Quantitative Impact Analysis of the Centralization of Firms in the Tokyo Metropolitan Area Considering Firm-to-Firm Trade Networks

When a firm makes a location decision, it considers only its own transportation costs and ignores the transportation costs of its trading partners, resulting in inefficient sparce locations of firms. Since Beckmann (1976), it has been known that such inefficient sparse locations occur in the canonical land use models with interactions between agents, and this externality is referred to as locational externality by Kanamoto (1990). We quantitatively analyze the scale of locational externalities using micro data of the listed firms located in the Tokyo metropolitan area and firm-to-firm trade network data. We show (1) which trade patterns involve locational externalities, (2) the ratio of trade generating locational externalities as a percentage of total trade is about 24%, (3) the transfer of a randomly-chosen 5% of firms to two business centers, Marunouchi and Shibuya, generates median external benefits of 1.9% and 1.3% in the total industry in terms of value-added, respectively, (4) benefits vary according to industry and location (e.g., about 10% in the case of firms located far from the centers, and about 5% in the case of firms in the information and communications industry).