JOURNAL OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT Editors in Chief: Khi V. Thai, Florida Atlantic University. E-mail: [email protected] Editors in Chief: Robert Kravchuk, Indiana University-Bloomington. E-mail: [email protected] Governmental Accounting Editor: Donald R. Deis, Texas A&M UniversityCorpus Christi. E-mail: [email protected] Healthcare & Nonprofit Organization Financial Management Editor: Dana Forgione, University of Texas at San Antanio. E-mail: [email protected] Copy Editors: Paula Altizer and Vivian Mydlarz Artistic Designer: Loy Nguy Editorial Board Roy Bahl, Georgia State University John Bartle, University of Nebraska at Omaha Jane Beckett-Camarata, Rutgers University Melvin V. Borland, Western Kentucky University Cynthia Jones Bowling, Auburn University Patricia Byrnes, University of Illinois at Springfield Cyril F. Chang, Memphis State University Carol Ebdon, University of Nebraska at Omaha Jerry Gianakis, Suffolk University, Sawyer Business School Merl M. Hackbart, University of Kentucky Jean Harris, Pennsylvania State University-Harrisburg Rebecca Hendrick, University of Illinois-Chicago W. Bartley Hildreth, Georgia State University Alfred Ho, Indiana University-Purdue University Indianapolis Beth Walter Honadle, University of Cincinnati Marc Holzer, Rutgers University-Newark Dennis S. Ippolito, Southern Methodist University Craig L. Johnson, Indiana University at Bloomington Janet Kelly, University of Louisville Aman Khan, Texas Tech University William Earle Klay, Florida State University Josephine M. Laplante, University of Southern Maine Thomas P. Lauth, University of Georgia Lance T. Le Loup, Washington State University Shuanglin Lin, Peking University, China and University of Nebraska at Omaha David Matkin, Florida State University Justin Marlowe, University of Washington Cliff McCue, Florida Atlantic University Charles E. Menifield, University of Memphis Hiroshi Mifune, Chuo University John L. Mikesell, Indiana University Gerald J. Miller, Arizona State University Patrick R. Mullen, University of Illinois-Springfield Jun Ping, University of Arizona Yuhua Qiao, Missouri State University ii J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT Christopher Reddick, University of Texas at San Antonio William C. Rivenbark, University of North Carolina at Chapel Hill Mark Robbins, University of Connecticut James D. Savage, University of Virginia Walt Schubert, La Salle University Joe Shepard, Florida Gulf Coast University Daniel Smith, New York University Kenneth A. Smith, University of Washington Douglas R. Snow, Suffolk University Charlie B. Tyer, University of South Carolina M. Peter Van Der Hoek, Erasmus University-Rotterdam Xiaohu Wang, University of Central Florida Douglas J. Watson, University of Texas at Dallas Jeffrey A. Weber, East Stroudsburg University Samuel J. Yeager, Wichita State University Jerry Zhao, University of Minnesota Governmental Accounting Section Editorial Board William R. Baber, George Washington University Rich Brooks, West Virginia University Rita Cheng, University of Wisconsin-Milwaukee Paul Copley, James Madison University Mary Fischer, University of Texas at Tyler Robert J. Freeman, Texas Tech University Gary A. Giroux, Texas A&M University James Guthrie, University of Sidney Rhoda C. Icerman, Florida State University Larry Johnson, Colorado State University Susan Kattelus, Eastern Michigan University Saleha B. Khumawala, University of Houston Suzanne Lowensohn, Colorado State University Santanu Mitra, Wayne State University Kris K. Raman, University of North Texas Jackie Reck, University of South Florida Robin W. Roberts, University of Central Florida Walter Robbins, University of Alabama Marc A. Rubin, Miami University (Ohio) Florence Sharp, Ohio University Pamela C. Smith, University of Texas at San Antonio Thomas Vermeer, University of Baltimore Jayaraman Vijayakumar, Virginia Commonwealth University JOURNAL OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT Volume 24, Number 3, Fall 2012 CONTENTS Changing and/or Funding OPEB Promises: A Typology of Local Government Responses to GASB 45 and the Realization of OPEB Liabilities ………….......…...................................................... 369 J. E. Yusuf and T. Musumeci Citizen Perceptions of General-Purpose and Special District Governments: A Comparative Analysis .............…………………………. 397 L. J. Killian and K. Le Two Accounting Standard Setters: Divergence Continues for Nonprofit Organizations ..…......................................................…….. 429 M. Fischer and T. Marsh The Sheriff of Nottingham's Favorite Tax: How Local Option Sales Taxes Exacerbate Budgetary Inequalities between Local Governments .…..…................................................................. 467 P. J. McHugh and G. J. Jolley Debt and Budget Gimmickry Revisited: The Upfront Bond Refinancing Savings Structure .…...….….…..…................................. 490 M. J. Luby PrAcademics Press (www.pracademics.com) Contributions to this journal are published free of charge INTERNATIONAL JOURNAL OF ORGANIZATION THEORY AND BEHAVIOR INVITATION TO AUTHORS International Journal of Organization Theory and Behavior (IJOTB) encourages practitioners and scholars to submit manuscripts dealing with the practice and study of public procurement at all levels of government in every country. Manuscript Submissions. All manuscripts should be submitted to Please see “Information for Contributors” at the end of this issue for manuscript style and submissions. Suggestions. JPBAFM invites readers to submit comments, communications and suggestions for the reprinting of informative government reports to the managing editor (Arthur Sementelli at [email protected]). For further information, please visit www.pracademics.com. J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT, 24 (3), 369-396 FALL 2012 CHANGING AND/OR FUNDING OPEB PROMISES: A TYPOLOGY OF LOCAL GOVERNMENT RESPONSES TO GASB 45 AND THE REALIZATION OF OPEB LIABILITIES Juita Elena (Wie) Yusuf and Thomas Musumeci* ABSTRACT. GASB Statement No. 45 addresses how governmental units account for employees’ other post-employment benefits (OPEB), requiring government employers to replace OPEB reporting on a pay-as-you-go basis with an accounting of the cost of current and future benefits. This requirement and the resulting OPEB liability may prompt government employers to reconsider key questions regarding their OPEB provision. The size of the OPEB liability depends on both the benefit promises made to employees and the assets to fund these promises. We propose a typology that defines four approaches for governments to respond to GASB 45 and their OPEB liabilities. These approaches represent different combinations of strategies involving OPEB promises and assets. We illustrate these strategies and responses using selected counties and nine mid-Atlantic cities. INTRODUCTION In June 2004, the Governmental Accounting Standards Board (GASB) issued Statement No. 45 (GASB 45) Accounting and Financial Reporting by Employers for Postemployment Benefits Other than Pensions to require clear and transparent reporting of the current ---------------------------------------* Juita-Elena (Wie) Yusuf, Ph.D., and Thomas Musumeci are an Assistant Professor, and a doctoral candidate, respectively, Department of Urban Studies and Public Administration, Old Dominion University. Dr. Yusuf’s research interests are in public budgeting and financial management, with a particular interest in transportation finance, and transparency, accountability and public participation. Musumeci’s research interests are in local government budgeting and finance. He is the Deputy Treasurer for the City of Virginia Beach, Virginia. Copyright © 2012 by PrAcademics Press 370 YUSUF & MUSUMECI value of other post-employment benefits (OPEB) promises in state and local government financial statements. These OPEB are defined as non-pension benefits provided after an individual leaves employment. They primarily include retiree healthcare, and often dental, vision, and prescription drug plans, but possibly also life insurance, disability, long-term care, and legal services if they are provided separately from a defined benefit pension plan. Given that health benefits are by far both the predominant and the most costly OPEB for government employers, our research focuses primarily on retiree health benefits. OPEB have traditionally been reported and paid for on a pay-asyou-go (PAYGO) basis. GASB 45, however, established new accounting standards for state and local governments as employers, requiring them to measure, recognize, and disclose the cost of their OPEB in their financial reports. Compliance with GASB 45 reporting guidelines provides employers with an understanding of their OPEB actuarial liability exposure. While intended solely as an accounting standard, GASB 45 may significantly affect the funding, sponsorship, and design of OPEB. This paper focuses on local government responses to GASB 45 and the information it discloses regarding OPEB liability, and summarizes decisions made by selected county and city governments in response to the realization of the magnitude of their OPEB liabilities following compliance with GASB 45. This focus on local governments is critical, as of the over 89,000 local governments in the US, approximately 77 percent provide OPEB to their retirees (Bell, 2006). We ask the following research question: How have local governments in the U.S. responded to GASB 45 and the resulting information about unfunded OPEB liabilities? Answering this research question is timely given the recent implementation of GASB 45. It is also relevant given the potential size of OPEB liabilities. Coupled with the current fiscal crisis, OPEB liabilities may prompt local governments to re-think the fundamental question of whether to continue to provide OPEB. Our research examines the decisions made by local governments regarding the continued sponsorship of retiree healthcare benefits, prefunding of such benefits, and subsequent benefit plan redesign and other cost cutting strategies to accommodate continued sponsorship in the face of significant liabilities and annual funding requirements. We illustrate, using a CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 371 four-cell typology, how local governments have responded to the informational content of GASB 45. We frame these responses in terms of changing OPEB promises and/or building assets to fulfill these promises. We begin with a brief overview of GASB 45 and discuss the potential impacts of GASB 45 on government employers. Against this backdrop we introduce four possible categories of responses to GASB 45. These responses involve different combinations of prefunding OPEB and maintaining benefit levels. We illustrate these responses using a study of counties by the National Center for the Study of Counties (NCSC) and examples of nine mid-Atlantic cities. We highlight how some local governments’ responses to GASB 45 have resulted in reduced OPEB. BACKGROUND ON GASB 45 GASB 45 is the governmental equivalent to Financial Accounting Standards Board Statement 106 (FAS 106) and International Accounting Standards Board Statement 19 (IASB 19). GASB 45 was partly motivated by economic and demographic factors, such as the impending baby boomers’ retirement and escalating healthcare costs that threaten to significantly increase the costs to government employers for retiree benefits (Marlowe, 2008). The traditional PAYGO approach to OPEB can result in inadequately prefunded benefits that may in turn contribute to mounting deferred liabilities and potentially adverse effects on the fiscal health of government entities. The importance of GASB 45 was spurred by most subnational governments’ not reporting or disclosing in their financial documents information pertaining to the amount of their OPEB obligations, making it impossible for the public to determine the true cost for government of providing these benefits. OPEB costs have traditionally been accounted for and funded on a PAYGO basis, with current year benefits and administrative costs paid out of current revenue. In establishing Statement No. 45, GASB (2004) noted that the PAYGO approach and its related reporting failed to (1) recognize the cost of benefits when the exchange of benefits and service takes place, (2) provide information about the actuarial accrued liabilities for promised benefits and the extent to which these benefits have been funded, and (3) provide information needed to assess potential demands on future cash flows. In 372 YUSUF & MUSUMECI essence, the problem with the PAYGO approach is that it allows governments to ignore future expenses associated with benefits promised to retirees, and to underreport substantial accumulated liabilities that have been incurred and will be incurred in the future. GASB 45 was driven by two inter-related issues (Governmental Accounting Standards Board, n.d.). First was the lack of reporting and information about the nature and size of long-term OPEB obligations and commitments, and second was the incomplete information with which to assess the cost of government and to analyze the financial position and long-run financial health of government. GASB 45 addressed concerns that the PAYGO approach was “not transparent and obscured the magnitude of government employers’ financial obligation … [and] the funded status of postretirement obligations was not disclosed and was not recognized on the financial statements” (Bell, 2006, p. 29). OPEB are a form of deferred compensation and should be recognized and recorded as the benefits are earned, rather than recognized in the future when the benefits are paid (Mead, 2008). GASB 45 establishes an amortization period so government employers can account for the cost of OPEB over the active service life of employees. From an informational perspective, GASB 45 dictates that employers offering defined benefit OPEB plans must measure and disclose their long-term OPEB costs and the extent to which the employers have contributed to meet those costs. In summary, GASB 45 requires that government employers produce statements for their OPEB using generally accepted accounting principles that present the estimated actuarial accrued liabilities and the annual required contributions necessary to cover the liabilities. In much the same way as GASB Statement No. 43 changed how subnational governments report the liabilities associated with their pension plans, GASB 45 changes how state and local governments report the liabilities associated with their OPEB plans. Informational Components of GASB 45 According to Wisniewski, “the new OPEB standards provide information about whether and to what extent promised benefits have been funded, as well as information about the potential impact by such benefits on the employer’s future cash flow” (2005, p. 106). GASB 45 can “[p]rovide information useful in assessing potential CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 373 demands on the employer’s future cash flows” (http://www.gasb.org/st/summary/gstsm45.html). It requires the measurement, recognition, and reporting of annual OPEB costs on an accrual accounting basis, which allows for the systematic and rational allocation of the present value of the OPEB over the working life of employees. The determination of the OPEB costs and liabilities begins with calculating the annual required contribution (ARC). There are two components to the ARC. The first is the employer’s normal costs or the present value of providing one year of the OPEB to all qualifying retirees. The second component of the ARC is the amortized costs of the OPEB liability. These amortized costs represent “[t]he portion of the actuarial present value allocated to prior years of employment and thus not provided for by normal costs in the current or future years” (Mead 2008, p. 290). The actuarial accrued liability (AAL) is the total amount of the OPEB earned by employees up to the date that GASB 45 is implemented and is amortized over thirty years. It represents “the amount that the organization should have been contributing in the periods prior to the implementation of the OPEB standard” (Voorhees, 2005, p. 64). Compliance with GASB 45 also involves determining the unfunded actuarial accrued liability (UAAL) which is the amount by which the actuarial value of the OPEB plan assets exceeds the AAL. For governments using PAYGO, the UAAL will be equal to the AAL as the OPEB plan will have no assets. The UAAL can also be defined as the cumulative annual ARC short- falls. The content of the information resulting from GASB 45 is daunting. Given that many governments have traditionally relied on the PAYGO approach to finance their OPEB, GASB 45 was expected to result in reporting of significant, eye-opening, unfunded liabilities. McKethan et al. (2006) reported that the AAL could amount to as much as $1 trillion nationwide. Keating and Berman (2007) cited estimates of OPEB liabilities in the $1 trillion (McTague, 2006) to $1.5 trillion (Zion & Varshney, 2007) range. Estimates by the Government Accountability Office (2007) showed an aggregate state and local government unfunded liability of between $500 billion and $1.6 trillion. Research by Credit Suisse estimated that the 25 largest cities had unfunded OPEB liabilities of over $90 billion (Zion & Varshney, 2007). OPEB liabilities for the City of Los Angeles alone 374 YUSUF & MUSUMECI were projected to be $93 billion, equivalent to $8,000 per resident (Marlowe, 2008). More recently, Coggburn and Kearney (2010) provided data on OPEB liabilities of individual states, which ranged from effectively $0 to $68 billion. Fourteen states had OPEB liabilities in excess of $10 billion. Clearly these liabilities, and the resulting funding requirements, can have serious consequences for governmental balance sheets, just as the implication of FAS 106 discussed next had serious consequences for the balance sheets of many private companies. POSSIBLE RESPONSES TO GASB 45 AND THE SUBSEQUENST REALIZATION OF OPEB LIABILITIES GASB 45 brings public sector accounting practices in closer alignment with the private sector pension accounting rules instituted in response to several highly publicized private sector pension fund collapses. Hurley et al. (2006) viewed GASB 45 as the public sector equivalent of FAS 106, to which many attributed a steep decline in private sector retiree healthcare benefits (Patterson, 2001). Learning from the FAS 106 Experience Pearson and Jerris (1995) argued that including liability reporting for OPEB reduced healthcare and other benefits received by private sector employees and retirees. FAS 106 became effective December 15, 1992, and statistics show that the private sector has since seen decreasing retiree health benefits. In 1997, 22 percent of private sector employers provided retiree health benefits, but by 2002, only 13 percent provided such benefits (Kilgour, 2009). Other studies documented this positive relationship between reduction in benefits and the implementation of FAS 106 (Mittelstaedt, Nichols, & Regier, 1995; Binnis & Riffe, 2000). The private sector experience with FAS 106 suggests that despite its intended role as simply an accounting and reporting standard that requires government employers to recognize OPEB costs, GASB 45 could result in these government employers reducing or even eliminating their OPEB. Kilgour suggested that “the major short run impact of GASB 45 will be to pressure employers to reduce costs by reducing retiree health and other benefits by curtailing access and benefit levels and shifting more of their cost to retirees” (2009, p. CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 375 32). Therefore, while GASB 45 is an accounting standard and does not mandate how a governmental entity conducts its operations, “[c]onsideration of the information produced by the application of the standard may encourage … governments to think about reducing such benefit programs (and their associated liabilities) in the future” (Wiesnewski, 2005, p. 115). Coggburn and Kearney (2010) argued that while GASB 45 has made OPEB costs more transparent, the resulting unfunded liabilities have prompted a concern “over the sustainability of many government benefit plans” (p. 106). Many of the questions being addressed by government employers today are equivalent to those resulting from FAS 106. These companies had to consider the sustainability of continuing to provide retiree healthcare benefits in their existing forms, and subsequently what types of plan design changes or cost control measures could be instituted and how they would be instituted. Companies also had to consider whether to prefund their retiree healthcare plans and how much to prefund. Both sets of questions are examined in this study. The former focuses on whether to continue providing OPEB and if so, what types of changes might be made to the OPEB plan. These questions are important because they affect the assumptions underpinning the OPEB actuarial valuation, which may result in “nontrivial differences” in the OPEB liabilities (Marlowe, 2008, p. 216). The latter question of prefunding deals with building the assets base for addressing future liabilities. The responses to these questions are important because they determine the extent of OPEB liability underfunding. Changing OPEB Promises: Continuing OPEB Provision and At What Levels? As private sector employers dealt with FAS 106, some considered discontinuing their retiree healthcare benefits or ending participation for future retirees or new hires. Several companies implemented changes to their OPEB plans, affecting benefit levels, eligibility, and the cost sharing relationship. Some private employers began to tie their retiree healthcare benefits to greater lengths of service with the company or introduced graded vesting and accruing retiree healthcare benefits over the employees’ working careers. Other elements of plan design that came under scrutiny were the provision of full benefits to spouses at no cost, the provision of full benefits at younger retirement ages, and the level of benefits provided (Feldman 376 YUSUF & MUSUMECI & Haynes, 2007). To ameliorate the financial impact of FAS 106, some companies capped the amount the employer would contribute toward retiree healthcare benefits. Also to reduce their financial burden, employers shifted away from a defined benefit (DB) plan to a defined contribution (DC) plan. These approaches had the key advantage of reducing both the employer’s liabilities and expenses for such benefits. Similar actions are expected to be taken by government employers as they deal with GASB 45. In situations where government employers may be considering plan redesign, the focus will typically be on preserving as much as possible the expectations of present retirees, while allowing the greatest changes to affect new future hires. While there is usually much more flexibility to change retiree benefits for new employees, public sector unions and the relevant labor laws may influence the government employer’s ability to change its retiree benefits (Moran, 2010). The power of government employees and their unions in particular has been shown to be an important factor influencing governments’ decisions regarding retiree benefits (McKethan et al., 2006). In some jurisdictions, unions are strong and can prevent unilateral action by the government employer. Other jurisdictions may be similarly constrained if their OPEB have constitutional, statutory, or judicial status that precludes or limits what governments can do, especially for retirees and incumbent employees. Building the Assets Base: Should the Government Employer Prefund? A key component of GASB 45 is the option of prefunding OPEB. Doing so provides a vehicle for building an asset base to offset the actuarial accrued liabilities and pay for the benefits as they come due in the future. The growth of the assets, theoretically, provides greater benefit security for retirees (Young, 2005) as more revenues into the plan, over time, come from investment income. The new OPEB reporting standards do not require that the OPEB be prefunded or that funds be held in trust. However, prefunding and setting up an irrevocable trust are recommended best practices for local governments implementing GASB 45 (Coe & Rivenbank, 2010). Government employers can choose to continue to address their OPEB on a PAYGO basis. But there are long-term consequences of doing so, in that the OPEB liability and net obligation for the employer will continue to grow unchecked. With the PAYGO approach “the CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 377 expensing is only out of pocket so there is not assistance against the accruing liability” (Bell, 2006, p. 34). However, there are significant obstacles to OPEB prefunding. For most employers the ARC payment is a large multiple of the PAYGO amount. Depending on the size of the OPEB plan, advance funding under the GASB 45 rules may add fiscal stress to the governmental entity (Young, 2005).1 Prefunding, because it usually involves higher short-term costs compared with PAYGO, may introduce additional financial obligations at an inopportune time. This may prompt a reconsideration of OPEB levels instead of prefunding existing levels of benefits (Wiesnewski, 2005). A Four-cell Typology of Government Responses to GASB 45 GASB 45 provides government employers with an understanding of their OPEB actuarial liability, which may raise the important issue of how to manage that liability (Young, 2005). The UAAL depends both on the promises to employees and the assets designed to fund these promises. Governments can respond to GASB 45 by focusing on both sides of the equation, by (1) changing benefits (i.e., the OPEB promises made to employees) and/or (2) building the OPEB assets needed to provide the benefits. Government responses to addressing the OPEB liabilities that result from compliance with GASB 45 can be organized into a fourcell typology. This typology categorizes government responses based on two dimensions (see Figure 1). The first dimension is the OPEB promise made to employees. Government employers can either maintain the same OPEB promise as before implementation of GASB 45, or reduce or eliminate benefits.2 The second dimension pertains to the assets and funding of OPEB. Here, governments also have two options, either to continue funding OPEB on a PAYGO basis, or to prefund OPEB. The four cells of the typology include a combination of these two dimensions. First, governments could maintain the status quo and make no changes to their OPEB benefits or to their PAYGO approach to funding OPEB (cell 1). Alternatively, government employers could prefund their OPEB and continue to provide the same OPEB benefits (cell 2). The remaining two response options both involve reductions in benefits, whether funding via PAYGO (cell 3) or by prefunding (cell 4). As the next section illustrates, local governments’ response to GASB 45 are spread across these different categories. 378 YUSUF & MUSUMECI FIGURE 1 Four-Cell Typology of Government Employers’ Responses to GASB 45 and Unfunded OPEB Liabilities OPEB Assets OPEB Promises PAYGO approach Prefund OPEB Maintain same benefits as pre-GASB 45 Reduce or eliminate benefits 1. MAINTAIN STATUS QUO 3. PAYGO WITH LOWER BENEFITS 2. PREFUND WITH SAME BENEFITS 4. PREFUND WITH LOWER BENEFITS HOW HAVE COUNTY GOVERNMENTS RESPONDED TO GASB 45 AND THE REALIZATION OF THEIR OPEB LIABILITIES? In 2007, the National Center for the Study of Counties (NCSC) examined how counties across the country were responding to GASB 45 through case studies of 15 counties (Sanford, 2007). The study’s findings indicate a range of responses, as shown in Appendix 1. Only a few selected examples from the NCSC study will be highlighted here to illustrate variations in how the responding counties addressed GASB 45 and its associated OPEB liabilities. Fairfax County, Virginia, had an unfunded OPEB liability of $143 million and an ARC of $15.2 million. The county did not make any changes to its retiree healthcare plan, but instead focused solely on funding the ARC and creating an OPEB trust. Montgomery County, Maryland, followed a similar approach. The FY 2008 unfunded liability was $2.6 billion and the ARC was $240 million. Partly because 70 percent of its employees are unionized, the county committed to full funding of the ARC over a 5 year period. Sonoma County, California, on the other hand, adopted a twopronged strategy: (1) provide personal, professional, and political recognition of healthcare costs and their importance, and (2) protect health retirement benefits over the long term by making them sustainable. This sustainability was defined as a benefit level where the county could provide services to citizens yet also offer a fair CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 379 compensation package to employees who deliver those services (Sanford, 2007). As a result, the county reduced its retiree healthcare contribution to a more sustainable level. Gwinnett County, Georgia, also made changes to its OPEB, but only changed the insurance funding mechanism and not access to healthcare (Sanford, 2007). County leaders, realizing that their OPEB liabilities were not sustainable, created an OPEB trust fund to set aside assets to cover the liabilities and changed from paying a percentage of the retiree healthcare premium to paying a defined monthly contribution. This strategy has allowed the county to essentially lock in its OPEB liabilities at more manageable levels. Finally, Chester County, Pennsylvania, had an unfunded OPEB liability of about $4.4 million, which prompted significant changes to its OPEB. To limit both exposure to such liabilities and the impact on taxpayers, the county eliminated the healthcare benefit for employees retiring after June 30, 2006. HOW HAVE CITY GOVERNMENTS RESPONDED? EXAMPLES FROM NINE MIDATLANTIC CITIES To provide a more comprehensive examination of how local governments have responded to GASB, we supplemented the finding of the NCSC study with examples of how cities have responded to the information concerning their liabilities. By combining these examples of nine Mid-Atlantic cities in Virginia, Maryland, and North Carolina with the fifteen case studies of counties, we are able to better represent the different types of local governments in the U.S. Three cities were selected from each of the three states: Baltimore, Gaithersburg, and Rockville in Maryland; Charlotte, Fayetteville, and Winston-Salem in North Carolina; and Alexandria, Richmond, and Virginia Beach in Virginia. These cities were selected to provide variety in terms of number of retirees, budget size, OPEB unfunded liability, and ARC. In addition we selected cities from both right-to-strike (Maryland) and right-to-work (Virginia and North Carolina) states. Table 1 summarizes the characteristics of the nine cities including government net assets, population, per capita income, UAAL and UAAL per capita. Appendix 2 provides a summary of the cities’ responses to GASB 45. As can be seen from Table 1 and Appendix 2, the cities selected for case study vary widely not only 380 YUSUF & MUSUMECI in terms of government size and affluence, but also with regard to OPEB-related characteristics such as the AAL, UAAL and UAAL per capita, ARC, and number of government retirees. The examples from the nine selected cities are discussed next. TABLE 1 OPEB Characteristics of Selected Cities (2008) Total Govt Net Assets (in billions) City Richmond, VA $2.489 Alexandria, VA $0.373 Virginia Beach, VA $3.093 Baltimore, MD $4.834 Gaithersburg, MD $0.165 Rockville, MD $0.271 Charlotte, NC $7.616 Fayetteville, NC $1.058 Winston-Salem, NC $1.413 Popula- Per Capita tion Income 200,123 141,000 434,072 637,455 59,912 63,170 695,995 181,453 224,889 UAAL (million) UAAL Per Capita $39,860 $76.0 $379.77 $65,141 $150.5 $1,067.38 $43,578 $127.7 $292.74 $32,445 $2,150.0 $3,161.01 $69,985 $18.0 $300.44 $53,754 $10.1 $160.46 $24,858 $229.7 $330.03 $34,245 $28.0 $149.33 $35,666 $79.0 $351.28 Analysis of the cities’ responses to GASB 45 and its informational content was based on examination of their respective comprehensive annual financial reports (FY 2006 through FY 2010) and other government documents. Data from FY 2006 through FY 2008 were used to determine how the cities responded to the implementation of GASB 45. Follow-up information using data for FY 2009 and FY 2010 was used to update the cities’ responses. City finance and budget directors were also surveyed to obtain information about compliance with GASB 45 and its impact on OPEB provision. Richmond, Virginia Of all the cities studied, Richmond, Virginia, had adopted the most proactive strategy to reduce its unfunded liability. The city’s original UAAL was $194 million, which prompted the city to make significant changes to its retiree healthcare program. This reduced the liability by almost $120 million to $76 million. The most CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 381 significant change was to eliminate the Medicare subsidy provided to its retirees. This was accomplished by enrolling retirees in the Medicare Advantage program instead of keeping them on the OPEB plan. Second, all employees hired after January 1, 1997, were moved into a DC retiree healthcare plan (in the form of a HSA), and those employees receive no retiree healthcare in the future. Finally, for current retirees and employees hired prior to January 1, 1997, the city froze its retiree healthcare contributions at the 2007 levels. By aggressively redesigning the OPEB, Richmond was able to significantly reduce and limit its OPEB liability. The city decided not to set up a trust fund to prefund the remaining benefits promised to their pre-1997 employees, but did set aside (in 2008) $650,000 towards future OPEB costs. In 2009 and 2010 the city set aside $1.4 million per year towards its OPEB liability. This amount was significantly less than the ARC and is roughly equal to the normal cost component of the ARC. Alexandria, Virginia The City of Alexandria, Virginia, decided to continue to provide retiree healthcare and to prefund the liability. This was accomplished by creating an OPEB trust fund and making an initial deposit of $5.6 million in 2008 as well as identifying an additional $10.7 million from the ending General Fund balance from FY 2008 for deposit into the trust fund. The city made several changes to its employee healthcare plan, which affected retirees, as they participate in the employee healthcare plan. The city’s monthly contribution is a fixed $260 per person toward the healthcare premium and the remaining balance is the responsibility of the retiree. Healthcare premiums were increased by 3.3 percent in FY 2007, and this increase was borne solely by retirees. In addition, retiree healthcare for employees hired after September 30, 2007, will be prorated based on the length of service. While not related to retiree healthcare, the city also decided that employees hired after July 1, 2008 would not receive the retiree life insurance benefit that had been previously provided. This had the effect of lowering the overall OPEB liability for the city. The city’s most recent actuarial valuation was on December 31, 2009 and at that time the OPEB trust fund was valued at $8.2 million and the AAL was $90.7 million (resulting in an UAAL of $82.5 million which is a funding ratio of 9%). This was after the city made the full ARC payment for both 2009 and 2010. 382 YUSUF & MUSUMECI Virginia Beach, Virginia The City of Virginia Beach decided that it would continue to provide retiree healthcare and that it would prefund the OPEB and establish a trust fund. In 2007 the city created an Employee Benefits Review Task Force, whose members represented City Council, city employees, public safety employees, and citizens and business owners. The Task Force developed an overall strategy of reducing the unfunded retiree healthcare liability and, to the extent practical, funding the ARC. Recommendations made by the task force became effective in FY 2009. The city’s first ARC was made on June 30, 2008, in the amount of $10.3 million. The city has contributed 100 percent of the ARC into the trust fund for all fiscal years since implementation of GASB 45. Reducing the OPEB liability involved significant changes to the OPEB promises. First, length of service necessary for retiree health insurance eligibility was increased from 5 years to 25 years. Second, the city shifted more costs to employees, in the form of reduced employer subsidies and increased employee co-pays, deductibles, and out-of-pocket maximums. The city eliminated the zero premium (i.e., 100% employer subsidy) healthcare option for single subscriber retirees, in addition to introducing a graduated subsidy plan capped at 50 percent. In FY 2010 the city offered employees the option to switch from the current DB plan to a DC plan with a HSA. However, limited enrollment in the HSA option indicates that it is unlikely to substantially reduce the unfunded liability. Baltimore, Maryland Baltimore has, by far, the largest liability of the cities included in this study. Maryland is a right-to-strike state, and Baltimore has union contracts that made it impossible for the city to consider changing or eliminating its OPEB to reduce the liability. Given the size of the UAAL ($2 billion for the city and its school system combined), carrying an unfunded liability of that magnitude on the financial disclosure and reporting documents could have a negative effect on the city’s bond ratings. Therefore prefunding was the only option, and in FY 2008 the city established an OPEB trust fund and contributed $183.3 million. This represented $18.7 million more than the ARC and was an increase of $78.6 million over the previous year’s PAYGO amount. For FY 2009 the city contributed $142.3 million to the OPEB CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 383 trust fund ($37 million less than the ARC). For FY 2010 the city again contributed $142.3 million to the OPEB trust fund (the ARC was $203.7 million). Despite the prefunding strategy, as of June 30, 2010, the UAAL for the city of Baltimore was $2.6 billion, an increase of $415 million since June 30, 2008. Gaithersburg, Maryland Gaithersburg, Maryland, adopted a proactive strategy, implementing GASB 45 a year earlier than required. The city decided to prefund its existing OPEB with no plan changes. The city pays 85percent of the retiree’s health and dental premiums as well as 100percent of the life insurance premiums. An employee must have at least 15 years of service to qualify for the OPEB. In FY 2007, the city established an OPEB trust fund with an initial deposit of $2.6 million (over two and a half times the ARC). However in subsequent years (FY 2008 through FY 2010) the city’s contribution to the trust fund was significantly below the ARC: 32 percent, 47percent, and 10percent of the ARC, respectively. According to the most recent actuarial valuation of the OPEB plan (completed July 1, 2009), the AAL was $24.9 million and the UAAL was $22.9 million. Rockville, Maryland The City of Rockville, Maryland, faced a very small OPEB liability, primarily because in order to qualify for the benefit, retirees must be at least 60 years old with ten years of service. This has had the effect of minimizing the “window” within which the retiree will receive OPEB before being eligible for Medicare at age 65. Because the benefits were already minimal, the city decided to make no plan changes. In 2009 the city established a trust fund and made the full ARC payment of $1.3 million. However, in the following year the city funded only $237,000 of the $1.4 million ARC. Charlotte, North Carolina Prior to GASB 45, the City of Charlotte, North Carolina, offered a modest OPEB in which retirees with at least 15 years of service were eligible for the healthcare plan and the city subsidized 85 percent of the premium. Employees who retired with at least ten but less than 15 years of service were allowed to participate in the plan but had to pay the full premium. Those with less than ten years of service 384 YUSUF & MUSUMECI received no benefits. With the implementation of GASB 45, no changes to those eligibility and participation requirements were made. Instead, the city adopted an aggressive prefunding strategy and focused on building the asset base needed to pay for future benefits. In FY 2008, the city established a trust fund and made a payment of $28.4 million (166% of the ARC). For 2009 and 2010 the city contributed 195 percent and 104 percent of the ARC payment, respectively. The most recent actuarial valuation (as of July 1, 2009) was $174 million, which suggests that the aggressive prefunding strategy has been able to reduce the OPEB unfunded liability. Fayetteville, North Carolina The City of Fayetteville, North Carolina, decided to continue to offer healthcare coverage to its retirees, but has made changes to the healthcare plan in an effort to reduce OPEB liabilities. For employees hired after February 1, 2008, eligibility to participate in the retiree healthcare plan is based on 20 years of credible service. This is a considerable increase from the previous requirement of only 5 years of service. In addition, retirees will only be able to carry a dependent on the plan if that dependent was on the plan for the consecutive five years prior to retirement, and the dependent must pay the full premium. However, the city has not yet decided whether to prefund the liability, and as of June 2010, the city was still using PAYGO. Despite continuing with the PAYGO approach and not prefunding, the city has seen a reduction in its UAAL, from $28 million in 2008 to $15 million in 2009, which could be largely due to changes to the plan. Winston-Salem, North Carolina The City of Winston–Salem, North Carolina, also decided to continue to provide retiree healthcare, while adopting an aggressive prefunding strategy. In January 2008 the city established an OPEB trust fund with $34.8 million (44.1% of the UAAL). In subsequent years (FY 2009 and 2010) the city contributed 100 percent of the ARC amount. The city also made plan changes to reduce its liability. For example, the city increased the required number of credible service CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 385 years to 15. Dependents will be allowed on the plan only if they pay the full group rate premium. For those retirees who qualify for the OPEB, the city limited its annual contribution to $2,400 per retiree, with increases in premiums, co-pays, and deductibles beyond this amount to be covered by the retiree. As of the most recent actuarial valuation on January 1, 2010, the OPEB liability was almost halffunded (48.5%), as the AAL was $80.6 million, and plans assets were valued at $39.1 million. CONCLUSION AND IMPLICATIONS The examples discussed in this paper show that local governments have responded to GASB 45 using a combination of strategies. The findings of the NCSC study of counties and the case studies of nine mid-Atlantic cities indicate that local governments have responded to GASB 45 and the knowledge of the magnitude of their OPEB liabilities in four different ways (see Figure 2). Improving the asset base appears to be a popular response to reducing the unfunded liability with almost 70 percent of the cities and counties included in this study taking steps to prefund their OPEB. However, some local governments (43.5%) have also responded by reducing or eliminating retiree healthcare. As illustrated in this paper, coming to grips with the OPEB unfunded liabilities has resulted in some local governments not only reconsidering their OPEB plan designs, but also contemplating eliminating their OPEB. While the severity of plan redesign and possible discontinuation of OPEB plan offerings by government employers will typically depend on the magnitude of the unfunded liability, the current fiscal environment facing governments is such that even those cities with smaller OPEB liabilities may consider reducing or eliminating their OPEB as a way to reduce future liabilities and outgoing cash flows. Several authors and studies have suggested measures for governments to manage their OPEB liabilities. For example, Russell (2008) suggested several approaches to controlling OPEB liabilities, including the following: - Modifying the cost-sharing philosophy by requiring retirees to pay an increased share. 386 YUSUF & MUSUMECI FIGURE 2 OPEB Liabilities in Four Different Ways PAYGO approach Prefund OPEB OPEB Assets OPEB Promises Maintain same benefits as preGASB 45 Maintain Status Quo Counties: Mecklenburg, NC; Multnomah, OR Prefund with Same Benefits Counties: Clark, NV; Fairfax, VA; Hillsborough, FL; Harris, TX; Montgomery, MD; Oakland, MI Cities: Baltimore, MD; Charlotte, NC; Gaithersburg, MD; Rockville, MD Reduce or eliminate benefits PAYGO with Lower Benefits Counties: Chester, PA; Marathon, WI Cities: Fayetteville, NC; Richmond, VA Prefund with Lower Benefits Counties: Gwinnet, GA; Shelby, TN; Sonoma, CA Cities: Alexandria, VA; Virginia Beach, VA; Winston-Salem, NC Note: Tulsa, OK was included in the NCSC study of county responses but is not included in this summary figure because the county does not offer a DB OPEB plan. Bernalillo, NM was also not included because its retirees participate in the state retiree health plan and the county has no liability. - Increasing eligibility requirements for receiving benefits or reducing benefits if retirement occurs after a certain date. - Limiting employer liability exposure by reducing duration of coverage, capping employer annual claims costs per retiree, and capping employer aggregate costs, or introducing DC plans. - Modifying benefit designs such as changes to deductible, copayments, co-insurance, and out-of-pocket limits. - Other cost management strategies such as switching retirees to Medicare advantage and Medicare supplement plans. Examples of how these approaches have been implemented by counties and cities in this study are presented in Table 2. CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 387 TABLE 2 Strategies and Approaches for Reducing Benefits Strategy/Approach Example Modifying the Cost-Sharing Philosophy Winston-Salem: Dependents pay full premium. Different subsidies for retirees vs. dependents Charlotte: Employees with 15 years of service Linking retiree contributions to years of are allowed to participate in the healthcare plan and the city pays 85 percent of the premium. service Employees who retire with at least ten but less than 15 years of service are allowed to participate in the plan but must pay the full premium. Increasing Eligibility Requirements for Receiving Benefits or Reducing Benefits If Retirement Occurs after a Certain Date Fayettevile: 20 years of credible service Increasing eligibility required for a retiree to participate in the requirements for healthcare plan. receiving benefits Chester County: Eliminate the healthcare Reducing benefits if retirement occurs after a benefit for employees retiring after June 30, 2006 certain date Limiting Employer Liability Exposure Limiting employer annual Winston-Salem: Limit the annual contribution to $2,400 per retiree, with increases in premiums, costs per retiree co-pays and deductibles beyond this amount to be covered by the retiree. Richmond: For current retirees and employees Capping employer hired prior to January 1, 1997, the city froze its aggregate costs contributions to retiree healthcare at the 2007 levels. Virginia Beach: Provide employees the option of Introducing DC plans switching from the current DB plan to a HSA (DC plan). – Changing deductibles, co- Virginia Beach: Increase retiree co-pays, payments, co-insurance, deductibles and out-of-pocket maximums. and/or out-of-pocket limits Richmond: Enroll retirees in Medicare – Switching retirees to Advantage program. Medicare supplement plans 388 YUSUF & MUSUMECI This study utilized a sample of 15 counties (from across the country) and nine Mid-Atlantic cities as the basis of the analysis of local government responses to GASB 45 and its informational contents regarding OPEB liabilities. The counties and cities varied in terms of size, measured as population, government net assets, and government workforce and retirees. Furthermore, the cities and counties represent both urban and rural areas, and right-to-work and right-to-strike states. While the cities are all from the mid-Atlantic region, there is no reason to believe that their responses to GASB 45 are driven by regional factors, particularly since many of their responses are similar to those of the counties. This suggests that the findings are, with some degree of confidence, generalizable to local governments throughout the country. Implications The goal of this paper was to highlight the variety of ways that local governments have responded to knowledge about their OPEB liabilities, and to show that some of these measures have resulted in reduced benefits for government employees. The next paragraphs discuss the implications of these responses to GASB 45. Clearly, the accounting and reporting standards of GASB 45 have serious implications for government employers. Clark (2008) argued that GASB 45 serves as “an important tool for policymakers and stakeholders in determining future compensation and employment policies and tax policies” (p. 3). In complying with GASB 45, governments are able to generate information about OPEB that is useful to policymakers, investors, and other stakeholders. While in the beginning GASB 45 may introduce staggering liabilities to government employers’ balance sheets, “in the longer term, it might bring into sharper focus for both the electorate and elected officials the fact that massive financial commitments have been made to public employees that will be progressively more difficult to meet” (Hurley et al., 2006, p. w203). Furthermore, information resulting from compliance with GASB 45 may shed light on instances where current and future funding of OPEB would strain a government’s operations or where conditions are such that governments are unable to fulfill their OPEB obligations (Bell, 2006). In the recent ICMA Survey of Local Government Employee Health Insurance Programs (ICMA, 2007), cities and counties were asked CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 389 the amount of their OPEB liabilities. Interestingly, 26 percent of respondents were not sure of their estimated OPEB liabilities. Clearly, the reporting requirements of GASB 45 are needed to ensure that those cities and counties are aware of the future funding needed to ensure that they can keep the promises made to their employees and retirees. As McKethan et al. (2006) suggested, “GASB 45 is an important and needed contribution to public accounting and oversight that will help ensure that public officials will take the future costs of such programs into account when setting retiree benefits and premiums” (p. 1523). It can be argued that the costs associated with reduced benefits are a reasonable tradeoff for the increased confidence of knowing that the needed assets have been put aside to deliver on those promises. For government employees, despite possibly experiencing reductions in OPEB, there is greater security in knowing that the retiree benefits retained will be more fully accounted for and better funded. It is also important to note that GASB 45 may not entirely be at fault for the reduction in retiree healthcare benefits. Beyond disclosing the liabilities and costs of governments’ OPEB, GASB 45 does not require governments to take any actions to manage their OPEB liabilities. Therefore, the resulting changes in benefits should not be blamed entirely on GASB 45. In many instances, the effect on OPEB of GASB 45 cannot be differentiated from other economic and fiscal factors. Clark (2008) argued that most government OPEB plans are regularly amended, as government employers periodically increase premiums, raise deductibles, increase co-payments, restrict choices, or raise eligibility criteria. Wiesnewski (2005, p. 118) suggested that implementation of GASB 45 is “less likely to influence the employer’s decision to continue to provide such benefits than other strong economic factors such as significant continued healthcare cost inflation and a continued deterioration in the active-to-retiree workforce ratio in state government employment.” In the face of the current fiscal crisis, it would not be surprising if more and more state and local governments turn to benefit plan redesign as a means of reducing their overall OPEB liabilities, not because of GASB 45 but because of economic pressures. Therefore, when coupled with factors such as the growth in medical and healthcare costs and the decreasing ratio of active employees to retirees in the public sector, 390 YUSUF & MUSUMECI GASB 45 may result in local governments reducing their OPEB in order to avoid additional liabilities. Finally, GASB 45 may also have implications for how governments think about their other unfunded liabilities. The Cobalt Community Research survey (2009) asked respondents whether their GASB 45 and OPEB experiences had heightened their awareness of other longterm liabilities. A majority (59%) of respondents answered in the affirmative; 27 percent indicated that GASB 45 has increased their awareness of other similar liabilities and had prompted them to begin planning for these liabilities. The other 32 percent, while being more aware, were insufficiently concerned about the liabilities to begin planning for them. Even while government employers deal with issues arising from the information provided through the implementation of GASB 45, several other concerns must also be addressed. Surrounding the overarching questions of whether to prefund (and the potential fiscal impact of higher contributions associated with prefunding) and whether the existing benefit structure will be maintained are key issues such as (1) the extent to which any decreases in benefits undercut government’s competitive ability to hire needed employees; (2) the affordability of prefunding; (3) the political hazards associated with addressing OPEB liabilities; (4) legal issues associated with OPEB levels and the possible reductions in such levels; and (5) the effect on bond ratings of decisions to address OPEB liabilities (Young, 2005). These questions were not addressed in this study. However, as more local and state governments are forced to come to grips with their OPEB plans and the associated liabilities, these issues will come increasingly to the forefront of policy discussion. These issues provide ample opportunity for continued future research on the financial, human resources, and political implications of GASB 45 and the recent realization by government employers of their OPEB liabilities. NOTES 1. It may have been the absence of both the equivalent standards (to GASB 45) and the greater disclosure and transparency resulting from such standards that precipitated the underlying financial problems. However, the additional financial pressures CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 391 associated with making the ARC payments could contribute to substantial additional stress. 2. While it is possible that another option would be to increase OPEB benefit levels, existing concern over sustainability of OPEB promises suggest that it is unlikely that government employers will raise benefits. REFERENCES Bell, L.L. (2006, March-April). “GASB 34 and GASB 45 OPEBs New Problems, Old Solutions?” Journal of Retirement Planning, 9 (2): 29-37. Binnis, H. & Riffe, S. (2000). “Cost/Benefit Tradeoffs Relating to Reductions in Postretirement Health Care Benefits in the SFAS Environment.” Journal of Applied Business Research, 16 (4): 95104. Clark, R.L. (2008). Financing Retiree Health Care: Assessing GASB 45 Estimates and Liabilities. Washington, DC: Center for State and Local Government Excellence. Cobalt Community Research (2009). Health and OPEB Funding Strategies: 2009 National Survey of Local Governments. Lansing, MI: Cobalt Community Research. Coe, C.K. & Rivenbark, W.C. (2010). “Implementing GASB 45: Recommended Best Practices in Local Government.” Public Budgeting and Finance, 30(4): 71-81. Coggburn, J.D. & Kearney, R.C. (2010). “Trouble Keeping Promises? An Analysis of Underfunding in State Retiree Benefits.” Public Administration Review, 70 (1): 97-108. Feldman, M. & Haynes, R. (2007). “Effect of New GASB 45 Accounting Rules: What We Can Learn from FAS 106.” Benefits & Compensation Digest, 44 (3): 18-21. Governmental Accounting Standards Board. (2004). Accounting and Financial Reporting by Employers for Post-Employment Benefits Other than Pensions: Statement No. 45. Norwalk, CT: GASB. Governmental Accounting Standard Board (n.d). Other Postemployment Benefits: A Plain-Language Summary of GASB 392 YUSUF & MUSUMECI Statements No. 43 and No. 45. [Online]. Available at http://www.gasb.org/cs/ContentServer?c=Document_C&pagena me=GASB%2FDocument_C%2FGASBDocumentPage&cid=11758 04850577. [Retrieved October 10, 2010]. Hurley, R.E., Felland, L., Gerland, A. & Pickreign, J. (2006). “Public Employees’ Health Benefits Survive Major Threats, So Far.” Health Affairs, 25 (3): w195-w203. International City/County Management Association (2007). Local Government Employee Health Insurance Programs, 2006. Washington, DC: International City/County Management Association. Keating, E.K. & Berman, E.S. (2007). “Unfunded Public Employee Health Care Benefits and GASB No. 45.” Accounting Horizons, 21 (3): 245-263. Kilgour, J.G. (2009). “Public Sector Retiree Benefits in California: Problems and Solutions.” Compensation Benefits Review, 41: 2732. Marlowe, J. (2008). “Financial Management Challenges of Other Postemployment Benefits.” In C.G. Reddick & J.D. Coggburn (Eds.), Handbook of Employee Benefits and Administration (pp. 211-234). Boca Raton, FL: CRC Press. McKethan, A., Gitterman, D., Feezor, A. & Enthoven, A. (2006). “New Directions for Public Health Care Purchasers: Responses to Looming Challenges.” Health Affairs, 25 (6): 1518-1528. McTague, J. (2006, March 13). “The Trillion Dollar Pothole.” Barron’s Magazine: 37. Mead, D.M. (2008). “Accounting and Financial Reporting by Governments for Retirement Benefits: Understanding and Using the Information in Audited Financial Reports.” In C.G. Reddick & J.D. Coggburn (Eds.), Handbook of Employee Benefits and Administration (pp. 285-307). Boca Raton, FL: CRC Press. Mittelstaedt, H., Nichols, W. & Regier, P. (1995). “SFAS 106 and Benefit Reductions in Employer-Sponsored Retiree Health Plans.” Accounting Review, 70(4): 535-556. CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 393 Moran, J.A. (2010). “The OPEB Tsunami: Riding the Wave of Public Sector Postemployment Health Benefits.” Buffalo Law Review, 58 (3): 677-713. Patterson, M.P. (2001). “State and Local Governments Plan Alternatives for Retiree Medical Benefits Funding.” Compensation and Benefits Review, 33 (2): 37-49 Pearson, T.A. & Jerris, S.I. (1995). “Accounting for OPEBs: Does Better Accounting Serve the Public Interests?” Journal of Applied Business Research, 11 (3): 1-7. Russell, R.E. (2008, December). “GASB 45: Governmental Employers are Seeking Balanced Solutions.” Benefits and Compensation Digest: 34-38. Sanford, P. (2007). The Implementation of GASB 45 Case Studies of 15 Counties. Report prepared for the National Center for the Study of Counties. Washington, DC: National Association of Counties. Voorhees, W.R. (2005). “Counting Retirement Expenditures Before they Hatch: GASB and the New Reporting Requirements for Other Postemployment Benefits.” Public Budgeting and Finance, 25 (4): 59-71. Wisniewski, S.C. (2005). “Potential State Government Practices Impact of the New GASB Accounting Standard for Retiree Health Benefits.” Public Budgeting and Finance, 25 (1): 104-118 Young, P. (2005). “Funding OPEB Liabilities: What Are Your Options?” Government Finance Review, 21 (6): 10-15. Zion, D. & Varshney, A. (2007). You Dropped a Bomb on Me, GASB: Uncovering $1.5 Trillion in Hidden OPEB Liabilities for State and Local Governments. New York: Credit Suisse Equity Research. 394 YUSUF & MUSUMECI APPENDIX 1 Summary Information for Counties in the NCSC Study(a) Government Response to GASB 45 and Employment Information About OPEB AAL ARC Liabilities County ($million) ($million) FTE County participates in the Bernalillo, state retiree healthcare NM n/a n/a 1,608 program and has no liability County already offered HSA, so only have minimal liability due to implicit rate subsidies Retirees have access to county health plan but pay 100% of premiums Tulsa, OK 1 n/a n/a Eliminated healthcare benefits for employees retiring after June 30, 2006 Chester, PA 4 1 2,290 County already offered HSA, so only have minimal liability Marathon, due to implicit rate subsidies WI 5 0.1 838 Cost-allocate the ARC across Hillsbothe department as part of the rough, FL 100 8.2 10,429 cost of each FTE Multnomah, PAYGO for 2008 OR 110 12.7 4,281 No plan changes Switched from a DB plan to a DC plan Created a locally controlled Gwinnett, OPEB trust and pension plan GA 140 13.6 4,586 No changes to benefits as the county recognizes that OPEBs are important Mecklenrecruitment/retention tools burg, NC 141 14.6 4,282 Focus on funding the ARC Created OPEB trust fund Fairfax, VA 143 15.2 10,999 No changes to health plans Significantly reduced retiree benefits package for new employees No changes for current employees Shelby, TN 267 28.2 6,277 Partially funding ARC Clark, NV 372 49.7 18,705 No changes to health plan CHANGING AND/OR FUNDING OTHER POST-EMPLOYMENT BENEFITS PROMISES 395 APPENDIX 1 (Continued) Government Response to GASB 45 and Employment Information About OPEB AAL ARC Liabilities County ($million) ($million) FTE Reduced employer contribution to the healthcare Sonoma, CA 382 37.2 4,154 premium Has been prefunding OPEB since 1980s and full AAL is advance funded Introduced HSA Oakland, MI 830 60.2 4,536 Significantly reduced retiree benefits package for new employees No changes for current employees Harris, TX 834 90.0 15,840 Full funding of ARC implemented over a five year Montgomery, period MD 2,600 240.0 9,089 Notes: Information about the number of government retirees is not available. Source: Research by the National Center for the Study of Counties (Sanford 2007). APPENDIX 2 Summary Information for Selected Cities (2008) Response to GASB 45 Govt AAL ARC Govt Employees and Information about ($million) ($million) Retirees OPEB Liabilities (FTE) Phasing out OPEB through significant plan Richmond, changes VA(a) 76.1 4.6 1,172 8,940 Established OPEB Trust Fund with initial deposit of $5.5 million Increased employee Alexandria, premiums VA(a) 88.1 17.9 1,015 4,472 Established OPEB Trust Fund and made the ARC payment Changed plan for Virginia retirees and employees Beach, VA 84.7 10.3 574 6,762 City 396 YUSUF & MUSUMECI APPENDIX 2 (Continued) Response to GASB 45 Govt ARC Govt AAL Employees and Information about ($million) ($million) Retirees OPEB Liabilities (FTE) Established OPEB Trust Fund with ARC plus $18.7 million Baltimore, 214,454 MD(a) .0 164.6 21,017 28,171 No changes to plans Established OPEB trust fund with an initial deposit of 250% of ARC Gaithers No changes to plan burg, MD 9.7 0.9 27 363 Established OPEB trust fund with ARC payment Rockville, No changes to plan MD 8.8(b) 1.3 18 535 Established OPEB Trust Fund with initial deposit of 166% of the ARC No changes to health Charlotte, plans NC 229.7 17.0 1,895 6,577 Set aside less than $1 million towards liability Fayetteville, NC 28.0 6.6 436 1,996 Made plan changes Established OPEB Trust Fund with initial deposit of $34.8 million (44% of UAAL) WinstonSalem, NC 79.1 $6.98 1,056 2,660 Made plan changes City Notes: (a) Figures are for city government and school district combined. (b) As of January 1, 2009. J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT, 24 (3), 397-428 FALL 2012 CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS: A COMPARATIVE ANALYSIS Larita J. Killian and Kimdy Le* ABSTRACT. Special districts comprise over 40% of local governments, prompting debate on the merits of general versus special entities. Previous research focused on relative cost-efficiency and tended to ignore how special districts impact government accountability. This study fills a critical need by testing how type of government (general versus special) impacts citizen awareness of and familiarity with government, a precondition for accountability. Drawing from two theoretical perspectives (institutional reform and public choice), we used survey research to assess familiarity with the goals and performance and financial practices of local governments in Indiana. Our study participants were more aware of and familiar with general than special governments, which supports the institutional reform perspective more than public choice. We conclude that efforts to improve accountability should be expanded to include special districts. INTRODUCTION Special districts are the growth area of government. Between 1952 and 2007, the number of independent school districts in the United States decreased by 81% and the number of general purpose local governments increased by just 5% (all growth was among municipalities), yet the number of special districts increased over 200%. Per census data, special districts now comprise nearly 42% of all local governments (U.S. Census Bureau, 1992; 2007a). The remarkable proliferation of special districts has prompted scholars -----------------------* Larita J. Killian, Ed.D, CPA, is an Assistant Professor, Division of Business, Indiana University at Columbus. Her teaching and research interests are in governmental accounting, special districts, and accounting pedagogy. Kimdy Le, Ph.D., is an Assistant Professor of Psychology at Indiana University at Columbus. His teaching and research interests are in personality theory, statistics, and measurement. Copyright © 2012 by PrAcademics Press 398 KILLIAN & LE and policymakers to debate the relative merits of general-purpose and special district governments. But a critical aspect of the debate, the relationship between local government structure and democratic accountability, suffers from a lack of empirical evidence. This study addresses that problem by testing citizen familiarity with the generalpurpose and special district governments that serve them -- a precondition for government accountability. Of course, the fact that citizens are aware of and familiarity with the governments that serve them does not guarantee accountability, but without such awareness and familiarity, citizens cannot hold governments accountable. This study uses the U.S. Census Bureau (2002a) definition of special districts, which includes many entities with “authority” in their titles. Several factors account for the growth of special districts, including inflexible or unsuitable municipal annexation laws (Bollens, 1957; Foster, 1997); the desire to circumvent debt limitations or personnel policies imposed on general-purpose governments (Bollens, 1957; Smith, 1969; Advisory Commission on Intergovernmental Relations (ACIR) 1987; Leigland, 1990; Axelrod, 1992); the desire to attract technical experts who prefer to avoid the glare of partisan general elections (Smith, 1969); and federal government policies that direct funds to special districts (Smith, 1969; Foster, 1997). Wood (1961) referred to tradition and political style as factors behind the increase of special districts. For instance, a growing community may face a critical decision on provision of water services for a new housing development: the new development can be annexed into the previous municipality, or a special water district can be created. Whatever course is chosen sets a precedent and becomes a learned response for future decisions. “As time goes on, the alternative first adopted tends to become a governmental tradition, a political style, and even may crystallize into law” (Wood, 1961, p. 66). Regardless of the explanations for the growth of special districts, a democracy cannot afford to ignore the impact on accountability. Government accountability is a multi-faceted and evolving concept. Most observers agree that accountability begins with adequate financial reporting, and in recent years, the need for performance reporting has gained acceptance (Harris, McKenzie & Rentfro, 2011). Ingram and Copeland (1981) found that accounting information is useful to voters in deciding whether to retain CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 399 incumbents. Guo, Fink and Frank (2009) found that financial reporting is a critical foundation of accountability and that the contemporary political and economic environment requires a higher quality of financial disclosure. They conclude, however, that the lack of norms (such as trend data or comparisons with referent groups) detracts from the ability of financial reports to enhance accountability (Guo, Fink & Frank, 2009). Lowery (2001) found that for citizens to exercise meaningful democratic control, they must have valid information on what government is doing and how well it is doing it. He asserted that a fragmented (versus consolidated) government structure provides more opportunities for public officials to manipulate reported information and thus weaken democratic accountability (Lowery, 2001). While adequate financial and performance reporting is necessary for accountability, it is not sufficient. Most definitions of accountability involve a dynamic exchange of information between government and citizens. If the exchange is ineffective, accountability is not achieved. The Governmental Accounting Standards Board (GASB) included the notion of active discourse in its definition of accountability: it includes the citizens’ “right to know” and to receive facts about government costs and operations that may “lead to public debate by the citizens and their elected representatives” (GASB, 1987). Wells and Scheff (1992) identified six measures of accountability, including “public accountability” which is indicated by the level of citizen participation in government affairs. Per Wells and Scheff, other measures of accountability include market accountability, political accountability, bureaucratic accountability, legal accountability, and professional accountability. Chan and Rubin (1987) wrote that governmental accounting and financial reporting stem from the normative theory of democracy which commends citizens to be well-informed; further, per demand models of government, public policies emerge from the direct and active demands of citizens. Williams (2002) found that accountability involves more than the provision of technical economic facts; it implies a moral imperative and “permits making explicit who is accountable to whom and for what” (2002, 5). Sen (1999) acknowledged the dynamic nature of accountability when he wrote that political participation is intrinsic to human well-being; to be prevented from participation is a “major deprivation.” Further, Sen (1999) found that public discussion and exchange of information are 400 KILLIAN & LE required for identifying and prioritizing economic needs and for keeping democratic governments responsible and accountable. From the above, we can identify key elements of government accountability. It involves the exchange of information between a government and its citizens; discussion or debate that serves to identify and prioritize civic needs; and voter judgments regarding the responsiveness of government officials. Government accountability is a growing concern, especially in the current climate where budget crises are forcing trade-offs and cutbacks in services, and where states such as Indiana, New York, and Ohio are working to streamline government and improve responsiveness.1 Efforts to assess or improve accountability, however, have tended to focus on generalpurpose governments and to by-pass special districts (Killian, 2011). Our research fills a critical need by extending the investigation of accountability to special districts. There are reasons to question the relative accountability of special districts as compared to general-purpose governments. After studying compliance with generally accepted accounting principles (GAAP) among state and local governments, GASB (2008) estimated that special districts have the lowest rate of financial reporting compliance of all local governments. Special districts also have a lower profile with regard to performance reporting. For instance, the Association of Government Accountants (AGA) posts information about performance reporting on its website, listing entities that have recently earned certificates of excellence for Service Efforts and Accomplishments (SEA) reporting and entities that have voluntarily submitted citizen-centric reports; special districts are rarely found on these lists. The U.S. Census Bureau reported that special districts had the lowest response rate to the organization survey for the 2002 census of governments (U.S. Census Bureau, 2002a). In short, while accountability is a concern for all types of government, citizens may be even less familiar with the finances and performance of special districts. This would magnify the “accountability deficit” for special districts relative to general-purpose governments. Most research on general versus special government has addressed the relative cost-efficiency of providing a given service through general or special government; questions of accountability have been vigorously debated but generally ignored in research. This situation obtains, in part, because it is more difficult to assess CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 401 accountability than to measure cost-efficiency (Lowery, 2001). We agree with Lowery (2001) that direct measures of accountability are difficult to obtain. We can contribute to the debate, however, by measuring a precondition for accountability: citizen familiarity with their governments. The key elements of accountability (information exchange between citizens and government, discussion and debate about priorities, voter judgments of officials) require citizen awareness of and familiarity with the governments that serve them. By comparing citizen familiarity with general-purpose and special district governments, we can test whether the form of government impacts this prerequisite for accountability. THEORETICAL PERSPECTIVES AND RESEARCH QUESTIONS Several theoretical perspectives have been applied to comparisons of general-purpose and special district governments (Foster, 1997). The “institutional reform” and “public choice” perspectives are most relevant to our research. Both perspectives compare general-purpose and special district governments along three dimensions: economy and efficiency of operations, policy alignment and allocation of resources, and democratic accountability. While a thorough analysis of these competing perspectives is beyond the scope of this paper, a brief summary is provided below. The institutional reform perspective originated with scholars such as Bolens (1957) and Smith (1969). It arose in response to the increasing fragmentation and layering of government that became apparent during the 1950’s and which continues, though at a reduced pace. Institutional reformers believe that a complex, fragmented government structure increases the total cost of government through increased administrative and coordination costs. Moreover, a fragmented structure confuses citizens and reduces their ability to hold public officials accountable. Institutional reformers often propose the consolidation of government entities to reduce fragmentation and improve efficiency. This perspective enjoyed early acceptance, though actual efforts to consolidate governments have usually ended in failure (Faulk and Hicks 2009). In recent decades, the public choice perspective has been ascendant. However, scholars such as Barlow (1991) and Lowery (2001) kept the institutional reform argument alive. Lowery (2001) identified an emergent, “neoprogressive” perspective that is grounded in the 402 KILLIAN & LE institutional reform tradition but places greater reliance on social science methods and empirical tools to combat the public choice argument. Recent efforts at government consolidation, such as proposals to consolidate or eliminate townships in Indiana (Indiana Commission on Local Government Reform (ICLGR), 2007) demonstrate that the institutional reform perspective retains currency. In contrast, public choice advocates believe that consolidation and centralization lead to monopolistic government with the power to force certain services upon citizens and require them to pay for the services, even if the services do not correspond to citizen preferences. The public choice argument stems from the famous Tiebout hypothesis. Tiebout (1956) argued that there is a market in governments, just as there is a market for commercial goods and services. Citizens have the power to “vote with their feet” and move to the community that offers the basket of government services they prefer at a price (tax levy) they can afford. Building upon the Tiebout hypothesis, public choice scholars such as Delorenzo (1981) and Mehay (1984) argued that the proliferation of governments results in the best value for consumer-citizens, if not always the lowest price. For instance, citizens might choose to pay more for higher quality or increased quantity of service. When institutional reformers claim that consolidated governments can provide services at less cost due to economies of scale, public choice scholars reply that small, specialized governments can also achieve economies through jointcontracting or service agreements. Moreover, Bish and Ostram (1973) found that some services, such as education and police protection, are highly sensitive to local conditions and that consolidation of these services on a large scale could lead to diseconomies. After decades of research, the argument between institutional reform and public choice is inconclusive. Studies can be found to support both perspectives; this is not surprising given the complexity of the subject matter, varying local contexts, and variations in state enabling laws for general and special governments (McCabe, 2000). Foster (1997) conducted a meta-analysis and reported that the preponderance of evidence supports the institutional reform perspective. A continuing problem, however, is that prior research has emphasized the cost-efficiency dimension while other CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 403 dimensions, such as the impact of government structure on accountability, remained virtually unexamined. Focus on Democratic Accountability As shown in the introduction, accountability involves the exchange of information between a government and its citizens; discussion or debate to identify and prioritize needs; and voter judgments regarding the responsiveness of government officials. At a minimum, this requires citizen awareness of which government entities are responsible for which services. Critics charge that the proliferation of special districts confuses citizens and detracts from democratic accountability. For instance, Bollens (1957) referred to fragmented local government as a three-ring circus that sows confusion and indifference among citizens, especially regarding special districts. “Twice removed” from special district officials, citizens are frequently unable to bring direct action against them (Bollens, 1957). Axelrod (1992) wrote that special districts insulate officials from voters; normal, democratic checks and balances do not apply. Peddie (2007) found that special district elections do not coincide with general elections, draw little voter interest, and are usually uncontested. Spitzer (2007) concluded that many citizens are simply unaware of the existence of special districts. The opposing, public choice argument is offered by scholars such as Wagner and Weber (1975), Mehay (1984), and the later ACIR (1987).2 They conclude that citizens can monitor small, specialized units of government more effectively than large, general governments. Wagner and Weber (1975) noted that a fragmented and overlapping government structure is preferable to having a reduced number of general governments that can act as monopolistic (and thus unresponsive) rather than competitive suppliers of services. Mehay (1984) found that citizens may be less able to monitor officials in general purpose entities because the (presumably) larger bureaucracy associated with general governments is harder to penetrate. In 1987, ACIR took a nuanced and carefully stated position: ACIR found that citizens can monitor a single, special district government with ease, though the proliferation of special districts exacerbates the accountability challenge. In summary, institutional reformers and public choice advocates agree that public awareness of service outcomes and costs, and the 404 KILLIAN & LE ability of voters to judge government officials, are essential to democratic accountability. They disagree, however, on the optimal government structure to achieve these ends. Our research addresses this problem by comparing citizen familiarity with general-purpose and special district governments. Research Questions In testing for comparative awareness and familiarity, we distinguish between a government’s goals and performance and its finances. GASB (1994) noted that citizens need information about government performance as well as government finances. Sanders (1994) agreed that citizens need both types of information, but noted that due to “fiscal illusion” citizens are more likely to have information about performance than finances. For instance, citizens can often experience or observe the quantity and quality of services provided, but they may lack information on the cost of those services. The problem is exacerbated when several small, overlapping taxing entities are consolidated onto a single county tax bill (Sanders, 1994). Since there is no objective standard for adequate civic information, we ask citizens to rate their own level of familiarity with the governments that serve them.3 This is appropriate for our research because the absolute level of familiarity is not the issue; rather, we are concerned with the comparative levels of familiarity regarding general-purpose and special district governments. We also want to test whether citizens perceive they have enough information about the governments that serve them, so we include questions on the ideal level of familiarity for a given government. Finally, we want to test whether the disparity between actual and ideal familiarity differs between general-purpose and special district governments. These goals lead to the following research questions: Goals and performance: Question 1: Do citizens differ in their familiarity with the goals and performance of general-purpose versus special district governments? Question 2: Do citizens differ in their perceptions of ideally how familiar they should be with the goals and performance of general-purpose versus special district governments? CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 405 Question 3: Is the disparity between actual and ideal familiarity with goals and performance different for general purpose versus special district governments? Finances: Question 4: Do citizens differ in their familiarity with the financial resources of general purpose versus special district governments? Question 5: Do citizens differ in their perceptions of ideally how familiar they should be with the financial resources of general purpose versus special district governments? Question 6: Is the disparity between actual and ideal familiarity with financial resources different for general purpose versus special district governments? METHODOLOGY The distribution of special districts varies appreciably from state to state, as do the powers granted to special districts in state enabling laws (McCabe 2000; Foster, 1997). This situation constrains the potential for valid, cross-state research. For this pilot study, we limited our sample to a single state, choosing the state of Indiana. In recent years, there have been high-level studies on how to streamline and improve Indiana government (ICLGR, 2007; Faulk and Hicks, 2009), and proposals to restructure local government have been repeatedly introduced into the state legislature. However, these past studies and proposals usually focused on general-purpose governments and bypassed special districts. A study that compares how general and special governments impact a precondition for accountability brings needed attention to this issue and contributes to the ongoing policy debate. Further, we live and work in Indiana and our university strongly encourages service to the region. By focusing on special districts in Indiana for this initial study, we are upholding an important institutional value. Research Instrument We designed an instrument to measure citizen awareness of the goals and performance and financial resources of local governments, shown in Appendix A. Our instrument is adapted from a survey that 406 KILLIAN & LE was commissioned by the Association of Government Accountants (AGA) in 2008. (AGA has since repeated the survey). The AGA survey addressed three levels of government (national, state, and local) but did not differentiate among entities at the local level (AGA, 2008). Our survey addresses local government exclusively and separates general purpose from special district entities. Like the AGA survey, our instrument uses a 10-point Likert-type response scale; this scale provides ample flexibility to participants yet avoids neutral (numerical middle) responses. We conducted a validity test of our instrument, and it was reviewed by academic staff members of the Center for Evaluation and Education Policy at Indiana University. Our analysis includes two general-purpose and two special district governments that serve every citizen of Indiana.4 The generalpurpose entities are counties and townships. Indiana is unique in that every citizen is served by both county and township government. Townships are distinct from cities and towns. All citizens are served by counties and townships, and some are also served by cities or towns. To minimize bias in research design (i.e., to avoid the more obscure special districts) we included two of the most common special districts in the state, solid waste management districts and soil and water conservation districts. The boundaries of these special districts coincide with county boundaries, though in a few cases several adjacent counties are joined in a single, solid waste management district. Indiana has 92 counties, 92 soil and water conservation districts, and 68 solid waste management districts (U.S. Census Bureau, 2007b). The counties, townships, and soil and water conservation districts are governed by an elected board or council; solid waste management districts have appointed boards comprised of officials who were elected to other posts. For instance, one member of the solid waste management board is appointed from elected members of the county council. For each government entity, the survey asks, Are you served by a (entity)? Participants indicated Yes, No, or Uncertain. The Uncertain option was included to prevent guessing and because the literature suggests significant confusion among citizens regarding which entities serve them (Axelrod, 1992; ICLGR, 2007; Spitzer, 2007; New York State Commission on Local Government Efficiency and Competitiveness (NYSCLGEC) 2008; Sanders, 1994). Thus, uncertainty is a legitimate response and a forced choice between Yes CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 407 and No would be invalid. When participants selected No or Uncertain to being served by an entity, they were prompted to move on to the next entity. When they selected “Yes” to being served, they were asked to rate their familiarity with the goals and performance of the entity and their familiarity with how the entity obtains and uses financial resources (separate questions). Responses were captured with a Likert-type response scale where 1 meant “Not at all familiar” and 10 meant “Very familiar.” Using the same scale, participants also indicated the ideal level of familiarity the average person should have with the goals and performance and financial resources of the given entity. To ensure participants were not influenced by order effects, each survey alternated presentation of general and special entities (general, special, general, special). We also created three different sequences (A, B, and C) of entity presentation. For both hard-copy and internet surveys, the three sequences were judiciously represented. The hard-copy surveys were collated so each sequence was presented in turn (A, B, C, A, B, C, etc.) For internet data collection, participants were randomly directed to one of the three sequences; there were 71 responses for sequences A and B, and 69 responses for sequence C. The research instrument also collected demographic data: gender and citizenship status were captured using simple checklists; educational attainment and household income were captured as interval data using checklists; and participants entered their zip code and age. The survey instrument was pretested on 40 diverse participants to determine whether it was too long, whether it was clear, whether participants would be willing to answer the questions honestly, and to examine whether the questions asked were reliable and valid and whether they correlated with each other in expected ways. Within each entity, familiarity ratings for goals and performance and financial resources were moderately to highly correlated, indicating that people who were familiar with one aspect of a particular government (goals and performance) were familiar with other aspects (financial resources). However, there were low correlations between government entities, indicating that knowledge of one government did not relate highly to knowledge of other governments. Thus, the pretest of the survey instrument yielded high internal consistency and 408 KILLIAN & LE some discriminant validity. We would not expect this pattern of results if participants were answering randomly to the questionnaire. Participants and Sample Size We collected responses from voting-eligible residents of Indiana using hard-copy and internet (electronic) surveys; both formats contained identical questions. We used a convenience sampling technique and gathered responses from several venues to obtain as large and diverse a population as possible. Hard-copy data were collected from volunteer participants at seven venues (an outdoor farmers’ market, a bingo game at a fraternal organization, a seniorcitizen center, a large grocery store, a non-profit community services center, a working-class shopping mall, and a college campus). For the internet questionnaire, we invited participants to respond via links in an online newspaper and via emails sent to a college alumni mailing list. Hard-copy participants received modest food incentives (snack bars); internet participants received no incentives. We strove for balance in our data collection: hard-copy collection on a college campus was balanced with hard-copy collection at a weekly bingo game, a human services center, and a shopping mall in a workingclass neighborhood. The alumni mailing list from a small public college (both graduates and drop-outs) was balanced with the mailing list of a popular regional newspaper. We collected responses from 444 participants (hard-copy and internet combined). To limit the survey to voting-eligible residents of Indiana, we excluded responses from six participants who were less than 18-years of age; from 75 participants who were not United States citizens or who provided no citizenship status; and from 16 participants who provided no zip code, provided a non-Indiana zip code, or provided a zip code found in Marion County. (As explained in Note 4, residents of Marion County were excluded from this study.) Finally, we excluded responses from 12 participants whose educational attainment was less than a high school diploma or equivalent: due to their low number, they constituted outliers, and we felt that excluding these outliers was appropriate even though it exerted upward pressure on educational attainment of the sample. The resulting data set contains responses from 335 participants, 190 hard-copy and 145 internet, drawn from 21 Indiana counties. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 409 We tested for demographic differences between the hard-copy and internet samples to see if, statistically, they were significantly different from each other. We conducted an independent samples ttest to test for age differences and chi-square tests to test for whether gender, educational attainment, and household income were different for the two samples. The only statistically significant difference occurred with age: the internet participants (Median age = 52) were a few years younger than the hard-copy participants (Median age = 55). Generally, the two groups did not differ demographically. Next, we tested to see whether the two samples differed from each other on familiarity ratings. There were no statistically significant differences in familiarity scores between the hard-copy and internet participants, so we combined them into a single sample for analysis. Appendix B summarizes the demographic profile of our sample and provides comparative data for registered and reported voters in Indiana, where available. “Reported voters” are those who indicated having voted in the 2008 general election when responding to the U.S. Census Bureau’s November 2008 Current Population Survey. Where Indiana-specific data are not available, we show comparative data for all registered and reported voters in the United States. The comparative data in Appendix B are from the U.S. Census Bureau (2008). The median age of our sample (52.3) falls within the median age range for registered and reported voters in Indiana (45-64). Likewise, the gender distribution of our sample (54% female) corresponds to the gender distribution of Indiana registered and reported voters (52.8% and 53.2% female, respectively). We do not have access to education and income data for Indiana voters, so Appendix B shows education and income data for all registered and reported U.S. voters. The median household incomes of our hardcopy and internet participants ($50,000-$74,999 and $75,000$99,999, respectively) correspond to the median household incomes of registered and reported U.S. voters, respectively. The median income of our total sample ($75,000-$99,999) corresponds to the median income of all U.S. voters. The U.S. Census Bureau (2009a) reports that median income for all Indiana households, not just voters, is $45,000-$49,999. This is below the median income for all U.S. households ($50,000-59,999), so we conclude that our participants, whose incomes correspond to 410 KILLIAN & LE all U.S. voters, have incomes somewhat above all Indiana voters. The educational attainment of our participants is above the educational attainment for all U.S. registered and reported voters. Since educational attainment in Indiana is slightly below the national average (U.S. Census Bureau, 2009b) we conclude that the educational attainment of our sample is above educational attainment for all Indiana voters. In summary, our survey participants correspond to all Indiana voters in both age and gender distribution; they have household incomes that correspond to U.S. voters but are somewhat above household incomes for Indiana voters; and they are more highly educated than the medians U.S. and Indiana voter. As explained in the Discussion section, we believe the demographic profile of our sample strengthens our conclusions. RESULTS Awareness of Government Entities For every government entity on the survey, the initial screening question is, Are you served by a (entity)? Results for this question are shown in Tables 1 and 2. All participants included in the analysis are served by all four entities. Yet as shown in Table 1, only 73 participants (22%) were aware of being served by all four entities. Twenty-one participants (6.3%) were unaware of being served by any of the entities, and about half (46%) were aware of being served by only one or two of the entities. TABLE 1 Positive (Yes) Responses on Being Served by One, Two, Three, or Four Entities (Frequency and Percent) N= (%) 0 Entity 21 (6.3%) 1 Entity 59 (18%) 2 Entities 95 (28%) 3 Entities 87 (26%) 4 Entities 73 (22%) Table 2 shows the Yes, No, and Uncertain responses for individual entities. With a sample of 335 individuals and an analysis that covers two general and two special governments, there were 670 opportunities for a Yes response on general governments and 670 CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 411 opportunities for a Yes response on special district governments. However, general entities received a total of 458 Yes responses and special entities received a total of 344 Yes responses. In the aggregate, citizens are more aware of being served by generalpurpose than by special district governments. TABLE 2 Awareness of Being Served by Individual Entities (Frequency and Percent) County N* 314 Township 314 Solid Waste Management Soil & Water Conservation 307 309 YES 269 (85.7%) 189 (60.2%) 203 (66.1%) 141 (45.6%) NO 22 (7%) 65 (20.7%) 43 (14%) 45 (14.6%) UNCERTAIN 23 (7.3%) 60 (19.1%) 60 (19.5%) 123 (39.8%) Note: * Number of valid responses varies because some participants did not respond to prompts regarding every entity. The aggregate pattern of awareness (i.e., citizens are more aware of being served by general than special governments) does not apply for each individual entity. Awareness of being served ranges from 85.7% for counties to 45.6% for soil and water conservation districts. However, 60.2% of respondents were aware of being served by townships while 66.1% were aware of being served by solid waste management districts. Interestingly, the highest No response was for townships, with 20.7% incorrectly saying they were not served by townships. The Uncertain responses for townships and solid waste management districts were comparable. Soil and water conservation districts received the highest “Uncertain” response (39.8%). Familiarity with Government Participants who responded “Yes” to being served by a specific entity were prompted to answer questions regarding their familiarity with that entity. Table 3 shows mean familiarity scores aggregated by 412 KILLIAN & LE type of government (general versus special), and Table 4 shows mean familiarity scores for individual entities. Since there is no objective standard for the appropriate level of familiarity, a single familiarity score lacks significance in isolation, but comparative scores allow us to assess relative familiarity with general and special governments. Tables 3 and 5, which show aggregate results, contain data from citizens who recognized being served by at least one general and at least one special entity. This ensures a robust comparative analysis of familiarity by type of government. As shown in Table 3, participants who recognize being served by at least one form of general and at least one form of special government are more familiar with general governments, and the differences are statistically significant. On a 10-point scale, the mean familiarity with the goals and performance of general governments is 4.84 while the mean for special district governments is 3.98. The mean familiarity with how governments obtain and use financial resources is 4.76 for general governments and 3.35 for special district governments. TABLE 3 Familiarity Ratings by Type of Government (General versus Special) Familiarity with Goals and Performance of Significance Test (Entity Type) N Mean Std. Deviation General 233 4.84 2.28 t(232) = 5.98, p < .05 Special 233 3.98 2.29 Familiarity with How (Entity Type) Obtains Significance Test and Uses Financial Resources N Mean Std. Deviation General 232 4.76 2.36 t(232) = 10.43, p < .05 Special 232 3.35 2.11 Note: To be included in Table 3 analysis, participants must have responded “Yes” to at least one general entity and “Yes” to at least one special district entity. Therefore, the N is smaller than the total number of participants. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 413 When individual entities are considered, in Table 4, the direction of results is the same: for both goals and performance and financial resources, citizens report more familiarity with general than with special entities. Participants are most familiar with counties, which received the highest, mean familiarity scores for both goals and performance (5.30) and financial resources (5.25). The familiarity scores for townships, on both goals and performance (4.36) and financial resources (4.3) are lower than for counties. Still, for both goals and performance and financial resources, participants are more familiar with townships than with the special districts. These results have face validity: counties score highest on awareness of being served (Table 2) as well as familiarity (Table 4), while soil and water conservation districts score lowest on awareness of being served and on both familiarity measures. TABLE 4 Familiarity Ratings for Individual Entities Familiarity with Goals and Performance of (Entity) N Mean Median Std. Deviation County 268 5.30 5 2.39 Township 188 4.36 5 2.63 Solid Waste Mgt. 202 4.31 5 2.63 Soil & Water 141 3.62 3 2.40 Familiarity with How (Entity) Obtains and Uses Financial Resources N Mean Median Std. Deviation County 268 5.25 5 2.53 Township 189 4.3 4 2.59 Solid Waste Mgt. 202 3.55 3 2.55 Soil & Water 141 3.18 3 2.28 Tables 5 and 6 show results for the ideal level of familiarity “the average person” should have with various governments. Per Table 5, the mean ideal familiarity with goals and performance of general governments is 7.50 while the mean for special district governments is 6.85. The mean ideal familiarity with the financial resources of general governments is 7.64 while the mean ideal for special district governments is 6.87. These differences are statistically significant, 414 KILLIAN & LE yet they converge more than the aggregate scores for actual familiarity. TABLE 5 Ideal Familiarity Ratings by Type of Government (General versus Special) Ideal familiarity with Goals and Performance of (Entity Type) N Mean Std. Deviation General 233 7.50 1.92 Special 233 6.85 2.06 Ideal familiarity with How (Entity Type) Obtains and Uses Financial Resources General 232 7.64 1.90 Special 232 6.87 Significance Test t(232) = 6.48, p<.05 Significance Test t(231) = 7.65, p<.05 2.14 Note: To be included in Table 5 analysis, participants must have responded “Yes” to at least one general entity and “Yes” to at least one special district entity. Therefore, the N is smaller than the total number of participants. Table 6 shows the ideal familiarity ratings for individual entities. In comparing the data in Tables 4 and 6, two important points should be noted. First, whether the topic is goals and performance or financial resources, and whether referring to general or special entities, citizens believe the ideal level of familiarity is higher than their actual level of familiarity. Second, the mean scores in Table 4 vary more than the mean scores in Table 6: there is more variation in actual than in ideal familiarity. Participants tend to converge in their perceptions of ideal levels of familiarity, but their actual scores are more varied due to their lower familiarity with special districts. Table 7 shows the results of gap analysis. The familiarity gap for each entity is computed by subtracting the actual familiarity score (Table 4) from the ideal familiarity score (Table 6). The familiarity gaps are statistically significant. The smallest gap (2.36 on a 10point scale) pertains to goals and performance for counties. The CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 415 TABLE 6 Ideal Familiarity Ratings for Individual Entities Ideal Familiarity with Goals and Performance of (Entity) N Mean Median Std. Deviation County 269 7.66 8 1.90 Township 188 7.53 8 2.04 Solid Waste Mgt. 202 7.05 7 2.12 Soil & Water 141 6.99 7 2.22 Ideal Familiarity with How (Entity) Obtains and Uses Financial Resources County 269 7.88 8 1.93 Township 188 7.65 8 2.11 Solid Waste Mgt. 202 6.90 7 2.30 Soil & Water 141 7.14 8 2.45 TABLE 7 Gap Analysis: Mean Familiarity minus Mean Ideal Familiarity Gap Between Actual and Ideal Familiarity* (on 10-point scale) Goals and Significance Test Financial Significance Test Resources Performance County 2.36 Township 3.18 Solid Waste Management Soil & Water 2.74 3.38 t(267) = 15.77, p < .05 t(186) = 15.06, p < .05 t(200) = 14.61, p < .05 t(140) = 15.32, p < .05 2.63 3.36 3.35 3.97 t(267) = 16.33, p < .05 t(187) = 15.46, p < .05 t(201) = 17.22, p < .05 t(140) = 16.22, p < .05 Note: *Perceived information gaps are computed using data from Tables 5 and 7. largest gap (3.97) relates to how soil and water conservation districts obtain and use financial resources. For goals and performance, participants report a larger familiarity gap for townships than for solid waste management districts (3.18 and 2.74, respectively), though these entities have similar familiarity gaps on financial resources (3.36 and 3.35, respectively). 416 KILLIAN & LE Table 7 reveals several important points. For every entity, the familiarity gap related to financial resources is greater than the gap for goals and performance. Further, the familiarity gaps for townships resemble special districts as much as they resemble counties. Finally, while participants place the “ideal” bar lower for special districts than for general governments, they perceive significant familiarity gaps for all entities in this analysis. DISCUSSION For citizens to hold their governments accountable, they must be aware of and familiar with the governments that serve them. Awareness and familiarity do not guarantee accountability, but where they are absent, accountability cannot ensue. We compared citizen familiarity with general purpose and special district governments in Indiana. The participants in our sample correspond to Indiana voters in both age and gender distribution; have incomes that correspond to all U.S. voters but are above the median for Indiana voters; and have educational attainment that is above the median for Indiana and U.S. voters. Since participation in political processes, whether voting or more active behavior, is positively associated with age, education, and income (Lijphart, 1997, U.S. Census Bureau, 2004 and 2008), the demographic profile of our participants strengthens the results. On measures of familiarity with the governments that serve them, it is reasonable to expect our participants to be as familiar as the general population, if not more so. We gathered responses using a pre-tested survey instrument. Table 1 indicates that the level of confusion among citizens is high and the awareness of being served by various local governments is low. Only 22% of participants recognize being served by all four entities, and less than half (48%) recognize being served by at least three of the four entities. Confusion is associated with both general and special governments, though Table 2 indicates that in the aggregate, citizens are more aware of being served by generalpurpose than by special district governments. The aggregate results do not apply to all individual entities; participants are more likely to recognize being served by solid waste management districts than by townships. The overall level of confusion is disconcerting, given that the boards or councils of these entities are comprised of elected officials, that three of the four entities can levy property taxes, and CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 417 that the fourth (soil and water conservation districts) can levy special assessments. In the aggregate (Table 3) participants who are aware of being served by at least one general-purpose and at least one special district government are more familiar with general-purpose governments. This finding applies for both goals and performance and financial resources, and the differences are statistically significant. When individual governments are considered (Table 4) the pattern holds: participants are more familiar with the goals and performance and financial resources of general-purpose governments. Tables 3 and 4 also support Sanders’ (1994) thesis on “fiscal illusion,” discussed earlier. In the aggregate and for individual entities, participants are less familiar with government finances than with government goals and performance. In terms of the ideal level of familiarity (Tables 5 and 6) participants set the bar higher for general-purpose entities. Across the board, however, participants want more information about the governments that serve them. There is more convergence on ideal than actual levels of familiarity. Ideally, participants want their familiarity with special districts to more closely resemble their familiarity with general-purpose governments. Table 7 shows the comparative gaps between actual and ideal familiarity. An important finding for government accountants and financial managers is that for each entity in this study, the familiarity gap on how the entity obtains and uses financial resources exceeds the gap on goals and performance. It is important to note that this study intentionally includes two of the most prominent types of special districts in the state.5 Solid waste management districts and soil and water conservation districts serve every citizen in Indiana and have boundaries that correspond with county boundaries. Thus, we expect them to be among the most familiar of the 25 types of special districts in the state (U.S. Census Bureau, 2002b). The recognition and familiarity levels for the special districts in this study do not bode well for the more obscure special districts. This research has near-term, practical significance. Every year from 2004 through 2011, bills were introduced into the Indiana legislature to reform or eliminate townships. In turns, the draft 418 KILLIAN & LE legislation called for consolidation of smaller townships into larger entities, or elimination of townships and transfer of their responsibilities to counties (Indiana Township Association (ITA, 2011). The drive to consolidate or eliminate townships is motivated by a desire to make government more visible and accountable and to reduce operating costs (ICLGR, 2007; Faulk and Hicks, 2009). Township defenders have managed to block reform efforts for eight consecutive years, and the ITA and other associations maintain a strong defense of township government. Yet only 60% of the participants in this study, whose demographic profile places them among the more politically active citizens, recognize being served by townships. This suggests that townships may be unable to mobilize sufficient public support to withstand reform efforts indefinitely. CONCLUSIONS The results of this study support the institutional reform perspective more than the public choice perspective. Public choice advocates assert that citizens can monitor smaller, specialized governments more effectively than larger, generalized governments, but our results contradict this claim. Participants were more familiar with the goals and performance and financial resources of generalpurpose than special district governments. Since this study included two of the most prominent special districts in the state, we expect the results would be even more pronounced for less prominent special districts. Further, institutional reformers assert that the proliferation of special districts leads to confusion among citizens as to which governments serve them, and we found this to be true in the aggregate: participants were more likely to recognize being served by the general-purpose than the special district governments. This was not always true at the individual level: participants were more likely to recognize being served by solid waste management districts than by townships. Citizens, public officials, and scholars who want to improve government accountability should pay more attention to special districts. At the very least, they should exert pressure on specialdistricts to publish financial reports that comply with GASB standards. As noted earlier, special districts have the lowest compliance rate for all local governments (GASB 2008). Of course, compliance needs to be improved among all local governments. Efforts to increase CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 419 performance reporting and citizen-centric reporting, such as the programs sponsored by the Association of Government Accountants (AGA), should be expanded to include special districts. Since external reports are of limited use without norms to support trend analysis and comparisons with other entities (Guo, Fink and Frank, 2009), scholars could make a significant contribution by developing norms, or at least regional benchmarks, for special districts as well as general governments. We note, however, that since accountability requires a dynamic exchange between citizens and their governments (discussed in the Introduction), good reporting is necessary but not sufficient: it must be accompanied by dialog and participation. We could add that civics education in public schools and other voter education programs (such as programs sponsored by the League of Women Voters) need to be more effective in promoting awareness of and familiarity with various governments. Ultimately, however, the problem may be structural. This study shows that it is challenging for citizens to monitor (even be aware of) multiple local governments, and that they are less familiar with special districts than general-purpose governments. As citizens and public officials consider structural changes to improve accountability and costefficiency, and as they weigh proposals to consolidate local governments, they should examine all local governments. Thus far, Indiana has devoted most of its time and energy to townships. It is time to examine the role and impact of special districts. Limitations The participants in our sample resemble “all Indiana voters” in age and gender and exceed “all Indiana voters” in income and educational attainment. As discussed earlier, we believe this strengthens our conclusions. To the extent that our participants are unaware of or unfamiliar with local governments, we expect the median Indiana voter to be even more so. The legal, political, and economic contexts for local government and the powers granted to general and special governments vary from state to state (McCabe, 2000; Foster, 1997). This limits the potential for cross-state conclusions and creates the need for local and regional research. Our research has direct implications for Indiana, and can inform and motivate research in other states. While the results cannot be generalized to other states, our research 420 KILLIAN & LE instrument can be adapted to measure awareness of and familiarity with local governments elsewhere. NOTES 1. The Indiana Commission on Local Government Reform was established in 2007, as was the New York State Commission on Local Government Efficiency and Competitiveness. Ohio established its Commission on Local Government Reform and Collaboration in 2008. 2. Throughout its existence, the U.S. Advisory Commission on Intergovernmental Relations (ACIR) adopted differing perspectives on special districts. In its 1987 report, ACIR endorsed policies consistent with the public choice perspective. 3. No objective norms have been established for how much citizens “should know” about government performance and finances, though secondary education standards might be a start. Selfreported familiarity is adequate for our research goals. Citizen perceptions would be inadequate for comparing actual costefficiencies of various governments. 4. Since 1970, the City of Indianapolis and Marion County have been merged under Uni-Gov. Marion County is included in the total count of Indiana counties (92), but due to Uni-Gov, Marion County does not share the same characteristics as all other Indiana counties. Therefore, residents of Marion County are excluded from this study. 5. We asked the Associate Director of the Indiana Advisory Commission on Intergovernmental Relations to help us select the most common special districts to minimize inherent bias. She suggested solid waste management districts, soil and water conservation districts, and library districts (Palmer 2009). Library districts were excluded from this analysis because they do not serve every citizen, as do the other entities. ACKNOWLEDGEMENTS We are indebted to Howard Frank, Patricia Patrick, and to anonymous reviewers for beneficial suggestions and for some references in this paper. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 421 REFERENCES Advisory Commission on Intergovernmental Relations (ACIR) (1987). The Organization of Local Public Economies. Washington, DC: Author. Association of Government Accountants (AGA) (2008). Government Accountability and Transparency 2008: A Survey Commissioned by the Association of Government Accountants. [Online]. Available at http://www.agacgfm.org/harrispoll2008.aspx. (Retrieved on April 20, 2008). Axelrod, D. (1992). Shadow Government: The Hidden World of Public Authorities. New York: John Wiley & Sons. Barlow, I. (1991). Metropolitan Government. London: Routledge. Bish, R., & Ostram, V. (1973). Understanding Urban Government. Washington, DC: American Enterprise Institute for Public Policy Research. Bollens, J. (1957). Special District Governments in the United States. Berkeley: University of California Press. Chan, J., & Rubin, M. (1987). “The Role of Information in a Democracy and in Government Operations: The Public Choice Methodology.” Research in Governmental and Nonprofit Accounting, 3: 3-28. Dilorenzo, T. (1981). “The Expenditure Effects of Restricting Competition in Local Public Service Industries: The case of Special Districts.” Public Choice, 37: 569-578. Faulk, D., & Hicks, M. (2009). Local Government Reform in Indiana. Center for Business and Economic research, Miller College of Business, Ball State University. Foster, K. (1997). The Political Economy of Special-Purpose Governments. Washington, DC: Georgetown University Press. (Republished in 2007). Governmental Accounting Standards Board (GASB) (1987). Summary of Concept Statement No. 1: Objectives of Financial Reporting. [Online]. Available at http://www.gasb.org/ pub/index.html. (Retrieved on June 10, 2010). 422 KILLIAN & LE Governmental Accounting Standards Board (GASB). (1994). Summary of Concepts Statement No. 2: Service Efforts and Accomplishments Reporting. [Online]. Available at http://www.gasb.org/pub/index.html. (Retrieved on June 10, 2010). Governmental Accounting Standards Board (GASB). (2008). State and Local Government Use of Generally Accepted Accounting Principles for General Purpose External Financial Reporting. [Online]. Available at http://www.gasb.org/cs/ContentServer?c= Document_C&pagename=GASB%2FDocument_C%2FGASBDocu mentPage&cid=1176156726669. (Retrieved on July 11, 2010). Guo, D., Fink, D., & Frank, H. (2009). “Disclosure Quality of Management Discussion and Analysis (MD&A): Evidence from Large Florida Cities.” Municipal Finance Journal, 30 (3): 53-72. Harris, J., McKenzie, K, & Rentfro R. (2011). “Performance Reporting: Assessing Citizen Access to Performance Measures on State Government Websites.” Journal of Public Budgeting, Accounting, and Financial Management, 23 (1): 117-138. Indiana Commission on Local Government Reform (ICLGR). (2007). Streamlining Local Government: We’ve Got to Stop Governing like This. [Online]. Available at http://indianalocalgovreform.iu. edu/assets/docs/Report_12-10-07.pdf. (Retrieved on January 22, 2010). Indiana Township Association (ITA) (2011). Indiana Bill Watch. [Online]. Available at http://indianatownshipassoc.org/ component/option,com_frontpage/Itemid,88/. (Retrieved on June 24, 2011). Ingram, R., & Copeland, R. (1981). “Municipal Accounting Information and Voting Behavior.” The Accounting Review, 56(4): 830-843. Killian, L. (2011). “The Continuing Problem of Special Districts in American Government.” Accounting and the Public Interest, 11 (1): 52-67. Leigland, J. (1990). “The Census Bureau’s Role in Research on Special Districts: A Critique.” The Western Political Quarterly, 43 (2): 367-380. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 423 Lijphart, A. (1997). “Unequal participation: Democracy’s Unresolved Dilemma.” The American Political Science Review, 91 (1): 1-14. Lowery, D. (2001). “Metropolitan Governance Structures from a Neoprogressive Perspective.” Swiss Political Science Review, 7: 130–136. McCabe, B. (2000). “Special-District Formation among the States.” State and Local Government Review, 32 (2): 121-131. Mehay, S. (1984). “The Effect of Government Structure on Special District Expenditures.” Public Choice, 44: 339-348. New York State Commission on Local Government Efficiency and Competitiveness (NYSCLGEC). (2008.) 21st Century Local Government: Report of the New York State Commission on Local Government Efficiency and Competitiveness. [Online]. Available at http://www.nyslocalgov.org/pdf/LGEC_Final_Report.pdf? pagemode=bookmarks. (Retrieved on January 22, 2010). Palmer, J. (2009, May). Personal communication with the authors. Indianapolis. Peddie, S. (2007, December 16). “Small Special Districts offer Big Pay, Perks.” Newsday. Sanders, G. (1994). “Tiebout Migrants and the Median Voter as Users of Accounting Information.” Research in Governmental and Nonprofit Accounting, 8: 137-155. Sen, A. (1999). “Democracy as a Universal Value.” Journal of Democracy, 10 (3): 3-17. Smith, R. (1969). Public Authorities in Urban Areas. Washington, DC: National Association of Counties Research Foundation. Spitzer, E. (2007). Executive Order No 11: Establishing the New York State Commission on Local Government Efficiency and Competitiveness. [Online]. Available at www.nyslocalgov.org/. (Retrieved on December 1, 2009). Tiebout, C. (1956). “A Pure Theory of Local Expenditures.” The Journal of Political Economy, 64 (5): 416-24. U.S. Census Bureau (1992). Census of Governments, Volume 1, Number 1: Government Organization. Table 5: Special Purpose Local Governments by State: Census years 1952-1992. [Online]. 424 KILLIAN & LE Available at http://www.census.gov/prod/2/gov/gc/gc92_1_1. pdf. (Retrieved on June 10, 2010). U.S. Census Bureau (2002a). Census of Governments, Volume 1, Number 1: Government Organization. [Online]. Available at http://www.census.gov/govs/www/cog2002.html. (Retrieved on June 10, 2010). U.S. Census Bureau (2002b). Census of Governments, Volume 1, Number 2: Individual State Descriptions. [Online]. Available at http://www.census.gov/prod/2005pubs/gc021x2.pdf. (Retrieved on June 20, 2010). U.S. Census Bureau (2004). Voting and Registration in the Election of November 2002. [Online]. Available at www.census.gov/ prod/2004pubs/p20-552.pdf. (Retrieved on June 21, 2010). U.S. Census Bureau (2007a). Local Governments and Public School Systems by Type and State:2007. [Online]. Available at http://www.census.gov/govs/cog/GovOrgTab03ss.html. (Retrieved on January 22, 2010). U.S. Census Bureau (2007b). 2007 Governments Integrated Directory. (Interactive data). [Online]. Available at http://harvester.census.gov/gid/gid_07/options.html. (Retrieved on June 21, 2010). U.S. Census Bureau (2008). Voting and Registration in the Election of November 2008 - Detailed Tables. [Online]. Available at http://www.census.gov/hhes/www/socdemo/voting/publications /p20/2008/tables.html. (Retrieved on June 24, 2011.) U.S. Census Bureau (2009a). American Community Survey 20052009. Table B19001: Household Income in the Past 12 Months. [Online]. Available at http://factfinder.census.gov/home/saff/ main.html?_ lang=en. (Retrieved on June 21, 2011). U.S. Census Bureau (2009b). Table B15002: Sex by Educational Attainment for the Population 25 Years and Over. [Online]. Available at http://factfinder.census.gov/home/saff/main.html? _ lang=en. (Retrieved on June 21, 2011). Wagner, R. & Weber, W. (1975). “Competition, Monopoly, and the Organization of Government in Metropolitan Areas.” The Journal of Law and Economics, 18: 661-84. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 425 Wells, D., & Scheff, R. (1992). “Performance Issues for Public Authorities in Georgia.” In J. Mitchell (Ed.), Public Authorities and Public Policy (pp. 167-176). New York: Praeger. Williams, P. (2002). “Accounting and the Moral Order: Justice, Accounting, and Legitimate Moral Authority.” Accounting and the Public Interest, 2 (1): 1-21. Wood, R. (1961). 1400 Governments: The Political Economy of the New York Metropolitan Region. Cambridge, MA: Harvard University Press. APPENDIX A Research Instrument Are you served by a select your answer) a. b. c. (Entity) government? (Circle the appropriate letter to Yes (please answer the questions below) No (skip to next page) Uncertain (skip to next page) 1. How familiar are you with the goals and performance of your (entity)? Not at all Very Somewhat Familiar familiar familiar 1 2 3 4 5 6 7 8 9 10 2. Ideally, how familiar should the average person be with the goals and performance of your (entity)? Not at all Very Somewhat Familiar familiar familiar 1 2 3 4 5 6 7 8 9 10 3. How familiar are you with how your (entity) obtains and uses financial resources? Not at all familiar 1 2 Somewhat Familiar 3 4 5 6 7 8 9 4. Ideally, how familiar should the average person be with how obtains and uses financial resources? Not at all Somewhat Familiar familiar 1 2 3 4 5 6 7 8 9 Very familiar 10 your (entity) Very familiar 10 426 KILLIAN & LE Demographic Data This information will tell us about the citizens who complete the survey. If you are not comfortable with providing any of the information, you may leave it blank. Are you male or female? Male Female Please enter your Zip Code __________ What is your age in years? __________ What is your household income, per year? (Check appropriate box) Less than $15,000 $15,000 to $24,999 $25,000 to $34,999 $35,000 to $49,999 $50,000 to $74,999 $75,000 to $99,999 $100,000 or over Prefer not to answer Place a check next to your education level. Less than high school Some high school High school diploma or equivalent (e.g., GED) Some college, but no college degree Associate's degree Bachelor’s degree Some graduate school, but no graduate degree Graduate degree (e.g., M.S., M.D., and Ph.D.) Prefer not to answer What is your citizenship status? Note: Identical questions were repeated for each type of government. Demographic questions appeared at the end of the survey. For the online survey, “skip logic” automatically progressed to the next entity if the participant selected No or Uncertain to being served by the given entity. CITIZEN PERCEPTIONS OF GENERAL-PURPOSE AND SPECIAL DISTRICT GOVERNMENTS 427 APPENDIX B* Demographic Data for Survey Participants and Registered and Reported Voters Percent (%) Educational Attainment High School or Some College Bachelor’s Advanced Equivalent or Associates Degree Degree 15.8 35.3 25.8 23.2 6.2 16.5 42.8 34.5 11.6 27.2 33.1 28.1 Hard-Copy Sample Internet Sample Total Sample U. S. Citizen Permanent resident, but not citizen Other Indiana Registered Indiana Voters All U.S. Registered 28.6 All U.S. Voters 27.3 31.4 31.6 Median Household Income Hard-Copy Sample Internet Sample Total Sample Indiana Registered Indiana Voters All U.S. Registered All U.S. Voters $50,000-74,999 $75,000 - 99,999. $75000-99,000 21.1 22.4 10.9 11.8 Gender (%) M F 42 58 51 49 46 54 47.2 52.8 46.8 53.2 Median Age 55 52 52.3 45-64 45-64 $50,000-74,999 $75,000-99,999 *Descriptions for Appendix B Data Elements. Indiana Voters and All U.S. Voters are those who reported voting in the 2008 general election in response to questions in the U.S. Census Bureau’s November 2008 Current Population Survey (CPS). Data for registered and reported voters in Indiana, and for all U.S. registered and reported voters, were extrapolated from various tables in the U.S. Census Bureau (2008), Voting and Registration in the Election of November 2008 - Detailed Tables. The specific tables are identified below. Where Indiana-specific data are not available, we report all U.S. data. Educational Attainment. Education attainment of our sample (shown for hard-copy, internet, and total) can be compared to educational attainment of all U.S. registered and reported voters. The “All U.S.” percentages are calculated from Table 5: Reported Voting and Registration, by Age, Sex, and Educational Attainment. We do not have access to data for education attainment for registered and reported voters in Indiana. 428 KILLIAN & LE Household Income. Median household income of our sample (shown for hard-copy, internet, and total) can be compared to median household income for all U.S. registered and reported voters. The “All U.S.” income is from Table 9. Reported Voting and Registration of Family Members, by Race, Hispanic Origin and Family Income: November 2008. Gender. Gender of our sample (shown for hard-copy, internet, and total) can be compared to gender of registered and reported voters in Indiana. The Indiana data is from Table 4b: Reported Voting and Registration of the Voting-Age Population, by Sex, Race and Hispanic Origin, for States. Age. Participants in our sample (shown for hard-copy, internet, and total) entered their age in years into a blank box on the survey. The median age for our sample can be compared to median age for registered and reported voters in Indiana, which is provided as interval data in Table 4c: Reported Voting and Registration of the Total Voting-Age Population, by Age, for States. J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT, 24 (3), 429-465 FALL 2012 TWO ACCOUNTING STANDARD SETTERS: DIVERGENCE CONTINUES FOR NONPROFIT ORGANIZATIONS Mary Fischer and Treba Marsh* ABSTRACT: The ability of financial statement users, investors, donors and academic researchers to compare financial information issued by nonprofit universities, hospitals, fund-raising organizations and government agencies is affected by their understanding of current accounting recognition and reporting guidance. Public nonprofit organizations report different financial results from private nonprofit organizations. This study looks at the events that brought about the divergence in nonprofit financial accounting recognition and reporting for higher education institutions, discusses specific differences, and offers a look at additional changes in recognition and reporting for the sector currently underway. INTRODUCTION Prior to the establishment of the Governmental Accounting Standards Board (GASB), financial recognition and reporting of both public and private not-for-profit organizations, i.e., colleges and universities, hospitals, utilities and fundraising organizations remained, similar in most respects (Fischer 1997). Since the mid1990s, much has been written about the divergence and growing recognition and reporting differences created by having two standard setters for the same industry (Van Daniker, 2009; Menditto, & -----------------------* Mary Fischer, Ph.D., CGFM, is Professor of Accounting, College of Business and Technology, The University of Texas at Tyler. Her research interests include financial accounting, nonprofit and governmental organizations, auditing and accounting education. Treba Marsh, DBA, CPA, is Professor of Accounting and Director of Gerald W. Schlief School of Accountancy, Stephen F. Austin State University. Her research interest includes state and local public finance, financial accounting, internal controls and public budgeting. Copyright © 2012 by PrAcademics Press 430 FISCHER & MARSH Gordon, 2008; Brown, 2005; Goldstein & Menditto, 2005; Robert, & Mohaghegh, 2004; Fischer et al., 2004; Nelson et al., 2003; Crawford, 2003; Fischer, & Gordon, 2002; Engstrom, & Esmond-Kiger 1997; Smith et al., 1997; Fischer, 1997). Accounting standards and guidance developed by the GASB and the Financial Accounting Standards Board (FASB) create differences for the public and private nonprofit sector accounting recognition and reporting. This paper reviews some of the history of the recognition and reporting divergence for the higher education nonprofit sector, identifies and discusses specific differences, and presents comments on current research agendas at the Boards that could lead to new divergence. BACKGROUND The American Institute of Certified Public Accountants’ (AICPA) College and University Audit Guide (1973) gave post-secondary institutions the fund-based reporting model. The model began changing in the 1980s when the FASB began its not-for-profit agenda that identified several projects. Those projects resulted in the issuance of Statement of Financial Accounting Standard (SFAS) No. 93 that addresses depreciation (FASB, 1987), SFAS No. 116 that provides contribution guidance (FASB, 1993a), SFAS No. 117 that establishes the reporting model (FASB, 1993b), SFAS No. 124 that directs investment recognition (FASB, 1995), SFAS No. 136 that provides guidance to recognize funds held by others (FASB, 1999), and SFAS No. 164 that provides guidance for not-for-profit combination and merger reporting (FASB, 2009a). The GASB was created in 1984 and soon after began issuing accounting standards affecting public colleges and universities. GASB Statement (GASBS) No. 8 (1988) addresses the recording and reporting of depreciation and GASBS No. 9 (1989) concerns the statement of cash flows. Both standards were reactions to standards issued by the FASB. GASBS No. 35 (1999b) was a landmark event in terms of higher education reporting and occurred as a consequence of the GASB abandoning the effort to develop a separate reporting model for higher education and electing to include public colleges and universities in the GASBS No. 34 model. GASBS No. 34 (1999a) identifies three options for reporting: business-type activities (BTA) that charge a fee for their services, governmental, and governmental TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 431 with a BTA. Most colleges and universities report as a BTA because (1) it is straightforward and (2) it matches well—though not perfectly— the reporting display followed by private institutions. The differences discussed in this paper are based on the GASB BTA reporting model. Some differences identified are not necessarily specific FASB action differences but result from actions by the AICPA addressing FASB guidance. The AICPA proposes guidance; the FASB and/or GASB review(s) the guidance and while they may not endorse the proposition, unless they object, the guidance becomes GAAP. In that situation, it is as if the FASB or GASB has issued the guidance. The AICPA reacted to the issuance of SFAS Nos. 116 and 117 (FASB, 1993a; 1993b) that fundamentally changed financial reporting for private higher education institutions, by issuing a new audit guide in 1996. The audit guide marked the official departure from the fundbased model and the introduction of an entity-based reporting model for not-for-profits. The capitalization of software is an example of AICPA (1997) action and non-action by the accounting boards that resulted in the establishment of GAAP guidance. Although different levels of GAAP exist, all guidance must be followed when preparing financial statements in order to obtain an unqualified audit opinion. With the adoption of the Sarbanes Oxley Act (2002), the role of the AICPA in accounting guidance development has been significantly reduced. The act created a five-member Public Company Accounting Oversight Board (PCAOB) that has the authority to set and enforce auditing, attestation, quality control and ethics guidance for public companies. The act also directed the FASB to assume the responsibility for issuing accounting guidance and the AICPA to no longer issue authoritative statements of position. PHILOSOPHY OF THE BOARDS Examining the philosophical approaches the Boards have taken may help to understand the various differences in standards issued by the two Boards. These philosophical approaches can be found in each Board’s respective concept statements. Concept statements are not accounting standards to which preparers must adhere when issuing financial statements. Instead, taken as a whole, concept statements form the framework used by the Boards when developing accounting standards. 432 FISCHER & MARSH The overall focus of the FASB is decision utility, i.e., attempting to provide the best information to influence decisions by investors, creditors, and others interested in commercial and not-for-profit activity. The FASB conceptual framework includes: - Objectives of Financial Reporting by Business Enterprises, - Qualitative Characteristics of Accounting Information, - Objectives of Financial Reporting by Nonbusiness Organizations, - Recognition and Measurement in Financial Statements of Business Enterprises, - Elements of Financial Statements, and - Using Cash Flow Information and Present Value in Accounting Measurement. The GASB’s overall focus is accountability as opposed to decision utility. This results from the heavy reliance on taxpayer support of governmental entities. The GASB conceptual framework includes: - Objectives of Financial Reporting, - Service Efforts and Accomplishments Reporting, - Communication Methods, and - Elements of Financial Statements. Future projects listed on the GASB agenda (www.gasb.org) continue to address recognition and measurement. The line between decision utility and accountability is blurred and the Boards’ propensity toward one does not make it exclusive of the other. Rather, it is a continuum on which the Boards operate. For example, the FASB in issuing FASB Statement No. 136 (1999), Transfers of Assets to a Not-for-Profit Organization or Charitable Trust That Raises or Holds Contributions for Others, was focused just as much on accountability as on decision utility. Similarly, the GASB’s efforts on other postemployment benefits, which resulted in GASB Statement No. 45 (2004), Accounting and Financial Reporting by Employers for Postemployment Benefits Other Than Pensions, clearly is focused as much on decision-making as it is accountability. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 433 AREAS OF POTENTIAL DIFFERENCES There are four major types of differences: recognition, measurement, display, and disclosure. Recognition differences deal with whether or not an item appears in a financial statement. An example is contributed services. SFAS No. 116 (1993a) addresses contributed services while GASBS No. 33 (1998) does not. Measurement differences refer to how items are included in the financial statements—at what amount and based on which criteria. For example, both the FASB and the GASB require recognition of pension liabilities but they have established different methods for measuring the amount to be recognized. An example of a display difference is the GASB requirement for a classified balance sheet that separates assets and liabilities into current and noncurrent components. The FASB requires listing assets and liabilities in order of liquidity, but does not require separating them into current and noncurrent categories. The Boards have significantly different requirements with respect to required note disclosures that are not addressed in this paper. EXHIBIT Ia Sample State University Statement Of Net Assets June 30, 2010 ASSETS Current Assets Cash and equivalents $ 76,975 Short term investment 78,856 Accounts receivable 33,891 Notes receivable Prepaid expenses Total current assets Noncurrent Assets Investments - bond trustee Endowment investments Pledge receivable 2,589 4,515 196,826 47,032 158,621 2,152 434 FISCHER & MARSH EXHIBIT Ia (Continued) Notes receivable Property and equipment Other assets Total noncurrent assets TOTAL ASSETS LIABILITIES Current Liabilities Accounts payable and accrued liabilities Deposits and deferred revenue Accrued benefits - current portion Long-term debt - current portion Total current liabilities Noncurrent Liabilities Life income agreements Government loan advances Accrued benefits Long-term debt Total noncurrent liabilities TOTAL LIABILITIES NET ASSETS Invested in capital assets net of related debt Restricted nonexpendable Endowment – campus Endowment - affiliated organizations Restricted expendable Endowment gains Funds functioning as endowments Life income and annuity contracts Gift, grants and contracts Loan funds Affiliated organizations Unrestricted net assets TOTAL NET ASSETS $ $ 18,746 439,120 2,282 667,953 864,779 43,469 24,514 7,397 1,938 77,318 $ 1,526 17,214 67,589 205,376 291,705 369,023 $ 279,974 38,839 66,230 $ 26,217 6,495 787 18,319 5,296 6,745 46,854 495,756 TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 435 EXHIBIT Ib Sample University Statement Of Financial Position June 30, 2010 (in Thousands) ASSETS Cash and equivalents Short term investments Accounts receivable Note receivable Pledge receivable Investments Investments held by others Prepaid expenses Property and equipment Other assets TOTAL ASSETS LIABILITIES Accounts payable Deposits and deferred revenue Compensated absences OPEB and other benefits payable Split-interest contracts Notes and bonds payable U.S. Government advance TOTAL LIABILITIES NET ASSETS Unrestricted Temporarily restricted Permanently restricted TOTAL NET ASSETS $ $ $ $ 82,158 120,705 33,891 21,335 15,810 87,789 70,832 4,515 439,120 2,282 878,437 43,469 24,514 13,534 61,452 1,526 207,314 17,214 369,023 $ 326,828 67,517 115,069 $ 509,414 436 FISCHER & MARSH EXHIBIT IIa Sample State University Statement of Revenue, Expenses and Changes To Net Assets For Fiscal Year Ended June 30, 2010 (in Thousands) Operating Revenue Tuition and Fees Discount Allowances Net Tuition and Fees Government Grants and Contracts Private Grants and Contracts Sale of Auxiliary Services Other Operating Revenue Total Operating Revenue Operating Expenses Salaries and Wages Fringe Benefits Supplies and Services Utilities Grant Subcontracts Depreciation Total Operating Expenses Operating Loss Non-Operating Revenue (Expense) State Appropriations Non-Capital Gifts Investment Income Interest Expense Total Non-Operating Revenue Net Income before Capital Items Capital and Other Items State Appropriations for Capital Capital Asset Gifts Endowment Gifts Endowment Gains, (losses) and Additions Total Capital Deductions Net Increase/Decrease in Net Assets Net Assets - Beginning of Year Net Assets - End of Year $ $ 193,491 -51,773 141,718 86,138 14,832 103,442 20,049 366,179 235,141 76,826 101,651 12,498 13,510 26,895 466,521 -100,342 79,134 11,989 16,494 -5,840 101,777 1,435 5,687 1,206 10,314 $ -26,922 -9,715 -8,280 504,036 495,756 TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 437 EXHIBIT IIb Sample University Statement of Activity For Fiscal Year Ended June 30, 2010 (in Thousands) Revenue Tuition and Fees Discount Allowances Net Tuition and Fees Government Grants and Contracts Private Grants and Contracts Private Gifts Other Investment Income/Losses Sale of Auxiliary Services Other Operating Revenue Total Revenue Reclassifications Net Revenue Expenses Salaries and Wages Fringe Benefits Supplies and Services Utilities Grant Subcontracts Depreciation Total Operating Expenses Change in Net Assets Net Assets - Beginning of Year Net Assets - End of Year Temporarily Permanently Unrestricted Restricted Restricted Net Assets Net Assets Net Assets Combined Total $278,312 -31,773 246,539 $278,312 -31,773 246,539 66,138 66,138 11,989 $1,206 3,658 $20,314 13,195 23,972 11,621 -7,439 222 4,404 20,536 20,536 103,442 20,049 477,739 0 477,739 20,536 94,533 $115,069 235,141 76,826 107,491 12,498 13,510 26,895 472,361 5,378 504,036 $509,414 103,442 20,049 459,778 11,906 471,684 235,141 76,826 107,491 12,498 13,510 26,895 472,361 -677 327,505 $326,828 -2,575 -11,906 -14,481 -14,481 81,998 $67,517 438 FISCHER & MARSH EXHIBIT IIIa Sample State University Statement Of Cash Flows For Fiscal Year Ended June 30, 2010 (in Thousands) Cash Flows from Operating Activities Tuition and fees Grants and contracts Sales of auxiliary services Other operating revenue Payments to employees Payments for employee benefits Payments to suppliers and vendors Net cash flow from operating activities Cash Flows from Noncapital Financing Activities State appropriations Noncapital gifts Net cash flow from noncapital financing activities Cash Flows from Capital Financing Activities State appropriations Capital gifts, grants and contracts Endowment gifts Capital assets acquired Proceeds from sales of capital assets New debt proceeds Debt principal payments Debt interest payments Net cash flow from capital financing activities Cash Flows from Investing Activities Investment sales and maturities Investment acquisitions Investment income Net cash flow from investing activities Decrease in cash and cash equivalents Beginning cash and cash equivalents Ending cash and cash equivalents $ 143,178 95,714 104,614 20,406 -232,781 -72,121 -117,214 -58,204 79,134 9,313 88,447 3,885 1,206 10,314 -74,634 198 43,153 -46,447 -9,169 -71,494 $ 100,771 -97,251 13,250 16,770 -24,481 101,456 76,975 TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 439 EXHIBIT IIIa (Continued) Reconciliation of Operating Loss to Net Cash Used by Operating Activities Operating loss Depreciation Changes in current assets and liabilities Accounts receivable Notes receivable Prepaid expenses and other current assets Accounts payable and accrued liabilities Deposits and deferred revenues Accrued employee benefits Net Cash Used in Operating Activities $ $ -100,342 26,895 -7,715 276 320 12,791 4,866 4,705 -58,204 EXHIBIT IIIb Sample University Statement of Cash Flows For Fiscal Year Ended June 30, 2010 (in Thousands) Cash Flows from Operating Activities Change in net assets Depreciation Changes in current assets and liabilities Accounts receivable Notes receivable Prepaid expenses and other current assets Accounts payable and accrued liabilities Deposits and deferred revenues Accrued employee benefits Net cash flow from operating activities Cash Flow from Investing Activities Investment sales and maturities Investment acquisitions Capital assets acquired Proceeds from sales of capital assets Loans disbursed Principal collected on loans $ 5,378 26,895 -7,715 276 320 12,791 4,866 4,705 47,516 $ 85,588 -87,251 -74,634 198 12,183 -12,188 440 FISCHER & MARSH EXHIBIT IIIb (Continued) Net cash flow from investing activities Cash Flow from Financing Activities New debt proceeds Principal payments Endowment gifts Net cash flow from financing activities Decrease in cash and cash equivalents Beginning cash and cash equivalents Ending cash and cash equivalents -76,104 $ $ 43,153 -46,447 10,314 7,020 -21,568 103,726 82,158 The Balance Sheet (Exhibits Ia and Ib), Activity Statement (Exhibits IIa and IIb), and Cash Flow (Exhibits IIIa and IIIb) illustrate the differences in the financial statement recognition and display following GASB and FASB guidance. All financial statements’ data are presented in thousands of dollars. The Sample State University statements follow GASB guidance while Sample University follows FASB guidance. Although the statements reflect the same transactions that occurred during the fiscal year, the statements reflect many differences in recognition (amounts), recording, and presentation. References are made to items within the exhibits throughout the paper to illustrate specific reporting differences. SPECIFIC DIFFERENCES Contributed Services The FASB set forth specific criteria for private institutions to identify and recognize contributed services. The GASB does not have criteria for recognizing contributed services, however, there may be a GASB project to address contributed services in the future. Currently, private institutions recognize contributed services that meet certain criteria established in SFAS No. 116 (1993a). The effect of this difference is the amount of gift revenues and expenses recognized and reported. In most cases, there is no impact on net revenues because the contribution revenues usually are offset by an equal expense. An exception occurs when the contributed services create or enhance nonfinancial assets. In those cases, net assets increase by the value of the services. Contributed services received by Sample TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 441 University are included in the gift revenue and the salary and wage expense displayed in Exhibit IIb. Restricted Cash Contributions For FASB institutions, restricted cash contributions are recognized as either temporarily or permanently restricted based on whether the restriction can be met or not – see Sample University revenue display in Exhibit IIb. For GASB institutions, time restricted cash contributions are recognized as deferred revenue displayed in Exhibit Ia until the restriction is met. This affects the amount of liabilities and gift revenues reported and thereby impacts net assets for both the public and private institutions. Endowment Pledges The difference in recognition of endowment pledges is quite noticeable. Endowment pledges to give are recognized by private organizations as a Pledge Receivable (Exhibit Ib) and as Permanently Restricted Revenue (Exhibit IIb). A comparable endowment pledge will not be recognized by a public institution. The FASB recognizes endowment pledges (promise to give) as permanently restricted as the endowment corpus cannot be spent. Rather, the income earned on the investment of the endowment is used according to the donor’s either restricted or unrestricted direction. The GASB prohibits recognition of endowment pledges based on its conclusion that a promise to give cannot satisfy the restriction because resources have not been received. This difference affects the recognition of assets, gift revenues, and net assets. Exhibit IIa illustrates that Sample State University (GASB) would recognize only $10.3 million as endowment gifts while Exhibit IIb for Sample University (FASB) recognized $20.3 million including an endowment pledge of $10 million. On the Balance Sheet (Exhibit Ia), Sample State University reports the endowment gift as a noncurrent investment and restricted nonexpendable endowment net assets held for its fundraising foundation affiliated organization ($66.2 million) while Exhibit Ib for Sample University (FASB) reports the endowment gift as an investment and permanently restricted net assets. 442 FISCHER & MARSH Restricted Non-Endowment Pledges The FASB recognizes restricted non-endowment pledges as temporarily restricted revenue (see Exhibit IIb Statement of Activity). The GASB prohibits recognition of pledges for future period use, under the same concept as endowment pledges – one cannot recognize the gift before it is available for use. This results in private institutions recognizing and reporting larger amounts of assets, gift revenues, and net assets than public institutions. Discounting of Pledges The FASB requires discounting if the pledge is collectible in installments over multiple years. That is, the net present value of the annuity is recognized when the installment gift is pledged and the pledge receivable is adjusted each year of the promise to reflect the time value of money as additional revenue is recognized. The GASB permits discounting, but does not require it. The difference in recognizing installment gifts affects assets, gift revenues and net assets. In Exhibit Ia, Sample State University (GASB) reports the noncurrent asset pledge receivable as $2.1 million while the Sample University (FASB) (Exhibit Ib) reports a current asset pledge receivable in the amount of $15.8 million. The differences with regard to pledges do not preclude GASB institutions from recognizing all pledges, only those that meet certain criteria, i.e., the eligibility to be utilized upon receipt of the pledge. The logic for the difference in discounting is APB Opinion 20 (AICPA 1971), which drives the recognition of discounts, and applies to exchange transactions. By definition, pledges are nonexchange transactions because the donor does not expect anything in return for the gift. Therefore, the GASB does not require discounting. Investment Income GASBS No. 31 (1997) establishes prescriptive treatment for investment income. Investment income and realized investment gains/losses must be reported as a single net amount illustrated by Exhibit IIa Sample State University reporting nonoperating investment income of $16.4 million. In most public institutions’ financial statements, investment income includes current yield and net realized gains and losses in addition to the net unrealized gains and TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 443 losses. Net unrealized gains and losses can be displayed separately from investment income by using a prescribed label: net increase (decrease) in the fair value of investments. Investment income cannot be reported as operating revenue by public colleges and universities unless it results from loans to students, which are classified as program loans. In fact, the only governmental entities that can report investment income as operating revenue are pooled-investment entities and fundraising foundations whose mission includes investment activities. The FASB does not have similar requirements related to investment income display or classification as nonoperating. Thus as displayed in Exhibit IIb, investment income and losses are combined and displayed as revenue for the related net asset class – unrestricted, temporarily restricted or permanently restricted. Pell Grants Funds received for Pell Grants are not reported as revenue by FASB institutions. Because the resources are funds held for the student, the funds are held in trust for the students in an agency account and not recognized as institutional revenue. It is a balance sheet cash and accounts payable to the student rather than an activity statement transaction. For GASB institutions, Pell Grant receipts are recognized as institutional revenue in accordance with GASBS No. 24 (1994) that directs all grant receipts to be reported as revenue. Whether the revenue is considered operating or nonoperating is somewhat controversial. The differing recognition treatment affects grants and contracts revenue, which will be higher for public institutions. Exhibit IIa indicates government grants and contracts revenue in the amount of $86.1 million for public universities versus $66.1 million for private institutions (Exhibit IIb). As a result of this recognition difference, tuition revenue will be higher for private institutions. Note that Exhibit IIb reports tuition and fee revenue of $278.3 million for private universities but only $193.4 million for public universities in Exhibit IIa due to the different recognition of Pell grant funds. In addition, liabilities and net assets may often vary. 444 FISCHER & MARSH Perkins Loan Program For private (FASB) institutions, federal advances for Perkins loans are reflected as liabilities on the balance sheet. Public (GASB) institutions have an option to reflect Perkins loan advances as liabilities on the balance sheet or revenues on the activities statement. Both are acceptable under current accounting guidance. The determination has to do with whether Perkins loan funds are treated as a net asset or a liability. Some believe the advances are a liability because the federal government can mandate their return. Others believe this possibility is so remote that they should be reported as revenues and net assets. National Association of College and University Business Officers (NACUBO) raised the issue with the GASB, suggesting that it is inappropriate to treat something as a net asset when an external party can require it to be returned. The GASB recognizes that different rules apply to different types of loans. Until the GASB adds the issue to its agenda and resolves the question, either approach is used. If public institutions reflect Perkins awards on the activities statement, it results in differences from private institutions for grants and contracts revenue, liabilities, and net assets. Both Sample State University and Sample University recognize Perkins student loan advances as liabilities in Exhibits Ia and Ib. Funds Held in Trust by Others Funds held in trust by others are recorded differently in terms of both recognition and display. When the private institution is an irrevocable beneficiary, the private (FASB) institution includes these resources as assets. The Statement of Financial Position in Exhibit Ib for Sample University reports investments held by others in the amount of $70.8 million while there is no comparable account for the Sample State University in Exhibit Ia. Public institutions do not treat funds held in trust by others as an asset; rather, if material, they are disclosed in the notes to the financial statements. However, under GASBS No. 39 (2002), if the external entity meets specified criteria, it is reported as an affiliated organization resource or component unit using a discrete presentation. That is, the financial statements for the affiliated organization will be included with those of the public institution in the annual financial report. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 445 Technically, public institutions can report blended component units when the unit holds assets in the institution’s name. Those assets would be included in the reported amounts but would not be labeled funds held in trust by others. The effect of this difference is that funds held in trust by others will appear on private institutions’ statements as a specific amount, but not on those of public institutions (unless they qualify as component units). In addition, private institutions display revenue/loss as a result of a change in the value of the funds held in trust that results in a change in net assets. Restriction Definition The FASB’s definition of who can establish restrictions is much narrower than the GASB. Under FASB, only donors can restrict resources held by the entity whereas under the GASB any external party – donors, creditors, legislation, contracts, and constitutional provisions can impose restrictions. This stems from the GASB’s philosophy toward accountability that if any restrictions are imposed, the GASB wants them to be clearly identified. The effect of the restriction is only on the categorization of revenues and net assets. Unless donor directed, resources would be displayed as unrestricted under the FASB guidance but would be displayed as restricted expendable under GASB guidance. There is no difference for true endowments, which are restricted nonexpendable for public institutions and permanently restricted for private institutions. Exhibit Ia shows the various restrictions for restricted expendable net assets of Sample State University while Exhibit Ib for Sample University does not give any details regarding the temporarily restricted net assets. Use of Restricted Funds The FASB’s guidance on the use of restricted funds is another area of difference and is considered one of the most controversial steps that the FASB has taken with respect to not-for-profit organizations. The FASB mandates the first dollar release method. Under first dollar release, restrictions are released if unrestricted resources are used for a purpose for which restricted resources are available. Although internal accounting and reporting may not be affected, the application of first dollar release will convert previously restricted resources into unrestricted resources simply because they 446 FISCHER & MARSH could have been expended. An amendment to SFAS No. 117 allows the requirement to be optional. The GASB considered the mandating position but concluded that it should not be required in a governmental environment. Instead they chose to recognize first dollar release as an acceptable method, making it optional. The GASB requires note disclosure of the policy applied in situations when both restricted and unrestricted resources are available for the same purpose. Again, this affects the categorization of net assets. Other Postemployment Benefits (OPEB) and Pensions The GASB uses the term OPEB when referring to other postemployment benefits which consist primarily of health care for retirees. The comparable term in FASB literature is other postretirement benefits. While the GASB deliberated recognition and reporting of OPEB and pensions, the Board concluded that similarities did exist between private and public sectors. The Board could have adopted the FASB guidance that was already in place but determined that significant relevant differences existed between for-profits and governments which warranted different guidance. For example, governmental entities can be assumed to have a ‘perpetual’ existence unlike businesses that can be liquidated. In addition, businesses’ prime concern is net income resulting from revenues exceeding expenses whereas governmental entities are concerned with compliance. While the GASB and the FASB have different approaches to calculating pension and other postemployment liabilities, they are consistent within their own methodologies for calculating these types of liabilities. This affects the measurement and recognition of those liabilities which impacts expense, liability, and net assets. Both Sample State University and Sample University display a liability for benefits payable but with different amounts in Exhibits Ia and Ib. Impairment The FASB requires a cash flow approach for determining impairment loss (loss based on measurement of expected cash flows). Thus private institutions must recognize an impairment cost TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 447 and reduction of a capital asset’s value when its carrying value will not be recovered by future cash flows. The GASB considered the cash flow approach but opted instead for a service utility approach when GASBS No. 42 (2003b) was issued. The impairment is determined by calculating the decline in service (years) based on the impairment event or change in circumstances. These events include technological obsolescence, environmental changes, or changes in the capital asset usage. The different methods for measuring an impairment loss also include (1) an estimate of the costs needed to restore the utility of the assets and (2) deflating the depreciation to reflect a replacement cost. Each of these approaches affects assets, expenses, losses, and net assets. Neither Sample State University nor Sample University reports an impairment expense. Endowment Losses The FASB requires that permanently restricted net assets remain intact—insulated from reductions caused by the recognition of losses. Therefore, SFAS No. 124 (1995) requires that endowment losses reduce temporarily restricted net assets to the extent of unspent appreciation. If losses exceed such amounts, the excess is reflected as a reduction in the unrestricted net assets. If the loss is considered temporary, a loss contra can be employed. However, when the Uniform Investment Act has been adopted in the state where the private institution resides, the act’s guidance prevails. The GASB has no similar requirement. Losses attributable to restricted nonexpendable net assets reduce the restricted nonexpendable net assets. Both Sample State University and Sample University report investment losses in Exhibits IIa and IIb. Sample State University reports the loss as Other Items while Sample University reports negative revenue in the temporarily restricted net asset class. Fair Value The SFAS No. 124 (1995) provides a definition of fair value and guidance for estimating the fair market value of investment in debt and equity securities. Private institutions report all debt and equity securities at fair value except for those investments accounted for by the equity method. Gains and losses, both realized and unrealized, 448 FISCHER & MARSH are recognized and reported in the statement of activities. This guidance results in increased consistency and comparability in fair value measurements. However, SFAS No. 157 (2006) and other accounting codification updates (2009c, 2010) introduced inconsistency as institutions sought to value investments other than equities and debt at their fair value. The GASB provides reporting and recognition guidance in its GASBS No. 31 (1997) that requires public institutions to report investment income (including changes in fair value) in the statement of revenue, expenses and changes in net assets. Money market investments with a maturity of less than one year may, however, be reported at their amortized cost rather than market value. Fair value recognition is reported by both Sample State University and Sample University as part of investment gains/losses in the Exhibits IIa and IIb. Endowments Holding Real Estate GASBS No. 52 (2007a) amended its earlier guidance to require public institutions to report their endowment investments of land and other real estate investment at fair value. In addition, it requires governments to report the changes in fair value as investment income and to disclose the methods and assumptions used to determine fair value. This reporting enhances the users’ ability to evaluate an entity’s investment decisions and performance. SFAS No. 157 (2006) provides similar guidance for long-term assets to be reported at fair value. Institutions typically use comparable estimates to determine the reportable values rather than relying on appraisals. The result of this recognition difference is not transparent in the financial statements. Rather, the financial statement user must rely on financial statement note disclosures to discern any change of the long-term asset value. Risk Assessment and Deposit Disclosures GASBS No. 40 (2003a) addresses various deposit and investment risks such as credit risk, concentration of credit risk, interest rate risk and foreign currency risk. It requires disclosure of TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 449 investments that have fair values that are highly sensitive to changes in interest rates. In addition, it requires disclosure of policies related to the various risks. The FASB has no similar requirement for private institutions. There is a benefit for publics with respect to concerns regarding financial transparency and risk assessment. Regardless of the risks, the financial statement user must rely on the financial statement note disclosures to identify any risk exposure. Early Retirement Plans Public institutions are obligated to report the financial obligations associated with early retirement plans when the plan is announced (GASB 2005). The obligations are allocated to the appropriate fiscal years when the plan is in place. This recognition causes public institution’s expenses and liabilities to be greater than a comparable private institution that also has an early retirement plan because there is no FASB reporting requirement for private institutions to recognize early retirement costs. Instead, private institutions report the early retirement related costs as they are incurred or expended. Neither Sample State University nor Sample University display any early retirement plan expenses or liability in Exhibits IIa or IIb. Intangible Assets Public institutions must report intangible assets, except goodwill, acquired in a combination transaction, as a capital asset (GASB 2007b). This recognition requirement ensures that nonfinancial assets without physical structure are displayed as noncurrent assets in the statement of net assets. Private institutions have no display restrictions but typically display intangible assets following the property, plant and equipment assets. Neither Sample State University nor Sample University report any intangible assets in Exhibits Ia or Ib. Collections The GASB and FASB do not require, but encourage, capitalization of collections (art, books, etc.) as long as the collections meet three conditions which are the same - held for educational purposes, cared for/protected, and proceeds used to add to the collection if items are 450 FISCHER & MARSH sold. If the collectibles do not meet these conditions, they must be capitalized. Exhaustible capitalized collectibles should be depreciated. Both the GASB and FASB require recognition of collectible contributions as revenue if the collectible is capitalized. However, if collectible is not capitalized the GASB (1999a) requires a charge to a program expense to offset any revenue. In contrast, the FASB (1993b) does not allow recognition of revenue for items contributed to noncapitalized collections. These differences can be identified with the comparison of expenses, revenues and net assets. Neither Sample State University nor Sample University capitalized collection is their balance sheet displayed in Exhibits Ia or Ib. Management Discussion and Analysis The GASB prescribes the reporting of a management’s discussion and analysis (MD&A) and outlines specific information that must be presented (GASB 1999a ¶11). The MD&A must discuss the relationships between each of the required financial statements as well as capital projects, functional expenses, change in long-term debt and the impact of known future events. The discussion should be easy to read and understand but avoid “boilerplate” language (GASB 1999a ¶11). The FASB has no similar requirement for private institutions. There is no substantive effect of this difference as private institutions tend to include descriptive transmittal letters from the president and financial leaders to convey the current state of the institution to the financial statement reader. Given this, there is a benefit for publics with respect to concerns regarding financial transparency. Disaggregation (Columns on Statements) The FASB allows line of business (e.g. academic, auxiliaries) or net asset class (i.e., unrestricted, temporarily restricted, permanently restricted) disaggregation. The GASB allows only line of business disaggregation that includes nonexpendable restricted and expendable restricted net assets. Unrestricted net asset class disaggregation is prohibited (GASB 1999a, ¶37) but disaggregation display is allowed as a note disclosure. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 451 Both Sample State University and Sample University display equity information by net asset class in Exhibits Ia and Ib. Only Sample University displays revenue and expenses in Exhibit IIb by net asset class. Balance Sheet Display The GASB (1999a) requires a classified balance sheet (i.e., current, noncurrent) and defines three net asset classes (capital assets, net of related debt – [that is a part of unrestricted new assets in FASB statements (1993b)], restricted and unrestricted). Exhibits Ia and Ib detail the different displays. If an institution has permanent endowments, it must classify restricted net assets as expendable or nonexpendable. Restricted expendable is very similar to temporarily restricted in FASB statements. It does not match exactly because the GASB’s broader definition of restriction results in some net assets being classified as restricted expendable by public institutions when the same net assets would be classified as unrestricted by private institutions. Capital assets that are being or have been depreciated are reported net of accumulated depreciation. When a government has a significant amount of nondepreciable capital assets such as land or infrastructure, the GASB requires that they be reported separately from depreciable capital assets (GASB 1999a ¶20). The FASB does not have a classification requirement. The FASB net asset classes are unrestricted, temporarily restricted, and permanently restricted. The FASB has no specific requirements related to the display of capital assets, although they are considered part of unrestricted net assets. Exhibits Ia and Ib display the different formats for the public and private institutions. Activities Statement Reclassifications The FASB (1993b) treats all expenses as unrestricted. To the extent that temporarily restricted provisions have been met and the resources are to be used to finance an expense, they must be reclassified from temporarily restricted to unrestricted net assets. 452 FISCHER & MARSH The GASB has no similar concept—expenses can be unrestricted, restricted, operating, or nonoperating (See Exhibits IIa and IIb). Operating Measure The GASB (1999a) requires a somewhat prescriptive operating measure. State appropriations, gifts, and investment income (except interest on program loans) must be reported as nonoperating. Institutions are allowed to decide how to treat other items. For example, gains and losses resulting from the disposition of capitalized equipment can be either operating or nonoperating. In institutions where the disposal of capital equipment occurs frequently, and is a routine part of doing business, it might be classified as operating. If it occurs infrequently, it likely will be reported as nonoperating. There has been a great deal of discussion with regard to an operating measure for private institutions. The FASB allows for a selfdefined operating measure and requires disclosure of the methodology for determining the measure if it is not obvious. See Fischer et al (2004) for a discussion of the variances found among college and university financial statements that report an operating measure, together with how the differences impact peer comparison and benchmarking. Exhibit IIb does not display an operating measure for Sample University. A quasi measure Net Income before Capital Items of $1.4 million is reported for Sample State University in Exhibit IIa. Expense Classification The FASB (1993b) allows natural classification of expenses but requires functional classification, either in the display or in the notes. The GASB (1999a) allows natural or functional expense classification display. NACUBO encourages a matrix reconciling functional and natural expense classification to be included in the notes. Both Sample State University and Sample University use natural classifications to report their expenses in Exhibits IIa and IIb. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 453 Specific Expenses The FASB (1993b) requires allocation of certain functional expenses, such as operations and maintenance of plant, depreciation, and interest. The GASB (1999a) allows, but does not require, allocation of depreciation. The GASB requires bad debts to be offset against the corresponding revenue source. The exception to this general rule is for bad debts resulting from loans because there is no revenue source. Since Sample University used natural classification to display expenses in Exhibit IIb, no allocation was necessary. Exception Items The FASB defines extraordinary items as those that are unusual and infrequent. The GASB has the same definition for extraordinary items and goes on to specify an additional category—special items. Special items are unusual or infrequent, and within management’s control. For example, if a public institution elects to abandon a software implementation in which significant dollars have been invested (and capitalized based on NACUBO guidance), it would report this as a special item because it is within management’s control. A comparable private institution might feel that this transaction is neither unusual nor infrequent and; therefore, record it as part of operating expenses. Cash Flow Statement There are significant differences in the cash flow statement display (see Exhibits IIIa and IIIb). The FASB (1993b) allows the indirect method and specifies three categories for reporting cash flows: operating, investing, and financing. Operating activities display the cash flow in the revenue and expense activities adjusted by the changes in the working capital accounts. The investing activities detail the cash flow changes in the nonworking capital and long-term asset accounts. The financing activities display the changes in the long-term liabilities and equity gifts for long-lived assets and endowments. In contrast, the GASB (1999a) mandates the direct method to display cash flows from operating activities that requires operating 454 FISCHER & MARSH cash flows (both revenue and expense) be reported by type. The GASB also requires a reconciliation of the operating cash to operating income as reported on the Statement of Revenues, Expenses, and Changes in Net Assets (Exhibit IIa). The GASB specifies the use of four categories for presenting cash flows: operating, investing, capital and related financing, and noncapital financing. These categories are different from those prescribed for private institutions and are based on activities being performed rather than balance sheet components. Exhibits IIIa and IIIb detail the different displays for Sample State University and Sample University. The GASB does not require the inclusion of the cash effects associated with nonoperating revenues and expenses as cash flows from operating activities. For example, cash paid for interest on borrowings is included in cash flows from noncapital financing activities or cash flows from capital and related financing activities depending on the circumstances. In Exhibit IIIa, the debt interest payment of $9.1 million made by Sample State University and reported as Cash Flows from Capital Financing Activities is part of the change in net assets within the Operating Activities reported by Sample University in Exhibit IIIb. Cash received from interest and dividends is displayed by Sample State University in the Cash Flows from Investing Activities while Sample University includes the interest income as part of their operating activities. Another major difference is the GASB requirement that cash payments to acquire capital assets be reported as Cash Flows from Capital Financing Activities and not Cash Flows from Investing used by Sample University to report the transaction. Exhibit IIIa displays Sample State University’s $74.6 million payment to acquire capital assets as part of Cash Flows from Capital Financing Activities while Sample University in Exhibit IIIb reports the same transaction/amount in the Cash Flows from Investing Activities. Additionally, cash received from the sale or disposal of capital assets is reported by Sample State University as cash flows from Capital Financing Activities while Sample University reports the same transaction as an investing activity. These are only a few of the differences between the FASB and GASB statement of cash flow display guidance. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 455 CURRENT GAAP REPORTING PROJECTS Several research topics currently on the GASB’s action agenda will expand the public versus private nonprofit recognition and divergence. One research item is monitoring the effect of the electronic media on information delivery and user needs to provide the GASB with a basis for evaluating the need to develop standards for electronic financial reporting. Another research item considers whether any additional information is useful for assessing a government’s economic condition and determining if it should be required or encouraged for inclusion in the financial report. Other research agenda items will re-examine current guidance pertaining to fair value measurement, pension accounting (GASB 2009b), and the reporting of component units (GASB 1991). The GASB recently issued Statement No. 58 (2009a) to provide accounting and financial reporting guidance for governments that have been granted protection from creditors under Chapter 9 of the United States Bankruptcy Code. It requires governments to remeasure liabilities that are adjusted in bankruptcy when the bankruptcy court confirms (approves) a new payment plan. Governmental nonprofit entities are not excluded from the guidance. Both the GASB and FASB had leases on their research agendas but the GASB put their investigation on hold pending the work currently underway at the FASB. The FASB is working with the International Accounting Standards Board (IASB) to reconcile lease recognition and accounting. Once this work is completed, the GASB will review the conclusion to determine whether to accept or develop specific guidance for US governmental entities. In addition to the joint project with IASB on leases, the FASB is working on a revision of the financial statement presentation. This project is considering a common approach to presenting information in financial statements using the current cash flow statement categories. To date, nonprofit entities have been specifically excluded from the deliberations. A FASB project that did impact the nongovernment nonprofit entities is the accounting standards codification (FASB 2009b). The codification is a single, authoritative, comprehensive integrated compilation of US accounting and financial reporting literature. It 456 FISCHER & MARSH became GAAP in June 2009 as a fully functional, online, real-time database with searching capability and retrieval. The GASB maintains a codification for GASB guidance in hardcopy paper format that can be purchased. For an annual fee, a searchable, digitized format of the GASB codifications is available. CONCLUSION As evidenced by the discussion of the specific differences between the financial recognition and reporting by public (governmental/GASB) and private (nongovernmental/FASB), a comparison between comparable institutions can lead to a very different financial conclusion. Table 1 presents common ratio and other analysis using financial information presented in Exhibits Ia, Ib, IIa and IIb. TABLE 1 Comparative Financial Statement Ratios Sample Sample State University University value value Liquidity current ratio current assets/current liabilities 2.546 3.483a quick ratio cash+ marketable security+ receivables/ current liabilities 2.454 3.483a days of cash on hand cash+ short term marketable securities/ (total operating expenses+ depreciation/365) 129.376 166.219 6.623 6.292 55.111 58.011 1.243 1.074 Asset turnover receivables revenue-tuition revenue/accounts receivable average 365/receivable turnover collection period fixed asset turnover total unrestricted revenue/net fixed assets TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 457 TABLE 1 (Continued) Sample Sample State University University value value total asset turnover total unrestricted revenue/total assets 0.631 0.537 Leverage debt ratio debt to equity Profitability total long-term debt/total assets total long-term debt/total net assets 0.337 0.588 0.343b 0.591b -0.019 0.011 -0.01 0.006 -0.017 0.011 increase in net assets/total unrestricted revenue and other support Return on assets increase in net assets/total assets return on net increase in net assets/total net assets assets total margin Notes: a. Without a classified balance sheet, this measure is an estimate that assumes assets and liabilities are presented in order of liquidity. b. Current portion of long-term debt may or may not be displayed in financial statement notes. Four categories of analysis are illustrated – liquidity, asset turnover, leverage and profitability-all available to the average financial statement user. Given the different reporting and recognition criteria used by Sample State University and Sample University, Sample University will typically report better liquidity measures because gift pledges at Sample State University cannot be recognized before the institution is eligible (based on time or purpose) to use the gift. Unless the current portion of long-term liabilities or obligations is known, Sample University analysis should yield better leverage values. With the immediate recognition of gift revenues, regardless of whether the gift is given as cash or a pledge, Sample University profitability analysis should exceed that of Sample State University. Sample State University, however, should report better turnover measures thanks to a lower amount of gift pledge receivables. Sample University’s tuition revenue will exceed that of Sample State University because Sample University will recognize the Pell Grant awards as tuition revenue rather than contract and grant revenue that Sample State recognizes. Without knowledge of why the 458 FISCHER & MARSH data differ, the financial statement user could make inappropriate conclusions about the financial health of an institution. Given current reporting and recognition guidance based on the fact that the FASB and the GASB treat time restriction differently, a comparable private institution will always report greater gift revenue than its comparable public institution due to the private’s recognition of pledged gifts. The time restriction treatment is the primary cause for some of the more significant reporting and recognition differences. Another significant difference is the requirement for public institutions to report state appropriations as nonoperating revenue that results in public institutions reporting a negative operating income. When a beneficial trust is held by a third party, the private institution will always report a larger amount of assets on the balance sheet than does the public institution. Thanks to the requirement for public institutions to recognize Pell Grant funds as revenue, comparable public institutions will always report less tuition revenue than private institutions. And the list of differences in reporting and recognition by nonprofit organizations goes on and on as outlined in Table 2. The divergence in financial recognition and reporting between private and public nonprofit institutions appears to be increasing rather than decreasing. TABLE 2 FASB GASB: Summary of Differences Panel A Financial Statements Governmental Nonprofit Organization Uses AICPA Government Audit Guide (1999) Nongovernment Notfor-profit Organization Uses AICPA Not-forProfit Audit Guide (1996) GASB 35/34 (1999a; 1999b) MD&A (RSI) Statement of Net Assets Statement of Revenues, Expenses and Change in Net Assets Statement of Cash Flow Notes Other Required Supplemental Information (RSI) FASB 117 (1993b) Statement of Financial Position Statement of Activities Statement of Cash Flows Notes TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 459 TABLE 2 (Continued) Panel B Governmental Nonprofit Organization Balance Sheet GASB 35/34 (1999a;1999b) Classified per ARB 43 Current Noncurrent A = L + NA A L NA (No totals) Within classificationsassets and liabilities presented in liquidity order Investments (including real estate) presented at fair value Funds held in trust by others reported in notes if significant Capital Assets net of related debt Restricted Nonexpendable Restricted Expendable Unrestricted (designation prohibited) Nongovernment Notfor-profit Organization FASB (1993b) Classification not Required A = L + NA Assets and liabilities presented in liquidity order Investments (real estate optional) presented at fair value Funds held in trust by others reported as assets and Permanently Restricted NA Unrestricted Net Assets Temporarily Restricted NA Permanently Restricted NA Panel C Income Statement Governmental Nonprofit Organization Expenses display in appropriate net asset class Any external party can restrict resources Restriction met when first $ spent - optional Nongovernment Notfor-profit Organization All expenses displayed as unrestricted Only donors can restrict Resources Restriction met when first $ spent - optional Met restrictions are reclassed into Unrestricted Net Assets Expenses displayed functionally or by natural class Expenses displayed functionally or by natural class 460 FISCHER & MARSH TABLE 2 (Continued) Plant operation and depreciation may be presented as functional designation Revenues presented by type Operating measure prescribed Realized and unrealized investment income presented separately State appropriations reported as non-operating revenue Plant operation, depreciation and interest must be allocated (no guidance provided) Revenues presented by type Operating measure-optional Realized and unrealized investment income may be netted Appropriations may be reported as operating revenue Panel D Cash Flow Governmental Nonprofit Organization GASB 9 (1989) Direct method mandated Reconciliation required (indirect method to report operating activities) Cash equivalents same as cash if <90 days and original investment maturity 90 days or less Four categories Operating Non-Capital Financing Transfers Subsidies Contributions Agency transactions Capital Financing All capital asset related Capital contributions New capital debt Capital debt obligations Sale of capital assets Investing All investing activities Acquisitions Sales Interest earned Nongovernment Notfor-profit Organization FASB 117 (1993b) May use direct or indirect method Cash equivalents same as cash if <90 days and original investment maturity 90 days or less Three categories Operating measureNet income +/- change in working capital accounts Investing +/- changes in long term assets Financing +/- long term liabilities New endowment gifts TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 461 TABLE 2 (Continued) MD&A Notes Required supplemental reporting - contents prescribed Required by various GASB guidance Not required Required by various FASB guidance Legends: A = Assets; L = Liabilities; NA = Net Assets. REFERENCES American Institute of Certified Public Accountants - Accounting Principles Board (1971). Accounting Changes. New York: Author. American Institute of Certified Public Accountants (AICPA) (1973). Audits of Colleges and Universities. New York: Author. American Institute of Certified Public Accountants (AICPA) (1996). The Not-for-Profit Organizations Audit and Accounting Guide. New York: Author. American Institute of Certified Public Accountants (AICPA) (1997). Accounting for the Costs of Computer Software Developed or Obtained for Internal Use. New York: Author. American Institute of Certified Public Accountants (AICPA) (1999). State and Local Governments – Audit and Accounting Guide. New York: Author. Brown, R. E. (2005). “Enron/Anderson: Crisis in U. S. Accounting and Lessons for Government.” Public Budgeting & Finance, 25 (3): 20-33. Crawford, M. A. (2003). “Closing the Gap with Non-GAAP State and Locals.” Journal of Accountancy, 196 (6): 24. Engstrom, J. H., & Esmond-Kiger, C. (1997) “Different Formats, Same User Needs: A Comparison of the FASB and GASB College and University Financial Reporting Models.” Accounting Horizons, 11(3): 16-35. Financial Accounting Standards Board (FASB) (1987). Recognition of Depreciation by Not-for-Profit Organizations. Statement of Financial Standards No. 93. Norwalk, CT: Author. 462 FISCHER & MARSH Financial Accounting Standards Board (FASB) (1993a). Accounting for Certain Investments in Debt and Equity Securities. Statement of Financial Standards No. 116. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (1993b). Financial Statements of Not-for-Profit Organizations. Statement of Financial Standards No. 117. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (1995). Accounting for Certain Investments held for Not-for-Profit Organizations. Statement of Financial Accounting Standards No. 124. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (1999). Transfer of Assets to a Not-for-Profit Organization or Charitable Trust that Raises or Holds Contributions for Others. Statement of Financial Accounting Standards No. 136. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (2006). Fair Value Measures. Statement of Financial Accounting Standards No. 157. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (2009a). Not-for-Profit Entities: Mergers and Acquisitions Including an Amendment of FASB Statement No. 142. Statement of Financial Accounting Standards No. 164. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (2009b). The FASB Accounting Standards Codification and the Hierarchy of Generally Accepted Accounting Principles – a replacement of FASB Standard 162. Statement of Financial Accounting Standards No. 168. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (2009c, August). Fair Value Measures and Disclosures (Codification Topic 820): Measuring Liabilities at Fair Value. Norwalk, CT: Author. Financial Accounting Standards Board (FASB) (2010, January). Fair Value Measures and Disclosures (Codification Topic 820): Improving Disclosures about Fair Value Measurements. Norwalk, CT: Author. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 463 Fischer, M. (1997). “Two Accounting Standard Setters: One Industry.” Journal of Public Budgeting, Accounting & Financial Management, 9 (2): 251-284. Fischer, M., & Gordon, T. P. (2002, August). “Inching Toward an Operating Measure.” NACUBO Business Officer, 36 (2): 39-43. Fischer, M., Gordon, T. P., Keating, E., & Greenlee, J. (2004, January). “When Operating Measures Don’t Add Up.” NACUBO Business Officer, 38 (7): 33-36. Goldstein, L., & Menditto, S. (2005, January). “GASB and FASB.” NACUBO Business Officer, 39 (7): 21-26. Governmental Accounting Standards Board Applicability of FASB Statement No. 93 Depreciation by Not-for-Profit Organizations” to Local Governmental Entities. Statement No. Author. (GASB) (1988). “Recognition of Certain State and 8. Norwalk, CT: Governmental Accounting Standards Board (GASB) (1989). Reporting Cash Flows of Proprietary and Nonexpendable Trust Funds and Governmental Entities That Use the Proprietary Fund Accounting. Statement No. 9. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1991). The Financial Reporting Entity. Statement No. 14. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1994). Accounting and Financial Reporting for Certain Grants and Other Financial Assistance. Statement No. 24. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1997). Accounting and Financial Reporting for Certain Investments and for External Investment Pools. Statement No. 31. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1998). Accounting and Financial Reporting for Nonexchange Transactions. Statement No. 33. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1999a). Basic Financial Statements – and Management’s Discussion and 464 FISCHER & MARSH Analysis – for State and Local Governments. Statement No. 34. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (1999b). Basic Financial Statements – and Management’s Discussion and Analysis – for Public Colleges and Universities. Statement No. 35. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2002). Determining Whether Certain Organizations Are Component Units. Statement No. 39. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2003a). Deposit and Investment Risk Disclosures. Statement No. 40. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2003b). Accounting and Financial Reporting for Impairment of Capital Assets and for Insurance Recoveries. Statement No. 42. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2004). Accounting and Financial Reporting by Employers for Postemployment Benefits Other than Pensions. Statement No. 45. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2005). Accounting for Termination Benefits. Statement No. 47. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2007a). Land and Other Real Estate Held as Investments by Endowments. Statement No. 52. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2007b). Accounting and Financial Reporting for Intangible Assets. Statement No. 51. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2009a). Accounting and Financial Reporting for Chapter 9 Bankruptcies. Statement No. 58. Norwalk, CT: Author. Governmental Accounting Standards Board (GASB) (2009b). Invitation to Comment: Pension Accounting and Financial Reporting. Norwalk, CT: Author. TWO ACCOUNTING STANDARD SETTERS FOR NONPROFIT ORGANIZATIONS 465 Menditto, S., & Gordon, T. P. (2008, July-August). “Finances in the Clear.” NACUBO Business Officer, 42 (1): 38-45. Nelson, M., Banks, W., & Fisher, J. (2003). “Improved Accountability Disclosures by Canadian Universities.” Canadian Accounting Perspectives, 2 (1): 77-108. Robert, N. l., & Mohaghegh, S. (2004). “A Comparison between Financial Reporting in Health Care versus Other Industries.” Journal of American Academy of Business Cambridge, 5(1/2): 333-343. Smith, G. R., Freeman, R. J., & Bryan, B. J. (1997). “GASB versus FASB; User Perceptions of the Alternative Formats for Preparing the Statement of Cash Flows.” Journal of Public Budgeting, Accounting & Financial Management, 9 (2): 224-250. US Congress (2002). Sarbanes-Oxley Act Public Law 107-204. Washington, DC: U S Congress. Van Daniker, R. P. (2009). “Remembering GASB’s Beginning.” The Journal of Government Financial Management, 58 (3): 4. J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT, 24 (3), 466-488 FALL 2012 THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX: HOW LOCAL OPTION SALES TAXES EXACERBATE BUDGETARY INEQUALITIES BETWEEN LOCAL GOVERNMENTS Patrick J. McHugh and G. Jason Jolley* ABSTRACT. This paper tests the theory that local option sales taxes (LOST) work to the disadvantage of poorer localities, particularly rural areas, where many residents commute to shop and work. We also hypothesize that LOST systems hurt struggling communities more than they help prospering ones. The LOST system is examined using multiple years of data from North Carolina, a state whose tax structure favors such an analysis. The results indicate that LOST systems exacerbate inequality between local communities by actively moving revenue from poorer communities to more wealthy ones. We find evidence that LOST systems cost poorer counties a greater percentage of their total budgets than is gained by the wealthy counties that attract retail activity. INTRODUCTION Local option sales taxes (LOST) came into vogue in the late 1970s and 1980s during the “property tax-revolt” and they now exist in more than thirty states. Local option sales taxes were initially proposed as an alternative revenue source that would allow local governments to provide property tax relief. Another touted strength of this policy is that each local government gets to decide for itself --------------------------* Patrick J. McHugh, Ph.D., is a Senior Research Associate, International Economic Development Council. His teaching and research interests include public policy, public administration, economic development, political theory and political communication. G. Jason Jolley, Ph.D., is the Senior Research Director, Center for Competitive Economies, and Adjunct Assistant Professor at the Kenan-Flagler Business School, University of North Carolina at Chapel Hill. His teaching and research interests include public administration, public policy, public finance, and economic development. Copyright © 2012 by PrAcademics Press THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 467 whether to adopt the tax, so reform is not being heavy-handedly imposed by state governments. While appealing for a number of reasons, the performance of local option sales taxes in the real world have been largely disappointing. First, evidence that LOST revenues allow localities to reduce property taxes is mixed, at best (Zhao & Jung, 2008; Sjoquist, Walker, & Wallace, 2005). In this paper, we focus on a second concern; there is growing evidence that local option sales taxes tend to shift revenue from poorer, often rural, communities to more affluent ones, making LOST systems regressive at both the individual and community levels. There is considerable evidence that poorer taxpayers pay a larger percentage of their income on sales taxes than do wealthier taxpayers (Rostvold, 1966; Ghazanfar, 1978; Schweitzer & Taylor, 2008; Shoup 1983; Siegfried & Smith, 1997). An effort to replace federal income taxes with a national sales tax in the 1990's never got off the ground because congressional Democrats and President Clinton saw the sales tax as a bad deal for the poor. Moreover, there is little evidence that states systematically use exemptions for particular goods to ease the regressive nature of sales taxation (Fletcher & Murray, 2006). We maintain that local option sales taxes tend to be regressive at the community level as well, and this paper tests whether local option sales taxes actively shift revenue from struggling communities to their more wealthy neighbors. Because retail activity is increasingly clustered in wealthier and more populous areas, we expect sales tax revenue to flow from more impoverished rural areas to centers of wealth and population that are capturing retail business. In a national sales tax, poorer citizens pay a larger percentage of their resources, but at least they are part of the population those funds serve. When sales tax revenues flow from one county to the next, the citizens in the losing county are actively paying for their more prosperous neighbors' roads, schools, and social services. How Local Option Taxes Work and Why they Reinforce Inequality Most of the attention to the inequalities of local governmental revenue has historically focused on property taxes. Imbalances in education finance, produced by imbalances in property tax revenue, have generated equal protection lawsuits in several states and substantial media attention. While property taxes are important, there 468 MCHUGH & JOLLEY is good reason to also examine how local option sales tax systems function and whether they reinforce inequalities between local communities. Sales tax policy is determined at the state level and most revenues go to state governments. Local option sales taxes exist where state governments give localities the option of levying an additional tax on top of the statewide rate. In most cases, state governments grant counties the authority to enact local sales taxes that will stay with the county governments. From the perspective of representation, this seems to be an equitable arrangement; communities that favor increased sales taxes can do so without forcing others go along for the ride. While LOST mechanisms may be intuitively appealing, the way they work in the real world raises serious concerns. Like property taxes, areas with more wealth generate more LOST revenues for local governments. While disparities based on wealth are to be expected, we contend that market forces shift revenue from poorer communities to more wealthy ones, making the final distribution even more uneven than underlying differences in wealth. Retailers follow rooftops, especially in wealthy populous areas. Malls, outlet stores, boutique shops, restaurants, and other retailers are all are drawn to communities that can afford their services. As a result, poorer communities usually have fewer retail choices, forcing residents to drive to the nearest retail centers. Particularly as big-box retailers, super-centers, and regional malls have replaced local shops and grocers, retail activity is more concentrated than it was several decades ago (Artz & Stone, 2003). In many rural areas and innercities, local shops have been squeezed out of business when larger retailers offering lower prices move into communities nearby. On top of differences in retail choices, many residents of poorer communities are forced to commute to find work. The sales tax revenues for whatever purchasing commuters do near their places of employment will stay where that shopping occurred. Whenever residents of one locality commuting, travel to shop, or shop elsewhere because it is near their work, they leave the LOST revenues behind. Because we expect residents of poorer communities to have more cause to travel to shop than their more affluent neighbors, we expect that LOST revenues to flow in the same direction. Unlike property taxes, whose variation across counties is purely an effect of differences in property THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 469 value, LOST revenues may be literally riding in wallets from poorer areas to more prosperous ones. While the inequities in LOST revenues have not been scrutinized as much as property taxes, the existing work supports the theory articulated here. Much of the existing work has focused on how LOST inequalities impact levels of education funding, just as has been the case for property taxes. Work on inequalities in education funding has indicated that LOST systems tend to reinforce the differences in revenue resulting from inequities in property tax base (Rubenstein and Freeman 2003). One study claimed that LOST inequalities violated the equal protection clause of Iowa's state constitution (Craft, 2002). Disturbing evidence about LOST inequalities is not limited to education funding. In a study of revenue inequalities between county governments in Georgia, Zhao and Hou found that LOST revenues were even more unevenly distributed than were property tax revenues (Zhao and Hou 2008). A study of LOST revenues in Iowa estimated that, over a ten year period, $800 million in LOST revenues would move from “retail poor” to “retail rich” counties (Artz & Stone, 2003). Another study in Tennessee found that when a rural county is in close proximity to two newly-opened urban malls, their sales tax revenues will fall an average of 16% (Chervin, Edmiston, & Murray, 2000). In this paper, we examine whether LOST systems are regressive in a variety of ways. Do local option sales taxes shift resources away from counties with poorer populations? Do LOST systems move revenue out of communities that are under more fiscal strain? Can we find direct evidence that LOST revenues move with the flow of traffic from poorer to more affluent communities? Finally, do poorer communities lose a greater proportion of their overall budget than is gained by their more affluent neighbors? For the first time, this paper shows that LOST systems tend to hurt poorer communities more than they help affluent ones, resulting in a net economic loss to the local economy. We examine LOST systems from a variety of vantage points, and find consistent evidence that the policy actively moves public resources away from where they are needed most. Outline of this Study The methods section discusses why North Carolina's sales tax policy makes it an appealing case. North Carolina's system is quite 470 MCHUGH & JOLLEY different from the traditional LOST setup, but these differences turn out to be analytically advantageous for studying how point of sale systems impact county budgets. We then describe the details of projecting a per-capita allocation for each county in each year and, by comparing that projection to the reported point of sale collections, sorting counties who are advantaged under the point of sale system from those that are disadvantaged by the point of sale system. The first section of results examines whether point of sale systems work against counties with less prosperous residents. In other words, we are less interested in whether LOST systems disadvantage “retail-poor” communities than whether they work against communities that are broadly less affluent. This section shows that the point of sale system tends to disadvantage counties that are more rural, have populations with lower income, and higher poverty rates. The next section presents evidence that LOST systems tend to undercut local governments that are under greater fiscal stress. We show that counties who are disadvantaged by the point of sale system have less cash on hand and have lower average credit scores than counties who prosper under the point of sale system. The next section tests the central theoretical argument outlined above. Are retail tax dollars literally moving from poorer communities to more affluent ones when sales tax revenues are realized on a point of sale basis? More wealth obviously generates more sales taxes but, if the theory outlined above is correct, LOST revenues should be distributed even more unevenly than absolute wealth. We find that out-commuting rates have a significant impact on sales tax revenue, even controlling for county income. These results provide direct evidence of the core theory driving this project. When county residents drive elsewhere to work, their home county loses sales tax revenue. When wealthier counties attract retail business, they have even more to spend on their citizens. The final section of results reports the ultimate consequences of distributing sales tax dollars at the point of sale versus allocating them according to county population. We show that moving from a point of sale to a per capita system would help the governments that are disadvantaged by the point of sale system more than it would hurt counties that prosper under point of sale. Finally, we show that this policy choice (per capita vs. point of sale) has a substantial aggregate THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 471 impact on where revenues are realized. Over a decade, the differences between point of sale and per capita systems can run into the billions. METHODS Why Study North Carolina? North Carolina's sales tax system is actually unlike most traditional LOST arrangements. Where other states give counties the option of levying a sales tax on top of the statewide rate, North Carolina's system was established and is run entirely at the state level. All sales taxes are collected by the state government and then part of those revenues are distributed back to the county governments. North Carolina does not have the traditional local option sales tax, but instead a statewide tax with revenue sharing built into the system. While North Carolina's system is not true to the original LOST idea, it is an advantageous case to study. At the most basic level, North Carolina's system ensures that differences in local sales-tax revenues are not due to differences in how the tax is administered. The choices of which goods to tax, how those revenues are collected, and how they are reported do not change from one county to the next.1 At a deeper level, North Carolina's system ensures that differences in retail activity between counties are not driven by differences between local tax structures. Previous research has found that differences in local sales tax rate affect the location of retail outlets (Fisher, 1980). In more traditional systems, some counties exercise the local sales tax option before others (Sjoquist et al., 2007), which may influence where retailers choose to set up shop. In addition, some jurisdictions may not enact the full allowable tax rate, again producing incentives for retailers to go where taxation is limited. Beyond influencing where retail outlets open, differences in sales tax rate between adjacent jurisdictions (e.g. neighboring states) have been shown to influence where consumers do their shopping (Fox, 1986; Walsh & Jones, 1988). Since the focus of this paper is on how market forces influence LOST revenues, and not on how LOST systems change market decisions, North Carolina's system removes a potential source of endogeneity from our analysis. In a purely local option system, some 472 MCHUGH & JOLLEY counties adopt the levy years after it is first instituted by other localities. During the period when some counties collect LOST revenues and others do not, we would expect retailers to factor these differences into where they choose to do business. As the market can be skewed by the sequence of adoption, North Carolina is an appealing case because these factors have not influenced where retail outlets are located. North Carolina also provides an example of how states can adjust for inequalities produced by point of sale systems without going to a purely per-capita alternative. Sales taxes revenue that is distributed to North Carolina counties is broken up into four independent articles enacted by the state legislature. Article 39 creates a 1% sales-tax where the revenues are distributed on a point of sale basis. Articles 40 and 42 both create ½% sales taxes that are distributed according to county population. Article 44 creates a final ½% sales tax with half of these revenues being distributed on a per-capita basis and the other half staying where sales occur. Articles 39, 40, and 42 apply to food purchases of home consumption, while Article 44 does not. Calculating LOST Revenues Distributed on Per-Capita vs. Point-of-Sale Bases The core of this paper is driven by a comparison between the revenues that each county would realize under a purely point of sale system, like those that exist in most states, and what each county would receive if sales-tax revenues were distributed on a purely percapita basis. Because each article of the North Carolina sales tax is reported independently, we can isolate the portion that is distributed on a purely point of sale basis. When we first began this project, the idea was to compare the revenues counties receive under the two ideal components of North Carolina's system (point of sale vs. per capita). It turns out that, at least in North Carolina, per-capita does not really mean per-capita. North Carolina law applies an “appropriate adjustment factor” to each county's population-based share of sales tax revenue, increasing or decreasing their actual allotment. These factors range from .81 (county would receive 81% of the share allotted to their population) to 1.49 (would receive 149% of per-capita share). To date, we have not found any specific explanation of how these factors were produced. While the “per-capita” component of the North Carolina tax is more THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 473 equitably distributed than the portion based on point of sale receipts, these adjustment factors create inequalities that run in the same direction. There is actually a positive correlation between county wealth and its “adjustment factor," meaning that wealthier counties are collecting more per-capita than are poorer counties. As a consequence, we could not use the “per-capita” portion of North Carolina's system to estimate what counties would receive if the entire system were actually distributing sales tax revenues according to population. Therefore, we use the purely point of sale component of North Carolina's system (Article 39), to calculate what each county would receive under an actually per-capita system. For each year, we summed the total revenues that all counties received from Article 39 and divided that by the statewide population. This produces the statewide average sales tax revenue generated by each resident of North Carolina. We then multiply this figure by each county's population to find the share of the statewide sales tax pot that each county would receive. With this in hand, we compare a county's estimated per-capita share to what they actually received under the point of sale system. Article 39 only constitutes 1% of the 2.5% of sales taxes that are distributed to county governments, so the real receipts and our estimate of per-capita share were both multiplied by 2.5 to produce an estimate of what each county government would receive if the entire system were run on a point of sale or per-capita basis. In the final section, we estimate the change in each county’s overall budget that is produced by moving from a point of sale to a per capita system under 1%, 2.5% and 5% sales tax rates. This approach accomplishes a few critical things. First, it allows us to separate who wins from who loses under these alternative systems. We know what counties actually received under a point of sale system, and now we can see whether they would have fared better or worse if these revenues were distributed based on population. This allows us to directly test whether the counties that suffer under a point of sale system really are poorer, more rural, and have more struggling populations. Second, we can directly estimate what percentage of the overall budget picture is at play. We can estimate how much each county would lose or gain if North Carolina moved from a purely point of sale system to a per-capita one. This allows us to test the claim that point of sale LOST systems cost 474 MCHUGH & JOLLEY poorer counties a larger percentage of their overall budget than wealthy counties stand to gain. RESULTS Point-of-Sale Systems Disadvantage Poor and Rural Communities Before anything else, it is important to look carefully at the profile of counties that are advantaged by a point of sale system compared to those that are disadvantaged by sales tax revenue being realized at the point of sale. Is it true that traditional LOST systems work against communities that are already under greater fiscal duress? This section presents evidence that counties who suffer under a point of sale LOST system are struggling rural communities. The next section shows that the same disadvantaged communities have fewer alternative sources to draw upon and are in greater fiscal distress. Looking at county finances from both ends, it is clear that point of sale systems hurt communities with greater needs and fewer resources to serve those demands. We argued at the outset that point of sale LOST systems disadvantage less affluent, often rural, counties because their residents' must travel elsewhere to work and shop. Table 1 supports the concern at the heart of this project. The first column reports the average population, median household income, and poverty rates for counties that are advantaged by a point of sale system (counties that collect more revenue on a point of sale basis than they would on a per-capita distribution). The second column shows the same averages for counties that disadvantaged by a point of sale allocation. These data were taken 2008 county revenue summaries provided by the North Carolina Treasurer (http://www.nctreasurer.com/lgc/units/ unitlistjs.htm). North Carolina has 100 counties, but 3 of these did not supply all of their data for 2008, so Table 1 reflects the 97 counties that did report. First off, it is clear that rural communities tend to be disadvantaged by the point of sale system. On average, counties that are disadvantaged when LOST revenues are based on where sales take place are significantly smaller than counties that prosper under a point of sale system. We suspect that a LOST system would also disadvantage purely inner-city counties, but North Carolina does not THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 475 TABLE 1 Demographic Differences between Counties Advantaged and Disadvantaged by the Point of Sale System (2008) Counties Counties Advantaged Disadvantaged by Point by Point of Sale of Sale Population std err Med House Inc Poverty (all) Poverty (under 18) Number of Counties 131,737 39,854 $45,071 1,595 14.1% 0.69 20.4% 0.90 22 80,422 13,120 41,050** 863 17.6%** 0.55 24.8%** 0.80 75 Note: Counties Advantaged by Point-of-Sale are counties that collect more point of sale revenues than they would on a per-capita basis; Counties Disadvantaged by Point-of-Sale collect less point of sale revenues than they would on a per-capita basis. ** Significant difference (.01 level) between counties advantaged by point of sale sales tax allocation and those disadvantaged. The differences reported as based on the entire population of North Carolina counties (excepting three missing cases), so significance levels are not necessary for discussing North Carolina's system. However, since we are interested in the theoretical behavior of LOST systems generally, we report significance levels for differences of means tests conducted on these data, effectively treating North Carolina as a sample of the population of all county governments. Source: Revenue and population data (North Carolina Treasurer 2011). Accessed at http://www.nctreasurer.com/lgc/units/unitlistjs.htm. Poverty rates (U.S. Census Bureau 2010). (Accessed at http://www.census.gov/cgi-bin/saipe/saipe.cgi). have any counties that match that profile. What we see here is the first evidence that residents of rural counties must travel to shop and, in so doing, leave their tax revenues behind in more populous areas. The claim that point of sale tax revenues flow from less prosperous counties to more wealthy ones will be more explicitly tested later, but these data do support the concern that LOST systems work against rural communities with limited retail activity. While it is important to know that point of sale systems work against rural counties, there would be less cause for alarm if those areas were prospering otherwise. However, many rural communities 476 MCHUGH & JOLLEY in North Carolina and throughout the county are struggling. If our central claim is correct, and LOST systems reinforce economic inequalities between local communities, average income should be higher in counties that are advantaged by a point of sale revenue allocation and poverty rates should be lower, which is precisely what the data in Table 1 show. TABLE 2 Per Capita Revenues for Counties Advantaged and Disadvantaged by the Point of Sale System (2008) Counties Counties Advantaged Disadvantaged by Point by Point of Sale of Sale Property Taxes std err Service Taxes Other Taxes Inter-Gov't Grants Misc Revenue All Non Sales Tax Revenue Number of Counties $727 $46 $181 $33 $114 $28 $332 $75 $103 $13 $1,722 $507** $13 $168** $13 $45** $3 $255** $10 $68** $4 $1192** $173 $41 22 75 Note: Counties Advantaged by Point of Sale are counties that collect more point of sale revenues than they would on a per-capita basis; Counties Disadvantaged by Point of Sale collect less point of sale revenues than they would on a per-capita basis. ** Significant difference (.01 level) between counties advantaged by point of sale sales tax allocation and those disadvantaged. Source: North Carolina Treasurer (2011); U.S. Census Bureau (2010). Point-of-Sale Systems Disadvantage Counties that are Under Greater Fiscal Stress Having seen that a point of sale revenue system tends to work against rural communities with lower income and more poverty, this THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 477 section examines whether these county governments are under greater fiscal stress. This section first reviews revenue streams generated by everything other than sales taxes, showing that counties who are disadvantaged by LOST collect significantly less revenue per resident from other funding sources than do counties who are advantaged by a point of sale system. Second, this section shows that counties who are disadvantaged by a point of sale system also have less cash on hand and have lower average credit ratings, both good indications that they are under greater fiscal stress. Table 2 shows that LOST systems work against counties that often face a more dismal revenue situation than the counties who are advantaged by the system. In 2008, North Carolina counties that did better when sales tax revenues were distributed on a point of sale basis also collected more per resident in property taxes, more in other taxes, more in service taxes, received more in intergovernmental grants, and collected more miscellaneous revenue. The final row of Table 2 reports the average revenue per resident derived from everything but sales taxes. In 2008, North Carolina counties that were advantaged by the point of sale system collected, on average, $530 more in revenue per resident than counties that were disadvantaged by the point of sale system. If we treat North Carolina counties as a sample, the difference in per-capita revenue between counties advantaged by a point of sale system and those disadvantaged by it is statistically significant at the .01 level. Showing raw differences in revenue does not, in itself, prove that LOST systems work against communities who are under greater fiscal stress. Given that we have already seen that point of sale systems work against rural communities, differences in county revenue could simply be driven by differences in the cost of living. Rural governments may need less revenue per citizen because the cost of providing services and infrastructure is lower. To account for this possibility, we examine two indicators of fiscal stress (bond ratings and fund balances) to test whether counties who are disadvantaged by point of sale systems are under greater fiscal stress than counties who are advantaged by how LOST systems generally function. Fund balances are one traditional measure of fiscal stress (Gold, 1986; Hendrick, 2004). Based on North Carolina Treasurer county revenue summaries, we calculated the fund balance available per resident for each North Carolina county in 2008. The first row of 478 MCHUGH & JOLLEY Table 3 shows that counties who were advantaged by a point of sale system had roughly twice as much cash on hand than did disadvantaged counties, a difference that almost certainly outstrips differences in cost of living and providing governmental services. TABLE 3 Per Capita Revenues for Counties Advantaged and Disadvantaged by the Point of Sale System (2008) Per-Capita Fund Balance Standard Error Average S&P Rating Obs for Fund Balance Obs for S&P Rating Counties Advantaged by Point of Sale Counties Disadvantaged by Point of Sale $407 $217** $51 5.94 0.44 22 16 $12 3.77** 0.25 75 49 Note: Counties Advantaged by Point of Sale are counties that collect more point of sale revenues than they would on a per-capita basis; Counties Disadvantaged by Point of Sale collect less point of sale revenues than they would on a per-capita basis. ** Significant difference (.01 level) between counties advantaged by point of sale sales tax allocation and those disadvantaged. Source: North Carolina Treasurer (2011). Bond ratings provide another test of whether point of sale LOST systems work against counties that are under greater fiscal stress, Bond ratings and changes in bond ratings are established measures of fiscal stress (Jesse Marquette, R Marquette, and Hinckle 1982; Grizzle 2010). Governmental bond ratings are designed to capture a government's ability to repay its debt, so counties that are on a solid fiscal footing receive higher bond ratings than do counties who are closer to insolvency (for a description of S&P bond rating methodology see http://www.standardandpoors.com/ratings/). The lowest rating reported for North Carolina counties in these data was BBB+ and the highest was AAA. We coded S&P ratings numerically in the following way: BBB+=1, A-=2, A=3, A+=4, AA-=5, AA=6, AA+=7, AAA-=8, AAA=9. Another coding was conducted, where +'s and -'s THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 479 were dropped and each major rating was placed on a scale from BBB=1 to AAA=4, which produced very similar results. The second row of Table 3 compares the average bond rating of counties advantaged and disadvantaged by a point of sale LOST system. Based on our coding, counties disadvantaged by a point of sale system have an average bond rating between A and A+, while counties advantaged by the system have an average rating AA, a difference of almost one major S&P rating. According to Standard and Poor's, an A rating mean “strong capacity to meet financial commitments, but somewhat susceptible to adverse economic conditions and changes in circumstances”. An AA rating means “Very strong capacity to meet financial commitments” (Standard and Poor's. Retrieved from http://www.standardandpoors.com/ratings/). Clearly, based on the best models of fiscal strength available, counties that are disadvantaged by a point of sale LOST system tend to be on more tenuous financial footing. The difference is not enormous but, given the variety of factors that feed into bond ratings, it is telling that differences in sales-tax revenue are so clearly correlated with measurable fiscal stress. It is worth noting that this result is not unique to the recession that set in during 2008; we ran this analysis on all years from 2003 to 2009 independently and the results were extremely similar. All told, there is substantial evidence that a point of sale LOST system works against counties that are harder pressed to meet their residents' needs. Counties that are disadvantaged by a point of sale system collect less revenue from other sources, have less cash on hand, and have lower bond ratings. Do LOST Revenues Move with the Flow of Traffic? To this point, we have focused on implications of the thesis behind this paper, but we have not explicitly tested the mechanism at the core of the theory. We argue that travel patterns reinforce revenue differences produced by resident income alone. As retail activity gravitates toward centers of spending power, residents of poorer communities do more of their working and shopping away from home. This section directly tests whether sales tax revenue literally moves with the flow of traffic when revenues are realized at the point of sale. 480 MCHUGH & JOLLEY TABLE 4 Predicting Per-Resident Revenue Each County Would Receive under 2.5% Point of Sale Tax Med House Inc ($1,000's) (Standard Error) Predictor Coefficients 4.51** (.163) 5.04** (1.53) -2.67** (.55) 18.08** (1.99) 30.95** (3.38) 16.38** (5.69) 13.08 (8.81) 8.92 (10.50) 67.46 (46.50) .41 Outcommuting Rate 2004 dummy* 2005 dummy 2006 dummy 2007 dummy 2008 dummy Constant r2 18.66** (2.08) 31.92** (3.44) 18.20** (5.99) 15.63 (9.29) 12.9 (11.06) -7.66 (53.42) 0.14 Note: Pooled linear regression model where the unit of analysis is the county budgetary year, with each year from 2003 to 2008 counted as unique observations. Standard errors are clustered on the county to account for correlations in each county's errors over years. Dependent variable is the tax revenue generated for each county under a 2.5% point of sale tax (Article 39*2.5), divided by the county's population in that year. Median houshold income and sales tax revenue taken from county summaries provided by the North Carolina Treasurer. Outcommuting rates for each county taken from 2000 U.S. Census (most recent year available). Sources: North Carolina Treasurer (2011); U.S. Census Bureau (2000). Table 4 reports the results of a pooled linear regression model predicting the revenues that each county would receive if North Carolina's sales tax system were distributed on a purely point of sale basis. The model includes data for each county from 2003 to 2008 (except where data was not reported in the NC Treasurer files), which are pooled. The dependent variable is the reported revenues from Article 39, multiplied by 2.5, and then divided by the county's THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 481 population. Median income data is taken from the US Census Small Income and Poverty Estimates (http://www.census.gov/cgibin/saipe/saipe.cgi). Outcommuting rates are taken from the 2000 census. Unfortunately, county outcommuting rates are not available for each year, so we use the 2000 estimates for each year of data.2 Because we are using pooled data, year dummies are included to account for differences in average sales tax revenues from year to year (2003 is the omitted category). In addition, because there are likely to be factors unique to each county's sales tax revenues, we cluster standard errors to account for any correlated errors for individual counties over years. The first model only uses median household income to predict the revenues each county would receive from point of sale taxes. As we would expect, counties with more disposable wealth also generate more sales tax revenues. The second model adds a county's outcommuting rate as a predictor of point of sale tax revenues. If our theory is correct, tax dollars should be moving with the flow of traffic. While outcommuting rates is an imperfect proxy for inter-county shopping, these data should capture travel patterns that influence shopping patterns. As predicted, outcommuting is negatively associated with sales tax revenues. When more of a county's residents work outside the county, and presumably do shopping while away from home, their home county collects less sales tax revenue per citizen. Critically, this effect holds even controlling for average income. This means that counties with high outcommuting rate collect less point of sale tax revenues than their income would predict and counties with low outcommuting rates collect more than they should, based on wealth alone. On top of this, the second model is a much better fit to the data (r2=.41 vs r2=.14). Even given that the addition of an explanatory variable necessarily inflates r2, the addition of outcommuting rates explains significantly more variance than the base model. If we want to understand why some counties collect more point of sale taxes per citizen than others, we must consider dominant travel patterns that influence where people do their shopping. 482 MCHUGH & JOLLEY Examining the Proportional and Aggregate Effect of Point of Sale versus Per Capita Systems This final section outlines what is at stake. Previous studies have shown that local option sales tax revenues are biased toward counties with more retail options. While one study did estimate the total leakage of revenues from “retail-poor” to “retail-rich" counties over a ten year span (Artz & Stone, 2003), no study has directly estimated the percentage change in each county's budget. Thus, we ask, how would county budgets change if North Carolina went from a purely point of sale system to one where proceeds are distributed according to population? We have seen that, on average, counties that are disadvantaged by a point of sale system are poorer, more rural, and generate less revenue through alternative sources, and have lower credit ratings. As a consequence, we hypothesize that the movements of revenue found in the previous section have the most impact on counties that lose retail activity. In other words, we expect LOST systems to reduce poor counties' budgets by a greater extent than they increase the budgets of wealthy communities. When residents of poorer counties drive to neighboring communities to shop, the lost revenues comprise a substantial portion of their home county's operating budget because these counties have less property tax valuation, lower incomes, and less economic activity to generate alternative sources of revenue. On the other hand, counties that attract retail business tend to already be much better off; that's why the retailers are keen to set up shop in those communities in the first place. While some counties would not do as well under a per-capita system as they do under a point of sale system, these more affluent counties have less at stake than do counties that are disadvantaged by the point of sale system. To test this hypothesis, we estimated the percentage change in each county's total budget that would result from exchanging a system where sales tax revenues stay where sales take place (point of sale) to one where revenues are distributed according to county population (per-capita). The equation used to accomplish this for county i in year j is given by: THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX ∆ 483 Where is the total estimated county budget if all sales tax revenues were distributed according to county population and is the estimated county budget under a purely point of is the percent change county ′ total sale system. Thus, ∆ budget, in year (positive or negative), that would be produced by moving from a point of sale to a per-capita system. Abandoning the local option sales tax, and replacing it with one where revenues are distributed according to population, would help struggling counties more than it would hurt their more affluent neighbors. We already have seen that point of sale systems penalize counties that can least afford it, and this exercise indicates that poorer counties lose proportionally more than the wealthier counties gain. Table 5 reports the average change in budget for counties that are advantaged by a point of sale system versus those that are disadvantaged. As local option sales tax rate varies across states, we provide three estimates depending on the allowable local tax rate. The first line reports the average gains and losses attached to changing from a point of sale to a per-capita system if the local rate is 1%. The second line reports the same estimates for states, like North TABLE 5 Average Impact of Exchanging the Point of Sale System for a PerCapita System Counties Counties Advantaged Disadvantaged by Point by Point of Sale of Sale Local Sales Tax Rate 1% 2.50% 5% Obs -1.73% -3.81% -6.43 179 +2.49% +5.84% +10.51 505 Note: Mean change in estimated county budgets that would result from moving to a per-capita allocation of sales taxes. Unit of analysis is the individual county budgetary year. Source: North Carolina Treasurer (2011). 484 MCHUGH & JOLLEY Carolina, with a 2.5% local sales tax. The final line reports the change that would occur in states where the local option tax rate is at 5%, as it is in Louisiana, which is the highest rate allowed by any state government. Even under a 1% LOST tax, counties that are disadvantaged by the point of sale system would see their budgets grow by 2.49%, while counties that benefit from the point of sale system would see their revenues shrink by 1.73%. Under a 2.5% local sales tax rate, counties that are disadvantaged by the point of sale system would see an average budgetary increase 5.84% while counties who do better under a point of sale system would lose 3.81% in revenue. In states that permit localities to tax sales up to 5%, shifting to a statewide per-capita system would increase budgets in counties that are disadvantaged under the existing system by an average of 10.51%, while counties that prosper under the current arrangement would see their revenues fall by roughly 6.43%. This is direct evidence that traditional LOST systems penalize the counties that can least afford it. When sales tax revenues flow from struggling rural counties to their more prosperous neighbors, the losses are felt more heavily than the corresponding gains. The coffers of wealthy counties that attract retail business are already being filled from a variety of larger revenue streams, while the revenue lost by rural counties is being taken out of much smaller sources. Finally, it is useful to useful to understand the total amount of revenue that hangs in the balance depending on whether sales tax revenues stay at the point of sale or are distributed according to population. The central point of this exercise is to demonstrate that we are talking about substantial amounts of revenue for local governments, which is what Table 6 shows. For example, if North Carolina had shifted from a 2.5% point of sale system to a 2.5% tax allocated according to county population, counties that would have gained from such a switch would have received an aggregate of $296,116,839 extra. Over the six year span captured in our data, the total revenue that would depend on per capita versus point of sale, for a 5% sales tax, would stack up to over $3 billion. All in all, it is clear that policy choices about sales tax revenue allocation have a substantial impact on local budgets. THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 485 TABLE 6 Total Revenue to Be Gained by Counties that Are Disadvantaged by the Point of Sale System if Sales Tax Revenues were Allocated According to Population Year 1% Tax 2003 2008 2003-2008 $89,663,472 $118,446,735 $633,183,285 2.5% Tax 5% Tax $224,158,681 $448,317,362 $269,116,839 $592,233,678 $1,582,958,213 $3,165,916,426 Note: Values reflect the total change in revenues depending on how sales taxes are allocated. For a shift to the per capita system, away from the point of sale system, the values represent the total revenues gained by counties who are disadvantaged by the point of sale system. Source: North Carolina Treasurer (2011). DISCUSSION There is growing evidence that local option sales taxes are a bad deal for impoverished communities with struggling local governments. Previous research has shown that LOST taxes are distributed more unequally than property taxes (Zhao & Hou, 2008), that they tend to flow from “retail poor” to "retail rich” counties (Artz and Stone 2003), and that over time these transfers added up to substantial sums of money. This paper reinforces all of the previous findings about local sales taxes and rounds out our understanding of the policy in a number of important ways. First, we show that local option sales taxes undermine the revenues of counties with struggling populations. Counties that are disadvantaged by a point of sale system tend to have fewer residents, lower average family incomes, and higher rates of poverty. We also showed that counties who are disadvantaged by a point of sale system generate less revenue per resident, have less cash on hand, and have lower credit ratings, all indications that these counties are under greater fiscal stress. Second, we provide direct evidence that variation in LOST revenue cannot be explained by differences in wealth alone. Counties where more residents work outside of the county collect less point of sale revenue than they we would expect, based on income and counties where fewer residents commute, collect more than they should, based on income. This is direct evidence that when counties 486 MCHUGH & JOLLEY attract business from their neighbors, they capture their neighbors' sales tax revenues as well. The final section demonstrates that local option sales taxes have a substantial impact on counties' overall budgets. Previous studies have shown that LOST revenues reinforce other inequalities in local funding, but do not provide a concrete picture of how much poor counties are losing and how much wealthy counties are gaining. One of the purposes of this study was to test our suspicion that proportional gains by counties that capture retail from their neighbors were smaller than the proportional losses to communities that leak retail. The evidence clearly shows that abandoning LOST systems and replacing them with systems that distribute sales tax revenue according to county population would help struggling communities more than it would hurt their more affluent neighbors. NOTES 1. There is a multiplicity of local option sales taxes that generate both general and specific-purpose revenues. Some localities exempt food purchases, housing expenditures, or transportation spending from their local sales taxes. 2. We expect that time mismatches between outcommuting rate and point of sale tax revenues should work to mute the relationship, so any error should be conservative. To be even more confident that this mismatch is not driving our results, we ran the model using only 2003 data, the closest in time to the outcommuting reading within these data, and the results were extremely similar. REFERENCES Artz, G. M., & Stone, K. E. (2003). “An Analysis of the Transfer of Funds from Weak Retail Counties to Strong Retail Counties in Iowa via Local Option Sales Taxes.” Presented at the 2003 Annual meeting, July 27-30, Montreal, Canada. Chervin, S., Edmiston, K., & Murray, M. N. (2000). “Urban Malls, Tax Base Migration, and State Intergovernmental Aid.” Public Finance Review, 28: 309-334. THE SHERIFF OF NOTTINGHAM'S FAVORITE TAX 487 Craft, M. M. (2002). “Lost and Found: The Unequal Distribution of Local Option Sales Tax Revenue among Iowa Schools.” Iowa Law Review, 88:199–1243. Fisher, R. C. (1980). “Local Sales Taxes: Tax Rate Differentials, Sales Loss, and Revenue Estimation.” Public Finance Review, 8: 171188. Fletcher, J. M., & Murray, M. N. (2006). “Competition over the Tax Base in the State Sales Tax.” Public Finance Review, 34: 258281. Fox, W. F. (1986). “Tax Structure and the Location of Economic Activity along State Borders.” National Tax Journal, 39: 387–401. Ghazanfar, S. M. (1978). “Sales Tax Equity Again: By Age Groups and Income Classes.” Public Finance Review, 6: 343-357. Gold, S. (1986). State Government Fund Balances, Financial Assets and Measures of Budget Surplus: Federal-State-Local Fiscal Relations. Washington, DC: National Council of State Legislator. Grizzle, C. (2010). “The Impact of Budget Stabilization Funds on State General Obligation Bond Ratings.” Public Budgeting & Finance, 30: 95-111. Hendrick, R. (2004). “Assessing and Measuring the Fiscal Health of Local Governments: Focus on Chicago Suburban Municipalities.” Urban Affairs Review, 40: 78-114. Marquette, Jesse, R Marquette, and Katherine Hinckle. 1982. “Bond Rating Changes and Urban Fiscal Stress: Linkage and Prediction.” Journal of Urban Affairs, 4: 81-95. North Carolina Treasurer (2011). North Carolina County and Municipal Financial Information. www.nctreasurer.com/lgc/ units/unitlistjs.htm. (Accessed February 6, 2011). Rostvold, G. N. (1967). “Distribution of Property, Retail Sales, and Personal Income Tax Burdens in California: An Empirical Analysis of Inequity in Taxation.” National Tax Journal, 20(1): 112-113. Rubenstein, R., & Freeman, C. (2003). “Do Local Sales Taxes for Education Increase Inequities? The Case of Georgia.” Journal of Education Finance, 28: 425–442. 488 MCHUGH & JOLLEY Schweitzer, L., & Taylor, B. D. (2008). “Just Pricing: The Distributional Effects of Congestion Pricing and Sales Taxes.” Transportation, 35: 797–812. Shoup, C. (1983). “The Property Tax versus Sales and Income Taxes.” Proceedings of the Academy of Political Science, 35 (1): 31–41. Siegfried, J. J., & Smith, P. A. (1997). “The Distributional Effects of a Sales Tax on Services.” In D. Netzer, & M. P. Drennan (Eds.), Readings in State & Local Public Finance (pp. 211-233). Cambridge, MA: Blackwell Publishers, Inc. Sjoquist, D. L, Smith, W. J., Walker, M. B., & Wallace, S. (2007). “An Analysis of the Time to Adoption of Local Sales Taxes: A Duration Model Approach.” Public Budgeting & Finance, 27 (1): 20–40. Sjoquist, D. L, Walker, M. B., & Wallace, S. (2005). “Estimating Differential Responses to Local Fiscal Conditions: A Mixture Model Analysis.” Public Finance Review, 33: 36-61. U.S. Census Bureau. 2000. 2000 U.S. Census. U.S. Census Bureau (2010). Small Area Income and Poverty Estimates. Washington, DC: Author. Walsh, M. J., & Jones, J. D. (1988). “More Evidence on the ‘Border Tax’ Effect: The Case of West Virginia, 1979-84.” National Tax Journal 41: 261–265. Zhao, Z. J., & Hou, Y. (2008). “Local Option Sales Taxes and Fiscal Disparity: The Case of Georgia Counties.” Public Budgeting & Finance, 28 (1): 39–57. Zhao, Z. J., & Jung, C. (2008). “Does Earmarked Revenue Provide Property Tax Relief? Long-Term Budgetary Effects of Georgia's Local Option Sales Tax.” Public Budgeting & Finance, 28 (1): 52– 70. J. OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT, 24 (3), 489-516 FALL 2012 DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS STRUCTURE Martin J. Luby* ABSTRACT. In recent years state governments have significantly increased the use of debt to fund their capital and operating budgets. As the size of government bond portfolios increased, the opportunities to refinance debt to realize substantial operating budget savings grew. To relieve short-term fiscal stress, some state governments have utilized debt refinancing structures that provide near term benefits with the same negative long-term budgetary impacts as other common “one-shot revenue” budget gimmicks. This paper thoroughly describes the strategy and mechanics of one such debt-related budget gimmick, the upfront bond refinancing savings structure, and details its usage by state governments over the last decade. Through descriptive research, this paper finds that states have only very modestly used this budget gimmick, which has theoretical and policy implications relating to principal-agent theory and the efficacy/necessity of debt restrictions, respectively. INTRODUCTION State governments across the country continue to struggle with sizable cyclical and structural budget deficits even as the Great Recession was “officially” declared over in June 2009. In the current 2012 fiscal year that started on July 1, estimates show that 44 states and the District of Columbia will report operating budget deficits totaling $112 billion (Center on Budget and Policy Priorities, 2011). This follows annual budget deficits of $110, $191 and $130 billion in fiscal years 2009, 2010 and 2011 (estimated), respectively. Moreover, state finances are only expected to slowly recover over the ------------------------* Martin J. Luby, Ph.D., is an assistant professor in the School of Public Service at DePaul University. His teaching and research interests are in state and local government capital markets, public financial management, and organizational behavior. Copyright © 2012 by PrAcademics Press 490 LUBY next few years as evidenced by a preliminary aggregate state budget deficit of $75 billion in fiscal year 2013 (Center on Budget and Policy Priorities, 2011). While the upcoming year constitutes one of the most fiscally challenging budget years on record, states have been regularly struggling with their finances for many years even before the Great Recession. For example, as a result of the last recession in 2001, state budget deficits totaled $40, $75, $80 and $45 billion in fiscal years 2002, 2003, 2004 and 2005, respectively. Unlike the federal government which has the ability to cover deficits with borrowing, states are constitutionally or statutorily restricted from running operating budget deficits. While the strictness of this restriction varies by state, states generally aim for balance, at least “on paper” (Briffault, 1996). The traditional measures used to address these budget gaps reside on both the revenue and expense sides of the operating budget and include such measures as implementing new or raising existing taxes, creating new or increasing existing user fees or other revenues, cutting program expenditures, reducing service levels, and consolidating programs and departments. Such measures were used extensively in the last two recessions. For example, during the Great Recession, 46 states reduced services to its residents and 30 states increased taxes with some of these increases quite substantial (Center on Budget and Policy Priorities, 2011). However, it has also been well documented that states have relied on other less traditional, and many argue more fiscally reckless, measures to achieve budgetary balance (Briffault, 1996; Petersen, 2003; Mikesell 2009). A common one of these measures involves the refinancing or restructuring of a government’s outstanding debt. The focus of this study is on one specific type of debt-related budget gimmick not extensively explored in the literature, the upfront bond refinancing savings structure. This budget gimmick involves a government structuring its bond refinancing so that the interest cost savings occur in the first year or in the first couple years after the refinancing with all subsequent years receiving minimal or zero savings. By taking all the savings upfront, the government is, in effect, employing a “one-shot revenue” to lessen the fiscal strain on its operating budget. The use of one- DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 491 shot revenues has generally been discouraged by public finance experts as it can lead to or exacerbate structural budget deficits. In order to shed greater light on this budgetary technique, this paper specifically describes the technique, details its usage by state governments and describes the implications of such usage level for public financial management theory (agency theory) and practice (efficacy of debt restrictions). More specifically, the paper proceeds as follows. First, the paper details how this analysis fits into the broader public finance literature on budget gimmickry and public administration literature using principal-agent theory to understand management behavior. Second, the paper details the rising importance of debt finance in recent decades and how this has provided an opportunity to use bond refinancing to help achieve short-term budgetary balance. This section will also provide a thorough description of the budget gimmick technique under study, the upfront bond refinancing structure. Third, the paper describes the methodology used in calculating the overall usage of upfront bond refinancing structures and then proceeds to detail the actual utilization level of this budget gimmick by state governments over the last decade. The final section describes the implications of such usage level for theory and practice. LITERATURE REVIEW This paper contributes to two different strands of applied empirical literature and one body of theoretical literature. First, this paper adds to the empirical budget gimmickry literature. This literature has generally been relatively descriptive in nature detailing the various budget gimmicks that subnational governments employ to achieve budgetary balance. Mikesell (2009) describes several budgetary gimmicks often employed by subnational governments including overestimation of revenues, asset sales, capitalization of unrecognized assets, interfund manipulations and borrowing, acceleration of revenues and delays in spending, capitalization of current costs and borrowing to cover them, and anticipated future savings. Briffault (1996) concurs with Mikesell and outlines a similar list of budgetary gimmicks in the context of the desirability of a federal balanced budget rule including the use of overly optimistic revenue forecasts, manipulation of fiscal periods, payment deferrals, pension fund reductions based on overly optimistic return yields, 492 LUBY asset sales, special fund transfers, and bond refinancing. Petersen (2003) focuses on what he calls “more exotic, interesting and more damaging budget gimmick methods” including sophisticated debt refinancings that amount to the “liquidation of an unrecognized asset.” The budgetary gimmick that is the focus of this paper, upfront refinancing savings structure, is included in the list of debt-related budget gimmicks briefly detailed in Petersen (2003). Second, this paper attempts to contribute to the empirical literature on bond refinancing of municipal debt, which, to date, is not very robust. Most of this literature has been prescriptive in nature detailing the optimal timing and structure of subnational bond refinancings (Dyl and Joehnk 1976; Babad and Speer 1978; Kalotay and May 1998; Kalotay, Yang and Fabozzi 2007; Zhang and Li 2004). There have also been a couple studies on the impact of bond refinancings and some of the factors associated with the refinancing decision (Luby 2011; Vijayakumar 1995). This paper details the actual technique and budgetary impact of a specific bond refinancing strategy. Third, this paper will also aim to add to the public management theoretical literature that utilizes economic models and, more specifically, agency theory as a means of understanding public financial management decision-making. There are several research studies that have at least tangentially utilized agency theory to explain public financial management decision-making as it relates to the use of financial advisors, bond method of sale, debt policies, and debt management networks (Bhagat and Frost 1986; Miller 1993; Leonard 1996; Simonsen and Kittredge 1997; Simonsen and Robbins 1996; Robbins and Dungan 2001; Peng and Brucato 2003; and Kriz 2003). A 2006 symposium in the Journal of Public Budgeting, Accounting and Financial Management included studies that seemed to refute the use of rational choice theories such as principal-agent theory in explaining public financial management behavior. The symposium presented four studies that offered alternative theoretical explanations on the connection between individual motives, organizational structures and financial results whereby public financial managers were found to be "satisficers" not entirely influenced by individual self-interest but by other factors including politics, professional norms and individual values (Carroll DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 493 and Marlowe 2006; Burke 2006; Rivenbark 2006; Robbins and Miller 2006; Stalebrink and Sacco 2006). Simonsen and Hill (1998) is the most prominent research study to date that has attempted to explicitly use agency theory to understand public sector debt management decisions. Their study utilizes a survey of city and county government finance officers in trying to determine whether there is evidence of a principal-agent problem in the issuance of municipal bonds. Based on the results of their survey instrument, Simonsen and Hill concluded that there is evidence of a significant principal-agent problem (Simonsen and Hill, 1998). This paper attempts to extend Simonsen and Hill’s study in drawing inferences from debt management decisions, not from survey responses. That is, rather than relying on survey research in determining whether agency problems plague debt management decisions, this paper looks at the actual behavior of public sector financial managers (i.e., whether to use an upfront refinancing savings structure) to observe evidence of the existence of a principalagent problem. BOND REFINANCING SAVINGS STRUCTURE DESCRIPTION The first part of this paper’s analysis entails describing the budgetary gimmick under study, the upfront bond refinancing savings structure. However, to provide broader context to the research study at hand, it is first important to briefly describe some recent subnational government bond finance trends in order to establish saliency of the issue. In recent years, state and local governments have significantly expanded their debt finance activities as a means of increasing and optimizing their capital and, sometimes, operating budget funding. In general, however, state and local government debt is mainly employed to finance capital activities such as the construction of roads, schools, and bridges. This is known as “new money” debt. As the size of their outstanding new money bond portfolios increase (which increases the interest cost component of the operating budget), the opportunities for state and local governments to refinance debt (depending on interest rate conditions) to realize substantial operating budget savings grows. Refinancing bonds are known as “refunding” debt. The actual acceleration in bond issuance is illustrated in Graph 1 which details the growth in bond volume by purpose, new money or refunding, for 494 LUBY GRAPH 1 State and Local Government Annual Aggregate Bond Issuance Par Amount by Purpose (New Money or Refunding) Source: The Bond Buyer. state and local governments between 1990 and 2010. In 1990, state and local governments sold $120 billion in new money bonds, $20 billion in refunding bonds and $5 billion in combined new money and refunding bonds. Twenty years later, in 2010, state and local governments sold $279 billion in new money bonds, $98 billion in refunding bonds and $55 billion in combined new money and refunding bonds. While this comparison only examines two discrete points in time, the graph clearly shows the significant upward growth trend over time in both new money and refunding bonds and, thus, the import of both new money and refunding debt to state and local governments. DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 495 As mentioned above, the most likely reason that a state or local government would decide to refinance all or a portion of its outstanding debt portfolio is due to changes in interest rates that make the refinancing economically beneficial on an interest cost savings basis. Under this type of refinancing, the state and local government issues new bonds and uses the proceeds to retire outstanding bonds that have higher interest costs. This exchange of new debt for old debt has the effect of replacing higher interest rate debt (outstanding bonds) with lower interest rate debt (new bonds). This is known as an economic refinancing/refunding and is congruent with an individual’s decision to refinance his home due to a reduction in interest rates from the time he originally obtained his mortgage resulting in monthly mortgage payment savings. In the case of a municipal bond economic refunding, the interest cost savings provide operating budget relief to the state or local government in the form of lower annual interest payments since these interest costs are considered a current expense and paid out of the operating budget. In the last twenty years, the increased use of debt finance combined with steadily declining interest rates has presented significant opportunities for state and local governments to refinance their debt to achieve interest cost savings. Graph 2 plots interest rates and refunding bond volume detailing the relationship between annual state and local government bond refunding volume and the level of interest rates as proxied by the annual average of the Bond Buyer’s 20 GO Index. As expected, a visual inspection of this chart illustrates a strong inverse relationship between refunding volume and the overall level of interest rates. This is further evidenced by the correlation coefficient of .583 between these two variables. Similar to an individual’s refinancing of his/her home mortgage, state and local government refinancings can be structured to produce equal annual interest cost savings to the government. In an equal annual savings structure, the subnational government receives the same interest cost savings each year over the course of the life of the bond refinancing. However, state and local governments are generally not required to structure the refinancing bonds to achieve equal annual interest cost savings. That is, many government entities can structure the refinancing such that the interest savings occur in a 496 LUBY GRAPH 2 State and Local Government Annual Average Refunding Bond Issuance Compared to Average Annual Interest Rate Source: The Bond Buyer. targeted year or period of years. For example, the refinancing can be structured so that the savings all occur in the first year or first few years after the refinancing bonds are issued with the subsequent years realizing little or no debt service savings. This type of refinancing structure is known as an “upfront savings structure” or, in the parlance of home mortgage finance, a “cash-out refinancing.” Compared to an upfront savings structure, an equal annual interest cost savings structure provides greater long-term budget relief and is more efficient from a present value savings perspective. An upfront savings structure offers enhanced budgetary flexibility in the short-term, thus making an upfront savings structure attractive to the often short-range budgetary perspective of politicians, and is often less efficient compared to an equal annual savings structure. DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 497 Upfront refinancing savings structures usually entail refunding nearterm principal maturities with these bonds known as “structuring bonds” in a refinancing. Utilizing these “structuring bonds” in a refinancing usually results in an efficiency loss represented by lower present value savings compared to an equal annual savings structure. The efficiency loss vis-à-vis an equal annual savings structure is due to two factors: 1) the refunding bonds are amortized slower with longer maturities generally carrying higher interest rates than shorter maturities and/or 2) bonds that would normally not be refinanced based on interest savings target thresholds are included in the refunding. Table 1 provides a hypothetical example of the difference in annual principal and interest payment savings under the equal annual savings and upfront savings structures. Graph 3 visually illustrates this comparison between bond refinancing structures. TABLE 1 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Total $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $100,000 $9,500 $9,500 $9,500 $9,500 $9,500 $9,500 $9,500 $9,500 $9,500 $9,500 $95,000 $8,500 $8,500 $8,500 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $10,000 $95,500 $500 $500 $500 $500 $500 $500 $500 $500 $500 $500 $5,000 Annual Savings (Upfront Savings Structure) Annual Savings (Equal Annual Savings Structure) Refinancing Bonds P&I (Upfront Savings Structure) Refinancing Bonds P&I (Equal Annual Savings Structure) Year Refinanced Bonds P&I Hypothetical Savings Structure Analysis; Annual Principal and Interest Savings by Refinancing Bond Structure (In $1,000) $1,500 $1,500 $1,500 $0 $0 $0 $0 $0 $0 $0 $4,500 498 LUBY GRAPH 3 Bond Refinancing Saving Structure Comparison As shown in Table 1 and Graph 3, compared to the equal annual savings structure, the state or local government receives greater savings in each of the first three years under the upfront savings structure ($1,500,000 compared to $500,000) but realizes lower savings in each of the subsequent years ($0 compared to $500,000). Moreover, the total savings over the course of the bond issue is less for the upfront savings structure ($4,500,000) than the equal annual savings structure ($5,000,000), which reflects the greater bond structuring efficiency of the equal annual savings structure. Efficiency considerations aside, the major budgetary problem with the upfront savings structure relates to structural budget deficits and the use of one-time revenue sources. By taking all of the refinancing savings upfront (or over the first couple years), the state or local government is effectively utilizing a one-time (or short-term) revenue “shot” to fund its budget. In this example, after year 3, the government entity will either have to replace the $1,500,000 in savings with some other revenue source or cut such amount from its DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 499 annual operating budget going forward since there will be no annual principal and interest savings associated with the refinancing starting in year 4. That is, the interest costs on the refinancing bonds will revert back to the debt service levels prior to the refinancing. In austere budget times, this use of the short-term budget shot may put further strain on the government’s operating budget when the annual savings disappear in year 4. Even in the presence of strong revenue growth and a balanced budget, the use of such a short-term revenue shot provides the perception that the government entity has greater resources than it actually does (at least on a longer-term basis) which provides an artificial incentive to increase spending from its operating budget in the short term. Analyzing an actual bond refinancing that utilized an upfront bond refinancing structure will further illustrate the issues described above. Table 2 details the interest cost savings analysis for the $225,655,000 Cook County, Illinois, General Obligation Refunding Bonds, Series 2004A issued in March 2004. These bonds were sold to refund a portion of Cook County’s 1993A, 1993B, 1999A, and 2001A bonds achieving $16.5 million in total interest cost savings throughout the term of the refinancing bonds. The refinancing produced savings of $5.11 million in fiscal year 2004, $10.18 million in fiscal year 2005, and $991,910 in fiscal year 2006. Savings in fiscal years 2007 through 2023 was no more than $50,000 with some years having negative savings and several years with zero savings. Due to an artifact associated with the time of year that the refinancing was executed, Cook County achieved significant savings in fiscal year 2004 as a result of the interest on the 2004 bonds being calculated over only a seven month period (end of March 2004 through November 30, 2004) whereas the refunded bonds interest was calculated over 12 months (November 30, 2003 through November 2004). The fiscal year 2005 savings was accomplished by refunding $9.225 million principal amount of the 1993B bonds scheduled to mature in fiscal year 2005 while not amortizing any of the 2004 bonds during this fiscal period. As described above, the $9.225 million in bonds represents the use of “structuring bonds” to achieve near-term budgetary relief. In essence, through the use of this bond refinancing structuring technique, Cook County was able to monetize most of the total interest cost benefits associated with the refinancing into the first three years. 500 LUBY TABLE 2 Refinancing Analysis General Obligation Refunding Bonds, Series 2004A, The County of Cook, Illinois Refunded Bonds Annual Principal and Interest FY 2004A P&I 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Total 6,931,352 11,090,163 15,515,163 15,713,163 18,481,163 10,685,763 21,778,888 10,125,138 10,131,438 10,131,525 22,505,500 33,955,500 33,921,250 43,681,750 27,469,250 27,454,750 13,819,250 13,836,000 13,853,500 13,865,250 374,945,752 1993A P&I 1993B P&I 1999A P&I 2001A P&I 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 3,998,500 13,783,500 13,794,250 13,810,000 13,819,250 13,836,000 13,853,500 13,865,250 148,742,250 Annual Savings 1,362,860 2,986,038 3,700,150 5,116,196 10,587,860 2,986,038 3,700,150 10,182,385 5,822,385 2,986,038 3,700,150 991,910 5,051,025 2,986,038 3,700,150 22,550 7,847,030 2,986,038 3,700,150 50,555 0 2,986,038 3,700,150 -1,075 0 14,076,038 3,700,150 -4,200 0 2,431,538 3,700,150 5,050 0 2,431,538 3,700,150 -1,250 0 2,431,538 3,700,150 -1,338 0 4,756,538 13,765,150 14,688 0 16,229,475 13,759,156 31,631 0 16,228,675 13,729,406 35,331 0 16,229,550 13,710,631 41,931 0 0 13,700,681 25,681 0 0 13,672,406 27,656 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30,671,160 92,731,113 119,338,931 16,537,702 DATA AND BOND REFINANCING SAVINGS STRUCTURE OPERATIONALIZATION The second part of this paper examines the usage of upfront refinancing savings structures in state government bond refundings. This section of the paper describes the data and methodology used in calculating the degree of upfront refinancing structure for a sample of refunding bond issues. The data consist of a random sample of all state government refinancing bond issues (both revenue and general obligation bond types) sold between January 1, 2002 through December 31, 2009, a period that covers two recessions and one period of economic expansion thus allowing for some variability in the fiscal environments of the states. The universe of state government refinancing bond issues was compiled through the Thomson Reuters SDC database. The sample was generated through the random number function in the computer spreadsheet program Microsoft Excel where each bond issue was assigned a number with the computer selecting the bond issues used in the sample. Refinancing DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 501 bond issues whose final maturity was more than 2 years after the final maturity of the refunded bonds were classified as “restructuring” bonds and were not included in the universe from which the sample was drawn since these refinancing transactions were not executed primarily on the basis of producing economic savings but more likely to explicitly provide short-term operating budgetary relief through the extension of debt maturities. Nine bond issues were excluded on this basis. In this study, the number of bond issues in the sample totaled 113 from a total population of 389.1 The information necessary for determining the degree of upfront refinancings savings variable was found in each bond issue’s official statement or through its escrow agreement. The official statement is a document prepared for a new municipal bond issue by or for the issuer of the debt. This document describes the issue, financial details about the issuer and other relevant facts and is the municipal bond market’s equivalent to the stock prospectus used in the equity markets. The official statement also provides information on the details, nature and purpose of the bond refinancing including information related to the bonds being refunded under the refinancing. The escrow agreement provides additional details regarding the refunding including information on the specific bonds being refunded. The official statements and escrow agreements were acquired through EMMA, the Municipal Securities Rulemaking Board’s (MSRB) market data repository website. The degree of upfront refinancing savings structure calculation is quite novel and somewhat technical. Specifically, this study has operationalized the variable as continuous in nature and calculated as the difference between the amortization (i.e., the gradual payment of bond principal) of the first three years of refunded bond principal as a percentage of entire bond issue principal minus the amortization of the first three years of refunding bond principal as a percentage of entire bond issue principal. This operationalization aims to capture whether the state is retiring the bond principal on its refinancing bonds as quickly or quicker as its refunded bonds or whether it is employing a structure to delay bond repayment on its refunding bonds as a means of pushing the refinancing savings into the near term. A three year amortization period was chosen based on the notion that elected officials will generally try to implement budget strategies that will provide them maximum financial flexibility in the 502 LUBY short-term often just long enough until the next election. A three year window for bond refinancing savings will usually satisfy this strategy since most administrative elected officials are elected for a four-year term. Under this operationalization, the greater the value of the variable, the more the savings is structured on an upfront basis. Conversely, the lower the value of the variable, the less the savings is structured on an upfront refinancing savings basis. Since this type of analysis has never been performed to date, the operationalization of the variable is novel to the municipal finance literature. However, it is justified based on basic mathematical/financial principles that underlie it, which is best illustrated through the use of an example. Table 3 details the calculation of the degree of upfront refinancing saving structure for the 2004 County of Cook, Illinois refinancing described in the TABLE 3 Refunding Analysis General Obligation Refunding Bonds, Series 2004A The County of Cook, Illinois (In $1,000) Refunding Bonds Refunded Bonds 3rd year Bonds Cumulative 3rd year Bonds Cumulative Amortization: $4,425 Amortization: $14,155 % of Bonds Amortized in First 3 years: % of Bonds Amortized in First 3 years: 1.96% 6.10% Cumulative Cumulative Maturity Maturity Amortization Date Par Date Amortization Par Amount Amount 3/1/2004 11/15/2004 0 11/15/2004 0 11/15/2005 0 11/15/2005 9,225 9,225 11/15/2006 4,425 4,425 11/15/2006 4,930 14,155 11/15/2007 4,800 9,225 11/15/2007 4,415 18,570 11/15/2008 7,760 16,985 11/15/2008 7,445 26,015 11/15/2009 275 17,260 11/15/2009 26,015 11/15/2010 11,375 28,635 11/15/2010 11,090 37,105 11/15/2011 290 28,925 11/15/2011 37,105 2012-2023 196,730 225,655 2012-2023 195,125 232,230 4.13% % Difference Between Refunded Bonds First 3 Years Amortization and Refunding Bonds First 3 Years Amortization DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 503 previous section of this paper. The table lists the maturity dates and principal amounts for the refunding bonds and the refunded bonds showing the cumulative amortization (in dollars) at any point in time as well as the percentage and dollar amount of bonds amortized in the first three years for the refunding and refunded bonds. The bottom of the table shows the percentage point difference between the cumulative three year amortization structures between the refunded and refunding bonds. By the end of the third fiscal year after the refinancing (i.e., 2006), the county had amortized 1.96% of its refunding bonds whereas the refunded bonds principal would have been 6.10% amortized if the refinancing did not occur. This produces a 4.13 percentage point difference between the amortization of the first three years of refunded bond principal as a percentage of entire bond issue principal and the amortization of the first three years of refunding bond principal as a percentage of entire bond issue principal. In dollar terms, the refunded bonds cumulative amortization by the end of fiscal year 2006 was $14,155,000 whereas the cumulative amortization of the refunding bonds was $4,425,000. This slower amortization of the refunding bonds vis-àvis the refunded bonds allowed the County of Cook, Illinois to delay approximately $9,730,000 in bond principal payments to provide short-term operating budget relief from FY 2004 through FY 2006 (in addition to the interest cost savings from the refinancing accruing in this period). Thus, this calculation comports well with Table 2 which clearly showed that the County of Cook, Illinois used an upfront savings structure on its 2004 refinancing. USE OF UPFRONT BOND REFINANCING SAVINGS STRUCTURES BY THE STATES This section of the paper proceeds to detail the actual utilization level of upfront bond refinancing savings structures by state governments over the last decade. As a means of establishing generalizability, the analysis must show that the sample is representative of the universe of state refinancing bond issues as it relates to various bond issue and issuer characteristics such as bond sale method type, bond size, use of a financial advisor, bond type, credit rating, and total state representation. Table 4 shows the summary statistics for the various characteristics of the bonds in the 504 LUBY TABLE 4 Summary Statistics: Upfront Refinancing Savings Structure (Random Sample of State Refinancing Bond Issues Sold Between 01/01/2002 – 12/31/2009) Variable REFSAVSTRUC (Degree of upfront refinancing savings1) BONDSIZE (in millions) SALEMETH (Obs. are number of negotiated sale bond issues) RATAAA (Obs. are number of bond issues from states rated AAA at time of refinancing) RATAA (Obs. are number of bond issues from states rated AA at time of refinancing) RATA (Obs. are number of bond issues from states rated A at time of refinancing) RATBAA (Obs. are number of bond issues from states rated BAA at time of refinancing) FINADV (obs. are number of bond issues that used a financial advisor) BONDTYPE (Obs. are number of revenue bond issues) Note: Obs. Mean Std. Dev. Min. 113 2.306283 7.73956 -6.68 Max. 45.54 113 330.1338 277.0992 6.365 1074.89 71 .6283186 .4854065 0 1 29 .2566372 .438723 0 1 79 .699115 .4606857 0 1 4 .0353982 .1856073 0 1 1 .0088496 .0940721 0 1 93 .8230088 .3833613 0 1 20 .1150442 .3204966 0 1 1 Continuous variable that represents the difference between the amortization of the first three years of refunded bond principal as a percentage of entire bond issue principal minus the amortization of the first three years of refunding bond principal as a percentage of entire bond issue principal). DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 505 sample. For the sample, 71 (or 63%) were sold on a negotiated basis, 93 (or 82%) used an independent financial advisor, 20 (or 18%) were revenue bonds, and 79 (or 70%) were sold by states rated in the AA credit rating category. The mean size of bond issues was $330 million reflecting a relatively large average size with bond issue sizes ranging from the tiniest issue of $6.365 million to the largest at $1.074 billion. Refinancing bond issues from 35 different states were utilized in the sample. For the entire universe of state government refinancing bond issues sold between January 1, 2002 and December 31, 2009, 246 (or 66%) were sold on a negotiated basis, 293 (or 78%) used an independent financial advisor, 97 (or 26%) were revenue bonds, and 311 (or 83%) were sold by states rated in the AA credit rating category. The mean size of bond issues was $259 million with bond issue sizes ranging from the tiniest issue of $1.975 million to the largest at $1.5 billion. This comparison demonstrates that the sample serves as a relatively good match with the universe as a whole along these issuer and issue characteristics (although revenue bonds and AA rated states may be slightly under sampled). Most importantly for the purposes of this study, Table 5 shows the descriptive breakdown of bond issues by degree of upfront savings structure. The greater the number for degree of upfront refinancing savings structure, the more the savings was structured on an upfront basis. 48 bond issues (42% of the sample) employed savings structures where the degree of upfront savings was actually negative (i.e., the states were amortizing the refinancing bonds faster than the original bonds amortization schedule). 71 bond issues (63% of the sample) utilized savings structures where the degree of upfront refinancing savings was 1% or lower (ranging from 1% to -6.68%). This signifies that 63% of the sample of bond issues utilized a bond structure that produced very little or no degree of upfront refinancing savings. Another seven bond issues (6% of the sample) utilized savings structures where the degree of upfront refinancing savings was between 1% and 2%. If we set a threshold whereby savings structures between 1% and 2% represent a relative modest amount of upfront refinancing savings, 69% of the sample employed savings structures that produced relatively modest or no upfront refinancing savings. 35 bond issues (31% of the sample) utilized savings structures where the degree of upfront refinancing savings was 2% or 506 LUBY TABLE 5 Upfront Refinancing Savings Structure Analysis Random Sample of State Refinancing Bond Issues Sold Between 01/01/2002 – 12/31/2009: Degree of Upfront Refinancing Savings >-1% -1% to 0% to 1% to 2% to 3% to 4% to 5%> 0% 1% 2% 3% 4% 5% Number of Bond Issues with this Savings Structure 24 24 23 7 14 Percentage of Bond Issues with this Savings Structure 21.2 % 21.2 % 20.4 % 6.2% 12.4 % 5.3% 1.8% 11.5 % Cumulative Percentage of Bond Issues 21.2 % 42.4 % 62.8 % 69% 81.4 % 86.7 % 6 2 13 88.5 100% % Note: * Categories represent the amount of bond issues where the difference between the amortization of the first three years of refunded bond principal as a percentage of entire bond issue principal minus the amortization of the first three years of refunding bond principal as a percentage of entire bond issue principal is in the stated range. The greater the number, the more the savings were structured on an upfront basis. greater (ranging from 2% to 45.54%). If we set a threshold whereby savings structures greater than 2% represent more considerable upfront refinancing savings, 31% of the sample employed savings structures that produced substantial upfront refinancing savings. In sum, a clear majority of the bond issues employed little or no upfront savings structure. However, some issuers did utilize this type of debt-related budgetary gimmick. While it is beyond the scope of this paper, future research should explore why these issuers chose an upfront refinancing bond savings structure. Nevertheless, even though the descriptive results represent “negative findings” in the sense that most states did not utilize this debt-related budgetary gimmick, the results do serve as a contribution to the public finance literature. Petersen (2003) suggested that the use of upfront bond refinancing savings structures was one way that states use fiscal alchemy to balance their books. The results on this paper did not DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 507 indicate that such type of budget gimmickry is pervasively employed by the states thus providing a quantitative estimate to the relative usage size and scope of the strategy Petersen described in previous research. In addition, the empirical findings have significant theoretical and policy implications. The next section of the paper examines such implications. THEORETICAL AND POLICY IMPLICATIONS The third part of this paper involves the application of the empirical findings to theory and practice. These findings have both theoretical and policy implications relating to principal-agent theory and the efficacy/necessity of debt restrictions, respectively. Generally-speaking, principal-agent theory revolves around conflicts of interest between two parties, a principal and an agent, whereby the principal purchases goods or services from an agent whose interests may not be congruent with the goals of the principal. The principalagent model used in this paper relies on the general agency theory framework and assumptions utilized in the debt management literature (Simonsen and Hill, 1998). From an agency theory perspective, in the case bond refinancing, the public is the principal and elected officials working through their financial managers are the agent. In general, the public favors efficiency in government finance activities and does not prefer debt management decisions that produce short-term benefits at long-term costs as current taxpayers do not want themselves or the future taxpaying public (which may include their children and grandchildren) to be burdened by shortterm reckless debt management decisions made by the prior generation’s elected officials and financial managers. Thus, the public’s goal is the long-term cost efficient management of state debt.2 The elected official, who ultimately oversees the state’s bond practices (usually the governor or state treasurer), is the agent to the public. The interests of the elected official are generally short-term and primarily consist of electoral self-preservation which by its very nature entails reelection. Reelection is achieved by keeping taxes low (i.e., maintaining or decreasing tax rates) while raising or maintaining government service levels, both of which can be facilitated by certain debt refinancing practices. The elected official’s interests are accomplished through the debt refinancing decisions made by his 508 LUBY financial manager who acts as an agent on behalf of his principal, the elected official. However, the bond issuance process is quite technical and the decisions made by elected officials working through their financial managers are often not transparent and depending on a state’s debt disclosure requirements may not easily be known or understood by the public. Thus, there is a strong information asymmetry between the principal (the public) and its agents (elected officials and financial managers) as it relates to the nuances of bond refinancing decision-making. Figure 1 details the principal-agent relationship between the parties in the municipal bond refinancing process as well each party’s goals and interests. Theoretically, principal-agent problems are said to exist when management activities involve delegated bureaucratic decisionmaking and are plagued by information asymmetry, and thus, allow for the prospect of moral hazard. The decision on how to structure refinancing bond savings, a subtle but important public financial management choice, meets these requirements. More specifically and as described earlier, many state and local governments allow public financial managers to capture the timing of refinancing bond savings in any permutation: upfront, equal in every year, proportional, back-loaded, etc. (i.e., there exists bureaucratic delegation). Moreover, the refinancing savings structure decision is not readily transparent to the principal since many states do not proactively FIGURE 1 Municipal Bond Refinancing Process Participants and their Goals and Interests DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 509 provide nuanced information about their bond sales to the public (i.e., information asymmetry is present). Finally, the temptation to structure refinancing savings upfront may be strong for public financial managers since it provides elected officials the ability to increase spending, cut taxes, or avoid painful spending cuts in the near term. However, the cash infusion from upfront refinancing savings effectively serves as one-time (or short-term) revenue infusion, which combined with any of the aforementioned budget decisions, will only exacerbate or help cause a structural budget deficit (i.e., prospect of moral hazard). Thus, the use of an upfront refinancing savings structure offers evidence of a classic principalagent problem in public financial management. In this policy environment, the elected official chooses an alternative policy option (debt refinancings with upfront savings structures) to meet his interests (i.e., lower principal and interest payments in the short-term so as to avoid the possible electoral wrath of the public often associated with tax increases or spending cuts) at the expense of his principal’s interests (i.e., lower principal and interest payments over the long-term). Figure 2 details in visual form how the use of upfront refinancing savings structures provides evidence of an agency problem. The results of the empirical analysis carry some interesting implications as it relates to the agency theory and the necessity of debt restrictions as applied in this context. First, the finding of only very moderate use of upfront refinancing savings structures (i.e., only 31% of all state government bond refinancings used savings structures where the three year amortization structure difference FIGURE 2 Agency Theory and Upfront Refinancing Bond Structures 510 LUBY between the refunded and refunding bonds was 2% or greater) questions the explanatory power of agency theory in understanding decision-making by public financial managers. That is, the use of upfront refinancing savings structures in bond refinancings provides agents (elected officials and debt managers) a perfect opportunity to make decisions that provide a short-term benefit to the agent at a long-term cost to the principal (the public) in the administrative management context where information asymmetry is present and delegated bureaucratic decision-making is available. Thus, a priori, agency theory would predict that upfront refinancing savings structures would be used extensively by state governments in this seemingly unending era of fiscal austerity. Since it has been shown that most state governments did not capitalize on such opportunity (i.e., they did not generally use an upfront refinancing savings structure), it seems that agency theory does not provide an adequate theoretical explanation for the decision-making behavior of most of these state governments. Thus, this finding does not provide support for prior research that directly addressed the issue of principal-agent problems in the municipal bond process finding that a substantial agency problem existed (Simonsen and Hill, 1998). As such, an alternative theoretical framework that stresses influences beyond individual self-interest may offer a better explanation for the behavior of these state governments. For example, based on this finding, one might hypothesize that state level public financial managers are generally motivated and influenced by their understanding of prudent debt policy as developed in their education and professional training in making bond refinancing decisions. As described in this study’s literature review, recent research that described public financial managers influenced by factors beyond self-interest such as professional norms and individual values may offer a better explanatory theoretical framework for understanding this type of public financial management decision-making behavior (Carroll and Marlowe 2006; Burke 2006; Rivenbark 2006; Robbins and Miller 2006; Stalebrink and Sacco 2006). However, further systematic research is needed in analyzing the use of upfront refinancing savings structures in local government refinancing transactions in order to definitively determine the extent of this debt management technique by all subnational governments. DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 511 Second, this finding related to principal-agent theory has direct policy implications as well. In recent years, several states have placed greater restrictions on their debt management practices due to the perceived existence of agency problems plaguing this area of financial management. That is, some observers in the municipal finance arena speculate that the interests of public financial managers (i.e., agent) acting on behalf of elected officials may not be aligned properly with the debt management goals of the public (i.e., principal). Therefore, statutory debt restrictions are needed to better ensure that a public financial manager’s actions reflect the desires and policy goals of the public, which are assumed to be the long-term cost efficient and effective management of state debt. However, reducing managerial autonomy is not costless and, thus, governments need to be sure such restrictions are truly necessary in order to determine their overall efficacy (Luby, 2009). Practically-speaking, the finding of very moderate use of upfront refinancing savings structures casts doubt on the wisdom of applying blanket restrictive managerial control mechanisms on all states’ debt managers as a means of controlling opportunistic managerial behavior and, thus, mitigating agency problems. That said, based on the fact that some states did, in fact, engage in some opportunistic behavior as it relates to bond refinancing, statutory restrictions may be appropriate in some states but certainly not in all. However, even in these states, more formalized/stricter debt policy guidance and greater disclosure requirements may accomplish the same goal as statutory restrictions without the costs associated with limiting financial managerial autonomy. CONCLUSION The escalating use of debt finance combined with steadily declining interest rates over the last decade have provided state and local governments ample bond refinancing opportunities that have the potential to substantially reduce the interest cost component of their operating budgets. However, the ever-present fiscal stress experienced during this time period has incented subnational governments to contemplate refinancing savings structures, including the upfront refinancing savings structure explored in this paper, that emphasize short-term budget relief at the expense of maximizing aggregate present value interest cost savings. These structures also 512 LUBY exacerbate or help create structural budget deficits as they act like other “one-shot” revenue techniques such as asset sales and payment deferrals. Thus, these debt refinancing structures effectively serve as a budget gimmick that state and local governments can use to achieve “short-term” budget balance. The first part of this paper offers substantial detail on the mechanics and impact of the upfront refinancing bond savings structure as a means of better understanding this type of fiscal alchemy. This paper’s empirical analysis of state government utilization of upfront bond refinancing savings structures attempts to shed light on managerial decision-making in an area of public financial management that generally lacks transparency yet offers bureaucratic autonomy and, thus, investigates a policy area where, at least theoretically, agency problems are most likely to be common and pervasive if they do, in fact, exist. However, the results of this study demonstrate that most state governments did not employ upfront refinancing savings structures in their bond refinancings. This finding provides evidence that a pervasive agency problem may not be present in the management of state debt. Thus, these results seem to question the grand use of agency theory in helping to understand this area of public financial management. In addition, from a policy perspective, this finding offers empirical support for generally minimizing statutory debt restrictions on public financial managers as this paper’s analyses seem to demonstrate that these managers usually make prudent financial decisions as it relates to bond refinancing. That said, some states did utilize an upfront refinancing savings structure and the wisdom of applying statutory debt management restrictions to these states is clearly still debatable. The states that did utilize upfront refinancing savings structures provide a fertile area of future research. That is, from a policy perspective, it would be instructive to determine the various factors that impact the refinancings savings structure decision as a means of better understanding decision-making in this area of public financial management and to offer some specific policy and management recommendations that could serve to mitigate the agency problem to the extent it does exist. For example, does the existence of formal or informal debt policies reduce the likelihood of utilizing upfront refinancing savings structures? Does the use of an independent DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 513 financial advisor impact decision-making in this area? Another area of inquiry may be whether the use of these structures is contingent on the level of fiscal stress or financial condition of the state government. Research of this type and scope would be very valuable in understanding the impact of various bond issue and issuer characteristics on the decision-making of public financial managers with an “eye” towards developing practices that encourage long-term fiscally prudent actions as it relates to debt management. NOTES 1. A sample was used in this study due to the significant time it takes to construct the dependent variable in this analysis. The data for the dependent variable comes from two source documents comparing the refunded and refunding bonds amortization. In addition to the substantial time it takes to acquire all the information on the refunded bonds from the source documents, the dependent variable for each bond issue needs to be constructed comparing and calculating the difference in amortization schedules between the refunded and refunding bonds, which also is time consuming. As stated elsewhere in the paper, the sample is relatively large (30% of the universe) and represents a good match of the universe as shown in Table 4, thus enhancing the external validity of the study. 2. One could argue that the public’s goal is not long-term, sustainable budgeting practices via the use of debt but rather to prefer debt management practices that allow for taxes not to be raised or services not to be cut in the short-term (e.g., using upfront refinancing structures) at the expense of long-term finances. However, the lack of the citizenry’s general knowledge of the tradeoffs associated with various debt refinancing structures makes it difficult to ascertain the public’s general preference. Thus, this paper’s model assumes citizen’s preferences for long-term prudent debt management activities. REFERENCES Babad, Y., & Speer, P. (1978). “Optimization of the Arbitrage Process in Advance Refunding of Municipal Bonds.” Management Science, 24 (10): 987-1000. 514 LUBY Bhagat, S., & Frost, P. A. (1986). “Issuing Cost to Existing Shareholders in Negotiated and Competitive Offerings of Public Utility Equity Issues.” Journal of Financial Economics, 15 (1-2): 233-259. Briffault, R. (1996). Balancing Acts: The Reality Behind State Balanced Budget Requirements (9th ed.). New York: Twentieth Century Fund Press. Burke, B. (2006). “Altruism, Accountability and Transparency: Cooperative Aspirations in Contemporary State-Level Budgeting.” Journal of Public Budgeting, Accounting and Financial Management, 18 (2): 197-222. Carroll, Deborah and Justin Marlowe. 2006. “Symposium on the New Organizational Economics of Public Financial Management: Part II.” Journal of Public Budgeting, Accounting & Financial Management, 18 (2): 193-196. Center on Budget and Policy Priorities (2011, March 9). States Continue to Feel Recession’s Impact. Washington, DC: Center on Budget and Policy Priorities. Dyl, E., & Joehnk, M. (1976). “Refunding Tax-Exempt Bonds.” Financial Management, 5 (2): 59-66. Kalotay, A., & May, W. (1998). “The Timing of Advance Refunding of Tax Exempt Municipal Bonds.” Municipal Finance Journal, 19 (3): 1-15. Kalotay, A., Yang, D., & Fabozzi, F. (2007). “Refunding Efficiency: A Generalized Approach.” Applied Financial Economics Letters, 3 (3): 141-146. Kriz, K. A. (2003). “Comparative Costs of Negotiated versus Competitive Bond Sales: New Evidence from State General Obligation Bonds.” The Quarterly Review of Economics and Finance, 43 (2): 191-211. Leonard, P. A. (1996). “An Empirical Analysis of Competitive Bid and Negotiated Offerings of Municipal Bonds.” Municipal Finance Journal, 17 (1): 37-67. Luby, M. J. (2009). “Reforming State Debt Management Practices: The Case of Illinois, 2004.” Municipal Finance Journal, 30 (9): 136. DEBT AND BUDGET GIMMICKRY REVISITED: THE UPFRONT BOND REFINANCING SAVINGS 515 Luby, M. J. (2011). “The Use of Financial Derivatives in State and Local Government Bond Refinancings: Playing with Fire or Prudent Debt Management.” Journal of Public Budgeting, Accounting & Financial Management, 24 (1): 1-31. Mikesell, J. (2009). Fiscal Administration (9th ed.). Chicago, IL: Harcourt Brace. Miller, G. J. (1993). “Debt Management Networks.” Public Administration Review, 53 (1): 50-58. Peng, J., & Brucato, P. Jr. (2003). “Another Look at the Effect of Method of Sale on the Interest Costs in the Municipal Bond Market – A Certification Model.” Public Budgeting and Finance, 23 (1): 73-95. Petersen, J. E. (2003). "Changing Red to Black: Deficit Closing Alchemy.” National Tax Journal, LVI (3): 567-577. Rivenbark, W. (2006). “Evolutionary Theory of Routine: Its Role in Results Based Management.” Journal of Public Budgeting, Accounting & Financial Management, 18 (2): 223-240. Robbins, M., & Dungan, C. (2001). “Debt Diligence: How States Manage the Borrowing Function.” Public Budgeting and Finance, 21 (2): 88-105. Robbins, D. and G. Miller. 2006. “Auction Off the Farm: Signaling, Politics and Economic Development.” Journal of Public Budgeting, Accounting & Financial Management, 18 (3): 307-350. Simonsen, B., & Hill, L. (1998). “Municipal Bond Issuance: Is There Evidence of a Principal-Agent Problem?” Public Budgeting and Finance, 18 (4): 71-100. Simonsen, W., & Robbins, M. B. (1996). “Does It Make Any Difference Anymore? Competitive versus Negotiated Municipal Bond Issuance.” Public Administration Review, 56 (1): 57-64. Simonsen, W., & Kittredge, W. (1997). “Competitive versus Negotiated Municipal Bond Sales: Why Issuers Choose One Method over the Other.” Municipal Finance Journal, 19 (2): 1-29. Stalebrink, O., & Stacco, J. (2006). “Public Sector Investment Failures: Theoretical Contributions from New Institutional and 516 LUBY Austrian Economic Theory.” Journal of Public Budgeting, Accounting and Financial Management, 18 (3): 351-375. Vijayakumar, J. (1995). “An Empirical Analysis of the Factors Influencing Call Decisions of Local Government Bonds.” Journal of Accounting and Public Policy, 14 (3): 203-231. Zhang, W. D., & Li, J. (2004). “The Value of the Advance Refunding Option and the Refunding Efficiency of Tax-Exempt Municipal Bonds.” Municipal Finance Journal, 25 (1): 17-31. JOURNAL OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT 2012 INFORMATION FOR CONTRIBUTORS EDITORIAL POLICY Published four times a year, Journal of Public Budgeting, Accounting & Financial Management (JPBAFM) is a refereed journal which aims at advancement and dissemination of research in the field of public budgeting and financial management. The journal concentrates on the development of theories and concepts so that the field's boundaries can be established. The cognate areas constituting the focus of this publication are disciplines which concern how a budget is prepared, decided and implemented. Practitioners and scholars are encouraged to submit manuscripts to the journal. Papers--whether empirical, field study, or conceptual--should help to serve the need for more active communication and greater exchange of thought, research and practical experiences among scholars and practitioners throughout the world. Appropriate topics for papers include various aspects of public budgeting and financial management such as (a) governmental accounting and financial reporting; (b) politics of budgeting, budgetary process and techniques; c) public financial management including cash management, risk management, debt management; (d) tax and expenditure policies, and (e) other issues related to governmental accounting, budgeting, financial management and fiscal policies. Priority will be given to papers having carefully developed methods, insightful conceptual development, and practical and analytical solutions to government financial management problems. Interdisciplinary approaches are welcome. WRITING TIPS FOR ARTICLES. The general format of the manuscript should be as follows: title of article, names of author, abstract, and text. The ABSTRACT should not have more than 120 words in length, covering (1) a statement of the background situation that led to the development of the manuscript; (2) a clear statement of the problem or the basic issues involved; (3) a brief summary of the key findings or conclusions of the research; and (4) a brief description of the methodology used. Whenever possible, the text discussion should be divided into such major sections as INTRODUCTION, METHODS, RESULTS, DISCUSSION, ACKNOWLEDGMENTS, and REFERENCES. Manuscripts should be submitted typed, double-spaced, on one side only. The entire typing area on the title page should be four and one-half inches wide by five and one-half inches long. The major headings should be separated from the text by two lines of space above and one line of space below. Each heading should be in capital letters, centered, and in bold. Secondary headings, if any, should be flush with the left margin, in bold characters, and have the first letter of all main words capitalized. Leave two lines of space above and one line of space below secondary headings. All manuscripts should be left- and right-hand margin justified. JOURNAL OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT 2012 ACKNOWLEDGMENTS of collaboration, sources of research funds, and address changes for an author should be listed in a separate section at the end of the paper after the section on References. EXPLANATORY FOOTNOTES should be kept to a minimum and be numbered consecutively throughout the text and aggregated in sequence under the heading NOTES, at the end of the text but before REFERENCES. REFERENCES. The references section serves to provide the reader with sufficient information so that he or she can easily locate the work cited in the research. Overall, each reference should include the following information author(s)’ name (first name, middle initial, and last name); title of work; journal, serial, proceedings, or book in which the work was published; volume and number of the issue [example: volume 1, number 1 would appear as 1(1)]; date the work was published; page numbers (in the case of journals, serials, and proceedings). See www.pracademics.com for detailed manuscript instructions. REVIEW PROCEDURE. All manuscripts are reviewed by three peer reviewers, who are selected on the basis of their specialized expertise by the editorial staff. Manuscripts are assigned a code number before being mailed to peer reviewers so the author(s) remain anonymous. The JPBAFM Editorial Board and peer reviewers consist of both academicians and practitioners, with national and international representation. Reviewers make suggestions to the editorial staff if and when a rewrite is needed. Rewrites are requested for approximately 70 to 80 percent of accepted articles. ACCEPTED MANUSCRIPT PREPARATION. All accepted manuscripts will be copy-edited by a professional copy editor. COPYRIGHT. Only original papers will be accepted, and copyright of published manuscripts will be vested in the publisher. In other words, contributors release the copyright of their articles to PrAcademics Press by signing a Copyright Release Form available for PDF download at www.pracademics.com. Please note that employees of certain governmental and profit entities may not be authorized to release the copyright of their articles. A SERVICE TO SCHOLARS WHOSE FIRST LANGUAGE IS NOT ENGLISH The IJOTB recognizes that to be a truly international journal, we must provide access to scholars whose first language is not English. To aid in this goal, we are offering copyright services to potential authors for a moderate fee. This will increase opportunities for publication, by enabling potential authors to communicate their ideas more clearly in written English. To take advantage of this service, please contact Helene Kreamer at [email protected] to initiate the process. MANUSCRIPT SUBMISSIONS AND INQUIRIES. Electronic submissions are encouraged. A cover letter must accompany each submission indicating the name, address, telephone number, and e-mail address of the corresponding author. JOURNAL OF PUBLIC BUDGETING, ACCOUNTING & FINANCIAL MANAGEMENT 2012 For questions concerning journal policies, manuscript submissions, and symposium proposals/submissions, please submit/contact appropriate editors as follows. Regular Manuscripts, Symposium Proposals/Submissions and other Inquiries should be e-mailed to Co-Editors in Chief as follows: Khi V. Thai, Ph.D., E-mail: [email protected] Robert Kravchuk, Ph.D., E-mail: [email protected] Governmental Accounting Manuscripts should be e-mailed to: Don Deis, DBA, Professor, Governmental Accounting Section Editor E-mail: [email protected]. Healthcare & Nonprofit Organizations Manuscripts should be e-mailed to: Dana Forgione, Healthcare & Nonprofit Organizations Manuscripts, Section Editor. E-mail: [email protected].
© Copyright 2024