal
government to intervene.84 While four states eventually repaid all of
their debt, three made only a partial repayment, and one,
Mississippi, never did.85
78 Alan J. Auerbach, Long-Term Objectives for Government Debt *10–13 (Swedish Fiscal
Policy Council Conference on Fiscal Policy and Labour Market Reforms, Feb 2008), online at
http://elsa.berkeley.edu/~auerbach/long_term_objectives_govt_dept.pdf (visited Nov 4, 2011).
79 Nadler and Hong, Political and Institutional Determinants at *4 (cited in note 75).
80 See id.
81 See Figure 4.
82 In this paper we ignore defaults by municipalities and other lower-tier governments.
These units of government do not have status in the US Constitution, and their status varies
from one state to the next, depending on state law.
83 See English, Understanding the Costs at 261–65 (cited in note 24).
84 See Rodden, Hamilton’s Paradox at 59–60 (cited in note 8).
85 English, Understanding the Costs at 265 table 3 (cited in note 24).
2012] Freedom to Fail 267
The boom-and-bust economy of the 1870s and 1880s provoked
another ten defaults, and Arkansas was unable to cover its debts
during the 1930s depression.86 Bondholders were not the only
creditors that states ignored during hard times. During the Great
Depression, Chicago teachers were on several occasions paid in
“scrip,” because the City could not find the cash in hand to
compensate them.87 Years later the “scrip” was made good, but few
teachers themselves ever received payment in full, as they had used
their highly discounted “scrip” to pay monthly bills.88
In none of these crises was there much hope that the federal
government would come to the rescue of states at risk of default.
During the 1840s, some political leaders invoked the precedent
established when Congress, prompted by Alexander Hamilton,
assumed the Revolutionary War debts incurred by some of the
states.89 But others argued that the precedent did not hold in that
revolutionary war debts had been incurred on behalf of a common
cause, while the state debts incurred prior to the 1840s were for the
purpose of setting up banking and transportation systems designed
mainly for the benefit of the state itself. Neither of the national
political parties saw any advantage in coming to the rescue of a few
states at the expense of the remainder.90
Nor has the US government guaranteed state debts in any
subsequent crisis. As a result, each state is held accountable by the
bond market in ways that lower-tier governments in most other
countries are not. Consider, for example, the differing response of
the bond market to the state bonds issued in the United States and in
the German Federal Republic. Even though the constitution of the
German Federal Republic, adopted in the aftermath of World War II,
was explicitly modeled on that of the United States and assigned
major responsibilities to state governments, the German federal
government has asserted control over state finances and guarantees
state debts.91 For that reason, the spread between German federal
and state securities is less than the spread between such securities in
the United States. As is shown in Figure 3, the 2008 financial crisis
86 Andrew Ang and Francis A. Longstaff, Systemic Sovereign Credit Risk: Lessons from
the U.S. and Europe *6 (NBER Working Paper No 16982, Apr 2011), online at http://www.nber.org
/papers/w16982 (visited Oct 8, 2011).
87 See Paul E. Peterson, The Politics of School Reform, 1870–1940 179 (Chicago 1985).
88 See William J. Grimshaw, Union Rule in the Schools: Big-City Politics in
Transformation 64 (Lexington 1979).
89 See Rodden, Hamilton’s Paradox at 57, 60 (cited in note 8).
90 See id at 62–63.
91 See id at 91. See also id at 83–87 table 4.1 (reporting that the variance in the credit
ratings of lower-tier governments in Germany is much smaller than in the United States).
268 The University of Chicago Law Review [79:253
had an impact on the perceived default risk of the average German
state relative to that of the German Federal Republic. The average
yield spread between the two types of securities from 36 basis points
to 112 basis points between June 2008, the eve of the financial crisis,
and December 2008, when the crisis was at its peak. But during that
same period, the average yield spread for US states relative to
federal securities increased from 19 basis points to 129 basis points.
Clearly, the bond market perceived that the risk of default by a state
in Germany was attenuated by the guarantee supplied by the
German Federal Republic.
In the United States, investors were willing to accept lower
interest rates on state debt securities relative to US Treasuries due to
their federal-tax-exempt status. After the financial crisis, however,
the yield on state bonds rose above that for comparable federal
securities, as any tax advantages were overwhelmed by perceived
increased risk.92 Rates of return on state bonds before the financial
shock trailed those for Treasury securities because federal taxes need
not be paid on the returns from most state and municipal bonds. But
after the financial crisis, the spread between state and federal bonds
turned from negative to positive, as the relative risk from state
investments outweighed any tax advantages. Moreover, the yield
spread between state and federal bonds varied significantly from state
to state, indicating that the market perceived greater default risk in
certain states.93
Notably, investors’ perceptions of the risk of default were
correlated with the unionization rate of the public-sector workforce.94
As shown in Figure 5, the relationship between union membership
and default risk was noticeably weaker in June 2008, prior to the
financial crisis, than it was over the next six months. Figure 5’s
vertical axis shows the spread for federal securities and state bonds
that will mature in one year, while the horizontal axis shows the
unionization rate for the state’s public sector. “The relationship
between the two variables, modest in June 2008, becomes
pronounced by June 2009, as bondholders became highly sensitive to
a state’s perceived political capacit[ies] to take actions needed to
bring budget deficits under control.”95 The differences in the
steepness of the slopes taken by the regression lines in Figure 5
92 Nadler and Hong, Political and Institutional Determinants at *8 (cited in note 75).
93 See Figure 4.
94 See Nadler and Hong, Political and Institutional Determinants at *8 (cited in note 75).
95 Id.
2012] Freedom to Fail 269
describe the strengthening of the simple relationship between the
union share of the public-sector workforce and default risk.96
This relationship persists when other factors are controlled, as
has been shown by Daniel Nadler and Sounman Hong.97 In this
study, unbiased estimates of the impact of political variables on state
default risks are estimated with models that solve for the
endogenous relationship between credit and yield, and that also take
into account the economic factors that James Poterba and Kim
Rueben have shown to be associated with a state’s default risk:
change in its unemployment rate, Gross Domestic Product (GDP),
and deficit-to-GDP ratio.98 Nadler and Hong present estimates of the
impact of a range of state-level political variables—such as the union
share of the public-sector workforce and partisan representation in
the legislature—on state municipal bond yield spreads in the context
of the unexpected deficit shocks seen following the 2008 financial
crisis. In particular, they evaluate the impact of the union share of
the public-sector workforce and partisan representation in the
legislature using “separate models, because unionism and partisan
balance are highly correlated with one another, making it difficult,
with the small number of observations available, to identify the
independent impact of each within a single model.”99
Their results are reproduced here in Table 1. According to
Nadler and Hong, unexpected deficit shocks of the size that took
place in 2008 especially affect state yield spreads when certain
political conditions are present. As can be seen in Table 1, a
1 percent difference in union membership in a state is associated
with an additional 2.02 basis point change in state borrowing costs, if
the state has experienced a billion-dollar change in its unexpected
deficit shock. In other words, a twenty percentage-point difference in
the share of the public-sector workforce that is unionized (one
standard deviation) is associated with an additional increase in the
96 Unlike the simple relationship between union share of the public workforce and bond
yield spreads, the simple relationship between yield spreads and the partisan composition of
the state legislature does not increase in the wake of the fiscal crisis. The impact of this political
factor is detectable only after estimating the model presented in Table 1.
97 Nadler and Hong, Political and Institutional Determinants at *6–7 (cited in note 75).
98 See id at *6, 8. Since the relationship between decisions to issue state bonds and yield
spreads are endogenous, an unbiased estimate of the factors affecting yield spreads cannot be
obtained by an ordinary least-squares regression. However, if a change in the size of the deficit
is unexpected, the demand for credit changes regardless of the price of the bond, permitting
unbiased estimates. The model thus estimates the interaction between deficit shocks and the
key political variables included in the models. The model follows those used by James M.
Poterba and Kim Rueben, Fiscal News, State Budget Rules, and Tax-Exempt Bond Yields, 50 J
Urban Econ 537, 539–44 (2001).
99 See Nadler and Hong, Political and Institutional Determinants at *9 (cited in note 75).
270 The University of Chicago Law Review [79:253
level of state bond spreads of 40.4 basis points, for every billiondollar
change in unexpected deficit shock that a state experiences.100
Similarly, Nadler and Hong found that a one-percentage-point
increase in the Democratic share of a state legislature is associated
with an additional 3.02 basis point increase in state borrowing costs
for every billion-dollar change in a state’s unexpected deficit shock.
This suggests that an increase in the Democratic legislative
representation by twenty percentage points is associated with a
60.4 basis point increase in the state-to-federal bond yield spread in
the context of a billion-dollar deficit shock. “The cost to the state
taxpayer of a standard deviation shift in either variable is, roughly
speaking, about one half of one percent on a five-year security
note.”101 As Nadler and Hong argue, “That amount is non-trivial. In
Illinois, an increase in the yield spread of that magnitude on its debt
of $145.5 billion amounts to $727 million dollars in additional
interest costs annually.”102
However, one should not reify these two indicators of a state’s
political situation. Union share of the public-sector workforce and
partisan representation in the legislature are actually indicators of a
broader set of factors affecting a state’s risk of default.103 As Nadler
and Hong’s work makes clear, the unionization rate of the publicsector
workforce is correlated with factors such as whether a state
has a right-to-work law and whether the legislature has permitted
public-sector collective bargaining, both of which are correlated with
the magnitude of bond yield spreads. In addition, the percentage of
the legislature that is Democratic is highly correlated with the
percentage that is Democratic in each house of the legislature, which
also are correlated with bond yield spreads.104 (However, as Nadler
and Hong note, that data shows that the partisan affiliation of the
governor is not correlated with yield spreads, “suggesting that
governors have broader constituencies than do members of the
legislature.”)105
The two interval variables emphasized by Nadler and Hong—
public-sector unionization and political orientation of the legislature—
100 Id.
101 Id.
102 Id.
103 Variations in state expenditures on Medicaid, a “highly redistributive program that
might be considered a default risk factor, were not correlated with yield” spreads. Bondholders
apparently think that expenditures incurred outside of collective bargaining agreements are
more easily managed. See Nadler and Hong, Political and Institutional Determinants at *9 n 16
(cited in note 75).
104 Id at *9.
105 Id.
2012] Freedom to Fail 271
should be understood as useful proxies for a broader set of collective
bargaining and partisan factors that affect bond yields. Economic
factors—growth, change in the deficit-to-GDP ratio, and change in
the unemployment rate—are also strongly associated with the
substantial interstate variation in yield spreads that occurred in the
wake of the financial crisis. But in addition to the impact of these
economic factors, political realities are clearly also taken into
account by bondholders: Nadler and Hong found that, when both are
entered into the same equation, the political variables seem to be at
least as important in explaining post-crisis interstate variation in
yield spreads as are core economic indicators such as state-level
growth in GDP and changes in state unemployment rates.106
VII. DEFAULT RISKS AND COMPETITIVE FEDERALISM
A system of competitive federalism has long been extolled as a
permanent feature of American government.107 Both early and
modern Supreme Court jurisprudence has recognized states as
sovereign entities that exercise autonomous power—and incur
concomitant risks—within the sphere allocated to them by the
Constitution.108 States and localities play a major role in the raising of
revenue, the delivery of services, and the servicing of public debt.
Nothing is as quintessentially American as the dual sovereignty
granted both to states and to the federal government.
Yet in Hamilton’s Paradox, a recent study of federal systems,
Jonathon Rodden argues that attempts to sustain systems of
competitive federalism usually fail, attributing to Alexander
Hamilton a similar appreciation of competitive federalism’s
fragility.109 When sovereignty is divided, lower-tier governments are
tempted to run debts that place themselves at grave risk of default in
times of financial crisis.110 And central governments, both to
safeguard their international credit rating and to respond to internal
political pressures, cannot resist providing the assistance necessary to
safeguard bondholders and other creditors from loss.111 Central
governments do not offer a helping hand without at the same time
asserting their authority, however. If they rescue states and localities
they will feel more than entitled to take preventative measures
106 See id at *10–12.
107 See note 12.
108 See, for example, Hans v Louisiana, 134 US 1, 21 (1890); New York, 505 US at 175.
109 Rodden, Hamilton’s Paradox at 2 (cited in note 8).
110 See id at 8.
111 See id at 78–79.
272 The University of Chicago Law Review [79:253
designed to preclude future defaults.112 Irresponsibility at the state
and local level thus undermines the dual sovereignty essential for the
survival of competitive federalism. Celebrated in theory as an
efficient government of Herculean proportions, competitive
federalism is but a ten-pound weakling in practice. To prove his
claims, Rodden inquires into the functioning of three large and
important federal systems—Australia, Brazil, and Germany. Based
on his findings, he recommends that new states be constructed either
as unitary systems or designed in such a way as to give the central
government undisputed fiscal authority.
So it is of considerable interest that Rodden does not apply his
argument to the United States. That exceptional nation, Rodden
suggests, can still enjoy the benefits of a system of competitive
federalism, because the arrangement has been woven so deeply into
the fabric of the society that it cannot be torn asunder. In the 1840s,
the national government stood aside when multiple states defaulted,
and it has never intervened to help them out in the decades since.113
Other factors have contributed to a stable system of competitive
federalism as well: the size of government has been relatively small,
the lower tiers of government have been responsible for a fairly large
share of all domestic spending, grants from the federal government
have remained only a moderate proportion of total state and local
spending, and debts have made up a small percentage of most states’
GDP.
Yet within five years of the publication date of Rodden’s
insightful study, competitive federalism in the United States seems
more fragile than it has ever been. Many of the stabilizing factors are
gradually being whittled away. In recent years, the size of federal,
state, and local government has grown from less than 30 percent to
over 35 percent of GDP, the federal share of overall domestic
expenditures has been on the increase,114 intergovernmental grants
are making up a greater share of total lower-tier expenditures, and
state and national debt is escalating at an astounding rate. In the
spring of 2009, Congress, as part of the stimulus package, transferred
hundreds of billions of dollars to states and localities, and tens of
billions in additional aid were appropriated the following year.115
Though presented as legislation that would protect public-sector
112 See id at 271.
113 See Rodden, Hamilton’s Paradox at 62, 67 (cited in note 8).
114 See Figure 1.
115 American Recovery and Reinvestment Act of 2009, Pub L No 111-5, 123 Stat 115.
2012] Freedom to Fail 273
jobs,116 the monies were most valued by governors and mayors as
mechanisms for reducing fiscal deficits.117 Though not a federal
guarantee against default, the stimulus packages provided a dramatic
example of the way in which federal aid can ameliorate state and
local distress when states find themselves at risk of default.
That sovereign entities may be at risk of default in the coming
decades is well understood. It is not just Greece, Ireland, Portugal,
Spain, and Italy whose debt situations have become a matter of
urgent concern. Even the US government is at risk, as the net foreign
debt of the US central government, in the absence of corrective
measures, is projected within the next twenty years to rise from
about $5 trillion dollars in 2011 (about 33 percent of GDP) to $50
trillion, or more than 140 percent of GDP,118 a level “far above any
levels that could be considered sustainable.”119 Those numbers do not
include the sovereign debts of the fifty states of the union, a debt
that is currently about 7 percent of GDP.120 Nor do they take into
account the value of the unfunded liabilities faced by public-sector
pension plans, officially estimated at $438 billion by states
themselves but which could in fact be as high as $3 trillion dollars,
about 20 percent of GDP.121
Within the United States the sovereign state default crisis is for
some states—Illinois, California, and New Jersey, for example—
serious enough that Washington policy makers are currently
debating the policy and constitutional implications of three
alternatives: federal loans that would bail out states at risk of
default,122 bankruptcy procedures,123 and simple defaults of the kind
116 See The White House, Press Release, State Governments Expected to Credit Recovery
Act with Creating, Saving at Least 250,000 Education Jobs Nationwide (Oct 19, 2009), online at
http://www.whitehouse.gov/the-press-office/state-governments-expected-credit-recovery-actwith-
creating-saving-least-250000-ed (visited Oct 9, 2011).
117 See William Murphy, Stimulus Helps Shrink Deficit, Newsday A20 (Aug 6, 2009).
118 William R. Cline, Long-Term Fiscal Imbalances, US External Liabilities, and Future
Living Standards, in C. Fred Bergsten, ed, The Long-Term International Economic Position of
the United States 11, 24 table 2.3 (Peterson Institute for International Economics 2009).
119 C. Fred Bergsten, The Global Crisis and the International Economic Position of the
United States, in Bergstern, ed, Long-Term Economic Position (cited in note 118). See also The
Current Account Deficit and US Foreign Debt Hearing before the Senate Budget Committee,
110th Cong, 1st Sess 3 (2007) (testimony of C. Fred Bergsten).
120 See Steven Maguire, State and Local Debt: An Analysis *5 (CRS Mar 31, 2011), online
at http://www.nasbo.org/LinkClick.aspx?fileticket=4sLYo0HTYI8%3D&tabid=81 (visited Oct
23, 2011).
121 Andrew G. Biggs, The Market Value of Public Sector Pension Deficits 5 (American
Enterprise Institute for Public Policy Research, Apr 2010), online at http://www.aei.org/files
/2010/04/06/2010RPOno1g.pdf (visited Nov 5, 2011).
122 See, for example, Douglas Turner, Serious Budget Troubles Brewing in Many States,
Buffalo News A8 (Jan 10, 2011).
274 The University of Chicago Law Review [79:253
that occurred during the 1840s.124 Representative Patrick McHenry,
chairman of a subcommittee of the Committee on Oversight and
Government Reform says that “already state and municipal
governments are coming to Washington, hat-in-hand, expecting a
federal bailout.”125 Berkeley School of Law Dean Christopher Edley
has proposed that the federal government bail out states by lending
them federal money at low interest in the expectation that it will be
paid in due course.126 The Obama administration’s proposal to loan
monies to states to help them cover deficiencies in their state
unemployment insurance accounts sets a precedent for larger and
more consequential federal actions in the future.
Bankruptcy protection has been proposed by University of
Pennsylvania law professor David Skeel. In his view, the country
needs a federal bankruptcy law designed specifically for sovereign
debts that would “enable a state to restructure [its] obligations.”127
Such a law, he argues, would be constitutional as long as state
sovereignty were protected by giving states the option to invoke
bankruptcy procedures rather than mandating them to enter
bankruptcy court if they would otherwise default.128 Voluntary
participation in bankruptcy procedures would give states the
opportunity to restructure their obligations to employees,
pensioners, and bondholders, much as bankrupt corporations may
continue to operate while under the protection of federal bankruptcy
law. Not only would bankruptcy give states the opportunity “to
restructure obligations that are [otherwise] extremely difficult to
restructure,” but it would “ensure[] that most or all of a state’s
constituencies make sacrifices, not just one or two.”129 Jeb Bush and
Newt Gingrich have proposed a similar plan that would give states
123 See, for example, Jeb Bush and Newt Gingrich, Let States Declare Bankruptcy:
Reorganization Allowing Breaking of Union Contracts May Be the Best Way for Some,
Baltimore Sun A13 (Jan 31, 2011).
124 See, for example, Jeff Segal, Martin Hutchinson, and Rob Cox, California’s Only
Option, NY Times B2 (June 10, 2009).
125 State and Municipal Debt: The Coming Crisis? Hearing before the Subcommittee on
TARP, Financial Services and Bailouts of Public and Private Programs of the House Committee
on Oversight and Government Reform, 112th Cong, 1st Sess 1 (2011) (“State and Municipal
Debt Hearings”) (Rep Patrick McHenry).
126 Christopher Edley, Let Treasury Rescue the States, NY Times A25 (July 8, 2010).
127 David A. Skeel Jr, States of Bankruptcy, 79 U Chi L Rev *18 (forthcoming 2012),
online at http://ssrn.com/abstract=1907774 (visited Oct 9, 2011).
128 Id at *23–25.
129 State and Municipal Debt Hearings at 5 (testimony of David A. Skeel Jr). See also
Skeel, States of Bankruptcy at *19–20 (cited in note 127).
2012] Freedom to Fail 275
the opportunity to seek bankruptcy protection in the event of a
deficit crisis.130
Nicole Gelinas of the Manhattan Institute argues that “state
bankruptcy would create more problems than it would solve.”131 Most
states do not owe their debt through a single entity, making it
difficult for any single bankruptcy court to handle the extraordinary
complexities involved. For example, pension obligations are typically
borne by local governments as well as by the state, adding to the
number of participants in bankruptcy procedures.132
None of the three proposed options are attractive, but if the
state fiscal crisis becomes increasingly severe, as could happen if
projected deficits in pension and health care accounts materialize,
then the federal loan option may prove to be the most politically
palatable. Multiple state and municipal defaults would likely
provoke a nationwide political crisis and could affect the credit of the
US government, especially if its debt-to-GDP ratio continues to rise.
Passage of bankruptcy legislation could allow for a more managed
imposition of costs on the full range of creditors, including
bondholders, pensioners, and beneficiaries of collective bargaining
agreements, but bankruptcy could also affect US credit in world
markets and would create a legal nightmare, given the complexity of
state contractual arrangements with its creditors. By comparison,
federal loans provide an attractive option to those elected officials
aligned with public-sector unions, a constituency at risk in any
bankruptcy proceeding. Even if power in Washington is divided
between the two political parties, the fear of international
consequences could induce compromises that require substantial
federal contributions to states along the lines of the stimulus package
passed in 2009.
The current contemporary flirtation with default, coupled with
demands for a federal rescue, poses a threat to the system of
competitive federalism. The threat comes not so much from the
accumulation of debt as the obligations that have been incurred as
part of the collective bargaining process, many of which may be
enforceable in court. So it is probably not surprising that a state’s
default risk, as judged by the contemporary bond market, is related
both to the share of the public-sector workforce that is unionized and
to the percentage of the members of the state legislature affiliated
130 See Jeb Bush and Newt Gingrich, Better Off Bankrupt: States Should Have the Option
of Bankruptcy Protection to Deal with Their Budget Crises, LA Times A19 (Jan 27, 2011).
131 State and Municipal Debt Hearings at 1 (testimony of Nicole Gelinas).
132 See id at 1–2.
276 The University of Chicago Law Review [79:253
with the Democratic Party.133 If that party is in control of the federal
government, it can be expected to look favorably on requests that it
rescue states in need. In the midst of the latest crisis, Warren Buffett,
a prominent investor with a large stake in the state and municipal
bond market, expressed the hope that such federal action would be
forthcoming, conceding that “[t]he bond insurers . . . have extraordinary
liabilities,” but doubting that “the federal government
[would] turn away a state that is having extreme financial difficulties
when in effect it honored” the debts of corporate entities, including
General Motors.134 Later, in an interview with the congressional
Financial Crisis Inquiry Commission, he qualified that assessment,
saying, “I don’t know how I would rate [state bond default risks]
myself. . . . It’s a bet on how the federal government will act over
time.”135
Making a bet on the federal response to a state sovereign debt
crisis is beyond the scope of this paper. We claim only that the
introduction of collective bargaining has magnified the risk of state
sovereign defaults, complicated the resolution of deficit problems
that provoke such crises, heightened the likelihood of a federal
intervention if such crises materialize, and set the conditions for a
transformation of the country’s federal system. The costs of such
actions are greater than just the dollar numbers explicitly on the
bargaining table. Within the past decade a system of competitive
federalism that once enjoyed an exalted, even Olympian, standing in
American political culture has now been placed at risk.
133 This Section draws upon the findings and analysis presented in Nadler and Hong,
Political and Institutional Determinants at 7 (cited in note 75). Their analysis is based on data
from the twenty states for which daily yield spreads are publicly available. It extends a previous
analysis of the economic and legal determinants of state bankruptcy risks by Poterba and
Rueben, 50 J Urban Econ at 537 (cited in note 98).
134 Svea Herbst-Bayliss and Jonathan Stempel, Buffett: US Can Bail Out States, Insurers
Pained, Reuters (May 1, 2010), online at http://www.reuters.com/article/2010/05/01/berkshirebuffett-
ratings-idUSN0118355720100501 (visited Oct 9, 2011).
135 Ianthe Jeanne Dugan, Investors Looking Past Red Flags in Muni Market, Wall St J C1
(June 14, 2010).
2012] Freedom to Fail 277
TABLE 1. THE EFFECT OF UNEXPECTED DEFICIT SHOCKS,
POLITICAL, AND LABOR INSTITUTIONS ON CHANGES IN
STATE BOND YIELDS (MODEL 2).
(1) (2)
Δ Defshock 24.17 -11.02
(14.25) (11.51)
Δ Defshock × Union membership 2.02***
(0.61)
Δ Defshock × Dem. share in State Legislature 3.02***
(0.96)
Δ Unemployment rate 37.41 57.65*
(23.11) (27.48)
Δ Real GDP 131.5 -128.3
(177.0) (139.2)
Δ Deficit to GDP 8.598** 13.23***
(3.335) (4.267)
Constant -100.3*** -131.9***
(31.96) (40.42)
N 20 20
R2 0.702 0.723
Source: Nadler and Hong, Political and Institutional Determinants at *16 (cited in note 75).