Agent-based modeling: A promising way for modeling entrepreneurial activity from location perspective

The location of new establishments is a major issue from the perspective of economic activity and the analysis of the territory. Although both its determinants and consequences are multiple and complex, most studies are focused on single factors or aggregate approaches. This paper presents an agent-based model (ABM) with the main objective to analyze, with high spatial accuracy, the effects of public policies on the behaviour and performance of firms. The starting point is the building, trough statistical matching techniques making use of microdata sources, of a database that replicates the attributes and location of all individuals and companies that are located in a specific spatial context. Then, individual behaviours are established for both companies and human beings who are interacting according to their preferences and endowments. Individuals are an essential part of the model as they operate as consumers, workers and entrepreneurs, conditioning the carrying out of firms, their entries and closures. In addition to these agents we include a raster of locations, built through with downscaling techniques, that includes for each cell the number of individuals, establishments or land use as well as other key variables such as distance to transport networks, land price or GDP generated. Patterns of mobility and consumption are identified with the matching of several sources of data and the building of relationships between agents. Furthermore, the current situation of different policies is displayed, in order to measure properly the changes later introduced to do simulations. Then we are able to forecast the growth of agglomeration economies and simulate the impact of changes in public policies on SMEs location and performance. A new contribution of this work is the inclusion of a complete modeling of entrepreneurial activity, from the generation of the idea to the decision of the location and the final implementation of the project. That is, it is possible a more realistic approach to the behaviour of Small and Medium Enterprises (SMEs) in which location decisions is related more to personal conditions of the employer that to business efficiency criteria. The model is implemented in MASON and the results are presented in stochastic terms thus enabling a better assessment of both policy outcomes and their risks. JEL: L25, L26, R12, R58 Keywords: Agent-based model, entrepreneurship, business location