TY - JOUR
T1 - An information-theoretic approach to the analysis of location and colocation patterns
AU - van Dam, Alje
AU - Gomez-Lievano, Andres
AU - Neffke, Frank
AU - Frenken, Koen
N1 - Funding Information: A.D. and K.F. are funded by the Netherlands Organisation for Scientific Research (NWO) under the Vici scheme, number 453‐14‐014. F.N. received financial support from the Austrian Research Agency (FFG), project #873927 (ESSENCSE). Publisher Copyright: © 2022 The Authors. Journal of Regional Science published by Wiley Periodicals LLC.
PY - 2023/1
Y1 - 2023/1
N2 - The study of location and colocation of economic activities lies at the heart of economic geography and related disciplines, but the indices used to quantify these patterns are often defined ad hoc and lack a clear statistical foundation. We propose a statistical framework to quantify location and colocation associations of economic activities using information-theoretic measures. We relate the resulting measures to existing measures of revealed comparative advantage, localization, specialization, and coagglomeration and show how different measures derive from the same general framework. To support the use of these measures in hypothesis testing and statistical inference, we develop a Bayesian estimation approach to provide measures of uncertainty and statistical significance of the estimated quantities. We illustrate this framework in an application to an analysis of location and colocation patterns of occupations in US cities.
AB - The study of location and colocation of economic activities lies at the heart of economic geography and related disciplines, but the indices used to quantify these patterns are often defined ad hoc and lack a clear statistical foundation. We propose a statistical framework to quantify location and colocation associations of economic activities using information-theoretic measures. We relate the resulting measures to existing measures of revealed comparative advantage, localization, specialization, and coagglomeration and show how different measures derive from the same general framework. To support the use of these measures in hypothesis testing and statistical inference, we develop a Bayesian estimation approach to provide measures of uncertainty and statistical significance of the estimated quantities. We illustrate this framework in an application to an analysis of location and colocation patterns of occupations in US cities.
KW - coagglomeration
KW - location quotient
KW - pointwise mutual information
KW - relatedness
KW - revealed comparative advantage
UR - http://www.scopus.com/inward/record.url?scp=85138997610&partnerID=8YFLogxK
U2 - https://doi.org/10.1111/jors.12621
DO - https://doi.org/10.1111/jors.12621
M3 - Article
SN - 0022-4146
VL - 63
SP - 173
EP - 213
JO - Journal of Regional Science
JF - Journal of Regional Science
IS - 1
ER -