This report marks the Volcker Alliance’s second annual assessment of US state budget practices. Covering the fiscal years 2016, 2017, and 2018, the study grades states’ success in pursuing transparent and fiscally sustainable procedures as they estimate their revenues and expenditures and attempt to keep them in balance not only at the start of the fiscal year but as it progresses.
Because the US federal system is composed of fifty sovereign states, it is essential to assess and compare the quality of their budget practices against a common set of standards. As we did in the 2017 report, we gave states grades of A to D-minus, the lowest possible mark, for their practices in five areas that are the building blocks of budgeting nationwide:
- budget forecasting, in which we evaluate how and whether states estimate revenues and expenditures for the coming fiscal year and the long term;
- budget maneuvers, in which we gauge dependence on one-time actions to offset recurring expenditures;
- legacy costs, in which we assess how well states are funding promises made to public employees to cover retirement costs, including pensions and retiree health care;
- reserve funds, in which we examine the condition of general fund reserves as well as rainy day funds and rules governing their use and replenishment; and
- budget transparency, in which we scrutinize disclosure of budget information, including debts, tax expenditures, and the estimated cost of deferred infrastructure maintenance.
In this report, we also compared states’ budgetary grades to marks given the year before and offer best practices in each of the five budget categories.
A state budget is as much a planning document as it is a financial one. The budget is intended to help a government remain fiscally sound and sustain its capacity to deliver services to the public. These goals may be unachievable if appropriate attention is not paid to estimating the possible future course of revenues and expenditures.
The use of a consensus method to estimate revenue is one of the best budget forecasting practices identified by the Volcker Alliance. The method attempts to avoid politically driven predictions by considering inputs from the executive and legislative branches and, sometimes, outside experts. Other best practices include providing a clear rationale to support forecasts and producing multiyear revenue and expenditure forecasts that can help policymakers spot and address long-term fiscal deficits. Yet only ten states received top average A grades in the budget forecasting category for fiscal 2016 through 2018, illustrating that most have considerable room to improve.
Among the findings:
- In 2018, nineteen states did not publish multiyear revenue forecasts in budget and planning documents covering at least three full fiscal years. That period is long enough to give policymakers a reasonable idea of potential fiscal trends.
- Twenty-seven states failed to present expenditure projections for the same period.
- Seventeen states had no multiyear forecasts for revenues or expenditures.
- Twenty-two states failed to use a consensus revenue estimate in the most recent year.
- Eight states did not provide a clear rationale to support their projections of revenue growth in the budget.


BUDGET FORECASTING
When States Make Midyear Budget Adjustments
States sometimes adjust their enacted budgets after the beginning of the fiscal year following an economic shock, such as a collapse in the housing market or drop in oil prices. Other times, however, midyear budgetary adjustments may be more a symptom of flaws in forecasting.
Though our 2017 report on state budgeting practices included the use of midyear budget adjustments in states’ annual forecasting grades, the 2018 report does not, as they usually show that policymakers are doing what is necessary to eliminate a shortfall. In fiscal 2017, sixteen states made such changes: California, Connecticut, Illinois, Iowa, Louisiana, Mississippi, Missouri, Nebraska, New Jersey, New Mexico, North Dakota, Oregon, South Dakota, Virginia, West Virginia, and Wyoming.
From July through October 2017 (the first four months of the fiscal year and the cutoff date for our 2018 report), only six states required midyear budget adjustments: Connecticut, Mississippi, Missouri, Montana, Nebraska, and Oklahoma. That number may not be a harbinger of the total number of states making adjustments in the fiscal year’s final eight months. Fiscal 2018 ended on June 30, 2018, for forty-six states. A final evaluation of budgetary processes covering the remainder of the fiscal year is scheduled to be published in 2019.
Mississippi, for example, with a three-year average of C in forecasting, made midyear budget adjustments in 2016, 2017, and 2018. In the first eight months of fiscal 2017, the state had a three-round series of cuts totaling $151 million. Part of this resulted from a $169 million overestimate of revenues and an accounting error made by legislators when writing the state spending plan. Among the agencies worst hit were the Mental Health and Health departments and the Mississippi Adequate Education Program.
California, meanwhile, did not require midyear budget adjustments in fiscal 2015 or 2016, or during the first four months of 2018. But a shortfall of $1.7 billion in general fund revenues in fiscal 2016 and 2017, stemming from overestimates of sales and corporation taxes, led to changes in the latter year. The shortfall was partly offset by upward revisions for personal income taxes.
Midyear budget adjustments may signal that projections were inaccurate, but they are not a bad practice in and of themselves. The frequent alternative is to resort to budget maneuvers, which have the effect of pushing shortfalls into future years, thus breaking a cardinal rule of good budgeting: matching a fiscal year’s revenues with expenditures in the same year.
It may be necessary to make midyear adjustments to avoid deficits and preserve bond ratings. But states should make them with care to avoid damaging effective programs.
Fiscal sustainability is key to successful budgeting. Using one-time actions to balance a budget will produce shortfalls in future years unless recurring revenues are increased or expenditures reduced. Similarly, deferring payment of planned expenditures to a future fiscal year can prompt a state to use maneuvers to pay for past-due bills.


BUDGET MANEUVERS
The Fund Transfer Trap
When Volcker Alliance researchers scrutinized states’ use of maneuvers to balance their budgets, the most widely used technique was making one-time transfers to the general fund from special funds to pay for recurring expenditures. Twenty-nine states engaged in this practice in fiscal 2018. While this indicates that these states are most likely caching spare cash in numerous places other than their official rainy day fund reserves, it poses a threat to the use of funds earmarked for specific purposes—say, clean energy retrofits or transportation. Additionally, reliance on fund transfers to balance budgets may not be repeatable indefinitely if the special funds are drained until they run dry.
Of the ten largest states by population, only California, Georgia, and Michigan eschewed such transfers in fiscal 2018. And while California and Georgia also avoided transfers in 2016 and 2017, Michigan broke ranks after the state Senate passed a bill permitting a one-time transfer to the general fund from the unemployment contingency fund, which is made up primarily of penalties and interest paid to the state because of fraud.
California has not always had such a clean record; it still carries liabilities from engaging in maneuvers in previous years. The state Department of Finance estimated that the balance of general fund liabilities to special funds was about $1.4 billion as of June 30, 2017.
Transfers involve significant sums in some states. New Jersey’s D average in budget maneuvers is largely due to its consistent use of this technique for all three years studied. In its fiscal 2018 budget, the state continued its longtime use of clean energy funds for general fund purposes, with the transfer of $161 million. There was a similar transfer of funds—totaling $204 million in the year—from the New Jersey Turnpike Authority to the general fund to help cover the operating expenses of New Jersey Transit, the state-owned commuter rail and bus system.
Similarly, Kansas shifted $198.4 million and $118.8 million in fiscal 2017 and 2018, respectively, to the general fund from the Pooled Money Investment Portfolio, which is made up of money from state agencies, local governments, and school districts and is invested by a state board. The funds are scheduled to be repaid over six years. With a three-year average of D in the category, Kansas has also made regular transfers of highway funds to the general fund, ranging from $173.5 million in 2015 to $288.3 million in fiscal 2018. The state may be able to lessen dependence on transfers and other budget maneuvers following a 2017 restoration of tax rates to the pre-2013 level.
In fiscal 2018, the New Mexico Senate raised recurring general fund expenditures by $90.1 million, using $81.4 million in transfers to cover most of that increase. Among the transfers, $71 million came from money that would have gone to the Severance Tax Permanent Fund, and $10.4 million came from the suspension of a severance tax bond distribution to the state water project fund.
Transfers from special funds are a particularly seductive approach to balancing a budget. They are nearly invisible without an exacting analysis of the budget and do not frequently show up in headlines. Still, they cannot go on indefinitely and may ultimately leave a state confronting a shortfall that can’t be fixed so quietly.
Trillions of dollars in unfunded liabilities for state public worker pensions and other postemployment benefits (OPEB), largely retiree health care, represent unpaid costs for services that governments delivered in the past. Efforts to cover these legacy costs present many states with challenges that often elude easy solutions, and the costs increasingly threaten to crowd out spending on education, infrastructure, and other critical needs.
While a 19 percent gain for the Standard & Poor’s 500 Index in 2017 helped state pension funds reduce their deficit by a small amount from the previous year, they still had a $1.35 trillion gap, according to data compiled by Bloomberg. This deficit was 6.5 percent larger than it was in 2015. Put another way, states in 2017 had set aside only $2.95 trillion to cover pension obligations totaling $4.23 trillion. On top of their pension obligations, states posted OPEB liabilities of $692 billion in 2016, the most recent year for which data are available, while amassing just $46 billion in assets. Perhaps it is not unexpected that in the legacy costs category, only eight states received average A grades for 2016 through 2018, while twenty-two were graded D or worse.


LEGACY COSTS
How States Created Today’s Pension Funding Gaps
State pension system shortfalls are often the result of decisions made years or decades ago.
In 2017, twenty-one states had pensions that were less than 70 percent funded. That stands in contrast to 2000, when half of state plans were at least 100 percent funded. Many states’ pension plans saw funding levels drop when technology stocks declined dramatically in 2000. The Great Recession had an even larger impact. Many states increased retirement benefits in the 1990s and early 2000s in lieu of salary increases, but not all of those raised government or worker contributions sufficiently to finance the additional largesse.
From 1997 to 2006, thirty-two states failed to make their full or nearly full actuarially determined or recommended annual pension contribution in at least one year. The shortfall left the underpayment to be addressed subsequently and the unpaid sums to compound at the funds’ assumed rate of return—currently about 7.5 percent. From 2016 to 2018, the years covered by this study, sixteen states missed making the full contribution in at least one year.
Kentucky and New Jersey are extreme examples of a precipitous drop in pension funding ratios. Both were at least 100 percent funded at the end of the twentieth century, but as of June 30, 2017, the unfunded liability for Kentucky’s pension systems was $42.9 billion, while New Jersey’s was $142.3 billion. That translates into a funding ratio of 33.9 percent and 35.8 percent, respectively.
New Jersey began steadily losing ground in its Public Employees’ Retirement System and Teachers’ Pension and Annuity Fund at the start of this century. But the state’s problems began even before that, when Governor Christine Todd Whitman signed into law the final part of a 30 percent income tax cut in 1995.
To make up for the revenue decline, the governor reduced the amount of money contributed to the pension funds and in 1997 sold $2.8 billion in pension obligation bonds at just under 8 percent interest. A plunge in stock prices between 2000 and 2002 consumed part of the funds raised through the debt sale, and New Jersey will continue to pay about $500 million annually in debt service costs for those bonds through 2029.
Despite the market losses, the state boosted pension benefits by 9 percent in 2001—without any plans for covering the additional cost. In fiscal 2018, New Jersey temporarily shifted ownership of the state lottery and its proceeds to the retirement funds for teachers, public employees, and police officers and firefighters. The move made the pension system appear better funded, but the net amount being injected by the state is not scheduled to reach the ARC level until fiscal 2023. The shortfall continues to deprive the pension system of any possible earnings on the sums that actuaries say should be contributed.
Kentucky has a similar story. In 2002, the Kentucky Retirement Systems, comprising five separate pension plans, was 100 percent funded. But starting in 2004 annual contributions ran consistently short of actuarial recommendations. The biggest problem was the state employee plan for those in nonhazardous jobs. At the height of the ARC underfunding, in fiscal 2012, the employer annual contribution for that plan fell short by $226.3 million—half of the actuarial recommendation. The legislature passed a bill in 2013 mandating full annual contributions for the Kentucky Retirement Systems. In the biennial budget that began July 1, 2016, and ended June 30, 2018, the state exceeded the ARC.
The state is still grappling with underfunding that stems from past actions. These actions include using an accounting process that allowed the state to backload its ADC so that costs would grow over time, using assumptions that turned out to be overly optimistic, and providing annual cost-of-living increases to retirees for decades without adequate plans for funding them. This occurred most recently in 2011. Two years later, legislators eliminated cost-of-living adjustments (COLAs) except in years when the assets of the system are greater than its liabilities. Still, just between 2008 and 2011, unfunded COLAs added $1.45 billion to the unfunded liability of the Kentucky Retirement Systems.
Recessions, commodity price swings, and natural disasters can wreak havoc on state budgets. Because states lack the federal government’s ability to print money, they try to maintain reserves, usually known as rainy day funds, as their bulwark against crisis-driven budget shortfalls.
Although a few states lack rainy day funds or have ones that are practically empty, the economic recovery that began in 2009 has helped most states restock their vaults with cash. At the end of fiscal 2018, the median state had reserves of 5.8 percent of general fund expenditures, the largest cash trove since rainy day funds hit a recent low of 1.9 percent in 2011, according to the National Association of State Budget Officers.


RESERVE FUNDS
Building Revenue Volatility into Rainy Day Fund Rules
For at least three decades, many states have followed a general rule that rainy day fund reserves should equal about 5 percent of the general fund balance. While the origins of this rule remain uncertain, it has become clear that it makes little sense to hold all fifty states to the same reserve standard without considering their individual revenue structures.
Largely due to advocacy efforts by the Pew Charitable Trusts, a growing number of states have acknowledged this issue and are tying their rainy day fund goals to the volatility of their revenue streams. Nineteen states follow this practice, thanks in part to recent legislative momentum. In 2017, Hawaii, Maryland, Montana, New Mexico, North Carolina, and North Dakota changed rainy day fund policies to increase their consideration of revenue volatility.
States with more or greater swings in revenue are more likely to need larger reserves. The Great Recession clearly showed states that they were unprepared for the extreme drops in revenue that were in part attributable to a highly volatile revenue stream.
For example, in November 2016, Maryland’s comptroller, and departments of Budget and Management and Legislative Services released a joint study looking at the state’s volatile revenue structure and recommending changes to its reserve fund policies. A bill passed in the 2017 legislative session that takes effect in 2020 will direct a portion of capital gains and other non-withholding income tax revenue to Maryland’s rainy day fund and, when reserve fund caps are reached, to a newly created Fiscal Responsibility Fund. The latter can be used for pay-as-you-go capital projects or allocated to state trust funds with unfunded liabilities.
One group of states with particularly fragile revenue streams are western ones, including Alaska, Montana, North Dakota, and Wyoming, that rely on severance fees and taxes on the production of oil, gas, coal, or other natural resources to help pay for services. Understandably, all four states connect funding of reserves to revenue volatility. The North Dakota legislature changes the rate of contribution to its fund based on annual revenue collection, which is closely connected to oil prices.
Some states that tie their rainy day fund policies to revenue volatility also cap the amount they are allowed to contribute, a practice that may partially negate the purpose of the volatility link. Virginia, for one, uses historic revenue growth in its major tax categories as an important factor in making decisions about its rainy day fund. This was a successful formula in the mid-2000s, driving large deposits during a time of economic expansion. But the state had a 10 percent cap on total deposits, a point it reached in fiscal 2006 and 2007. The reserve enabled the state to cover just 15 percent of the shortfalls that occurred between 2008 and 2010. Virginia voters in 2010 approved a measure raising the cap to 15 percent, which would have given it an additional $594 million to face revenue losses and spending demands during and shortly after the Great Recession.
There is a good reason a growing number of states are tying the amounts held in rainy day funds to volatility rather than keeping a set percentage of general fund revenues on hand. The practice helps avoid salting away too much cash in states that have a less volatile tax structure or skimping on reserves in states likely to experience more revenue swings.
Budget information is worth little to elected officials, policy advocates, and the public if they can’t find it. Yet only three states won top average A grades for fiscal 2016 through 2018 for budget transparency. The lack of comprehensive budgetary information on the cost of deferred infrastructure maintenance in forty-six states explains part of the result, but many states also trail in other critical areas.
In addition to the infrastructure disclosure, a state budget transparency strategy should include a consolidated website that provides easy access to budget procedures and timing. It is also important to post links to information needed to form spending plans, such as reports on unfunded liabilities for long-term pension or retiree health care, capital budgeting, economic forecasts, or reserves. To follow best practices, states should also produce clear, accessible, and detailed tables of outstanding debt and debt service costs, and an annual or biennial accounting of the cost of tax exemptions, credits, and abatements. The good news is that thirty-eight states received a B average for the three-year period, showing reasonable efforts at disclosure in these areas.


TRANSPARENCY
The Ins and Outs of Tax Expenditures
Federal, state, and local governments devote billions of dollars in public resources to tax expenditures. These include exemptions, deductions, credits, and other exclusions from levies that would otherwise be paid by individuals or businesses.
While many such expenditures by states are focused on economic development and jobs, they also may be aimed at reducing the cost of basic consumer necessities such as food and clothing, subsidizing low-income senior citizens or veterans, or bringing deductions in line with those in the federal tax code. While forty-two states provide at least some regularly updated information on tax expenditures, their reports vary in scope.
Tax expenditure reports should be used to measure the costs of the abatements against their benefits, thus helping government officials and policy advocates evaluate which programs should be curtailed, maintained, or expanded. They should be produced annually or biennially and be publicly available on budget websites, or on easily accessible transparency or revenue department sites.
A useful state report includes comprehensive data pertaining to all major streams of tax revenue, including personal and corporate income taxes, sale and use taxes, real and personal property taxes, and excise and gross receipts taxes.
Georgia, which has a three-year B average for transparency, has one of the more complete reports. Published annually, it is available on the website of the Governor’s Office of Planning and Budget. It does not provide the total value of tax expenditures but does show estimates for three fiscal years for hundreds of credits, exemptions, and deductions for individuals and companies. The comparative fiscal year data enable readers to see how the amounts involved change from year to year. For instance, the fiscal 2018 report shows that the film tax credit cost the state $414 million, up 22 percent from 2016.
Ohio is another state winning a B. Its comprehensive tax expenditure report is published as Book Two of the biennial budget and can be found on the Office of Budget and Management website under Operating Budget. The report for the 2018–19 biennium, released in November 2016, includes estimates for fiscal 2016 through fiscal 2019. It lists $9.1 billion of tax expenditures in fiscal 2018, about 28 percent of general fund spending for that year. The state’s earned income tax credit for those making less than $10,000 annually cost the state $3 million in fiscal 2018. A separate job creation credit for businesses cost $113 million that year.
Although we gave credit to any state that issued comprehensive tax expenditure reports regularly, individual reports provide varying levels of detail. Ohio estimates credits for upcoming budget years, and Colorado has historical reports. Maryland provides narrative descriptions for overall categories of tax expenditures rather than for individual ones, while Arizona provides very detailed descriptions of each provision.
Differences in how states compile tax expenditure reports make it difficult to compare them. But within each state, the reports provide crucial information on the expenditure of resources via tax breaks and on the individuals and industries that benefit. While the reports themselves do not generally provide information on whether tax expenditures achieve their statutory purpose—creating jobs is a common one—they provide an important tool for further analysis.
