Final Installment in the Four-Part Performance Management Series: PERFORMANCE BUDGETING in Federal, State & Local Government
Rivenbark, William C., Kelly, Janet M., The Journal of Government Financial Management
The premise of performance budgeting is simple: Resource allocation should follow achievement in efficiency and effectiveness of service delivery. Mostpublic managers would agree that a performance approach to budgeting is useful. In fact, we assert that public budgeting systems often have a performance component. Some systems are formal, encompassing most agencies and departments; others are informal, applying only to selected programs. Budget analysts routinely consider how well a program has performed in the prior and current years of operation when they offer a recommendation forthe next fiscal year. Senior administrators also consider performance, including how programmatic goals align with the organization's mission when they review budget recommendations. We know less about the deliberative process of elected officials, but it seems reasonable to assume that program performance is germane to it.
Yet if we polled these budget officers, executives and elected officials, many of them would say that they are not engaged in performance budgeting because the final budget decision is not determined solely by performance. These public servants have an outcome definition of performance budgeting.
They believe that when high-performing programs are not rewarded with more resources (or when resources are not withdrawn from low-performing programs) performance budgeting has failed. We argue that performance budgeting has been successful when program performance information is part of the budget deliberation process, even if counterintuitive outcomes result. We offer a process definition of performance budgeting, one that recognizes that public officials operate in an environment of constrained resources and public preferences. An attempt to force any determinant budget process into this environment is destined to fail. Moreover, it is undemocratic.1
The roots of performance budgeting are deep. Most budgeting scholars trace them back to the Hoover Commission in the late 1940s, but we found evidence of a performance-based budget approach in the work of the New York Bureau of Municipal Research around the turn of the 20th century. The roots of performance budgeting also are wide. Certain federal agencies, states and localities have been experimenting with some form of performance budgeting since the early 1950s. If you examine the major federal budget reforms (planning-programming-budgeting, management-by-objectives and zero-based budgeting), the link between program and budget outcomes has been an important component in each reform. The reason many public budgeting scholars and practitioners believe that these reforms have failed has been their outcome-based definition of success. For example, if budget allocation patterns did not change significantly after zerobased budgeting was implemented, then it must not have been effective as a budget reform. But what evidence do we have that resources were misallocated prior to the adoption of zero-based budgeting? Similarly, if budget allocation patterns do not change significantly after implementing a performance budgeting program, we might conclude that the existing pattern was consistent with legal mandates, fiscal constraints and public preferences. This is hardly a damning indictment of performance budgeting.
The purpose of this article is to address the current status of performance budgeting in federal, state and local government. However, we judge the success of performance budgeting within the context of the budget process rather than a budget outcome. Therefore, we begin with a process definition of performance budgeting to guide our findings. We conclude with limitations and possibilities of performance budgeting and offer some suggestions for expanding our understanding of performance budgeting in all government levels.
PERFORMANCE BUDGETING: DISCUSSED AND DEFINED
Performance budgeting is an element of performance management, where program performance is relevant to every managerial decision, not just to resource allocation. The performance management toolbox for public organizations contains strategic planning, performance measurement, benchmarking and performance auditing.2 It also can include other management tools that have proved useful over the years, like program budgeting, management-by-objectives, continuous process improvement and total quality management. Imprecise definitions, counting issues and the "presence-absence" problem confound researchers when they try to assess the extent to which governments have embraced performance management. We have already described the way an outcome definition versus a process definition of performance budgeting impacts responses. The counting issue goes back to the tools in the toolbox. If the government has a strategic plan and a performance measurement system, does that mean they are engaged in performance budgeting? If they have all four tools for performance management, does that mean they are engaged in performance management? The answer is no, not until the government is using the tools of performance to make management decisions.
The presence-absence problem is especially thorny and we take special care of it because it applies to what we know about performance budgeting in all government levels. It can take several forms. First, if the government has enacted a law requiring performance budgeting, are they engaged in performance budgeting? No, adoption is meaningless without implementation. Second, if the government has implemented performance budgeting, but is not using results for decisionmaking, are they engaged in performance budgeting? The answer is again no if you accept our process definition of performance budgeting. Consider the following survey question: "Please identify the type of budgeting system used in your organization." The respondent is offered the standard menu of choices-line-item, program, zero-based, performance, other. Because performance budgeting may incorporate elements of all these systems, the "correct" answer is not apparent. Our research indicates that most local governments engaged in performance budgeting use the line-item approach. The presence-absence problem exemplified by a forced-choice question might lead the researcher to the wrong conclusion about the government's use of performance budgeting.
We define performance budgeting as a budget preparation and adoption process that emphasizes performance management, allowing allocation decisions to be made in part on the efficiency and effectiveness of service delivery.3 The first part of the definition suggests that performance budgeting does not begin or end in the budget office. Program managers are the key to performance budgeting. They start the process by identifying service delivery goals, quantifiable objectives and performance measures (outputs, outcomes and efficiencies). Program managers must then use the information produced by performance measures to support their decision-making processes (planning, organizing, staffing, directing, etc.), which constitutes performance management. Other management tools, such as strategic planning, benchmarking and performance auditing, also are used to support performance management. Performance budgeting occurs when program managers use performance results to justify their budget requests and their annual work plans during budget preparation and adoption.
The second part of the definition-allowing allocation decisions to be made in part on the efficiency and effectiveness of service delivery-goes to the weakness of an outcome-based definition of performance budgeting. Program performance informs the budget process, taking its place at the same decision-making table with public preferences and fiscal constraints. In other words, operational accountability must interact with political accountability and financial accountability during budget preparation and adoption. The goal of any meaningful performance budgeting system is the incorporation of performance information in the budget deliberation process even when the final allocation decision is at odds with the performance of service delivery. The notion that operational accountability trumps political and financial accountability during budget preparation and adoption is simply at odds with democratic government. Reconciling multiple accountabilities is the responsibility of program managers, budget directors, senior administrators and elected officials; incorporating a performance component during budget preparation and adoption is simply good management.
The federal approach to performance budgeting must be evaluated from the perspective of its underlying assumption; specifically, that a process that aligns efforts and outcomes yields higher quality services at lower costs. This assumption was evident in the work of the National Performance Review.4 The National Performance Review called for a link between program performance and program funding. However, it was intended to be a performance management reform and not just a performance budgeting reform. The Government Performance and Results Act (GPRA) of 1993 (PL. 103-62), enacted while the work of the National Performance Review was under way, provided for a phased implementation of a performance management system. The GPRA requires program goals and performance measures that address outcomes in addition to outputs, including a link between the program plan and the program budget. TWs link represents GPRA's performance budgeting component. The U.S. General Accounting Office (GAO) updated the progress of federal agencies in linking program goals with program budgets in July 1999 and found modest progress in the agencies it reviewed. However, more progress has been achieved since.
Significant differences exist in how performance budgeting is being approached today in federal agencies as compared to previous budget reforms. Recall the wave of federal budget reforms that started in the early 1960s with planning-programming-budgeting, which was replaced by management-by-objectives and zero-based budgeting in the 1970s.5 These initiatives were based primarily on budget reform, not management reform. The difference between GPRA and prior initiatives is the focus on performance management infrastructure. Federal agencies are required to develop five-year strategic plans, to prepare annual performance plans that contain service delivery goals, quantifiable objectives and performance measures (performance measurement), and to prepare annual performance reports on their progress. If one accepts that a process definition of performance budgeting is appropriate, agencies have made significant progress toward improving their use and value of performance information. The challenge is that additional work is needed before agencies can clearly communicate the relationship between performance, budget requests and resources consumed.6
The U.S. Office of Management and Budget (OMB) responded to this challenge with the creation of the Program Assessment Rating Tool (PART), which is a systematic process for rating federal agencies on program effectiveness.7 PART was used to rate approximately 20 percent of federal programs during preparation of the Fiscal Year 2004 budget and an additional 20 percent for the Fiscal Year 2005 budget. OMB maintains that an agency's PART score only represents one decision-making factor during budget deliberation, suggesting that PART is a diagnostic tool for program improvement as opposed to a tool for automatic funding decisions. PART also provides federal agencies with another mechanism for linking their GPRA plans to the annual budget process.
Another difference between GPRA and prior initiatives is how the reform was structured for implementation. GPRA started with pilot programs to gain experience and to identify successes and failures. Organizations considering the adoption of performance measurement often ask whether to initiate an organizationwide approach or a pilot program approach. We have always advocated a pilot program approach, given the steep learning curve associated with tracking operational accountability. It is important to develop performance management champions within the organization and to expand management reform over time.
A final difference between GPRA and prior initiatives deals directly with the outcome focus of the reform, or the extent to which resources were reallocated after adoption of the reform. Both planning-programming-budgeting and zero-based budgeting were judged failures because the federal budget looked essentially the same before and after their adoption. The verdict on performance budgeting is different, though the evidence is similar. Prior to 2002, the federal budget did not change much. During 2002, budget outcomes changed significantly, but owing much more to the tragic events of September 2001 than to the transforming power of performance budgeting. The GAO acknowledged as much in 2002. Comptroller General David Walker affirmed that linking "resources with results" was still a priority for the Bush Administration in testimony before the House Government Reform Subcommittee on Government Efficiency, but cautioned that the prospect of a long-term fiscal imbalance changed the calculus for budgetary decision-making. The comptroller maintained that performance will not provide mechanistic answers for allocation decisions nor will it eliminate the need for good judgment and political choice.8 Operational accountability had taken its place alongside political and financial accountability.
The record on performance budgeting in state government is not as clear when compared to the federal government. This is not surprising given that most studies on performance budgeting in state government are based on survey research, and many of those survey instruments counted tools and suffered from the presence-absence problem described earlier. Therefore, we approached our assessment of the current status of performance budgeting in state government by defining "adopted," "implemented" and "used." Adoption means a state passed some form of legislation that requires performance budgeting. Implementation means a state created the necessary infrastructure to engage in performance budgeting,at least in larger agencies and programs. Used means the budget preparation and adoption process emphasized performance management, allowing allocation decisions to be made in part on the efficiency and effectiveness of service delivery, our definition of performance budgeting. Again, we are concerned with the budget process, not a budget outcome, given the interaction of multiple accountabilities in all government levels.
One study on performance budgeting in state government reported that 47 states had adopted performance budgeting requirements-31 with legislation requiring performance budgeting and 16 with administrative policy requiring it through budget guidelines and instructions.9 These results provide clear evidence that states are interested in the benefits of considering operational accountability during budget preparation and adoption. However, they do not tell us anything about the capacity for performance budgeting or about the use of performance information for allocating resources.
Another study found that out of the 47 states that had adopted performance budgeting, 29 had implemented the tools for performance management, and therefore had the capacity for performance budgeting.10 A key to this study is that it defined performance budgeting as a process that requires strategic planning (agency mission, goals and objectives) and that requests performance data on programmatic outcomes. In other words, the definition is focused on budget process rather than on budget outcome. Other findings from the study were that nine states did not have the capacity for performance budgeting and that the remaining states could not agree on whether implementation of the tools had actually taken place. That knowledgeable respondents could not agree on whether performance budgeting had been implemented highlights the definitional problems that have hindered research on performance budgeting over the years.
We now turn to another state-level study to assess the extent to which performance management is used during budget preparation and adoption for the allocation of resources, our definition of performance budgeting." The study's authors defined performance budgeting as performance funding, allocating or distributing a percentage of the appropriated funds contingent upon the assessment of performance measures. Note that the definition for performance funding includes a percentage of the appropriated funds, allowing allocation decisions to be made in part on the performance of service delivery and also on other factors. The study found that the following 10 states used performance budgeting for resource allocation: Arkansas, Hawaii, Illinois, Louisiana, New Hampshire, New Jersey, Texas, Virginia, Washington and Wisconsin.
At first glance, it may seem discouraging that only 10 states are engaged in performance budgeting based on the distinction we drew between adoption, implementation and used. However, it suggests that 20 percent of the states are using their capacity for performance management to enhance their budget preparation and adoption processes, allowing allocation decisions to be informed by the performance of service delivery. Similar to the federal government's recent approach, these 10 states are making progress toward performance budgeting based on the benefits of performance management. In this regard, we feel that the outlook for performance budgeting in state government is promising.
Discussing the status of performance budgeting in local government is by far the most difficult of the three government levels. Many of the pioneers in performance measurement can be found in local government, including San Diego, CA, Clark County, WA, Dallas County, TX, and Portland, OR.12 We also would add Charlotte, NC, Fairfax County, VA, and Phoenix, AZ, to the list of measurement pioneers. Case studies from these and other local governments have provided the most detail on performance budgeting in local government, but their experiences may not be comparable to other jurisdictions. As for survey research, the same problems we described with survey research on performance budgeting in states are magnified in local government.
Most of the research on local government counts the tools of performance management (strategic planning, performance measurement, performance auditing, etc.), but seldom deals specifically with performance budgeting. Additionally, surveys that count the tools of performance management have generally been limited to counties with populations of 50,000 and above and to municipalities with populations of 25,000 and above. There has been a tendency to ignore smaller jurisdictions in these surveys despite the fact that most municipalities are quite small. A generous assessment of the state of knowledge concluded that, "different research approaches using different definitions and applied to different samples have yielded a wide range of estimates."13 In other words, we are not sure what we know, but what we have learned is from larger counties and municipalities.
We conducted our own survey in late 2002 to determine the commitment to performance budgeting in municipalities with populations of 2,500 and above, stratified for size distribution.14 Our instrument tried to surmount the problems with distinguishing between adoption, implementation and use by asking how performance measures inform budget preparation and adoption. We found that approximately 23 percent of the respondents had formally adopted a performance budgeting system, requiring program managers to submit performance measures as part of the annual budget process. Another 8 percent responded that programs often submit performance measures during the budget process even though they are not required to submit them. We did not frame the question like the studies of state government, which focused on whether legislation had been passed requiring performance budgeting. Performance budgeting at the local level is more administratively driven as opposed to being required by a mandate from elected officials.
The next line of inquiry focused on implementation, determining if municipalities had created the necessary infrastructure to engage in performance budgeting. We found that 30 percent of municipalities tracked output measures of performance, 23 percent tracked outcome measures of performance and 17 percent tracked program efficiency measures. These findings are consistent with other studies on performance measurement in municipalities, especially the likelihood that municipalities are more likely to collect output measures over any other form of performance measure. Output measures are the least expensive and easiest to create and track over time. Outcome measures, which provide feedback on the quality of services being provided, follow output measures in terms of frequency. Recall that one study on state government defined performance budgeting as a process that required performance data on programmatic outcomes. According to that definition, 23 percent of municipalities with populations of 2,500 and above have the capacity to engage in performance budgeting. It is not surprising that efficiency measures, the relationship between input and outputs, were adopted only by 17 percent of the respondents. Efficiency measures often require cost accounting, an area of financial management rarely emphasized in local government.
The goal of performance budgeting, however, is to ensure that performance is part of allocation decisions. Approximately 12 percent of the respondents reported that while budget requests are evaluated on a variety of issues, performance data are usually part of every allocation decision. If we generalize this finding across the 6,981 municipalities with populations of 2,500 and above, it means that approximately 838 of municipalities in the United States are engaged in performance budgeting when we define it as a budget process rather than a budget outcome.
Compared to federal and state government, the outlook on performance budgeting in local government is not as promising. A plausible explanation for this goes to the reason why so many surveys of municipalities are skewed toward larger jurisdictions even though an overwhelming majority of localities in the United States have population of less than 25,000. Performance budgeting requires capacity that smaller jurisdictions often do not have. Performance budgeting requires performance measurement, which demands technology, analytical skills, staff support and periodic training. While it makes sense to go looking for progress where one is most likely to find it, the need to monitor management and budget innovation in smaller municipalities is just as great as in larger municipalities.
LIMITATIONS & POSSIBILITIES
Our research suggests that performance budgeting is not being embraced as a stand-alone budgeting technique and is not the purview of the budget office, as were previous budget reforms. Since the early 1990s, the three levels of government have focused on adoption (requirement for performance budgeting) and implementation (infrastructure for performance budgeting). We now have evidence that some organizations are working on using performance data to expand their budget processes. But remember, performance budgeting is not determinate. We will be perpetually disappointed if we define the success of performance budgeting by how much high performers gain and low performers lose. In the classic example of the dilemma, rising crime rates do not signal the need to withdraw resources from the police department.
For all its promise, performance budgeting is limited as to what it can accomplish in any government. We offer these:
* Performance budgeting cannot take the politics out of budgeting.
* Performance budgeting cannot reduce the influence of interest groups.
* Performance budgeting cannot refocus citizen priorities.
* Performance budgeting cannot solve a fiscal crisis.
* Performance budgeting cannot prevent poor managerial decisions.
Notice how the first three limitations deal with political accountability. In a democratic government, performance does not trump the need for elected officials to be accountable to their constituents. The fourth limitation addresses financial accountability. No matter how efficient programs of service delivery are, the demand for services typically exceeds the supply. The last limitation is the most meaningful from an administrative perspective. Given that performance budgeting cannot prevent poor managerial decisions, why should program managers continue to invest their time and energy in operational accountability other than it represents good management? We offer four reasons why managers should use performance budgeting to enhance operational accountability:
* Performance budgeting can align service priorities and service spending.
* Performance budgeting can add an information dimension to budget deliberations.
* Performance budgeting can motivate program managers and employees by recording their progress toward service delivery goals.
* Performance budgeting can help demonstrate to citizens that their public service providers are interested in improving service quality.
The basic model for managing any public or private program is aligning service efforts with service accomplishments. Service efforts represent resources like dollars, full-time equivalent positions, equipment and technology. Service accomplishments represent outputs, outcomes and efficiencies as identified from the program's mission statement, service delivery goals and quantifiable objectives. Without this basic understanding of how service priorities are aligned with service spending, it becomes very difficult to communicate meaningful budget requests. Of course, neither the limitations nor the possibilities listed are exhaustive. They simply demonstrate why performance budgeting is so attractive to public managers engaged in professional standards of service delivery, and what those public managers need to realize before promoting performance budgeting in their government.
We conclude with a few thoughts on how we can further our understanding of performance budgeting in federal, state and local government. First, researchers must move beyond limited scope surveys and one-jurisdictional case studies to longitudinal, multi-jurisdictional case studies. This approach will help us develop a more complete understanding of how performance data are converted into information useful for decision-making and how that information is used during budget preparation and adoption.15 The drawback to multijurisdictional cases studies, however, is that they are often prohibitively expensive. The most ambitious effort to date, the Governmental Accounting Standards Board's (GASB) research on performance measurement in local government, was made possible through a large, multi-year grant from the Sloan Foundation.16 Second, we need a common language for performance budgeting. As evidenced by this article, not differentiating among terms like adoption, implementation and use complicate discussions of any management tool in public organizations. An inconsistent language also makes training difficult because program managers across different organizations and within the same organization often are exposed to very different approaches to performance budgeting. The need for professional standards in regard to language is being addressed by organizations like GASB. The recent initiative by the Association of Government Accountants to launch the Certificate of Excellence in SEA Reporting also should assist with establishing professional standards.17
Finally, we offer a rather revolutionary proposal: Performance budgeting should be subjected to program evaluation within the organizations that embrace it.18 If the goal of performance budgeting is a budget preparation and adoption process that emphasizes performance management, we should evaluate our progress toward that goal. The results of the evaluation also should include whether the marginal benefits of performance budgeting outweigh the cost of including performance information in the budget process. Accountants routinely weigh the cost of acquiring information against its value for decision-making; we suggest a similar approach. Of course, we offer this idea with some trepidation. Operational accountability is essential to public organizations to maintain the public trust. On the other hand, we think that previous budget reforms may have been judged failures because the improvements in managerial capacity they provided was overpowered by the expectations resulting from an enthusiasm for reform for reform's sake. Performance budgeting is particularly susceptible to this enthusiasm because it is so intuitive and appealing. A process whereby performance budgeting is evaluated like any other program would help a government identify the costs and benefits of the process, and facilitate process improvement, just like any other program.
1. Joyce, Philip G., "Appraising Budget Appraisal: Can You Take Politics out of Budgeting?" Public Budgeting & Finance, Winter 1996, pp. 21-25.
2. For information on performance auditing, see Daniel E. Schultz and Richard E. Brown, "Performance Auditing in Ohio: A Customer Service Orientation." Journal of Government Financial Management, Summer 2003, pp. 58-63.
3. Kelly, Janet M. and William C. Rivenbark, Performance Budgeting in State and Local Government, M.E. Sharpe, 2003.
4. National Performance Review, From Red Tape to Results: Creating a Government that Works Better and Costs Less, Government Printing Office, 1993.
5. Henry, Nicholas, Public Administration & Public Affairs, Prentice Hall, 1992.
6. For more information on the Government Performance and Results Act, see Jonathan D. Breul, "The Government Performance and Results ActlO Years Later." Journal of Government financial Management, Spring 2003, pp. 58-64.
7. Executive Office of the President, Office of Management and Budgeting, M-02-10, Program Performance Assessments for the FY 2004 Budget, July 16,2002. For more information on PART, see www.whitehouse.gov/omb/.
8. Testimony of Comptroller General David M. Walker before the Government Reform Subcommittee on Government Efficiency, Financial Management and Intergovernmental Relations, Rules on Legislative and Budget Process, House of Representatives (GAS-02-1106T), General Accounting Office, 2002.
9. Melkers, Julie and Katherine Willoughby, "The State of the States: Performance-Based Budgeting Requirements in 47 out of 50." Public Administration Review, January/February 1998, pp. 66-73.
10. Willoughby, Katherine G. and Julie E. Melkers, "Implementing PBB: Conflicting Views of Success." Public Budgeting & Finance, Spring 2000, pp. 105-120.
11. Jordan, Meagan M. and Merl M. Hackbart, "Performance Budgeting and Performance Funding in the States: A Status Assessment." Public Budgeting & Finance, Spring 1999, pp. 68-88.
12. Wendland, Timothy A. "Measuring Governmental Performance: The Accountants Weigh In." Journal of Government Financial Management, Fall 2003, pp. 56-60.
13. Poister, Theodore H. and Gregory Streib, "Performance Measurement in Municipal Government: Assessing the State of the Practice." Public Administration Review, July/August 1999, pp. 325-335.
14. A survey questionnaire was mailed to a stratified sample of 1,143 municipalities with populations of 2,500 and above in September 2002. The potential respondents were asked to return their completed questionnaires in self-addressed, stamped envelopes that were included in the initial mailing. Postcards were mailed two weeks after the initial mailing to the entire sample of 1,143, encouraging the potential respondents to participate in the study and giving them the option of downloading another copy of the survey from an identified website if their original copy was misplaced. This methodology produced 346 usable survey responses, providing a response rate of slightly over 30 percent.
15. For more information on case studies, common language and evaluation of performance measurement implementation research, see Howard A. Frank and Jayesh D'Souza, "Twelve Years into the Performance Measurement Revolution: Where We Need To Go in Implementation Research." Paper presented at the Southeastern Conference for Public Administration, Savannah, Georgia, October 2003.
16. For more information on the Governmental Accounting Standards Board's performance measurement initiative, see www.gasb.org.
17. For more information on the Association of Government Accountants' Certificate of Excellence in SEA Reporting, see wipw.agacgfm.org and page 17 in this issue.
18. Frank, Howard A. and Jayesh D'Souza, "Twelve Years into the Performance Measurement Revolution: Where We Need To Go in Implementation Research." Paper presented at the Southeastern Conference for Public Administration, Savannah, Georgia, October 2003.
William C. Rivenbark, Ph.D., and Janet M. Kelly, Ph.D.
William C. Rivenbark, Ph.D., is an assistant professor in the School of Government at the University of North Carolina at Chapel Hill. He specializes in local government organization and administration, focusing primarily on performance and financial management. He has published in numerous academic and professional journals and is the co-author of Performance Budgeting for State and Local Government, M.E. Sharpe, 2003.
Janet M. Kelly, Ph.D., is the Albert A. Levin Professor of Public Service in the Maxine Goodman Levin College of Urban Affairs at Cleveland State University. She specializes in intergovernmental fiscal relations, performance management in state and local government and financial administration for nonprofit organizations. She has published in numerous academic and professional journals and is the co-author of Performance Budgeting for State and Local Government, M.E. Sharpe, 2003.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Final Installment in the Four-Part Performance Management Series: PERFORMANCE BUDGETING in Federal, State & Local Government. Contributors: Rivenbark, William C. - Author, Kelly, Janet M. - Author. Journal title: The Journal of Government Financial Management. Volume: 53. Issue: 2 Publication date: Summer 2004. Page number: 50+. © Association of Government Accountants Winter 2008. Provided by ProQuest LLC. All Rights Reserved.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.