In February, Congressmen Devin Nunes and Darrell Issa (both California Republicans) and Paul Ryan (R-Wis) introduced legislation (H.R. 567) to require states and local governments to report their pension liabilities to the federal government using what is known as a “riskless rate” — an interest rate tied to the Treasury bond rate — to determine those liabilities.[1] Any state or local government not complying with this requirement would face a penalty of great severity — it would lose the ability to issue tax exempt bonds, which states and localities rely upon to finance infrastructure construction and related purposes.
There is clearly a need to standardize state and local pension reporting and make it more transparent. Extensive work on these matters by an expert, non-political board that enjoys respect in financial markets — the Governmental Accounting Standards Board — is now nearing completion. GASB is on track to issue new standards next year for state and local government reporting on the financial status of their pension funds.
The Nunes bill, however, would effectively short-circuit and override the GASB process by issuing a federal edict on how pension funds are to report liabilities. It would be unsound policy to substitute heavy-handed and unnecessary federal intrusion (which seems designed in part to advance ideological goals, as explained below) for the GASB standards and the financial market discipline that induces state and local governments to comply with those standards.
There are a number of other, related problems with the Nunes proposal, as well, including:
- It could sow public confusion, as it would likely lead the public and many policymakers to believe that the amounts that states and localities need to deposit in their pension plans each year are substantially larger than the amounts actually needed. Misunderstanding of the level of contributions required could lead to excessive cuts in other parts of state budgets (such as education) and/or greater-than-needed tax increases in order to free up more room in state budgets for the greatly increased pension contributions; it also could generate pressure to end defined-benefit pension plans.
- It would fail to accomplish its stated objective of allowing observers to compare the funding status of pensions on an “apples-to-apples” (or standardized) basis because it would set standards for only two of the array of variables that are used to calculate pension-fund liabilities.
- A new federal bureaucracy would have to be created to gather, process, and verify the information for the nation’s 2,550 state and local pension plans and to apply and enforce any penalties that would have to be applied under the legislation. The new bureaucracy would be created at a time when the federal government is trying to cut costs.
- The legislation could unnecessarily spook the bond markets, leading to higher borrowing costs for states and localities.
This analysis first explains the underlying issues and why the legislation would likely sow confusion and fail to enable meaningful comparison of different pension plans. It then examines the GASB process that represents a far superior approach. Finally, it looks at the bill’s impact in necessitating a new federal bureaucracy and its potential to rattle bond markets.
A debate has begun over what assumptions public pension plans should use for their “discount rate” — that is, the interest rate they use to translate future benefit obligations into today’s dollars.[2] The decision on what discount rate assumption to use affects the estimates of the future liabilities that pension funds face, and it may also affect the level of required annual contributions by state and local governments to the pension funds.
The “findings” section of the Nunes bill declares, as though it were a clear and agreed-upon fact, that state pension plans have unfunded liabilities of $3.23 trillion. In reality, there is considerable debate and disagreement about the best way to measure unfunded liabilities, a matter for which there is no simple right or wrong answer, since estimates of liabilities depend on the assumptions that are employed. The $3.23 trillion estimate of unfunded liabilities which the Nunes bill would enshrine in law derives from calculating liabilities using what is known as the “riskless discount rate,” essentially the rate for U.S. Treasury obligations. In contrast, standard analyses based on accepted state and local accounting rules — which calculate liabilities using the historical average return on pension plans’ assets, which is higher because pensions diversify their investments rather than investing solely in bonds — put the unfunded liability at about a quarter of that amount, a more manageable (though still concerning) $700 billion.
Economists generally support use of the riskless rate in valuing state and local pension liabilities because the constitutions and laws of most states prevent major changes in pension promises to current employees or retirees; they argue that definite promises should be valued as if invested in financial instruments with a guaranteed rate of return, like Treasury bonds. State and local pension funds, however, historically have invested in a diversified market basket of private securities and received average rates of return much higher than the riskless rate – 8 percent over the past two decades. (A “riskless rate” would be about 4 percent or possibly less.) It also should be noted that economists are not arguing that state and local pension funds should change their investment practices, liquidate their equity portfolios, and invest solely in bonds.
A key point to understand is that the issues of: 1) how states and localities should value their pension liabilities; and 2) how much they should contribute to their pension funds each year to meet their pension obligations are two separate issues, although they obviously are related. The estimate of more than $3 trillion in unfunded liabilities that results from use of the “riskless rate” does not mean states and localities should have to contribute that amount to their pension funds, since the pension funds very likely will earn higher rates of return over time than the Treasury bond rate — and that, in turn, means states can achieve pension fund balances adequate to meet future obligations without adding the full $3 trillion to the funds. In fact, two of the leading economists who advocate valuing state pension fund assets at the riskless rate have observed, “…the question of optimal funding levels…is entirely separate from the valuation question.”[3]
Given this complex situation, there is risk that a federal rule that mandates valuation on the basis of the “riskless” rate would sow confusion between the unfunded liability valuation and the amounts necessary for state and local government to contribute to the plans. The public is unlikely to be able to distinguish between the valuation required under the legislation and the required annual contribution levels, especially if the large unfunded liability numbers derived under the legislation’s reporting requirements are widely publicized. This may lead some policymakers to conclude that the levels of required contributions are unaffordable and the pension plans no longer viable when that is not the case.
More specifically, if states and localities are pressured to increase their annual pension contributions to meet the much larger contribution amounts that would be required if the “riskless” discount rate were used mechanically to calculate the contribution levels, any of several deleterious effects could result. States could end up cutting education or other priority investments in order to free up room in their budgets for pension contribution levels that exceed the amounts needed to cover future pension liabilities. Or, states could raise taxes more than is needed. Moreover, the overfunding of pension plans that ultimately would result could lead to demands for increased pension benefits that would not represent a sound use of resources.
Alternatively, if states and localities are pressured to raise their pension contributions to the levels that would be needed if pension funds actually invested solely in bonds, that could induce more states to abandon defined benefit pension plans altogether. That appears to be one of the goals of some of the bill’s sponsors. Commenting on an identical bill introduced in the last Congress, the Wall Street Journal said, “Their bill would encourage governments to switch to defined-contribution plans by revealing the true magnitude of their unfunded liabilities.” [4] Rep. Nunes also told a group of California government officials, “So what this will only set up, what the folks in the private sector have figured out a long time ago, was that you have to get away from the defined benefit plan (pensions) and somehow get to a defined contribution (401(k)-style plan).”[5]
To be sure, many states and localities do need to increase the amount of funding they deposit in their pension funds to address their unfunded liabilities. But they do not need the massive increases in contributions that would be required if pension funds were to invest solely in bonds and to receive the bond rate of return.
Legislation Would Not Standardize Pension Reporting
Congressman Nunes has said that an important purpose of his bill is to allow someone to “…compare the roughly 2,500 public employee pensions against each other to know which plan is better funded compared to other plans.”[6] The legislation, however, would not accomplish that objective.
The bill requires states and localities to provide a variety of information about the funding status of their pension plans, presented in the normal ways they make those calculations. These calculations generally are made by states and localities based on the guidance provided by the Governmental Accounting Standards Board. GASB’s current standards (which, as described below, are in the process of being changed) allow a range of variation on a number of dimensions of pension accounting. Variables affecting plan valuation and the current value of unfunded liabilities that may differ from plan to plan include the liability accrual method (with GASB currently allowing seven different methods, of which three are in common use), the assumptions about wage inflation and (if there is an automatic COLA) price inflation, the valuation of assets, and the interest rate (i.e., the assumed rate of earnings that the pension fund earns on its assets). These types of variations make it difficult to compare the funding status of one plan to another.
The Nunes bill specifies that each state and locality must include a supplementary report in which plan assets are valued using fair market value and the interest rate is the rate on Treasury bonds. But standardizing just these two dimensions of pension funding calculations will not create comparability among the reports of different state and local governments, because other elements of the calculation that result in large differences across plans will continue to be reported on different bases.
For example, different accrual methods yield different estimates of funding status with different interpretations. It is also the case that a state with an automatic COLA that assumes lower price inflation in the future will look better funded than a similar plan that assumes higher inflation. Moreover, considering the size of unfunded liabilities alone does not give a full picture of the status of a state’s pension fund. Also important is the annual pension expense relative to some measure of affordability. This can differ with the number of years over which unfunded liabilities are amortized. For example, Maine amortizes unfunded liabilities over 10 future years, while Illinois amortizes over 40 years. Other states fall in between. For any given amount of unfunded liabilities, a state with a shorter amortization period will look as though it has higher annual pension expenses relative to its budget or economy than a state that uses a longer period.
As described below, the proposed revisions in the Governmental Accounting Standards Board’s pension guidance for state and local governments attempt to standardize all of these dimensions to create comparability among plans.
The Legislation Would Short-Circuit and Conflict with the GASB Process
For the past four years, the Governmental Accounting Standards Board — which sets standards for financial accounting for governments, just as the Financial Accounting Standards Board (FASB) does for private sector businesses — has been conducting extensive research and consultation with well over 100 stakeholders, including public hearings, to develop new pension financial reporting standards. The standards are expected to be promulgated next year. While GASB has no authority to require states and localities to follow its standards, bond raters and financial analysts generally look askance at governmental entities that do not comport with the standards. Thus, GASB standards typically become the de facto financial accounting rules for state and local governments, because of the discipline of the financial markets.[7]
The new GASB standards have been issued in draft form. Assuming that the final standards will be similar to the most recent GASB draft, the new standards will provide for many fewer choices of methods and move state and local governments to issue annual statements of the financial position of their pension funds that are comparable to one another. Unlike the Nunes bill, the forthcoming GASB standard addresses a wide array of areas of pension financial reporting in its effort to achieve accuracy and much greater uniformity.
For example, the draft of the GASB standard requires all pension funds to report on the same liability accrual basis, known as the “entry age” method. This would assure that the reported funding status (i.e., the percentage of liabilities for which assets are available) is comparable from plan to plan. [8] The GASB standard also would eliminate most of the discrepancy in the number of years over which unfunded liabilities are amortized. The Nunes bill, by contrast, does not address differences in accrual methods or amortization periods.
In addition, while the Nunes bill requires use of the interest rate for Treasury bonds to determine the funding status of pension funds, the draft GASB standard explains that the liability amount that results from using the “riskless rate” to calculate fund liabilities does not reflect the amount that state and local governments need to deposit in their pension funds.[9] To estimate the amount of pension fund assets actually expected to be available to finance pension payments, the draft GASB standard calls for use of the actual expected rate of return, because it better reflects the level of contributions that will be needed. The GASB draft does require use of a lower interest rate for projecting benefit payments expected to have to be made after the projected point at which a plan’s assets are expected to be depleted. For the portion of future liabilities for which current assets or expected contributions and associated earnings cannot be identified, the GASB draft standard uses an interest rate derived from an index rate for state and local governmental bonds of high quality.
Accordingly, the overall rate that the GASB draft would use to estimate the level of contributions that state and local governments need to make to their pension funds would be based on a blend of the expected rate of return on a fund’s existing and expected assets and the rate of return on high-quality municipal bonds (which would be applied to the additional funding that will be needed). The precise blend would depend on various factors related to the funding status of the pension fund.
A large amount of analysis and thought has gone into the development of the new GASB standards. The process of developing these standards has been one into which all stakeholders and interested parties have had an opportunity to provide input and analysis, and the GASB Board has spent considerable time scrutinizing these submissions and various other analyses. When the final standard is issued, it is likely to have strong compliance. [10]
There are 2,550 distinct state and local pension plans in the country, as Rep. Nunes has noted.[11] Each state or locality provides information about its plan. The federal government does not collect or compile this information.
As a result, instituting the Nunes bill would necessitate creating a new bureaucracy within the Treasury Department to maintain information on active pension plans, collect relevant reports from state and local governments, assure the timeliness of those reports, check them for accuracy and for compliance with the terms of the legislation and the implementing regulations, communicate with the states and localities regarding any discrepancies, and enforce the penalties (described below) on states and localities that are out of compliance.
As noted above, it is financial analysts and the securities markets that enforce discipline on the pension reporting of states and localities. Once the new GASB standards are in place, the strong market incentives for compliance with those standards will assure that the reported financial status of pension plans is comparable across plans. This is a clear example of a situation in which the operation and discipline of markets is preferable to the imposition of federal regulations.
The Legislation Could Spook Bond Markets
The state and local bond markets have been volatile recently. In particular, the appearance on the television show 60 Minutes by a financial analyst predicting 50 to 100 significant bond defaults this year — a claim later shown not to be supported by solid analysis — is thought to have played a major role in causing some investors to lose confidence in municipal bonds at the end of 2010 and beginning of 2011, resulting in a sizeable sell-off of those securities.[12] States and municipalities have had to offer higher interest rates to sell their bonds because of this loss of confidence, a situation that is only now slowly turning around.
The potential of large and confusing unfunded liability numbers appearing on a federal government website could have a similar affect on the bond market, especially since those numbers would not represent the amounts that states and localities actually need to contribute to their pension funds to cover their obligations. Not every financial analyst, bond trader, or investor understands the complexity of state and local pension financing and thus would know how to interpret these data. The legislation and the data it requires hold the potential to frighten potential investors unnecessarily and thereby result in higher interest costs for states and localities.
The penalty for failure to comply with these reporting requirements would be the loss of a state’s or municipality’s ability to issue tax-exempt bonds. That would be a massive — and unprecedented — penalty that would render it virtually impossible for the state or local government in question to attract investors. While it is unlikely that matters would get to a point where such a penalty would actually be imposed on a state, such a development is not impossible. And it is conceivable, even if unlikely, that the penalty could be imposed as a result of legitimate differences of opinion between a city or state and employees of the Treasury Department over the best data to use in particular circumstances and reports.
States and localities issue tax-exempt bonds largely to finance infrastructure. The bonds allow them to build and maintain bridges and highways, build schools and other public buildings, and subsidize the building of low-income housing, to name some of the most common uses. The country has a large backlog of unmet infrastructure needs, and, over time, our deteriorating infrastructure will have a negative effect on economic growth and development. Threatening state and local governments’ ability to support infrastructure to compel compliance with a federally-dictated reporting requirement would be excessive and inappropriate.
No state or locality has asked the federal government to “bail out” its pension plan, nor is any state likely to do so in the future. A number of states and localities clearly need to institute pension reforms. But they have the tools needed to restore their pension plans to an appropriate level of funding over time as markets and the economy improve.
Pension financial reporting standards are about to be overhauled as GASB completes its process with input from stakeholders and consideration of an array of issues related to pension financing and reporting. No useful purpose is served by arbitrarily imposing different federal standards before the GASB process is completed. Doing so would likely sow public confusion, require the creation of a new (and potentially intrusive) federal bureaucracy, and risk spooking the bond markets and threatening the ability of states and localities to invest in needed infrastructure improvements.
