Dynamics of Intergovernmental Networks
1 . Introduction
This article develops a patternoriented , agent-based model ( ABM ) to simulate alternative institutional rule scenarios in order to assess the alternate outcomes of transportation funding . The ABM simulates an intergovernmental network consisting of federal resources , state decision makers , and regional and local representatives . The ABM is designed to generate experimental simulations for addressing the following two specific questions , which broadly pertain to the design and governance of intergovernmental transportation policy implementation networks and their predicted impacts on policy outcomes :
1 . Federalism versus Regionalism versus Localism Debate : How the allocation and distribution of federal and state funding resources for transportation infrastructure development projects changes under different configurations of weighting state versus regional versus local government priorities ?
2 . Resilience Assessment : How do different shocks to the intergovernmental policy system , such as increased frequency and intensity of extreme weather events and / or federally mandated funding sequestrations , influence the distribution of financial resources across regional and local governments ?
This article demonstrates that a “ pattern-oriented ,” ABM approach can provide a viable empirical and computational methodology for effectively designing the functions and resource allocation decisionmaking processes of intergovernmental policy implementation networks . Through such policy informatics platforms , the emergent and self-organizing phenomena observed in policy and governance systems , as well as lags and inertia in designing and implementing programs and projects , can be more accurately modeled and simulated . The development of such policy informatics platforms requires “ patternoriented ” computational modeling of the system that “ generates ” observed patterns of resource flows , resource allocations , and resource distributions under an observed configuration of intergovernmental networks authorized to implement specific public policies ( see Epstein ( 2006 ) for generative social science applications ).
The capacity of computer models of complex governance networks to lead to accurate forecasting and prediction of particular policy outcomes is predicated on a “ deep uncertainty ” that characterizes our current state of understanding of complex social systems and associated wicked planning problems ( Rittel & Webber , 1973 , 1984 ). Bankes ( 2002 , p . 7263 ) characterizes this deep uncertainty arising as , “ the result of pragmatic limitations in our ability to use the presentational formalisms of statistical decision theory to express all that we know about complex adaptive systems and their associated policy problems .”
To cope with the inherent complexity and uncertainty in the social complexity of governance networks , we undertake a variation of “ pattern-oriented ” modeling . Pattern-oriented models are described by Grimm et al . ( 2005 ) as “ bottom-up ” models that emphasize the applicability of models to real problem solving .
Grimm et al . ( 2005 ) describe pattern-oriented models this way :
52