ABSTRACT
This chapter presents the explanation of substantively weighted analytical techniques (SWAT) using a dataset of 534 school districts in Texas and deals with both the successes and the failures of public management. It presents the differences among risk-averse bureaucracies, failing bureaucracies, and performance-optimizing bureaucracies. The research literature on bureaucracy generally assumes that bureaucrats are risk averse, and some work goes further to characterize bureaucrats as risk avoiding and conservative. The chapter presents the analyses with generalized substantively reweighted least squares (GSRLS) and shows that more than one group of interest can be isolated with SWAT analysis. It also shows that such analysis may reveal interesting differences in how these groups use the resources at their disposal. Since the GSRLS models are in effect controlling for a specified phenomenon, residual behavior, the analogy is appropriate. The chapter looks at four general classes of educational bureaucracies: typical, superior, poor, and trepidant.
