Replication package for "Industrial Parks, Local Employment, and Land Values: Evidence from Korea"

Replication package for "Industrial Parks, Local Employment, and Land Values: Evidence from Korea" (Hyeongjin Kim), submitted to the Journal of Public Economics in October 2026. The paper links the universe of registered manufacturing plants in Korea for 2006–2024 to municipal electricity use and to digitized boundaries, assessed land values, and project costs for 577 industrial parks completed in that period, and estimates the effects of park completion with staggered difference-in-differences, boundary, and distance designs. The package contains one Stata do-file per table and figure of the manuscript and appendix (do/table_.do and do/figure_.do, each runs on its own in Stata 17 or later with the community packages listed in the README; all of them take about 30 minutes) and the 13 Stata analysis data sets they read: the municipal panel, the park master file, park-by-band-by-year cells of plant density, parcel-year assessed land prices around park boundaries, stand-alone plant-years with distance bands, the composition of park employment, the classification of park entrants, and the aggregated wage panel. The raw administrative files (factory registry, park master and boundary map, cadastral map, assessed land prices, electricity statistics, regional accounts, and labor force survey microdata) are obtained from the issuing agencies under their terms of use and are not redistributed; README.md lists each source and access route and documents every data set. Version 2.1 (29 October 2026) renumbers the appendix tables and figures in the order in which the text first cites them, adds five do-files for the appendix tables moved out of the main text (Tables A.7, A.12, A.14, A.17, and A.18, each assembled from the output of the do-file of the corresponding main-text table), and documents in the README where the manuscript prints a subset or a rearrangement of a tex file; no estimate changed. The package holds 40 do-files and 14 Stata data sets.