Media reports convey that the prime minister had invited leaders of the main opposition party, the Bharatiya Janata Party, for a meal and, hopefully, to discuss some of the outstanding policy issues.

Can we hope that the occasion represents the beginnings of a reasoned dialogue and debate between the two main political parties, on issues crucial to our future - Kashmir, the nuclear pact with the US, the Indo-Chinese border dispute and, even more importantly for most of us, economic policy and reforms?

For, there clearly is a need for such dialogue and debate - as it is, in the last few years, our Parliament has spent barely 20 per cent of the time it was in session, in discussing legislation, presumably the raison d'etre of its existence.

But this apart, what little discussion does take place, is often purely qualitative - whether to export iron ore or not; agricultural land for industry; divestment/privatisation; using forex reserves for infrastructure; etc - rarely quantitative or analytical.

But let me elaborate: one basic quantitative assumption I make is that there are limits to the resources that can be invested, and these equal the sum of domestic savings and the deficit on the current account.

The other point is that the single-most important problem facing the economy is the creation of jobs, particularly for the less skilled and educated youth entering the employment market.

Take the question of exporting iron ore, by definition a limited resource. Exports imply that the value addition will take place somewhere else. The ministry and the steel industry seem to favour a ban on exporting ore, while companies like Posco and Mittal may not invest in the steel plants unless they get permission to export ore as well.

The steel industry obviously has a vested interest in not allowing the export of ore. However, as a matter of public policy, and given the need to create jobs for the less skilled or educated, is converting ore into steel the optimum use of the limited capital resources? How many jobs does a steel plant directly create and how many of them are for the unskilled? (The downstream jobs need not be considered, as they can as well be created by importing steel.

Again, Bharat Forge, one of the great successes in the engineering industry, now employs only graduates and engineers on its shop floor, which is what a steel plant may also do.) How many jobs would be created by using the same resources in, say, manufacturing toys or garments, which require far less energy and also have much less environmental impact?

Take, again, the question of using agricultural land for industry. Once Sonia Gandhi articulated her views on the issue, all Congressmen are now dutifully echoing them. The problems in acquiring land for the Tata Motors factory in West Bengal are continuing to make headlines.

What I am puzzled about are the numbers, which nobody seems to talk about. What is the percentage of agricultural land that may conceivably need to be used for non-agricultural purposes, over the next ten years? Assuming that the gross income of a farmer is, say, Rs 50,000 per acre, should the land not be sold even at Rs 10 lakh (Rs 1 million) per acre, the risk-free return on which is more than the agricultural income? How many direct and indirect jobs does one acre under agriculture produce? What is the corresponding number for industrial use?

Take the question of disinvestment/privatisation. The Left insists that no profit-making unit in the public sector should be sold. Is it the argument that, so long as an asset is producing a return of, say, Rs 10 lakh, it should not be sold even if it could be sold for Rs 10 crore?

Surely, this is an absurd argument to make, at least in financial terms. Again, given the constraint on resources, what is the cost to society of the forgone resources of Rs 10 crore (Rs 100 million) - in terms of the number of primary schools, the kilometres of rural roads or water taps, etc that will not be built/provided in the absence of resources?

Note that the latter investments would give a far higher return to the economy than keeping the unit in the public sector because it is profitable, even if the profit is one per cent of the sale price! Surely it cannot be argued that these are irrelevant issues, and one would like to debate and analyse them using numbers and not merely whether a unit is making profit or not.

The issue of using forex reserves for infrastructure seems to have been raised once again, this time apparently by the Finance Ministry: a couple of years back, the proponent was the Planning Commission, but not much came from the idea.

Prima facie, it looks so attractive: on the one hand, we have $160 bn earning barely 3-4 per cent p.a. and, on the other, there is a crying need for infrastructure investment. What better than to use the reserves and provide the funds?

It is such a plausible solution, but like too many plausible ideas, wrong. First, are resources the constraint in infrastructure investment or the regulatory, legal and other issues limiting progress?

Secondly, the reserves are not "idle": an equivalent number of rupees, say Rs 7 lakh crore (less sterilisation), have been released in the economy when the dollars were bought. And these rupees are meeting the economy's financing needs.

Thirdly, we would, in any case, be using reserves to the extent of perhaps $20 bn in the current year, for financing the deficit on the current account. The deficit has not arisen purely because of infrastructure investments but the point is that, for using forex reserves for whatever purpose, the deficit on the current account will have to go up correspondingly.

What is our comfort level on that - 2.8 per cent of GDP as the Approach Paper to the 11th Plan seems to suggest? This would mean an annual average of $30 bn over the next five years. At what point of time would the foreign investors start feeling uncomfortable, and stop financing it?