Indirect cost allocations and the incentive to report high program service expenditures in Form 990: an examination of donors' perceptions.

ABSTRACT

The Internet presence of Forms 990 has provided nonprofits with increased incentives to report high program service expenditures. Organizations can artificially inflate these expenditures by allocating large amounts of indirect costs to program services. An experiment investigating donors' perceptions regarding program services and indirect cost allocations finds that nonprofits' attempts to mislead donors through cost allocations may succeed, even when signs of potentially manipulative allocations appear on functional expense statements. Donors possessing financial analysis skills may, however, detect these attempts, suggesting that web sites may need to emphasize a careful examination of the expense composition of program services.

1. INTRODUCTION

As a result of the 1996 Taxpayer Bill of Rights 2, most nonprofit organizations must now provide copies of their three most recent Forms 990, the comprehensive yearly informational return filed with the Internal Revenue Service (IRS), to donors requesting them. This mandate has led to the creation of two web sites, www.ncos.urban.org and www.guidestar.org, where donors can now view thousands of nonprofit organizations' Forms 990. Despite Form 990's role as one of the most commonly used sources of nonprofit financial information today (The Urban Institute, 2002a), many in the nonprofit field fear that the informational return may mislead donors (Moore and Williams, 1998). One area of special concern lies in the reporting of program service expenditures and the use of cost allocations to artificially inflate their magnitude. This paper discusses an experiment conducted to investigate donors' perceptions of potentially manipulative indirect cost allocations reported in Form 990. In providing a preliminary indication of the likely success or failure of organizations' attempts to mislead donors through cost allocations, the paper offers some practical suggestions for web sites housing Forms 990.

2. LITERATURE REVIEW

The federal government has long supported the establishment of nonprofit organizations, enacting legislation conducive to their formation and granting thousands of them tax-exempt status (Wilson, Hay and Kattelus, 1999). As an overseer of the nonprofit sector, the government has sought to ensure that nonprofit organizations have not engaged in activities that have violated their tax-exempt status. To fulfill this duty, the federal government currently requires most tax-exempt nonprofit organizations (501(c)(3) organizations) to file a yearly informational return, the Form 990. Form 990 is a comprehensive report consisting of nine parts and one set of schedules. It discloses not only revenue sources and organizational expenses, but also nonfinancial information, such as major nonfinancial program service accomplishments. In its reporting of organizational expenses, Form 990 uses one of three categories to classify all expenses: program services; management and general; and fundraising. According to the IRS, organizations completing Form 990 should use the same accounting procedures they employ in their day-to-day business operations.

Because of the increased accessibility of Forms 990, nonprofit entities face incentives to disclose high program service expenses and low management and fundraising expenses (overhead) to ensure a donor's favorable impression of the organization (Gordon, Greenlee and Nitterhouse, 1999). Organizations may respond to these incentives by allocating large amounts of indirect costs to the program service expense classification to inflate its magnitude, a practice that, according to critics, charitable organizations have committed in the past (Gordon, Greenlee and Nitterhouse, 1999). While the IRS encourages organizations to allocate indirect costs among the three expense classifications (program services, management and fundraising), it fails to specify an appropriate allocation method. Further, professional accounting organizations offer only limited cost allocation guidance. For example, the American Institute of Certified Public Accountants' (AICPA) Statement of Position (SOP) 98-2, Accounting for Costs of Activities of Not-for-Profit Organizations and State and Local Governmental Entities That Include Fund Raising (1998) establishes general guidelines for the allocation of joint fund raising costs but requires only footnote disclosure of the allocation. Despite the limited guidance, nonprofit organizations continue to recognize the importance of cost classification and assignment. One professional website dedicated to Form 990 preparation advises nonprofit organizations (The Urban Institute, 2002b), "If the percentages for either fundraising or management and general appear too high, go back and make sure that your organization used appropriate guidelines when classifying expenses."

A Statement of Functional Expenses included in Form 990 may provide assistance to donors in detecting an organization's efforts to manipulate program service results. This statement reveals the individual expenses, both direct and indirect, comprising total program service, management and general, and fundraising, and reports, in columnar format, the manner in which an organization allocated its indirect expenses among the three expense classifications. Some have maintained that this statement is the most important financial report of a nonprofit organization, since it may disclose unusual cost allocations (Khumawala and Gordon, 1997). How actual donors interpret this statement and its disclosure of potentially manipulative cost allocations remains, however, unclear.

3. DEVELOPMENT OF RESEARCH HYPOTHESES

When evaluating the performance of nonprofit organizations, donors and public service agencies rank the ratio of total program service expenditures to total organizational expenditures high in importance (Wilson, Hay and Kattelus, 1999). Donors can calculate this ratio from data appearing in Part I of Form 990, the Revenue, Expenses, and Changes in Net Assets or Fund Balances statement. This statement reports an organization's revenue sources and the total dollar amounts spent on program services. management and general, and fundraising. Given that donors likely perceive high amounts of total expenses devoted to program services as indicative of an organization's use of contributions to support program service activities, the following hypothesis, stated in alternative form, is offered:

H1: Other factors being equal, donors receiving only the Revenue, Expenses and Changes in Net Assets or Fund Balances statements, Part I of Form 990, of two competing organizations will judge the entity reporting a higher proportion of total organizational expenses classified as program services as more likely to use a financial contribution to fund activities directly related to the organization's tax-exempt purpose.

Through its disclosures of the amounts of indirect costs (overhead) allocated to the program service classification, the Statement of Functional Expenses, Part II of Form 990, can reveal attempts by a nonprofit entity to inflate program service expenditures. Prior research has found considerable interest among donors in functional expense information (Khumawala and Gordon, 1997). Thus, it is likely that donors will closely scrutinize the expense composition of program services appearing in a statement of functional expenses and will prefer those organizations reporting proportionately more (less) directly (indirectly) related program service expenses. A second hypothesis, stated in alternate form, is offered below:

H2: Other factors being equal, donors receiving Statements of Functional Expenses, Part II of Form 990, of two competing organizations will judge the entity reporting a higher proportion of total organizational expenses directly related to program services as more likely to use a financial contribution to fund activities directly related to the institution's tax-exempt purpose.

4. EXPERIMENTAL MATERIALS AND METHOD

An experiment was conducted to test the two hypotheses. Sixty-four donors, consisting of 38 males and 26 females with an average age of 34, participated in the study. Subjects had an average of 9.5 years of business experience and were employed full time. Average household income fell in the $50,000 to $75,000 range. All participants had contributed money, or had donated their time, to at least one nonprofit organization. In addition, all had completed at least one, three semester credit hour introductory accounting course and were enrolled in an evening MBA program when the experiment was administered. Excluding nine participants with undergraduate accounting degrees, subjects averaged 5.5 semester credit hours of undergraduate accounting. Participants were randomly assigned to a control group (33 subjects) and an experimental group (31 subjects).

Following Gordon and Khumawala (1999), case materials instructed participants to assume that their company's board of directors had authorized a substantial, unrestricted contribution to one of two liberal arts colleges. Subjects were told: that each college served approximately the same total number of students; that they had no ties to either institution; and that their superiors had requested them to review sections of each institution's Form 990. After reviewing the sections of Form 990 they received, participants indicated which institution they believed would more likely utilize a financial contribution to fund activities directly related to the organization's tax-exempt purpose.

To perform the evaluation, participants in both groups received Parts I and III of Form 990 of each college for three years. These reports were based loosely on actual Forms 990 prepared by existing liberal arts colleges. Part I, the Revenue, Expenses, and Changes in Net Assets or Fund Balances statement, listed each institution's revenues and total amounts of program services expenses, management and general expenses, and fundraising expenses. Each college reported nearly identical amounts of total revenues and total expenses for each year. These amounts grew at a yearly rate of three percent. For each year, Institution A (B) reported approximately 87.5 (71.9) percent of its total expenses devoted to program services, 8.3 (23.8) percent to management and general, and 4.2 (4.3) percent to fundraising. Thus, from the perspective of the Revenue, Expenses, and Changes in Net Assets or Fund Balances statement, Institution A appeared to have spent more on tax exempt (program services) activities than Institution B. Part III, a Statement of Program Service Accomplishments, informed subjects of the tax-exempt purpose of each college and the general accomplishments of each organization, both of which were the same for each institution.

In addition to Parts I and III, experimental group participants also received Part II, the Statement of Functional Expenses, of each college for each year. Part II revealed the actual expense composition of the totals appearing in Part I, and, in turn, how each institution allocated its major indirect costs. Four allocated expenses were manipulated in the experiment: compensation of officers and directors; occupancy; interest; and depreciation and depletion. Institution A (B), which reported significantly greater (lower) totals for each of the four indirect expenses, allocated 91 (20) percent of them to program services, eight (78) percent to management and general, and one (two) percent to fundraising. Institutions A and B disclosed approximately equal total amounts of all other indirect expenses and allocated nearly identical amounts of them to the three cost classifications.

Institution A's significant dollar magnitude of the four indirect expenses, combined with the high percentages it allocated to program services, allowed it to report a high proportion of total expenses devoted to program services in Part I of Form 990. Without these four allocations, Institution A's (B's) proportion of program service expenses dropped to 51.8 (67.9) percent of total expenses. The Statement of Functional Expenses also revealed that Institution B spent over twice as much on grants and allocations to students and significantly more on other directly related educational activities. Thus, Institution B, with less overhead in total, lower amounts of overhead allocated to program services, and higher amounts of directly related program expenditures, appeared to report in the Statement of Functional Expenses a higher percentage of total organizational expenses devoted to activities directly related to its tax-exempt purpose.

5. TESTS OF HYPOTHESES

According to the study's two hypotheses, control group participants, who received only the Revenue, Expenses, and Changes in Net Assets or Fund Balances statements, should have selected Institution A (the institution reporting a greater proportion of total expenses devoted to program services) as the college more likely to use a contribution to fund activities directly related to its tax exempt purpose; experimental group participants, who received Statements of Functional Expenses, should have favored Institution B (the institution reporting a greater percentage of directly related program service expenditures). Two separate chi-square tests, one for each group, were performed to evaluate the two hypotheses. These are shown in Table 1.

Control group subjects demonstrated a statistically significant preference for Institution A, evidence supportive of the study's first hypothesis H1. Experimental group participants, however, failed to demonstrate a preference for either college, a finding not supportive of the study's second hypothesis H2.

A closer examination of the academic and professional backgrounds of experimental group subjects revealed that nine participants had an accounting or finance background (defined as either a baccalaureate degree in accounting, finance, or economics or work experience in accounting or finance). These participants comprised nearly 44 percent of all subjects selecting Institution B, the institution reporting a greater percentage of directly related program service expenditures. Since those with a financial background may have possessed a greater familiarity with financial statement analysis, a separate chi-square test examining only these subjects' decisions was performed. A chi-square test statistic of 2.7777 (p=.10) resulting from the 7 participants selecting Institution B and the 2 choosing Institution A provided some statistical support of a preference for Institution B among these participants, evidence supportive of the study's second hypothesis, H2.

6. DISCUSSION

The IRS grants nonprofit organizations considerable discretion in their determination of program service, management and general, and fundraising expenditures. It encourages these organizations to allocate its indirect costs among the three expense classifications, although it provides only limited guidance as to the allocation techniques or methods. Regulators and oversight agencies have long considered the Statement of Functional Expenses of primary importance in revealing questionable cost allocations. Evidence from this study suggests, however, that the functional expense statement's disclosure of a nonprofit organization's potentially manipulative use of indirect cost allocations to inflate program service expenditures may go undetected, implying that organizations may find it relatively easy to mislead donors using Form 990.

Participants with a financial background, however, exhibited a modest, statistically significant preference for the institution reporting proportionately higher directly related program costs in the Statement of Functional Expenses. Additional education, or experience, in financial analysis may enhance Form 990 users' ability to detect attempts to manipulate financial results. Currently, online Form 990 databases offer short tutorials explaining the Statement of Functional Expenses and encouraging users to examine attachments. Donors may also benefit from information explaining the potential financial effects of manipulative cost allocations and emphasizing the need to examine carefully the expense composition of program services.

7. LIMITATIONS AND CONCLUSION

Three limitations characterize the design and administration of this study. First, participants received only portions of each organization's Form 990. Second, subjects may have had more accounting knowledge than the average user of Form 990. Third, participants were not randomly selected from the general population of donors.

Despite limitations future investigations might encounter, the study of nonprofit accounting and Form 990 offer abundant research opportunities. Clearly, Form 990 can no longer be viewed simply as an informational return designed to satisfy IRS requirements. Its widespread distribution and its potential importance to donors support, and argue for, continued research.

TABLE 1

CHI-SQUARE TEST RESULTS OF PARTICIPANTS' CHOICES

Panel A: Control Group

Number Selecting   Number Selecting
Institution A      Institution B     Total     Chi-Square

     29                   4            33        18.9393

Panel B: Experimental Group

Number Selecting   Number Selecting
Institution A      Institution B     Total     Chi-Square

     15                 16            31         .0333

Panel A: Control Group

Number Selecting
Institution A      Degrees of Freedom          p-value

     29                  1                       <.01

Panel B: Experimental Group

Number Selecting
Institution A      Degrees of Freedom          p-value

     15                  1                       >.5

8. REFERENCES

American Institute of Certified Public Accountants (AICPA), Statement of Position No. 98-2: Accounting for Costs of Activities of Not-for-Profit Organizations and State and Local Governmental Entities that Include Fund Raising, AICPA, New York, New York, 1996.

Gordon, T. P., Greenlee, J.S., and Nitterhouse, D., "Tax-Exempt Organization Financial Data: Availability and Limitations," Accounting Horizons, Vol. 13 (2), 1999, 113-128.

Gordon, T. P., and Khumawala, S.B., "The Impact of Joint Cost Allocation Disclosures in the Decision to Support Not-for-Profit Organizations," delivered at the 1999 American Accounting Association Annual Meeting, August 1999, San Diego, California.

Khumawala, S. B., and Gordon, T P., "Bridging the Credibility of GAAP: Individual Donors and the New Accounting Standards for Nonprofit Organizations," Accounting Horizons, Vol. 11(3), 1997, 45-68.

Moore, J., and Williams, G., "Return of the Future: Disclosure Laws and Internet Access Spur New Focus on Charity Tax Forms," The Chronicle of Philanthropy, December 17, 1998; The Urban Institute, http://www.qua1990.org/return of the future.html.

The Urban Institute, Welcome to Quality 990, 2002a; The Urban Institute, http://www.qua1990.org/index.html

The Urban Institute, Six Things the Board President Should Check Before the 990 Is Filed, 2002b; The Urban Institute, http://www.qua1990.orq/support center.html.

Wilson, E.R., Hay, L.E, and Kattelus S.C., Accounting for Governmental and Nonprofit Entities (11th ed.), Irwin/McGraw-Hill Publishing, Burr Ridge, IL, 1999.

Dr. R. Steven Flynn earned his Ph.D. at the University of Cincinnati in 1987. Currently he is an associate professor of accountancy and chair of the Department of Accountancy at Thomas More College. Dr. Flynn is also a Certified Public Accountant (CPA).