Demand-pull versus cost-push: monocentric equilibrium in a spatial network

The present model combines insights from two different approaches in the literature. The first one focuses on scale economies and aggregate demand externality. The second approach is without scale economies, and the focus is on the accessibility of markets where production is location-specific. We assume the existence of multiple regions connected by a pre-existing network. In our framework, the manufacturing sector is mobile, and production can occur under both scale economies and without. On the other hand, agriculture is location-specific, and each area produces a particular variety. Only manufacturing goods incur transportation costs in our first model (demand-pool). We illustrate that agglomeration can only proceed in the presence of scale economies. Scale economies require a large pool of workers; there is greater demand in the metropolitan area. Since there is no cost of trading within the metro area, higher demand leads to agglomeration. When aggregate demand externality is a factor, geographical and history become pertinent in selecting an urban manufacturing center. In the second model (cost-push), the transportation cost for agriculture is positive, while manufacturing goods can be transported without costs across regions. Hence, aggregate demand externality is no longer present. Since agriculture is heterogeneous and consumers prefer to consume all varieties, locating in regions with easier access to rural hinterlands is more profitable. Consequently, geographical factors alone determine the location choice.