A quarter century of state budgeting practices.
The analysis is based on surveys of state budget offices conducted in 1975, 1980, 1985, 1990, and 1995. Information also is available for 1970 because the first survey asked about that year as well as 1975.(1) Responses from the states and the District of Columbia were high-50 in 1970, 50 in 1975, 48 in 1980, 43 in 1985, 47 in 1990, and 51 in 1995.
This study is perhaps the longest running one on public budgeting (Lee, 1991a). Although survey data suffer from the necessary subjectivity of respondents, this set of surveys provides insights into how state budget offices perceive their operations and those of other state agencies operating in the realm of budgeting and finance. As will be seen, noticeable trends emerge from the data, and it is significant that these trends are the result of different budget directors and political administrations responding over time to the same set of questions. (An alternative to the survey method is the case study; see Clynch and Lauth, 1991.)
Budget Preparation and Budget Documents
Preparing a budget proposal that suggests a set of recommended policies and stays within whatever financial limits are considered politically realistic has been a prominent issue in public budgeting (Bunch and Straussman, 1993). One approach is to set specific dollar ceilings on budget requests. This approach has the distinct advantage of making agencies prepare requests that include only financially feasible options (Lee, 1992). This method is often called fixed-ceiling budgeting. The opposite approach is to provide no ceilings of any sort, allowing agencies to recommend funding programs at levels they consider appropriate.
Table 1 shows these two approaches (first and last rows). The dollar-level-ceiling technique has been steadily gaining in popularity. Nearly half of the states now report using such a method. The no-ceiling approach, in contrast, has gone up and down in popularity but the general trend is downward. No-ceiling budgeting was reported in 59 percent of the states in 1970 and only 28 percent in 1995.
Table 1 Budget Ceilings Issued by Governor for Use by Agencies in Request Preparation, Selected Years, 1970-1995 Percentage Providing Ceiling Type of Ceiling 1970 Specific dollar-level ceilings 0 Policy ceilings, which encourage or retard the growth of certain types of programs 16 Instructions to maintain only a minimum level of services -- Instructions to maintain the current level of services -- Instructions to include necessary program improvements -- A requirement of a ranking of priorities among programs -- No ceilings of any kind 59 Type of Ceiling 1975 Specific dollar-level ceilings 18 Policy ceilings, which encourage or retard the growth of certain types of programs 42 Instructions to maintain only a minimum level of services 0 Instructions to maintain the current level of services 17 Instructions to include necessary program improvements 26 A requirement of a ranking of priorities among programs 28 No ceilings of any kind 43 Type of Ceiling 1980 Specific dollar-level ceilings 23 Policy ceilings, which encourage or retard the growth of certain types of programs 45 Instructions to maintain only a minimum level of services 20 Instructions to maintain the current level of services 44 Instructions to include necessary program improvements 50 A requirement of a ranking of priorities among programs 73 No ceilings of any kind 28 Type of Ceiling 1985 Specific dollar-level ceilings 37 Policy ceilings, which encourage or retard the growth of certain types of programs 49 Instructions to maintain only a minimum level of services 24 Instructions to maintain the current level of services 56 Instructions to include necessary program improvements 61 A requirement of a ranking of priorities among programs 68 No ceilings of any kind 33 Type of Ceiling 1990 Specific dollar-level ceilings 45 Policy ceilings, which encourage or retard the growth of certain types of programs 60 Instructions to maintain only a minimum level of services 24 Instructions to maintain the current level of services 61 Instructions to include necessary program improvements 65 A requirement of a ranking of priorities among programs 89 No ceilings of any kind 11 Type of Ceiling 1995 Specific dollar-level ceilings 48 Policy ceilings, which encourage or retard the growth of certain types of programs 63 Instructions to maintain only a minimum level of services 26 Instructions to maintain the current level of services 62 Instructions to include necessary program improvements 61 A requirement of a ranking of priorities among programs 78 No ceilings of any kind 28 -- These questions were not asked for 1970.
With today's widespread use of computers, data can be compiled and manipulated in many ways, making it possible to present a variety of budget scenarios without committing massive amounts of personnel time to laborious mathematical calculations. As a result, a state may use several approaches to setting limits on agencies in addition to a dollar ceiling. The survey covered five such approaches. All but one are used by 60 percent or more of the states. The option receiving low support (26 percent) was that of instructing agencies to calculate requests based on maintaining only a minimum level of services. Support for that approach has increased from 20 to 26 percent between 1980 and 1995 (Table 1). Some governments have introduced the minimum-service-level approach as a form of zero-base budgeting, but the approach dearly has shown little gain among the states since 1985.
Support for the other types of ceilings changed comparatively little between 1990 and 1995. Policy ceilings gained three percentage points (63 percent); instructions to maintain the current level of services gained one percentage point (62 percent); and instructions to include necessary program improvements went down four percentage points (61 percent).
An important change is the drop of 11 percentage points in support for ranking programs based on priorities (Table 1). Ranking programs was particularly popular in the zero-base budgeting reform movement, and its decline may signal a slow reversal away from that approach. However, the rankings approach remains the most popular form of budget ceiling, used in 78 percent of the states.
Budget ceilings instruct agencies to request next year's budgets based on certain assumptions, such as their set of priorities regarding rankings. Another approach is for the governor to provide policy guidance in terms of overall priorities and/or guidance by major programs. The data in Table 2 show recent declines in three of the four types of program policy guidance--overall guidance in writing, overall guidance by other means, and major program guidance by means other than in writing. Even though there have been declines, these methods are used in 60 to 70 percent of the states. The one area of growth regarding policy guidance is that of written guidance by major program, rising from 30 percent in 1990 to 39 percent in 1995. This type of guidance started with the lowest percentage of all forms (2 percent in 1970) and remains the least used, but it has shown steady growth over the last 25 years.
Table 2 Program Policy Guidance Provided by Governor for Agency Budget Request Preparation, Selected 1970-1995 Percentage Providing Guidance Type of Guidance 1970 1975 1980 Overall policy guidance--in writing 33 46 62 Overall policy guidance--by other means 22 23 43 Guidance by major program--in writing 2 25 23 Guidance by major program--by other means 27 30 45 Type of Guidance 1985 1990 1995 Overall policy guidance--in writing 67 80 70 Overall policy guidance--by other means 60 77 69 Guidance by major program--in writing 29 30 39 Guidance by major program--by other means 53 71 60
Efforts to include program information in the budget preparation process were common even before the 1950s. During the 1990s, several sets of events have given greater emphasis to this general area. The total-quality-management (TQM) movement has suggested that attention should focus on the results of governmental programs. The focus on results gave rise to the need for measuring whether programs are achieving their intended purposes (Lee and Johnson, 1994). Osborne and Gaebler's popular work, Reinventing Government, has suggested that governmental activities need to be reexamined with an eye to measuring their results; operations should then be reconfigured to improve results (Osborne and Gaebler, 1992).
Reinventing Government, other writings, and a general concern for improving the management of government have led to several efforts at the national level. Currently, the federal government is engaged in a series of reforms aimed at gauging the accomplishments of its programs (Bowsher, 1992; Joyce, 1993; U.S. Congressional Budget Office, 1993; U.S. General Accounting Office, 1992, 1994a, 1994b). The National Performance Review, spearheaded by Vice President Gore, has endeavored to "reinvent" the federal government (National Performance Review, 1994; U.S. General Accounting Office, 1994c). Further increasing the emphasis on program information was congressional passage of the Government Performance and Results Act of 1993 (P.L. 103-62), which requires agencies under the tutelage of the Office of Management and Budget to develop annual performance plans that contain multiyear program information (Finch, 1995; U.S. General Accounting Office, 1995a). In addition, Executive Order 12862 issued by President Clinton in 1993 requires agencies to develop customer service standards and then measure whether those standards are being met.
While these important efforts at reform have been underway in the federal government parallel efforts have been noticeable regarding state and local governments. The Governmental Accounting and Standards Board (GASB), the agency that sets accounting standards for state and local government accounting, has been moving toward recommending "service efforts and accomplishments" (SEA) reporting as part of accounting systems (Governmental Accounting Standards Board, 1994; Harris, 1995). The thrust of this work is to devise means of measuring the results of government programs and then to link those results with accounting information.
Serious methodological and political problems exist in any effort to engage in service efforts and accomplishments reporting. For example, the linkage between police efforts and the control of crime may be tenuous in many situations. Increases in criminal activity, such as drug trafficking, seem to have little or no relationship to the vigilance of police departments. Politically, such measures present problems in that comparisons across political jurisdictions might be attempted but inappropriate. A mayor's effectiveness might be challenged, for instance, when the city's high crime rates were compared with lower rates in other cities of comparable size even though the mayor may have limited control over crime rates.
Regardless of the debates over service efforts and accomplishments reporting, program measurement has gained popularity during the 1990s (Carter, 1994; Connelly and Tompkins, 1989; U.S. General Accounting Office, 1993, 1994d). But its link to budgeting is unclear. To what extent has program performance been incorporated into state budgeting? Is it a growing component of the budget request process? Since its inception the survey has divided program information into two categories. Effectiveness measures define and quantify the ultimate effects of government activities on society or the environment, and productivity measures define and quantify the goods or services being produced. In terms of GASB reporting, effectiveness is equivalent to "outcomes" and productivity, to "outputs."
The survey results for the use of program information in budget requests and revisions in programs are presented in Table 3. Despite all of the recent attention given to program information, budget offices reported a distinct decline in the use of such information in budget requests. Whereas almost all states reported in 1990 that program effectiveness estimates were necessary when agencies asked for new programs or revisions in current ones, only about three-fourths took that position in 1995 (95 percent in 1990 and 72 percent in 1995). Similarly, reported use of productivity estimates in budget requests declined from 86 percent of the states to 67 percent. Why these declines occurred is unclear. One possible explanation is that budget offices were dissatisfied with the information that line agencies produced in justifying their budget requests. They may have viewed having no information as better than having inaccurate or false information. Also governors and legislators may have ignored such information, thereby signaling to budget offices that the information had low utility.
Table 3 Program Information in Budget Requests and Revisions, Selected Years, 1970-1995 Percentage Including Information in Request and revision Type of Information 1970 1975 1980 Program effectiveness estimates included in budget requests for new or revised programs 24 78 79 Productivity estimates included in budget requests for new and revised programs 29 73 74 Effectiveness measures revised when funding levels changed 13 46 49 Productivity measures revised when funding levels changed 22 56 52 Type of Information 1985 1990 1995 Program effectiveness estimates included in budget requests for new or revised programs 95 95 72 Productivity estimates included in budget requests for new and revised programs 79 86 67 Effectiveness measures revised when funding levels changed 57 49 57 Productivity measures revised when funding levels changed 61 51 60
The use of effectiveness and productivity measures did show an increase in one important respect, namely funding level changes. A problem can exist if agencies base expected program results on one set of funding levels and then the actual levels diverge. Agencies routinely receive less funding than they request, creating a need for revising the projected program accomplishments. The 1995 survey found an increase in revising these numbers. Approximately 60 percent of the budget offices (up from about 50 percent in 1990) reported that effectiveness and productivity measures were revised when funding levels changed (Table 3).
Computer technology has fully permeated state budget preparation (Stevens. et al., 1991). All states reported using computers in preparing agency budget requests, in the analysis of agency requests by the central budget office, and in preparing the proposed budget document (Table 4). (The advancing use of computers has necessitated a thorough revision of this segment of the survey instrument for the cycle in the year 2000.)
Table 4 Computer Usage in State Budgeting Selected Years, 1970-1995 Percentage Using Computers Function 1970 1975 1980 1985 1990 Agency budget request preparation 28 60 68 93 98 Agency budget request analysis by central budget office 16 38 52 93 100 Executive budget preparation 16 60 75 95 100
Collecting program information in the budget-request process is one matter, but does that information find its way into the budget narratives or elsewhere into the budget document? Productivity, which is much easier to measure than effectiveness, is twice as likely to be discussed in budget document narrative sections (79 percent for productivity compared with 42 percent for effectiveness). While discussion of productivity showed an increase between 1990 and 1995, budget narratives focusing on effectiveness dropped. Indeed, effectiveness narratives dropped to the level reported in 1980 (Table 5).
Table 5 Inclusion of Program Information in Budget Document, Selected Years, 1970-1995
Percentage Providing Information Type of Information 1970 1975 1980 Effectiveness discussed in budget narrative 10 38 42 Productivity discussed in budget narrative 24 58 63 Effectiveness measures incl. in budget document Most agencies 2 24 28 Some agencies 27 40 43 Total 29 64 71 Productivity measures incl. in budget document Most agencies 8 26 28 Some agencies 37 48 49 Total 45 74 77 Future year projections of effectiveness measures included in budget document 2 18 19 Future year projections of productivity measures included in budget document 8 20 19 Type of Information 1985 1990 1995 Effectiveness discussed in budget narrative 51 50 42 Productivity discussed in budget narrative 67 73 79 Effectiveness measures incl. in budget document Most agencies 26 24 24 Some agencies 49 41 49 Total 75 65 73 Productivity measures incl. in budget document Most agencies 24 33 41 Some agencies 52 44 39 Total 76 77 80 Future year projections of effectiveness measures included in budget document 13 26 33 Future year projections of productivity measures included in budget document 18 37 44
Productivity measures elsewhere in the document have shown a gradual increase since 1975, now reportedly included in 80 percent of state budget documents, although about half the documents display such information for only some agencies. The use of effectiveness measures also showed an increase between 1990 and 1995, but the increase brought them back only to the level achieved 10 years earlier. For 1995, 73 percent of the states reported using effectiveness measures for some or all agencies in their budget documents.
A long-time concern in budgeting is that it should extend beyond the immediate budget cycle to focus upon future years. Multiyear projections are commonly justified as providing decisionmakers with a better perspective on budgetary commitments than annual or biennial budgets. Table 5 shows modest gains between 1990 and 1995 in the reporting of future year projections of effectiveness and productivity. However, productivity forecasts are still more common than effectiveness forecasts (44 percent compared with 33 percent).
The use of program analysis in budgetary decisionmaking has been a popular recommendation for reform, especially since the planning-programming-budgeting (PPB) movement of the 1960s (Lee and Staffeldt, 1977; Posavac and Carey, 1992; Rossi and Freeman, 1993; Sylvia, Meier, and Gunn, 1991; Stevens and Lee, 1981; U.S. General Accounting Office, 1995b). The terms program effectiveness analysis and program productivity analysis were used in the surveys of state budget offices. Effectiveness analysis was defined as determining the relationship between governmental activities and the effects of those activities on society or the environment. Productivity analysis was defined as determining the relationship between program resources (costs) and program goods and services.
Survey results suggest that state governments may have undergone considerable change between 1990 and 1995 regarding which state organizational units conduct program analysis (Table 6). Sharp drops in reported conduct of analysis were reported for central budget offices. Effectiveness analysis dropped from 66 percent of these offices in 1990 to 46 percent in 1995; productivity analysis dropped from 94 percent to 72 percent. These results indicate a reversal in a pattern that has been evident since 1970, namely budget offices increasing their conduct of analysis.
Table 6 Conduct of Program Analysis, Selected Years, 1970-1995 Percentage Conducting Analysis Organizational Unit 1970 1975 1980 1985 Central budget office conducts effectiveness analysis 18 65 53 63 Central budget office conducts productivity analysis 31 65 75 81 Other central staff orgs. conduct effectiveness analysis 2 18 28 22 Other central staff orgs. conduct productivity analysis 8 22 26 22 Most major agencies conduct effectiveness analysis 14 32 44 42 Most major agencies conduct productivity analysis 20 38 41 44 Legislature conducts effectiveness analysis 14 34 42 50 Legislature conducts productivity analysis 16 28 39 50 Post auditor conducts program analysis 4 35 39 38 Organizational Unit 1990 1995 Central budget office conducts effectiveness analysis 66 46 Central budget office conducts productivity analysis 94 72 Other central staff orgs. conduct effectiveness analysis 28 50 Other central staff orgs. conduct productivity analysis 30 51 Most major agencies conduct effectiveness analysis 45 38 Most major agencies conduct productivity analysis 49 35 Legislature conducts effectiveness analysis 51 43 Legislature conducts productivity analysis 51 52 Post auditor conducts program analysis 38 61
Budget offices also reported that major agencies have retreated on their conduct of analysis. Effectiveness analysis done in most major agencies dropped from 45 percent of the states in 1990 to 38 percent in 1995. Similarly, productivity analysis dropped from 49 percent to 35 percent.
According to the budget offices, legislatures have changed little in the last decade regarding productivity analysis, with about half conducting productivity analysis in 1995. Effectiveness analysis conducted by the legislature dipped from 51 percent in 1990 to 43 percent in 1995.
The conduct of program analysis has grown in other central executive agencies and in the post-auditor's office (Walton and Brown, 1990). Other central staff organizations conducting effectiveness and productivity analysis rose sharply from approximately 30 percent in 1990 to about 50 percent in 1995, suggesting that budget offices have shifted this responsibility to other units. Also, the post-auditor's office as the locus of analysis increased from 38 percent in 1990 to 61 percent in 1995 (one of the largest changes to be found in all of the data). Worthy of further research is whether state auditors are taking on the role of program evaluator, similar to the role long performed by the General Accounting Office for the federal government.
Whether analysis is used in decision-making, of course, is a different matter from whether it is conducted (Abney and Lauth, 1987; Stanford, 1992). Analysis can be conducted and promptly ignored, or in other situations analysis might be conducted by a budget office and ignored by that office but then used by the legislative branch in decision-making. Survey results regarding the use of analysis by the executive and legislative branches are presented in Table 7. Executive usage in 1995 was down from 1990. Productivity analysis was used substantially or to some degree in 84 percent of the executive offices in 1995 in contrast with 100 percent in 1990. The decline in executive usage of effectiveness analysis was even sharper, down from 90 percent in 1990 to 72 percent in 1995. The 1995 figure is lower than that of every year except 1970. In contrast, legislative usage remained the same between 1990 and 1995 at 83 percent for effectiveness analysis, but dropped 15 percentage points for usage of productivity analysis (94 percent compared with 79 percent).
Table 7 Use of Program Analysis in Decision Making, Selected Years, 1970-1995 Percentage use Type of Use 1970 1975 1980 1985 1990 Executive budget decisions based on effectiveness analysis Substantial degree 15 20 11 21 26 Some degree 23 59 70 64 64 Total 38 79 81 85 90 Executive budget decisions based on productivity analysis Substantial degree 19 27 21 29 45 Some degree 32 54 64 67 55 Total 51 81 85 96 100 Legislative decisions based on effectiveness analysis Substantial degree 25 20 12 12 9 Some degree 19 48 66 73 74 Total 44 68 78 85 83 Legislative decisions based on productivity analysis Substantial degree 19 20 10 15 13 Some degree 25 32 56 76 81 Total 44 52 66 91 94 Type of Use 1995 Executive budget decisions based on effectiveness analysis Substantial degree 18 Some degree 53 Total 72 Executive budget decisions based on productivity analysis Substantial degree 30 Some degree 54 Total 84 Legislative decisions based on effectiveness analysis Substantial degree 11 Some degree 72 Total 83 Legislative decisions based on productivity analysis Substantial degree 10 Some degree 69 Total 79
The data seem to suggest that the legislative and executive branches have balanced out in their use of analysis. In previous times, such as in 1975, state executives may have had a reputation for being consumers of analysis, while legislatures had a reputation for playing-loose politics. Now, at least in the view of state budget offices, that pattern has changed. The survey data suggest that executives have slid backward in their use of productivity and effectiveness analysis to the level of usage by legislatures.
What changes could be expected in the field of accounting over the 25-year period from 1970 to 1995? As state governments grew in size during this period, one might expect increases in their abilities to track financial activities. One might also expect the use of computers to be widespread in accounting. A particular advantage of computer databases over paper records is the ability to track and manipulate data in a variety of ways (Burk, 1994; Norvelle, 1994).
Just as the application of computers has permeated all major aspects of budget preparation, almost all state accounting systems have been upgraded to provide data in a variety of formats. Accounting information is now available in almost every state on at least five bases--appropriations, departments, organizational units within departments, programs when different from organizational units, and specific levels within programs (Table 8).
Table 8 Bases on Which Accounting System Provides Information, Selected Years, 1970-1995 Percentage Providing Accounting Information Basis of Accounting 1970 1975 Appropriations 72 74 Departments 76 80 Organizational units within departments 62 72 Programs, if different from organizational units 24 60 Specific levels within programs 28 50 Cost in relation to work/tasks performed 4 10 (i.e., cost per unit of work) Basis of Accounting 1980 1985 Appropriations 85 95 Departments 91 100 Organizational units within departments 83 90 Programs, if different from organizational units 74 81 Specific levels within programs 61 65 Cost in relation to work/tasks performed 6 10 (i.e., cost per unit of work) Basis of Accounting 1990 1995 Appropriations 98 98 Departments 96 100 Organizational units within departments 91 96 Programs, if different from organizational units 80 93 Specific levels within programs 69 94 Cost in relation to work/tasks performed 0 2 (i.e., cost per unit of work)
The lack of progress in accounting is notable in only one area Although basically all states report having accounting systems that provide a multitude of information, only one has a cost accounting system (a system that compares costs with the work or tasks performed). Cost accounting, a fundamental aspect of accounting in the private sector where there are great incentives to determine whether a firm is making a profit, has yet to become even a small trend in state governments (U.S. General Accounting Office, 1990).
Collecting and monitoring accounting information are different from controlling operations based on that information. A temptation may be to control when the ability presents itself. In an earlier time when accounting systems were relatively simple, a budget office would have little choice but to allow agencies to operate rather freely. However, as accounting has moved from the paper ledgers of departmental accountants to high-speed computers, a budget office now has access to detailed information and may begin to control spending accordingly.
A pattern of increasing control of the central budget office over expenditures is evident. Basically all state budget offices control by appropriations and almost all by departments. Between 1990 and 1995, control by organizational units within departments increased (from 50 percent in 1990 to 68 percent in 1995). This change suggests that departments, and the people who head them, may have lost some autonomy to the central budget offices. While reform movements such as the reinventing government movement may champion the empowerment of administrators to use innovative techniques in meeting agency missions, the survey data suggest increased centralization. Even control in the minor area of expenditure level has increased (from 20 percent in 1990 to 26 percent in 1995). Some budget offices may be micromanaging line agencies.
Budget Office Personnel
The ability of a state budget office to perform the tasks expected of it necessarily depends upon the capabilities of its personnel. Two key aspects are the level of education and the academic disciplines of the personnel (Hackbart and Ramsey, 1994; Lee, 1991b; Lee and Staffeldt, 1976; Yunker, 1990).
Table 10 indicates the level of education for budget office personnel over the 25-year period. The proportion of professional staff with less than a baccalaureate degree has been under 10 percent since the beginning of the survey, with the current figure at 4 percent. The percentage of staff with no more than a baccalaureate degree has shown a steady decline (from 72 percent in 1970 to 47 percent in 1995). As would be expected, master's degrees are becoming standard. Nearly half of the professional staff members now have a master's degree. Doctorates remain relatively rare--only 3 percent in 1995, while the figure was 2 percent 25 years earlier.
Table 10 Levels of Education of Central Budget Office Professional Employees, Selected Years, 1970-1995
Means in Percent Education Level 1970 1975 1980 1985 1990 1995 Two-year degree or less 9 6 4 5 6 4 Baccalaureate degree 72 59 56 50 49 47 Master's degree 17 30 36 42 41 48 Doctoral degree 2 3 3 4 4 3
The academic disciplines of professional staff in state budget offices show minor fluctuations between 1990 and 1995. Accounting continued its downward slide, dropping 17 percentage points, from 29 percent of the staff in 1970 to 12 percent in 1995. This decline is possibly the result of push-and-pull forces. Accountants may have been pulled away from the public sector to the private sector where salaries are higher, and may have been pushed out of state government to make room for other academic disciplines, most notably public administration.
Business and public administration together accounted for 51 percent of the staff in contrast with 52 percent in 1970. However business administration was more than twice the size of public administration in 1970 (37 percent compared with 15 percent), whereas in 1990, the two disciplines were evenly balanced. Economics remained at about 10 percent of the staff.
Discussion and Conclusion
Major changes in state budgeting have been recorded in the surveys between 1970 and 1995. Decision-making processes use program information and analysis more extensively. Computers have thoroughly invaded all aspects of budgeting. Accounting systems in 1995 were considerably more sophisticated than in 1970. A major concern raised by Schick in his classic 1971 book, Budget Innovation in the States, was being able to crosswalk information from one format to another (203-211). In accounting that problem has been largely resolved, undoubtedly with major assistance from increased computer capabilities. Another major set of changes has involved the increasing educational levels of central budget office staff.
Change on other fronts has been slower. Howard gave special attention at the end of his ground-breaking 1973 book, changing State Budgeting to the need for multiyear projections of information in the budget process (266-368). The survey results show that the inclusion of such projections of effectiveness and productivity measures has grown slowly but steadily. Although state accounting systems changed considerably in the quarter century studied, one aspect showed virtually no change. States simply have not developed accounting systems that track the costs of performing work or providing services. States apparently have felt few incentives to devise such systems and have concluded that their costs outweigh their potential benefits.
The period between 1990 and 1995 saw a mixture of continuing trends and possible reversals. Developments were noted in several respects. States are now more likely to revise effectiveness and productivity measures when funding levels change. Such a development is important in keeping expected program results in balance with available funding. Another trend, and one that has reached its limit of 100 percent saturation, is the use of computers. Effectiveness measures have gained increasing prominence in budget documents over the five-year period. Major growth was recorded in effectiveness and productivity analysis conducted by central staff units other than the budget office and program analysis conducted by post auditors. Gains in accounting control have increased, raising the question of whether budget offices, with more sophisticated accounting systems available, are using these new powers to exert greater centralized control over agencies.
Although previously identified trends continued between 1990 and 1995, there were some noticeable exceptions. Whether these changes constitute mere blips or signal real change will not be known until the next survey is conducted in the year 2000. The ranking of priorities, which has long been a common practice in budgeting, took a tumble. Policy guidance, provided by budget offices to agencies regarding their budget requests, declined whether it was written guidance, other means of guidance, or guidance by mayor program other than in writing. Budget offices retreated somewhat from demanding that agencies report effectiveness and productivity estimates when requesting approval for new programs or revisions in current ones. Effectiveness in budget document narratives also declined, and budget offices reduced their conduct of effectiveness and productivity analysis. The executive in budget decision-making was said to rely less in 1995 on effectiveness and productivity analysis. Budget offices reported that legislatures also relied less on productivity analysis in budget decision-making.
These findings raise the familiar questions: What next? and Why? A change in pattern from one survey to the next does not necessarily signal a reversal in a trend. For instance, the inclusion of effectiveness measures in budget documents dropped between 1985 and 1990 but rebounded between 1990 and 1995 (Table 5). The "why" question cannot be answered with the data from this survey, a fact that suggests the need for some good case studies.
Despite these methodological concerns, the 1995 results raise important questions about the future course of state budgeting practices. An earlier interpretation of the surveys through 1990 was that many states had advanced in the development of their budget systems and that these changes might be considered irreversible (Lee, 1991a). One can safely conclude that budget offices are highly unlikely to revert to professional staffs with less educational background, such as the level of education in 1970. The changes in accounting may also be largely irreversible as accounting standards have risen over the last 25 years.
Are other changes irreversible? An interpretation of the 1970 through 1990 survey data was that some changes may have been irreversible but had reached a plateau. States that had achieved changes were unlikely to reverse them, but perhaps few additional states were likely to adopt these same changes in the near future. With the 1995 data now available, however, a revised interpretation is that some reversal of trends may have occurred. A reversal may be at hand for the once steadily increasing pattern of budget offices providing policy guidance in writing or by other means to agencies as they prepared their budgets. At the same time, the budget offices have continued to use fiscally related ceilings in budget preparation. The data raise the question of whether budget office staff feel they can, or must, rely more on financial than programmatic constraints in soliciting budget requests from agencies.
The survey results are quite difficult to interpret because mixed trends are indicated. Although both budget offices and major line agencies have reduced their conduct of analysis, other central staff organizations and the post auditors have increased their conduct of analysis. This change may simply indicate a realignment of responsibilities for analysis among state agencies in the 1990s, but a competing explanation is that the daily demands upon budget offices to deal with immediate problems drive out longer-range work such as program analysis.
More troubling for anyone interested in having analysis play a prominent role in decision-making are the 1995 results indicating less reliance by executives on analysis. Although all of the states responding in 1990 said the executive used productivity analysis to some degree or to a substantial degree in decision-making, only 84 percent reported that extent of reliance in 1995. The increase in the conduct of analysis by other units of state government has seemingly had no positive influence on executive decision-making. Analyses conducted by these offices may have little influence on the executive, a phenomenon that may bode poorly for the future of such analysis. Why conduct analysis if it has little influence on decision-making?
The survey findings leave unanswered such major questions as who brought about the changes that have occurred and what difference the changes have made. Schick (1971, 193) wrote of a "budgetocracy" that was resistant to change, but Howard (1973, 362) noted that budget staff could be agents for change. One can safely conclude that the changes of the last quarter century have been implemented by budget staffs, and any resistance to change that Schick observed was overcome. The change toward increased reliance on budget personnel with master's degrees in public administration most likely has encouraged greater use of program information and analysis, given the curricula of MPA programs. Contrary to what Howard thought should happen, productivity has tended to be emphasized over effectiveness, but that pattern is most likely a function of the fact that productivity data are both easier to gather and more reliable. Regardless of their academic training, budget staff may tend to favor reliable productivity data over less reliable effectiveness data.
With all of the changes of the last two and a half decades, do states now make better budget decisions? That question is simply unanswerable from the data available. What can be surmised is that budgetary decision-making is different, and presumably, then, the outcomes of the decision process are different. As Wildavsky (1964) wrote more than 30 years ago, "No significant change can be made in the budgetary process without affecting the political process" (132). The relative powers or influences on budget making have been changed for legislators, governors, budget offices, line agencies, and other units such as post-auditor offices. The seemingly extreme rationalism of planning-programming-budgeting of the 1960s has long been abandoned, but as both Howard and Schick noted, rationalism could be increased in budget deliberations without the wholesale abandoning of budget practices, and their replacement with new ones (Howard, 1973, 359; Schick, 1971, 203).
The millennium now approaches. If previous surveys are a good indication of the future, a few general comments can be made. The trends of the past will continue, but some backsliding may occur. An explanation for a reversal of trends is not readily available. One explanation may be that budgets are leaner with budget offices expected to reduce expenses just as line agencies must cut back their operations. Budget offices may feel compelled to reduce their effectiveness and productivity analysis activities.
A forecaster from 1970 might be surprised by the status of state budgeting in 1995. He or she might not have expected computer applications to have advanced as much as they have, or for legislatures to have made as many improvements in budgeting as they have.
On the other hand, a forecaster in 1970 might well have expected state budget offices to be more advanced than they were in 1995 regarding the use of program information and analysis. Indeed, a forecaster from that earlier period is likely to be dismayed by what most likely will be the status of state budgeting in the year 2000. Unless some dramatic unforeseen reform movement materializes in the next few years, the year 2000 will be reached with many states still making limited use of program information and analysis, and with program information being included in budget documents for only some agencies or none at all. The revolution in budgeting that was foreseen in the name of planning-programming-budgeting in the 1960s, and which has its roots in the early 1900s, almost certainly will be incomplete by the year 2000.
Table 9 Budget Office Control over Expenditure of Appropriated Funds, Selected Years, 1970-1995 Percentage Controlling Basis of Control 1970 1975 Appropriations 76 74 Departments 59 60 Organizational units within departments 40 38 Programs, if different from organizational units 19 46 Major objects of expenditure 59 66 Minor objects of expenditure 10 12 Basis of Control 1980 1985 Appropriations 87 90 Departments 75 71 Organizational units within departments 57 48 Programs, if different from organizational units 59 60 Major objects of expenditure 73 63 Minor objects of expenditure 21 10 Basis of Control 1990 1995 Appropriations 91 98 Departments 72 91 Organizational units within departments 50 68 Programs, if different from organizational units 62 65 Major objects of expenditure 72 71 Minor objects of expenditure 20 26
Table 11 Academic Disciplines of Central Budget Office Professional Employees, Selected Years, 1970-1995
Means in Percent Discipline 1970 1975 1980 1985 1990 Accounting 29 20 21 18 15 Business administration 37 28 23 22 24 Public administration 15 19 22 22 21 Economics 3 10 9 9 10 Other social sciences 2 5 9 10 9 General liberal arts 9 8 5 5 7 Other professional disciplines 2 5 5 8 8 Math, science, humanities, and other 3 7 6 6 7 Discipline 1995 Accounting 12 Business administration 26 Public administration 25 Economics 10 Other social sciences 8 General liberal arts 4 Other professional disciplines 8 Math, science, humanities, and other 8
(1.) The first four surveys were conducted by the Department of Public Administration and the last by the School of Hotel, Restaurant, and Recreation Management of the Pennsylvania State University. Some of the early surveys were cosponsored to varying degrees by the National Association of State Budget Officers and the Pennsylvania Budget Office.
Abney, Glenn and Thomas P. Lauth (1987). "Perceptions of the Impact of Governors and Legislatures in the State Appropriations Process." Western Political Quarterly 40 (June): 335-342.
Botner, Stanley B. (1985). "The Use of Budgeting/Management Tools by State Governments." Public Administration Review 45 (September/October): 616620.
Bowsher, Charles A. (1992). Performance Measurement: An Important Tool in Managing for Results. Testimony before the Committee on Governmental Affairs, U.S. Senate. Washington, D.C.: Government Printing Office.
Bunch, Beverly S. and Jeffrey D. Straussman (1993). "State Budgetary Processes: The Two Faces of Theory." Public Budgeting and Financial Management 5(1): 9-36.
Burk, Karen B. (1994). Activity Based Costing and Performance. Fairfax, VA: American Management Systems.
Carter, Karen (1994). The Performance Budget Revised: A Report on State Budget Reform. Washington, D.C.: National Conference of State Legislatures.
Clynch, Edward J. and Thomas P. Lauth, eds. (1991). Governors, Legislatures, and Budgets. New York: Greenwood Press.
Connelly, Michael and Gary L. Tompkins (1989). "Does Performance Matter? A Study of State Budgeting." Policy Studies Review 8 (Winter): 288-299.
Finch, Johnny C. (1995). Managing for Results: Status of the Government Performance and Results Act. Testimony before the Subcommittee on Government Management, Information and Technology, U.S. House of Representatives. Washington, D.C.: Government Printing Office.
Governmental Accounting Standards Board (1994). Concept Statement No.2: Service Efforts and Accomplishments Reporting. Norwalk, CT: Governmental Accounting Standards Board.
Hackbart, Merl M. and James R. Ramsey (1994). "Budgeting Concepts and Processes: Perceived Value for State Budgeting." Public Budgeting and Financial Management 6(1): 28-51.
Harris, Jean, ed. (1995). "Special Issue on Service Efforts and Accomplishments (SEA) Reporting." International Journal of Public Administration 18(2-3): 249-608.
Howard, S. Kenneth (1973). Changing State Budgeting. Lexington, KY: Council of State Governments.
Joyce, Philip G. (1993). "Using Performance Measures for Federal Budgeting: Proposals and Prospects." Public Budgeting & Finance 13 (Winter): 3-17.
Lauth, Thomas P., ed. (1993). "Theory, Research, and Practice in State Budgeting." Public Budgeting and Financial Management 5(1): 1-130.
Lee, Robert D., Jr. (1991a). "Developments in State Budgeting: Trends of Two Decades." Public Administration Review 51 (May/June): 254-262.
-- (1991b). "Educational Characteristics of Budget Office Personnel and State Budgetary Processes." Public Budgeting & Finance 11 (Fall): 69-79.
-- (1992). "The Use of Executive Guidance in State Budget Preparation." Public Budgeting & Finance 12 (Fall): 19-31
Lee, Robert D., Jr. and Ronald W. Johnson (1994). Public Budgeting Systems. 5th ed. Gaithersburg, MD: Aspen Publishers.
Lee, Robert D., Jr. and Raymond J. Staffeldt (1976). "Educational Characteristics of State Budget Office Personnel." Public Administration Review 36 (July/August): 424-428.
-- (1977). "Executive and Legislative Use of Policy Analysis in the State Budgetary Process: Survey Results." Policy Analysis 3 (Summer): 394-405.
National Performance Review (1994). Putting Customers First: Standards for Serving the American People. Washington, D.C.: Government Printing Office.
Norvelle, Joan W. (1994). Introduction to Fund Accounting. 5th ed. Tucson, AZ: Thoth Books.
Osborne, David and Ted Gaebler (1992). Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector. Reading, MA: Addison-Wesley.
Posavac, Emil J. and Raymond G. Carey (1992). Program Evaluation: Methods and Case Studies. 4th ed. Englewood Cliffs, NJ: Prentice-Hall.
Rossi, Peter H. and Howard E. Freeman (1993). Evaluation: A Systematic Approach 5th ed. Newbury Park, CA: Sage.
Rubin, Irene S. (1992). "Budget Reform and Political Reform: Conclusions from Six Cities." Public Administration Review 52 (September/October): 454-466.
Schick, Allen (1971). Budget Innovation in the Stases Washington, DC: Brookings Institution.
Stanford, Karen A. (1992). "State Budget Deliberations: Do Legislators Have a Strategy?" Public Administration Review 52 (January/February): 16-26.
Stevens, John M. and Robert D. Lee, Jr. (1981). "Patterns of Policy Analysis Use for State Governments: A Contingency and Demand Perspective." Public Administration Review 46 (November/December): 636-644.
Stevens, John M., et al. (1991). "The Utilization of Information Systems Technology in State Financial Management: An Empirical Assessment." Journal of Management Information Systems 8 (Summer): 107-128.
Sylvia, Ronald D., Kenneth J. Meier, and Elizabeth M. Gunn (1991). Program Planning and Evaluation for the Public Manager. Prospect Heights, IL: Waveland Press.
U.S. Congressional Budget Office (1993). Using Performance Measures in the Federal Budget Process. Washington, D.C.: Government Printing Office.
U.S. General Accounting Office (1990). Cost Accounting Issues: Survey of Cost Accounting Practices at Selected Agencies. Washington, D.C.: Government Printing Office.
-- (1992). Program Performance Measures: Federal Agency Collection and Use of Performance Data. Washington, D.C.: Government Printing Office.
-- (1993). Performance Budgeting: State Experiences and Implications for the Federal Government. Washington, D.C.: Government Printing Office.
-- (1994a). Executive Guide: Improving MIssion Performance Through Strategic Information Management and Technology. Washington, D.C.: Government Printing Office.
-- (1994b). Government Auditing Standards. Washington, D.C.: Government Printing Office.
-- (1994c). Management Reform: Implementation of the National Performance Review's Recommendations. Washington, D.C.: Government Printing Office.
-- (1994d). Managing for Results: State Experiences Provide Insights for Federal Management Reforms. Washington, D.C.: Government Printing Office.
-- (1995a). Managing for Results: The Department of Justice's Initial Efforts to Implement GPRA. Washington, D.C.: Government Printing Office.
-- (1995b). Program Evaluation: Improving the Flow of Information to the Congress. Washington, D.C.: Government Printing Office.
Walton, Karen S. and Richard E. Brown (1990). "State Legislators and State Auditors: Is There an Inherent Role Conflict?" Public Budgeting & finance 10 (Spring): 3-12,
Wildavsky, A. (1964) . The Politics of the Budgetary Process. Boston: Little, Brown.
Yunker, Jon (1990). "Managing a Budget Office." Public Budgeting & Finance 10 (Summer): 96-101.
Robert D. Lee, Jr., is a professor of public administration and professor of hotel, restaurant, and recreation management at the Pennsylvania State University, University Park. He is coauthor of Public Budgeting Systems, 5th ed. (Aspen Publishers, 1994) and author of Public Personnel Systems, 3rd ed. (Aspen Publishers, 1993). He is chair of the Association for Budgeting and Financial Management, a section of ASPA.
|Printer friendly Cite/link Email Feedback|
|Author:||Lee, Robert D., Jr.|
|Publication:||Public Administration Review|
|Date:||Mar 1, 1997|
|Previous Article:||Public deliberation: an alternative approach to crafting policy and setting direction.|
|Next Article:||The pitfalls of political decentralization and proposals for reform: the case of New York City public schools.|