First in an Occasional Series

Performance Budgeting:Reshaping the State's Budget Process

The Clinton Administration

recently released Its Report Of

the National Performance Re­

view, which proposes ways to

Improve the operations of the

federal government The idea of

restructurlng governmental

operations is equally applicable

at the state level.

This Is the first In an

occasional series of papers

discussing opportunities to

make California government

work better.

Legislative Analyst's Office

October 25, 1993

'II BACKGROUND

It is hard for most people to get excited about the state budgetprocess with its mountain of numbers, intricate details, and end­less jargon. It is a fact of life, however, that the budget processdetermines who gets what services and who pays for them. TheGovernor's 1993-94 Budget proposed to change the state's budget­ing process by pilot testing performance budgeting in four statedepartments because the state's traditional budget process "hasbecome seriously dysfunctional."

- FINDINGS

We have reviewed the use of performance budgeting in otherstates and the Governor's proposal for California. We concludethat nine months after being proposed by the administration, thepilot project still lacks sufficient details and its schedule hasslipped. Nevertheless, we believe that performance budgeting hasmerit and is worth pilot testing. This is because it focuses onprogram results, thereby offering the potential of improving thedelivery of services.

- RECOMMENDATIONS

If the potential benefits of performance budgeting are to be real­ized, the Legislature will need to "break the mold" in how itappropriates funds for the pilot departments and conducts legisla­tive oversight. It will have to focus on longer-term program goalsand outcomes, rather than immediate process and service de­mands. Accordingly, we recommend the establishment of a jointlegislative committee-including representation from the fiscaland policy committees of both houses-to oversee the pilot projectand review the performance and budgets of the pilot projectdepartments.

Legislative Analyst's Office

2

"Performancebudgeting differsfrom the tradi­tional approachto budgeting inthat it focuses onoutcomes ratherthat inputs orprocesses..."

BACKGROUND

What Is PerformanceBudgeting?Stated simply, performance bud­geting is the allocation of re­sources based on an expectationof performance levels, whereperformance is measured in spe­cific, meaningful terms. It differsfrom the traditional approach tobudgeting in that it focuses onoutcomes rather than inputs orprocesses when deciding how toallocate resources. For example,under the traditional budgetingapproach, the number of parksmanaged by the Department ofParks and Recreation and thenumber of state park employeesare input factors that help deter­mine the department's budget.Under performance budgeting,the level of satisfaction ex­pressed by visitors to state parkswould be one of a number ofoutcome measures. Visitor satis­faction could be determined bya simple survey of park-goers,with the results providing anindication as to how well thedepartment was carrying out itsresponsibilities. This measure ofperformance could be used inevaluating a budget request.

Performance budgeting is not anew concept. The City of LosAngeles implemented a versionof it in the early 1950s. Recently,however, there has been re-

newed interest in this budgetingapproach at the state and fed­eral levels. For example, theNational Conference of StateLegislatures identifies 23 statesthat are involved to some de­gree with performance budget­ing.

The federal government hasrecently shown interest in theuse of performance measures, akey component of performancebudgeting. For example, in Feb­ruary 1993 the U.S. General Ac­counting Office (GAO) releaseda report on the uses and limita­tions of performance budgetingin selected states. A primaryobjective of the GAO study wastogather information on theimplications of state experiencesfor federal performance budget­ing efforts. In July 1993, Con­gress passed the GovernmentalPerformance and Results Act,which provides for three-yearpilot projects that will embodymany of the components of per­formance budgeting. The Na­tional Performance Review, asix-month study of the federalgovernment requested by Presi­dent Clinton, resulted in a re­port issued by the Vice Presi­dent in early September 1993,which likewise embraces aspectsof performance budgeting.

Making Government Work Better

3

What Has Been the Experiencein Other States?In its review of performancebudgeting, the GAO focused onfive states (Connecticut, Hawaii,Iowa, Louisiana, and North Car­olina) that had the followingcharacteristics:

• Each regularly publishedperformance measures in itsbudget documents.

• Each reported using a vari­ety of performance mea­sures, including effectivenessand productivity measures.

• Each had used performancemeasures through at leasttwo budget cycles.

Results Have Been Mixed. AsFigure 1 shows, the GAO con­cludedthatperfo~ncebudge~

ing produced mixed results inthe five states reviewed. In addi­tion, the GAO pointed out thatperformance measures-a keycomponent of performance bud­geting-took time to developand implement.

As Figure 1 suggests, while per­fo~nce budgeting may offerpromise, states have had diffi­culty realizing that promise. Partof the problem has been thereluctance of executives tochange their budget decision­making from one based on pol­icy and political considerationsto one based on performance. Inaddition, Legislatures,

It provided helpful budgetary decision­making information, but did not funda­mentally change the bUdget process (seeFigure 2 for details).

It was not the ''final arbiter" of fundingdecisions given the political nature of thebUdget process.

It gave managers greater decision­making flexibility.

Performance bUdgeting complicated thebudget process by highlighting trade-oftsamong programs competing for limitedresources.

Time, resources, and data constraintslimited the use of performance informa­tion by the legislative and executivebranches.

Legislative and executive budget decisionmakers were dissatisfied with and ques­tioned the reliability of performance mea­sures.

Legislative Analyst's Office

4

"...while perfor­mance budgetingmay offer prom­ise, states havehad difficultyrealizing that

• /Ipromlse.

used to considerable control inapproving budgets through em­phasis on process and proce­dure, have had difficulty em­bracing the fundamentally dif­ferent approach of performancebudgeting, which stresses pro­gram missions, goals, and out­comes.

Figure 2 summarizes the specificreasons why performance bud­geting has not fundamentally

WHAT Is CALIFORNIAPROPOSING?

How the AdministrationDescribes PerformanceBudgetingThe Governor proposed in Janu­ary 1993, as part of his 1993-94

changed the budget process inthe five states reviewed by theGAO.

Other States Continue to SeePromise. While results weremixed in the states the GAOchose to examine, many otherstates, including California, havedecided to pursue performancebudgeting, believing that therewill be net benefits.

budget, to pilot test performancebudgeting in four departments.According to the administration,performance budgeting, alongwith quality improvement pilotprojects, offers the potential forsubstantial savings, improvedprogram performance, enhanced

Legislative Analyst's Office

6

"...the Legislaturehas not been pro­videdsufficient detailsregarding the ad­ministration'sperformance bud­getingproject."

How the Pilot DepartmentsWere Selected. The administra­tion has listed the followingcharacteristics of "ideal" candi­dates for pilot departments:

• Department must be com­mitted to all aspects of theproject.

• Department must be wellmanaged currently.

• Department must be pre­pared to initiate strategicplanning activities.

• Department must be me­dium-sized.

• Pilot departments shouldinclude both internal serviceagencies and public serviceagencies.

Based on these criteria, the De­partment of Finance (DOF),which has been given responsi­bility for the performance bud­geting pilot project, selected thefollowing four departments toparticipate in the pilot test: theDepartments of Consumer Af­fairs, General Services, andParks and Recreation, and theStephen P. Teale Data Center.

Legislature Has Shown Interestin Performance BudgetingIn addition to theadministration's interest in per­formance budgeting, similar

interest has been expressed bythe Legislature. Specifically, theLegislature passed SB 500 (Hill),the Performance and Results Actof 1993, which embodies severalaspects of the administration'spilot project. A primary differ­ence, however, is that SB 500(Ch 641/93) establishes a spe­cific timetable for the projectand requires the OOF to evalu­ate project results and report itsfindings to the Legislature byJanuary 1, 1996.

LAO ASSESSMENT OFTHE PILOT PROJECT

In preparing this issue paper,we reviewed available docu­mentation concerning the ad­ministration's performance bud­geting proposal and discussedthe status of the pilot projectwith pilot departments and alsothe DOF. We also reviewed lit­erature on performance budget­ing, focusing on recent experi­ences and current initiatives.Our assessment of the adminis­tration's pilot project to date issummarized in Figure 4 anddiscussed in greater detail be­low.

Definition LackingAt the time the Governor's Bud­get was released in January1993, the administration had notdeveloped the details of the per­formance budgeting pilot

Making Government Work Better

5

citizen satisfaction, and greateraccountability in the delivery ofstate services.

In the administration's view,performance budgeting hasseven essential elements, aslisted in Figure 3.

Annual budgetary contracts between leg­islative budget writers and theadministration.

Operational flexibility, which could includerelief from statutory requirements.

Incentives for performance andefficiency, including the ability toreinvest 50 percent of any savingsinto discretionary activities.

An emphasis on long-term strategicplanning.

~ Development of performance measures.

Benchmarks for measuring operationalefficiency.

[i2f' A commitment to quality improvement.

Making Government Work Better

7

Pilot project lacks sufficient definition.

Despite project schedule slippage,implementation should not be rushed.

Participating departments are only par­tially representative of the range of de­partments in state government.

Implementation costs will occur, andshould be bUdgeted for.

Performance needs to be verifiedindependently.

Sanctions for poor performance shouldbe considered.

Departments may need additional moti­vation to ensure a fair test of perfor­mance budgeting.

project. To date, the Legislaturehas not been provided sufficientdetails regarding the administra­tion's performance budgetingproject. Specifically, the Legisla­ture has not been provided thefollowing information:

• A detailed implementationschedule showing majortasks and milestones.

• How the administration andthe Legislature will be in­volved in developing perfor­mance measures.

• How the actual performanceof a pilot department will beevaluated.

• The form budget contractswill take and how they willbe submitted to the Legisla­ture.

• An estimate of the cost toimplement perfonnance bud­geting and a proposal onhow it will be funded.

Legislative Analyst's Office

8

"Some of the de­lay (of the pilot)may be appropri­ate,if it reflectsa cautionaryapproach to get­ting the pilotdone right ,"

We recommend that the admin­istration provide this informa­tion to the Legislature as soonas possible, but no later than thesubmission of the 1994-95 bud­get.

Project Schedule Has SlippedThe administration initially indi­cated that the pilot projectwould be implemented in thecurrent fiscal year through bud­get "contracts" with the Legisla­ture. The administration advisedthat the contracts would proba­bly be in the form of a letter tothe Legislature, in which theadministration would commit tocertain results in exchange for anapproved budget.

At this time, however, the OOFhas no budget contracts plannedfor the current fiscal year. SenateBill 500 requires implementationof such contracts in 1994-95.

As noted above, the OOF'sschedule has slipped. In part,this is due to the fact that thedepartment had not developedthe type of up-front planningthat it should have. Some of thedelay, however, may be appro­priate, if it reflects a cautionaryapproach to getting the pilotdone right.

Pilot DepartmentsCould Be More DiverseThe four pilot departments se­lected by the administration are

only partially representative ofthe range of departments instate government.

Two of the four pilot depart­ments, General Services and theStephen P. Teale Data Center,are primarily departments thatserve other state departments.Of the remaining two, onlyParks and Recreation has a sig­nificant General Fund allocation(approximately $44 million),which is relatively small whencompared to many other statedepartments. Given the continu­ing difficulties with the GeneralFund, it would seem desirableto include in the performancebudgeting pilot project a depart­ment with a more significantGeneral Fund allocation (forexample, the Department ofJustice). Alternatively, the Legis­lature could consider adding adepartment that is a traditionalcaseload budget (for example,the Department of Rehabilita­tion).

Only Up-FrontCosts Are CertainExperience in other states and areview of the OOF's proposalfor California indicates that,while significant benefits aretypically anticipated, the onlycertainty is that there will becosts to implement the pilotproject. These costs reflect:

"Performancebudgeting, like

many approachesto reinventinggovernment, isneither new nora panacea foraddressing allthe ills ofgovernment."

• Staff time devoted to estab­lishing and maintaining per­formance budgeting.

• The development and main­tenance of strategic plans.

• The development and main­tenance of management in­formation systems to collect,monitor, and evaluate per­formance.

• The implementation andmaintenance of new pro­grams designed to contin­uously improve the qualityof work performed.

The administration's pilot pro­ject proposal mentions activitiesfor which there will have to bean expenditure of resources, butthe proposal does not indicatewhether project costs will bebudgeted or are to be absorbedby participating departments.Such costs could be significant.

Proposal Should IncludeIndependent VerificationWith respect to verifying theresults of the performance bud­geting pilot, the administrationhas not yet spelled out how thisis to occur. Much of the litera­ture on performance budgetingnotes the problems associatedwith the credibility of perfor­mance reports. The difficultywith credibility has oc­curred in other states prirnarily

Making Government Work Better

9

in the legislative branches,where there has been reluctanceto accept performance assess­ments made by departments.This is because legislative staffgenerally have not had the timeor resources to assess the reli­ability of the reported results. Inorder to avoid this situation, webelieve that verification shouldbe performed by an indepen­dent entity, such as the Bureauof State Audits.

Sanctions For Poor Perfor­mance Should Be ConsideredThe administration's proposedpilot project includes rewards asincentives for participating de­partments to do a good job. Inaddition to being able to retainand redirect to other "discretion­ary" activities 50 percent of anysavings achieved during theyear, pilot departments wouldbe freed from certain externalcontrols.

Despite these incentives, it ispossible that a department maynot deliver the promised results.Indeed, increasing managerialflexibility carries with it a poten­tial for increased failure. In suchinstances where this may occur,and depending on the magni­tude and reasons for failure, itmay be appropriate to applysanctions. We think that suchsanctions should not take theform of budgetary or adminis­trative constraints, which could

Legislative Analyst's Office

10

have an adverse impact on de­partmental programs, but bemore in the nature of sanctionsapplicable to those making thepromises and those possessingthe authority to fulfill them (forexample, not granting a payincrease or, in extreme cases,removal from a position).

Additional Motivators and Lon­ger-Term Commitments May BeNeededAs noted above, the administra­tion proposes to allowparticipating departments toreinvest 50 percent of their sav­ings into discretionary activities.It is not clear whether such dis­cretionary activities would allowemployees to share monetarilyin the savings (for example,through a program similar toexisting state programs that al­low monetary awards for exem­plary work). If such sharingwere contemplated, this wouldlikely provide additional moti­vation, and a greater level ofcommitment to perfonnancebudgeting.

While managers may be moti­vated simply by the relaxationof external controls, unless cer­tain controls are relaxed on alonger-tenn basis, the level ofcommitment by managers maybe short-lived. For example, if,pilot participants are providedan exemption from Departmentof Personnel Administration

(DPA) requirements governingorganizational structure, whatwill happen to a departmentthat which has reorganized in amanner not consistent with theDPA's rules once the pilot pro­ject is over? This is an importantconsideration, given that mana­gerial classification and pay lev­els are based generally on cur­rent organizational models.Therefore, if the pilot test is tobe as meaningful as possible intenns of demonstrating the po­tential of perfonnance budget­ing, consideration will have tobe given to extending some ex­emptions for some period oftime beyond the termination ofthe pilot project.

SOME CAUTIONARYOBSERVATIONS

Solid Foundation Is ImportantAs we have noted earlier, anadequate test of performancebudgeting will require severalyears and therefore could span acouple of administrations. Giventhe necessity for both gubernato­rial appointees and other high­level managers to buy into thepilot program, a solid founda­tion is reqUired in order to en­sure the appropriate level ofcommitment from current andfuture administrations. A well­defined pilot program, with ad­ministration and legislative roles

Making Government Work Better

11

clearly agreed upon, would pro­vide such a foundation.

Performance BudgetingIs Not a PanaceaPerfonnance budgeting, likemany approaches to reinventinggovernment, is neither new nora panacea for addressing all theills of government. Thus, theLegislature needs to be realisticabout this approach to changinggovernmental operations. Whilethis approach helps the adminis­tration and Legislature to focuson the results of programs, itdoes not guarantee that depart­ments will improve. There maybe other important changesneeded before a department canimprove its performance. Forexample, performance budgetingwill have limited success if thedepartment has poor managers,inadequately trained staff, prob

lems filling key job classifications, or conflicting objectives.

THE LEGISLATURE'SROLE IN PERFORMANCEBUDGETING

AChange In PerspectiveIs NeededDespite the limited progress todate and the cautionary notesdiscussed above, we believe thatperfonnance budgeting hasmerit and is worth pilot testing.This is because it offers the po­tential of improving the deliveryof services by focusing on pro­gram results. In order to realizethese potential improvements,however, we believe the Legisla­ture will need to change its gen­eral perspective toward the bud­get process, as shown in Figure5.

It needs to be willing to relinquish somecontrols over departments and programs.

It needs to focus on program mission,goals, and outcomes, not on inputs andprocesses.

It must be willing to accept a longer-termview of implementation and results.

Legislative Analys~s Office

12

~ Review proposed budget contracts.

Review and approve performancemeasures.

Consider whether sanctions should beincluded in the pilot project.

the mission, direction, and out­comes of the agency. This is notan easy task, as has been shownin those states which haveimplemented performance bud­geting. However, these changesare a necessary precondition ifperformance budgeting is tohave a chance of success.

Review proposals to exempt pilotdepartments from statutes.

Obtain more specific information from theadministration on its plans to implementperformance budgeting.

Consider adding to the pilot projectother state departments.

In short the steps shown in thefigure indicate that the Legisla­ture has to give up the tradi­tional type of budgetary controlit exercises over departments. Itwould relinquish the more de­tailed short-term control it nowhas over the resources given anagency/hopefully in return formore longer-term control over

If•••it's importantthat theLegislature1Jreaks the mold /

in the way thatit appropriatesfunds forthe pilotdepartments."

Making Government Work Better

13

Different ApproachNeeded to ImplementPerformance BudgetingA changed perspective on thebudget process alone may notbe enough to ensure successfulimplementation of performancebudgeting. This is because per­formance budgeting is such asignificant departure from theway the state has budgeted re­sources in the past. As a result,we recommend that the Legisla­ture consider a completely dif­ferent method to budget re- .sources for the pilot depart­ments. For instance, it is criticalthat both houses provide uni­form direction to the pilot de­partments. In addition, it's im­portant that the Legislature''breaks the mold" in the way

that it appropriates funds forthe pilot departments. In orderto address these unique con­cerns, we recommend the estab­lishment of a joint legislativecommittee to oversee· the pilotproject and review the budgetsof the pilot project departments.The joint committee would in­clude representation from thefiscalcornmittees and relevantpolicy committees of both hous­es. Figure 6 lists the specificactivities we recommend thecommittee carry out.

One of the key functions of thecommittee would be to reviewand approve the performancemeasures. Figure 7 identifies keyingredients for designing theperformance measures.

LiZ

Measures need to focus on outcomes notprocess.

Measures must be relevantto theperformance being measured.

Measures should be customized to fitspecific programs.

Multiple measures should be developedto capture the complexity of programs.

Measures must be reliable-that is, pro­duce accurate and verifiable information.

Legislative Analyst's Office

14

This joint committee would beacting in lieu of the normal bud­geting process. As noted above,this is a significant departurefrom current practice. In makingsuch a recommendation, we areattempting to offer a possiblealternative that still meets theLegislature's needs for input andinvolvement from the member­ship.

CONCLUSION

The Governor's proposal to pilottest performance budgeting pro­vides an opportunity to demon­strate improved governmentaleffectiveness through a newapproach to developing andmanaging budgets. In order toprovide the most meaningfultest of the Governor's proposal,both the executive and legisla-

tive branches will need to coop­erate in specific ways through­out the pilot test. Both branchesalso will need to be willing tochange the manner in whichthey have traditionally dealtwith budgets. In order to enablea cooperative relationship be­tween the executive and legisla­tive branches to work well, theadministration needs to provideconsiderably more detail con­cerning its plans for perfor­mance budgeting. The joint leg­islative oversight committee ,which we recommend, based onthe desire to ensure a meaning­ful test of performance budget­ing, would provide the forumfor a discussion of the adminis­tration's plans, and facilitate theLegislature's participation in thepilot project.

This report was prepared by Bob Dell'Agostino, under the supervision of Craig Cornett. Foradditional copies, contact the Legislative Analyst's Office, State of California, 925 L Street, Suite1000, Sacramento,CA 95814, (916) 445-2375.