How pay for performance has emerged as the new model for federal human resources pay practice and how executives and managers will be challenged to solve its complexities.

 A. C. Hyde

This year’s public Management Fad of the Year proves that a past record of futility does not preclude ultimate triumph. Pay for performance has been proposed, debated, and dismissed for over fifty years in public personnel management. And now, although legal challenges have been filed, pay for performance emerges as the new model for federal human resources pay practice and the cornerstone of federal public management strategy.

But first, a brief aside is in order to restate the premise for this award. For our first issue in the millennium, The Public Manager inaugurated the idea of bestowing a Management Fad of the Year (like Time’s Woman or Man of the Year) on whatever management concept, idea, process, technique, and strategy most dominated the federal public management landscape. Choosing the title “Fad of the Year” was, we explained back then, “…deliberate, hopefully provocative, and not intended as derogatory.”

And with each year’s award, we’ve respectfully repeated the caveat that fads have an impact and are important. As we explained then, “Looking back over the last two decades, one can see a number of ‘dominant’ management concepts—from total quality management and re-engineering to reinvention and downsizing that once commanded the public management headlines. None of these has been relegated to the trash heap. While some have been recycled, modularized, or even redefined, they have become part of our basic management tool kit and our vocabulary.”

And the Nominees Are…

Perhaps it’s the timing or coincidence, but this annual article has come to be more and more influenced by the Academy Awards in making the final selection and bestowing the award. Instead of just launching into the discussion of the merits and prospects for the Management Fad of the Year, it’s become customary to note and say a few words about the other nominees. And since we’ve been at this for five years now, we’re even tempted to suggest that there should be a “lifetime achievement award” to contenders that perennially come up short.

Anyway, this year’s nominees are an interesting group. Supply chain management (SCM) was a new nominee that made quite a splash on the public management landscape, aided by a three-issue double article forum in the Harvard Business Review lauding “truly revolutionary changes” to a core process that is…“accomplished by means of mind-boggling technological innovations, clever new applications of old ideas, seemingly magical mathematics, powerful software, and old-fashioned concrete, steel, and muscle.” SCM also got its own special issue in the California Management Review. Surely SCM will be a contender to deal with in 2005.

Another nominee was network governance. That may not be the right name to give to this rising issue area loosely centered around migrating from hierarchical bureaucratic organizations to more collaborative alliances of public, nonprofit, and private organizations. Much of this growing movement is driven by public policy advocates who have been arguing for a decade that government is out, governance is in. A recent work (reviewed in the last issue of The Public Manager), Goldsmith and Eggers’ Government by Network, uses the subtitle—The New Shape of the Public Sector—to describe the significance of what’s involved here.

The last nominee was one that has come up every year; 2004 was an interesting year for the Government Performance and Results Act (GPRA) and results management. It saw a first wave (well, maybe a wavelet) of public management textbooks and academic journal articles that viewed results management as core management as opposed to a legislated reporting requirement or the usual executive mandate. These works and others promise a more serious discussion of what high performance is and what different models exist for performance results measurement. With a little rebranding, one can see results management being recast as “high-performance management.” At a minimum, it should be given our first aforementioned lifetime achievement award to public management.

And the winner is pay for performance—a rather different kind of choice among the nominees. For one thing, pay for performance is only a part of a larger management reform: civil service modernization or pay modernization. However, it’s the most controversial and significant part.

Furthermore, while viewed as a radical change, it is not described as a reform. Indeed the word “reform” seems to have somehow disappeared from the public management vocabulary, as if it had become politically incorrect. Both the influential U.S. Office of Personnel Management (OPM) White Paper on Federal Pay (2002) and the Volcker II Commission Report (2003) omit the word reform in their advocacy of change. OPM’s White Paper is entitled “A Fresh Start for Federal Pay: The Case for Modernization.” Volcker II’s report is entitled Urgent Business for America: Revitalizing the Federal Government for the 21st Century. The word reform does not appear in any of the fourteen recommendations and the recommendation on pay (number eleven) states simply: “More flexible personnel management systems should be developed by operating agencies to meet their special needs.” One would think from these inferences that all that’s involved here is a little loosening up and getting a new wardrobe. Obviously, public managers know that’s not the case.

The Merits of Pay for Performance in Historical Context

Personnel textbooks have ranted about the terrible effects of public-sector pay systems for years. The compensation chapter of the first edition of Personnel Management in Government in 1978 states, “It is very difficult to convince employees that their pay is fairly arrived at when they have before them on a daily basis other more highly paid employees, who serve not as role models that one should strive to emulate, but rather as glaring examples of the inequities of the pay program.”

That was fairly typical of the “personnel bashing” style of many of the behavorialists critiquing public personnel. What we lacked in data to validate our observations, we made up for verbally. But even that assessment was primarily aimed at the dysfunctionality of the classification system and an evaluation system that made grade inflation at Harvard looked tame. For most of the twentieth century, federal agencies evaluated individuals in terms of merit based on a list of behavioral traits. They developed performance appraisal systems tied to behaviorally anchored rating scales that defined basic levels of service from unsatisfactory to outstanding. Almost inevitably, the result was rating inflation in which 90 percent of employees rated were assessed at above satisfactory ratings, drawing the ire of critics who contended that this facet alone invalidated the entire performance appraisal system.

Performance-based pay was touted as the solution. As Table 1, taken from a recent U.S. Merit Systems Protection Board (MSPB) presentation shows, it began with efforts to provide small annual bonuses (incentive awards and quality step increases). The first big reform steps were taken as key parts of the Civil Service Reform Act of 1978— establishing larger cash awards and managerial bonuses and later limiting the percentage of employees who could get awards. But within a decade, the pay-for-performance movement had basically faltered. In its 1991 assessment Pay for Performance, the National Research Council concluded that there were fundamental underlying difficulties to resolving tensions “between the potential benefits of pay for performance and the reality of federal personnel and compensation systems.”

Balkanization

But the 1990s saw a major shift in federal management strategy. With the advent of performance results budgeting through GPRA and other new public management approaches being advocated—like performance-based organizations—management strategists were arguing that organizations had to relate organizational outcomes or results measures to individual performance objectives. Rather than try to force through a system-wide reform, the management strategy shifted to individual agencies. When the Federal Aviation Administration and the Internal Revenue Service were beset by management crises, part of the solution was to give each agency authority to create its own pay system. As the “Balkanization” of the General Schedule continued into the Bush administration, the Departments of Homeland Security and Defense were given authority to create their own personnel and pay systems. Both made pay for performance centerpieces for “modernization.”

Table 1.

Source: Office of Policy Evaluation, U.S. Merit Systems Protection Board.

Pay-for-performance systems based on results make several key assumptions. First, the organization has readily measurable results that can be transferred from organizational levels to managerial levels and ultimately work groups and individuals. Second, managers can and will make both fair and candid assessments of their subordinates. Third, individuals will be motivated by pay levels that differentiate between those who carry the true workload of the organization and those—recently identified by The Wall Street Journal—as employees who are “actively disengaged” at work. The latter concept is perhaps a new way of looking at the attitudes of those who once might be called “poor performers” but are in fact individuals who see a job as time spent on the job as opposed to time spent doing significant work. Here lies the true test of pay for performance: ensuring sufficient variability in pay so outstanding performers get large rewards, average performers get smaller raises (to “maintain buying power”), and poor performers get no increase.

Not every public-sector manager or employee is going to like the new brand of pay-for-performance management. Indeed, the most recent (2000) MSPB survey of federal employees showed the desire for a good performance rating as a distant ninth on the list of fifteen top motivators for performance—being important to only 10 percent of those surveyed. Of course, the survey can’t separate employee disrespect for current inflated appraisal systems from disinterest. More interesting is the third-place rating given to monetary awards, which appealed to 27 percent of those surveyed. There may be more interest in variable pay levels among public-sector employees than suspected. Of course, pay-for-performance advocates will be quick to point out that the system is supposed to be fair but should not please all employees.

Issues for Today and the Future

Clearly, federal agencies will enter into this brave new world of pay for performance/results with some trepidation. To begin with—as a study by the MSPB has noted—past trends in performance appraisal systems (with the active and strong encouragement of federal unions) have been exactly in the opposite direction. From 1995 to 2000, the number of performance appraisal systems that have only two levels (essentially pass-fail) increased from less than 5 percent to nearly 25 percent. Clearly, increased training budgets to prepare managers and the long lead time to implement pay for performance will be needed.

Another complicating factor will be the pressures of “competitive government.” Managers taking charge of agencies or enterprises, who must compete with contractors or even other governmental competitors, are going to push hard for pay-for-performance flexibilities to reward successful results, pay for new innovations, and possibly most important of all, lower compensation levels when failure occurs.

Another issue will be all the intangibles that go into federal public-service work. Beth Asch’s recent review of “The Economic Complexities of Incentive Reforms” (High-Performance Government, Rand, 2005) reminds public managers that if pay for performance is going to work, gains in productivity have to exceed costs of performance measurement. And performance measurement in federal work is not exactly simple. Much of the work activity performed by federal workers is multidimensional, done in teams, subject to multiple employers and multiple objectives, and longer term. Any number of studies by research psychologists testify that linking pay to individual results has negative consequences—undermining teamwork, levels of cooperation, and even relationships among teams within an organization.

Asch looks at one other critical issue that has to be factored in—promotion rates. One aspect is simply the separate impact on promotions in raising pay and attracting and retaining talent. Equally important, however, is the fact that measurement costs for promotion are lower than pay for performance. As she concludes, “What is required in a promotion system is that the supervisor determine who has the best performance, not the exact level of performance of each employee. It may be considerably easier to determine ranking than to determine the precise level of performance.” One shouldn’t forget, of course, that ranking is usually not allowed in pay-for-performance systems.

Perhaps the most important issue for the future will be making the system fair and transparent. What that really means is another issue. Consider Steve Barr’s October 2004 article in The Washington Post:

Fairness counts, especially when it comes to bonuses. An inspector general review found that the Transportation Security Administration [TSA] handed out bonuses last year to 76 percent of its executives but to only 3 percent of its rank-and-file employees. The TSA also gave time off from work as a reward to an additional 7 percent of the “rank-and-file.” The inspector general’s report called the pronounced tilt toward management “a substantial inequity” and recommended that the TSA “provide more equitable treatment for lower graded employees when making performance award decisions.”

Essentially, an inspector general concludes, 75 percent plus for managers and 10 percent for employees is unfair. Would TSA have been criticized if it had given cash awards to 50 percent of its rank-and-file employees? What kind of management team could TSA hope to retain if it awarded only 50 percent of its managers bonuses? What should the percentages be? In discussing this issue with some union leaders, their perspective on fair and transparent emerges something like this. Transparent means a system you can see right through in order to determine how the decisions were made. Employees need to be able to see if they were treated fairly, and shouldn’t we want employees to be able to see what they need to do better to get a reward next time?

Going beyond that focus on the individual, unions also want to be able to see how various classes of employees did. This is not only a concern for the “protected classes” of various civil rights laws, but also the nonprotected classes. For example, did all the money go to a few occupations or grades at the expense of lower-graded minorities? Did a disproportionate share go to political appointees? The new system should be run in a way that there is nothing to hide. If someone is trying to hide any portion of it, the obvious message is that they are embarrassed by it or can’t explain to others what did happen.

Philosophical Challenge

Beyond these issues lies a final, more philosophical challenge to pay for performance. As public management strategists celebrate the final acceptance of the concept, they will devote more time to working the details and detailing the working parts of the new systems. Certainly, public-sector pay can be reshaped as a powerful tool in pushing public management into a new era. But two costs should be kept in mind. Beth Asch would remind us that public-sector work is more complicated and that the cost of performance measurement must be clearly outweighed by some measurable increase in organizational productivity. The second is a simple reminder of the first tension between the potential of pay for performance and the reality of federal personnel systems noted by the National Research Council in its 1991 report—that of the potential impact on the neutral competence of the public service. As political factions in this country become increasingly partisan, “the centrality of the principle of neutral competence in the public service” may be something we regret subjecting to the economic complexities of compensation incentive reforms. They never had a conversation about the expected outcomes of their joint efforts.

Input from stakeholders is essential for quality outcomes and evaluations. Their experiences and expertise can raise issues that should be addressed when developing outcomes. Their input ensures that relevant information is collected to measure outcomes and complete evaluations. In addition, collaboration supports a cycle of continuous learning because stakeholders are encouraged to use feedback from results measurement to guide their work.

Integrate Results Measurement into Program Planning

It is unfortunate that most evaluation planning occurs toward the end of a funding cycle. This adds to the perception that conducting evaluations is an extra burden for the organization. This perception is accurate because by waiting until the end to plan for an evaluation, the staff must now find time to create evaluation questions, review documents, find program participants to interview, analyze data, and write a report. This burden diminishes when evaluation is integrated into the process of program development and review. If outcomes and evaluation questions are identified when a program begins, it is easier to create useful data collection tools that can be modified throughout the life of the program. There should be a priority on reviewing reporting requirements to see how they can be used to support current evaluation needs. What staff members resent most is an onerous process that asks them to fill out too many forms. When evaluation activities are incorporated into the normal work day, employees understand why information is being collected and how it will be used.

Tell Your Story, Communicate Your Findings

Evaluation helps an organization share its successes and failures. I encourage organizations to develop a dissemination plan that will help them create and take advantage of opportunities to share conclusions from their results management activities. This information should not be left in the form of a final report. Instead, the report content should be “cut and pasted” to inform different audiences about the findings they will find most interesting. Every stakeholder is a different audience. They should review the array of written materials produced in their organizations. Results information can be integrated into newsletters, press releases, Web sites, mailings, solicitations, grant proposals, brochures, and annual reports. Don’t rely upon written materials. Consider presentations at workshops and conferences. If an organization is included in an evaluation with other organizations, ask whether each can have a separate report summarizing its relevant findings. To encourage participation in a multiyear evaluation of two pilot projects, I was asked to prepare shorter and separate evaluation reports for each project. Though everyone was interested in the overall success of the initiative, the executive directors for the pilot organizations saw how they could use the evaluation findings to support their program.

A. C. Hyde is a senior staff consultant with Brookings Executive Education, formerly the Center for Public Policy Education. Special thanks to John Crum of the U.S. Merit Systems Protection Board (MSPB) Office of Policy Evaluation for his recent superb presentation at the 4th annual San Francisco U.S. Government Accountability Office–American Society for Public Administration spring forum on performance and having shaped the issues, but of course not the content, of this article.

References

Innovation Network. “Evaluation in Nonprofit Organizations.” Presentation by Monica Heuer and Veena Pankaj. April 2004.

Patton, Michael Quinn. Utilization-Focused Evaluation. 3rd ed. (SAGE Publications, 1996).

United Way of America. Agency Experiences with Outcome Measurement: Survey Findings 2000, http://national.unitedway.org/ outcomes/files/agencyom.pdf.

———. United Way Outcome Measurement Resource Network, Frequently Asked Questions, http://national.unitedway.org/outcomes/. 2005