Functional Expense Reporting for Nonprofits
The Accounting Profession’s Next Scandal?

By Kennard Wing, Teresa Gordon, Mark Hager, Thomas Pollak, and Patrick Rooney

E-mail Story Print Story

Unfortunately, nonprofit financial reporting represents a potential ticking time bomb for the profession. The authors’ Nonprofit Overhead Cost Project (see the Sidebar) found serious and widespread errors in IRS Forms 990 and audited financial statements prepared or attested to by CPAs. The dollar amounts involved may be smaller than for large public companies, and the errors less likely the result of malfeasance, but these errors can and should be corrected. This article shows where the problems are, and presents one approach to doing something about them.

The Importance of Functional Expense Reporting

The most widespread and serious problems in nonprofits concern expenses by functional classification (program, management and general, and fundraising). Many nonprofits cannot provide good information about their relative effectiveness in fulfilling their mission. As a result, donors, funders, and charity watchdog organizations have placed undue reliance on financial indicators, many of which are based on reported functional expenses.

Two commonly used financial indicators are the program-spending ratio (total program expenses ÷ total expenses) and the fundraising-efficiency ratio (fundraising costs ÷ total contributions). Many articles aimed at donors use these ratios as their primary basis for evaluating and ranking charities. However, such ratios are only as good as the numbers used to calculate them, which generally come from an organization’s IRS Form 990.

IRS Form 990

Publicly traded companies’ audited financial statements are public documents, while the tax return is a private matter between the corporation and the IRS. For nonprofits, the opposite is the case. A nonprofit’s Form 990 is a public document. Nearly 2.4 million that have been imaged by the IRS are available online at www.guidestar.org. In contrast, nonprofits’ audited financial statements are required to be released by only a minority of states’ charity registration offices.

Because the data are readily available, most donors, funders, and charity watchdog agencies calculate program-spending and fundraising-efficiency ratios using Form 990 data. The authors’ study of this data found widespread reporting that defies plausibility in the functional expenses used to make those calculations. Examples include the following:

• 37% of nonprofits with at least $50,000 in contributions report zero fundraising costs.
• One-fourth of nonprofits reporting $1 million to $5 million in contributions report zero fundraising costs, as do nearly one-fifth of those reporting more than $5 million in contributions.
• 13% of nonprofits report zero management-and-general expenses.
• 7% charged all accounting fees to programs, and another 20% split them across more than one category—despite the fact that Form 990 instructions use accounting fees as an example of what is meant by management-and-general expenses.

Because of the large number of nonprofits reporting zero fundraising costs, the authors included two such organizations in a set of nine in-depth case studies they performed. The first organization erroneously reported zero fundraising costs in its audited financials, and on its Form 990. The organization had a part-time employee who worked exclusively on fundraising, the executive director was involved in fundraising, and the organization did some direct-mail fundraising. In the second case, the tax professional at the audit firm prepared a Form 990 with zero fundraising cost despite the fact that the financials behind the opinion letter of the audit team showed more than $500,000 in fundraising expenses. When the authors’ research team asked about this almost a year later, no one inside the organization had noticed, and all expressed bewilderment as to how the error could have occurred.

For a limited number of organizations, including the nine case studies, the authors looked closely at the Form 990 reporting. Aside from the functional expense problems, the authors found several other issues.

The case studies turned up two different ways that nonprofits report restricted contributions in Part I of Form 990. Users rely on reported contribution amounts to calculate fundraising-efficiency ratios. Those ratios may lead users to draw false conclusions if not all organizations report contributions the same way. Most organizations report the total of unrestricted, temporarily restricted, and permanently restricted contributions on Line 1d. Yet the authors also found that some organizations report only unrestricted contributions on Line 1d, and report the change in restricted net assets on Line 20 as an “Other Change in Net Assets.” A review of Form 990 and its instructions suggests that this problem arises because the form does not provide separate blanks for unrestricted, temporarily restricted, and permanently restricted revenue. The authors recommend that all contributions be reported on Line 1d.

The authors also found that organizations do not always report donated space and services properly. Organizations that leverage significant amounts of such in-kind donations can appear to have unusually high overhead because their value is excluded from revenue and expenses (unlike donated goods). One organization in the case study had been told by a funder reviewing its Form 990 that its grant would not be renewed because overhead was more than 30% of total expenses. Based on GAAP financials, the organization’s overhead consumed only 12% of total expenses. Donated space and professional services accounted for the difference. Given the importance of overhead reporting to public users, it is important for preparers of Form 990 to include the value of these in-kind donations in the appropriate places. It should appear in Part IV-A Line b(2) and Part IV-B Line b(1), where the Form 990 values are reconciled to the audited financials. This value, plus any other donated services not valued under GAAP, should also be reported in Part VI Line 82(b).

Form 990 reporting for nonprofits composed of multiple, affiliated legal entities also makes overhead and fundraising cost analysis problematic for users. The majority of organizations in the case study, and five of the six with more than $1.5 million in annual revenue, consisted of such conglomerates. Unless the entities are covered by a group exemption letter, the IRS requires separate reporting for each entity. In three of the five largest cases, all or almost all management-and-general and fundraising costs were reported in a single entity’s Form 990, leaving zero or very low nonprogram costs in the other entities. Given such practices, the overhead and fundraising costs of nonprofits with complex legal structures cannot be accurately assessed using Form 990 data. The case studies suggest this is not uncommon among large nonprofits. The best thing for users would be for the IRS to switch to consolidated reporting, such as required by GAAP. Until then, tax professionals can make Form 990 information more accurate and useful by allocating fair shares of management and general and fundraising costs to all reporting entities.

In addition to these problems with Form 990, the case studies also turned up a few gross errors in audited financial statements. The aforementioned organization whose statement of activities erroneously reported zero fundraising costs is one such error. Another organization’s audited financials placed the statement of functional expenses in with the supplemental information, despite the fact that SFAS 117, Financial Statements of Not-for-Profit Organizations, clearly states that it is a required part of the core financial statements for this type of organization.

Problematic Accounting Methods

Upstream from the implausible functional expense numbers reported on Forms 990 and financial statements lie accounting methods that range from inadequate to incorrect. The authors’ national survey of a representative sample of more than 1,500 nonprofit organizations, for example, found that only 25% of nonprofits that receive foundation grants properly classify those proposal-writing costs as fundraising expenses. Only 17% of nonprofits that receive government grants properly report those proposal-writing costs as fundraising expenses. This percentage is probably low, given that many government grants provide only incidental benefits to the grantor. This type of grant is equivalent to a charitable contribution, and proposal-writing costs should be treated as a fundraising expense. When the primary beneficiary is the governmental unit providing the funding (i.e., an exchange transaction of money for services), proposal-writing costs are properly classified as management-and-general expenses.

Personnel costs form the largest expense at many nonprofits, and how those costs are allocated across the program, management-and-general, and fundraising categories can make a huge difference in their program-spending and fundraising-efficiency ratios. In the authors’ survey, barely one-third of nonprofits said they track staff time by functional expense category for each payroll period. Similarly, in the case studies, three of the nine organizations had a paper or automated time-tracking system that was capable of serving as the basis for functional expense tracking. Unfortunately, only one of those three used it for that purpose, and in that case the fundraising staffer charged proposal-writing time to the grant-funded program rather than properly accounting for it as fundraising costs. (Interestingly, this organization had adopted its timesheet system only at the urging of its auditor.)

At the other eight organizations, the vast majority of employees were classified as falling wholly within one of the three functional expense categories. The rest of the staff made a retrospective judgment at year-end about how they had spent their time, and this was used to allocate their personnel costs across the functional categories. The accuracy of such judgments is open to question, and given the emphasis that users place on low overhead and low fundraising costs, it is not surprising that such judgments tended to result in low percentages for management-and-general costs, and especially fundraising costs.

Root Causes

Uniformly, the nonprofit organizations surveyed reported that they used a CPA firm with a regional practice and at least one partner who specialized in nonprofits. Typically, the same audit firm had been used for some years and was familiar with the organization’s structure and finances. Most Form 990s were prepared by a tax professional at the audit firm. Therefore, the authors’ findings cannot be readily explained away by lack of experience at these public accounting firms.

This research suggests that a number of factors contribute to the current state of nonprofit reporting:

• Many nonprofits, especially small ones and those with a majority of donor-restricted funds, have inadequate accounting staff and systems.
• The overemphasis on overhead expenses by donors, funders, and charity watchdogs gives nonprofits a significant incentive to underreport overhead.
• The accounting profession continues to think of the audited financials as the public document, and has not applied its audit and attest standards to the Form 990.
• The accounting profession does not consider the method by which expenses are allocated into functional classifications to be an important audit issue.

Although the first two of these are primarily the responsibility of a nonprofit’s board and management, clearly the public accounting profession is responsible for the latter two, and has a role to play in addressing the first two as well. The underreporting of overhead that is so widespread in the nonprofit sector could not exist without the tolerance of the auditors. And many nonprofits rely on their auditors to advise them on accounting issues.

What to Do

Given the public nature of Form 990, the profession must reorient its thinking and begin to apply the same assurance standards to that document as it applies to audit and attest work. Next, the profession must recognize the importance that users place on expenses by functional classification and make that an important audit issue.

At that point, auditors will have to face both the inadequate accounting systems of most nonprofits, and their incentive to underreport overhead expenses. An auditor’s desire to maintain a positive working relationship with a client is going to be in conflict with the need to maintain public reputation and trust. As the Enron–Arthur Andersen case demonstrated, public reputation and trust are the more valuable assets at risk here. Nevertheless, a graduated response is possible.

In some cases, the auditor will be able to identify misallocations across functional classifications and can require adjusting entries to correct the problem. In other cases, the auditor will recognize weak accounting methods that “happen to” result in underreported overhead, but will be unable to determine the appropriate allocation of expenses after the fact.

The minimum response in these cases is to use the management letter to identify and raise concerns about the quality of accounting for functional expenses. If the same problems persist the next year, the auditor’s next response would be to point out that notes to the financial statements are required to explain significant accounting policies, especially where alternatives exist, and that because functional expense reporting is so important to users, methods of accounting for overhead require footnote disclosure. This is less effective in the nonprofit world than in the corporate world, because the notes are not tied to the public Form 990, but nonetheless, it should succeed in getting management to resolve the problems.

If the problems remain the third year, the appropriate response for an auditor to consider is rendering a qualified opinion letter that narrowly focuses on problematic allocation of costs among functional categories. This action would likely upset the nonprofit client tremendously, but is the best way to protect the auditor’s public reputation and trust.

Unfortunately, there is currently no way for a CPA to offer a similarly qualified opinion of the numbers on Form 990. IRS enrolled agent standards (Circular 230, particularly sections 10.21, 10.22, and 10.33), however, could provide support to a strong position taken by a CPA firm attempting to ensure that the Form 990 return fairly represents the financial position of a tax-exempt client. The AICPA’s Statements on Standards for Tax Services (SSTS) could be amended to clarify that preparation of returns based on audited financial statements should properly reflect audited revenues and expenses with appropriate reconciliation to equivalent figures on the return. Most important, state and national accounting organizations could serve as a strong positive force by educating their members as to the importance of Form 990 and the functional expense allocations used by so many donors.

How the Profession Can Respond

All nonprofit organizations are required by SFAS 117 to report expenses by functional classification. The many users that emphasize program-spending and fundraising-efficiency ratios are relying on these numbers, and assuming that they fairly reflect the activities of the organization. Taken collectively, the findings of the authors’ Nonprofit Overhead Cost Project suggest that this information—often prepared by or attested to by CPAs—is in many cases incomplete, misleading, or inaccurate.

Members of the public accounting profession can respond in three ways:

• Recognize the emphasis that the public places on functional expense classification and make it an important audit issue.
• Recognize that Form 990 is the major public financial document in the nonprofit sector and apply the same standards to it that are applied to audit and attest work.
• Confront the weak accounting and underreporting at nonprofit clients through the graduated response suggested here.

For more of this month's Special Focus on Nonprofit Management, see "Advising Nonprofit Organizations".