Crossing the capability threshold: ecosystem capacity and technology-driven growth
Description
This repository contains the reproduction package for the paper "Crossing the capability threshold: ecosystem capacity and technology-driven growth" (S. Kim and Y.-W. Sawng). Research hypothesis. The growth payoff to technology is conditional on a country's accumulated innovation-ecosystem capability. We hypothesize that the transmission from total factor productivity (TFP) growth to manufacturing value-added growth is gated by prior, accumulated capability along three complementary dimensions - business R&D intensity, institutional quality, and human capital - so that the same productivity gain raises manufacturing growth more strongly in higher-capability economies (a capability-threshold mechanism). Data. One country-year panel in long format covering 24 OECD countries, 2002-2019 (432 rows; 374 used after listwise deletion). Variables: manufacturing value-added growth (g_m, OECD STAN); TFP growth (Conference Board Total Economy Database); the lagged change in a patent-based technology index and the lagged change in an export-sophistication index (controls); and three capability measures entered as grand-mean-centered country period means - business R&D share (OECD MSTI), institutional quality (first principal component of the six World Bank Worldwide Governance Indicators), and researchers per million (OECD MSTI and UNESCO UIS). Moderators, lags, interactions, and capability terciles are constructed within the script, so the construction is fully transparent. Notable findings. With country and year fixed effects and country-clustered standard errors (N = 374), TFP growth is the dominant driver of manufacturing growth (baseline coefficient 1.33, p < 0.01). Its interaction with each capability dimension is positive and significant (TFP x business R&D 2.85; x institutional quality 0.14; x human capital 0.0002). Sorting the 24 countries into capability terciles, the implied TFP-to-growth coefficient roughly doubles from the low to the high tercile (about 1.1-1.2 to 2.0-2.2); the high-tercile amplification is strongest and most significant for human capital. A specificity check shows that replacing TFP growth with the lagged technology change in the same interactions yields no significant moderation (all p > 0.38), indicating that capability gates the realized-productivity path rather than raw technological output. Interpretation and use. Results should be read as conditional (moderated) associations under a two-way fixed-effects design, not as causal parameters, and as evidence on the direction and pattern of the threshold rather than exact magnitudes. Running P3_reproduce.R in R (4.0 or later) with the fixest and dplyr packages reproduces every number reported in the paper; expected outputs are annotated in the script and listed in README. Reported p-values come from fixest; other software may apply slightly different small-sample corrections to clustered standard errors.
Files
Steps to reproduce
Install R (version 4.0 or later) and the packages "fixest" and "dplyr" (install.packages(c("fixest","dplyr"))). Place P3_reproduce.R and main_panel_P3.csv in the same working directory. Run the script: source("P3_reproduce.R"). It reads the CSV, constructs the moderators (country period means, grand-mean centered), lags, interactions, and capability terciles internally.
Institutions
- Konkuk UniversitySeoul, Seoul