Services Exports, Not Just Software Exports

In the short space of three years, the software industry has come to symbolise the hopes of rapid economic growth in India. I can recall the scenario as of Jan/1997, and how different things seemed to be at the time. We have seen remarkable changes in these three years in three respects :

Gains from Trade. It gives me great pleasure, as a card carrying economist, to say that we knew this all along. The reasoning that is working here is part of the essential intuition of modern economics. This reasoning has important implications well beyond the software sector.

The simple logic that is at work in software is the way this industry harnesses India's strengths (cheap labour) while avoiding India's weakness (scarce, expensive capital). Software versus hardware is an ideal illustration of the benefits of international trade. In principle, India can produce hardware just as we are able to produce software. However, we would suffer a high opportunity cost making hardware; the labour and capital that is used to make hardware would generate a higher output if it was put into producing software instead. Hence, the efficient path is to produce software, export it, and use the proceeds to buy hardware. This violates "self-reliance", but it maximises the revenues of our labour and capital.

The slogan "computer chips, not potato chips" sounds clever, but it is exactly wrong. There is much economic sense to producing potato chips (a labour-intensive product which harnesses India's strength in agriculture) when compared with computer chips (a highly capital intensive area).

Where are the engineers? The software industry is considered an "ideal" situation for India since the main "raw material" is labour, which is cheap in India. However, the achilles heel in the growth of the industry is the shortage of high quality labour. It would not be an exaggeration to say that the quality of software production in India today is amongst the poorest in the world. I have met numerous software professionals in India today, but I can count the well educated ones on my fingertips. The output of serious universities in India in this field is pitifully small, and most of the labour that is in the industry are technicians who have learned by tinkering with manuals as opposed to learning from books. There is remarkably little that exists by way of a conceptual understanding, of a deeper vision about how things work, and an ability to design and plan complex systems in an efficient way.

The rise of private sector education (e.g. NIIT, Aptech, etc.) has generated a vast increase in the supply of educational services, and these firms have undoubtedly made the growth of the software industry possible at a time when the public sector universities were simply unable to swiftly reorient themselves. However this education is oriented towards the lowest grade of blue-collar software workers. The private sector educational initiatives teach an apallingly narrow-minded use of packages, which is very different from a knowledge of computer science and engineering.

I believe that this starvation of high quality engineers will prove to be a serious bottleneck. It serves to block the efforts of Indian software firms to walk up the "ladder of quality" into higher value added. The revenues per man-month that are earned by Indian software companies are roughly the lowest in the world. In the absence of a way to migrate our great mass of technicians into engineers, this is hard to change.

Services exports, not software exports. To most people who think about software in India today, the future is a happy march forward up the ladder of quality, where Indian software firms rise from billing rates like $500/manmonth to $6,000/manmonth, and the software industry yields tens of billions of dollars of software exports. If the bottleneck in converting from technicians to engineers is a serious constraint -- as I believe it is -- then this scenario is unlikely to come about.

We have to face the fact that higher education of note is restricted to five to eight universities. In this case, our competitive advantage reduces to a great mass of unskilled labour. What labour-intensive exports can we obtain out of this? The answers are fascinating, and often well-removed from computer software:

  1. Medical transcription: Pools of secretaries working in India transcribe notes for doctors located anywhere in the world.
  2. Call centers: Calls from all over the world are routed to a location in India where hundreds of telephones are manned by cheap Indian telephone operators.
  3. Back office and accounting: The accounting and paperwork for the global operations of a corporation are handled from an office in India.

For an extended treatment of this reasoning, see The second generation of reforms, by Kirit Parikh and Ajay Shah, in India Development Report : 1999-2000. The strength of each of these areas, from an Indian perspective, is that for a firm to succeed, labour with a high quality education is not required. The functions in each of these areas can be met by a person out of high school with a smattering of job-training (e.g. the labour which is going into our software industry today).

In conclusion, there is a great future for services exports for India, in firms which are able to organise large teams of low skill labour for export purposes. Software is the first of these, but similar opportunities lie in many other areas. In areas where high quality staff is required, and where a high quality university education is called for, India has a distinct disadvantage owing to the existence of only five to eight good universities. However, the opportunities in low-tech services exports are extremely large when compared with GDP. We could see hundreds of thousands of workers moving into jobs which have low wages (by world standards), yielding tens of billions of dollars of export revenues.

Back up to Ajay Shah's media page