Country Survey: Defense Policy and Military Spending in Taiwan, 1952–2009

Abstract

In this article we survey military spending, the evolution of national defense policy, and the role that defense spending played in economic activities in Taiwan from 1952 to 2009. With the ongoing cross-Strait dialogue and closer commercial ties between Taiwan and China, the share of military spending in government expenditures (or GDP) has tended to gradually decline, which is in accordance with the change in defense policy from an offensive posture to a defensive-oriented attitude. In addition, we investigate the defense spending–inflation nexus in Taiwan. The multivariate structural change test identifies two break points and three regimes are categorized accordingly. The Granger causality test based on the VAR model reveals that the heavy defense burden in Taiwan during 1952–71 (the first regime) is essentially an important factor causing the higher price levels. However, when the tension across the Taiwan Strait is alleviated as well as the defense-oriented policy adopted, the military spending no longer plays a crucial role in determining inflation in the recent two regimes. Sensitivity analysis confirms that our results are robust to different model specifications.

JEL Codes:

• C32
• H56
• P24

Keywords:

• Defense policy
• Inflation
• Military spending
• Taiwan

Acknowledgements

We are grateful to the Editor Christos Kollias for drawing our attention to conducting this project. We also thank Christos Kollias, Hamid E. Ali and anonymous referees for their helpful comments and suggestions. Financial support from the National Science Council in the form of grant NSC 99-2410-H-007-022 is gratefully acknowledged.

Notes

¹In what follows, we will refer to the political regime that relocated to the island of Taiwan since 1949 as Taiwan, the Republic of China (ROC), or the Kuomintang (KMT), while mainland China (or simply China) and the People’s Republic of China (PRC) are used interchangeably to represent the country ruled by the Chinese Communist Party.

²The Min Nan language belongs to a family of Chinese languages spoken in southern Fujian (in mainland China) and its neighboring regions, including Taiwan.

³The Human Development Index is a statistical measure that gauges a country’s level of human development according to the United Nations. Unlike GDP per capita or per capita income, the HDI takes into account how income is turned ‘into education and health opportunities and therefore into higher levels of human development.’ For more details, see http://hdr.undp.org/en/.

⁴Please refer to International Institute for Strategic Studies (IISS (International Institute for Strategic Studies), 2010) at http://www.iiss.org/ and Stockholm International Peace Research Institute (SIPRI (Stockholm International Peace Research Institute), 2009) at http://milexdata.sipri.org/.

⁵Please refer to the following website: http://163.29.3.66/english/index_01.html.

⁶The US Congress passed the China Aid Act in 1948 to appropriate $500 million to the ROC government (ruled by the KMT) for reconstruction and development (Jacoby, 1966).

⁷The Taiwan Relations Act also states that it is US policy to provide Taiwan with such weapons as may be necessary for its security and an adequate defensive capability. Based on this Act, the US sent two carrier battle groups (CBGs) (including the Independence and Nimitz) to Taiwan and resolved the so-called 1995–6 Taiwan Strait Missile Crisis, which refers to the two sets of missiles fired by the PRC during 1995 and 1996 to send a strong signal to the ROC government for its moving away from the One-China policy and to intimidate the Taiwanese electorate in the upcoming 1996 presidential election.

⁸More specifically, this was in order to reconstruct the ROC Armed Forces through streamlining command levels, renovating logistic systems, merging military schools and upper-ranking staff units, eliminating overlapping staff units and reducing the total number of military manpower as well as transferring non-military tasks to organizations outside the MND.

⁹For instance, the ‘921’ earthquake was a 7.3 Ms or 7.6 Mw earthquake which occurred on 21 September 1999 in Jiji, Nantou County, Taiwan. It resulted in 2,415 people killed, 11,305 injured, and $10 billion US dollars worth of damage. On 8 August 2009, Typhoon Morakot brought an unprecedented level of rainfall to the southern part of Taiwan, dumping more than two meters of rain on the island in just two days and causing the worst flooding and mud-slides in 50 years. Hsiao Lin Village in Kaohsiung County was buried by the mud-slides. The total death toll officially reported was over 700.

¹⁰Note that the Nationalist Party governed both regions, mainland China and Taiwan Island, from 1945 to 1949. The data on WPI, the government budget deficit, total money supply, and the share of military spending in government spending in China are drawn from Chang’s (1958) classic book, which offers a comprehensive collection of the desired data in mainland China from 1939 to 1950. In addition, the Taiwanese WPI in these periods originates from Wu and Gau (1991).

¹¹The Communist government (PRC) has been the official government ruling mainland China since 1949, while the Nationalist government (ROC) governed Taiwan.

¹²The decline in defense spending is closely related to the major military conflicts between China and Taiwan from 1949 to 1965. Please refer to Table I for the summarized conflicts.

¹³Available military spending in AREMOS is the same as that in Taiwan’s Yearbook of Financial Statistics, officially published by the Ministry of Finance in Taiwan.

¹⁴Please refer to the following websites: http://eng.stat.gov.tw/mp.asp?mp=5 and http://www.datastream.com.

¹⁵The M1 growth rates in Wu and Gau (1991) and AREMOS are similar to each other during the 1962–82 period.

¹⁶Please refer to: http://research.stlouisfed.org/fred2/series/OILPRICE.

¹⁷Please refer to: http://www.ceicdata.com/.

¹⁸The data appendix of Reinhart and Rogoff (2004) can be downloaded at: http://terpconnect.umd.edu/creinhar/Data/0L_black_official_PREMIA_22.xls. The link to International Financial Statistics can be found at: http://www.imfstatistics.org/imf/logon.aspx.

¹⁹Let n be the dimension of Y_t We have n=4 in this case.

²⁰More specifically, to construct the per capita real GDP, we divide the nominal GDP by the GDP deflator and the size of the population.

²¹It is true that an error correction model may be a favorable modeling strategy when data display non-stationarity and there is likely a long-run relationship among the variables examined. Hence, the following two procedures have been considered to address this issue (results are available upon request from the authors). First, based on the visual inspection of the trending characteristics in data and various unit-root tests, we conclude that all variables under consideration in this study are non-stationary (i.e. I(1)). Second, utilizing the co-integration test developed by Johansen (1995), no stable long-run relationship among the variables can be found in this study. Therefore, it motivates our usage of differenced variables in a VAR framework to examine the causality between defense spending and economic activities.

²²vecr(Π) creates a column vector by appending the transposed rows of Π to each other.

²³Most of the crude oil used in Taiwan is imported from abroad. See the 2010 US Department of State website for more details: http://www.state.gov/r/pa/ei/bgn/35855.htm.

²⁴During our empirical studies, a one-period lag term is selected based on the AIC.

²⁵Please refer to Qu and Perron (2007) for more details.

²⁶ where the definition of a_m can be found in Bai and Perron (1998).

²⁷Nowadays, oil refining has become one of the major industries in Taiwan. For example, Formosa Petrochemical Corporation (FPC) has invested approximately US$7.4 billion in refineries. It is reported that FPC earned profits when the oil price soared in the early 2000s. See the following website: http://www.fpcc.com.tw.

²⁸We divide Taiwan’s money supply by its population to obtain per capita money supply.