Change, Persistence and Convergence in NATO Member Defense Spending

Abstract

This study presents an analysis and quantitative summary of 18 NATO member country defense spending over the period 1953-2020. Using recently developed econometric techniques, we explore the time series properties of change, persistence and convergence in two indicators of NATO defense spending data typically used in the literature: real (2019) US dollars and Percent of GDP. Our two indicator variables display a mix of positive, negative and zero trends over the sample period. The only NATO countries with more than 2% defense spending after the 2006 and 2014 Summits are: Greece, Turkey, UK, USA and Poland. Using the fractional difference-based persistence tests of Martins and Rodrigues, we find only UK, Hungary and Poland dollar Defense Spending reject the null of a constant fractional difference (unit root) for the entire sample period; while seven NATO members reject the null of a constant fractional difference (unit root) for Percent of GDP. The mixed set of positive, negative and zero trends render the popular relative convergence test of Phillips and Sul inappropriate for our data. Using the more appropriate weak sigma-convergence test of Kong, Phillips and Sul, we find mixed evidence for convergence of our indicator variables. Our quantitative results present a mixed picture of statistical consistency and coherence for NATO defense spending. Our tests of persistence suggest major changes in the defense policies and spending of NATO members will have a lasting effect in most cases. The emergence and growth of serious threats and potential threats from Russia (and now potentially China as well) will require the NATO allies to address these issues.