The services-manufacturing debate

The pattern of India?s strong growth since the 1990s has raised somewhat of a puzzle for economists. At least since the 1950s, following the detailed empirical work of Nobel prize winner Simon Kuznets, the pattern of development has been understood to be in a sequence from agriculture to manufacturing to services. India, on the other hand, seemed to be growing rapidly (among the fastest in the world) without robust manufacturing sector growth. Observers have been divided on the nature and consequences of this phenomenon. Some suggested that India was following a new pattern of services-led industrialisation. Others (notably former chief economic adviser Shankar Acharya) cautioned against ?hype,? questioning the reality, robustness and sustainability of the apparent services boom.

A significant recent event, as well as much new research, together suggest returning to the debate on India?s pattern of development. The major news came last month, when China revised its GDP figures upwards by a whopping one-sixth, with most of that increase coming from a revaluation of services sector activity. At a stroke, the estimated contribution of services to China?s GDP went up by nine percentage points, crossing 40%. This is a much more plausible figure than previous estimates, even though still short of the average for countries with a similar per capita income levels. India, on the other hand, calculates its share of services at about half of GDP, somewhat above the average for its income class. That discrepancy has arisen essentially within the last decade. The questions for researchers are: can we believe the numbers? What are the sources of services growth? What are the consequences for India?s current and future growth?

Can we believe the numbers? Shankar Acharya has pointed out the fragility of India?s services sector numbers, suggesting some deficiencies in CSO methodologies and resulting overstatement. On the other hand, China?s example indicates that there might be biases in the other direction. To take one case, a survey of doctors in Delhi once found massive under-reporting of income to evade taxes. It is easy to imagine that much service activity of this nature has not been fully captured in the national accounts. There are other problems with India?s economic statistics, and China?s bold example should encourage India to put its own statistical house in order. But this is a general need and prescription, going beyond accounting for services activity.

What are the sources of services growth? Shankar Acharya and others have pointed out that the excitement over software and IT-enabled services is not matched by numbers. While business services grew fastest in the 1990s among service categories, this was from a tiny base. The biggest contributors to the change in the share of services in GDP were, in order, banking, wholesale and retail trade, community services and communication. However, business services were not far behind the last two in their increase, and this suggests that some of the optimism about services is not misplaced.

The last and most important question has to do with consequences. Sanjay Hansda, of the Reserve Bank of India, conducted a detailed input-output analysis (using 1993-94 data) and suggested that linkages from services to industry were strong. Taking a quite different approach, Rashmi Banga and Bishwanath Goldar, in an Icrier working paper, estimated manufacturing production functions for data from 1980-81 to 1997-98, and found that services inputs made a significant contribution to manufac- turing output. Neither of these studies posits a growth mechanism, but they support the intuition that services growth and innovation can spill over positively to manufacturing, just as China?s national accounts revision accords with the idea that its manufacturing prowess has also fueled domestic services growth. In fact, linkages work both ways: automobiles need servicing, while call centres need computers.

There are other reasons to be optimistic about services. Several analysts have highlighted the importance of the nature of services growth. In particular, to the extent that parts of the services sector are becoming more like industry, with the use of modern technology, routinisation and the reaping of economies of scale, services-led growth becomes a more plausible model of industrialisation. Further- more, the tradability of some of these kinds of services has increased, on the back of modern communications technology?thus, services are making a well-known and strong contribution to India?s balance of payments. Finally, the development of more efficient corporate organisational cultures and institutions, which arose in India?s software services industry, has had a salutary effect on how Indian industry runs. Interestingly, if one goes back in time before the industrial revolution, it was a merchant and finance (i.e., services) revolution that got the whole process going, as Sir John Hicks argued in his Theory of Economic History.

For policymakers, the lessons ought to be clear. Rather than targeting manufacturing in isolation, policy should focus on areas where the government can and should make a difference. Software got going because it was not strangled by government, but at the same time it benefited from government investments in higher education. Allowing both manufacturing and services to flourish by reforming misdirected regulation and removing constraints in infrastructure and human capital is the way forward for job-friendly growth. Given the parlous state of public finances and the poor efficiency of public delivery, more private sector involvement in these core areas should be encouraged. Ultimately, making the conditions right for enterprise and innovation throughout the economy matters more than the services-manufacturing debate.

The writer is professor of economics, University of California, Santa Cruz