Posted by François Michel
“Program budgeting as a platform for performance-based budgeting has been approached entirely from the standpoint of strategic planning; however, programs interact with a host of other “structures,” including the structures of control associated with appropriation bills, accounting systems, and organizational structures.” This is the beginning of the conclusion of the illuminating article that Carolyn Bourdeaux, from Georgia State University, dedicates to the State of Georgia’s effort to develop a program-based budget (Public Budgeting and Finance, Summer 2008).
Ms. Bourdeaux leads us to adopt successively the different perspectives of the legislature, agencies, and the executive office of the Governor on how programs should be structured. In the program design process, each body strives to maximize its powers in that principal/agent chain that goes from the legislature (principal), the executive (agent of the legislature, but principal of agencies), to agencies (agents). Ms. Bourdeaux vividly shows that up to the point where they acknowledge that too strict controls discourage entrepreneurialism and may be challenging to manage (e.g. a large budget bill), principals favor disaggregated program structures (or, more generally, budget classifications); in contrast, agents want large, fungible programs. Logically, the analysis then turns to the State’s budget control framework.
One interesting aspect of the article is the focus on the number of programs as a vector of budgetary control, with a trade off between transparency and flexibility in the design of the program structure (p. 26). This focus would not be relevant to analyze international examples of program budgets, where the absence of standards on a desirable number of programs is well acknowledged. The major reason for this is that most countries implement a variety of budgetary controls over the budget, or over budget managers, in addition to program structures. For instance, many countries who have implemented program-based budgets, such as Sweden, have maintained some form of line-item budgeting. But even in other cases, there is often a disconnect between the program structure and the “legal level of control”, or the level of specification of appropriations (as Ms. Bourdeaux considers in theory, p.48). In fact, budgetary controls from the legislative branch or from the ministry of finance can be set at any level of the budget classification.
Ms. Bourdeaux also rightly points out that little analytical work is available on how to implement program budget structures in practice, an unexplainable lack in the enormous literature on budget reforms inspired by New Public Management (NPM) methods. Recent works on the topic, including those of Diamond (2003), Robinson and van Eden (2007), and Schick, Dorotinsky, Kim, and Sarraf (2007) (see the recent post: From line item to program budgeting), allow identifying some major principles for a successful design of programs, such as: respecting existing administrative structures (this departs from the initial movement of NPM reforms), limiting the number of sub-programs, allocating resources to define consistent cost centers, etc. However, a definitive account of practical issues involved in program design has yet to be issued.
At this juncture, it is crucial to remember that designing a program structure typically involves defining budget control systems not only at the center (between program managers and their principals from the executive or the legislative branch), but also within programs (between program managers and the managers of administrative units that will collaborate to the execution of the program). That being said, an account of the major practical difficulties arising in developing program structures could look at the four following areas:
- First, programs need to provide a sound basis for policy formulation, or strategic planning. This is where we find appropriation issues between outcome and outputs, or conceptual issues surrounding medium and long-term outcomes; but it is also the only sector where we find international standards for the classification of governments’ functions (COFOG for instance). For agencies, a typical problem is how program objectives (and associated performance indicators) should percolate to front-line managers of local administrative units.
- Second, a program structure cannot be analyzed without the entire array of associated budgetary controls. Such controls include caps over transfers, but also the very notion of the level of constraining specification of appropriations in the budget nomenclature. In practice, programs constrain agencies not only through their size but also through their composition.
- Third, designing a program budget structure will inevitably involve addressing the question of the relationship between programs and existing administrative units, and of the role of program managers. A related issue is defining the respective responsibilities of program managers and of managers of administrative units that participate in the programs.
- Fourth, other factors influence how programs impact the operations of agencies, which should be taken into account in their design. These include, in particular, the adequacy of accounting, IT, or HR procedures.
References
Bourdeaux, Carolyn, "The Problem with Programs: Multiple Perspectives on Program Structures in Program-Based Performance-Oriented Budgets", Public Budgeting and Finance, Summer 2008
Diamond, Jack, "From Program to Performance Budgeting The Challenge for Emerging Market Economies" (June 2003). IMF Working Paper No. 03/169
Robinson, Marc and van Eden, Holger, “Program Classification” in Robinson, Marc (ed.) “Performance Budgeting – Linking Funding and Results” (2007). See in particular Box 5.2: “Loosening input control without losing control”.
Schick, Allen; Dorotinsky, Bil; Kim, Dong; Sarraf, Feridoun, “Paths Toward Successful Introduction of Program Budgeting in Korea”, in Kim (ed.), “From Line-item to Program Budgeting”, Korean Institute of Public Finance and World Bank (Seoul, 2007).
