STATE BUDGET EXECUTION: ADMINISTRATIVE DISCRETION AND FISCAL CONTROL (AUTONOMY, GOVERNMENT, FLEXIBILITY)

This dissertation examines the conflicting values of fiscal control and administrative discretion in budget execution in the American states. Based upon a mail survey of state government central budget officers and agency administrators, the perceptions of those surveyed are analyzed. The existing control techniques as well as controls that respondents perceive can best serve the needs of their agencies are identified. The frequency and magnitude of budgetary change in state agencies are reviewed for two reasons. First, to determine if a relationship exists between the frequency of changes during budget execution and the magnitude of those changes. Second, to analyze the impact of controls on the frequency and magnitude of budgetary changes.; It is confirmed that the majority of agency administrators prefer substantial levels of discretion whereas the majority of central budget officers prefer high levels of control. It is also observed that agency administrators and central budget officers are in general agreement with regard to the kinds of control techniques and discretionary practices that they perceive to be necessary for performing their budgetary responsibilities.; On the basis of this research, it has been possible to identify at least three categories of controls. Controls which appear to be associated with low levels of change during budget execution. These controls are classified as effective in that they have the expected effect on budget change. Certain other controls seem to be inversely associated with levels of change during budget execution. Where such controls are present, budget change during execution is both more frequent and more substantial than when the control did not exist. These controls are classified as ineffective in that they have unexpected and somewhat counterintuitive effect on budget change. The presence of such controls may encourage budgeters to circumvent them. Finally, there are controls which appear not to affect either positively or negatively the frequency or the magnitude of budgetary change. These controls are classified as neutral controls.; It is also found that there is a substantial difference between existing techniques and practices and those which budgeters perceive to be necessary in the budget execution phase. Most controls that budgeters prefer are absent and most of those which they disfavor exist. In addition, most of the existing and disfavored controls appear to be ineffective or neutral.