Citation

Lee CC, Chen ST. Def. Peace Econ. 2007; 18(6): 537-555.

Copyright

(Copyright © 2007, Informa - Taylor and Francis Group)

DOI

10.1080/10242690601055762

PMID

unavailable

Abstract

To gain better insight into the debate concerning the relationship between defence expenditure and economic growth, this paper empirically applies Hansen's (1996) threshold regression model to examine the threshold effect between the two variables. Allowing defence expenditure to have a non-linear effect sheds considerable new light on the characteristics of the defence-growth link. The single subject, Taiwan, is tested using both a neoclassical one-sector aggregate production function model and Feder's (1982) two-sector production function model over the 1960 to 2002 period. The empirical analysis indicates that the threshold effect strongly exists in 1973 and 1982 coinciding with the occurrence of two energy crises, and the non-linear relationship is solidly supported with both models. While positive externalities seem to prevail for moderate levels of defence spending, the positive effect of defence expenditure on economic growth disappears when the level is relatively large. Thus, only when defence expenditure is small, does the Benoit hypothesis hold. It is demonstrated that a threshold regression provides a far superior empirical model than the standard linear model.