Optimal Fracturing

DOES size matter? This is an age-old question, for countries no less than companies. I remember the United Nations used to pass resolutions urging special concessions for small, island and land-locked countries in the 1970s and 1980, to help out the poor chaps.

Then TN Srinivasan came out with a devastating paper showing that, if anything, such countries seemed to have an advantage. The richest developing countries are the tiny island states of Singapore and Hong Kong.

The fastest growing countries in Africa are Botswana (which is land-locked) and Mauritius (which is an island state). The richest country by far in South Asia is the tiny island state of Maldives. But, large states like China can become miracle economies too.

Look at corporations. Socialists have long dreaded large corporations and urged curbs on them. Yet only one of the 30 companies in the original Dow Jones index survives today. Microsoft and Wipro grew from tiny outfits to among the biggest in the US and India respectively.

So size is irrelevant. What matters is how a country or corporation is organised. In one of the greatest books of the last decade, Guns Germs and Steel, Jared Diamond proposes an optimal fragmentation principle. Companies and countries should be neither too centralised nor decentralised.

To illustrate, consider China, which for most of history was the global leader. Yet, after the 1400s, China lost ground and Europe took over. China had an industrial revolution long before Europe. It led the world in inventions like cast iron, gunpowder, printing and the compass. Why did China lose this lead to late-starting Europe?

Because of over-centralisation, says Diamond. By 1400, China had the biggest and best ships in the world. But suddenly it lapsed into isolationism.

A court faction persuaded the Chinese Emperor that ship-building and naval spending were wasteful and should be stopped. Other similar imperial decisions stopped the development of clocks and water-driven machinery. The centralised power of the emperor could stop all innovative ventures by fiat.

By contrast, Europe was broken into dozens of competing states. Christopher Columbus, an Italian, left his country to seek foreign support for charting a new route to India. Nine European princes refused him, but the tenth — the Spanish King — took the risk. No Chinese equivalent of Columbus could have succeeded, because that region had only one ruler, not several.

Competition between European states was a prime reason why technology and commerce prospered in the region. Whenever one ruler tried to shut out innovations — and there were many such — this merely gave an opportunity to other states to take advantage of the innovations.

However, over-fracturing can be just as bad as over-centralisation. Consider India, which was a leading scientific power in ancient days. Whereas China was centralised under one emperor, India developed thousands of warring kingdoms. The Mughal Empire brought some limited stability, but never remotely approached the comprehensive power of Chinese emperors. So, over-fractured India slipped behind Europe, no less than over-centralised China.

Jared Diamond gives similar examples in the corporate sector. The Japanese food processing industry is fragmented into small local monopolies reinforced by government regulations. So, the productivity of this industry is one third that of its US counterpart.

By contrast, Japan’s steel, automobile and electronics industries are large companies with internal decentralisation mechanisms like quality circles, and combine economies of scale with the flexibility and multiplicity of ideas yielded by small quality circles. Japan’s productivity exceeds America’s in all these industries.

In sum, the best form of organisation avoids both excessive and insufficient centralisation. You should break up a large entity into smaller groups that have flexibility, yet communicate freely. Unity within diversity means that people, materials and machinery can go unhindered to that part of a country or organisation where they are most productive.

What are the lessons for India? First, that Nehruvian planning was a predictable disaster. It discouraged entrepreneurship and innovation, and forced all to conform to one model. No business equivalent of Columbus could have progressed in the licence-permit raj. But when the controls were lifted after 1991, the software industry came out of nowhere and soared.

Narayana Murthy of Infosys says he struggled in the 1980s because it took one year to get a phone connection, two years to get an import licence for a computer, 15 days to get currency for each foreign trip. Liberalisation removed the straitjacket, and Infosys took off.

Indian businessmen could experiment with all sorts of ideas planners had never thought of, and that enabled the software industry to become a stunning, unanticipated success.

The second lesson is that some of the biggest gains from liberalisation will come from competition between Indian states. The old planning model made it impossible for any state to strike out on its own. Today different states are going different ways. Gujarat has a port-based model. Karnataka and Andhra Pradesh emphasise computer software. Madhya Pradesh emphasises decentralisation. All these experiments have yielded gains.

The third lesson is that in some ways the Indian market is still excessively fractured. We do not have free internal movement of goods. Hundreds of restrictions continue on the movement of agricultural items. States impose entry taxes, which are equivalent to import duties, and give huge sales tax concessions valid only in their territory. So the price of Indian goods varies enormously from state to state.

We need to learn from the USA, which prohibits hurdles to inter-state movement even while giving its 50 states great autonomy in other matters. This give the US 50 vibrant experimental laboratories within one unified system. We too should aim for that.

What do you think?