 |
|
||
This first appeared as an ASPANET On-Line column in November 2002.
The Unsolved Puzzle: Performance Budgeting at the Federal Level (Part 1)
By John M. Kamensky
This four-part series describes some of the challenges of moving to a performance budget. The first part lays out the context and the basics of the federal budget; the second part describes a series of ongoing OMB initiatives; the third part analyzes some of the challenges facing OMB and the agencies that stem from these initiatives; and the concluding part offers some guideposts for future action.
Office of Management and Budget (OMB) Director Mitch Daniels at the June 2002 Performance Consortium conference declared that acting on performance and budget integration under the President’s Management Agenda “is Job One.” And there has been tremendous progress. In the past 18 months, OMB has created a broad framework for how this might be done. And it has detailed some of the specifics in guidance to agencies.
However, at this point, agencies are facing enormous challenges in solving the puzzle of getting the various pieces of guidance to fit together. One somewhat exasperated official in one department told me: “If they [OMB] are saying all this, surely they must have a plan.” Maybe. Maybe not!
This exasperation can be expected, in part, because of the inherently confusing nature of the budget process; in part, because OMB’s directives are targeted to different “players” in agencies that traditionally have not had to work together; and in part, because some conceptual details have not been sorted out. Jonathan Breul, a former OMB official, said at a recent conference: “Linking funding decisions to program performance was a key purpose of GPRA [the Government Performance and Results Act]. However, the challenge of translating this objective into policy and practice may prove to be more difficult and time consuming than some might assume. One major reason is that performance planning, budgeting, and cost accounting generally operate in separate, parallel structures. In addition, performance plans, budget presentations, and cost accounting vary considerably across the departments and agencies in the federal government, depending on missions, organizational arrangements, and other specific operating characteristics of the entity.”
Sorting these out these practical issues over the next couple of years will be essential to making the performance budgeting initiative work. Part of the challenge is clarifying some basic concepts, and part is increasing communication among the various leads of initiatives within OMB and within agencies – and in some cases, resolving differences that cannot be bridged.
As OMB approaches the replacement of its aging computer system that manages the entire federal budget system, it may want to resolve some of these differences before it approaches the technical design work of what the new system should look like. But even more important, OMB must give agencies a better sense of the “big picture” so the overall implementation of performance budgeting is not paralysed by conflicting implementation initiatives. Right now, different actors in the budget process are pursuing their own visions of what the system should look like. There needs to be an agreed-upon vision, or at a minimum, agreed-upon standards and definitions within the Executive Branch and, ideally, the Legislative Branch. This was the hallmark of success in the Texas performance budgeting process – the two branches agreed upon the process, definitions, and the data at the technical level and the political policy makers interpreted the data based on their respective political values.
Some Background of the Federal Budget Process. The federal budget process is inherently complex because at any one point in time, a budget office will be addressing four different budget cycles simultaneously and the “rules” for each may differ. For example:
This series of overlaps are one of the contributing factors to why budget offices strive for consistency! But in addition to the budget cycle, there are a number of management laws adopted in the 1990s that do not easily integrate with each other or the budget process – such as the Chief Financial Officers Act, the Government Performance and Results Act, and the Clinger-Cohen/Information Technology Management Reform Act. However, the elements of these laws are critical to the success of performance budgeting and their successful integration is an important element in moving to a managing-for-results culture.
Paul Posner, director of budget issues at the General Accounting Office (GAO), says that another challenge in integrating budget and performance is bridging the differences between the budgeting and planning cultures. Planners, he says, look to the long-term, are open to new possibilities, focus on outcomes, and tend to be holistic. Budgeteers, on the other hand, tend to focus on the short-term (the coming budget year), on incremental changes (rather than major new departures), on outputs (which can be quantified more easily than outcomes), and tend to “chunk” programs or policies into digestible pieces so decisionmakers can make choices.
Given these dynamics between laws and organizational cultures, however, there is a certain structure to the process. Budgets are generally organized first by department, then by agency, then programs, then activities (and of course there are always exceptions to every rule). For presentation purposes in the President’s Budget, programs have also been historically organized by budget functions that display funding in broad categories reaching across agency boundaries. Beginning in FY 2003, however, OMB has presented the budget by department/agency, not by budget function.
The Budget Account Structure. The President’s budget narrative is organized by department and agency – for example the Department of the Treasury -- and within each agency, there is one or more Budget Account. There is a Program and Finance Schedule for each budget account that details program activities. And the definition of “program activities” varies widely. Former OMB executive, Jonathan Breul, says “Program activities, like budget accounts, may represent programmatic, process, organizational, or other orientations.” For example, for FY 2003, the Treasury’s Bureau of Alcohol, Tobacco, and Firearms had three budget accounts: reduce violent crime, collect revenue, and protect the public. In FY2001, it had two budget accounts: salaries and expenses, and internal revenue collections for Puerto Rico. Many agencies’ budgets still reflect the old approach. OMB’s goal is to move more agencies to the new performance-oriented approach, organized around agency goals, rather than the programmatic activities that are not related to goals articulated in agency plans or, in some cases, the traditional object-classification.
The 13 congressional appropriations subcommittees, however, often have their own approaches, so there are many exceptions to these generalizations. Appropriation bills can have different account structures than those reflected in the President’s budget, so the Congressional Justifications that agencies submit to Congress along with their budget requests must reflect these variations. One agency budget director told me: “The Appropriators like the fact that the account structure is so disjointed – they have more control that way.” Nevertheless, for FY 2004, OMB has encouraged agencies to submit their Congressional Justifications in a performance-budget format.
How many budget accounts are there? GAO conducts periodic surveys of the budget account structure. For FY 2001, it identified 1,159 budget accounts.
In addition to organizing the budget by agency, OMB has historically summarized proposed spending by about 20 major “budget functions.” These functions aggregate spending into major policy areas, such as national defense, income security, transportation, and agriculture. Theoretically, these could be converted to national goals or results areas, but there has been no effort to do this. GAO also wrote a helpful guide about the historical evolution of the budget function classification structure. The subtext of that report is that the existing framework is out of date. To spur changes, GAO will soon wade into this debate by creating a forum that could result in a possible framework for organizing the federal government’s activities around major outcomes.
“Rules of the Budget” Puzzle. To integrate budget and performance information, however, requires actions by more than just the budget offices in agencies. Key players include the planning, financial management, information technology, and the program evaluation offices. Each has been given guidance by OMB that contributes to the implementation of performance budgeting, but not always through common communication channels. As one agency official told me, “Agencies want to comply, but OMB doesn’t make it easy. We don’t get the game rules and they keep moving the goal posts.”
OMB’s basic guidance to agencies on how to prepare their budgets is detailed in Circular A-11 (caution: 8.4 MB file!), which is updated annually in mid-Summer. The FY 2004 guidance was issued in June 2002. In that guidance, and in other directives, OMB has given directions in ten areas that touch on the success of its performance budgeting initiative:
How do these initiatives fit together? Do they? Should they? Who suffers the consequences? Stay tuned . . .
- - - - - -
The
Unsolved Puzzle: Performance Budgeting at the
Federal Level (Part 2)
by John
M. Kamensky
As described in my previous article, OMB has given guidance to agency budget, planning, finance, and technology offices to better integrate budgets with performance. This article describes ten ongoing OMB initiatives to implement this initiative:
The significance of these initiatives is that they reach across the traditional lines of responsibility of the budget, planning, technology, financial management, and evaluation functions in each agency. However, because these initiatives were developed in different parts of OMB and are targeted to their respective functional communities, there may be inadequate communication across these functions and individually they may not see how their efforts need to be coordinated or connected.
Performance-based reporting in the budget, FY 2004. OMB’s FY2004 guidance requires each program activity in an agency’s Program and Finance (P&F) Schedules in the President’s Budget Appendix be covered by a performance goal or indicator in the accompanying Government Performance and Results (GPRA) annual performance plan (see OMB Circular A-11, Section 221). It also notes that “The program activity structure has served as the foundation for aligning performance goals with resources. In many instances, however, the current titles and listings of program activities describe neither true programs nor performance measures. This greatly complicates linking resources with performance measures.”
As a result, agencies are refining and re-naming many of their program activities to more accurately reflect what is going on – as seen in the Bureau of Alcohol, Tobacco, and Firearms example in my last article. Many have historically advocated restructuring the budget account system, but this requires formal congressional support. Restructuring – and in many cases just renaming – program activities beneath each budget account may achieve the same end, but this can often be done via staff-level congressional involvement.
Some agencies describe their obligations by program activity differently in their GPRA annual performance plan differently than they do in their budgets. These “GPRA program activity structures” can consolidate, aggregate, or disaggregate program activities in the P&F Schedules. While it may make understanding an agency’s performance commitments clearer, it adds a layer of complexity in explaining how funding is related to results because the dollars come from individual program activities in different budget accounts. A good example of this is the clean air strategic objectives pursued by the Environmental Protection Agency (as described by GAO in its January 2002 report, page 19), which is comprised of various program activities in three different budget accounts.
Performance-based reporting in the budget, FY 2005. OMB says that agencies should expect an extensive restructuring of the FY 2005 budget display that will show a “much greater alignment between resources and performance.” Agencies will be asked to both rename more of their obligations by program activity listed in their P&F schedules as well as restructure entire budget accounts – and develop informational tables matching resources with outputs and outcomes. (see A-11, Section 221.2) OMB also provides some sample formats of what this might look like, such as a “before” and an “after” example for the Fish and Wildlife Service, and also promises more guidance in the future.
Full Costing for Program and Finance Activities, FY 2004. OMB directed agencies to determine the full costs for each program activity, as defined in the President’s budget. Agencies are to prepare their budget requests on that basis (see A-11, Section 221.3). This means that agencies will have to reallocate costs currently charged to other budget accounts or other obligation by program activities. This has implications for agencies’ financial statements as well, especially as agencies increasingly move toward activity-based costing. As agencies begin asking for budget authority where resources are used, this should simplify the relationships between budget and accounting systems, since accounting records resources where they are used.
Performance Assessment Rating Tool, FY 2004. In the FY 2003 budget, OMB for the first time assessed the performance of about 125 programs. For FY 2004, OMB has developed a standardized tool for conducting this assessment and identified a list of about 250 programs to which it will be applied – about 20 percent of all agency programs. Jonathan Breul, a former OMB official, says this initiative is OMB’s attempt “to shift the burden of proof, asking agencies and advocates to supply evidence of program effectiveness, instead of assuming effectiveness in the absence of evidence to the contrary.” OMB says it will expand these assessments in future years and use this information to help make budget allocation decisions.
The definition of a “program” in this assessment, however, is developed by OMB in consultation with each agency. It does not necessarily conform to the definition used by agencies in preparing their GPRA Performance Plans. It also does not align with the Business Reference Model discussed later.
Government Performance and Results Act Strategic Plan Update, FY 2003. According to GPRA, agency strategic plans must be updated every three years. While most agency strategic plans will not need to submit an update until September 30, 2003, OMB has directed all agencies to submit updated drafts to OMB by March 1, 2003 “so that efforts to restructure the agency’s budget account and define informational displays in the budget documents may commence as soon as practicable” for FY 2005. (see A-11, Section 212.1).
Since these plans require consultation with OMB, the Congress, and stakeholders, they offer a useful opportunity to discuss an appropriate framework for a performance budget.
Government Performance and Results Act Annual Performance Plan Guidance, FY 2004. For FY 2004, OMB required agencies to develop their annual performance plans so that performance information will be collected and reported for each major “program activity,” as defined in the “program and financing schedule” in the President’s Budget (see A-11, Section 221.1).
Integrating Financial and Performance Reports and Accelerating Reporting, FY 2002. OMB has directed agencies to, for the current and future fiscal years, combine their statutorily-required financial reports and GPRA performance reports into a single document. It has also directed agencies to accelerate the release of these reports from their statutory deadlines of March 31 to February 1. By FY 2005, these reports are to be released by November 15 (see OMB memorandum, September 25, 2001).
This is an enormous challenge for most agencies. However, OMB sees the value of combining and accelerating these reports as the ability to make this information available for use in Presidential budgeting decisions, which occurs in December and January before the budget is sent to Congress in February. Also, by requiring agencies to accelerate their reporting, they will be encouraged to create new accounting systems that may provide agency managers more current financial information that can be used during the course of the year for management purposes, not just at the end of the year for accountability purposes.
Business Reference Model and Form 300, FY 2004. In addition to asking agencies to develop performance information by program activity as is reflected in each agencies’ GPRA plans, OMB is also asking agencies to classify all of their activities in the context of a Federal Enterprise Architecture (see OMB Release, July 24, 2002). This Architecture is being developed by OMB as a way of classifying and guiding future technology investments but has a broader implication for how agencies organize and describe their investments in other areas. This Architecture had in it a “Business Reference Model” that identifies three business areas (service to citizens, support delivery of services, and internal operations and infrastructure) that are further decomposed into 137 sub-functions or Lines of Business in which agencies commonly engage. This framework could be the new framework for how the government organizes and budgets its service delivery functions.
OMB requires agencies to submit their capital investment requests along with their overall budget requests. This is done via a “Section 300 Form” which is ultimately linked to the Business Reference Model via OMB’s internal review processes. Norm Lorenz, OMB’s chief technology officer, says that agencies will have to show how their proposed technology purchases will help them meet their performance goals, starting with the FY 2005 budget
Common
Performance Measures, FY 2004.
Separate from the OMB Business Reference Model
initiative, OMB is in the process of identifying
common performance measures for six program
areas. These areas include:
These differ from the Business Reference Model in that they are organized around common outcomes, rather than common business processes. It is not clear how these will fit into the Performance Reference Model, described as part of the Federal Enterprise Architecture.
Guidance on “Getting to Green” in Budget and Performance Integration. OMB included in its FY 2004 A-11 guidance to agencies a section on what it will take for agencies to achieve a “green” on their OMB scorecards for the President’s Management Agenda in the area related to better integrating budget and performance (see A-11, Section 225). This guidance defines examples of good practice in five areas:
It also spells out future consequences for agency managers, that this information will be used to shift dollars to more effective programs and to improve efficiency in program and support functions. The guidance also describes greater year-round collaboration with programs, the development of an integrated performance budget, and the alignment of budget accounts and sub-accounts with outputs and outcomes.
Congressional Justifications. Of course, all of the OMB guidance in the world cannot resolve a key element of budget presentation: What do the Congressional Appropriations Committees want? The agencies’ key budget documents are those submitted to Congress, called the Congressional Justification, and these are formatted, for the most part, based on case-by-case individual negotiations with the staffs of the various Appropriations subcommittees.
Most agencies treat these as property of the Appropriations committees and do not release them publicly until the Appropriations committee publishes them as part of their hearings process. In some cases, the format between the materials submitted to, and used by, OMB are similar. But not in the majority of cases.
However, OMB is putting a stake in the ground on this practice. It encourages agencies to merge their GPRA annual performance plans with their Congressional Justifications (see A-11, Section 221.1 (c)).
But “so what?” Can’t this just evolve through trial-and-error and massive differences in ability and interest between agencies? Conflicting guidance always worked in the past, didn’t it? Stay tuned . . . .
The Unsolved Puzzle: Performance Budgeting at the Federal Level (Part 3)
by John M. Kamensky
In a presentation on performance budgeting in March 2002 at the National Academy for Public Administration, OMB senior economist Justine Rodriguez painted a vision of what it would take to create a federal performance budget. She said that “the key to performance budgeting is to align obligations by program activity within accounts and staff units with outputs” so that program activities match the cost of programs with results, not with inputs or processes. “This puts authority and funding for the full annual cost of programs in the hands of line managers and gives them flexibility in resource use to achieve better results,” she noted. To do this, agencies will have to realign their budget sub-accounts, charge resources to the place they are used, separate program and support accounts, and develop a performance-oriented budget format.
And the formal mechanisms in OMB have been producing guidance to attempt to do just this. However, given the prolific issuance of OMB guidance on budget and performance integration, many federal agencies seem less, rather than more, enlightened. This lack of clarity may be driven by the perception that some of the OMB initiatives have not been placed into a larger conceptual framework.
The important thing, though, is – what do agency leaders need to have clarity on when they go to implement the performance budgeting elements of the President’s Management Agenda? Here are some of the issues that seem to still be unresolved that OMB is either grappling with, or needs to grapple with, so agency chief operating officers can pursue (and invest in new approaches) with better confidence:
Defining the “Performance Budgeting” Pieces of the Puzzle. What does a successful performance budget look like? Even though OMB has issued several elements of guidance, both in the President’s FY2003 budget and in OMB’s FY2004 A-11 budget guidance (Section 225), it is still not clear for many agencies just what is expected of them. GAO’s Paul Posner says the debate over “what is success” in performance budgeting is really about “how high the bar should be set for our expectations.” He says there are four performance budgeting models and that “the political community needs to settle on the model that is the most realistic way to think about performance budgeting.” The four models for assessing the success of performance budgeting include:
In the end, the selection of an appropriate model may be agency-by-agency, developed in discussions between an agency, its OMB examiner, and its congressional appropriations staff. But being clear about the parameters up front would help sort out this piece of the puzzle.
OMB senior economist Justine Rodriguez says a successful performance budget “. . . goes back to the use of a revised Strategic Plan as a template. What does the agency want to accomplish? What are its strategic goals and outcome measures by which it gauges progress?” This then cascades to the outputs and activities that an agency uses to influence each outcome and the logic model, or strategy, used to determine the right activities. She continued, “ The government needs resources to produce outputs; it purpose in producing outputs is to influence outcomes. The relationship between resources and outputs is like an accounting ‘match;’ the relationship between outputs and outcomes is like an econometric equation with sometimes long and variable lags, and with other independent variables and environmental factors. So calculating [the results of a performance budget] in two steps is much more precise.”
Defining the “Scope” Pieces of the Puzzle. How should the overall budget be organized? Should it be organized around intended outcomes? Around specific agencies? Around key business processes? Or all three? The tension here is how do you deal with accountability for spending and management in the context of a performance budget designed to create accountability for results.
Organizing budget priorities around intended results has been done in other governments (such as Oregon, Missouri, and New Zealand). It makes sense because most policy results are not in the jurisdiction of any one, single agency, such as clean air or reducing violent crime. But this requires substantial, up-front engagement of top political leadership.
OMB’s current approach is to organize the President’s budget narrative around agencies. This is reinforced by GPRA’s requirement (Section 1115) that directs “. . . each agency to prepare an annual performance plan covering each program activity set forth in the budget of such agency.”
Reinforcing this approach in the FY 2003 budget, OMB reformatted its budget presentation around major agencies instead of the traditional format around budget function codes. The advantage of this type of presentation is greater accountability and clarity in presentation. The disadvantage is that results are generally achieved by a combination of agencies and the collaboration needed to achieve large policy objectives are further obscured.
OMB recognizes this and to address this weakness in the reformatting, OMB is sponsoring a series of common cross-cutting measures in selected policy areas, such as job training, wildland fire management, etc. In past OMB cross-cutting measurement efforts, metrics were largely around common administrative processes. This effort is around program results.
Nevertheless, the current OMB approach does not create a results-oriented framework for the government. There are vehicles for doing this, such as the creation of a governmentwide performance plan. The staff of the now-defunct National Partnership for Reinventing Government catalogued the agency goals reported in the first annual GPRA performance reports and identified more than 1,000 goals and organized them by major outcomes. So this approach is not impossible, but because it is an inherently value-laden exercise, it would require top-level political leadership involvement.
Organizing the budget around business processes has some appeal because it allows greater cross-agency comparisons for assessing process efficiency. This, however, could be seen as subsuming the importance of program outcome goals in favor of program process efficiencies. Again, this depends on what is seen as the intended vision of “success” for the overall performance budgeting initiative.
Defining the “Program” Pieces of the Puzzle. If program results or agencies are used as the basis for organizing the budget, then obligations by program activity are probably the basis for defining the sub-elements of the budget. For example, the National Science Foundation has 20 major program activities within 5 budget accounts. However, defining “what is a program activity” isn’t as easy as it sounds. There are constant discussions on this between the Foundation, OMB, and Congress – without clear agreements as to what the program activities should be, and how they align with NSF’s intended outcomes.
Generally this micro-level of budgeting is described in an agency’s Congressional Justification, with the audience being the Appropriations Subcommittee staffs. A July 2002 GAO report on the appropriate structure and content of the Internal Revenue
Service’s (IRS) Congressional Justification shows that agencies have the flexibility to provide more performance-oriented materials around program activities than previously thought.
OMB’s FY2004 A-11 guidance defines the link between the budget and agencies’ GPRA annual plans as “being based on the program activities in the Program and Financing (P&F) Schedules in the President’s Budget.” The P&F program activities, however, vary widely between agencies, with the trend being toward more outcome-oriented activities that align with agency GPRA plans. However, agencies also have the discretion to, instead, organize program activities around a “GPRA program activity structure” which can be developed by consolidating, aggregating, or disaggregating program activities in the P&F schedules.
Examples of the range of approaches in the FY2003 budget include:
In addition to the lack of clarity as to what constitutes a “program activity” in the budget and in an agency’s GPRA performance plan, OMB has added another layer of complexity via its PART reviews. These reviews are being conducted on programs, but these seem to be defined differently than the guidance used for defining programs in either the budget or GPRA programs.
In the end, whatever is defined as a “program” needs to be a unit of measure that is helpful in making decisions, and is measurable. But it helps if different budget initiatives use the same definitions.
Defining the “Targets” Pieces of the Puzzle.
Once a goal (either an outcome or an output) is defined, and a metric is developed to measure progress toward that goal, agencies need to develop a target level of performance to be achieved; this will help define the level of resources to be devoted. For example, in the Fish and Wildlife Service, the goal might be to save endangered species, the metric may be defined as an endangered species (and specific ones might be defined), the strategy might be to restore certain lands where the species lives to a wild state, and the target might be to preserve a certain number of acres of land. The resources would be the cost of restoring that number of acres.
This chain of connections between goals and targets is what seems to be missing from the current budget construct. For example, the Bureau of Alcohol, Tobacco, and Firearms (BATF) may have a goal of reducing violent crime, and a target of tracing 225,000 firearms per year. But why this target was set and other contributing factors for “reducing violent crime” are not available for decisionmakers to weigh in judging whether this is the right metric or the right target.
However, BATF does reflect its logic for this target in its performance report and Congressional Justification, where it describes its strategy. But this piece of context is not reflected in the President’s Budget. Absent an understanding of BATF’s strategy, it is difficult for policymakers to identify potential tradeoffs between different strategies for common outcomes across agencies. While this may be too much information to put into the President’s Budget, it should be readily available, say via a website, so decisionmakers have access to a better understanding of the context as to what an agency’s overall contribution might be to a broader national goal, or its rationale as to why the targets it proposes makes sense. For example, in the case of BATF, it cannot by itself “reduce violent crime,” but it does make an important contribution, and tracing firearms is a useful strategy. But is “tracing 225,000 firearms” too much, too little, or just right?
Defining the “Cost” Pieces of the Puzzle.
Full-costing of a “program” depends in large part on the ability of agencies to first be abale to define “program activities” and “performance targets.” However, this does not mean agencies should wait. They can begin by laying the groundwork with conceptual issues like costing the individual process elements, charge-backs between administrative support services, etc. The important element is ensuring the financial management staffs coordinate their creation of a new “chart of accounts” that align with program activities defined in the budget so there is a clear alignment between the two.
OMB says there are four areas that need to be considered when developing full costs of programs: (1) retirement pensions and retiree health benefits, (2) the cost of support services and goods provided centrally, and (3) annual capital usage charges and rent, and (4) the accruing costs of hazardous waste cleanup. OMB sent legislation to Congress in 2001 (S.1612, Title II) to address the first element, full costing of employee pensions and benefits. However, no action was taken. OMB had promised to send legislation on the second component to Congress, but it seems to have stalled in the interagency clearance process.
But agencies do not have to wait for legislation to begin full-costing their programs, if they are clearly defined. The Patent and Trademark Office has done this for support services and its patent and trademark processing processes, and so have a number of other programs that are currently fee-based. It is not an easy process, but an essential step to making budget and performance integration possible.
Strategy is a leader’s hypothesis of how to achieve a pre-determined result. Sometimes this is framed as an agency’s “priorities.” For example, at BATF, the goal of reducing violent crime is pursued by focusing on two strategies – reducing the future number and cost of violent crimes. This in turn is being pursued in four programs: denying criminals access to firearms, safeguarding the public from arson and explosive incidents, removing violent offenders from communities, and preventing violence through community outreach.
For the most part, strategies are not clearly articulated in most agency budgets or plans. Essentially, in the OMB budget process, these pieces are missing. This, however, is where most managers and decisionmakers can influence progress towards goals. This probably needs to be highlighted in the budget decisionmaking process better than it is now. Some agencies successfully use logic models to do this.
Twenty-three “Lines of Business Activities” are being defined by OMB’s technology group as a way of rationalizing technology investments across agencies. But this exercise has a much broader implication for performance budgeting (and ultimately activity-based costing). In fact, the 23 lines of business seem to be an alternative to OMB’s historical Budget Functions.
To date, OMB has developed a “Business Reference Model” with mappings to each agency. For example, subfunctions in the “Services to Citizens, Health” line of business are defined as: illness prevention, immunization management, and public health monitoring.
While many agency Chief Information Officers are using this guidance to help them frame their agency enterprise architectures and investment spending plans, it could also be helpful to engage the budget, performance, and financial management offices in the evolution of this structure to ensure there is an alignment across related performance budgeting activities. This becomes even more important as OMB’s technology group begins to develop its planned “Performance Reference Model.” The proposed Performance Reference Model, as currently conceptualized, focuses on efficiency standards such as phone response times, not performance in elation to program outcome.
There are too many puzzle pieces – and yet not enough. Can this puzzle be solved? Should it be? Stay tuned . . .
The Unsolved Puzzle: Performance Budgeting at the Federal Level (Part 4)
by John M. Kamensky
Can the Puzzle be Solved?
Given the many initiatives OMB has underway to implement performance budgeting, what does it look like from an agency’s perspective? Again, take the case of the National Science Foundation (NSF). It has 3 goals, 5 priority areas, 9 operating divisions, 5 budget accounts, 20 program activities, and 3 “lines of business” activities. None of these are clearly articulated in any one, single location. Its priorities change annually. What should budget decisionmakers focus their attention on? Making their decisions by goals? by priorities? by organizational units, by program activities? And once a budget is in place, how should program managers be held accountable? Again, by organizational units, by program, by goals?, by program activity? by lines of business activity? by how well they pursue the priorities? And how should the public and Congress assess its performance? Again, by organizational units? by program activity, etc.
This is not an NSF dilemma. It reaches across the government. Former OMB executive Jonathan Breul says: “OMB will have to develop a governmentwide architecture for relating budget and performance data.” But is this an analytical, value-neutral exercise that can be conducted by OMB or is this an exercise that must be conducted on the political stage because the framework is inherently rooted in political values over the appropriate role of government?
GAO’s Paul Posner says, “The basic question is how to think about the unit of analysis for budget decisonmaking.” And there probably isn’t one, single way. However, he continues, “The weakness now is there are at least seven voices but no choir leader and no sheet music telling us how all the instruments fit together or could fit together.” He continues, “The defining challenge is how to articulate a broader framework that can take what we are already developing for agencies, programs, unit costs, and business lines – which are important initiatives in themselves – and put them in a broader context of important results that cut across each of these units of analysis.”
While OMB has not yet taken a step in articulating a broader framework, the General Accounting Office plans to do so. GAO in early 2003 will launch an effort that may lead to a possible framework for characterizing government outcomes. Comptroller General David Walker, the head of GAO, will convene a forum in late February 2003 to “bring together key national leaders and experts representing major independent producers, users, and funders of public information to discuss the development of a ‘next generation’ shared information resource on key performance indicators of national position and progress.” This would be done through public-private partnerships building on the nation’s existing statistical system “to enhance public decisionmaking, enhance civic engagement, and reinforce accountability in our democracy,” according to GAO.
We will then be able to see if this is an analytical, value-neutral exercise or if it will spark a political debate over the appropriate role of government. Part of the debate will be – who besides the President should be the governmentwide “outcome owners” for those outcomes reaching across traditional agency boundaries?
Experiences of State Governments. Looking at the experiences of some states may help inspire a federal approach, but any resulting framework has to be something developed by the stakeholders in the federal budget process – in which OMB and/or GAO may want to serve as the convenor.
State government examples, however, are instructive as to the range of approaches that the OMB/GAO stakeholders’ group might consider. For example, Oregon in the mid-1990s developed a series of citizen-based outcomes that were used to inform its budget decisionmaking. This process was led by the governor and was developed bottom-up with citizen participation. It is called Oregon Benchmarks, and the Legislature, for a period, used it as a framework for developing budget priorities. The process, however, seems not to have been institutionalized because it was not formally integrated into the state’s budget decisionmaking processes.
Governor Mel Carnahan of Missouri in 1999 used a different approach. He developed a series of priority outcomes based on a top-down analytic process. This did not last beyond his term in office, though.
The performance budget structures in both Florida and Texas are based on specific outputs and targets, but at the agency, not outcome levels. In the Florida case, detailed program-level performance information is developed for the Legislature to use in assessing the Governor’s budget proposals. In the Texas Strategic Budgeting model, program targets, and specific strategies used to pursue outputs and outcomes by agency, are written into statute. The governor also has a quarterly progress tracking system (see an earlier column on this topic).
In all four of these states however, the common thread was the creation of a common vocabulary among key budget stakeholders. This implies the need to describe the current system in terms all the stakeholders can agree upon, the need to develop a common agreement on terms among the key stakeholders, like “what is a program?” or “what are the key elements that should be included in a budget request?” This cannot be done at the technical level. There is too much that is value-laden in this definitional process – just look at the experience that GAO uncovered when it was asked to examine the IRS congressional budget justification format (as discussed in Part 3 of this series). Developing this description needs to be a collaborative effort, so the stakeholders “own” the resulting standards and definitions and at least implicitly agree to use them.
Given the range of initiatives underway to move toward a performance-based budget, there needs to be a more focused dialog and action around several unresolved issues before the pieces of the puzzle can be put together.
First, if President Bush truly wants an outcome-oriented, performance-based budget, his Administration will have to define broad results areas. This is a necessary prerequisite to restructuring the budget accounts, defining programs activities, developing the financial chart of accounts, applying the business reference model, etc. Since this definition process is probably inherently value-driven, there may have to be a commitment on the White House national agenda to do this. Alternatively, the 50-year old existing budget function structure could be used, but that may not be the best construct for a 21st century government. Alternatively, GAO may drive this debate via its forthcoming initiative on national outcomes to the new Congress.
Second, and related to the first, would be the development of cross-cutting performance goals and metrics – a larger version of the OMB Common Measures project. This may be something that evolves from either the GAO initiative on governmentwide outcomes, or the OMB Enterprise Architecture initiative.
Third, OMB will need to work with agencies to better define “program activities,” probably in the context of what strategies or priorities will be applied. The “strategies” piece is currently missing, yet that is probably what most policy and budget decisionmakers need to have explicitly articulated in order to have a robust discussion of how to achieve intended outcomes. The Texas budget model may be a helpful guide in this regard; since the Secretary of the Cabinet, Albert Hawkins, used to be the Texas Budget Director, someone in the White House knows how to do this.
And finally, defining business activities will be a useful frame for making investment choices, but probably not budget choices. The Business Reference Model effort needs to be cross-walked to the program activity structure, as it gets defined. Likewise, the chart of accounts developed as part of agency activity-based costing efforts needs to parallel the program activity structure as well, so full costs are charged back to programs in a clear way.
President Bush has really put himself on the line to make performance budgeting work. The new Congress may be willing to address at least some of the legislative aspects, such as full costing legislation, and possibly better attention by appropriation committees to performance information being provided through the budget process.
In advance of dealing with these big issues, what signals should observers look for to see if OMB is serious? There are probably two.
Depending on the reader, this series of articles may have been too deep or not deep enough. In either case, I hope it has been thought-provoking. The IBM Center for The Business of Government has recently committed to a line of research on performance-based budgeting, so check there for more on this topic in the coming months.