The interaction of policy and management presents a close relationship in administrative practices. This dissertation estimates the impacts on use of local performance measures related to local service contracting. The previous research only tested individual state level or local level, rather than estimated how state factors influence local practices. For example, the hierarchical rules, mandates, and laws made by state levels might affect the adoption rates of local management innovation. As well, the previous efforts only used cross-sectional data to understand the adoption of local management innovation, which may overlook the changes over time that take place in local management performance measurement innovation. In this dissertation, the propensity to "implement" the adoption of a new management instrument has been studied under the rubric of management innovation, specifically as adopting a variety of performance measurement in local service delivery contracting. To analyze the "diffusion" characteristic of space and time, the model strategies employed in this dissertation include the local level and state level with a time growth curve estimated by Hierarchical Linear Model (HLM) and Hierarchical Generalized Linear Model (HGLM). To enhance understanding of the adoption and the process of management innovation, the study employs panel data drawn from the survey of 1992, 1997, and 2002 International City and County Management Association (ICMA) city administrators' responses to local service delivery programs on municipal and county governments nested in state levels to test the adoption of local performance measures related to service contracting. This dissertation intends to estimate the relationship about how contract management capacity and state factors influence the rate of use of performance measurement, and to compare two group differences of private contracting, and nonprofit contracting employed by local governments nested in state levels. Beyond our expectations, the diffusion of local performance measurement nested in state level did not significantly change over time, but the findings with the growth curve models showed that the adoption of performance measurement indeed had grown over our observed time. In general, the local contract management capacity (e.g., feasibility assessment, evaluation, and implementation) significantly matters to the adoption of three types of performance measurements (e.g., citizen satisfaction, cost, contract compliance). In individual, the contract management capacity and state factors (e.g., state divided government, state reinventions) significantly impacts the use of performance measurement for private contracting. However, the state factors weakly influence the adoption of performance measurements when local governments employed nonprofit contracting. Several factors--structural and institutional heterogeneity-- can explain these differences in use of performance measures related to private contracting, and nonprofit contracting. The empirical findings also show that state factors have a large impact on the adoption of performance measurement related to service delivery contracting. The findings suggest that local practices are indeed embedded in multilevel diffusion that may be unobserved by the previous studies in public administration research.
