Development and Validation of Firm Demography Model - Case Study Analysis in Phoenix and Tucson Megaregion
Corresponding Author: Srinath Ravulaparthy, Cambridge Systematics, Inc.
Presented By: Arun Kuppam, Cambridge Systematics
Understanding and forecasting freight flows is critical to determining the need for future transportation capacity on the regional highways or other modal infrastructure. Maricopa Association of Governments (MAG) has identified a need for a new generation regional freight model that involves firm synthesis, supply chain formation and tour-based models for urban truck travel.
Integrated land-use and transportation models address regional issues related to congestion, mobility and economic competitiveness. Essential to this is freight transport, in the form of economic vitality along with distribution and complex interactions between various firms. Within this context, it is critical to capture and accurately describe the behavioral dynamics of these firms in the region including modeling firm demographic events of birth (formation), death (dissolution), growth (expansion), decline (contraction), and migration (relocation) patterns.
Firm demography is a phenomenon that captures the evolution of firms in time and space in a given region. Firm evolution model captures firm births, deaths, migration, and growth/decline over a period of time. This presentation will describe the methodology for modeling firm demographics using a microsimulation approach for the Phoenix-Tucson megaregion. The model is based on principles of firmographics that predicts the location, magnitude and size of firms in the study region. This model uses the National Establishment Time Series (NETS) database as a seed table. A series of econometric models are estimated to simulate the firm events that consider determinants such as firm internal attributes (size, age and growth) and external attributes (market area characteristics, transportation costs, agglomeration economies).
The simulated results are validated with observed firm demographic trends along with zonal-level employment estimated using various goodness-of-fit measures. For instance, using a 20% validation sample, the firm synthesizer accurately projects firm events of birth, death and relocation from 2007 to 2012. In addition, total zonal employment predicted in the region also closely resembles the observed trend, thereby also capturing the observed spatial distribution of employment in the megaregion.