Modeling Federal Implementation as a System: The Clean Air Case

Several studies have examined implementation determinants for parts of a U.S. federal policy structure, either national or subnational. However, no study treats federal implementation comprehensively as a single system by examining both national and subnational outputs in mutual interdependence. This article reports an analysis of Clean Air Act enforcements by EPA and the 50 states, separately and in relative dependence with one another. Granger causal and pooled time series regression analysis explain the vigor of federal activity from 1977 to 1985. The results show that federal implementation is not easily understood by examining parts of a federal structure; rather, more insight comes by examining implementation as a system. For clean air, national and subnational outputs were inertial, interdependent, and linked to both vertical and horizontal influences. The implication is that federal structures are adaptive entities that provide greater democratic representation than other administrative forms.