LITERATURE REVIEW

The first study which explored the above relationship was conducted by ^{Benoit (1973},78) among the less developed countries, and the findings suggested that military expenditure has a direct impact on the economic growth of less developed countries. Military expenditure promotes infrastructure development, employment opportunities, and helps in economic development. Contrary to Benoit, the study conducted by ^{Lim, (1983)} and ^{Biswas and Ram, (1986)} suggested that an increase in defence spending is more favorable for developed economies as compared to less developed economies. Studies that support the above views are ^{Deger (1986)} and ^{Yakovlev (2007)}.

^{Batchelor (2000)} conducted a study on the economic benefits of military spending on the South African countries by employing the ARDL approach. The findings of the study suggest that military spending boosts the manufacturing sector; thus, in the long-run, military spending has a direct influence on economic development. On the contrary, a study conducted by ^{Dunne et al. (2001)}, using modern vector autoregressive (VAR) between Greece and Turkey, concluded that defence expenditure has an inverse relationship with the fundamental development of Turkish economies.

^{Yang et al. (2011)} explored the nexus between defence and economic growth expenditure. The results of the paper suggested that 23 countries that have an income below a threshold limit of U.S. $ 475.93 conclude that there is a negative relationship between the two variables, whereas when the income increases above the threshold limit, economic growth tends to show a decreasing trend. Similar findings were suggested by ^{Dunne and Tian (2013)}, who employed a dynamic panel and exogenous growth model in 106 countries, covering the period between 1988 and 2010.

^{Aye et al. (2014)} analyzed the association between military spending and economic growth in African nations during the period 1951-2010 by employing a bootstrap rolling window. The findings suggested that for the periods 1966-1972, 1973-1975, 1975, 1977, and 1979-1984, military spending had a direct influence on economic growth; however, for the rest of the period, military spending had an indirect relationship with economic growth. ^{Mosikari and Matlwa (2014)} also supported the findings by employing the Johnsen and Eagle cointegration technique and concluding that military spending has a direct influence on economic growth.

One of the latest studies based on the above relationship was conducted by ^{Raju et al. (2019)} using cointegration and causality test on the panel data of three South Asian countries. The findings of the study suggested that there is a positive long-run relationship between military expenditure and economic growth. The study also concluded that there is a unidirectional long-run causality among the variables in all the three South Asian countries. Similar findings were also reported by ^{Khalid and Mustapha (2014)}, who analyzed the relationship between military expenditure and economic growth of India using an autoregressive distribution lag approach. Findings suggested that military expenditure has a direct relationship with economic growth in the short-run. The study also concluded that there is a unidirectional relationship among the variables as per causality analysis.

Studies that stated a positive relationship between military spending and economic growth include the ones by ^{Kollias, Mylonidis, and Paleologou (2007)}; ^{Awaworyi and Yew (2014)} and ^{Sezgin (2001)}. Studies that concluded the negative relationship between military spending and economic growth are those by ^{Lebovic and Ishaq (1987)}; Vougas (1999); ^{Dunne and Tian (2013)}. Apart from the positive and negative relationship, studies that have pointed out that military spending has no impact on economic growth are the ones conducted by ^{Abu-Qarn and Abu- Bader (2003)}; ^{Habibullah, Law and Dayang-Afizzah (2008)}, and ^{Pradhan (2010)}.

Theoretical background

There are two contrary schools of theories that provide a theoretical background for this work. Keynesian schools state that military expenditure boosts economic growth by offering investment and employment opportunities, and providing technological advancement. Contrary to this hypothesis of defence burden postulates, Benoit (1978) says that huge defence expenditure not only promotes tax burden on the society but also crowds out private investment opportunities. Figure 3 shows the theoretical background, and based on these theories, the study further explores the above topic.

Source: author's elaboration.

Figure 3 Theoretical background 

Research gap

Although various studies have investigated the impact of military spending on economic growth, this study is quite different on the following grounds from the previous work. First, it incorporates South Asian countries, which was not fully explored in previous works. Second, the main limitations of the above studies were that they have assumed time series to be linear and have just checked the symmetric effect by employing a linear ARDL approach. ^{Anoruo (2011)} postulated the non-linear nature of time series. Thus, to find the hidden cointegration, as advised by ^{Yoon (2002)}, this study has investigated both non-linear and linear aspects of the variables involved by employing an asymmetric ARDL approach (NARDL). Lastly, this study also tried to relate the work with military Keynesianism, which was also not explored in previous work, together with examining the role of military expenditure on industrial productivity.

METHODOLOGY AND DATA ESTIMATION TECHNIQUE

This study analyses the asymmetric relationship between economic growth and military expenditure in India, China, and Pakistan, covering the period from 1990 to 2018. The data for Economic growth(EG), Inflation (In), and Industrial Productivity(IIP) were collected from the International Monetary Fund statistics database, and the data for military expenditure (ME) were retrieved from Stockholm International Peace Research Institute (SIPRI). This study is only limited to India, Pakistan, and China, as they are the most important South Asian countries and also considered as the three lions of South Asia. Also, these countries have a convoluted history of unresolved border disputes, with military confrontations between India and China at the Line of Actual Control (LAC), and India and Pakistan at the Line of Control (LoC), occurring intermittently. Therefore, studying the impact of military expenditure on economic growth and industrial productivity will help in better understanding military confrontations and their implications.

This study has explored the asymmetric relationship between the variables using the NARDL approach. NARDL approach is employed as it jointly verifies the asymmetry and cointegration of the variables under study (Ullah et al., 2020). Moreover, they also consider hidden cointegration (^{Shahzad et al., 2017}). The NARDL model jointly investigates long run and short asymmetric effects along with the flexibility of order of integration: I(0), (^{Shin et al., 2014}). The NARDL model is also much superior to other error correction models as it jointly explores asymmetries and cointegration, and provides robust findings even with small samples (^{Granger, 2002}).

The following equation was used for long-run analysis:

Where EG represents Economic Growth; ME shows Military Expenditure; In, inflation; IP, Industrial Productivity, and β _i . shows the coefficient vector representing the long-run period. We have further segregated the military expenditure into negative and positive components to obtain asymmetric results.

Hence, Eq (1) is written as:

Where ME _t ⁺ and ME_t ^- represent positive and negative changes in Military Expenditure, and Ω_i shows the long-run coefficient vector.

Equation 2 has been rewritten in the ARDL framework as stated by ^{Shin et al. (2014)}

Where a, b, c, d, and e are the respective lags.

Before proceeding with the results of NARDL and ARDL, certain pre-tests are required. The order of integration has to be I (0), I (1) for applying the NARDL model; therefore, for checking the order of integration, the Augmented Dicky Fuller test is employed, as suggested by ^{Pesaran et al., (2001)}. As the Augmented Dicky Fuller test may provide specious results due to the problem of structural breaks, we have also used Zivot and Andrews test (^{Qazi et al., 2017}; ^{Raza et al. 2017}).

Table 1 and Table 2 show the results of the Augmented Dicky Fuller and Zivot and Andrews test. As the unit root test confirms that none of the variables is I (2), so the study further proceeded with the NARDL estimation, Schwartz Information Criteria (SIC), and the FPss test, used for optimum lag selection and for checking cointegration. In the FPss test null hypothesis represents no cointegration against the alternative hypothesis of cointegration. In the FPss test if the calculated value is more than the upper bound value, then, in that case, the null hypothesis is rejected and vice versa. The results show rejection of the null hypothesis, meaning variables have asymmetric cointegration. In addition to the above test, other diagnostic tests are also performed. Annex 1 depicts the results of the diagnostic tests. Lastly, the WALD test is also employed to find out the short and long-run asymmetric effect of military spending on economic growth. Annex 2 shows the result of the WALD test. The null hypothesis, H₀: No long-run asymmetric relationship, and H₀: No short-run asymmetric relationship was checked against their alternative hypothesis.

CONCLUSION AND MANAGERIAL IMPLICATIONS

This study investigated the asymmetric relationship between economic growth and military spending on the panel data of India, China, and Pakistan, covering the period from 1990 to 2018. Using the asymmetric relationship, the objective of this research was also to examine the Military Keynesianism philosophy on the panel of the three countries separately. Unit root, structural break, and BDS test are employed to analyze the above relationship. The findings of the FPss confirm an asymmetric cointegration between military spending and economic growth for India and China. Using NARDL, as proposed by Shin et al., the study concluded that military spending has a significant long-run impact on the economic development of India and China along with the fact that positive influence of military expenditure on economic growth is much more powerful than the negative shock. Meaning the linear model will show the hidden cointegration among the variable in the long run. On the other hand, in the short run, this relationship turns out to be symmetric.

Findings suggest that policymakers should pay more attention to nonlinearity for military spending and economic growth, and accordingly devise appropriate policies for the betterment of the economy. The study shows that for India and China, military expenditure is beneficial in the long run as, according to Keynesian demand, high military spending will promote high demand for labour, infrastructure development, and resource allocation. It will also help in providing employment opportunities, along with investment opportunities. Research for better weapons and advanced technology boost research and development funds, which can be used in different research work related to industrial and domestic productivity. As military spending promotes national security, thus countries that are at a stable pace of development like India and China should keep fuelling their military spending as this will contribute to economic growth, as postulated by the Military Keynesianism approach.

But countries like Pakistan who are still struggling to maintain their economic stability should focus more on their internal domestic development and other sectors apart from just concentrating on the defence sector. The results show that military spending in Pakistan is not so significant to economic growth. Thus, too much military spending crowds out the economic budget, which results in the budget constraint for other sectors like education, health, and industrial productivity. Thus, from this study, it is concluded that military expenditure is beneficial for developed or developing countries that have a stable growth rate.

Limitations/ future scope of the study

This work focuses only on the panel of three countries: India, Pakistan, and China, thus allowing replicating the same study with other countries like the U.S or Russia, which are spending significantly in their defence sector. This study has taken only selected variables such as economic growth, industrial productivity and inflation, which also provide an opportunity to see the effect of military spending on other key macro-economic variables.

Practical implication for Asian countries

Over the years, the defence expenditure of most South Asian countries is showing a downturn. This fall in the defence budget is also affecting the research and development funds of South Asian countries. This downturn could reflect the end of a military-modernization cycle that began after the 1997-98 Asian financial crisis. In 2012 the rate of defence expenditure growth of South Asian countries was higher than that of the whole of Europe. But over the years, defence expenditure is showing a downfall which is negatively affecting the economic development of South Asian countries as well, as this study projects that appropriate defence spending or military expenditure is beneficial for the economic growth of a country as a whole. An increased defence sector budget not only helps in infrastructure development and employment opportunities but also helps in the development of better technologies and the know-how, which can be used in the creation of countries' business, production, and overall economic growth.

Thus the findings of this study will help in understanding the importance of defence budget for the overall growth of a country based on the Military Keynesianism Postulate. As India and China are the most emerging markets among South Asian countries, the outcome of this study will help to understand the significance of the defence budget on business and economic growth in overall South Asian countries, from a broader perspective.