Master Public Administration

Chapter31.pdf

Home >Government homework help >Master Public Administration

3 The Partial Adoption of Performance Management Reforms in State Government

This chapter reviews current knowledge on performance management imple- mentation at the state government level.1 From one perspective it looks optimistic. A number of surveys show that state governments have been busy creating per- formance management systems. State governments have mandated that agencies create and disseminate performance reporting requirements, but they have not provided the type of personnel and budgeting flexibility that performance man- agement doctrine suggests is needed. Therefore, we see only a partial adoption of performance management doctrine. In addition, there is little evidence that this information is being used among decision makers in the governor’s office or in the legislature. I illustrate this point with case evidence from my research on the use of performance management in Vermont, Virginia, and Alabama.

The Rise of Performance Information Systems

At the federal level agencies have been producing strategic plans in accordance with GPRA since 1997, performance goals since 1999, and performance results since 2000. At the state level, the 1990s saw a period of adoption of similar types of performance reporting requirements. Surveys have found a high level of adop- tion of performance information systems that create strategic goals and perfor - mance data, based on measurements of formal requirements for performance reporting or surveys of budgeteers or administrators.2 By the late 1990s, thirty-one states had legislative requirements creating performance information systems, and sixteen states had similar administrative requirements. By 2004, the level of adop- tion grew to include all states. Thirty-three states had a performance management statute on the books (listed in table 3.1), and the remaining seventeen states had an administrative requirement.

Co py ri gh t © 2 00 8. G eo rg et ow n Un iv er si ty P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 11/14/2013 4:03 PM via LINDENWOOD UNIV AN: 228430 ; Moynihan, Donald P..; The Dynamics of Performance Management : Constructing Information and Reform Account: 092-800

Evidence collected as part of the GPP confirmed the widespread adoption of per- formance management policies and offered an additional insight. By using content analyses of performance management documents across the states rather than meas- uring requirements for performance management, or relying on self-reporting, it was possible to verify the actual existence of performance information in state doc- umentation and to demonstrate wide variation among states in terms of the range

40 Chapter Three

Table 3.1 Performance management legislation in the states

State Legislation Year Passed

Alabama State Code 41–19-11 1995 Alaska State Code 37.07.010 2002 Arizona State Code 35–113-115.5 1997 Arkansas Act 1463 of 2003 2003 Colorado State Code 2–3-207 2001 Connecticut Sec. 4–7(b) CGS 1985–86 Delaware State Code 70 Ch. 492 and Title 29 Part V Ch. 60B 1996 Florida State Code Ch. 216 1996 Georgia State Code 45–121 1993 Hawaii State Code 101 Sec. 26.8 1970 Idaho State Code 67–19 1994 Iowa State Code Ch. 8.22 2001 Kentucky HB 502 Part 3 Section 35 2002 Louisiana State Code 39–87.2 2003 Maine State Code Title 5 Ch.151-C, Sec. 17.10K-Q 1999 Minnesota State Code Ch. 16A.10 2003 Mississippi State Code 27–103-153 through 27–103-159 1996 Missouri Revised Statutes, Chapter 33.210 2003 Montana State Code 17–7-111 1999 Nevada State Code 353.205 1996 New Mexico MNSA 6–3A-1 2001 Oklahoma State Code 74–9.11 1975 Oregon Oregon State Code 285a.150 1993 Rhode Island State Code 35–3-24, Section 16 Article 1 1996 South Carolina State Code 1–1-820 1995 South Dakota State Code 4–71972 amended 1985;

4–7-35–38 enacted 1994; repealed 1999

Tennessee Chapter 874 of Public Acts 2002 Texas State Code 322.011 under General Government 1993 Utah State Code Title 36 and Title 62A 1997 Vermont State Code Title 32, 307c 1993 Virginia State Code 2.2–5510 and 2.2–1501, -1509, -1511 2003 Wisconsin Act 27 91561997 1997 Wyoming Code Title 28, Section 28 115–116 1995

Source: Melkers and Willoughby, Staying the Course: The Use of Performance Measurement in State Government.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

of performance information produced.3 Table 3.2 provides the result of content analyses of budget, strategic planning, and performance reports. States with higher numerical scores in table 3.2 have built better performance information systems. States increase their standing on the scale if they feature a range of performance in- formation such as quantitative measures/targets and qualitative mission/vision statements and goals, link responsibility for achievement of goals to specific actors, have clear goals and specific measures, show consistency between goals and meas- ures, track performance measures across time and against preset targets, and main- tain consistency of goals across different policy documents.

While all states say they have some type of performance management system in place, and most states require agencies to produce strategic plans and/or performance

Partial Adoption of Performance Management Reforms 41

Table 3.2 Focus on performance information: Range of documented performance data

Range of Range of performance performance

State information State information

Alabama 13.00 Montana 45.00 Alaska 10.00 Nebraska 71.00 Arizona 101.00 Nevada 43.00 Arkansas .00 New Hampshire 0.00 California 33.00 New Jersey 44.00 Colorado 41.00 New Mexico 57.00 Connecticut 38.00 New York 21.00 Delaware 82.00 North Carolina 33.00 Florida 69.00 North Dakota 50.00 Georgia 68.00 Ohio 48.00 Hawaii 63.00 Oklahoma 22.00 Idaho 66.00 Oregon 63.00 Illinois 53.00 Pennsylvania 55.00 Indiana 36.00 Rhode Island 56.00 Iowa 77.00 South Carolina 44.00 Kansas 19.00 South Dakota 24.00 Kentucky 52.00 Tennessee 59.00 Louisiana 98.00 Texas 89.00 Maine 69.00 Utah 52.00 Maryland 55.00 Vermont 47.00 Massachusetts 14.00 Virginia 79.00 Michigan 27.00 Washington 61.00 Minnesota 72.00 West Virginia 21.00 Mississippi 20.00 Wisconsin 16.00 Missouri 81.00 Wyoming 47.00

State average 48.08

Source: Moynihan, “Managing for Results in State Government: Evaluating a Decade of Reform.”

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

reports, the availability and quality of the resulting information can look very dif- ferent from state to state. Even with this caveat in mind, the fact that all state gov- ernments have willingly accepted one of the basic recommendations of performance management doctrine—to build a performance information system—is indicative of the success of at least this part of the doctrine. However, states have been much more reluctant to embrace the second recommendation of this doctrine, that is, to foster managerial flexibility.

Managerial Flexibility: Financial Controls

Many states claim to have loosened managerial authority with financial tools. However, either central executive branch agencies or the legislative branch remains involved in the use of these tools, eliminating any real increase in discretion at the agency or operational level. Levels of centralization on financial controls that con- strain process change—procurement, contracting, and the use of resources—are reported here, based on 2001 GPP survey responses from state financial manage- ment officers. The potential to carry over cost savings across years, a means to en- courage efficiency in operations, is also discussed.

In the area of procurement and contracting, agencies have a fixed dollar sum under which they have discretion (see box 3.1). These sums are relatively low, with an average of $18,300 for procurement and $24,567 for contracting. Above such levels a formal bidding process will be required, over which agencies rarely have dis- cretion without central approval. While the agency may be involved in the process, the governor and/or the executive budget office usually makes the final decision. In only seven states do agencies have complete discretion in formal procurement bids, and only four states for contracting.

A similar pattern exists for discretion in resource allocation. Efforts to switch money between programs, object classifications, and line items may be prohibited or subject to ex-ante approval by the Office of the Governor, the finance depart- ment, or the legislature. In only seven states is there limited discretionary power for agencies to switch money among programs, and this discretion is unlimited in only one state. Controls become more decentralized for smaller resource categories, such as object classifications and line items. Thirteen states claim limited discre- tionary power for agencies to switch between object classifications, and thirteen other states claim unlimited discretion. Eleven states claim limited discretionary power for agencies to switch between line items, and nineteen states claim unlim- ited discretion. In some cases the discretion applies to some but not all agencies. Another frequent and significant limit on resource discretion is the explicit prohi- bition to affect wages or salaries, protecting the largest departmental operating ex-

42 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

penses from change and reinforcing human resource (HR) controls. A final form of financial control is the requirement to return unspent appropriations rather than allow carryovers, denying a possible incentive for greater efficiency in spend- ing resources. While twelve states allow the possibility of carryovers, it is usually subject to uncertain ex-ante central permission. Only two states claim to offer agencies unlimited discretion in carrying over unspent funds.

Human Resource Management Controls

Recent analyses of HR trends in state governments note increased decentralization, most significantly in personnel classifications.4 However, this decentralization is rel- ative, largely reflecting a shift in control of personnel functions from state-level HR offices to agency-level central HR offices.5 Central agencies at the statewide and agency level clearly define the framework within which managers can work, usually guided by some form of civil service legislation. GPP survey data find that managers have high levels of autonomy in establishing performance expectations and are largely responsible for administering performance appraisals. However, they are closely guided by centralized performance-appraisal instruments and scoring sys- tems, and they have little control in determining compensation to reflect perfor - mance. Similarly, classification systems that detail the duties and status of employees are largely centralized at the state level, beyond the control of individual managers. In terms of hiring, managers are usually involved in approval to fill a position, in- terviewing candidates, recommending appointments, and making appointment de- cisions. However, a mixture of the agency and statewide HR agencies first establish

Partial Adoption of Performance Management Reforms 43

Box 3.1 Agency discretion in procurement/contracting in fifty states

Agency discretion in nonformal contracts Yes = 80%

Limit before formal contract required $24,567.69 (average)

Agency discretion for formal contracting Yes = 8%

Agency discretion in nonformal bidding process Yes = 90%

Limit before formal bidding process required $18,300 (average)

Agency discretion for formal bidding process Yes = 14%

External oversight of procurement process Yes = 96%

Source: Moynihan, “Managing for Results in State Government: Evaluating a Decade of

Reform.”

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

the screening, ranking, and selection of a candidate pool. Table 3.3 illustrates these findings, providing average state score on HR discretion, where a six represents managerial control and a one represents central statewide agency control.6

Survey evidence supports the picture painted of statewide performance man- agement systems. Melkers and Willoughby’s survey of executive and legislative budgeteers caution that actual implementation of performance management is still a work in progress, partly due to diverging expectations for use and success among different actors.7 Surveys of senior administrators reported by Brudney, Hebert, and Wright find that the most intensively adopted reforms of the 1990s were related to creating performance information systems, while the least inten- sively adopted reforms were the reduction of HR and financial management con- trols.8 Later work by Burke, Cho, Brudney, and Wright supported this earlier research, finding that factor analysis confirmed the division of 1990s state gov- ernment reforms into two distinct categories relating to managerial authority and results, and that the managerial authority category lagged behind the results cate- gory in terms of adoption.9

Perhaps most telling is the survey results of Anders, who surveyed administra- tors and central budget officers in 1996 and again in 2004. The results show that

44 Chapter Three

Table 3.3 Degree of managerial control of specific HR functions

HR function Average

Recommending appointments 5.40 Establishing performance expectations 5.34 Making appointments 5.26 Administering performance appraisal 4.86 Determining promotions 4.80 Interviewing 4.76 Approval to fill position 3.76 Determining appraisal grading/scoring systems 2.77 Screening candidates 2.59 Ranking candidates 2.55 Developing performance appraisal instruments 2.18 Establishing candidate list 2.12 Developing tests 2.04 Administering tests 2.04 Scoring tests 2.02 Determining compensation 2.00 Developing classification 1.24 Conducting classification 1.94

Note: Managerial authority on a 1 to 6 scale, where 6 is complete managerial control and 1 is complete control by statewide central agency actor. Source: Moynihan, “Managing for Results in State Government: Evaluating a Decade of Reform.”

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

most public managers see the arrival of performance information systems as hav- ing added an administrative burden but not an increase in discretion. These neg- ative perceptions of performance information systems are hardening over time. In both periods a vast majority agreed that performance information systems in- creased the number of rules and instructions they had to comply with (72 percent in 1996 and 73 percent in 2004) and increased the resources needed to comply with them (61 percent in 1996 and 68 percent in 2004). In 1996, 66 percent of respondents disagreed that the introduction of performance information systems had reduced the number of rules involved in budget execution, increasing to 79 percent by 2004. In 2004, only 14 percent agreed that the new performance in- formation systems were accompanied with legislative grants of discretion in budget execution, down from 26 percent in 1996.10

Managers have a variety of different types of discretion, of which discretion over HR and financial matters is only one. Indeed, chapter 5 shows the ability of man- agers to use performance management reforms even without any increases in man- agement discretion. The point, therefore, is not that managers have no authority to use performance information, but that among U.S. states, the increasing emphasis on performance data has not been accompanied by increases in managerial author- ity, contrary to the premise of performance management doctrine.

The Partial Adoption of Performance Management Doctrine

States have embraced the creation of performance information systems, but they have been reluctant to increase managerial authority. This suggests a partial adop- tion of performance management doctrine. HR authority remains largely shared be- tween statewide and agency-level specialists, and there have not been significant increases in authority for operational managers. Evidence suggests that deregulation of key controls largely implied shifting more authority to agency-level personnel specialists, not to managers themselves.11 In short, increases in managerial author- ity did occur, but not to the degree called for by performance management doctrine.

On the other hand, there has been a good deal of activity in focusing on per- formance information systems in recent years. GPP surveys also found state gov- ernments emphasizing performance management as a new wave of reform, and previous reviews suggest limited state government interest in performance man- agement prior to the 1990s, followed by a renewed interest in the topic in imple- mentation of performance information systems during the 1990s.12 Table 3.2 suggests that some states have advanced more quickly than others in producing a wide range of performance information.

If states were pursuing a complete performance management doctrine, we would expect to see a significant positive correlation between the focus on results scale

Partial Adoption of Performance Management Reforms 45

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

indicated in table 3.2 and different measures of managerial authority. We find a pos- itive but far from significant correlation in each state’s focus on performance infor- mation and the overall measure of HR authority (a summary of the data used in table 3.3) of 0.073.13 The relationship between financial management authority and focus on performance information is stronger but still less than implied by per- formance management doctrine. An overall weighted measure of financial author- ity that includes program reallocation powers and carryover power has a 0.265 correlation with focus on results, not significant at 5 percent in a two-tailed test.14

The individual measures of financial authority provide a mixed picture. There are significant or near-significant positive correlations between the limited discretion over object classifications and line items and the focus on performance information measure. However, there is no relationship between the more substantive program measure of reallocation discretion and focus on performance information. Fur- thermore, there is only a weak positive relationship between the ability to carry over funds and focus on performance information. It appears that states that adopt cre- ate a wide range of performance information are more likely to provide a narrow range of increased financial discretion, but not broad increases in either financial or personnel authority. Even the modest indicators on financial authority are over- stated, since low personnel authority severely limits financial discretion. The most prominent limitation on the use of resources is that any changes cannot affect per- sonnel, where most public organizations spend the majority of their budget. Even if managers gain new financial authority, this limitation means that they can redi- rect only at the margins of their operating budget.

This evidence undercuts the proposition that performance management reform is being implemented in stages and that managerial authority simply lags the im- plementation of performance information systems. States that have led the way in focusing on results have not taken on significant parallel reforms of authority and have not viewed the two aspects of performance management as closely connected. Such a connection may be realized in the future as the limitations of the current approach to performance management become evident, but it does not appear to be the current blueprint for reform.

The evidence presented thus far focuses on the states. But what about the federal level? Here, the evidence is not as clear-cut but nonetheless suggests a similar pattern. As other chapters will discuss, the federal government has adopted successive rounds of performance management reforms, first with the Government Performance and Results Act and later in the form of the Program Assessment Rating Tool. At the same time, there has not been an overhaul of governmentwide federal personnel laws, de- spite multiple efforts by both the Clinton and George W. Bush administrations. Under President Bill Clinton and Vice President Al Gore, efforts to reform person- nel rules were met with opposition from both public sector unions and Republicans

46 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

in the legislature.15 The White House tried to find alternative ways to provide flexi- bility to innovators, but without a statutory basis, the effect of such efforts was lim- ited. One exception was the area of procurement, where legislative change did have an impact. But overall, as Rubin and Kelly note, “throughout the 1990s the U.S. at- tempted, but largely failed, to introduce an expenditure management system with fewer input controls in exchange for performance contracts.”16

The Bush White House also proposed expanding managerial discretion and struggled with these efforts until the aftermath of September 11 allowed them to push such management reforms in the Department of Homeland Security and the Department of Defense.17 The pattern follows a more piecemeal approach to HR reform, where specific agencies have won varying degrees of exception from the standard personnel system, to the point that it is difficult to say that a single per- sonnel system still operates. However, this patchwork model of personnel has not, as yet, evolved into anything resembling the performance management ideal model identified in figure 2.1 or seen among NPM benchmark countries. Indeed, focus groups of federal managers reported concerns similar to those expressed by state government counterparts in Anders’s survey: They perceive that they have no greater discretion than before, even as they are being held to more explicit per- formance expectations.18 A survey of federal managers provides a similar finding: 57 percent of non–Senior Executive Service (SES) managers and 61 percent of SES managers felt they were being held accountable for outcomes, but signifi- cantly less—38 percent and 40 percent, respectively—felt that they had the authority to achieve these results.19

High Results Focus and Restricted Authority: Implications for Public Management

Performance management doctrines did indeed exert an influence in the United States, leading to a new configuration of governmental systems. This finding is in itself significant, but what is more interesting is that the outcome did not mirror the performance management ideal-type called for in performance management doctrine and pursued by NPM benchmark countries. Instead, U.S. state govern- ments shifted toward a configuration characterized by a high focus on results but constrained managerial authority, illustrated by box 4 in figure 2.1.

What are the implications of this constrained performance system as a new con- figuration for governance? The performance management ideal-type portrays the potential for improved efficiency and results-based accountability, but performance management doctrine also emphasizes that efficiency gains through process im- provement are inherently limited by constraints on managerial authority. Likewise,

Partial Adoption of Performance Management Reforms 47

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

results-based accountability will not occur when managers lack authority over processes and, ultimately, outputs. Whether performance management has any pos- itive benefit at the managerial level depends largely on the willingness of agency leaders and managers to search for ways to utilize performance management within existing constraints.

Ultimately, therefore, the potential for performance systems where managers seek to improve performance and are formally held accountable to results is undermined by the continued existence of financial and personnel control systems that empha- size compliance and error avoidance. The existence of performance information has not replaced these systems as the fundamental way in which state governments at- tempt to control managerial behavior. The ability of public managers to achieve high-level outcomes is already stretched by factors beyond governmental control, but internal management controls further restrict the ability to direct and improve more modest organizational outputs. There are many valid arguments to be made for such controls, in terms of the need for probity and avoidance of improper use of public power. However, state governments have lurched headlong into the pursuit of results-based government hoping for improved efficiency and results-based ac- countability, while only partially implementing the reforms necessary to achieve these goals and maintaining control systems contradictory to them.

A performance information system ensures that data is available to decision makers. But the availability of performance information does little to improve the capacity or incentives for decision makers to use this information, and their exis- tence tells us little about how decisions are actually made. Providing performance information is a necessary, but not sufficient condition to ensure its actual use. Ul- timately, the most telling standard for performance management is not whether performance information exists, but whether it is used in various decision-making venues in government, from day-to-day management of programs to high-level re- source allocation decisions. It is by incorporating performance data into decisions that governments move from simply measuring results to managing them.

Ascertaining whether and to what degree performance data was actually used in decision processes is difficult. Self-reported surveys tend to be more positive on whether information is being used than evidence from case research. In respond- ing to surveys such as the GPP or those undertaken by the GAO or the Govern- mental Accounting Standards Board, state officials are usually able to give examples of how performance data influenced specific decisions but cannot show any sys- tematic use of performance data.20 At the same time, legislators complain of in- formation overload and poor data quality, expressing mistrust toward the performance data given to them.21

This book relies primarily on case study evidence to try to ascertain if per- formance information is being used and if it matches the doctrinal claims laid out

48 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

in chapter 2, including expected uses in the areas of resource allocation, account- ability, and technical efficiency. The following chapters examine whether these claims have been satisfied and try to explain factors that lead to their occurrence or not. Chapter 4 discusses the potential for political oversight and policy control using performance management benefits of reform, and chapter 5 includes a dis- cussion of responsibility accountability. Chapter 6 deals with the case evidence on resource allocation, and both chapters 5 and 9 examine managerial uses of per- formance information in greater detail. The balance of this chapter introduces the three cases and previews case findings to illustrate how the partial adoption of per- formance management systems affected the doctrinal claims of performance man- agement in Alabama, Virginia, and Vermont.

Performance Management Adoption in Alabama

Performance management in Alabama lacks a glorious past, although it has made some progress. In the 1999 assessment of performance management practices in state government, the GPP issued the state an F grade, finding little evidence of any sort of strategic planning or performance reporting. Such practices had been introduced in previous administrations, but faded away once these governors left office. The only vestiges of previous efforts were a pronounced sense of cyn- icism among long-term employees toward results-based reform and a barely im- plemented legislative requirement for performance management. Under legislation, performance reporting occurs in conjunction with the biennial budget cycle: The governor is required to develop a statewide strategic plan, and agencies are required to produce goals, objectives, and plans for implementation and performance measures. These requirements, passed under Governor Folsom (1993–95), were largely ignored or led to pro forma completion when he failed to win the following election. Subsequent governors did not create strategic plans. Some agencies reported a handful of performance measures or nothing at all. In instances when performance information was reported, it was widely ac- knowledged that such information had little impact on either resource allocation or management, and neither have more recent reform efforts. Performance meas- ures were no longer included as part of the printed budget by 1999, eliminated, says a budget official, because “it just wasn’t worth what it was costing to print extra pages.”

Governor Siegelman (1999–2003) tried to reignite performance management in Alabama. A statewide strategic plan, Achieve: Achieving Accountability for Ala- bama, was created by the governor, policy staff, and officials from the Department of Finance in 1999 and 2000. Achieve and subsequent agency planning adopted a balanced scorecard approach, a technique suggested by the consultants hired by

Partial Adoption of Performance Management Reforms 49

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

the state. Achieve itself explicitly mentions many of the performance management doctrinal benefits identified in chapter 2: improved accountability, efficiency, align- ing individual actions with overall state expectations and strategies, communicating strategic objectives clearly, and linking objectives to resource allocation.22 Following the development of Achieve, Governor Siegelman required five pilot agencies, in- cluding the Alabama DOC, to undertake performance management.

The creation of a statewide strategic plan was seen as the first step in building what was called a Strategic Plan and Performance Measurement System that “aligns individual actions with overall state expectations and strategies.”23 The main link- ing mechanism to provide goal congruence between statewide goals and individ- ual action was the creation of agency-level plans. A logical order is presented in the selection of goals, suggesting that choices are first made at a governmentwide level and become increasingly specific. According to Achieve, the Strategic Plan and Per- formance Measurement System is patterned on the following chronological order of actions:

• Create a mission statement for Alabama. • Determine strategic issues areas, called “focal areas of great importance.” • Determine strategic objectives that provide direction for activities

occurring at the agency level. • Establish the long-term goals Alabama would like to achieve. • Identify performance measures that match the strategic objectives and will

help determine levels of success, prioritize allocation of resources, and create needed accountability.

• Create agency action plans to implement goals. • Results, which are intended to be reviewed and to feed back into every

element of the process for future adjustment of the system.

While this research focuses largely on the Siegelman reforms, it is notable that Alabama has continued to search for new modes of performance management after Siegelman lost his reelection bid in 2002. Siegelman’s successor, Governor Bob Riley, abandoned Achieve and has introduced another attempt at performance budgeting, in the form of SMART, which is short for Specific, Measurable, Ac- countable, Responsible, and Transparent.

Performance Management Adoption in Virginia

Virginia is arguably the foremost exponent of performance management among U.S. state governments, pursuing results-based government in one form or an- other since the 1970s. Central agencies contain experienced specialists on per-

50 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

formance management, and the state has consistently received an A- grade for its performance management efforts across three GPP surveys, the only state to do so. The GAO, the National Performance Review, and others have pointed to Virginia as an exemplar that the federal government could learn from.

By the time Virginia legislated performance management in House Bill 1065 in 2000, it was, unlike Alabama or Vermont, codifying existing practices of per- formance reporting rather than seeking to initiate new policies. H.B. 1065 does not describe the nature of the reporting requirements for agencies, but it gives dis- cretion to the Department of Planning and Budget to determine what perfor - mance information is to be reported with the budget. The Department of Planning and Budget has been fulfilling this role since 1995, at various times requiring issue analysis, environmental scanning, discussion of customer service needs, linkage of performance targets to budget requests, and activity-based costing. H.B.1065 was followed by more comprehensive legislation in 2003, H.B. 2097, known as the Virginia Government Performance and Results Act, which further formalized the process by mandating that each agency develop a strategic plan that included goals and objectives on a three-year basis. Governor Gilmore (1998–2002) also devel- oped a statewide strategic plan intended to provide guiding principles for agency goal-setting and operations. This was followed by Governor Warner’s (2002–6) Roadmap for Virginia’s Future.

Virginia has developed a performance management system that claims to link strategic planning, performance measurement, evaluation, and performance budg- eting. Budget analysts are called on to assess performance information in terms of resource allocation, management decision making, and program improvement. Publicly available performance information is everywhere: in the budget, the gov- ernor’s statewide plans, annual statewide performance reports, agency strategic plans, and in the case of the Virginia DOC, subagency strategic plans. The De- partment of Planning and Budget has developed a web-based central database for such information, whose information has been assessed and audited by one or more of three central agencies with oversight for performance management.

Performance Management Adoption in Vermont

The state of Vermont did not have a performance management system as com- prehensive as Virginia, but it did have greater experience than Alabama. At the statewide level, legislators in 1994 adjusted the appropriations process to require that budget submissions contain a strategic plan for each state agency and depart- ment, including a statement of mission and goals, a description of indicators used to measure outputs and outcomes, identification of the groups of people served by programs, and the strategies for meeting the needs of the agency or program.

Partial Adoption of Performance Management Reforms 51

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

Unlike the other states surveyed, there was no single statewide strategic plan in Vermont. Governor Howard Dean (1991–2003) was not a strong supporter of for- mal statewide planning, preferring to promote policy goals through his policy staff and weekly cabinet meetings, and his successor Jim Douglas proved little differ- ent. The most formal statement of the governor’s goals appeared in the budget. The main vehicle for producing performance information came from the five overar- ching agencies and subordinate departments through their statutory reporting re- quirements as part of the budget, although the legislation did not lead to systematic use of performance information in resource allocation. Staff at the Department of Finance and Management were all too aware of the political nature of decision making, even at department-level budget preparation decisions, where budgeteers allocate resources based on their expectations of the reactions of other budgeteers at Finance and Management and the legislative branch. However, the Department of Finance and Management pursued performance management anyway. Accord- ing to one department official, this was “because it makes sense,” it was how re- sources should be allocated, and the Department of Finance and Management had a responsibility to make that aspiration as close to a reality as possible.

The Department of Finance and Management has also emphasized the possi- bility of using performance information for managerial decisions, with perfor - mance information becoming part of the decision-making culture of agencies. However, efforts to prompt agency-level managerial uses have been limited. The main vehicle for trying to ingrain performance information into departmental cul- ture and management decisions remains the statutory requirement to report such information in budget submissions. It is not surprising, according to one Depart- ment of Finance and Management employee, that many departments “don’t treat it as a management tool. They treat it as an exercise they do once a year in their budget.”

Matching the Promise of Doctrine with Reality

Chapter 2 identified the promise of performance management, including specific claims about resource allocation, accountability, and performance improvement. This section briefly describes whether performance management reforms met these goals, summarized in table 3.4.

For resource allocation, the findings suggest that the use of performance data is episodic, especially by legislators. In the states examined, performance information was discussed in some instances, but usually prompted by an advocate using in- formation to support a preestablished position, rather than a careful consideration by unbiased actors on what the data meant. This case evidence is supported by other research. For instance, Joyce and Tompkins find some limited evidence of use

52 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

among budgeteers in the executive branch but almost negligible use among legis- lators.24 Melkers and Willoughby found implementation of performance man- agement proceeding slowly. They also found that performance information had not, thus far, affected budget decision outcomes, noting that “while governments continue to strongly emphasize integrating budgeting with performance assess- ment, there is little recognition of how performance measurement applies to budget balancing.”25 Indeed, the most significant problem in the application of performance measurement as cited by state budgeteers was lack of regular use by top management and elected officials and the low priority assigned to performance measures in the budget process.26 The GAO found that even in states with excep- tional performance, budgets use data to inform but not to determine budget allo- cations.27 Anders found that 22 percent of administrators and central budget officers surveyed in 2004 believed that performance information had an influence on political decision makers, down from 25 percent in 1996.28 Other in-depth case studies of performance management efforts have found little evidence that statewide actors use performance information.29

The case findings on agency-level performance improvement are similar. All three DOCs had strategic goals and matching performance targets, and they re- ported actual performance on a regular basis. However, there was little evidence that organizations systematically evaluated this knowledge in the search for im- proved alternative organizational processes. There were some uses of performance

Partial Adoption of Performance Management Reforms 53

Table 3.4 Promised benefits of performance management compared with case findings

Benefit Description

Resource allocation Performance information available and occasionally discussed in budget process in each state, but not used in a systematic way.

External accountability Performance information available to public in all states; range, quality, and accessibility of data better in Vermont and especially Virginia. Little evidence of public use of this information, with exception of Vermont DOC, where some members of public could participate in goal-setting.

Internal accountability Availability of performance information provides opportunity for elected officials to exert greater oversight and control, but there is little evidence that they use performance information to the degree necessary to achieve such control. No evidence of management systems or actions consistent with goal of responsibility accountability.

Performance improvement Some evidence that agency staff used performance data for operational improvement, but its use was not systematic.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

information by managers, explored in chapter 9, but there was little evidence of managers changing processes for the purposes of performance improvement.

For external accountability, the case evidence suggests that, while performance information is indeed provided in often impressive detail, it is uncertain whether the general public has been really demanding it or knows what to do with it. An- other and somewhat higher standard of external accountability would be that the public was involved in setting goals, a standard largely unmet by the states studied.

For internal accountability, performance information provides improved ca- pacity for elected officials to exert both oversight and policy control. However, elected officials have largely shown a disinclination to actually use performance management for these purposes. In part, this is because they are unlikely to be a willing audience for performance data for any purposes, and performance infor- mation is not always of a quality to provide useful oversight. The process of se- lecting goals provides an opportunity for elected officials to exert explicit policy control by requiring agencies to ensure that their strategic plans are consistent with their goals. Central agencies usually review these plans, but direct interven- tion is the exception rather than the rule. There are few goals critical to the gov- ernor, and the agency is likely to already have made an effort to reflect this. In any case, broad gubernatorial goals can usually be easily satisfied with the existing policy objectives of agencies.

Performance management doctrine also suggests the possibility of responsibility accountability, where bureaucrats would be directly held to account for performance. The doctrinal claim is that improvement comes when the bureaucrat has clear goals to achieve, with a set amount of resources, and is deemed responsible for the achieve- ment of these goals, which may be tied to financial incentives. The motivated bu- reaucrat figures out more efficient means of delivering these goals. This model of responsibility accountability has clearly not been implemented in U.S. state govern- ments. Senior managers in agencies have performance goals to achieve, but there is not an expectation that ex-post performance will be carefully examined and have an impact on their tenure or finances. In part, this is because strategic goals are not treated as a contractual promise, and performance results do not bring about bene- fits or punishments. Indeed, it is hard to imagine how this might work in the U.S. system, where strategic goals are established prior to legislative resource allocation and not renegotiated after budgeting decisions are made.

Conclusion

Evidence on the adoption of performance management offers a number of in- sights. States have been creating performance information systems, although the

54 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

range and quality of the information that results varies a great deal. States have been less active in adopting recommendations to provide managers with more author- ity, suggesting a partial adoption of performance management doctrine. On the critical issue of use, evidence from the three cases suggests that performance in- formation was not widely used, and as a result, the predicted benefits of perfor - mance management were largely unfulfilled. Later chapters expand on this point, but they also suggest that there are additional uses of performance information that are positive, though largely unanticipated by performance management doctrine.

State governments have not begun to identify and respond to the weaknesses of the current approach to performance management. The 2005 GPP survey found that aspects of performance management that can be converted into standard pro- cedures are spreading. Aspects of performance management that require behavioral change on the part of users remain weak, in particular the issue of performance in- formation use. More states are creating performance information than ever before. But in some cases, they are failing to build off past efforts. Instead, a new model of results-based government is presented by successive governors. For instance, the most recent round of GPP analysis credits Governor Bob Riley’s efforts to intro- duce performance budgeting in Alabama. But this effort is the third such per- formance budgeting reform in the last decade. The previous efforts of Governor Siegelman and Governor Folsom were introduced with fanfare, but they never seemed to affect agency decision making and died quietly.

A skeptical viewpoint suggests that such an episodic approach to performance management is doomed to a familiar cycle of failure. Lots of time and energy is de- voted to creating a performance information system that fails to be used by agency officials, who assume it will disappear over time and who are proved correct when it is abandoned by a new governor anxious to put his or her own mark on govern- ment reform. While Alabama may be an outlier in terms of repeated failures, the 2005 GPP reports that many states are in the same position: starting to embrace performance management; in the process of planning, building, or improving a performance information system; optimistic about the future of results-based gov- ernment; and in some cases able to offer examples of how performance informa- tion has made a difference. The 2005 GPP report offers numerous examples. Alaska has “positive momentum”; Wyoming’s “recent adoption of a results-based accountability model are promising developments”; “although Arkansas does not currently have a statewide strategic plan, there is momentum in the state due to the Governor’s new performance-based budgeting system”; in Georgia “the agency strategic planning system appears somewhat fragmented, but new requirements just passed may add value to the process”; planning in Illinois and Kansas are both described as “a work in progress”; “Maryland became in engaged in strategic plan- ning in 1997, but has yet to fully implement the process.”30

Partial Adoption of Performance Management Reforms 55

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

For many states, the promise of performance management looms large, and the major gains and benefits of performance management appear just a couple of years away. But at this stage in the development of performance management systems in state government, it is reasonable to expect that states would have progressed be- yond this point.

Notes

1. Some of the cross-state comparative analysis in this chapter previously appeared in Moynihan, “Managing for Results in State Government.”

2. Anders, “Performance Measures as Strategic Management Tools”; Brudney, Hebert, and Wright, “Reinventing Government in the American States”; Melkers and Willoughby, “The State of the States”; Melkers and Willoughby, Staying the Course; Office of Program Policy Analy- sis and Government Accountability, A Report on Performance-Based Program Budgeting.

3. For more detail on how the measures were constructed to track performance informa- tion systems, human resource authority decentralization, and financial control, see Moynihan, “Managing for Results in State Government.” The data drew from public documentation and GPP survey responses, and the content analyses of public documents employed multiple coders for each documenting, with intercoder reliability of 0.83.

4. Selden, Ingraham, and Jacobson, “Human Resources Practices in State Government.” 5. Moynihan, “Managing for Results in State Government.” 6. Details of how these measures were developed can be found in Moynihan, “Managing

for Results in State Government.” 7. Melkers and Willoughby, “Budgeters’ Views of State Performance-Budgeting Systems.” 8. Brudney, Hebert, and Wright, “Reinventing Government in the American States.” 9. Burke et al., “No ‘One Best Way’ to Manage Change.”

10. Anders, “Performance Measures as Strategic Management Tools.” 11. Selden, Ingraham, and Jacobson, “Human Resources Practices in State Governments.” 12. Denhardt, “Strategic Planning in State and Local Government,” 174–79. Other support

for this claim comes from table 3.1, where of the thirty-three states listed, all but two saw the legislative changes occur during the 1990s or since 2000. Brudney, Hebert, and Wright, “Re - inventing Government in the American States”; Snell and Grooters, Governing-for-Results: Legislation in the States; Seong, “Adoption of Innovation.”

13. Georgia did not offer responses to the questions upon which the HR data was based. The state has significantly decentralized its personnel system, essentially eliminating its civil service system. It could be argued that Georgia most closely resembles Texas in this regard, and if we were to give Georgia a Texas score of 4.83, then the correlation rises to 0.148.

14. A simple average of the four measures of financial authority (program reallocation pow- ers, objection classification reallocation, line-item reallocation, and carryover power) has a 0.308 correlation with focus on results, significant at 5 percent in a two-tailed test. However, the use of a simple average as an overall measure would be misleading, incorrectly implying that the four

56 Chapter Three

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

measures are of equivalent importance, e.g., that the ability to move around resources among line items is equivalent to the power to move resources among programs. The weighted meas- ure reflects differences between the measures. Details on how this measure was created can be found in Moynihan, “Managing for Results in State Government.”

15. Moynihan, “Public Management Policy Change in the United States 1993–2001.” 16. Rubin and Kelly, “Budget and Accounting Reforms,” 576. 17. Moynihan, “Protection versus Flexibility.” 18. Anders, “Performance Measures as Strategic Management Tools”; Dull, The Politics of

Results. 19. U.S. General Accounting Office, Observations on the Use of OMB’s Program Assessment

Rating Tool, 23; Frederickson and Frederickson, Measuring the Performance of the Hollow State. 20. U.S. Government Accountability Office, Performance Budgeting: States’ Experiences Can

Inform Federal Efforts. 21. Joyce and Tompkins, “Using Performance Information for Budgeting.” 22. Siegelman, Achieve, 3, 5. 23. Ibid., 5. 24. Joyce and Tompkins, “Using Performance Information for Budgeting.” 25. Melkers and Willoughby, Staying the Course, 17. 26. Ibid., 26. 27. U.S. GAO, Performance Budgeting: States’ Experiences Can Inform Federal Efforts. 28. Anders, “Performance Measures as Strategic Management Tools.” 29. Aristigueta, Managing for Results in State Government; Franklin, “An Examination of Bu-

reaucratic Reactions to Institutional Controls.” 30. Quotes taken from state reports prepared by Barrett et al., “Grading the States ’05.”

Partial Adoption of Performance Management Reforms 57

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .