In this study, I have argued that by simultaneously monitoring
program processes and documenting management
and program changes in diverse settings and across multiple
levels of government, federal managers might use the
performance data more effectively to evaluate the effects
of different policies and approaches to managing and delivering
government services. Empirically, this might be
accomplished in the same way that federal job-training
programs already use statistical methodologies (such as
multiple regression) to control for local population characteristics
and economic conditions in evaluating performance
by adding controls in their models for state and local
program structures, policies, and management practices.