Is a silent economic crisis lurking and growing in the balance sheets of the 89,000 state and local governments in the United States? For decades, states, municipalities, and towns have been underfunding their pension liabilities, resulting in a gap between the promises made to future retirees and the funds put aside to meet those promises that some analysts estimate could total $4 trillion. That's 27% of GDP. Closing that large a gap could have a severe impact on the U.S. economy.
The rules for such funds allow governments to set the expected rate of return on money allocated to pensions, and many states and cities have made overly optimistic predictions. Significant losses could also be hidden for a time through the delayed reporting allowed by income smoothing.
In June 2012, the Governmental Accounting Standards Board (GASB) approved new standards for public employee pension accounting. The rules, which will roll out over the next two years, seek to increase transparency and bring public-sector accounting standards more in line with those used in the private sector. The regulations don't impact the amount that should be set aside for pension funds, but they alter the way that pension obligations are recognized and disclosed.
In an email Q&A, Rick Antle, the William S. Beinecke Professor of Accounting at Yale SOM, and Stanley J. Garstka, professor in the practice of management at Yale SOM, explain the new rules, give context to the public pension funding crisis, and discuss why it may not be the biggest funding crisis facing states and municipalities.
Q: What are the challenges in estimating pension liabilities? Why do estimates of the gap range from $1 to $4 trillion?
Pensions have a few basic features that combine to make estimating pension liabilities a tricky business. Pension obligations are large, stretch over a long time, and are impacted by economic variables that are difficult to predict. Pension accounting and pension finance calculations therefore make many assumptions, and the amounts of dollars involved mean that the ultimate estimates are very sensitive to these assumptions.
For example, one important variable is the assumed rate of return on assets invested to fund future pension obligations. A second is the interest rate used to discount the estimated future pension obligations to compute the present value of the pension liability. Some estimates use a discount rate equal to the expected rate of return on the assets invested to fund these obligations. Other estimates use (correctly!) a lower rate to discount the future obligations, since these obligations are risk free to the retirees (the government contractually guarantees to pay the pension). Since the future obligations are large in absolute terms, the present value of the obligation is much larger using a lower discount rate. Hence the amount currently in the pot of assets to fund the obligation is much smaller relative to the present value of the obligation—the unfunded gap is larger.
Although the funding problem has been brewing for a while now, the financial crisis of the past several years has both exacerbated and raised awareness of the financial difficulties faced by states and municipalities (decreasing revenue streams and fixed, sticky costs). Since one of the largest government expenses is for pensions, it logically received scrutiny. Given the magnitudes of the numbers, this scrutiny quickly leads to a closer examination of the underlying assumptions of the various calculations.
Q: How do you describe the new rules to non-accountants?
The new accounting rules for pensions affect three areas of accounting: recognition, estimation and disclosure.
Recognition refers to what must be formally included in the financial statements themselves. For cities, these financial statements include the government-wide Statement of Net Assets and the government-wide Statement of Activities. The Statement of Net Assets is the balance sheet, and shows assets, liabilities, and their difference, net assets. The Statement of Activities in the counterpart to the income statement, and shows flows of resources that affected the net assets over the period.
Estimation refers to the methods used to arrive at the numbers to be formally included in the financial statements.
Disclosure refers primarily to the material in notes to the financial statements containing additional or explanatory information about items recognized in the financial statements or items that are omitted from the financial statements.
Overall, the new pension rules brought recognition, estimation and disclosure closer to what is done in the private sector. To go into more detail, we should first make more clear exactly what we're talking about here. There are basically two types of pension plans: defined benefit plans and defined contribution plans.
In a defined contribution plan, the employer is responsible for making specified contributions to the pension fund and that's the end of it for the employer. What the beneficiaries ultimately receive is not specified and the employer makes no promises in that regard. Defined contribution plans do not require any fancy accounting for the contributor because its obligation begins and ends with the contributions. There is no need to estimate the eventual benefits received—it is up to the beneficiaries to keep track of their own situations.
In a defined benefit plan, the employer makes promises about the benefits to be received by beneficiaries and is responsible for fulfilling these promises (the employer takes all of the investment risk). These promised benefits entail a financial liability for the employer. This is where funding comes in—if benefits are promised, what is being done to make sure these promises can be kept? In the private sector, both ERISA [Employee Retirement Income Security Act] requirements and the tax law, which determines the deductibility of contributions to pension funds, influence the funding decisions. Tax deductibility is not relevant for governments, and they are not subject to ERISA. Funding of government-promised pensions is a political matter, not a tax or regulatory one.
So the new rules are aimed at defined benefit plans, and make required accounting recognition and disclosures much closer to what is done in the for-profit sector. The net pension obligation, i.e., the excess of the estimated liability for making promised payments over the value of the funds set aside for fulfilling this liability, will be recognized and shown on the statement of net assets (balance sheet). The amount of recognized pension expense in the statement of activities (income statement) will increase, because it will include interest on the net pension obligation, as well as the present value of the future pension benefits earned in the current year by employees. Pension expense will also take into account any substantive changes in the liability due to changes in the terms of the plans. As in the private sector, these will be amortized into pension expense over a period of time that depends on the number of years the beneficiaries of the pension plan are expected to work.
Q: What are the key changes in the new rules?
As we discussed above, the new rules will require recognition of the net pension obligation in the statement of activities. This amount is not currently recognized. They also affect the recognition of pension expenses. Some information about these amounts is disclosed in the notes to the financial statements, but it is hard to dig out.
For example, consider the financial statements of the city of New Haven for the year ended June 30, 2011. The amount of long-term liabilities recognized on New Haven's statement of activities at June 30, 2011 is $679.5 million (see Note 10). Of this amount, $13.3 million is the unfunded pension obligation. However, the disclosures in Note 12 show that the city is responsible for two main defined benefit pension plans: one for city employees, another for policemen and firemen. The unfunded pension obligation for city employees is estimated to be $203.9 million. For policemen and firemen, it is estimated to be $266.8 million. The total of these estimates is $470.7 million. This all means that under current accounting rules, $470.7 million - $13.3 million = $457.4 million of the estimated amount underfunded, while shown in the footnote, is not recognized in the financial statements. The new rules will close this gap in the GAAP for governments.
Our exploration above depends on numbers currently disclosed in the notes to the financial statements. The new rules also affect how amounts currently disclosed must be calculated. It is possible that the biggest change will be the level of discipline and consistency brought to the calculation of these numbers by state and local governments. This will make differences in various governments' obligations easier to identify, and may bring a better approach to both funding decisions and the decisions to provide or alter defined benefits. However, in such a complex area with such a complicated set of rules, it is always difficult to predict what will occur. However, by imposing some consistent discipline on the accounting, we can get people to focus on the issues in a more systematic way.
That said, the requirements for choosing discount rates may ultimately have the biggest effect on the reported numbers.
Q: Are changes to reporting rules likely to affect how pension plans are actually funded?
These new accounting rules, in and of themselves, have nothing to do with funding. Accounting policy makers cannot make law, or make anyone do anything other than account for what they do in a specific way. However, as we saw in the for-profit sector thirty years ago when there were similar pension accounting changes, the awareness of the magnitude of future pension obligations may prompt, or at least contribute to, a more broadly shared focus on the issues and their potential resolution. This more broadly shared focus provides better conditions under which the key players can engage each other and address what really has to be done for a seriously underfunded defined benefit plan, whether that be additional funding or restructuring of promised benefits.
Of course, one thing we have seen in the private sector is a shift to defined contribution plans from defined benefit plans, and governments may end up doing the same thing. With a defined contribution plan, the funding is essentially immediate and transparent. That is from the providers' point of view. The recipients now have the problem of figuring out what their incomes in retirement will be.
Q: Were the changes needed?
Part of the answer here depends on how we view recognition versus disclosure. In our experience, requiring amounts to be recognized, instead of only disclosed, results in greater care in estimating the amounts and in greater attention paid to these amounts by various parties, including management. Also, we should point out that the GASB spent a lot of time and resources making these rules, and its members must have felt that it was worth the commitment of effort going in. It seems to us that they were right—so many state and local governments are under financial stress stemming largely from pension obligations that focus on the area was warranted.
Whether these rules result in amounts in the financial statements that better reflect, in a statistical sense, the ultimate cash flow stream from pension obligations and funding decisions remains to be seen. Also, we will have to see whether the assumptions imposed in doing the calculations turn out to be better connected with what actually happens. For example, many estimates use what seems now to be a pretty high estimate of expected returns on pension assets, and justify it by past results. Whether future returns will be that high is an open question.
Q: The practice of recognizing market gains and losses over several years is eliminated in the new standards. What was the benefit of “smoothing”?
Arguments for smoothing in the area of pensions in part rely on the fact that the amounts are large, spread over a long period of time, and are sensitive to assumptions. Having reported numbers swing around wildly because of small changes in assumptions is distasteful, so procedures are imposed to “smooth” things out over time. The hope is that this smoothing results in a better picture of the long run relative to what would be portrayed without smoothing.
Q: To what degree are public pensions in or approaching a crisis?
Most knowledgeable people believe that public pensions are in crisis right now. Not to open another can of worms, but a likely even larger state and municipal funding crisis looms: the funding of employee health benefit costs in retirement. Many believe that this number is larger and more unfunded than pension obligations. For example, the notes to the City of New Haven's June 30, 2011, financial statements contain an estimated of the unfunded amount for these other postemployment benefits to be $414.0 million, only $90.7 million of which is currently recognized in the financial statements themselves. But we will leave that discussion for another time.
With respect to the economics and finance of pension obligations, an interested reader should consult a great new book on the subject, Pension Finance, Putting the Risks and Costs of Defined Benefit Plans Back Under Your Control, by Barton Waring '87. This is a great book and we are happy to say that Barton is an SOM alumnus.
Q: Is more accurate reporting likely to influence how governments do approach pension plan funding?
Whether the new rules have this effect or not, we hope they are at least a step in the right direction. However, there are a few issues impacting the effect of the rules. One is the time required to implement any systematic change in accounting. The new rules for state and local governments are slated to take full effect for fiscal years beginning after June 15, 2014. Earlier adoption is “encouraged.” Please understand that we are not saying that the new rules should be imposed earlier—it is often the case with complex accounting standards that a long lead time is required. But the numbers themselves will not be available to even have an impact for quite some time.
A second issue is whether the effect of the rules on estimation and disclosure of amounts is even greater than their impact on recognition of amounts in the financial statements proper. This may seem arcane, but it is important. If the only impact is on recognition, then any effects must be generated from commitments that the state or local government have that are explicitly tied to their financial statements. Such commitments are common in the private sector, but we believe they are rare in the public sector.
If the impact is primarily on estimation and disclosure, then there are two possible issues: whether new information is produced as a result of the rules or whether there will be a common-knowledge effect. The assumptions that have been used in estimating pension obligations and expenses are disclosed in financial statements, and it may be that we will not learn that much that we couldn't have figured out under the old disclosure rules. But even so, there may be an effect caused by the fact that the accounting for pensions and its results will be common knowledge—i.e., known by all the players, and all the players know they all know, etc. Game theory teaches us that making something common knowledge can have dramatic effects on the outcomes of negotiations.
We hope the new rules will help us budget for and deal with the funding crisis that is surely looming. Otherwise, as was pointed out to us by current New Haven mayor John DeStefano, the event that causes us to finally confront the pension issue will be a cash flow crisis—i.e., some states or municipalities will not be able to pay their bills when they come due—instead of a budget crisis. We hope that the new rules result in budgeting and planning being taken seriously enough so that the problem will be addressed before it becomes an actual cash flow crisis. Unfortunately, the current financial distress of many cities suggests that budgeting and planning are not sufficient, and that a cash flow crisis must occur before meaningful progress can be made.