TIME-VARYING CHARACTERISTICS OF SOUTH KOREA-UNITED STATES AND JAPAN-UNITED STATES MILITARY ALLIANCES UNDER CHINESE THREAT: A PUBLIC GOOD APPROACH

Abstract

This paper investigates the nature of two military alliances under Chinese threat. The findings are as follows: First, South Korea does not consider China a significant threat while Japan and the United States have recognized China as a serious threat since the 1990s and the 2000s, respectively. Second, the relationship between South Korea and the United States is a true military alliance for all time periods, but the nature of the alliance has changed since the 1970s. Third, although Japan began to form an alliance relationship with the United States in the 1990s, Japan is considered a more significant ally by the United States. This paper implies that, should China provoke a military confrontation, it might be difficult to deduce a common solution among the three countries because of the different response to military threats from China.

Keywords:

• Military alliance
• Defense goods
• Threat elasticity
• Spill‐in elasticity
• Public good

JEL Codes:

• H410
• H560
• O530

ACKNOWLEDGEMENTS

We gratefully thank an anonymous referee for helpful comments. The second author is also grateful for the financial support provided by the Hankuk University of Foreign Studies Research Fund of 2010.

Notes

¹ GDP is based on purchase power parity valuation from the International Monetary Fund (2009).

² Even though a joint product model was applied to our study, the results would be qualitatively the same. In the model of no alliance with other countries, the reaction (or elasticity) to the outside threat is the biggest because the country without alliance relationship with other countries should directly cope with the outside threat by its own defence forces. The smallest reaction to the threat occurs in the pure public goods model. It is because defence goods consumed (not purchased) is the biggest among the three models. However, the signs of the threat elasticity and spill‐in elasticity might be identical among the three models.

³ Certainly there are some potential roles of the U.S. military power around the world for the East Asian regional defence. However, if we include the whole U.S. defence budget for the consumed defence goods for Korea and Japan, it was ultimately impossible to estimate the equation system due to a singular problem in the estimation. This could be a practical reason for the definitions of consumed defence goods for South Korea and Japan in this paper. The total U.S. defence budget is so large relative to the Korean and Japanese defence budgets that it causes a singular problem when total military expenditure of the United States is considered as public goods to its allies, South Korea and Japan. The military expenditure of South Korea is only 3–5% of the U.S. military expenditure during 1970–2005. It was 3–13% for the case of Japan.

⁴ The term are referred to as the supernumerary and subsistence incomes, respectively.

⁵ The Chinese threat could have an endogenous nature in that the Chinese defence expenditure would respond to a military buildup by the U.S. and its allies in Northeast Asia. If this kind of endogeneity is significant, we may need to develop a game theoretical framework to encompass the interactions between China and the U.S. alliances. However, a game theoretical approach is beyond the scope of this paper and could be a future research topic.

⁶ When defence good deflator (P ₁) is not available (South Korea and Japan in our sample), we use the deflator for government expenditures. We use the government expenditure deflator instead of the overall GDP deflator since the government expenditure deflator is more closely related to the defence deflator than the overall GDP deflator, which could be found from the U.S. case where the defence deflator data are available. With the U.S. data for the period of 1970–2005, the correlation coefficient between the government expenditure deflator and the national defence deflator is 0.999, while the correlation coefficient between the GDP deflator and the national defence deflator is 0.995.

⁷ The data for the number of U.S. military personnel stationed in South Korea are from Korea Ministry of National Defence (2008) and those stationed in Japan are from Heritage Foundation (2009).

⁸ The use of military expenditure as a measure of threat might be questioned because of the possible inaccuracy, but it appears that we are lack of other reliable data. Murdoch and Sandler (1982, 1984), Okamura (1991) and Goo and Kim (2009) also used military expenditures to measure the size of threat of the opposing countries.

⁹ It is not easy to interpret the estimated parameters for understanding the exact relationship between military expenditures and exogenous variables in nonlinear models.