Enterprise Hall, 318
March 12, 2009, 08:00 PM to 07:00 PM
Improved technology and increased transparency and accountability have made the shortcomings of federal government processes and results more apparent, but they have not helped diagnose or solve the problems commonly attributed to bureaucracy. I submit that this may be due to outdated “machine” models of government that neglect implementation challenges. What if the problem is that we have too many controls? Analyzing the federal government as a complex system leads to new questions and insights that have profound implications for theory and practice of public choice.
Using relationships as a unit of analysis, I construct a graphic model of a hypothetical agency, which I call a Kaleidic Hyperstructure, to demonstrate how rule changes affect behavior of the system. A case study of formation and reorganization of the Department of Homeland Security illustrates how this method of analysis can inform institutional design. I conclude that real control and coordination (i.e., effective policy implementation) will only occur if accountability for results is accompanied by flexibility to allocate resources – a relaxation of some constraints. Trading control for results works because agents value choice that allows them to use dispersed and emergent knowledge. Enabling (and requiring) agencies to pay more than lip service to policy can benefit the public not just by more effective administration but by providing feedback on how policies impact program performance and by inducing innovation and adaptation in government.