A Word to the Wise About Executive Compensation: Lessons from the IRS Colleges and Universities Compliance Project

By    //    Volume 21,  Number 4   //    July/August 2013

Are colleges and universities following best governance practices concerning the compensation of the president and other institutional leaders? For the most part, yes, according to the Internal Revenue Service (IRS), which released a long-awaited final report on its Colleges and Universities Compliance Project in April.

The final report focuses on two aspects of the compliance project: first, unrelated business income (UBI), and second, executive compensation. This article will examine the governance conclusions of the report regarding setting executive compensation.

On balance, the final report offers a picture of colleges and universities that are doing an excellent job in following best practices for governance of the compensation of the president and other senior leaders. The report suggests, however, that a minority of institutions have made errors with respect to how they determine and use comparability data in establishing reasonableness of compensation. In so doing, they subject themselves to heightened scrutiny by regulators and a greater burden of demonstrating that the compensation paid is reasonable. There are action steps to help boards and their institutions manage this risk, which this article will provide.

The IRS Colleges and Universities Compliance Project

In 2008, the IRS commenced its Colleges and Universities Compliance Project, part of a larger IRS effort to review and better understand the largest, most complex organizations in the tax-exempt sector. It sent a compliance check questionnaire to approximately 400 colleges and universities, most of which were four-year, public or private, degree-granting institutions. This questionnaire focused on institutional governance practices, endowment management, employee compensation, and activities that generated unrelated business income during the tax year ending in 2006. The IRS subsequently examined (audited) 34 of these institutions.

AGB and NACUBO Analysis

Upon the announcement of the IRS Compliance Project, AGB and the National Association of College and University Business Officers (NACUBO) partnered on a voluntary parallel analysis of questionnaire data submitted by their members to assess the information collected and offer their own conclusions. The independent accounting firm Ernst & Young conducted the analysis, which was based on responses from 146 of the 400 institutions the IRS surveyed.

The AGB/NACUBO analysis indicated that college and university governance practices related to determining presidential compensation are fundamentally sound. For example, 97 percent of those surveyed reported using conflict-of-interest policies to govern top management officials; 91 percent make their audited financial statements available to the public; and 80 percent of boards use comparative data.

AGB and NACUBO concluded, however, that colleges and universities need to pay more attention to satisfying the “rebuttable presumption of reasonableness” procedure set forth in IRS regulations to determine compensation of chief executives. The rebuttable presumption of reasonableness is a safe harbor found in IRS regulations for demonstrating reasonableness of compensation and is widely considered a governance best practice. Compliance with the process is a matter of public record for private institutions that file the IRS Form 990. The AGB/NACUBO findings showed that only about half of the institutions eligible to use the IRS rebuttable presumption procedure actually did so for their top executives.

The Final Report

The IRS’s final report presents new information based on deeper statistical analysis of data presented in an interim report it released in 2010, analysis of certain questionnaire responses that were not included in the interim report, and IRS examinations of 34 colleges and universities. The results of the examinations produce the most noteworthy information in the report. Because the examinations permit a much closer look at institutional behavior than is possible through answers to a questionnaire, these findings are significant.

It is important to note, however, the limitations of the IRS’s audit findings. The institutions examined constituted an extremely small sample, less than 1 percent of colleges and universities nationwide. Moreover, as the IRS takes care to point out in the final report, these institutions were not randomly selected. The IRS states:

“These institutions were selected for examination because the returns and questionnaires indicated potential noncompliance in these areas. They are not a representative sample of all colleges and universities, and readers should not make assumptions about the UBI and compensation practices of other colleges and universities based on these examination results.”

Notwithstanding that warning, some observers have been quick to assert that the final report suggests widespread noncompliance and alarming findings. But a careful reading of the final report, even with its limitations, makes clear that governance of compensation by colleges and universities is to a large extent compliant with IRS rules and guidance, and therefore worthy of acclaim rather than alarm. At the same time, deficiencies in the compensation-setting process by some institutions were identified that should be avoided or, where necessary, corrected.

Once the IRS made adjustments in its examinations to ensure the proper comparability of data, it analyzed where the compensation of college and university executives fell in the spectrum. The results suggest the full bloom of the “Lake Wobegone Effect,” in which everyone is above average. The IRS found that compensation for most executive positions it analyzed was set in the range of the 75th percentile. Indeed, among the 34 examined colleges and universities, compensation set at or above the 90th percentile was much more common than compensation set in the range of the 50th percentile.

These findings are at odds with those of the Chronicle of Higher Education and of various higher education salary surveys that have reported salaries falling lower on the scale. And as noted, the data reported by the IRS comes from a small sample size. Nevertheless, this data and their skew towards the higher end of reasonable compensation limits are likely to add more fuel to the continuing public debate about salaries for higher education leaders.

A Story of Good Governance Practices

Too easily lost in the statistics provided in the final report is a compelling story of compliance and good governance by colleges and universities.

The IRS found that, to a large extent, the colleges and universities examined followed governance practices in setting compensation for officers, directors, trustees, or key employees (which the IRS commonly refers to as ODTKEs) that satisfied the best practice and legal safe harbor for ensuring that compensation is reasonable. The numbers and good governance compliance practices are impressive:

  • Compensation for 94 percent of ODTKEs at examined colleges and universities was established by using a procedure intended to satisfy the rebuttable presumption of reasonableness.
  • Nearly two-thirds of the institutions examined had formal compensation policies in place. Of those ODTKEs who were not subject to a formal policy during the years examined, more than half were covered by such a policy in the years immediately following.
  • Either the board of trustees or the compensation committee of the board set compensation levels for each ODTKE.
  • Compensation was set by the board according to procedures designed to avoid conflicts of interest. The compensation of each ODTKE was approved in advance by individuals on the board who did not have a conflict of interest. In almost every case, individuals recused themselves from discussions about their own compensation.
  • In nearly every case, the college or university documented the basis for setting compensation before the ODTKE received the compensation.

The IRS determined that about half of the institutions examined used an outside compensation consultant to assist them with setting compensation levels. Current college and university compensation surveys (by position) were the most commonly used tool for determining comparability of compensation levels, whether or not a compensation consultant was used.

Compensation Amounts

The IRS reported on the compensation data it analyzed based on its survey of 400 institutions and examination of 34. Overall, the average and median base salary for the top management official (usually the president) of the colleges and universities examined, both public and private, were as follows:

The average and median total compensation for the top management official of the colleges and universities examined were as follows:

Sports coaches and investment managers received the highest average base salary:

  • Investment managers: $894,214
  • Sports coaches: $884,746

The average compensation of the most highly paid positions (excluding medical doctors) other than officers, directors, trustees, and key employees was as follows:

  • Heads of departments: $279,770
  • Faculty: $215,854
  • Administrative/ Managerial: $381,745

Problems with Comparability of Data for Some

In examining the 34 institutions, the IRS sought to determine whether the comparability data they used was truly comparable—specifically, was it based on similarly situated institutions and functionally comparable positions? The IRS answered that question with observations on what it called “weaknesses” in the comparability data. Specifically, the IRS found:

  • About 20 percent of the private colleges and universities included institutions in their data set that were not similarly situated.
  • Compensation studies provided by the colleges and universities often did not document the selection criteria or offer an explanation as to why the chosen institutions were deemed comparable to the college or university relying on the study and under examination.
  • Many colleges and universities relied on a compensation survey compiled by an independent firm in which their compensation data were included. However, the survey itself was not limited to institutions that were sufficiently similar to be comparable to each other.
  • Compensation surveys relied on for comparability data often did not specify whether amounts reported included only salary or factored in other types of compensation to equal total compensation.

Action Steps

The IRS has stated in the final report that it will be addressing, through education and examinations, the problems that it has identified about the comparability of data. College and university trustees can take action now to follow lessons learned from the final report and to be prepared in the event of an IRS audit. They should:

Make Apples-to-Apples Comparisons. For boards wondering how to avoid the “weaknesses in comparability data” found by the IRS in establishing compensation comparables among institutions, this is one area where IRS guidance is clear. Under the IRS regulations for establishing the rebuttable presumption of reasonableness, a college or university board must obtain and rely upon appropriate data as to comparability before making its compensation decisions.

According to the IRS, relevant information includes compensation levels paid by similarly situated organizations (both taxable and tax-exempt) for functionally comparable positions; the availability of similar services in the geographic area of the institution; current compensation surveys compiled by independent firms; and actual written offers from similar institutions competing for the services of the executive. (See box below for a helpful example from IRS regulations.)

The findings in the final report provide some additional detail as to what the IRS will look for in verifying that true apples-to-apples comparisons are being made. The report states that during the examinations, the IRS looked to factors such as: type, e.g., private or public, liberal arts, college, research university; size, e.g., undergraduate enrollment, faculty size; location, e.g., urban, rural, suburban, geographic region; endowment size; tuition and cost to attend; selectivity, e.g., SAT ranges; and the age of the institution. In the examinations, the IRS found that 20 percent of institutions were not comparable based on at least one of the following factors: location, endowment size, revenues, total net assets, number of students, and selectivity.

Document Survey Selection Criteria. Another key to achieving compliance is to address the IRS’s concern that the noncompliant institutions often did not document the selection criteria for the colleges and universities in the surveys provided and did not offer an explanation as to why those institutions were deemed comparable to the institution relying on the study and under examination.

Particularly where the executive compensation analysis is done internally, and not performed by an independent survey firm, the board must provide adequate information in the record—meeting minutes, memos to the file, research notes, and so on—to enable an auditor or regulator to validate how comparables were determined. In some cases, it is difficult to find a significant number of comparables because of rural locations, specialty of the institution, and so on. When that happens, the board or board committee should explain this problem in the records of its deliberations and the rationale for selecting the institutions that were chosen. Likewise, the record should explain why it was necessary to provide presidential compensation at higher levels relative to the mean, such as when the president has served for many years, has unique academic credentials, has built a particularly strong student body or endowment, or where recruitment has been problematic.

The IRS also determined that many of the colleges and universities it examined relied on a compensation survey compiled by an independent firm in which the data were not limited to institutions that were sufficiently similar to be comparable to each other (and presumably did not document the reason for including this data). Some boards used these survey results without any adjustment; others removed institutions that they determined were not sufficiently comparable.

Conduct an Independent Survey. An independent compensation survey by a competent firm is perhaps the best tool a board can use to properly determine that compensation has been set at fair market value. Such surveys are time-consuming and not inexpensive, so a professional result should be expected. It would be appropriate in selecting a survey firm to ask how the firm identifies comparable institutions and whether it has access to adequate data for the institution’s needs. Boards should not accept such reports at face value without further evaluation. The initial survey report should ideally be prepared in draft form, so that an executive or compensation committee can vet its findings and ensure that the compensation data relied upon is from legitimately comparable institutions.

Salary surveys are also commonly used, although these are not in abundant supply in higher education. Boards relying upon such surveys should ensure that they are current, utilize a data base of meaningful size, contain adequate differentiation of positions surveyed, clearly identify the types and elements of compensation being reported, and include only comparable institutions or at least permit the subtraction of data from noncomparables.

Conclusion

The IRS’s final report on its Colleges and Universities Compliance Project marks the end of its compliance check, but only the beginning of action on its findings. It highlights noteworthy governance successes in higher education when it comes to setting compensation for presidents and other officers and key employees. It also shows the need for additional progress in developing the processes used to identify reasonable compensation. A word to the wise is sufficient.

Bad Facts/Good Facts

The IRS regulations provide a helpful example—using different facts—of a tax-exempt, private university that has obtained appropriate data as to comparability:

“Bad facts.” The university is negotiating a new contract with its president, because the old contract will expire at the end of the year. In setting the president’s compensation at $600,000 per year, the executive committee of the board of trustees relies solely on a national survey of compensation for university presidents that reports that university presidents receive annual compensation in the range of $100,000 to $700,000. The survey does not break down its data based on any criteria, such as the number of students served by the institution, annual revenues, academic ranking, or geographic location. Many members of the executive committee have significant business experience; however, none has any particular expertise in higher education compensation matters.

The IRS concludes in the example that because the survey did not provide information specific to comparable universities, and because no other information was presented, the executive committee’s decision on the president’s compensation was not based upon appropriate data as to comparability.

“Good facts.” The example’s facts are then changed so that the national compensation survey divides the data regarding compensation for university presidents into categories based on various university-specific factors, including the size of the institution (by number of students and revenues) and geographic area. The survey data show that university presidents at institutions comparable to and in the same geographic area as the institution in question receive annual compensation in the range of $200,000 to $300,000. The executive committee of the board of trustees relies on this survey data and its evaluation of the president’s many years of service as a tenured professor and high-ranking university official in setting the president’s compensation at $275,000 annually. The IRS concludes that the data relied upon by the executive committee constitute appropriate data as to comparability.