510 forum

ON THE LACK OF A BUDGET THEORY

BARBARA L. NEUBY State University of West Georgia

ABSTRACT

Theories of behavior exist in most subfields of political science yet, after seventy years of research, there is sitill no theory of budget- ing. Lack of consensus on the proper focus approach to and status of budgeting theory as well as disagreement on a basic definition of the concept have militated the conceptual discourse and iterative debate necessary to develop theory. The degree of effort required and rewards of theory development also retard theory formation. Conse- quently, progress in budget formation, administration, and politics has not advanced as it might if theory were available to examine and explain budget processes.

INTRODUCTION

Theories are found in nearly every subfiield of political science and public administration, yet budgeting, as an integral part of both or perhaps a subfield in its own right, is basically devoid of theoreti- cal guidance. A comprehensive search of the literature of the past seventy years shows that the bulk of articles do not deal with the topic at all or only mention it in passing. Dozens of works describe one or more aspects of a particular city's budget process, but very, very few discuss theory.

Noted authors and experts in the field of budgeting have ob- served this lack of theory. As early as 1926, Seligman stated that fiscal theory and practice has not been explained nor connected in any clear, detailed fashion. V.O. Key (1940), in what is perhaps the most popular and most often quoted article on the subject, reiterat- ed the general lack of theory. More recent authors and practitioners follow suit (Thurmaier, 1941; Rubin, 1990; Bretschneider and Straussman, 1988; Schick, 1988; Wildavsky, 1988, 1975; Caiden, 1985; Premchand, 1983; White, 1985; Wanat, 1974).

Caiden (1985:501) summarizes the status of theoretical efforts by stating, "The interacting aspects of pubHc budgeting and the uneasy

(132) PAQ SUMMER 1997

tension between empirical and normative concerns preclude a single theory of budgeting." Schick (1988:59) notes that theory is "difficult if not impossible" and Premchand (1983:40), while attempting to deal with theory, still feels there is no overarching paradigm. White (1985:624) concludes with, "In short, budgeting is nearly everything and far more than anyone can comprehend ... These multiple and conflicting purposes make action difficult if nol impossible." Litera- ture revolves around discussions of the concepts and aspects of budgeting, the approach theory should lake, and the status of the theoretical effort.

CONCEPTS AND ASPECTS OF BUDGETING

With previous disclaimers in mind, why is there no commonly accepted and debated theory of budgeting? Implicit in the literary debate are several factors inhibiting theory development. There is no consensus as to what budgeting actually is, whether it is a political, economic or social process. There is no agreement on the scope of a budget theory, no agreement upon the dominant methods or tech- niques of budgeting, disagreement about the approach a budget theory should take, and no agreement about the status of theory itself.

Process

There is no agreement on the "type" of process budgeting is. The literature is filled with references to a political process (Thurmaier, 1995; Bretschneider and Straussman, 1993; Rubin, 1990) or an economic process (Axlerod, 1995; Wildavsky, 1992, 1988, 1975; Pigou, 1947). Schick (1988) and Seligman (1926) have noted that budgeting encompasses so much that it has become a social process. Key (1940) believes that budgeting equals a fuU-blown theory of government, a supremely political process and Wildavsky followed suit in his spirited The Politics of the Budgeting Process (1992, 1988). Figou's economic rationale states that budgeting is based on an equalization of the marginal utility of resources (in Key, 1940:1137) and most of the other authors give some attention to economic factors. Anthony Downs (1957:71) combines the two approaches when he states "budget decisions result from the calculation of which alternative w\\\ gain the most votes" while Rubin (1990) notes that budgeting is a coopted field. Schick (1988) and Premchand (1983)

PAQ SUMMER 1997 (133)

feel budgeting is at a disciplinary crossroads. The general focus is on the allocation of resources but there is

no agreement on whether that is primarily a political or economic function. Effective communication necessaiy for theory building is limited when theorists view the topic within an exclusive framework. Feedback in successive iterations, usually necessary to promote conceptual consensus, therefore is slow in coming, if at all.

Scope of Theory

Another aspect of the budget debate centers on the scope of theory and the number of actors to be included in theory. Many and varied are the functions of government. Whether classified as a strictly political or economic process, budgeting deals with many of these functions and operations. The literature is filled with works describing multi-year planning, "top down" and "bottom up" budget- ing, socially-responsive budgeting, fiscal policy making, conflict reso- lution, uncontrollable and deficit reduction, to name but a few of the proper functions of budgeting (Axlerod, 19'35; Schick, 1995; Thur- maier, 1995; Rubin, 1990; LeLoup, 1988; Pitsvada and Draper, 1984; Premchand, 1983; Crecine, 1969).

Other accepted budget functions include oversight and control of agencies, priority determination, accounting management, and budgeting is considered as an arena for settling political power struggles (Axlerod, 1995; Thurmaier, 1995; Schuman, 1984). Wildav- sky (1992, 1988) notes that budgeting must provide continuity for planning, flexibility to respond to crises, rigidity to control spending, and openness for accountability. The scope of budgeting functions seems tremendously broad.

The range of eligible actors has broadened as well to include nearly every bureaucrat and interest group under the sun (Axlerod, 1995; Rubin, 1990; Caiden, 1985; Sokolow and Honadle, 1984; Premchand, 1983; Crecine, 1969; Downs, 1957). The general public is also implicitly included in broad discussions of priority setting and resource allocation (Crecine, 1969). Budgeting at all three levels of government is no longer monopolized by a few elected officials and agency officials as an infinite variety of actors tries to pass go and collect their two hundred dollars. The multiplicity of actors and the assumed macro-boundaries of budgeting lessen the likelihood of a single theoretical response ( Caiden, 1985).

(134) PAQ SUMMER 1997

Methods and Techniques

What techniques are actually used in budgeting and do they structure the process definitively? No agreement has been reached about the proper role of incrementalism but it remains the dominant budgeting method enjoying the most, sometimes heated, debate. Many have pointed out incrementalism's shortcomings as a "theory" of budgeting and have offered situational-specific alternatives to escape its dominance. Wildavsky (1992, 1988), Lewis (1988), and Lindblom (1%5) form the main camp holding fast to incrementalism as theory as they believe the process dominates reality. The process of changing the base slightly incorporates Lindblom's decision- making of piecemeal, mutual adjustment which allows parties to cooperate within the conflictual process of debating goals and values (Wildavsky, 1992,1988; Braybrooke and Lindblom, 1%5).

Davis, Dempster, and Wildavsky (1966) also use mathematics to demonstrate how incrementalism operates in the budgetary process. Pitsvada and Draper (1984) believe the process is the dominant method because the practices of indexing, multi-year budgeting, continuing resolutions, and preparation of current budgets generally begin with some established base. Budget textbooks (Axelrod, 1995, 1988; Lynch, 1990; Lee and Johnson, 1989) and numerous articles all emphasize the concept of budgetary "base" as a starling point from which increments are added to or subtracted from. Agencies use their preceding year's budgets.

Incrementalism is supported by a common sense view that the budget refiects societal wants which do not change dramatically from year to year (Kamlet and Mowery, 1980). Lewis (1988) further notes that recent budgetary reforms such as Zero-Based Budgeting and Planning Programming and Performance budgeting have not altered the use of a base figure and increments.

Sliding down the scale of acceptance, Schick (1988) and others criticize incrementalism as a sometimes-useful method but believe the broader budgetary scope, with the advent of uncontrollables, off- budget spending, automatic deficit reduction, and other mechanisms lessen the technique's importance (Bretschneider and Straussman, 1988; Gist, 1977; Wanat, 1974). Casting aside incrementalism as a theoretical construct entirely are Rubin (1990), LeLoup (1988), Bozeman and Straussman (1982), and Gist (1977). LeLoup points out that incrementalism declined in the 1970s as the theoretical shift from micro to macro-level processes began. Rubin (1988) declared

PAQ SUMMER 1997 (135)

that the process was "dead" because it was too "global." Criticizing incrementalism as being overly deterministic, Rubin (1990) and Schick (1990) believe that, as an explanatory model, it did not de- scribe the process well. Noting that a budgetary base is not always defined, Bozeman and Straussman (1982) and Gist (1977) strike at the heart of incrementalism theory.

Status of Efforts

Also lacking is a consensus on the stat us of theoretical efforts. Views alternate between those stating theoiy has not kept pace mth events and those believing it has. Much of the recent work described one or more situationally-specific aspects of a budget process promoting same as "theory." Straussman (1985:345) notes much attention to normative theory and Magner and Johnson (1995:452) assert that their particular municipal findings are "consistent with theory." LeLoup (1988) asserts that theory has, generally, kept up with the times and Rubin (1989:80) agrees by stating "recent theory is expanding its scope to include politics of revenues, decision proc- esses, balanced budget attempts, and budget implementation." Rubin later qualifies the behef by stating "we need more descriptive theory" and in the same work also says that budgeting is so complex that no simple theory will be adequate.

Bozeman and Straussman (1982), Schick (1988), and Wanat (1974) all believe that events have outpaced theory. Wildavsky (1988), although the father of incrementalism as the predominant method, realizes that there is no current theory. Premchand (1983) mirrors this approach by saying that, despite historical accounts of budgeting in all societies, recent literature has still produced no overall theory.

Reviewing the literature, it appears that there is no consensus among theories from different disciplines as to an acceptable, opera- tional concept of budgeting, on the status of budgeting as a singular field, on how broad the theory's focus should be, what functions are included or which actors play a role, and on the status of efforts to develop theory. While incrementalism remains the dominant method discussed in budgeting circles, there is mucti diversity as to its suit- ability.

(136) PAQ SUMMER 1997

AN INCONSISTENT APPROACH

The lack of consensus in budgeting concepts spills over into a bipolar split as to the proper approach to theoretical efforts. On the one hand, a suitable approach must be positive and empirical, and on the other hand, it is claimed that budget theory must be norma- tive to be explanatory. Normative t h e o r i s t s stress the all- encompassing nature of budgeting, believing that budget theory should incorporate and explain interaction between public wants and allocation of resources in terms of accepted norms (Bretschneider and Straussman, 1988:305; LeLoup, 1988; Wildavsky, 1988; Crecine, 1969; Musgrave, 1959). For Wildavsky (1975), theory must relate to the values of those budgeting. Key (1940) felt that budgeting would approximate a whole theory of government—a supremely normative process.

Rubin called for more descriptive theory and criticizes Witdav- sky's and Lindblom's incrementalism because it promotes reliance on norms. Schick and Crecine take the only known, published empir- ical and positive approach by searching for elements basic to the generic budgeting process.

Problems are inherent in each approach. Without consensus on operational concepts, empirical theory is difficult, if not impossible, to test (Schick, 1988). Based on varying notions of what budgeting "should be" or "should explain," any normative theory is also impos- sible to prove. The fact that theoretical "efforts" arc divided along these two approaches further militates unified theory.

ESSENCE OE THE DEBATE

A distillation of the literature produces a framework for further research. Among scholars there is no agreement on whether budget processes are political, economic or even social; there is no agree- ment on what functions are included in the process or on the posi- tion of actors; on method or techniques; on whether budget theory is or should be normative or empirical. Last but not least, there is no agreement on the status of theoretical efforts. Rubin (1990) and LeLoup (1988) assert that theory has kept pace with events but neither says what that theory is. The literature revolves around quasi-normative and situationaliy-specific studies. From such one can only note that budgeting is a "process" of allocating resources among conflicting claims, involving many factors and participants.

PAQ SUMMER 1997 (137)

That is all the literature affords. It is not enough. Exposition of budgetary aspects is not theory. Social science is

criticized for not being scientific, for deriving spurious, untestable hypotheses from biased data. Physical science is considered science because it investigates relationship between forces and bodies in their most elemental level. It is positive and empirical, not norma- tive. Because of its positive approach, it is perceived as legitimate by its practitioners and by society. Social science struggles to achieve this status, in part, by attempting a positive approach in a social world full of vague operational concepts and difficult to measure variables.

Budgeting occurs in every society, therefore discernible elements must exist. As a social science effort, if budget scholars want to get on the scientific bandwagon, foster their own legitimacy and become useful, an empirical, core concept of budgeting must be developed that describes the process, its elements, and predicts change. Theo- rists have to stop beating around the budget bush.

Scholars ignore the basic concept of "budgetness" and instead describe and debate the various aspects of a process whose elements they cannot even agree upon. Premchand (1983) noted that there was no settled body of knowledge. Where is the search for common, identifiable elements of the process? Budgeting occurs in every society, therefore discernible elements must exist.

The work of Schick and earlier work of Crecine (1%9) represent such a search for basic, root elements. Schick (1988, 1990:206) be- lieves we must begin by identilying the smallest actions that consti- tute budgeting. His studies induce him to offer "claiming" and "allo- cating resource" as two basic elements of every budget process. His studies show that these elements are subject to rules and procedures which, if uncovered, can predict change.

Crecine (1969) reiterates the need for an atomistic theory stating that theory based on norms can never be proven since norms cannot be tested and are political questions. Norms also vary across socie- ties and over time. To build a useful, testable theory it must be applicable to all societies. This is a tremendous task.

Is empirical, scientific theory impossible to develop and test? Whicker et al (1993) note that the entire fields of political science and public administration are not very amenable to laboratory exper- imentation in an environment of shifting human phenomena and fiuid social relationships. The scholars note only recent mathemati- cal grounding in these two social science fields. The excuse that

(138) PAQ SUMMER 1997

budgeting is too broad a topic in which to build theory does not wash as theories of grander topics, i.e., the cosmos, etc., exist with mathe- matical models in tandem. Or are budgeting scholars incapable of the inductive/deductive theoretical disciplines necessary to tackle a topic subordinate to the cosmos?

Where are the regression analyses, the posit/logic studies, the factor analyses of common budgetary elements? In seventy years of study there have been very few testable hypotheses reported in the literature. Cyert and March (1963) suggested that there may be heuristic rules for complex budget decisions, but did not test them in real-life situations. Sacks and Helmuth (1961) performed one of three regression analyses found in the literature that linked budget outcomes to population, community wealth, and service demand. Their model explained between ten and eighty percent of the vari- ance. They stopped their efforts at this point.

Davis, Dempster, and Wildavsky (1966) attempted the behemoth task of explaining congressional budgeting using linear regression, demonstrating the prominent position of incrementalism. Their findings do not constitute a generalizable theory. Crecine's (1969) book, Govemmental Problem Solving, remains the only major work known applying a testable set of hypotheses with real municipal data to them. Crecine tests the Davis, Dempster, and Wildavsky model using data from Cleveland, Detroit, and Pittsburgh. Crecine's regres- sion models test the theory that budget decisions are made using historical precedents, simple fiscal decision rules, incrementalism, and satisfycing techniques.

He assumes a restrictive revenue environment, stable govern- mental structure, staffmg continuity, and a stable community power network. He devotes 250 pages to explaining model characteristics and questions of validity and provides descriptions of computer sub- routines used. Crecine found that a hierarchy of budgeting rules does exist and becomes simpler as problem complexity increases. He also found that last year's budget provides the base from which decisions are made but that external political constraints are not controlling. Revenues are the primary external limitation as munici- pal budgeting is mainly an internal process. He provides 78 pages of appendices of all data used.

Crecine's work is not perfect but il remains the primary work of tested budgeting theory and provides a framework around which more might be done. Crecine devoted several years of effort to a work of that magnitude. Scholars could use and expand upon his

PAQ SUMMER 1997 (139)

approach. Theory development is time consuming, requiring pa- tience. Budgeting is a human process subject to an infinite variety of potentially irrational and arbitrary decisions, making theory difficult.

Given the normative, descriptive, and atheoretical nature of most budgeting articles, it appears that scholarship need not be scientific. The tenure-producing capital that accrues fiom a Crecine-style effort is apparently an insufficient drawing card to warrant such an effort. Just as much capital and academic legitimacy flow from the normative, descriptive essays that discuss the situation in the city of Bumpsville. Interest in and rewards for theoiry development may not exist. Without the search for root elements of the budget process and the iterative debate that typically accompanies theory develop- ment, theory will not be produced. So what?

Theories in other governmental disciplines such as personnel management and organizational behavior explain human motivation and agency practices, as examples. Practical applied theory has shed a whole new light on public personnel matrers and organizational practice resulting in greater understanding and positive change in those realms.

If we could understand budgeting's root elements and interac- tions as well as we claim to understand personnel motivation and bureaucratic procedures, might we not be on our way to better budgeting? If the average budget process became even slightly more streamlined, lesti rancorous, and divisive, theoretical efforts would be worth their weight in gold. New information might foster better training. How many governments might benefit from these new insights? Perhaps budgeting is so mired in politics not even usable theory can shake it loose to offer practical guidance. Until scholarly disensus becomes consensus and combined theoretical effort is begun with suitable rewards, we, as scholars, budgeters, and gov- ernments will, as Lindblom (1965) says, continue to "muddle through."

REFERENCES

Axelrod, Donald (1995). BUDGimNG FOR MODHRN GC^VRRNMEIVT. New York: St. Martin's Press.

(1988). BUDGRTING FOR MODHRN GOVERNMENT. New York: St. Martin's Press.

Bozeman, Barry and Jcffery D. Straussman (1982). "Shrinking Budgets and the Shrinking of Budget Theory." I'UBLIC ADMINISTRATION REVIEW 42

(140) PAQ SUMMER 1997

(November-December):509-515. Braybrooke, David and Charles E. Lindblom (!963). A STRAIEGY OF DECI-

SION MAKING. New York: Free Press. Brelschneider, Stuart, Jeffery D. Straussman, and Daniel Mullins (1988). "Do

Revenues Forecast Effective Budget Setting?" POLICY SCIENCES 21 (Novem- ber-December) :305-325.

C'̂ iden, Naomi (1969). "The Boundaries of Public Budgeting." PUBLIC ADMINIS- TRATION REVIEW 45 (July/August):495-50L

Crecine, John (1969). GOVERNMENF PROBLEM SOLVING. Chicago: Rand McNally.

Cyert, Richard M. and James G. March (1963). A BEHAVIORAL JHEORY OF LHE FIRM. New York: Prentice-Hall.

Davis, Otto IL, M.A. H. Dempster, and Aaron Wildavsky (1966). "A 'l"heory of the Budget Process." AMERICAN POLITICAL SCIENCE REVIEW 60 (September): 529-547.

Downs, Anthony (1957). AN ECONOMIC THEORY OF DEMOCRACY. New York: Harper and Row.

Gist, John R. (1977). "'Increment' and 'Base' m the Congressional Appropriations Process." AMERICAN JOURNAL OF POLHICAL SCIENCE 21 (May)..Ml-352.

Kamlet, Mark S. and David C. Mowery (1980). "The Budgetary Base in Federal Resource Allocations." AMERICAN JOURNAL OF POLITICAL SCIENCE 24 (November): 814-820.

Key, Vladimer O. (1940). "The Lack of a Budgetary Theory." AMERICAN POLIT- ICAL SCIENCE REVIEW 34 (Dccember):1137-n44.

Lee, Robert D., Jr. (1991). "Developments in State Budgeling Lrends in Two Decades." PUBLIC ADMINISTRATION REVIEW 51 (May/June):254-262.

Lee, Robert, Jr. and Ronald W. Johnson (1989). PUBLIC BUDGELING SYS- TE.MS. Rockville, MD: Aspen.

Ixlxiup, I^nce (1988). "From Micro-Budgeting to Macro-Budgeting Evolution in ITieory and f'ractice," m Irene Rubin (ed). NEW DIRECrnONS IN BUDCif-ri- ING THEORY. Albany: Slale University of New York Press.

Ixwis, Carol (1994). "Budgetary Balance: llie Norm, ("oncepi and Practice in l̂ arge U.S. Cities." PUBLIC ADMINISI RATION REVIEW 54 (November/December): 515-523,

Lewis, Verne (1988). "Reflections on Budget Systems." PUBLIC BUDGETING AND FINANCE 8 (Spring):4-19.

(1952). "Toward a Theory of Budgeting." PUBLIC ADMINISTRATION REVIEW 12 (Win(er):42-54.

Lindblom, Charles E. (1965). IHE INTELLKJENCE OF DEMOCRACY New York: Free Press.

Lynch. Thomas D. (1990). PUBLIC BUDGETING IN AMERICA. New York:

PAQ SUMMER 1997

Pretitice-Hall. Magner, Nace and Ciary G. Johnson (1995). "Municipal Official's

Reaction to Justice in Budgetary Resource AUocation." PUBLIC

ADMINISTRATION OUARIT-RLY 18 (Winter):439^56. Miller, Richard A. (1987). FACL AND MEFHOD. Princeton: Princeton University

Press. Musgrave, Richard A. (1959). A THEORY OF PUBLIC FINANCE. New York:

McGraw-Hill Pigou, Arthur C. (1947). A STUDY IN PUBLIC FINANCE, l^ndon: MacMillan. Pitsvada, Bernard and Frank D. Draper (1984). "Making Sense of the Federal

Budget the Old-Fashioned Way-lncremenlally." PUBLIC ADMINISTRATION REVIRW 44 (September/October):401-106.

Premchand, Aaron (1983). GOVERNMENT BUDGETING AND EXPENDI- TURE CONTROL: THEORY AND PRACTICE. Washington. D.C.: Interna- tional Monetary Fund.

Rubin, Irene (1990). "Budget Reform in St. Louis: Why Does Budgeting Change?" PUBLIC ADMINISTRATION REVIEW 50 (July/August):420-426.

(1990). "Budget Theory and Practice: How Good the Fit?" PUBLIC ADMlMS^rRAnON REVIEW 50 (March/April):179-189.

(1990). m E POLinCS OF PUBLIC BUDCiE'HNCi. Chatham, NJ: Cha- tham House.

(1989). "Aaron Wildavsky and the Demise of Incrementalism." PUBLIC ADMINISIRAIION REVIEW 49 (January/February):78-82.

(1988). NEW t̂ IRECTIOMS IN BUDGETARY THE.ORY. Albany, IVY:

State University of New York Press. Sacks, Seymour and William F. Helmuth (1961). FINANCING GOVERNMENT IN

A METRO AREA. New York: Free Press. Schick, Allen (19W). THE CAPACII"V TO BUDGET. Washington, D.C.: Urban

Institute Pres.s. (1988). "Inquiry into the Possibility of Budgeting Theory," in Irene Rubin

(ed). NEW DIRECTIONS IN BUDGETARY THEORY. Albany, NY: State University of New York.

Schuman, Howard E. (1988). POLITICS AND THE BUDGET: THE STRUGGLE BETWEEN THE PRESIDENT AND CONGRESS. Englewood Cliffs, NJ: Pren- tice-Halt

Seligman, Edwin R.A, (1926)- "ITic Social Theory of Fiscal Science." POLITICAL SCIENCE QUARTERLY 41 (September):354-387.

Simon, Herbert A. (1965) ADMINISTRATIVE BEHAVIOR. New York: Free Press.

Sokolow, Alvin D, and Beth W. Honadle (1984). "How Rural Local Governments Budget: Alternatives to Executive Preparation." PUBLIC ADMINISTRATION

(142) PAQ SUMMER 1997

REVIEW 44 (September/October):373-383-

Straussman, Jeffery D. (1985). "V.O. Key's 'The Lack of a Budgetary Theory -

Where Are We Now?" INTERNATIONAL JOURNAL OF PUBLIC ADMINIS-

TRATION 7(4):345-374.

Thurmaier, Kurt (1995). "Decisive Decision Making in the Executive Budget Pnx:-

ess: Analyzing the Political and Economic Properties of Central Budget Bureau

Analysts." PUBLIC ADMINISTRATION REVIEW .55 (September/Octo-

ber):448-460.

Wanat, John (1974). "Bases of Budgetary Incrementalism." AMERICAN POLTH-

CAL SCIENCE REVIEW 68 (September):l221-1228.

Whicker, Marsha U, RuthAnn Strickland, and Dorothy Olshfski (1993). "lTie Trou-

blesome Cleft: Public Administration and Political Science." PUBLIC ADMINIS-

TRATION REVIEW 53 (November/December):531-541.

While, Joe (1985), "Much Ado about Everything—Making Sen.se of Federal Budget-

ing." PUBLIC ADMINISTRATION REVII-W 45 (September/October):623-630.

Wildavsky, A a r o n (1992). T H E NEW POLITICS O F T H I ! B U D G E T A R Y

PROCESS. Glenview, IL: Scott, Foresman.

_ _ _ _ (1988). P O L m C S O F THE BUDGETARY PROCE.SS. Gtenview, IL: Scot!,

Foresman.

_____ (1975). A COMPARATIVE THEORY OF BUDGETARY PROCESSES.

Boston: Little, Brown.