Top
Reference

State and local government debt

This page tracks outstanding debt of US states and local governments. The main question is the sustainability of current trends.

Summary

5 Jan 2014.

Explicit state and local government debt is in the neighborhood of $3 trillion, and unfunded retirement obligations add a similar amount. The sum is something like three times aggregate income. So, if the obligations were spread evenly, and if politicians were to begin grappling with the problem immediately, this would be a serious but manageable problem. Unfortunately politicians continue to delay facing the issues, and the obligations are not spread evenly, so sporadic state and local governments insolvencies, attempts to wiggle out of retirement obligations, and periodic cutbacks in services, are likely to be with us for the long term.

Graph

5 Jan 2014. Data through 2011.

Outstanding debt is from the Census Bureau (see Sources); total receipts are from the NIPA, specifically table 3.3.

Highlights

Census accounting conventions (19 Jul 2010) Debt as counted by the Census Bureau includes most classes of debt that one would think of – for example short-term interest-bearing obligations as well as long-term bonds, and revenue bonds as well as general obligation bonds. But the following are among the items not included in the total debt reported:

  • Accounts payable (so, e.g., Illinois' $4 billion in delayed vendor payments do not show up)
  • Loans from the savings of one branch of government, to another branch, that then spends them (even though the money is gone and will have to be raised again later)
  • Sales of state assets that are then leased back (even though the rent is a future obligation that did not exist before the sale)
  • Borrowings from the US Treasury to cover shortfalls in unemployment compensation
  • And, most importantly, unfunded retirement obligations



Delayed vendor payments a small part of overall debt (19 Jul 2010) In 2010 California and Illinois were the two states with the greatest probability of default. In 2009 California delayed $5.6 billion in payments, an amount equal to about 3% of 2008 revenues. In 2009 Illinois delayed $3.9 billion in payments, an amount equal to about 7% of 2008 revenue.


Local government retirement funds (18 Nov 2010) Most studies of unfunded state and local government retirement obligations concentrate on state plans, sometimes including a small representative sample of local plans. This is necessary because, as of the 2008 Census Bureau inventory, there were 2332 local government pension plans, and it is more than challenging to track down data on all of them.

Local plans, though numerous, claim only 10% of the combined state and local pension plan members, and 16% of assets. As a consequence, it suffices to have a detailed knowledge of the condition of state plans, and a more cursory knowledge of local plans. Most studies simply assume that local plans are more or less in the same condition as state plans. Is this assumption justified?

For pensions, the overall condition of local plans does indeed seem to be very similar to that of state plans. A 2008 Boston College study, using 2006 data, looked at 84 local government plans across 38 states, covering about 58% of local plan assets. For plans providing data on unfunded obligations, this study found an average funding ratio of 85%, compared to 84% for state plans. To bring contributions for all 84 of the local plan contributions up to the level of the Actuarial Required Contribution would have required an adjustment, in aggregate, of 1.6% of payroll, compared to 1.8% for state plans. (Note: that was before the crash; the numbers are worse now.)

A 2010 study by Rauh and Novy-Marx, extrapolating from 77 of the largest local plans, found that the unfunded pension liability at the local level was, as a percentage of revenue, almost precisely the same as that for the states.

There is much less data on retirement health care obligations. One Government Accountability Office study of retirement health care benefits, including all state plans and 39 large local government plans, found that the unfunded liability of the local plans was $129 billion, 24% of the combined $530 billion for both state and local. These 39 plans probably covered less than half of all local government plan members. If we double the $129, just to get a rough idea of what the total might be for all local plans, this would bring the local amount to about 30% of the combined total unfunded obligations. Given that local pension plans only have 10% of combined state and local pension plan members, this suggests that local retirement health care plans are in considerably worse shape than state plans.

This conclusion is supported by data on at least one state. In a study of all New York health care retirement plans, the two state plans had, together, unfunded liabilities of $60 billion, while all local plans had, in aggregate, unfunded liabilities of $145 billion – almost two and a half times greater. In contrast, Census data show that in 2008 state pension plans had assets of $277 billion and local pension plans had assets of $116 billion, suggesting that, as in most states, the state plans have far more members.


State retirement funds (23 Oct 2010) Unfunded obligations of state retirement funds, including both pensions and retiree health care, are in the neighborhood of $3 trillion. This is something like two years' worth of aggregate state income. On average, then, this is a heavy, but not crushing, debt load. Because politicians have delayed facing the issue, annual expenditures will rise. One reasonable projection puts pension plan contributions needing to increase by about 5% of aggregate budgets. Health plan expenditures are less well characterized, but might add another 2%, bringing the total adjustment to 7% of state budgets. The difficulties in both pension and retiree health plans are spread very unevenly. Some states face minor adjustment, and some will likely face adjustments of perhaps twice the average. For comparison, the drop in state tax revenues due to the recent crisis averaged about 10%.


Unemployment compensation (5 Jan 2014) States are meant to build up reserves in the UI Trust Fund, but reserves as a fraction of covered wages have been falling for decades. Reserves were insufficient for unemployment payments in the great recession, and as of 1 Apr 2010, states carried loans of $39 billion from the US Treasury to cover unemployment compensation. As of 3 Jan 2014, the total amount of loans to the state was $21bn.


Sources

The only attempt to provide a comprehensive view of state and local debt is in the Census Bureau's government survey data. See

Note that the Census Bureau covers only defined benefit pension plans, not defined contribution plans or post-retirement healthcare.

See also

Selected commentary:

  • 23 Aug 2010. State asset sales not a small issue, but not the largest States and cities are selling and leasing back property. This is, in essence, borrowing. The scale is probably in the tens of billions of dollars per year. Although not a small amount in absolute terms, it is small compared to unfunded pension liabilities, which remains the big issue.
  • 27 Oct 2010. Unemployment insurance borrowing is not a key state problem States have had to borrow from the federal government in order to pay unemployment insurance claims. These borrowings have made the news now and then, adding to a long list of worries about the states. But the unemployment insurance trust fund problem, though discouraging, is relatively small and not worth too much attention from investors.
  • 30 Oct 2010. States stiffing vendors Based on a quick look at four states with well-known budget problems, it would appear that Illinois is the only state were delayed payments are a sizable fraction of the state budget. And even there, delayed payments are a small fraction of unfunded retirement obligations.
  • 2 Nov 2010. State and local governments take the lump sum option In several different ways state and local governments are taking lump sum payments now and, in exchange, either losing a stream of regular future revenue or agreeing to pay regular future rents. While this does provide yet another example of disappointing budget gimmickry, it is not, quantitatively and in aggregate, a major problem. Looking at (1) securitization of tobacco settlement revenue, (2) leasing of toll roads, parking systems and airports, and (3) selling buildings and leasing them back, I estimate a total of about $70 billion is involved. This is very small compared to either annual state and local government budgets ($2.1 trillion) or unfunded state retirement obligations ($3 trillion).
  • 10 Nov 2010. Another creative way to hide government debt A trick to hide debt, used by many municipal, regional, and sovereign governments in recent years, was to disguise loans by incorporating them within interest rate hedges. This was one of the key ways Greece hid its true deficit for so long. Now evidence has come out that the same tricks were used by cities and states in the US. Although quite ugly, this is not, however, one of the big problems facing state and local governments, at least in aggregate.
  • 21 Dec 2010. Quasi-governmental debt from "independent authorities" is substantial State and local governments often set up independent authorities that issue debt to fund utilities, sports and entertainment centers, and other projects. While governments often have no legal obligation to stand behind such debt, a default would lower public ratings, and so there is, in practice, some level of obligation. The amount of such debt outstanding is on the order of $500 billion. It is extremely difficult to judge what fraction of that debt might end up defaulting and coming back to the governments that set up the issuing authorities. What one can say is that this is the one area of hidden state and local government debt that might be, if not as large as primary debt and unfunded retirement obligations, at least on the same scale.
  • 10 Jan 2012. US state and local government debt, as a percentage of income, grew substantially after the crisis. State and local government debt began to grow faster than income after the 2000 recession, and the trend accelerated in 2008 and 2009. This is a long-term problem, not a short-term crisis, but it is important to keep an eye on (1) in deciding where to live, and (2) in choosing municipal bonds.
  • 5 Jan 2013. State and local government debt worsened slightly in 2010 As a reminder, here is a short summary on the main features of state and local government debt in the US …

Clippings below were used in the construction of this page

Covered through 18 Jul 2010.

Treatment of debt by Census

2005 (?). Classification Manual.

http://ftp2.census.gov/govs/class06/ch_6.pdf (reached via http://www.census.gov/govs/classification/)

“Indebtedness”

“This edition of the Classification Manual contains revisions to the classification system used by the Census Bureau for government debt statistics. The revisions were the result of the 2005 redesign of the government finance statistics program. In keeping with the theme of the 2005 redesign, revisions were made to enhance internal consistency in statistical categories and to simplify the classification categories across levels and types of governments. To meet this objective, the number of categories used to classify state and local government regular debt statistics was reduced from sixty-six to eight, with the former categories combined into broader groupings.

Even though the debt classification categories were simplified, there was no change to the basic concepts and coverage, or to the overall definitions. Rather, the revisions represented a simplification of the classification structure, applied to the manner in which the statistics are presented and reported. The 2005 changes to the classification system for debt statistics are in sharp contrast to the revisions made to government debt statistics that took effect with the 1988 government finance program. …

6.1.1 Definitions – Debt Instruments

Government debt comprises all interest-bearing short-term credit obligations and all long-term credit obligations incurred in the name of the government and all its dependent agencies. This definition includes all debt, whether backed by the government’s full faith and credit or nonguaranteed. It includes tax-exempt as well as taxable public debt.

Government debt also includes judgments, mortgages, “revenue” and “earning” bonds, and special assessment obligations, as well as the more traditional general obligation bonds, notes, and interest- bearing short-term warrants. It includes not only government debt for public improvements (roads, sewers, airports, etc.), but also debt issued for the direct benefit of the private sector entities (industrial development, mortgage revenue, pollution control and abatement, etc.).

6.1.2 Liabilities Outside Census Bureau Definition

The concept of government debt does not cover every type of liability listed on the balance sheets or elsewhere in government finance reports. The following types of liabilities are excluded from Census Bureau statistics:

  • Non-interest-bearing short-term obligations, such as accounts payable, non-interest-bearing warrants, and the like. Note, however, that non-interest-bearing long-term obligations having a formal debt instrument (such as a fixed repayment schedule) are classified as government debt.
  • Interfund advances, loans, or other obligations transacted between accounting funds of the same government (see Section 3.9 for a discussion of intragovernmental transactions). As a practical matter, these debt statistics do include formal debt instruments of a government that are held as investments by its own agencies or fund on behalf of others.
  • Amounts held and owned by a government in a trust or agency capacity on behalf of others (see Section 3.10).
  • Contingent loans and advances from other governments, which are classified as intergovernmental transactions (see Section 3.7).
  • Rights of individuals to benefits from retirement funds or other social insurance trust systems.
  • Unfunded liabilities of retirement systems, unpaid annual or sick leave, or other obligations without a formal debt instrument specifying terms, length, interest rate, etc.
  • Leases, including both capital and operating leases (see Section 6.4.5).
  • Loan guarantees and nonguaranteed obligations of the Federal Government.

Chart 6.A provides a sample of the most common types of government liabilities that are included and excluded from the Census Bureau’s definition of public debt.”

[See also pages 4-52 and 4-62 in the full classification manual, which together make it clear that advances from the Treasury for unemployment compensation are treated as revenue, and paying back these advances is treated as negative revenue.]

Unemployment insurance loans from the US Treasury

6 May 2010. GAO report to Congress, #GAO-10-440.

http://www.gao.gov/cgi-bin/getrpt?GAO-10-440

“UNEMPLOYMENT INSURANCE TRUST FUNDS: Long-standing State Financing Policies Have Increased Risk of Insolvency.”

“Extended benefits (EB) are provided to workers who have exhausted regular unemployment insurance benefits during periods of high unemployment.6 The EB program is financed in approximately [equal] shares by the states and the federal government. …

The UI program was designed to be forward funded and self-financed by states, with each trust fund building up reserves from employer taxes during periods of economic expansion in order to pay UI benefits during economic downturns. Because unemployment can vary substantially during a business cycle, it is important that states build sufficient trust funds to remain solvent during recessionary times. The program is financed primarily by taxes levied on employers. Each state sets UI tax rates to finance regular UI benefits. In addition, employers pay a Federal Unemployment Tax Act (FUTA) tax. The FUTA tax on employers is 6.2 percent on the first $7,000 of each employee’s annual pay. Employers in states whose UI programs comply with federal requirements receive a tax rate credit of 5.4 percent, resulting in an effective rate as low as 0.8 percent, or a maximum of $56 per worker per year. The FUTA tax is used to fund:

  1. federal and state UI administration costs;
  2. the federal share of EB;
  3. Title XII loans to state trust funds when they cannot pay benefits;
  4. benefits under federal supplemental and emergency programs;
  5. labor exchange services, employment and trainingf or veterans; and
  6. some labor market information programs.

States choose both a taxable wage base, the annual earnings per worker on which employers pay UI taxes, and statutory tax rates that apply to the base. …

The Unemployment Insurance Trust Fund (UTF) in the U.S. Treasury consists of 53 state accounts, including one each for the District of Columbia, the Virgin Islands, and Puerto Rico, plus 6 federal accounts that are dedicated for special purposes. Federal taxes go into the Employment Security Administration Account (ESAA), the Extended Unemployment Compensation Account (EUCA), and the Federal Unemployment Account (FUA), and state taxes go into their individual state accounts …

As UI is forward-funded, states collect trust fund reserves in advance to pay benefits. However, during exceptional periods when states exhaust their UI reserves, they may borrow from the federal government. …

ETA reports reserve ratios, or UI trust fund levels as a percentage of total annual statewide [covered] wages, as well as high cost multiple (HCM), which divides the reserve ratio by the high cost rate, the highest historical ratio of benefits to wages for a 12-month period in that state. An HCM of 1.0 corresponds to sufficient reserves to pay benefits at the high cost rate for 1 year. …

By any measure, UI trust funds nationwide are in historically poor financial condition. As of the fourth quarter of 2009, reserves in state trust funds totaled $11.1 billion, lower than any end-of-year total (inflation adjusted) in the program’s history and down sharply from the $30.0 billion in aggregate reserves at the end of 2008. Reserve levels look even weaker when one considers that fund levels are buoyed by federal loans, which surged during 2009 and continue to grow. As of April 1, 2010, 34 state trust funds had taken out federal loans totaling $38.9 billion; this total loan balance is up almost 50 percent since just December 31, 2009. By comparison, 24 states required loans during the recession of the early 1980s, during which unemployment nationally approached 11 percent. Aggregate net reserves (reserves less loans) as of December 31, 2009, measured -$15.4 billion, the first such deficit since the end of 1983 and the lowest level in the program’s history. …

A 2009 state survey found that 35 states would increase their UI taxes on employers in 2010, with … a median projected contribution level increase of 27.5 percent. …

DOL has projected that trust fund account balances, net of loans, will fall to -$88.4 billion at the end of fiscal year 2012 before starting to grow again, with net balances not becoming positive until well beyond fiscal year 2014. DOL anticipates that the number of outstanding UI loans that states have from the federal government will increase until 2012, when they could total $90 billion. Employers in states that cannot make their loan payments within the required 2-year period could lose some of their FUTA tax credit and pay increasing tax rates each year until the loan is repaid. …

from 1938 to 1973, state UI trust funds held average year-end reserves, net of loans, equal to 5.1 percent of wages, and never dropped below 2 percent. From 1974 to 2008, that average fell to 1.0 percent of wages and has never been as high as 2 percent. Therefore, states have had less of a financial buffer in their trust funds to withstand a high-cost benefit period. Prior to the recent recession, the aggregate HCM nationwide was only 0.35, corresponding to enough reserves for about 4 months of benefits at a high-cost rate; therefore even a much milder recession was likely to have caused widespread trust fund insolvency …

UI tax contributions as a percentage of UI-covered wages have trended downward in recent decades, from an annual national average of 1.15 percent (1979 to 1988) to 0.79 percent (1989 to 1998) and in the past decade to 0.65 percent (1999 to 2008).”

States with the biggest loans, and total loans across all states (millions of dollars):

State 31 Dec 2009 balance 1 Apr 2010 balance
California 5,984.7 8,456.5
Michigan 3,159.1 3,797.1
New York 2,160.2 3,032.0
Pennsylvania 1,871.5 2,814.3
Ohio 1,727.9 2,229.5
Illinois 1,168.5 2,081.1
North Carolina 1,606.7 2,138.7
Indiana 1,488.6 1,807.7
Texas 1,322.5 2,035.0
US total 26,470.2 38,881.4

Delayed vendor payments -- Illinois

11 Jun 2010. BusinessWire.

http://www.businesswire.com/portal/site/home/permalink/?ndmViewId=news_view&newsId=20100611005969&newsLang=en

“Fitch Downgrades Illinois' GO Bonds to 'A' from 'A+'”

“The state's general fund cash position is weak and the state is increasingly relying on short-term borrowing and deferring vendor payments to provide liquidity. Accounts payable at the end of the 2009 fiscal year were approximately $3.9 billion, equal to more than a three-month delay in payments. This compares to June 30, 2008, when there was $975 million in accounts payable. …

Longer term challenges remain, in particular, continued significant growth in funding requirements to address the pension systems' large unfunded liabilities. As of June 30, 2009, the unfunded accrued actuarial liability was reported at $62.4 billion, resulting in a 50.6% funded ratio, a level which reflects the use of five-year smoothing for the first time.”

[So delayed vendor payments are only a few percent of unfunded pension liabilities.]

[For another comparison, total state revenue in 2008, according to Census, was $55bn; so delayed vendor payments were about 7% of 2008 revenue.]

As of Jun 2010, CA and IL were the states with greatest risk of default

22 Jun 2010. CreditWriteDowns.

http://www.creditwritedowns.com/2010/06/illinois-default-woes-pass-california.html

“Stat of the day: Illinois takes over from California as state most likely to default. Posted by Edward Harrison”

Delayed payments, California

18 Jul 2010. Sunshine Review.

http://sunshinereview.org/index.php/California_state_budget

“California State Budget”

“The California controller delayed payments in February 2009 and issued IOUs in July and August 2009. The February and July budget packages were not enacted early enough to prevent the Controller from: (1) delaying over $3 billion of scheduled payments (mainly tax refunds) in February 2009 and (2) issuing 449,000 registered warrants (also known as IOUs) for a total of $2.6 billion of payments in July and August 2009. The February 2009 delayed payments generally were paid in March 2009, and the IOUs were able to be redeemed by recipients beginning on September 4, 2009. This was only the second time since the Depression that the state issued IOUs for some of its budgeted payments. In effect, the IOUs forced recipients (such as state vendors and local governments) to provide the state with a loan involuntarily. The IOUs were redeemable with interest, paid at a 3.75 percent annual rate. “Priority payments”—including school, payroll, and debt service payments—were not subject to IOUs”

[By way of comparison, California revenues, in 2008, according to Census, were $194 billion. So the delayed payments described above were about 3% of 2008 revenue.]

Clarification on what Census covers

21 Oct 2010. email exchange.

I asked:

“From the definitions page, it seems that the “State and Local Public Retirement Systems” data covers only pension plans, not defined contribution or post-retirement healthcare plans. But I see that the supplemental form at the state level also asks for data on the latter two. So what, exactly, is included in the main summary data?”

And Erika H. Becker-Medina, Chief, Employment & Benefit Statistics Branch, answered,

“You are correct, the State and Local Public Employee Retirement Systems summary tables only include data for defined benefit plans - not defined contribution or post-employment healthcare plans. The supplemental form that you are referring to is part of the new data collection that we are still testing. Therefore, we have not published those results.”

Overview of arguments that government workers are not overpaid

15 Nov 2010. Rortybomb.

http://rortybomb.wordpress.com/2010/11/15/are-government-employees-overpaid-still-no/

“Are Government Employees Overpaid? Still No”

Update on Illinois stiffing vendors

5 Mar 2012. FT.

http://www.ft.com/intl/cms/s/0/eec9b472-6285-11e1-872e-00144feabdc0.html?ftcamp=published_links/rss/companies_financials/feed//product

“Illinois state ‘on brink of collapse’. Hal Weitzman, Nicole Bullock”

”… debts totalling more than $280bn, an $86bn hole in its public pension funds and a $9bn backlog of unpaid bills. …

To stem the growing shortfall in its public pension funds, Illinois pledged to pay its full pension obligations – $5.2bn next year – without borrowing, something budget watchdog groups have long urged it to do. …

Illinois – which has the most underfunded public pension system in the US – could be a leading indicator for a pension crunch that will hit other cash-strapped states. “We’re going to see more and more states cutting spending to meet pension obligations mandated by the state’s own statutes,” says Prof Rauh. …

Illinois is only remaining solvent by not paying its bills on time, a practice that hurts thousands of small contractors that do business with the state. When payments do arrive, they are often incomplete. If Illinois continues this practice, it will rack up $35bn in unpaid bills by 2017, according to the Civic Federation, a Chicago think-tank. …

Prof Rauh reckons that ultimately, fiscal discipline will only come when the bond market demands it. Moody’s, the rating agency, downgraded Illinois in January to the lowest credit rating of any US state. Yet less than a week later, the state auctioned $525m of general obligations bonds at 3.9 per cent, the lowest rate on such a bond in recent history.”

State of the Illinois payments backlog

9 Apr 2012. Bloomberg.

http://www.bloomberg.com/news/2012-04-09/illinois-is-pension-basket-case-you-forgot-about.html

“Illinois Is Pension Basket Case You Forgot About. Josh Barro”

“Starting after the 2010 election, with Blagojevich gone, the Legislature did raise the income tax, from a flat 3 percent to a flat 5 percent. But Illinois’s budget situation was so dire that even a 66 percent increase, which raised general fund revenues by about 20 percent, still left an annual gap of more than $1 billion.

In 2010, Republicans gained seats in the Legislature and have blocked any further general-obligation bond issuances to close budget gaps. As a result, the unpaid bills backlog has reached more than $8 billion, and there is no plan to pay it off. Because of the tax increase, at least the backlog is no longer growing very much, and vendors can expect to be paid on a rolling basis after waiting about five months. ”

GASB likely to adopt more realistic accounting for state pensions

23 Jun 2012. FT.com.

http://www.ft.com/intl/cms/s/0/503ccc36-bd5c-11e1-838c-00144feabdc0.html

“Accounting change bloats US pension gap. Nicole Bullock and Hal Weitzmanin”

“new accounting standards likely to be approved on Monday …

The new standards “will better reflect the economic reality,” said Robert Attmore, chairman of the Governmental Accounting Standards Board (GASB).

One GASB proposal would affect how public pension funds report their assets. Currently, funds measure their assets using “smoothing”, a controversial accounting technique that has enabled governments to spread losses over several years and prevent reported assets from fluctuating along with their market values. The GASB proposal would require them to mark to market annually.

“That change is going to be fairly dramatic,” said Laura Quinby, research associate at the Center for Retirement Research at Boston College. “Currently actuarial assets are higher than market assets because the full losses from 2008-9 have not been phased in yet.”

Another GASB proposal could result in higher reported liabilities. Currently, public pension fund liabilities are typically discounted by 8 per cent annually – the projected long-term rate of return on investments for most public pension funds.

Under the new standards, states would project how long current assets, expected returns and contributions will cover expected benefit payments. For that period, they will be permitted to use their own discount rate. Beyond that, states would have to assume the much lower rate of a municipal bond – currently around 3 per cent.

The measures could enlarge the current shortfall in public pension funds from 24 per cent to 43 per cent, according to the Center for Retirement Research, pushing overall liabilities from $3.4tn to $4tn. That assumes that states continue to contribute to public pension funds in the same way.

The GASB standards if approved would begin to take effect from June 15, 2013.”