This though, seems a particularly bad time to get indignant on behalf of theIndian software industry. As I write this, the stock market has lost morethan 500 points in around a week's time. Much of it is, I am assured,because foreign institutional investors have decided to sell their shares inIndian software firms and head for the hills. But they will, at a morepropitious time return, whenever that may be. Questions about the future ofour software industry are not entirely out of place. And part of thisscepticism (mild scepticism, I hasten to add) rises from the question ofthese firms' fortunes in a post-Y2K world. Nearly 40 per cent of allrevenues in the Indian IT sector can be directly attributed to Y2K-relatedconsulting and solutions. This means trouble on two fronts. The obviousproblem will be in the mad scramble for new sources of revenue in theimmediate future. But equally importantly, in an industry that competes onits intellectual capital, vast numbers of our engineers who have so farconcentrated on menial coding tasks will have to be retrained to add valuein other functions in the software industry.

But there are more fundamental reasons to worry about the future of theindustry here. Start with the most obvious, and the best elucidated,problem, that of an efficient infrastructure for the production of software.

Among the most vital inputs into the software industry are the services of atelecommunications sector which is competent at delivering high qualityservices at the lowest prices. On both counts, Indian telecommunicationshave had a horrible record. Estimates suggest that the average Indian firmmay consequently spend up to a third more in such costs than a comparableAmerican firm. This is especially dreadful considering the fact that Indianfirms can usually compete only on a cost basis with foreign firms. Thisbrings us to the next, and arguably even more important weakness of oursoftware industry. Our firms have so far not made much of a foray intodeveloping new products for rich country markets. They have instead,concentrated on providing a few labour-intensive services at internationallycompetitive prices. Only about 10 per cent of revenues come from the sale ofstandardised products and packages. As much as 55 per cent of revenues comefrom providing support and maintenance services, where the advantages oflow-cost labour is often decisive.

This may primarily be a result of a lack of effective demand for suchservices in the domestic economy. Remember it is only recently that thesector started enjoying large trade flows. Software is an industry with ahigh income elasticity of demand. This obviously meant that the demand forsuch services in a low-income market such as India's was on the lower side.

Another vital factor that limited domestic demand has been the lax standardsof intellectual property rights protection. This limits the opportunitiesavailable to our firms from learning-by-doing and from exploiting economiesof scale in researching and developing new product lines. The low-coststrategy would have been sustainable over time if it had been based oncontinuous process innovations which reduce costs, rather than on a singulardependence on access to a large army of technically skilled labour force. Ifthere were to be bottlenecks in the supply of skilled labour, we might findmany of our advantages being eroded away rather quickly. And this, if recentstudies are any indicator, may be exactly what is happening.

On average, around 30 per cent of our new software engineers end up going toAmerica. In the recent past, we have found that that numbers left behind arefast becoming insufficient to meet the needs of the industry. This isleading to a situation where salaries in the sector are being rapidly bidupwards. This hurts us badly in an industry that relies on cheaper skilledlabour to compete with others. The only way firms could maintain revenuesper worker employed in a labour-intensive industry such as this might be tomove into a higher value-added sector in the industry. The other alternativeof course, is to suffer the consequent losses. So unless our software firmsstart moving rapidly into more profitable segments of the industry, thevaluations put on IT stocks on Dalal Street might begin to look decidedlyoptimistic.

This is absolutely not a suggestion that software does not deserve all theattention it gets now. It seems more than likely that software as anindustry, is likely to outpace general industrial growth for a good manyyears to come. But when people start pricing companies (again, we are nottalking about any specific companies) at a 100 times current earnings, justbecause it happens to be in the software sector, I hope they know somethingI do not.