INDIA’S services revolution has dazzled businesses in the rich world, turning Indian companies into global competitors and backwater cities such as Hyderabad into affluent, sophisticated technology centres. Yet economists have been less star-struck, clinging to the received wisdom that has prevailed since the industrial revolution: modernisation runs from agriculture through manufacturing and only later to services. Now some have broken ranks.
The logic supporting the conventional path towards an advanced economy is straightforward. Development typically involves moving workers from low-productivity activities such as subsistence farming to high-productivity sectors. That points to a shift into manufacturing because it lends itself to specialisation and economies of scale, both essential for rising output per worker. As first Japan, then Taiwan and South Korea, and now China have demonstrated, manufacturing can also accelerate development because its output can be exported to rich countries.
Services, in contrast, appear to be a graveyard for productivity. Because a haircut or a restaurant meal has to be delivered in person, there is almost no potential to exploit economies of scale and to export. People consume more services not when technological advance lowers their price but when they have reached a level of affluence that satisfies most of their other needs. Indeed William Baumol famously argued in the 1960s that as countries grew richer and their citizens became keener on buying services, their productivity growth would inevitably slow.
That conventional wisdom is now under fire, in a book edited by Ejaz Ghani of the World Bank and a related article he wrote with Homi Kharas of the Brookings Institution and Arti Grover also of the World Bank on the VoxEU website. The authors argue that technology and outsourcing are enabling services to overcome their former handicaps. Traditional services such as trade, hotels, restaurants and public administration remain largely bound by the old constraints. But modern services, such as software development, call centres and outsourced business processes (from insurance claims to transcribing medical records), use skilled workers, exploit economies of scale and can be exported. In other words, they are just like manufacturing. If that is the case, then poor countries should be able to go straight from agriculture to services, leapfrogging manufacturing.
And that is precisely what seems to be happening. India may be the most prominent example but it is far from being the only pathfinder. Pakistan, Sri Lanka and Nepal have imitated India, albeit less spectacularly. In poor countries as a whole, services have contributed more to growth since 1980 than has industry. Productivity growth in services has also outpaced that of industry in India, Pakistan and Sri Lanka. In all three, the level of productivity (measured at purchasing-power parities) is higher in services than in industry. In Nepal, productivity is three times higher in services. The opposite pattern prevails in East Asia. As Mr Ghani writes, “South Asia resembles the growth patterns of Ireland and Norway, rather than that of China and Malaysia.”
Underlining their role as an engine for, rather than product of, development, exports have swelled from roughly 6% of services output in poor countries in 1985 to almost 10% in 2005. Burundi, Swaziland and Rwanda have all recorded growth of more than 25% a year in services exports between 1995 and 2008. Kenya exports professional services such as accounting to its neighbours.
Services offer several advantages over manufacturing. They can more readily employ women and are less likely to despoil the environment. Located in big cities, they accelerate urbanisation. Modern services are arguably less vulnerable to protectionism than either traditional services, such as lawyers, or goods, both of which require physical entry to the foreign market.
Services, however, may not be the answer for all countries. South Asia benefited from a good deal of luck. India’s leading software exporters were founded by engineers educated in America who had returned home. The prevalence of English speakers helps to sell services in America. Many other developing countries lack these advantages.
Most problematic of all, modern services require skilled workers, not the unskilled type that poor countries have in abundance. In South Asia, service workers typically have one to three more years of education than industry workers. In modern services, school grades or a university degree are often necessary. The flip side of their high productivity is that modern services employ relatively few people. Just 2m of India’s population of 1.2 billion work in information technology; in the rest of South Asia, only 100,000 do. That is one reason why India is still keen to promote manufacturing, which is also booming.
Indeed, for many countries, the success of services is an indictment of their failure in manufacturing. In India and Sri Lanka, restrictive labour laws have hamstrung the emergence of a more competitive manufacturing base. In contrast India helped its information-technology sector by declaring it an essential industry and lifting the prohibition on operating around the clock in some states. In South Asia services have benefited from investment in telecoms infrastructure, as measured by the number of phone lines and personal computers per 100 people, whereas manufacturing is held back by a shortage of paved roads.
This suggests that for countries that avoid those problems the conventional wisdom is still right: manufacturing holds the most promise for millions of reasonably well-paying jobs. For those not so lucky, at least there’s an alternative.