Technological Innovation and Agricultural Productivity Under CPEC: A Local Projection Analysis of Patent Activity in Pakistan
Description
The study uses annual data for Pakistan from 1991–2021, based on the first year with agricultural‑employment data and the last year with complete patent records. The sample includes 31 observations: 24 pre‑treatment years (1991–2014) and 7 post‑treatment years (2015–2021). All variables come from well‑known international data sources. Agricultural value added, GDP, rural population, agricultural employment, and patent applications are taken from the World Bank’s WDI. Temperature and precipitation data come from NASA POWER, and fertilizer use (nitrogen, phosphate, potash) is taken from FAO. Agricultural value added (% of GDP) is calculated by dividing agricultural value added by GDP and multiplying by 100. The CPEC‑2015 treatment variable equals 0 for 1991–2014 and 1 for 2015–2021. Because FDI inflows increased sharply after 2015 and move closely with this indicator, FDI is treated as part of the treatment effect. Technological innovation is measured using total resident patent applications. Agricultural‑specific patent data are unavailable, so overall patent activity is used as a proxy for innovation relevant to agriculture. One missing value for 2009 is filled using linear interpolation.
Files
Steps to reproduce
Stata version 17 was used for all data processing and interpolation, and the single missing year of patent data for 2009—covering both resident and non‑resident applications—was completed using linear interpolation.