Agglomeration Externalities, Spatial Policies and Welfare

This project draws on rich Indonesian data to study the externalities and impacts of agglomeration in order for relevant insight into spatially targeted policies for manufacturing firms in Low-Income Countries (LICs). 

Most manufacturing firms in Low-Income Countries (LICs) are characterized by small scale and low growth, and are critically constrained by restricted access to capital and technology, as well as an unfavourable investment climate. In order to tackle this issue and to promote industrial growth, many LICs are implementing spatially-targeted policies. These policies are designed with the hope that clustering firms can result in agglomeration economies. This project sets out to investigate aspects of agglomeration externalities, such as existence, magnitude, dynamics, and impacts on firm productivity. The researchers also investigate whether government subsidies aimed at encouraging clustering, are effective in their goals, and whether there is an identifiable optimal subsidy.

By using an annual survey combined with several other data sources, the researchers will estimate firm-level production functions to examine the magnitude and scope of agglomeration spillovers, allowing them to postulate a model for the relationship between individual firm output and the output of other firms. They will also estimate the impact of a spatial targeting policy on firm entry, output, productivity, and wages. Beyond this, they will build and estimate a structured general equilibrium model in which firms and workers choose locations based on government policies and spatial features.

This project will yield valuable data, which will help with very relevant analysis of LICs with economic zone policies, such as Bangladesh. The industry level analysis will help understand how Bangladesh’s key industry of textiles and garments would be affected by spatial targeting and the model developed can be calibrated to Bangladeshi data to examine the impacts of local policies.