Dynamics of Intergovernmental Networks
Figure 8b . Simulated success rates of roadway projects across local towns under higher number of new projects and decreased funding availability .
6 . Limitations of the Current ABM and Next Steps
All models are abstractions of reality .
It is the function and purpose of designing computer simulation models that broadly define the boundaries of the dynamic policy and governance system that is abstracted out of the complex reality . The main purpose of designing the ABM presented in this study is to simulate the decision-making dynamics of project prioritization processes among multilevel government agencies . The baseline institutional designs and rules that govern the interactions of intergovernmental governance dynamics are the focus of this study , yet the system boundaries of the ABM presented in this study led us to simplify and in some cases even ignore the institutional rules that might have direct bearing on the model outcomes presented in this study . The geophysical boundaries of ABM are limited to Vermont . We intentionally excluded five other project classes and just focused on roadways . While this simplification allowed us to focus on modeling the institutional rules across different levels of government , the competing dynamics that occur in terms of allocation of funding across different
project classes ( i . e ., between bridges and roadway and / or bike / pedestrian and park and rides ) have been ignored in setting up the system boundaries .
There are many possible ways , both vertically and horizontally , to extend and generalize the ABM presented in this study . Within the current transport policy domain , the ABM could be made spatially explicit by adding Geographic Information System ( GIS ) layers . Furthermore , the ABM could be extended to all six transport project classes . Explicit rules of federal programs , such as Surface Transportation Program ( STP ), Interstate Maintenance ( IM ) and others , could be incorporated in an upward expansion of the model . Similarly , in a downward expansion of the model , the complex network dynamics that occur in local towns and their planning commissions / boards could be explicitly captured using social network analysis techniques ( Wasserman & Faust , 1994 ). Furthermore , even at a finer grain level , the outcomes of the funded projects could be ascertained by coupling integrated land-use transportation models such as UrbanSim ( Waddell , 2002 ), with the policy implementation network simulation model presented in this study . Finally , as well known in the policy implementation literature since Pressman
68