To test the hypothesis that court orders adversely affect government finances, three California counties (Orange, Santa Clara, and Contra Costa) were studied using time series data, case histories, and interviews with government and corrections officials. In addition, cross- sectional data from two constructed samples of matched counties compared expenditures in counties under and not under court orders. Findings suggested that judges were sometimes persistent in seeking reform, even ordering direct expenditures on jails. Statistical investigation, however, revealed that the effects of judicial intervention were relatively strong in some cases but weak in others. Counties under court orders tended to have larger increases in correctional expenditures relative to matched controls. Nonetheless, in some cases, court-ordered counties actually showed lower expenditures than their matched controls. In Contra Costa County, proactive use of intermediate sanctions likely prevented more intrusive judicial intervention and helped keep jail expenditures down. In Orange County, no direct orders to expand jail capacity were issued, but officials perceived that capacity was a primary cause of and solution to judicial intervention. In Santa Clara County, explicit orders by two judges to allocate expenditures, increase jail capacity, and hire staff contributed to substantial increases in expenditures. It was determined that the effects of court orders on correctional policies are heterogeneous and that judicial "power of the purse" is limited by various legal and pragmatic constraints. Appendixes contain supplemental data on study procedures. 62 references, 6 notes, 4 tables, and 3 figures