be the most important factors affecting organizational performance, although their analysis is limited by the fact that organizational performance is not directly measured but constructed from employees’ survey responses.
Following the approach of Lynn, Heinrich and Hill (2000), other studies employ multilevel modeling methods to investigate the influence of these variables on organizational outcomes. Roderick, Jacob and Bryk (2000), Heinrich and Lynn (2000), and Bloom, Hill and Riccio (2001) operationalize structural, management and primary work components to empirically investigate their relationships to school performance (students’ test scores), job-training program outcomes (trainees’ earnings), and welfare-to-work program outcomes (individual earnings and welfare participation), respectively. Roderick, Jacob and Bryk, for example, build a model showing how environmental, school, teacher and student characteristics interact to influence the implementation of educational processes (instruction, school learning environment, parent/school relations and other aspects of school management) and ultimately, student achievement. The model tested by Heinrich and Lynn includes measures of administrative arrangements, centralization of authority in program management, service technologies and contracts for service delivery, management strategies such as performance requirements, and controls for individual and local labor market characteristics. The Bloom, Hill and Riccio model includes operationalizations of service technology, organizational climate (based on staff perceptions as measured in a survey), and managerial processes (such as commitment to program mission, staff discretion and recognition, etc.), and measures of environmental and individual client characteristics. Empirical results confirmed the importance of these structural, management and policy/technology factors on organizational outcomes.
In Table 1, six common model components among public management theories of