Military expenditures, economic growth and spatial spillovers

Abstract

The relationship between economic growth and military expenditure has been the subject of a large literature in defence economics. This study analyses the influence of military expenditures on economic growth in a global perspective for the time period 2000–2010 taking spatial dimension into account. The augmented Solow model is employed to investigate the defence-growth nexus using the cross-sectional data relating to 128 countries. Following a traditional regression analysis, spatial variations in the relationships are examined utilizing different spatial econometric specifications estimated by maximum likelihood. The regressions are compared with each other via likelihood ratio tests, and the spatial Durbin model is found to be the most appropriate one suggesting that the typical least-squares model is misspecified. Empirical evidence indicates that military expenditure has a positive effect on economic growth with a significant spatial dependence for the time period under consideration.

Keywords:

• Military expenditure
• Economic growth
• Spatial data analysis

JEL Codes:

• C21
• H5
• O41
• O47

Acknowledgements

The paper benefited from the comments by two anonymous referees. The usual disclaimer applies.

Notes

¹ For an extensive reviews of the empirical literature, please see Deger and Sen (1995), Ram (1995), Dunne, Smith, and Willenbockel (2005) and Alptekin and Levine (2012).

² For a review of spatial drivers in conflict studies, please see Stephenne, Burnley, and Ehrlich (2009).

³ For technical derivations and the selection of optimal instruments, please see Kelejian and Prucha (1998) and Kelejian and Robinson (1993).

⁴ The list of countries included in the analysis is presented in Appendix 1.

⁵ Please see Sachs and Warner (1995), Collier and Goderis (2008), van der Ploeg and Venables (2009) and Caselli and Cunningham (2009) for an extensive examination of natural resource curse paradox.

⁶ The Urdal and Ellingsen data-sets, which end in 2000, are from http://www.prio.no/ CSCW/Datasets/Economic-and-Socio-Demographic/.

⁷ In 2008 the world has experienced a severe economic crisis. There has been a decline in the growth rates of all countries in the last two years. However, 2000–2010 time period is not long enough to capture the full impact of this economic crisis.

⁸ In the Hausman test, OLS is consistent but inefficient whereas the SEM is efficient; the difference between their estimated parameters is tested. Under the hull hypothesis, there is no significant difference between these two models, which can be interpreted as the omitted variable does not cause an important problem or it is not correlated with the explanatory variables (Le Sage and Pace 2009).

⁹ Please see Kelejian and Prucha (1998, 1999) for technical derivations and the selection of optimal instruments.