The empirical study of new economic geography

• Main Author: Toh, Ci En
• Other Authors: Fedor Duzhin
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Science::Mathematics; Social sciences::Economic theory
• Online Access: https://hdl.handle.net/10356/153388
• Institution: Nanyang Technological University

Many empirical studies had proposed scientific methods to apply New Economic Geography (NEG) theory and models to explain uneven spatial distributions of population and economic activities in real-world economy. This implies that NEG in general, can be a scientific theory that makes verifiable predictions. Although Krugman(1999)'s NEG theory in [13] is considered the most fundamental form of NEG, most empirical studies only used parts of his proposed NEG models and little has been done specifically on any of his full model in [13]. This paper aims to specifically use Krugman's full monocentric economy model in [13] to prove that Krugman's NEG theory is scientific and hence, emphasise that NEG theory in general is scientific. We used empirical data of New York Metropolitan area from years 1990 to 2005 and Least Squares method and Particle Swarm Optimization to estimate the parameters for the market potential function in the full model. The model predicted that agglomeration equilibrium will always be sustainable in New York county, the chosen city in this paper. While the prediction might not fully align with the real-world state of agglomeration, the prediction can be refuted by logically observable events such as a possible shift of the central business district away from the New York county or decrease in population in New York county in the future. The model can also be tested against metropolitan areas where we know clearly when and where a new city has been formed. In this way, we had proven that Krugman's NEG theory is falsifiable and hence, scientific since its model is able to make verifiable predictions. Thus, underscored that NEG theory in general can be scientific.