In the coming days, Finance Minister Jaswant Singh will have to struggle with ways to reduce the Budget deficit in his Budget proposals for 2003-04. The task will obviously not be easy.

The Bharatiya Janata party has clearly asked for a cherry picking of certain Kelkar proposals, which favour its constituency. This would be simply imprudent fiscally.

But while the debates on tax exemptions occupy the salaried sections who are among those who would be affected the most, an important aspect of the Budget is getting inadequate attention. The need for looking at the expenditure side of the Budget is equally important. After all, rupee saved on expenses is as good as a rupee from additional tax revenue.

The focus on raising the tax-GDP ratio is actually inappropriate as long as public spending is not streamlined. Currently, what is keeping public expenditure from bloating limitlessly is the Budget deficit.

If tax collections increase, Parliament could just spend these gains in inventing more programmes which would transfer these resources to pet constituencies. There are important issues about expenditure which have not yet been addressed.

While there is consensus that wasteful expenditure should be curtailed, there are no clear guidelines about what is wasteful. This is not surprising because the definition of wasteful comes from the role of government.

While the role of the government has been redefined under reforms, the discussion about the role of government departments and staff that should accompany this change has not really got adequate attention.

The debates on expenditure have tended to think in terms of distinctions like plan versus non-plan expenditure and current versus capital expenditures. Somehow, 'non-plan' and 'current' expenditures are bad. But this is bad economics. Governments spend well when they produce public goods.

Paying salaries to the police or the army is non-plan and current expenditure, but it is certainly about governance, it is certainly about producing public goods. In contrast, silly expenses like buying planes for Air-India might be plan expenditure and might be capital expenditure, but this has nothing to do with legitimate goals of governance.

One of the crucial changes that must accompany the process of reforms is the role of the bureaucracy. With increasing liberalisation of both trade and industry, a number of ministries and functions that were created during the period of import restrictions and industrial licensing have lost their meaning.

With privatisation and divestment, the role of the departments that monitored and controlled these public sector enterprises has become redundant. Yet, while we speak of privatisation, we are failing to dismantle the babudom that controls the public sector units.

India has an extremely bloated set of ministries by world standards. It is striking to see that in Japan, there are only ten ministries.

They are (a) Public Management, Home Affairs, Posts and Telecommunications, (b) Justice, (c) Foreign Affairs, (d) Finance, (e) Education, Culture, Sports, Science and Technology, (f) Health, Labour and Welfare, (g) Agriculture, Forestry and Fisheries, (h) Economy, Trade and Industry, (i) Land, Infrastructure and Transport and (j) Environment. All these titles seem to link up to genuine governance concerns and producing public goods.

In India, is it not amazing that the government has done away with import licensing but the ministry to implement import licensing remains? The size of the staff of the commerce ministry has hardly been cut. The department of commerce has a number of departments associated with it that add up to a total of over 6000 employees.

'Redefining' roles is not sufficient. The role of the commerce ministry in the times of import licensing, foreign exchange shortages and under the regime of export promotion to Russia and East European Countries perhaps justified its size. Today its regulatory role is non-existent.

Instead, its genuine governance functions are now limited to World Trade Organisation-related issues. The change in its role has not resulted in a significant restructuring and reduction in its size. Perhaps the ministry can be closed down, and the WTO negotiations group can be moved into the ministry of finance.

The case of the ministry of steel is similar. Perhaps its staff had something useful to do when it was expected to coordinate growth and production in the steel industry.

With the ushering in of reforms in the 1990s, over half of the steel production is now in the private sector. The role of the joint plant committee was to make guidelines for the production and distribution of steel materials.

There needs to be adequate recognition of the fact that this ministry and its departments have no reason to exist. As with steel, we have a series of ministries for specific industries. But industrial policy, in the sense of government taking interest in which industry does well and which industry does badly, is now obsolete.

These ministries now merely serve to do lobbying for their industries, while extracting rents from them. Each of these ministries should be closed down. Prime examples of these include the ministry of heavy industries, the ministry of food processing, and the ministry of chemicals and fertilisers.

The ministry of information and broadcasting is a particularly important one in targeting closure. It is animated by an Orwellian vision which should be anathematic given India's dreams of becoming a liberal democracy. Such a ministry belongs in illiberal states like China or Pakistan or Saudi Arabia, not in free India.

Doordarshan and All India Radio should be converted into corporations and handed over to the ministry of divestment, and then this ministry should be closed down. Closing down ministries will save money directly.

In an era when meddling by government in markets is being reduced, it is incomprehensible that the size of the government, particularly in aspects where the government is meant to interfere in markets, should be maintained. But that is not all.

The presence of a bureaucracy whose rent seeking interests are aligned with the continuation of controls can create resistance to reforms.

Privatisation is now firmly on the agenda of economic policy today -- we all agree that the job of the government is not to be in business. But it should be accompanied by a complete elimination of the associated bureaucratic machinery.

The reforms process should aim to get a lean set of ministries, all of which are focused upon the job of producing public goods, and upon the genuine problems of governance. This will save money, and reduce the confused process through which government sets about trying to do governance.

How might Vajpayee go about this? A two stage procedure which combines economics and politics might make sense. First, Vajpayee should ask his Economic Advisory Council to pick twenty ministries which fail to produce public goods, which have the least connection to legitimate concerns of governance.

Then, he should ask his senior ministers to choose ten of these that can safely get the axe. And in this, let us all hope (as citizens of free India) that information and broadcasting makes it into the shortlist!