ARMS RACE AND ECONOMIC GROWTH: THE CASE OF INDIA AND PAKISTAN*

*An earlier version of this paper has been presented at the Seventh Annual International Conference on Economics and Security held in Bristol, UK, on 26–28 June 2003.

Abstract

The hostility between India and Pakistan is believed to have led to an arms race between the two countries, which might have contributed to their retarded economic growth. This paper investigates this twin problem of arms race and economic growth for the time period 1949–2003. The empirical results suggest that there is a mutual causal relationship between the military expenditures of India and Pakistan. Even though military expenditure does not Granger cause economic growth in Pakistan, there is causality from military expenditure to economic growth in India. A VAR analysis revealed that military expenditure hinders economic growth in India in the long‐run, but it has a growth promoting effect in the short‐run.

*An earlier version of this paper has been presented at the Seventh Annual International Conference on Economics and Security held in Bristol, UK, on 26–28 June 2003.

Keywords:

• Military expenditure
• Arms race
• Growth
• India and Pakistan

Notes

*An earlier version of this paper has been presented at the Seventh Annual International Conference on Economics and Security held in Bristol, UK, on 26–28 June 2003.

¹See Alexander (1987), Deger and Sen (1990), Tibbett and Akram‐Lodhi (1997) and Ganguly (1995) for the historical review of the conflicts between the two nations.

²For a brief literature review see Oren (1994) and the references therein.

³See for example Deger and Sen (1990), Dunne et al. (1999), Öcal (2003) for the India–Pakistani arms‐race models.

⁴See for example Sezgin and Yıldırım (2002) for Turkey; Kollias and Paleologou (2003) for Greece; Dunne and Perlo‐Freeman (2003) for a group of developing countries.

⁵A similar model has been adopted by Kollias and Paleologou (2003).

⁶These results are not reported here to conserve space but are available from the authors.

⁷In order to establish the robustness of the results, the tests of causality were conducted using several lag lengths for each model. The results, which are not presented here to conserve space, are not sensitive to the choice of the lag length.

⁸Please see Sandler and Hartley (1995) and Ram (1995) for comprehensive empirical literature.

⁹See Hendry (1995) for a discussion of inclusion of trend in the cointegrating space.