DUE 4/4

13

345

PerformanCe budgetIng

Political rationality is the fundamental kind of reason, because it deals with the preservation and improvement of decision structures, and decision structures are the source of all deci- sions. . . . There can be no conflict between political rationality and . . . economic rationality, because the solution of political problems makes possible an attack on any other problem. . . . In a political decision . . . action is never based on the merits of a proposal, but always on who makes it and who opposes it. . . . Compromise is always a rational procedure, even when the compromise is between a good and a bad proposal.

(Paul Diesing 1982, 198, 203–204, 232)

Government performance needs to be viewed from the perspective of the organic wholeness of a political system in which the public, private, and nonprofit sectors work together to create the uniqueness of a given political community. This view emphasizes the synergistic influence of history, social institutions, and culture in creating a shared system of values, agreement on governance processes and structures and the respective roles that the private and nonprofit sectors play in the creation of the common good.

(Douglas F. Morgan and Craig Shinn, 2014, 6)

As the opening epigraphs for this chapter illustrate, what counts for performance is very much a contested issue. Like the Great Wall of China or a simple ax, it depends on whose perspective is being used to test its performance. The opening teaching case for Part III of this book illustrates the concrete ways in which differences in perspective take on organizational expression in the budgeting process. Budget and Finance director Spiro Augustine understands performance im- provement as a process of organization and process reform. City department and program directors likely see the new performance effort as conflicting with their current, often successful, efforts to improve productivity. City budget analysts wonder what the new mayor’s performance initiative means for the procedures and requirements of budget preparation, and social service advocates and rank-and-file city staff readily interpret the new initiative as an excuse to undertake more and deeper cutbacks. While a performance orientation to budgeting may be in the eye of the beholder, there is general agreement that performance budgeting provides an analytic perspective and ap- proach to budgeting that draws on data and information derived from performance measures and comparisons. Our primary purpose in this chapter is to build a basic understanding of performance measurement and explain how it can integrate into the budgeting system and support the prepara- tion of a budget request through performance budgeting.

Our secondary purpose is to demonstrate that a performance approach to budgeting is neces- sarily a political question: What kind of information should be collected and how should it be assessed to improve the public good? The etymological origins of the term performance reflect the broadening perspectives by which we have come to judge the performance of government and its various activities. Originally, the term performance was attached to carrying out a promise or

346 EXPENDITURE FORMATS FOR DECISION AND CONTROL

duty, but over time it has broadened to include anything that is performed, especially entertainment. The democratization of the term performance has important implications for public budgeting, where the terms budgeting and performance have become increasingly linked. By connecting the two, performance becomes fundamentally a political question of deciding what people value and whether they are getting their money’s worth.

The discussion of performance in this chapter is arranged in four parts. In the first part, we describe the performance budgeting technique and how it relates to our earlier discussions of PPBS and program budgeting. We give examples to illustrate what it looks like in practice. In the second part, we review the purpose, history, and contextual setting that creates an enabling environment for a performance orientation to budgeting. In the third part, we discuss some of the enabling conditions for making performance budgeting work in practice. In the fourth part, we assess the strengths and weaknesses of performance budgeting and discuss strategies for building a performance orientation into a range of other activities that government undertakes.

ThE TEChNIqUE OF PERFORMANCE BUDGETING

Performance budgeting is a term that applies broadly to any number of budgeting strategies and formats that attempt to incorporate the measurement of results and costs as an important consider- ation in the allocation of budget resources. Unlike PPBS and program budgeting, where the goal is to assemble and budget to program objectives or goals, performance budgeting systematically incorporates performance and cost measurement information into the budgeting process and uses this information to allocate scarce public resources. You can have program budgeting without performance measurement and you can have performance measurement without program budget- ing. For example, provision of services to juveniles can be organized into programmatic sets of distinct activities with clear and separable objectives and overall costs without a commitment to systematic measurement. By contrast, you can have systematic measurement of process costs and, say, apply that to the number of miles of road paved by a department of transportation without necessarily having these activities organized along programmatic lines. Or you can measure the number of HIV cases successfully treated with refined cost information without having a distinct and separate HIV program. In short, if outcomes and process cost data are measured but are not taken into consideration in the allocation of public resources, you may have some kind of performance system; however, you do not yet have performance budgeting. For example, many jurisdictions may have a strategic plan or a community visions document that sets forth the aspirations and the priorities of a community, but unless these documents play an important role in determining the level of resources to allocate to various programs and agencies, the jurisdiction does not have performance budgeting.

Examples of Performance-Based Budgeting

Exhibit 13.1 presents a simple performance budgeting example that builds on the community cor- rections example we used at the beginning of the previous chapter to illustrate how the planning, programming, budgeting system (PPBS) works. Extending the PPBS to include performance indicators makes performance budgeting the natural and final step in the preparation of goals and objectives. While performance measures can be used without systematically linking them to a causally linked hierarchy of goals and objectives, this is not the preferred approach when doing performance-based budgeting. In the community corrections example, notice the need for several performance measures for a given objective and the need for several objectives for each goal. All performance measurements should include cost information, since few activities undertaken to achieve objectives are cost free. Performance measures can also relate to published standards.

PERFORMANCE BUDGETING 347

Using standard measures allows the comparison of program performance across jurisdictions and organizations with respect to efficiency, effectiveness, and other performance indicators. The number of recommend police offers per 1,000 residents or the recommended criteria for replace- ment of motor vehicles are examples of these standard-based forms of measurement. There may be valid explanations that justify lower performance standards, but inefficient process performance is often the root cause of high costs for reduced outputs.

Compare the example in Exhibit 13.1 to the program example in the previous chapter on program budgeting. The difference between the two examples illustrates the difference between program budgeting and performance budgeting. In program budgeting, we identified the total costs of meeting goals and objectives. The cost figure often represents the historical cost of providing the desired services. But in Exhibit 13.1, we add detailed performance measurements and criteria as the basis for allocating scarce public resources. When performance measures are added to a set of goals and objectives, there are several important considerations to keep in mind:

• One needs to develop measures of both input and output in order to have a comprehensive understanding of performance. For example, if your goal is to pave additional miles of road over the current fiscal year, it is not enough to know what the outcome performance measures are going to be (for example, 1,000 miles of road paved at x standard of asphalt). A more complete answer to a performance-based orientation is that you are able to assess what you are accomplishing in relationship to what it will cost to do the additional work.

• The cost of accomplishing an identifiable set of activities needs to be translated into the im- plications for next year’s budget. For example, if you need two additional field employees to achieve your projected work objective of paving an additional 1,000 miles of road, you need

Exhibit 13.1

Community Corrections Budget for FY5

Goal: To rehabilitate clients on parole and probation so they can return as productive members of society.

1. Objective: To increase FY2 employment references of clients by 10 percent over FY2 with no budget increase over what was allocated for these activities in FY1.

Performance Measures:

a. Number of FY2 employment references are at least 10 percent more than FY1, and b. at a total cost that is equal to or less than what was allocated in FY1

2. Objective: To double total mental health consultations experienced by clients in FY2 compared to FY1 at a cost of 1.0 FTE Mental Health Nurse.

Performance Measures:

a. Total mental health consultations in FY2 is double the total mental health consultations in FY1 and b. costs no more than $65,000

3. Objective: To limit FY2 parole and probation violations to 15 percent of Corrections clientele by introducing a bracelet monitoring program at a cost that does not exceed the total budget allocation expended in FY1 for parole and probation monitoring.

Performance Measures:

a. FY2 parole and probation violations are 15 percent or less than the FY1 parole and probation violations and b. the total costs do not exceed what was allocated in FY1 for the bracelet program

348 EXPENDITURE FORMATS FOR DECISION AND CONTROL

to know the amount of costs associated with accomplishing your objective. This includes the salary and benefit costs of the new employees, but also additional supplies, equipment, and staff support.

• Once you have linked program performance with budget information, you have lots of discretionary choices concerning how to frame this linkage, as the following discussion of performance, productivity, and other types of performance information will illustrate.

Types of Performance and Cost Measures

As the previous discussion makes clear, if you want to do performance-based budgeting, you need to have increasingly sophisticated cost-related information. This information has grown in urgency as public bodies are attempting to offset shrinking resources and an inflation-related loss of purchasing power with improvements in productivity. The legal restrictions that have been put in place to limit new revenues, coupled with a faltering economy, have increased the incentive for local government jurisdictions to develop some clear measures of performance for the alloca- tion of tax dollars. There are many ways to approach and define performance, as the following sections will illustrate.

Productivity and Performance

While the concepts of productivity and performance measurement are not synonymous, they are very closely linked. Without a system for measuring the changes in actual performance, it is very difficult to determine if any strides are being made toward increasing productivity. The concept of productivity implies a ratio of the quantity and/or quality of results (output) to the resources (input) invested to achieve them. Productivity has two dimensions: effectiveness and efficiency. Effectiveness focuses on the extent to which governments achieve their program objectives, and it can be assessed through a combination of qualitative and quantitative measures. Qualitative measures collect in- formation on the satisfaction and quality of a given product or service, while quantitative measures collect information on the number and kind of services delivered to x number of clients.

Whether one is concerned with measuring the quality or the quantity of an objective, the ob- jective itself can take several different forms. For example, an objective may focus on impacts, outcomes, outputs, or results. The combination of ways in which objectives can be defined and quantitatively and/or qualitatively measured produces many discretionary possibilities in taking a performance-based approach to budgeting. For example, the assessment of a local school reading program could be defined in terms of quantitative scores on a nationally standardized examination. Or it could be measured in terms of the qualitative impact the program has on a student’s sense of well-being and improved performance in other areas of development. Or it could be measured in terms of successful performance in college or in one’s career.

In contrast to effectiveness measures, which focus on outputs, outcomes, results, or impacts, efficiency focuses on the ratio of a unit of goods and/or services produced in relationship to the amount of resources required to produce it. For example, a reading program may prove to be highly effective in achieving a set of predefined objectives, but the cost per student could turn out to be significantly higher than other alternative programs. If this proved to be the case, we would describe the program as being effective, but not as efficient as other alternative programs that are operated at a lower cost per student. In short, a concern for productivity does not predetermine the pathway to success. For example, higher productivity can be achieved by:

1. keeping expenditures constant and improving effectiveness and/or efficiency; 2. keeping effectiveness and/or efficiency constant and reducing expenditures; or

PERFORMANCE BUDGETING 349

3. increasing expenditures but improving effectiveness and/or efficiency at a higher pro- portional rate.

As this list of productivity improvement strategies suggests, productivity that focuses only on the relationship of inputs to outputs is a limited perspective at best. In addition to reaching and maintaining a given service-level target for the community, it is also important that the qualitative nature of that production does not diminish and that costs per unit of service do not skyrocket in the process. In summary, performance measurement involves monitoring the varying degrees of effectiveness and efficiency in performance that result from the production of public goods and services. When measurement of these dimensions of public service are systematically related to planned goals and objectives, performance measurement becomes an integral part of the man- agement planning cycle—that is, the development of goals and objectives (a plan for resource allocation), process implementation (the comparison of inputs to outputs in order to determine productivity), and performance measurement (efficiency and effectiveness determination). In fact, performance budgeting completes the planning cycle that is an integral part of PPBS.

Any indicator that seeks to measure the success of a government unit’s achievement of its goals or objectives can be considered a performance measure. The three types of performance measurements are (1) effectiveness measures, (2) efficiency measures, and (3) workload measures. Definitions and examples of each are provided in the sections that follow.

Measures of Effectiveness

Effectiveness is a result-oriented concept. It focuses on how well a goal or objective is accom- plished independent of cost or other inputs. Examples may include:

• Percentage of improved neighborhood traffic realignment projects in which through traffic is no greater than 1,000 vehicles per day one year after installation of the improvement.

• The average number of highway deaths from drunken driving in FY2 compared to the aver- age rates in FY1.

Neither of the above examples provides very precise information on the activities to be under- taken to produce the named results. And they do not provide information on the cost or efficiency of producing the result. They only give us information about outcomes, outputs, results, or impacts. When the effectiveness of a specific goal is explicitly connected to an intentional set of activities, the conditions are in place to create one of the following four kinds of effectiveness measures.

Targeted Goal or “Percent of Perfect” Performance Measures. This is the most desirable measure if it can be developed for an activity. The actual result is expressed as a target rate—a percentage of the most desirable result. Example: the percentage of neighborhood locations with traffic bar- riers in which through traffic is greater than 1,000 vehicles per day one year after barriers have been installed.

Result Characteristic Measures. When a base standard cannot be established, then one can use various characteristics of an activity to determine meaningful measures of the effectiveness of the activity. Examples include:

• Service complaints per year per 1,000 clients. • Response to complaints within x days. • Successfully resolved complaints measured by x criteria.

350 EXPENDITURE FORMATS FOR DECISION AND CONTROL

Compliance with Standards Measure. With many routinely provided services, it is possible to develop performance standards that measure the success of that activity against the developed standards. No additional credit is given for exceeding the standard. Note that the measure is similar to the “percent of perfect measure.” The important difference is that in using a “compliance with standards measure,” 100 percent is theoretically achievable and may be expected. Examples of compliance measures with standards include the following:

• 100 percent compliance with EPA and EQC Clean Air regulatory standards within one year for a minimum of 95 percent of the individual emission standards established by the regula- tory process.

• Number of citations issued by regulatory agencies for permit violations per million gallons of wastewater processed (used to measure 100 percent compliance with federal and state public health and regulatory agency requirements).

Subjective Measures. At times it is necessary to rely on the subjective views and attitudes of those persons involved in the activity. Although less reliable, these measures should not be ignored. In some areas, it is extremely difficult to develop other acceptable measures. When subjective measures are used, the following steps should be taken to improve consistency and reliability:

1. Identify the fact that a subjective effectiveness measure is being used. 2. Develop a checklist of points to be evaluated for each measure. 3. Try to have evaluators use simple scales to help organize their evaluation, for example:

outstanding, very good, good, average, below average, and poor. 4. Use more than one evaluator and take samples at a number of points, then average the

results. 5. Have the evaluation made by people other than those responsible for the activity. For

example, in surveying the satisfaction level of the business community with parking en- forcement, the evaluation should be undertaken by someone other than those responsible for enforcement activities.

Examples of subjective measures include the following:

• Percentage of business people surveyed who perceive the enforcement of parking regula- tions as helpful.

• Percentage of customers/clients reporting satisfactory service delivery by agency personnel.

Measures of Efficiency

Efficiency is a process-oriented concept. It focuses on how well resources are used without par- ticular regard for the amount of service output produced in relationship to the amount of resources required. Process efficiency measures can be defined in an engineering context by the amount of ingredients, time, and effort needed to produce an output. But purchasing or valuing inputs rede- fines many efficiency measures into costs that define the relationship between the funds needed for process inputs and a corresponding output or outcome. Carefully constructed cost measures can evaluate process performance, compare performance to published standards, and provide a basis of comparison across budget needs. Examples of efficiency measures include:

• Cost per participant in a tuition reimbursement program.

PERFORMANCE BUDGETING 351

• Cost per million gallons of wastewater processed. • Cost per installation of traffic control devices.

Efficiency measures relate output to the applicable input, generally dollars, and tend to fall into two groups:

Cost per Unit of Output Measures. This measure is the most direct. It relates the cost to the actual output or result and is expressed in dollars per unit of output. Examples include the following:

• Cost per dry ton of sludge processed. • Cost per person hour to maintain and operate one pumping station.

Cost per Other Base Measure. This measure is used where the cost cannot be directly related to output (as in prevention-oriented activities). Rather than measuring on a per individual basis, costs are scaled to appropriate orders of magnitude, such as the costs per one hundred, per thousand, per ten thousand, or per million basis. Using costs on a scaled basis results in a more easily understood measure rather than do costs with excessive leading zeros or of large size. Examples of costs per base measure include the following:

• Cost per 1,000 licensed drivers for traffic safety. • Cost per million lane miles for traffic fatalities. • Cost per million gallons of water treated.

Measures of Workload

Workload measures, as the name implies, track the amount of work that is done within a specified period. While common, they do not provide information on issues of efficiency, effectiveness, or how to reallocate resources to higher priority public goals. However, workload measures are essential for managers and unions who are interested in standardizing expectations with regard to what can be accomplished in a given amount of time. It is common, for example, to use workload measures to judge the amount of resources needed in managing permit centers, emergency rooms, and case managers in probation/parole/social service programs. Examples of workload measures include the following:

• Number of traffic lights installed. • Number of public requests for parking enforcement resolved. • Number of calls answered.

Workload measures are most commonly used by supervisors and serve various operational purposes, but they indicate little about the quality of the work performed. However, workload measures can be utilized to develop effectiveness measures by conversion to a rate or normalized basis, as the following example illustrates: Number of public requests resolved within the estab- lished response time compared to the number received. Workload measures can also be utilized to develop efficiency measures by conversion to a cost or input: Cost per traffic light installed.

Types of Costs as a Refinement of Performance Measurement

Costs are a specific type of efficiency performance measure that relates inputs of resources and funds to measured outputs and outcomes. Costs, however, reflect several different aspects of

352 EXPENDITURE FORMATS FOR DECISION AND CONTROL

resources used in the production process. Understanding the different types of cost is critical for the development of budget estimates and for gaining productivity improvements. Efficiency and workload measures are often related by using one or more of the following cost measures (Lyden and Lindenberg 1983, 70–74):

• Fixed Costs. A fixed cost input covers the resources needed to operate the program or service at any level of service or complexity. A program must have basic office space, utilities, in- formation technology and communications accessibility, supporting administrative services, and organization administrative relationships and oversight.

• Variable Costs. Costs that change proportionally as the size of the program or service in- creases or decreases are considered variable costs. Variable costs are often defined on a per individual basis. Typically, as more customers or individuals are served, the costs increase in a continuous curve. Variable costs increase with increasing numbers of students at a school, patients in a hospital, or inmates in a prison. For example, a state youth authority pays in- creasing food costs in direct positive proportion to the number of offenders and at-risk youth in a given population.

• Incremental or Step-Increase Costs. These costs are a variation of the variable cost examples described above. Instead of changing on a per individual basis, incremental costs change by a defined unit of capacity. For example, a state youth administration buys 40 passenger buses to transport offenders. Once the first bus is filled, the next increment is another 40 passenger bus with its added capital cost, driver, and maintenance.

The total of all fixed, variable, and incremental costs represents the total cost of providing a service to a specified level or number of individuals. The total cost at any specified level contains all fixed costs, the proportional variable costs, and the lowest level of incremental costs needed to meet service-level needs. Based on the per unit cost rates, analysts can estimate the costs for smaller or larger program levels. This forms the basis for allocating resources for next year’s budget. The level of service represents the amount of workload accomplished to a specified stan- dard within a given time span. The gross amount of product or service can be calculated using the following equation:

Service Level = Workload × Specified Standard × Duration.

Two examples of the application of this formula follow:

• Number of 911 calls responded to within 15 minutes during FYx. • Percent of customer complaints successfully resolved within seven days during FYx.

The service-level calculation in the above examples refines the level of service costs used in program (PPBS) budgeting (see discussion in chapter 12). The total cost to meet a level of service reflects the number of units in the workload and the cost per unit to meet the specified standard over a specified duration. It is important to note that as the level of service rises, fixed costs can be shared over a larger number of individuals, resulting in a smaller share per individual. For example, the construction and start-up costs of a new school auditorium and theater complex can be shared or amortized over the student population in an inverse relationship. At the same time, variable and incremental costs will increase in proportion to the number of individuals served.

Defining a cost in terms of the total fixed, variable, and incremental costs per number of in- dividuals served represents an average cost value for each individual. This may provide a useful model for many applications, including budget estimation. For example, hundreds of offenders in a youth corrections facility include both heavy teenage eaters and light eaters, but on balance, an

PERFORMANCE BUDGETING 353

average level of food consumption supports effective purchasing and budgeting. However, effec- tive performance and cost analysis often must move beyond average values to identify classes of individuals that incur relatively high or low costs. Cost averages mask important variations that lead to effective cost reduction. For example, Oregon Health Science researchers (Dorr 2007) estimate that 3 to 5 percent of Oregon Medicare patients account for about 50 percent of state Medicare expenditures. These patients have an intense level of service need and are typically elderly, with multiple—and oftentimes related—chronic diseases, along with other social and financial issues. Establishing a medical “home” for these patients and instituting coordinated team-based care may be one pathway for controlling these concentrated high costs. Focusing performance improvement and cost reduction efforts on the most expensive 5 percent would bring the greatest return for the improvement effort. Determining costs by categories of clients is important for water, sewer, fire, public health, and police services.

Benchmarks and Comparison to Established Standards

Performance measures are often custom designed to the context, program structure, and client base of a government or nonprofit program. These measures may provide effective information for budgeting and for cost analysis and monitoring. But without some sort of comparison to other providers or some type of published benchmark, administrators have no idea if their organization is meeting the highest possible level of efficiency and effectiveness. Standardized performance measures, or benchmarks, set by national professional organizations, state regulatory standards, federal agencies or federal grantors, major foundations, or published research can be used as a means of making comparisons across peer organizations. At least one of the performance measures developed for a program should match or reflect a standard measure or benchmark. Express- ing performance measures as a standard ratio and scaling the ratio to a standard form enhances comparisons.

In chapter 11 on line-item budgeting, we introduced the concept of a current service baseline as the level of funding needed to continue the quantity and quality of this year’s services into the next fiscal year. In the line-item case, we assume that organization and program productivity will remain the same year to year. But with a greater understanding of productivity and performance measurement, we can question that assumption and compare an organization’s performance to the performance of peer organizations. The current service level approach implies that a jurisdic- tion can benchmark its performance in various categories of public service against what it should be performing or what other comparable jurisdictions are providing. For a good example of this approach, see the annual Current Services Level Reports prepared by the City of Portland Audi- tor’s Office (City of Portland, Auditor’s Office 2010). Exhibit 13.2 provides an example of the performance of the Parks and Recreation Bureau in 2010 on a variety of workload, efficiency, and effectiveness measures, including its performance in comparison to other cities.

As Exhibit 13.2 illustrates, a variety of workload, efficiency, effectiveness, and comparative performance information can be collected and published that is of potential value to citizens, managers, and elected officials. Exhibit 13.2 also illustrates how a variety of different ways of measuring performance can be linked to budget information. But there is an ongoing debate as to how closely this information should be linked to the budget development and allocation process. The most extreme linkage would use the information in Exhibit 13.2 to determine how much more or how much less the Parks Bureau should get for next year’s budget. In the sections that follow, we use the history of a performance orientation to governance to explore in greater detail the enabling conditions that need to be in place to make performance-based budgeting possible and some of the technical considerations that make it possible to link performance metrics to resource allocation decisions.

354 EXPENDITURE FORMATS FOR DECISION AND CONTROL

Exhibit 13.2

Example Performance Report

Portland Parks & Recreation

MISSION Portland Parks & Recreation contributes to the City’s vitality by:

• Establishing and safeguarding the parks, natural resources, and urban forest, thereby ensuring that green spaces are accessible to all.

• Developing and maintaining excellent facilities and places for public recreation, and building community by providing opportunities for play, relaxation, gathering, and solitude.

• Providing and coordinating recreation services and programs created for diverse ages and abilities that contribute to the health and well-being of community members.

INPUT MEASURES 05-06 06-07 07-08 08-09 09-10

Expenditures (millions, adjusted):1

Operating $62.0 $63.2 $65.9 $1 $73.3 Capital $7.2 $13.5 $26.8 $1 $8.7 TOTAL $69.2 $76.7 $92.7 $1 $82.1

Permanent staffing (FTEs) 412 408 414 437 445 Seasonal staffing (FTEs) 284 298 320 335 381 Volunteers (FTEs) 219 221 223 224 222 Total volunteer hours 457,307 461,274 462,877 465,353 460,746 Total paid staff hours (millions) 1.4 1.4 1.5 1.6 1.7

WORKLOAD MEASURES 05-06 06-07 07-08 08-09 09-10

Service population 556,370 562,690 568,380 575,930 582,130 Number of Parks & Facilities:

Sports fields2 — — — 351 351 Community centers 12 12 12 12 12 Arts centers 6 6 6 6 6 Pools 13 13 13 13 13 Golf courses 5 5 5 5 5 Off-leash dog areas 31 31 32 32 32 Skate parks 1 2 3 5 5 Community gardens 31 31 32 32 35

Park acres: Developed parks 3,257 3,260 3,272 3,274 3,417 Natural areas 7,074 7,140 7,263 7,287 7,523 Undeveloped 282 285 228 234 207 TOTAL 10,613 10,685 10,763 10,795 11,147

Building square footage 1,081,712 1,081,712 1,091,944 1,117,922 1,120,035 Estimated recreation visits (millions):

PP&R–sponsored recreation programs and facilities2 — — — — 3.7 Sports programs using PP&R– managed fields 2.0 2.1 2.2 2.3 2.3

PERFORMANCE BUDGETING 355

EFFICIENCY MEASURES 05-06 06-07 07-08 08-09 09-10 Operating spending per capita

(adjusted)1 $111 $112 $116 —1 $126 Capital spending per capita

(adjusted)1 $13 $24 $47 —1 $15 Cost recovery for fee-supported

programs1 33% 34% 33% —1 34% Workers compensation claims/100

workers 7.3 8.3 8.4 8.3 7.6 Percent of maintenance done that is

preventive (hours spent, goal: 52%) 55% 53% 49% 46% 58%

Volunteers hours as percent of paid staff 33% 32% 30% 29% 27%

EFFECTIVENESS MEASURES 05-06 06-07 07-08 08-09 09-10 Residents living within ½ mile of

park (goal: 100%) 75% 75% 76% 76% 77% Park acres per thousand residents (goal: 19 acres per thousand

residents) 19.1 19.0 18.9 18.7 19.1 Facilities Condition Index

(0.05–0.10 = good) 0.05 — — 0.05 0.06

2006 2007 2008 2009 2010 Customer ratings (Community centers & pools): Percent rating overall quality good or very good — — — 96% 95%

COMPARISON TO OTHER CITIES 05-06 06-07 07-08 08-09 09-10 Parks operating budget per capita

(adjusted):3

City of Portland $94 $91 $98 $102 $109 6-city average4 $103 $111 $113 $103 $96

1 Data not reliable for FY 2008–2009, due to midyear accounting system change. 2 The Bureau improved the methodology for counting sports fields (2009) and recreation visits

(2010). Prior year data is not shown because the numbers are not comparable. 3 For comparison purposes, enterprise activities such as Portland International Raceway are

excluded from these numbers. 4 Charlotte, Cincinnati, Denver, Kansas City, Sacramento, and Seattle are the cities we used for

comparison.

For more information about Portland Parks & Recreation, click or go to: www.portlandoregon.gov/parks/

356 EXPENDITURE FORMATS FOR DECISION AND CONTROL

hISTORICAL PERSPECTIVE ON PERFORMANCE- BASED BUDGETING AND GOVERNANCE1

In the previous section, our discussion of performance-based budgeting focused on the con- ceptual categories, techniques, and methodologies required when taking a performance-based orientation to the budget allocation process. But, as we argued in our introduction to this chapter and as illustrated in the master case to Part III (Expenditure Formats for Decision and Control), performance is a matter of reconciling competing perspectives as to what truly counts for performance. We have argued that performance-centered budgeting is not simply a matter of mastering the technical expertise to understand the conceptual categories and methodologi- cal techniques of performance and measurement; rather, it is a matter of deciding what counts for the successful political performance of the governance system as a whole. And here, the perspective of citizens is important, as one would assume in a democratic system of govern- ment. But citizens have not been of equal mind on this issue, as a brief history of what counts for performance will illustrate.

Early Concerns of Public Administration for a Performance Orientation

The concern for performance is not something new. In fact, one can argue that the founding of the American political system was driven by a concern for performance. The English governance system in place in the 13 colonies was not performing to the satisfaction of colonists, so they took the initiative to alter the status quo. The American Revolution began to reveal the weaknesses of the Articles of Confederation, so there was another initiative to alter the system in ways that would bring it into better alignment with the performance needs and expectations of citizens.

In fact, from a historical perspective, the history of American government can be viewed as a nearly constant effort of performance realignment and change. The Jacksonian Era introduced rotation in office as a performance tool to check the disproportionate influence of commercial elites on the policy-making and implementation process. The Populist Era introduced civil service reform, the initiative, referendum, recall, nonpartisan local elections, citizen boards and commis- sions, and an expansion in the number of elected offices at the state and county levels of govern- ment out of a concern for better government performance. This tradition was continued with the Progressive movement at the turn of the nineteenth century, which introduced a wide variety of new performance initiatives (Morgan et al. 2013, chap. 5; Stever 1988). These included research bureaus that for the first time emphasized data-based policy recommendations and evaluation of results. Research bureaus laid the groundwork for performance measurement of managerial opera- tions, which could then be fed into the budget allocation process. The Progressives championed the council-manager form of government as a governing structure that would bring increased attention to issues of business efficiency and effectiveness, especially as an alternative to the commission form of government that was currently in vogue (Morgan, Nishishiba, and Vizzini 2010, 280–283). The council-manager model embodies the Progressives’ reliance on profession- als to meet the needs of citizens, whether through social work, public health, public works, law enforcement, or the management of the general affairs of the community (Stever 1988; Stivers 2000).

The New Deal continued America’s historical preoccupation with finding new ways of im- proving government performance. The vision that guided the New Deal was clearly articulated in a famous report by one of the intellectual leaders of the New Deal, Louis Brownlow. The Brownlow Committee Report (Roosevelt 1937) provides the underlying rationale for more than 100 executive-level reorganization plans submitted to Congress since 1939.2 It places priority on the dual principles of centralized political accountability and administrative efficiency.

PERFORMANCE BUDGETING 357

The Rise of New Public Management (NPM)

A concerted effort was made from 1980 through 2010 to reframe the model of public administration from a rule-centered system of accountability to one more focused on businesslike performance. This movement, called new public management (NPM), strives to make the services provided by government more responsive and accountable to citizens by applying businesslike manage- ment techniques with a strong focus on competition, customer satisfaction, and measurement of performance. There is general agreement among scholars and practitioners alike that these efforts to measure what gets done have produced a variety of worthwhile results:

• Measures of efficiency and effectiveness have inspired managers, supervisors, and frontline employees to improve their capacity to diagnose and correct operational problems (Ammons and Rivenbark 2008).

• An emphasis on performance measurement has heightened interest in creating systems that improve the overall management and governance of political entities (Hatry 2002, 2010; Moynihan and Pandey 2005; Wholey and Hatry 1992).

• Performance measurement systems have expanded to embrace an increasingly larger ar- ray of values, including cost and efficiency measures; various effectiveness measures like outputs, outcomes, and impacts; and qualitative measures like satisfaction, responsiveness, and quality of life (O’Flynn 2007; Stoker 2006; Jørgensen and Bozeman 2007).

But there are two major weaknesses in using private-sector business principles to improve the performance of government, regardless of what gets measured. First, there is no common denominator—like profit, market share, or return on investment—that can serve as a common comparator across the very wide range of criteria that defines government performance. These criteria and the way they are handled can either build or erode people’s sense of trust and legitimacy in their political institutions. Efficiency, effectiveness, and responsiveness to citizens (customers) do not exhaust the legitimating possibilities. Values like fairness, equity, protection of rights, and transparency play important roles in determining the legitimacy of political institutions, processes, and outcomes (Cooper 2003; Rosenbloom 2003; Moe 1994; Moe and Gilmour 1995; Lynn 1998; Kelly 1998; Moore 1994, 1995; Hefetz and Warner 2004). How do you put these more elusive values on the same plane as those that can be more easily measured? And how do you create a common denominator that allows for meaningful comparisons?

The second weakness inherent in using private-sector models for improving government performance is that the public sector consists of increasingly fragmented structures of authority that confound the possibility and even desirability of moving in a straight line from goals and objectives to instrumentally and rationally linked performance measures. The proliferation of local governments we described in Part I of this text, coupled with the entrepreneurial push to outsource public service to nonprofit- and private-sector organizations, shifts performance away from a single-minded focus on the efficient, effective, and customer-sensitive achievement of goals and objectives. Organization now must also focus on achieving and sustaining agreements between and across structures of authority. This has been described as “leading in a power-shared world” (see Bryson and Crosby 1992; Morgan et al. 2013, chap. 12) and is the focus of a body of literature called new public governance (NPG), which began to emerge in the early 1990s (Osborne 2010).

To summarize, the NPM agenda is based on a truncated view of the purpose of government. Government is not simply the agent of accountability to citizens or the agent of efficiency and effectiveness; it is also responsible for collecting the values of the community and creating inte- grated responses to these values across increasingly fragmented government systems (Kelly 1998;

358 EXPENDITURE FORMATS FOR DECISION AND CONTROL

Lynn 1998). When the values of the community are in conflict, or when they are contrary to the interests of significant minorities, government officials have a fiduciary responsibility to make decisions that are in the larger public interest. This kind of accountability model has faded into the background with the emergence of NPM. In a previous day, whether it was the scientific manage- ment movement of the early twentieth century or the efficiency and effectiveness initiatives of the Brownlow and Hoover Commission reports in the 1930s, 1940s, and 1950s, the discussion of performance was framed within a larger discussion. This larger discussion focused on the respective roles of the executive and legislative branches in making the overall system of government work (see especially Cook 1996, 119–120; Stever 1988, 1990). Performance was not stripped down to the simple issue of business efficiency and effectiveness. It was treated as an important part of a larger set of governance issues—issues that were an integral part of building trust and legitimacy in governing institutions and their leaders. This larger, more holistic and organic perspective has been largely missing from the proponents of the NPM movement over the past three decades. These concerns have spawned the rise of the new public governance movement (NPG).

Performance and the Rise of New Public Governance (NPG)

Concerns with NPM spawned a countermovement by both practitioners and academics to place substantive political values more at the center of the governance debate than has been the case with NPM’s narrow instrumental focus. This countermovement emphasizes three characteristics of public governance that are important for building trust and legitimacy but are undervalued by NPM. First, NPG is value-centered. It argues that the goal of government is to promote the larger common good. Mark Moore has called this new emphasis “the public value” approach (Moore 1994, 1995), while Stoker and others have called it “the collective preference” approach (Stoker 2006; Alford 2002). NPG is interested in advancing the value created by the whole of government activities, not just improved efficiency, effectiveness, or responsiveness in the implementation of a given program. This shift has broadened the objectives of performance measurement and man- agement to include service outputs, satisfaction, outcomes, a wide array of substantive political values, and, ultimately, citizen trust and the very legitimacy of government itself.

A second characteristic of NPG is that it emphasizes the importance of creating government processes that facilitate the generation of implementable agreements among wide-ranging stake- holders who may disagree on what courses of action will produce the maximum public value (Yang and Holzer 2006; Sanger 2008). This is because NPG views politics as the politically mediated expression of collectively determined preferences that the citizenry determines to be valuable (Alford 2002; Moore 1995; O’Flynn 2007). This contrasts with NPM, which views politics as the aggregation of individual preferences. The consequence of this difference is well-illustrated by the major steps the United Kingdom (UK) has taken over the past decade to reform the delivery of its social, education, medical, and justice services to citizens at the local government level. In recent years, public officials in the UK have chosen to treat government performance not as a set of rationally planned objectives but as a process of political mediation among contending stakeholders. These stakeholders have very legitimate differences regarding the public values that need to be preserved to ensure the integrity of the larger public good. Such efforts have resulted in the creation of a wide variety of new policy instruments, negotiated agreements, and performance measures that would have been difficult, if not unthinkable, under NPM (Brookes and Grint 2010; Osborne 2010, especially chaps. 16, 19, and 22).

A final characteristic of the NPG movement is that it views the creation of the public good as a coproduction process involving the public, private, and nonprofit sectors (see chapter 1 discussion on intersectoral interdependence; see also Crosby and Bryson 2005; O’Toole 1997, 2006; Osborne 2008). Under this model, the role of government is not simply to regulate, distribute, or redistribute

PERFORMANCE BUDGETING 359

public benefits but to serve as a catalytic agent to invest private and nonprofit stakeholders in shared ownership of the public good. This can take the simple form of community policing programs or a much more complicated form of networked governance such as watershed management over a very large geographic area involving multiple stakeholders and structures of authority. It is important to distinguish NPG’s approach to partnerships with the private and nonprofit sectors, compared to NPM’s approach. The latter is primarily interested in using the private and nonprofit sectors to deliver a service cheaply, efficiently, and effectively (Osborne and Gaebler 1992), while NPG is interested in enhancing the capacity of local organizations as a means of building civic infrastructure and the overall capacity of a community to be self-authoring (Smith and Lipsky 1993; Smith and Smyth 2010).

The three characteristics of NPG discussed above emanate from a common belief that govern- ment performance needs to be viewed from the perspective of the organic wholeness of a political system. In such a system, the public, private, and nonprofit sectors work together to contribute to the distinctive way of life of a political community. This view emphasizes the synergistic influence of history, institutions, and culture in creating a shared system of values and shared agreement on governance processes and structures—both formal and informal. In chapter 2, we introduced the concept of polity budgeting to capture this more holistic and integrated approach to public budget- ing. It draws from the resurgence of scholarship that uses polity or regime as the unit of analysis for understanding performance, political change, governance, and leadership development (Rohr 1989; Ozawa 2005; Johnson 2002; Stone 1989, 1993; Leo 1997, 1998; Lauria 1997). We believe this framework is especially useful in broadening the focus of performance away from program metrics to include building bridges across sectors and across organizations, thereby leveraging the maximum resources possible to promote the larger good of the community.

Current State of Performance-Based Budgeting: Performance as Ideology

Over the past four decades, a steady ideological shift has arisen that calls into question the para- digm represented by the New Deal. With this shift has come an increased interest in a performance orientation to the budgeting process and growing doubts about the effectiveness of the New Deal model of an activist government. As noted in our discussion of new public management and new public governance, there is general agreement between the two major political parties that the government has a central role in managing the economy, but this agreement does not extend to other roles, especially those that make government an active agent of managing social service, public health, environmental pollution, and consumer protection programs. Such initiatives can often compromise economic prosperity. For example, adding government regulations to control pollution, protect public health, and ensure consumer safety frequently increases the cost of doing business, even though long-term health costs for the nation as a whole may be considerably less.

Consider, for example, the information found in Exhibit 13.3. In his last year in office, President George H.W. Bush presented a federal budget for FY1992 that included the informa- tion in Exhibit 13.3. The information on first glance is shocking. Federal regulatory efforts are extraordinarily costly. Whether the information contained in Exhibit 13.3 is accurate or paints a fair picture of the overall federal regulatory environment is not the point for purposes of discussion here. Viewed in the larger context of an ideological debate over the size, role, and expense of government that was occurring during this period, the budget information in Exhibit 13.3 is part of a larger systematic campaign by presidents to alter the implementation of public policy in the United States.

The ideological shift to reduce the role, size, and cost of government has continued unabated since the 1980s and has played an important role in the budgeting processes of both democratic and republican administrations. For example, shortly after becoming president, Bill Clinton promulgated Executive Order 12866 (1993), the key provisions of which are summarized in Exhibit 13.4. This order sets out a new philosophy intended to guide the work of federal administrative agencies.

360 EXPENDITURE FORMATS FOR DECISION AND CONTROL

Exhibit 13.3

Cost Effectiveness of Selected Regulations

Regulation Agency

Cost per premature death averted (in

millions of $ in 1990)

Space Heater Ban CPSC 0.1 Aircraft Cabin Fire Protection Standard FAA 0.1 Auto Passive Restraint/Seat Belt Standards NHTSA 0.1 Steering Column Protection Standard NHTSA 0.1 Underground Construction Standards OSHA 0.1 Trihalomethane Drinking Water Standards EPA 0.2 Aircraft Seat Cushion Flammability Standard FAA 0.4 Alcohol and Drug Control Standards FRA 0.4 Auto Fuel-System Integrity Standard NHTSA 0.4 Standards for Servicing Auto Wheel Rims OSHA 0.4 Aircraft Floor Emergency Lighting Standard FAA 0.6 Concrete and Masonry Construction Standards OSHA 0.6 Passive Restraints for Trucks & Buses (Proposed) OSHA 0.7 Crane Suspended Personnel Platform Standard NHTSA 0.7 Children’s Sleepwear Flammability Ban CPSC 0.8 Auto Side Door Support Standards NHTSA 0.8 Side-Impact Standards for Autos (Dynamic) NHTSA 0.8 Low Altitude Wind Shear Equipment and Training Standards FAA 1.3 Electrical Equipment Standards (Metal Mines) MSHA 1.4 Traffic Alert and Collision Avoidance (TCAS) Systems FAA 1.5 Trenching and Excavation Standards OSHA 1.5 Hazard Communication Standard OSHA 1.6 Side-Impact Standards for Trucks, Buses, and MPVs (Proposed) NHTSA 2.2 Grain Dust Explosion Prevention Standards OSHA 2.8 Rear Lap/Shoulder Belts for Autos NHTSA 3.2 Benzene NESHAP (Original: Fugitive Emissions) EPA 3.4 Radionuclides in Uranium Mines EPA 3.4 Ethylene Dibromide Drinking Water Standard EPA 5.7 Benzene NESHAP (Revised: Coke Byproducts) EPA 6.1 Asbestos Occupational Exposure Limit OSHA 8.3 Benzene Occupational Exposure Limit OSHA 8.9 Electrical Equipment Standards (Coal Mines) MSHA 9.2 Arsenic Emission Standards for Glass Plants EPA 13.5 Ethylene Oxide Occupational Exposure Limit OSHA 20.5 Arsenic/Copper NESHAP EPA 23.0 Hazardous Waste Listing for Petroleum Refining Sludge EPA 27.6 Cover/Move Uranium Mill Tailings (Inactive Sites) EPA 31.7 Benzene NESHAP (Revised: Transfer Operations) EPA 32.9 Cover/Move Uranium Mill Tailings (Active Sites) EPA 45.0 Acrylonitrile Occupational Exposure Limit OSHA 51.5 Coke Ovens Occupational Exposure Limit OSHA 63.5 Lockout/Tagout OSHA 70.9 Asbestos Occupational Exposure Limit OSHA 74.0 Arsenic Occupational Exposure Limit OSHA 106.9 Asbestos Ban EPA 110.7 Diethylstilbestrol (DES) Cattle Feed Ban FDA 124.8 Benzene NESHAP (Revised: Waste Operations) EPA 168.2 1,2 Dichloropropane Drinking Water Standard EPA 653.0 Hazardous Waste Land Disposal Ban (1st 3rd) EPA 4,190.4 Municipal Solid Waste Landfill Standards (Proposed) EPA 19,107.0 Formaldehyde Occupational Exposure Limit OSHA 86,201.8 Atrazine/Alachlor Drinking Water Standard EPA 92,069.7 Hazardous Waste Listing for Wood Preserving EPA 5,700,000.0

Source: From the U.S. Budget for Fiscal Year 1992, Table C-2, Part II, p. 370.

PERFORMANCE BUDGETING 361

Exhibit 13.4

Excerpt from Executive Order 12866

Regulatory Planning and Review

September 30, 1993

The American people deserve a regulatory system that works for them, not against them: a regulatory system that protects and improves their health, safety, environment, and well-being and improves the performance of the economy without imposing unacceptable or unreasonable costs on society; regulatory policies that recognize that the private sector and private markets are the best engine for economic growth; regulatory approaches that respect the role of State, local, and tribal governments; and regulations that are effective, consistent, sensible, and understandable. We do not have such a regulatory system today.

With this Executive order, the Federal Government begins a program to reform and make more efficient the regulatory process. The objectives of this Executive order are to enhance planning and coordination with respect to both new and existing regulations; to reaffirm the primacy of Federal agencies in the regulatory decision-making process; to restore the integrity and legitimacy of regulatory review and oversight; and to make the process more accessible and open to the public. In pursuing these objectives, the regulatory process shall be conducted so as to meet applicable statutory requirements and with due regard to the discretion that has been entrusted to the Federal agencies.

Section 1. Statement of Regulatory Philosophy and Principles

(a) The Regulatory Philosophy

Federal agencies should promulgate only such regulations as are required by law, are necessary to interpret the law, or are made necessary by compelling public need, such as material failures of private markets to protect or improve the health and safety of the public, the environment, or the well-being of the American people. In deciding whether and how to regulate, agencies should assess all costs and benefits of available regulatory alternatives, including the alternative of not regulating. Costs and benefits shall be understood to include both quantifiable measures (to the fullest extent that these can be usefully estimated) and qualitative measures of costs and benefits that are difficult to quantify, but nevertheless essential to consider. Further, in choosing among alternative regulatory approaches, agencies should select those approaches that maximize net benefits (including potential economic, environmental, public health and safety, and other advantages; distributive impacts; and equity), unless a statute requires another regulatory approach.

(b) The Principles of Regulation

To ensure that the agencies’ regulatory programs are consistent with the philosophy set forth above, agencies should adhere to the following principles, to the extent permitted by law and where applicable:

(1) Each agency shall identify the problem that it intends to address (including, where applicable, the failures of private markets or public institutions that warrant new agency action) as well as assess the significance of that problem.

(2) Each agency shall examine whether existing regulations (or other law) have created, or contributed to, the problem that a new regulation is intended to correct and whether those regulations (or other law) should be modified to achieve the intended goal of regulation more effectively.

(3) Each agency shall identify and assess available alternatives to direct regulation, including providing economic incentives to encourage the desired behavior, such as user fees or marketable permits, or providing information upon which choices can be made by the public.

362 EXPENDITURE FORMATS FOR DECISION AND CONTROL

(4) In setting regulatory priorities, each agency shall consider, to the extent reasonable, the degree and nature of the risks posed by various substances or activities within its jurisdiction.

(5) When an agency determines that a regulation is the best available method of achieving the regulatory objective, it shall design its regulations in the most cost-effective manner to achieve the regulatory objective. In doing so, each agency shall consider incentives for innovation, consistency, predictability, the costs of enforcement and compliance (to the government, regulated entities, and the public), flexibility, distributive impacts, and equity.

(6) Each agency shall assess both the costs and the benefits of the intended regulation . . . [and] adopt a regulation only upon a reasoned determination that the benefits of the intended regulation justify its costs.

(7) Each agency shall base its decisions on the best reasonably obtainable scientific, technical, economic, and other information concerning the need for, and consequences of, the intended regulation.

(8) Each agency shall identify and assess alternative forms of regulation and shall, to the extent feasible, specify performance objectives, rather than specifying the behavior or manner of compliance that regulated entities must adopt.

(9) Wherever feasible, agencies shall seek views of appropriate State, local, and tribal officials before imposing regulatory requirements that might significantly or uniquely affect those governmental entities. [A]s appropriate, agencies shall seek to harmonize Federal regulatory actions with related State, local, and tribal regulatory and other governmental functions.

(10) Each agency shall avoid regulations that are inconsistent, incompatible, or duplicative with its other regulations or those of other Federal agencies.

(11) Each agency shall tailor its regulations to impose the least burden on society . . . taking into account, among other things, and to the extent practicable, the costs of cumulative regulations.

(12) Each agency shall draft its regulations to be simple and easy to understand, with the goal of minimizing the potential for uncertainty and litigation arising from such uncertainty.

The philosophy embodied in Executive Order 12866 puts all federal administrators on notice that they bear a much heavier burden in justifying all regulatory activity than had been the case in the past. It clearly expresses the countervailing values that should be given increased weight in the regulatory balance: protecting the private marketplace from unnecessary rules, reducing the costs of government regulations, placing higher priority on the role of states and Native American tribes, reducing paperwork, improving the readability of rules, and providing greater attention to the liberty and privacy interests of those being regulated. Another step in President Clinton’s campaign to change the implementation philosophy of administrative agencies was the designa- tion of Vice President Al Gore to implement Executive Order 12866 and carry out what came to be called the National Performance Review Initiative.3 This initiative received widespread me- dia coverage and formed an important cornerstone of Gore’s subsequent campaign for the U.S. presidency. Taken together, the executive order, the National Performance Review Initiative, and the systematic media campaign succeeded in significantly altering the implementation of public policy at the national level.

The performance focus of the Clinton and the elder Bush’s presidential initiatives has existed in all subsequent presidential administrations. President George W. Bush used the Office of Man- agement and Budget to create a new Program Assessment Rating Tool (PART) that facilitated the establishment of program-level performance measures across federal programs (Mullen 2006; Kamensky 2011; although no longer an actively maintained website, see ExpectMore.gov for an explanation and illustration of PART). Each PART questionnaire includes 25 questions divided into four sections.

PERFORMANCE BUDGETING 363

• The first section of questions asks whether a program’s purpose is clear and whether it is well designed to achieve its objectives.

• The second section involves strategic planning and weighs whether the agency establishes valid annual and long-term goals for its programs.

• The third section rates the management of an agency’s program, including financial oversight and program improvement efforts.

• The fourth section of questions focuses on results that programs can report with accuracy and consistency.

The answers to questions in each of the four sections result in a numerical score for each sec- tion from zero to 100 (100 being the best score). Because reporting a single weighted numerical rating could suggest false precision or draw attention away from the very areas most in need of improvement, numerical scores were translated into qualitative ratings. Exhibit 13.5 presents a summary of the bands and associated ratings used in the PART process.

Exhibit 13.5

PART Rating Range

Effective 85–100 Moderately effective 70–84 Adequate 50–69 Ineffective 0–49

President Obama and Congress have moved beyond the PART approach and replaced it with the Government Performance and Results Modernization Act of 2010 (GPRA Modernization Act 2010). GPRMA does not grade programs as being successful or unsuccessful; instead, it requires agency leaders to set priority goals, demonstrate quarterly progress in achieving goals, and explain performance trends. The GPRMA definition of goal reflects both cross-agency federal priority goals and initiatives, and agency-level priority strategic goals (Kamensky 2011). The GPRMA calls for each agency of the federal government to make a strategic plan available on its public website. The plan must contain a comprehensive mission statement covering the major operations of the agency; general goals and objectives, including outcome-oriented goals; a description of how the goals and objectives are to be achieved; and an identification of key factors external to the agency that could significantly affect the achievement of the goals and objectives. The act also calls for the establishment of a balanced set of performance indicators to be used in measuring or assessing progress toward each performance goal, including customer service, efficiency, output, and outcome indicators. The means used to verify and validate measures must be described, as well as how the agency will ensure the accuracy and reliability of the data. An additional effect of GPRMA is that it requires states receiving federal funds to measure their performance with those funds and to report on a regular basis to the responsible federal agency. GPRMA requirements for quarterly reporting and Internet-based communication of results are transforming the timescale of performance reporting into near real-time information. This short time frame contrasts with the static, annual performance reports used in most budget decision making. Reinforcing direct performance reporting, the American Recovery and Reinvestment Act of 2009 (Recovery Act) included an enhanced accountability system to track oversight and employment performance at the grant and contract level (Kamensky 2011).

The aforementioned examples from the last three decades illustrate the heightened importance of performance information to justify the continued existence of many public programs. But even

364 EXPENDITURE FORMATS FOR DECISION AND CONTROL

when data does not support ideological predispositions, it can be used under certain conditions to force a rethinking of public policy that was originally driven by an ideological agenda. As the following section illustrates, this rethinking is occurring with the mandatory prison sentencing policies that swept across most states in the 1980s; such policies were part of a larger, ideologi- cally driven bandwagon of reform supported by very little real data.

Performance Measurement as a Mediating Influence on Ideology: The Prison Problem

During the last quarter of the twentieth century, the states experienced a heightened interest in strategies that would reduce the growing crime rate. The most common response was to pass mandatory minimum sentences for different categories of crime, thus reducing the discretion of judges. One of the most famous mandatory sentencing initiatives was California’s “three strikes law,” which was the first mandatory sentencing law to gain widespread publicity. The law required imprisonment for a minimum term of 25 years after a defendant was convicted of a third felony. Similar laws were subsequently adopted in most states.

These laws were passed without much analysis of the long-term financial consequences, which have been tectonic. States are now spending $50 billion a year nationwide on incarceration. Cor- rections is the second-largest state expenditure behind Medicaid. One out of every 15 state dollars is spent on corrections. Not coincidentally, one in 31 American adults is incarcerated. It costs taxpayers, on average, $7.47 per day for parole and $78.95 for incarceration. The total annual cost per prisoner in California is $44,563, nearly the same price as a year at a major university with room and board (Skolnick 2011).

In South Carolina, the prison population tripled between 1983 and 2009, with a 500 percent increase in spending totaling $393 million (Skolnick 2011). California’s prison system has been in crisis mode during this period as well. With the need to cut the budget, its prisons filled to over 160 percent capacity. Thirty-three of the state prisons were operating at double capacity in 2011. Temporary beds, half of which were filled by probation or parole violators, were triple stacked in gyms and classrooms and crowding out rehabilitation programs. Sometimes two and three inmates shared a cell designed for one person. These conditions resulted in a lawsuit in which the U.S. Supreme Court mandated that the California Department of Corrections and Rehabilitation (CDCR) reduce the prison population to 137.5 percent capacity by 2012 (Brown v. Plata 2011). As a result of the decision, nearly 23,000 prisoners in California were released, all of whom were nonviolent and deemed to be low risk re-offenders (Vara and White 2011).

The problems described above have triggered the production of an avalanche of prison budget performance data: medical and mental health costs; labor and benefit costs; recidivism rates; post-prison employment rates; capital development and operating costs; the comparative costs and benefits of alternative strategies, including parole, probation, community service, employ- ment, education, and others; and contracting out services to private prisons.4 The collection of this information is having some dramatic policy consequences as each state scrambles to control its runaway corrections costs (Zaitz 2010; Skolnick 2011; Vara and White 2011). The most dramatic reversal is occurring in South Carolina, which, with help from the Pew Foundation for the States, passed the Omnibus Crime Reduction and Sentencing Reform Act of 2010. The act redirects investments in prisons to investments in probation, parole, and alternative service programs. It is projected to save $175 million in construction and $66 million in operational costs over the next five years (Pew Center on the States 2010b).

The summary point we wish to make in this section is that performance information that is connected to the budget allocation process can add up to a difference that counts in how taxpayer money gets spent—and on what kinds of programs. This can occur even when there are serious

PERFORMANCE BUDGETING 365

ideological differences of opinion among citizens and organized interest groups. But elected of- ficials have to be convinced that the performance information being collected is aligned with a political agenda that can acquire the support of those upon whom they depend for their political and financial electoral success. It is much easier for this to occur at the local level than at the state and federal levels of government.

REDEEMING PERFORMANCE FROM IDEOLOGy: whAT LOCAL GOVERNMENTS CAN DO TO MAkE PERFORMANCE BUDGETING SUCCESSFUL

Performance budgeting has enjoyed much greater success in local jurisdictions than at the state and federal levels (O’Toole and Stipak 1995; O’Toole and Marshall 1990; O’Toole, Marshall, and Grewe 1996).The reasons include the smaller scale and complexity of many local govern- ments, the absence of highly partisan political parties, and the existence of strong mayor and council-manager forms of government that allow for more centralized control and direction over administrative agencies by the executive branch. When administrative agencies operate with more than one master (as is the case at the federal level, where strong interest groups, divided government, and the separation of powers system vie for influence), it becomes difficult for agencies to sustain the will to implement performance budgeting over an extended period. This is especially the case where there is concern by strong unions that performance budgeting may threaten jobs and pay increases for employees. In the sections that follow, we will go beyond our discussion of the importance of performance information in budget decision making to focus specifically on what it takes to create and implement a performance-based budgeting system at the local budget level.

Practical questions for Doing Performance Budgeting

Performance budgeting is partly art and partly science. It is therefore useful to learn from the experience of those who have developed performance metrics, especially with the goal of tying them to the budget allocation process. The following “baker’s dozen” list of questions is based on our compilation of the experience of practitioners who have taken our public budgeting course over the past 20 years and who possess considerable experience with various forms of performance budgeting.

1. Is there a strong causal connection between the performance measure and the objec- tive? For example, if reduction of crime is the objective, it would make little sense to use increased police response time as a performance measure if that did not have much influence on the reduction of crime. On the other hand, if the objective is to reduce the fear of crime, then a quicker response time may have a more causal influence on the achievement of the objective. In other words, when selecting a performance measure, you need to ask whether there is any real connection between the performance measure you have chosen and the objective you are seeking to accomplish.

2. Does the performance measure overstate or understate how much of the objective can be achieved by the chosen measure? For example, how much can AIDS be reduced by a public education campaign? By a needle disposal program at selected sites throughout a jurisdiction? By an aggressive safe sex program? These questions illustrate one of the reasons for the resistance by some career administrators and program managers to per- formance measures. They resist being held accountable for the achievement of results through limited specified performance measures.

366 EXPENDITURE FORMATS FOR DECISION AND CONTROL

3. Are your performance measures minimally necessary? Frequently, more than one mea- sure needs to be developed for an objective, especially if you are seeking to incorporate quantitative, qualitative, and cost information. But ask yourself, How many measures are minimally necessary to determine how well my organization is achieving its stated objective? Data collection, management, and analysis are a cumulative cost to the orga- nization. So collecting information that is not needed—or where less expensive data is sufficient—is an important consideration.

4. Is the measure valid? Does the indicator measure the effectiveness or efficiency of the objective?

5. Can the measure be quantified? The better quantified the measure, the easier it will be to use.

6. Is the data required to develop the measure available? The more available the measure, the easier it will be to implement and operate.

7. Is the measure as simple as you can make it? The simpler the measure, the easier it will be for all concerned to understand and use.

8. Is the measure acceptable to those who are likely to care about its use? Ask whether people will accept it as a valid measure.

9. Does the measure give you timely information? Will the measure reveal problems soon enough to permit corrective action before significant loss of service occurs?

10. Can the information be collected and sustained over a broad enough sample and pe- riod to provide you with a reliable measure? How dependable is the data collection system?

11. Can the information be collected at a reasonable cost? 12. Are you spending so much time on the development of performance measures that service

delivery is suffering? It is possible to become so involved in improving efficiency that service objectives are nearly forgotten and effectiveness impaired.

13. Previous measures may no longer be adequate, and they should be changed and refined to a point of general managerial acceptance. Experience may show managers that the information being tracked is irrelevant. If any of these problems occur in the year or two after objectives have been established and performance measures have been developed, time needs to be built into the process to correct the situation prior to the following fiscal year’s budget submission. This will enable the necessary changes to be reflected in the budgetary plan for the upcoming fiscal year. The important thing to remember is that there should be flexibility in any program planning and performance measurement system.

Some Examples of how to Link Performance Measures to Objectives

One of the more difficult challenges in doing performance budgeting is choosing measures that are causally linked to the objectives you are trying to achieve. For example, a public safety de- partment may have an objective of “reducing homicide crimes by 10 percent over the previous year.” But if the department chooses to measure the response time to calls, this may have little or no impact on the actual reduction of homicide crimes. Similarly, the timely response to citizen and victim complaints may have little influence on the factors that contribute to the complaints in the first place. It is important, then, to select performance measures that are causally linked to the actual achievement of the stated objective. Exhibit 13.6 provides some examples that illustrate this causal connection.

PERFORMANCE BUDGETING 367

Exhibit 13.6

Example Causal Connections

1. Housing and Community Conservation

Objective: To reduce abandoned vehicle violations in FY2 by 10 percent or greater over FY1 in each neighborhood with a high number of violations, at a cost no greater than 5 percent above FY1.

Performance Measure: Ratio and cost of abandoned vehicle violations in high violation neighborhoods compared to the previous year in the same neighborhood.

2. Performing Arts

Objective: To increase potential usage of all related facilities by referring inquiries when Performing Arts Auditorium is unable to accommodate.

Performance Measure: Percentage of inquiries referred in FY1 compared to FY2.

3. Public Works

Objective: To operate treatment facilities at a cost no greater than the previous three years’ average cost, with 100 percent compliance with federal and state public health requirements.

Performance Measure: Cost of operation in FY4 compared to average cost of FY1–FY3.

Performance Measure: Number of citations issued by public health agencies for permit violation per million gallons of wastewater processed.

4. Financial Management Services

Objective: To provide revenue estimates that are within a 5 percent range of accuracy at a cost no greater than the previous year.

Performance Measure: Accuracy of revenue projections within a variance of 5 percent.

Performance Measure: Total cost of operation in FY2 compared to total cost in FY1.

5. Police

Objective: To ensure that the average response time for all calls for service does not exceed x minutes.

Performance Measure: Response time = Average response time per call

Number of calls

6. Fire

Objective: To contain the dollar and life loss per fire at a level not to exceed x percent of the average for the past three years.

Performance Measure: The dollar loss per fire and the life loss per fire for next year as compared with the average dollar loss per fire and life loss per fire for the previous three years.

7. Executive

Objective: Ensure that progressive discipline training is provided to 50 percent of new supervisory personnel appointed in FY5 at a unit cost of $100 per person.

Performance Measure: The percentage of new supervisors appointed in FY5 who received progressive discipline training.

Performance Measure: The total cost of progressive discipline training for FY5 divided by the total number of new supervisors trained in FY5.

368 EXPENDITURE FORMATS FOR DECISION AND CONTROL

Performance Budgeting Exercise

One of the best ways to appreciate the intricacies and complexities of performance budgeting is to actually experience the task of doing it. Using the form provided in Exhibit 13.7, prepare some performance measures for goals and objectives you have developed. You can use an example from your own experience, build on the examples from chapters 11 and 12, or draw from the data provided on the accompanying website for this book. Remember to tightly link the performance measures to the objectives and to include a performance measure on costs. Costs may not be as relevant or as easy to quantify when dealing with mandated or constitutionally required activities, such as first and fourth amendment rights, due process, and regulatory requirements. But even in these cases, there are discretionary administrative choices that need to be made within the required parameters. From doing the exercise, one can quickly see why it takes three to four years to put such a system into place. This cannot be accomplished without incentives for elected officials to use the performance information to make budget allocations easier rather than more difficult.

Creating the Enabling Conditions for a Performance Budgeting Orientation

More often than not, practitioners are motivated to collect and use performance information when it helps solve practical public policy and budget problems rather than make the problems more difficult. This may sound obvious, but it is an essential starting point for understanding what it takes to create incentives for public officials (both elected and career) to place increased emphasis on the use of measurement in the budget allocation process. It is quite rational for public officials to resist spending their time and the public’s money to undertake extensive performance analysis when there is strong political support to continue with business as usual. And it is even more ra- tional for them to resist analytic information when it generates considerable political resistance, as Budget and Finance director Spiro Augustine came to realize in the teaching case for Part III. As Kettl (2000) points out, “Most discussions of public management reform begin with a flawed premise: that management reform is most fundamentally about management. . . . Elected officials do not pursue management reform for its own sake but because they believe it helps them achieve a broader political purpose” (67). This means that important incentives are needed in the political environment for performance information to be collected and used in the budget allocation process (Hatry 2010; Lynn 1998). There are several external factors currently at play that create positive incentives in favor of a performance approach to budgeting.

External Drivers Create Incentives to Undertake Performance Budgeting

We write this book on local public budgeting in the midst of widespread ideological dissensus over what American citizens want their government to do. But there is growing evidence that all parties in the debate are relying more and more on the collection and use of performance information to make their case. This dependence on performance information has been significantly aided and abetted by a growing bipartisan recognition that many state and local governments simply can’t afford to continue doing what they have done in the past; without important changes in the way all levels of government run, Americans may feel they are being pushed to the edge of bankruptcy and tethering future generations to very high rates of taxation, fees, and charges for government programs and services. Together, these conditions create incentives for elected officials to make use of performance data and to tie the measurements to the budget allocation process. While the debate over what to cut and what not to cut will remain fierce at the national level, where party partisanship will continue to use major government programs (i.e., social security, health care, food stamps, business tax credit, etc.) as symbolic “weapons” of mass destruction against their

PERFORMANCE BUDGETING 369

opponents, at the local levels of government these debates are significantly tempered by a focus on how to make basic services more efficient, affordable, and accessible. For these reasons, we are optimistic that the trend for local jurisdictions to rely on the collection and analysis of per- formance information in making budget allocations will continue to grow in the future. In fact, according to an April 2013 report by the Center for American Progress (Shah and Costa 2013),

Exhibit 13.7

Practical Exercise: Performance Budgeting

Goal No. 1: _______________________________________________________________________

__________________________________________________________________________________

__________________________________________________________________________________

Objective No. 1: _____________________________________________________________________

__________________________________________________________________________________

__________________________________________________________________________________

Performance Measure No. 1: _______________________________________________________

_______________________________________________________________________________

_______________________________________________________________________________

Performance Measure No. 2: _______________________________________________________

_______________________________________________________________________________

_______________________________________________________________________________

Objective No. 2: _____________________________________________________________________

__________________________________________________________________________________

__________________________________________________________________________________

Performance Measure No. 1: _______________________________________________________

_______________________________________________________________________________

_______________________________________________________________________________

Performance Measure No. 2: _______________________________________________________

_______________________________________________________________________________

_______________________________________________________________________________

370 EXPENDITURE FORMATS FOR DECISION AND CONTROL

New York City is now piloting the use of social impact bonds, a relatively new outcome-based strategy in which the government pays for a contracted service only after there has been an audit by a third party verifying the achievement of the agreed-upon contract outcomes or results. Under this performance-based model, a government agency or agencies define a specific, measurable social outcome they want achieved in a given population over a certain period of time, such as reducing the number of asthma-related emergencies in low-income children by 15 percent over five years in a small city. They then contract with an external organization that pledges to achieve that outcome within the specified time. The external organization—sometimes called an intermediary— raises money from private investors to fund the interventions and actions needed to achieve the outcome. If an independent assessor verifies through a rigorous evaluation that the outcome has been achieved, the government releases an agreed-upon sum of money for the outcome, and the external organization repays its investors with a slight return for shouldering the financial risk. If the external organization fails to achieve the outcome, the government doesn’t pay, and the investors stand to lose their capital (Shah and Costa 2013). We anticipate that such performance-oriented innovations will become increasingly more common at the local government level.

The Importance of Getting Agreement on Goals: The Example of Using Government to Manage the Economy

As we argued earlier, unless elected officials can reach agreement on goals and priorities, it will be very difficult to make performance budgeting a way of life. Our example in the previous section of getting agreement to reduce asthma-related emergencies in certain target populations is relatively easy to achieve compared to getting policy agreement on how to manage and provide services to the local homeless population. But once this agreement is achieved, a performance orientation can quickly become the norm for doing government business. A dramatic example that illustrates this point is the political agreement among both Democrats and Republicans following World War II on using government to manage the economy. For the first half of the twentieth century, there was profound partisan disagreement over the role of the government in managing the economy, as evidenced by the New Deal debate over whether the government programs created during this period represented the advancing guard of socialism. But this partisan divide ended in 1946, with the passage of the Full Employment Act. In this bipartisan postwar act, both Democrats and Republicans agreed to lay the responsibility for economic stability (as measured by inflation and unemployment) onto the federal government (Stein 1969). The act did not specify how this was to be done, thus leaving open the question of whether it should be done through Keynesian policies of increased government spending during troubled times or fiscal policies that focused on control of the money supply through interest rates and taxation.

The Full Employment Act created the Council of Economic Advisers (attached to the White House), which provides analysis and recommendations, as well as the Joint Economic Com- mittee of Congress, consisting of members of both houses and both major political parties. This bicameral and bipartisan structure has played a powerful role in norming expectations about the role of government in the economy and the role of analysis in measuring government economic performance. This is in evidence on a daily basis in our media coverage. For example, when a report is issued on the current unemployment rate, the inflation rate, or the rate of growth in a given sector of the economy, there is seldom ideological debate over the accuracy of the numbers. Instead, there is debate over who is at fault for the inadequate progress or which strategies will work best to increase economic performance.

The larger point to be made from this example is that the chances for linking performance metrics to budget allocation increase significantly where there is agreement on the goal. Without agreement between Democrats and Republicans that government has the central role in manag-

PERFORMANCE BUDGETING 371

ing the economy, there would be far more debate over the meaning and value of the data that is collected, its use in measuring and managing the state of the economy, and paying for the salaries and programs to make this management effective. The lesson for local government officials is that political agreement on goals is critical to the success of performance budgeting. If goals are not understood, or if there is disagreement over their value or priority, no amount of data is likely to make much difference. Our discussion of planning and goal setting in chapter 10 and priority- based budgeting in chapter 14 have important implications for the success of performance-based budgeting. These chapters set forth various strategies and techniques for getting political and community agreement on budget priorities, which is the necessary precondition for the long-term political survival of performance-based approaches to budgeting.

ADVANTAGES AND DISADVANTAGES OF PERFORMANCE BUDGETING

The completion of the budget exercise in the previous section provides the basis for summarizing the key advantages and disadvantages of doing performance budgeting.

why a Focus on Performance Is Important

There are many reasons for focusing on performance in the public sector. The conventional view is that public-sector organizations are much more concerned about creating and running programs that meet the pleasure of elected officials and vested interest groups than about being efficient and effective. Therefore, one of the primary purposes of performance budgeting is to temper the influence of political pressures on the resource allocation process with a strong dose of rational assessment of what is being purchased with tax dollars. This concern is real, especially at the federal and state levels of government.

Frequently, elected officials and their constituents press for passage of programs that do not necessarily represent an effective or efficient expenditure of tax dollars. As we discussed earlier, mandatory sentencing is an example of where this has occurred, resulting in extensive expansions of correctional facilities that are now too expensive to maintain. Had there been clear and long-term analysis of the budget consequences of these decisions before they were made, maybe the outcomes would have been different . . . but maybe not, since such laws frequently are passed when the heat of public passions are high and where there is little time or patience for preparing and debating various kinds of performance data. As we pointed out in chapters 1 and 7, the taxpayer revolts starting in the 1970s did not provide the opportunity for passions to be tempered by the hard facts of the longer-term reality that would result from property tax limitations. However, one of the secondary consequences of both of these populist initiatives is that they have heightened interest in performance measures as a way of demonstrating to citizens what their taxes are purchasing in the way of both outcomes and efficiencies. One way of viewing the politics of performance bud- geting is to see it as a struggle between a constituent service/clientele and a taxpayer perspective, with the latter more interested in performance measures than the former.

In addition to these political reasons for undertaking performance measurement, there are important management purposes served by performance budgeting. For example, performance information provides an opportunity for managers to determine whether an organization is achiev- ing its general and specific goals. Because of resource limitations, a manager must be able, at least periodically, to assess the current situation against some agreed-upon criteria. As achieve- ment is measured, this information provides the basis upon which to make program adjustments. Measuring the achievement of an objective can also provide an improved basis for developing succeeding managerial strategies, alternative budget proposals, and service level objectives. In short, a performance system has several managerial benefits:

372 EXPENDITURE FORMATS FOR DECISION AND CONTROL

• Performance measurements can be used to evaluate specific program or project proposals by comparing planned to actual achievements. Programs can be evaluated in terms of improved performance while at the same time considering whether changes are cost effective.

• Performance measurements can be used to forecast resource requirements. • Performance information provides an important rationale to justify budgetary resource requests.

Performance-based rationales supplement, and sometimes counter, the arguments made by political interest groups in favor of and in opposition to public programs and services. This is especially important during a period of declining resources with pressures to reduce personnel costs. With- out productivity information, it becomes difficult for both labor and management to defend the number of staff on the payroll without some kind of productivity measurement system.

The Barriers and Disadvantages of Performance Budgeting

Despite the seemingly obvious advantages of using performance measures, little widespread inter- est has surfaced in systematically using such information in the budget allocation process. There are several reasons for this. First, performance measurement takes considerable time and energy to execute over time, especially if reliability is an important concern. The rule of thumb is that it takes three years to develop, implement, and test a performance-based budgeting system. This is about the length of an election cycle, which means that sustained political support over time is needed to create and maintain a performance budgeting system.

There are many factors that contribute to the need to take time in the development of performance budgeting. Not only is the technical aspect of the task difficult and time-consuming, as we have already illustrated in previous sections of this chapter, but the development of performance mea- sures may be threatening to program managers and their staff, especially if there is the perception that the performance measures will be used to justify program and staff cuts. But managers also might resist performance budgeting for the same reason they resist program budgeting: It limits their flexibility to shift dollars from one set of activities to another. As we noted in our discussion of program budgeting in the previous chapter, over time strong constituency support can develop around defined sets of activities and programs, thus constraining managerial flexibility.

Another barrier is that many units of government may find it difficult to measure outcomes, especially in the short run. For example, there are a variety of social service programs for children, addicts, battered women, and the mentally ill that encounter difficulty crafting adequate, valid, and reliable longitudinal measurements. Even when such measurements can be constructed, the time and expertise required to do so can end up reducing the resources available for direct service. Finally, some public programs are not nearly as popular as others. When this is the case, program supporters and advocates frequently argue that high measures of performance will not fundamen- tally change the support they are likely to receive from the general citizenry. For instance, there is abundant evidence that building more prisons and expanding mandatory punishment for various categories of criminal activity will not reduce the crime rate or do much good for the incarcerated. Despite this performance information, the general public has generally supported incarceration over treatment or rehabilitation.

Finally, and perhaps most important, performance budgeting has been resisted in some cases by elected officials and their constituents. Elected officials and their staff frequently judge a program negatively if they receive a large number of complaints about it from their constituents. Conversely, a program is judged positively if it has considerable constituency support. The de- velopment and use of performance measures may significantly upset this political calculus. For example, programs that may have strong support from a powerful few may not stack up well on the basis of performance measures. Likewise, politically weak constituents may have a strong case to be made in their favor on the basis of performance measures.

PERFORMANCE BUDGETING 373

Performance Budgeting in a Community Network Context

A results-oriented, outcome-driven focus on performance quickly moves beyond the boundaries of a single local governmental organization and into its service-delivery network of intergovern- mental, nonprofit, and for-profit partners.5 New public management and new public governance encourage these extensive external partnerships to focus attention on creating greater governmental efficiency and effectiveness. However, partnered service delivery brings an array of challenges in measuring, reporting, and assessing partner and network performance. Network governance decisions to revise service delivery programs and arrangements, to change previous allocations of grant or contract funding, or to eliminate inefficiencies in performance often bring political and even legal challenges.

As in the organizational context, effective performance budgeting by a community network requires planning and agreement on joint missions, goals, and work objectives. The PPBS/ program budgeting format provides an umbrella structure under which many community actors can contribute policy recommendations, funding, service, and administrative capacity. Effective planning identifies the mission and goals for the community network as an entity. Additionally, plans must allocate resource and performance expectations to each member organization and to its subnetwork constellation of contractors, grantees, subgrantees, and collaborative partners. For performance budgeting to work, all members of the network must consistently collect detailed cost, accomplishment, and productivity information, and communicate that information to budget analysts with the network lead organizations.

The criteria for measuring efficient and effective network performance and accountability vary with an observer’s perspective and role responsibility. Network member organizations and the network leadership must respond to and communicate performance accomplishments and quality with at least the following three perspectives in mind (Provan and Milward 2001; Bardach and Lesser 1996):

• an external view of the network as a community responder and system; • an internal view of the network as a single integrated entity of members, resources, admin-

istrative systems, and service delivery systems; and • an internal view of the network that takes account of the individual organization/participant

as a member of the network.

The criteria for measuring successful accountability and performance will vary according to the needs of the three perspectives just listed. Lead administrative organizations in a network, including lead local governments and community intermediaries and foundations, will need ac- complishment, cost, and productivity information to support their annual budgeting processes and budget decision making. The budget decisions made by these lead organizations must respond primarily to the external community and to the function of the network as an integrated system. The network and its leadership may need to make compromises and trade-offs to demonstrate successful performance among the three levels. Slippage on the part of one member’s perfor- mance measures may be necessary to ensure the performance of the larger network (Bardach and Lesser 1996, 204). Performance measurement under the NPM framework would place the great- est emphasis on the community-level outsider’s viewpoint. NPM would ask how efficiently and effectively the partner or network arrangement responded to the community’s issues and needs. While the NPG framework recognizes the importance of the external community viewpoint, it also takes into account the health of the network as a civic entity and the well-being of the member organizations as long-term contributors to the civil society. Bardach and Lesser (1996, 201) argue that networks are accountable not only for results but also for the governance issues of choosing priorities wisely, targeting their resources and professional discretion, and overseeing system modification and redesign.

374 EXPENDITURE FORMATS FOR DECISION AND CONTROL

The capacity to measure the performance of individual network members and of an entire network rests on the authorities and requirements embedded in the relationships between and among network members. Network relationships fall into four broad categories: intergovernmental agreements (IGAs), short- to medium-term contractual relationships, long-term public-private partnerships, and collab- orative partnerships. Cost data in the form of budget line-item funding and contract/grant costs are typically available from governmental or nonprofit organization procurement and finance systems. IGAs for services between local governments do not necessarily require performance measurement or reporting features, although local government councils, commissions, or governance boards could add agreement provisions to develop program goals and objectives and to collect and report accomplishment, cost, and productivity information. With this information, government analysts, executives, and elected officials can then apply various performance budgeting techniques.

Accomplishment and performance measures may be readily available in many short- and medium-term contractual relationships. Contract and grant performance requirements and as- sessments may provide work accomplishment and quality measures that closely reflect efficiency and effectiveness measures (see chapter 17; Farris 2009). Contract reporting requirements ensure a steady flow of information to the contracting organization. However, to determine whether external contracting is more or less efficient and effective than in-house program delivery, the contracting government or organization needs to collect internal costs, industry benchmark costs and performance data, as well as contractor costs and performance results (see U.S. Government Accountability Office [GAO] 2007).

Long-term public-private partnerships are often results oriented and outcome focused, and partner performance may be measured by required criteria. No matter the exact structure of these formal partnerships, the contracting or granting government or the lead network nonprofit must collect, organize, and report on the partnership performance and costs for its subsequent use in performance budgeting. Collaborative partnerships present the most challenging relationships on which to measure performance. The very mention of performance measurement and verification can transmit a lack of trust and confidence in the partner or relationship. Many smaller collaborat- ing nonprofits may rely extensively on volunteer labor, which does not lend itself to performance enforcement. Negotiated agreements of accountability between partners can help to provide a limited means of performance responsibility between the partners. Reliance on the partner’s or- ganizational procedures and reporting may provide an information source on which to construct network performance (Morgan et al. 2013, Figure 14.8, 448).

The performance-based budgeting decisions by the lead government agencies, community foundations, or federated intermediaries may have implications throughout the community network. Public perception of poor performance by one network member tends to implicate all the other network members (Bardach and Lesser 1996, 208). Making governance decisions to revise the structure of the joint network or to replace a network partner to increase efficiency or effectiveness requires both procedural transparency and extensive communication between network members and the public. For a participant-governed network (Provan and Kenis 2007) of relatively equal organizations, a decision to modify the network likely requires dialogue and broad agreement. For networks organized around a lead organization or external network administrator, the policies and procedures of the lead organization would provide network governance procedures. Federal or state contract or grant requirements may provide additional procedures for managing subcontracts or subagreements with service providers.

The results of performance measurement and analysis may imply the need for revised fund- ing levels and revised policies. Network governance decisions that change network structures to increase performance or to eliminate a poorly performing member require careful attention to timing and the potential reaction of participants across the network. Contract or grant completion and consideration for renewal provide a natural opportunity for participants in network and inter-

PERFORMANCE BUDGETING 375

governmental agreements to rethink and restructure partnerships, cost structures, and performance requirements. This may be the time to undertake an objective analysis and use the information to replace a poorly performing partner or to reconfigure the network to increase efficiency and effectiveness. But frequently, the contracting and grant processes operate within a relatively short time frame—one that is not well suited to a performance-based analysis and discussion. Unless the performance criteria have been established at the outset of the contract, grant, or intergovernmental agreement, it will be difficult to use the end of the process to acquire the kind of information that is needed for a thorough performance-based dialogue and a consideration of alternatives strategies. And pushing changes without the data to support it may open the contracting or granting organi- zation to legal liability. It is also frequently the case that replacing major governmental or major nonprofit network funders or participants is extremely difficult, even in the presence of rival large organizations (Bardach and Lesser 1996, 205). Large community foundations and intermediary organizations have a major presence in the community polity.

In the face of the difficulties of linking performance to contract and grant renewal, community leaders and elected officials may provide an alternative remedy for poor performance by major organizations. Contracts and memoranda of agreement sometimes serve as a means to clarify responsibilities and roles and to ensure performance and accountability among major network members. Negotiations and political pressure may be the only means of encouraging more ef- ficient and effective performance by collaborative members of a network with interpersonal and informal ties to the larger group.

CONCLUSION

We conclude this chapter with the observation that performance budgeting will become increasingly important in the future, despite all of the barriers that stand in the way of its successful application in the public sector. There are two reasons for this conclusion. The first has to do with the state of the economy for the foreseeable future. When the economy is growing, the justification for adding new public programs is commonly a matter of generating the political support needed to authorize it and get it funded; however, when the economy is declining, public budgets and programs have to be cut. Political support, however important, is not enough to save the day. This is especially the case when program cuts operate against the popular will or prevailing ideology of the day, as has been the case with prison reform. In these cases, performance information becomes an increasingly urgent and even normal part of the budgeting process.

There is a second reason that performance budgeting will become more important in the future. Despite the barriers to implementing performance budgeting on a systemwide basis and getting agreement by elected officials on community-wide goals and priorities, individual program man- agers will still be motivated to use performance budgeting to achieve a variety of administrative objectives within their scope of authority. Their ability to do so will vary, of course, with the type of work they are managing. For example, it would probably be much easier for the manager of a road crew to develop performance measures than it would be for the manager of a group of neighborhood associations. But the general pressure on managers “to do more with less” and to justify their existing base budgets creates the kind of incentives that make performance budgeting here to stay—if not for elected officials, at least for professional career managers.

STUDy qUESTIONS

Completion of the performance budgeting exercise in Exhibit 13.1 and making use of appropriate data on the textbook website (www.pdx.edu/cps/budget-book) may be helpful in answering one or several of the following study questions.

376 EXPENDITURE FORMATS FOR DECISION AND CONTROL

1. What is meant by the term performance budgeting”? When, why, and by whom was performance budgeting first used extensively in the public sphere?

2. What are the differences (the strengths and weakness) between performance budgeting and PPBS?

3. What are the various ways of measuring the performance of an organization so that it can be tied to budget allocation?

4. How does the role of the chief executive (CEO) and the legislature differ in performance budgeting and PPBS?

5. What conditions are necessary for the successful implementation of performance budgeting?

6. Who tends to gain and who tends to lose in adopting performance budgeting? 7. What efforts have been made by your agency in recent years to implement a performance

budgeting system? What have been the results of that experience, or what might the results be if your agency were to attempt to implement performance budgeting?

8. What factors account for the success of productivity-oriented budgeting approaches at the local government level?

9. What explains the differences in performance-oriented budgeting efforts at the federal government level from those attempted at the state and local levels?

10. To what extent do you believe performance measurement can be successfully incorporated into your own organization’s budget development and implementation process?

11. To what extent can/should performance improvement be included as a major purpose of the public budgeting process?

NOTES

1. For this section, we have drawn from previous work published in the Foundations of Public Service (Morgan et al. 2013, chap. 5).

2. See Franklin Roosevelt’s “Summary of the Report of the Committee on Administrative Management,” January 12, 1937; Richard Nixon’s “Special Message to the Congress on Executive Branch Reorganization,” March 25, 1971; Mansfield (1969); Nathan (1976); Rohr (1986, chap. 9); National Academy of Public Ad- ministration (1983); and Gore (1993).

3. For a history of the National Performance Review Initiative, relevant documents, and its accomplish- ments, see Kamensky 2001. Significant groundwork for the National Performance Review Initiative had been laid in the previous administration of President George H.W. Bush, who established a Council on Competitiveness in 1989. Bush charged Vice President Quayle with responsibility for reducing the regulatory burden on the economy. On June 15, 1990, the president expressly directed the council to exercise the same authority over regulatory issues as the former Task Force on Regulatory Relief. Executive Orders 12291 and 12498, issued on February 17, 1981, and January 4, 1985, respectively, set forth the specific procedures for the regulatory review process and served as the legal authority for council activities. Initially, the council considered issues in four different areas: preserving free enterprise, access to capital, bringing science to market, and development of human resources. In President Bush’s 1992 State of the Union address, he an- nounced a 90-day regulatory moratorium followed by a 120-day moratorium extension on agency rulemak- ing. During this period, federal agencies submitted lists of then-current rulemaking and cost estimates to the council for review and consideration.

4. There are a variety of analytic and performance-based studies and reports that have been undertaken by foundations (see Pew Foundation Public Safety Performance Project at www.pewstates.org/projects/ public-safety-performance-project-328068), private-sector consulting companies, the U.S. Department of Justice, the National Institute for Justice and state-level studies undertaken by the department of corrections. These sources provide good examples to illustrate a performance-based approach intended to make significant policy changes and to shift public budgeting priorities.

5. We draw on Morgan et al. 2013 (chap. 14, 438–449) to support this discussion.