While the debate on the usefulness of government expenditure as a fiscal policy instrument to correct short-term fluctuations in economic activity has largely been resolved , the question of existence and direction of relationship between various components of expenditure and economic growth is far from settled.

One of the most interesting and controversial results on the subject is that "countries with heavy defence budget generally had the most rapid rate of growth, and those with lowest defence burdens tend to show the lowest growth rates.

These unexpected results provided a strong impetus for the research on the relationship between defence expenditure and economic growth. However, the empirical evidence on the subject remained controversial largely on account of specific factors. The major country specific factors include the role of military in non-military affairs, internal security needs, military in rival countries and geopolitical position in the country.

Pakistan has also attracted the attention of the researchers both within and outside the country due to its huge defence expenditure in an effort to maintain a creditable level of deterrence, a vital geopolitical position position in Afghanistan war, pursuance of nuclear capabilities and a long standing dispute over the territory of Kashmir with India. In fact, large size of defence expenditure in overall budgetary outlay(in presence of high budget deficits) remained a hot topic of debate among the domestic researchers as well.

Surprisingly, Robert E. Looney, a professor of National Security Affairs at Naval Post Graduate School, USA concluded that Pakistan has been using its defence expenditures "as a stabilization, especially during the periods of relative peace with India." The author further argued that defence expenditure seems to be controlled by economic environment instead of exogenous factors. Looney has written frequently on Pakistan's defence spending.

Over the years, defence spending of Pakistan also remained important topic of research. The large size of the defence expenditure in presence of high budget deficits, declining development expenditure and increasing debt servicing on account of exploding public debt focused the attention of the domestic and foreign economists on the subject.

These factors together with Pakistan's pursuit of nuclear capability, responses to (although weak) increased Indian defence expenditures and prevalence of considerable poverty attracted the attention of the researchers outside the country.

This study was aimed to analyse the casual relationship, if any, among the defence expenditure, development expenditure, inflation and economic growth. The result indicate the presence of a long-term relationship among defence spending, development, inflation and GDP. The variables are connected in Granger- casual chains, however, short-term and long- term-chains differ with each other.

Although there is a long-run temporal causality relationship among the variables, defence expenditure cannot be used for stabilisation purpose due to absence of any significant temporal relationship in the short-run. Furthermore evidence also suggests that the defence expenditure cannot be used to counter inflationary pressures in the short-run.

Significantly, there is no significant causal relationship between development and defence expenditure. This undermines the popular perception that increases are generally accompanied with the development expenditure.

The popular perception that defence expenditure hurts economic growth does not seem to hold over the estimation period (FY51-FY2003) but the military Keynesian hypothesis also does not hold in case of Pakistan. Nevertheless, one should not forget that defence expenditure is primarily used for securing external defence, which has an undisputed intrinsic worth for any nation.

Pakistan's defence spending remained one of the largest components of total government expenditures since independence. Although sizeable variation in defence expenditure to GDP ratio has been witnessed over the past five decades and the ratio declined significantly toward the end of the 20th century, the absolute size of defence expenditure is considered still very high. The defence expenditure was 17.7 percent of the government's budgetary outlay for FY03.

Specifically, the shares of defence and development expenditures in overall budgetary outlay were 33.5 percent and 17.7 percent respectively during FY51. This exceptionally high share of defence expenditures in early years of independence may be largely attributable to the government efforts to achieve a minimum level of deterrence, necessitated by the conflict on disputed territory of Kashmir and a war with India in 1948.

Later, the share of defence expenditure in total expenditure saw a considerable decline with some fluctuations before spiking up again in FY66 on account of 1965 war with India.

In the post-1965 war era, the defence spending to the total expenditure ratio saw a sharp reversal in FY67. However, this decline proved short lived, as the ratio surged again in FY72 due to 1971 war before dipping down to pre-1965 war trend.

After FY72, the ratio gradually declined to 23.2 per cent by FY80. However, the declining trend once again was reversed in FY81 following the high tension in Afghanistan (Pakistan as a front line state).

During first half of 1980s, the share of defence expenditure averaged 26.8 percent of total expenditure, indicating an increase of 4.1 percentage points on average ratio for the second half of 1970s.

The withdrawal of Russian forces from Afghanistan coupled with the prevalence of high fiscal deficits propelled government to revisit its defence spending. As a result, the decade of 1990s recorded considerable decline in the share of defence expenditure. The decline in second half of 1990s was more pronounced compared to the first half.

Despite tensions on borders with Afghanistan (following the September 11) and India (due to incident of December 13), the share of defence expenditure continued to decline and averaged 18.7 percent for FY01-03. This recent declining trend amidst a few episodes of tension, is largely underpinned by the nuclear capabilities of Pakistan and neighbouring India. The possession of nuclear weapons by both countries seems to be an important factor in minimizing the chances of war, even with the enlargement of traditional weapons.

It is generally perceived that massive decline in development expenditure to GDP ratio over the past two decades is primarily on account of huge defence expenditure. However, the data do not lend credence to this view, as both defence and development expenditures are on decline since mid-1980s, not only in term of their ratios to total expenditures but also in terms of GDP. The decline in defence and development expenditures over the past one and half decades is largely attributable to the consolidation of public sector.

Furthermore, the rise in defence expenditure generally accompanied the increase in overall expenditures, instead of decline in development expenditure. Specially, there are only seven years (out of 52 years) in which the increases in defence expenditure to GDP ratio does not accompany the increases in total expenditures to GDP ratio, but moves with the declines in development expenditures to GDP ratio.

In sum, the war driven changes in defence expenditure in Pakistan hardly support the Military Keynesianism Hypothesis (MKH) and lend more credence to the view that the defence expenditure is used hardly support the MKH and lend more credence to the view that the defence expenditure is used mainly for securing external turn cyclical movements in the economy.

Furthermore, the negative relation between the defence and development expenditure does not seem to be hold.

Extracts from a working paper on "Defence Expenditure and Macroeconomic Stabilisation: Causality Evidence from Pakistan" by a central bank analyst. The paper has been put on the website of the State Bank of Pakistan for comments and suggestions though the author says the paper reflects his personal views and not necessarily that of the organization.