«Public Pension Accounting Rules and Economic Outcomes James Naughton j-naughton 847-491-2672 Kellogg School of Management ...»
Public Pension Accounting Rules and Economic Outcomes
Kellogg School of Management
2001 Sheridan Road, Room 6223
Evanston, IL 60208
Sloan School of Management
Massachusetts Institute of Technology
100 Main Street, E62-662
Cambridge, MA 02142
Sloan School of Management
100 Main Street, E62-667
Cambridge, MA 02142 September 2014 We thank Keith Brainard of the National Association of State Retirement Administrators for kindly providing pension data. We thank Mary Barth, Sudipta Basu, Rajib Banker, Natalya Shnitser and seminar participants at the 2013 American Accounting Association annual meeting, the 2014 University of Minnesota Empirical Conference, Baconni University, College of William & Mary, Georgetown University, University of Minnesota, Massachusetts Institute of Technology, Northwestern University, Southern Methodist University, Temple University, University of British Columbia, and University of Washington for helpful comments and suggestions. Rita Strang provided excellent research assistance.
Abstract We provide evidence that the accounting rules prescribed by the Governmental Accounting Standards Board (GASB) and the choices states make when implementing these rules allow states to understate pension funding gaps, especially during times of fiscal stress. We also find that the funding gap understatement is negatively associated with states’ decisions to increase taxes and cut spending, and that these understatements are associated with higher future labor costs. Importantly, we find that the positive association between the funding gap understatement and future labor costs is attributable to the inherent methodology in the GASB rules, which systematically understate the funding gap, and not to opportunistic reporting by state governments. Therefore, it is not only the case that the GASB pension accounting rules pose intergenerational fairness issues (by not requiring sufficient pension contributions), but also that they are associated with policy choices (such as increased labor expenditures) that have the potential to exacerbate future fiscal problems.
1.Introduction With the financial crisis in 2007 and 2008, the subsequent recession, and the resulting loss of jobs, there has been a heightened interest in the financial outlook of state governments.
In particular, politicians, voters, capital providers, and regulatory bodies have all focused on whether the deterioration in states’ fiscal wellbeing and burgeoning debt balances will affect their tax policies and their ability to provide entitlement programs. In this paper, we conjecture that politicians will view defaulting on debt, raising taxes, or cutting entitlement programs to be costly political responses to fiscal stress, and as a result, politicians will use mechanisms like accounting discretion to mask deficits. We also conjecture that there are economic outcomes associated with the use of accounting discretion, as outputs from the accounting system serve as inputs into a variety of economic decisions made by governmental entities.
We focus our analysis on whether states engage in actions to mask the size of pension funding gaps and thus reduce the size of pension contributions during periods of fiscal stress and whether there are economic outcomes associated with those actions. Pension obligations are among the largest obligations states face, and taking accounting actions to improve the funded status of these obligations may reduce a state’s expenditures on pensions and provide states with the flexibility to avoid raising taxes or cutting entitlement programs during economic downturns.
However, using accounting discretion to reduce funding gaps is likely to result in an understated cost of labor being reflected in the state’s accounting system. This is particularly important as the information that is generated by the states’ accounting system serves as inputs into states’ budgets, appropriations, and other control mechanisms. When an accounting system understates the costs associated with pension benefits, states may invest more in labor, as the “true-cost” of each worker is not reflected in the decision-maker’s information set.
We argue that there are several elements of GASB pension reporting rules that allow states to understate their pension funding gaps, and thus lower their annual pension related expenditures. In particular, governmental entities discount future pension benefits using the expected investment return on assets held in the pension trust and have the discretion to amortize investment gains and losses over future reporting periods. Jointly, these rules provide financial reporting discretion to politicians so they can reduce funding gaps and required contributions to the pension fund, which may alleviate budgetary stress and the need to increase taxes and cut spending. Thus, the main hypotheses we investigate in the paper are whether states understate their pension funding gaps in periods of fiscal stress, whether understated pension funding gaps are associated with states’ decisions to engage in mid-year spending cuts and tax increases, and whether the extent to which states understate their pension funding gaps is associated with future employment costs.
To provide evidence on these hypotheses, we collect data on the state’s pension obligations from two sources: the Public Pension Coordinating Council’s PENDAT Survey of State and Local Government Employee Retirement Systems and the Boston College Center for
collecting information for missing plan-years directly from the state’s Comprehensive Annual Financial Reports (CAFRs) and pension plan valuation reports. After accounting for missing data, we have 984 state years with pension data.
We start our analysis by investigating the relation between the extent to which states understate their pension funding gap and measures of the states’ fiscal condition.1 We develop a We focus on developing a measure of the extent to which states understate their funding gaps, and argue that the extent to which states understate their funding gaps will affect the annual expenditures states make for their pension obligations. As we discuss below, understated funding gaps can reduce expenditures for both the normal and amortization components of the Annual Required Contributions (ARC) for a state pension plan.
measure of the extent to which states understate their pension funding gaps by comparing the funding gap reported in the state’s accounting records under the GASB’s rules to the funding gap they would have reported had they followed the Financial Accounting Standards Board’s (FASB) rules. We use the funding gap calculated under the FASB rules as our benchmark because it approximates the settlement cost of the pension liability and uses the market value of the pension assets. We believe this number more accurately reflects the true size of the funding gap. The FASB approach is also consistent with the one advocated by academics (see for example, Novy-Marx and Rauh, 2011) and adopted by most other developed countries, including Canada and many European countries.
We then decompose the total understatement of the pension funding gap into three pieces: the portion that is related to the use of discretion to overestimate the expected return on plan assets, the portion that is attributable to the amortization of realized and unrealized investment returns, and the portion that reflects differences between the GASB’s and FASB’s rules for the selection of the appropriate discount rate. We measure the extent to which a state is fiscally stressed using two variables. Our first variable measures the extent to which the state is running a budget deficit in the current fiscal year and our second variable measures the extent to which a state has reserves to meet any fiscal shortfall.
We find that the extent to which a state is fiscally stressed is associated with the magnitude of the pension funding gap understatement. Specifically, we find that states over estimate the expected return on plan assets to a greater extent during periods when they are running budget deficits and when they have a relatively smaller cushion in their general fund balances. A higher expected return on assets will produce a lower pension liability, and hence a lower funding gap. These results indicate that states use the discretion allowed under the GASB rules to understate the funding gap when they are financially constrained. We also find that states overestimate the expected return on plan assets during periods in which the governor is in a relatively more competitive election and during periods in which the state issues debt. These results are consistent with findings in prior studies that states have incentives to manipulate the outputs of the accounting system to influence the outcomes of elections (Kido et al., 2012) and to influence debt costs (Baber and Gore, 2008). We do not find any evidence that states opportunistically set the amortization period for investment gains and losses to influence elections, reduce debt costs, or to inflate fiscal performance.
Our first set of tests provides indirect evidence that states manipulate their funding gaps to avoid raising taxes or cutting expenditures. Our second set of tests examines this question more directly by investigating whether the accounting discretion states use to understate their funding gaps reduces the extent to which fiscal stress is associated with states’ decisions to raise taxes or cut expenditures. Consistent with our expectation, we find that states that overestimate the expected return on plan assets to a greater extent are less likely to engage in midyear tax increases and midyear expenditure cuts when they face fiscal stress. These results suggest that using accounting discretion to understate pension funding gaps can reduce the extent to which states have to engage in other more politically costly actions like increasing taxes or cutting expenditures in response to financial difficulties.
We conclude our paper by investigating whether there is a relation between states’ future payroll costs and the extent to which states understate their pension funding gaps. In this analysis we model future payroll costs to be a function of the reported funding gap, the unreported funding gap, and various control variables for the demand for public service and the state’s economic condition. The reported funding gap is obtained from the financial statements prepared under the GASB rules. The unreported funding gap is equal to the understatement calculated as we describe above. We measure states’ future payroll costs over a one, three, and five-year horizon from the year the state sets its pension assumptions. We find a weak, positive relation between the expenditures on state payrolls in year T+1 and the reported pension funding gaps in year T. However, as we move out in time to year T+3 and year T+5, we find that this relation becomes insignificant. These results suggest that states with larger reported pension funding gaps tend to spend more on employment in the short term. However, over time, states respond to the reported funding gaps and reduce spending on their employment.
We also find a positive relation between the expenditures on state payrolls in year T+1 and the unreported pension funding gaps in year T. However, unlike our tests of the reported funding gap, we find that the relation between the unreported funding gap and payroll persists three and five years into the future. These results suggest that by not reporting the true economic costs of pension obligations, states tend to spend more on labor both in the short and long terms.
In addition, we find that this positive relation is attributable to the understatement related to the design of the GASB rules and the results are robust to a change specification. Overall, these findings highlight that discretionary actions, like distorting expected rates of returns on assets, or choosing inappropriate investment smoothing horizons are relatively transparent, and thus do not affect employment decisions. In contrast, the non-discretionary portion of the funding gap understatement masks the true cost of each employee and government officials do not (or cannot) undo the understatement in their hiring decisions.
Our findings extend the literature on public pensions. The debate on public pensions is driven by the concern that states use a discount rate that is too large. As a result, states understate their pension funding gaps, resulting in insufficient current contributions, and hence an unfair shifting of the cost of these plans to future generations. We find that states are more likely to understate pension funding gaps during periods of fiscal stress. We also find that funding gap understatements are associated with a reduced likelihood of tax increases and expenditure cuts during periods of fiscal stress, and that these understatements are associated with future increases in payroll expenditures. Importantly, the increases in payroll expenditures appear related to the inherent design of the GASB’s rules. Therefore, it is not only the case that the current GASB regime poses intergenerational fairness issues (by not requiring sufficient pension contributions), but also that it is associated with policy choices (such as increased labor expenditures) that have the potential to exacerbate public sector finances.
The paper proceeds as follows. Section 2 provides background information on financial reporting for public pension plans and Section 3 develops the hypothesis. Section 4 discusses our research design and Section 5 describes our sample. Section 6 presents our empirical findings and Section 7 concludes.
2.1 GASB’s Pension Accounting Rules During the period of our study the financial reporting rules for pension plans of governmental entities were codified under GASB Statement No. 25, Financial Reporting for Defined Benefit Pension Plans and Note Disclosures for Defined Contribution Plans, and GASB Statement No. 27, Accounting for Pensions by State and Local Governmental Employers. These rules define how states determine the discount rate used in calculating the present value of their pension obligations and the methods states use to determine the values of the assets that are being held to satisfy these obligations.