The Economic Cost of A Nuclear Weapon: A Synthetic Control Approach

Abstract

ABSTRACT This study estimates the economic effects of nuclear weapons development efforts in Pakistan using synthetic control group methods. Pakistan started its nuclear weapons program in 1972 and conducted its first test in 1998. This paper focuses on the growth impacts during the 1973 to 1997 period, before Pakistan established itself as a nuclear power. I create a synthetic control group for Pakistan using Per capita Gross Domestic Product (GDP) from the 1950 to 1972. The impact of the nuclear weapons development program is measured as a treatment dummy for the years 1973–1997 in a Difference-in-Difference model. I find that Pakistan’s per capita GDP would have been an average of about $718 per year higher had the country not undertaken the effort to produce a nuclear weapon. This equates to per capita GDP being 27.8% lower on average over the 25-year weapons-development period. Results are robust to several alternative specifications, including country exclusion, sparse synthetic controls, non-outcome characteristics as predictors of GDP, and in-space placebo experiments of differing specifications.