Rewarmed or Reformed? Minnesota Relaunches Performance Based Budgeting

Advocates say it's a recipe for productivity and efficiency improvements in government and a foundation of long-term budget sustainability. Skeptics says it is politically naive and administratively utopian with potential hazards. Experience says it's best to ratchet way back on any transformational expectations. November-December 2012.

Rewarmed or Reformed? Minnesota Relaunches Performance Based Budgeting

In rancorous debates over government spending it’s the place of common ground. Amidst the ideological battles it’s the Kumbaya element of Minnesota public finance – a recommendation jointly advocated by the state’s preeminent progressive think tanks, conservative think tanks, good government organizations, business organizations, opinion leaders and most everyone else.

The “it” is performance-based budgeting or “budgeting for outcomes.” 1 Advocates say it is the foundation for productivity and efficiency improvements in government and a cornerstone of long-term budget sustainability. Skeptics say it is politically naïve and administratively utopian with unappreciated potential consequences and hazards. Experience says that regardless of your perspective, it’s best to ratchet way back on any transformational expectations.

The outcomes are coming

The basic concept of budgeting for outcomes is fairly simple. Broadly identify the outcomes or results which should be achieved by government and develop performance measures across all government agencies and programs in support of these outcomes. Then use performance against these measures to guide budget-decisions and determine resource allocation in government. Since performance measures would ideally capture both cost and outcomes, the practice offers the promise of improving both efficiency and productivity in providing public services. Such an approach contrasts with traditional public budgeting, which too-often focuses predominantly on inputs but often includes little measurement of the results or policy outcomes achieved.

The conceptual appeal is obvious, which is one reason why the practice of performance-based budgeting is not new. Under pressure to improve public sector services and attain fiscal sustainability, a wave of enthusiasm for this practice took hold around the world in the 1990s. By 2005, nearly 75% of OECD countries had introduced some form of performance information into the budgeting process. 2 Sub-national governments also began to take notice and experiment with the idea. In the United States, Texas was one of the first states to embrace the practice requiring performance measures for all state agencies in its 1993 general appropriations bill.

Considering today’s budgetary pressures and the enthusiastic support coming from almost every political and organizational direction, it’s not surprising that the Dayton administration has embarked on building a framework for performance based budgeting for the 2014-15 biennium and beyond. According to MMB Commissioner Jim Schowalter, the department has identified nine primary outcomes and is constructing a “dashboard” of several dozen numerical performance markers to guide the governor’s proposals for allocating state resources in fiscal 2014 and beyond.

If at first you don’t succeed…

If this seems familiar, it’s because Minnesota has flirted with performance-based budgeting before. Its roots can be traced to at least 1969 when the legislature said future budgets should be stated “in terms of accomplishment.” In 1975, the Department of Finance encouraged agencies to develop and report performance measures in their budget instructions. In 1992, the publication of Minnesota Milestones provided the richest environment to date for pursuing performance based budgeting. The plan contained 20 goals and 79 performance indicators addressing the state’s well-being and indirectly the performance of government. Department of Finance budget instructions to agencies in 1993 stated that Minnesota Milestones along with an agency’s own performance measures would “provide the basis for budget decisions.”3

Through these early years, the promise and most of the expected benefits of performance budgeting never materialized. A 1993 report from the Governor’s Commission on Reform and Efficiency concluded that the state’s budget system was not sufficiently oriented toward agency missions and program outcomes. A subsequent 1994 program evaluation report by the Office of the Legislative Auditor on performance budgeting provided important insights into what was and was not happening:

  • “For the most part, the performance information presented in the Governor’s 1994-1995 budget proposal had little impact on key discussions or decisions by the executive and legislative branches.”
  • “We found that there were very limited discussions of performance in executive branch budget meetings convened by the Department of Finance and no instances in which the department reduced agency base funding on the basis of performance information.”
  • “Legislators and their staff told us that the performance based budgeting approach had little impact on public discussion of the proposed budget or legislative decisions.”
  • “Agency staff said agencies have viewed performance measurement as an idea that was not relevant to decision makers and would not outlive each existing administration.”

Why did it fail to deliver? The OLA flagged a number of issues including insufficient training in performance measurement, the time needed to develop consensus and collect reliable data, political distrust, lack of agreement on agency mission and goals, and lack of legislator confidence in the quality of the metrics. Theoretically at least, all of these concerns might be addressed with patience, time, and additional investment leading the OLA to conclude the development of a performance-based budget “should be considered a multi-year process.”

But Minnesota is not alone in experiencing results that failed to live up to the ambitions and expectations. A 2000 Governmental Accounting Standards Board (GASB) survey of state budgeting officials (both executive branch and legislative branch) found that only 7.5% believed that performance measures were “very effective” or “effective” in changing budget appropriations.4 Among the findings from GASB’s state-specific case studies:

  • “interviewees…saw little impact on re source allocation” (Wisconsin)
  • “not one individual provided an example of a direct relationship between the use of performance measures and funding levels” (Arizona)
  • “there has not been a measurable result in terms of dramatic allocation changes or trends realized” (Illinois)

At the federal level a GAO survey of 1,300 agency managers and supervisors found that only 21% of respondents reported performance measures were used to a “great” or “very great” extent in developing agency budgets. Similarly, only 20% of respondents indicated that performance measures factored into actual funding decisions.5

Other research findings conclude that even when there is evidence that budget allocations have been affected by performance evaluation, success in meeting a performance target did not necessarily translate into better outcomes. For example, in a series of over 10,000 paired observations of Texas’s very mature performance budgeting system over a 10 year period, only 52% of the key performance measures reflected in the state’s budget process showed improvement over time.6

Managing Expectations

“Proponents of performance measurement in government need to reevaluate how they sell this stuff. …The fact is governments are never going to achieve anything like what performance measurement proponents view as textbook performance-based budgeting.

- “Management and the Holy Grail” Jonathan Walters, Executive Editor, Governing, May, 2000

Such results suggest that the claims and expectations associated with performance-based budgeting demand closer scrutiny. There is unambiguous support in public administration for the practice of stronger performance measurement and results tracking within government. Substantial amounts of research have documented benefits from creating and using measures to improve internal processes and agency performance – often with significant budget implications within a particular agency.

However, it’s the further extension of this practice to support comparative program evaluation and ultimately the prioritization and allocation of resources across government that gives many practitioners and scholars significant pause. In other words, there is a clear endorsement for performance-informed budgeting but considerable skepticism for performance-based budgeting.

Why the doubt? In addition to all the practical issues flagged by the Minnesota OLA in 1994, critics note that measurement remains inherently imperfect and subjective. Given the political context of budgeting, those with principled beliefs concerning the appropriate roles and activities of government may not be receptive to information suggesting appropriations that conflict with those beliefs. Causality between performance and outcomes is far from certain and may have much more to do with factors completely outside of government’s control than with spending decisions. In many core functions of government it is extremely difficult to specify intended outcomes and their relationship with outputs and activities. Moreover, there is inherent ambiguity surrounding the budgetary implications of performance measurement. Does poor performance mean a program should be dismantled or reformed, or does it instead deserve additional resources to better pursue its mission? If a program performs well, does that mean it deserves more funds or does it show that it is sufficiently supported? Is it possible the same results could actually be achieved at lower cost, and will performance budgeting prevent us from finding that out?

Many public administration scholars actually see government-wide resource allocation on the basis of program performance as a “risky undertaking.”7 They express concern that imperfect output measures “may focus effort but may send bureaucrats marching in the wrong direction” leading to an erosion of both quality and outcomes.8 According to an International Monetary Fund working paper reviewing the empirical literature on performance budgeting, this is an argument “which appears to have considerable basis not only in the academic literature but also to reflect the lessons of history.”9 Still others express concern about creating incentives for data manipulation and misreporting and the degree to which the energies of civil service are sapped by the infrastructure needed to support this endeavor.

But perhaps the most important reservation lies in the potential disconnect between the performance budgeting system and the realities, behaviors, and motivations within government. Human and financial resource allocation in government is already constrained by a variety of institutional and statutory forces making any change based on performance evaluation difficult. But equally important to success is the existence of structural incentives for people to “do the right thing” at the micro level.

As one scholar has put it, “if organizational conditions discourage performance, the budget will not be oriented to performance … it is futile to reform budgeting without first reforming the overall managerial framework.”10 There’s a real danger that overemphasizing a sophisticated budgeting process will distract policymakers’ and citizens’ attention from more serious matters and fruitful investigations into internal issues having a far greater impact on the effective and efficient delivery of public services.

Transparency, Not Allocation

In 1995, the Brandl/Weber “Agenda for Reform” stated that “stressing outcomes not inputs” and “installing the latest management fads… are condemned to have but fleeting effect.” The report was highly critical of the notion that if we only had better internal management and planning processes we could neutralize the role of politics (and the self-interests guiding politics and bureaucracies) in establishing priorities within government and allocating resources. It’s doubtful any other report better captures the uphill climb outcome-based budgeting faces in this state.

What seems clear to us is that the primary benefit of performance indicators is the transparency benefits they provide to educate citizens and policy makers and inform, not determine, resource allocation. Ratcheting way back on the extraordinary claims, and expectations surrounding this concept is perhaps the best way to ensure something beneficial and useful actually results from the effort.

  • 1 We use the terms “performance-based budgeting” and “budgeting for outcomes” interchangeably. They build on identical ideas even though some may quibble technically that they are not exactly the same. Since the academic literature refers to this general approach as “performance based budgeting” we do as well in this article.
  • 2 “Performance Information in the Budget Process: Results of the OECD 2005 Questionnaire” OECD Journal on Budgeting, 5 (2):87-131
  • 3 “Performance Budgeting” Minnesota Office of the Legislative Auditor, 1994.
  • 4 Performance Measure at the State and Local Levels: A Summary of Survey Results, GASB 2002.
  • 5 Performance Budgeting: Initial Experiences Under the Results Act in Linking Plans With Budgets, GAO/AIMD/GGD-99-67
  • 6 Using a Performance Budgeting System: Lessons from the Texas Experience, John C. Stennis Institute of Government, Mississippi State University
  • 7 “Reward the Winners, Punish the Losers? Performance Budgeting Reconsidered” Dr. Miriam Allem, EIPA Scope, 2007 Volume 3.
  • 8 “Assessing the Performance of Performance Standards in Public Bureaucracies” American Economic Review Vol 87, No. 2
  • 9 “Does Performance Budgeting Work? An Analytical Review of the Empirical Literature” International Monetary Fund, November 2005.
  • 10 “The Performing State: Reflections on an Idea Whose Time Has Come but whose Implementation Has Not” OECD Journal on Budgeting 3 (2):71-103