A recent Reuters report indicates that China is preparing to spend $1.3 trillion to promote key technology. This money will be directed to state-owned enterprises. Is this a wise move? Should India emulate it? History provides two major models by way of exemplars. The first may be called the MITI model and the second the NSF model.

MITI or the Ministry for International Trade and Industry was created by Japan in 1946 to be the hub of industrial policy, allocation of foreign exchange (until 1971, Japan had fixed exchange rates and foreign exchange was rationed), guidance on investments and competition and international trade. So powerful was it that right up to the 1980s, no politician could aspire to be the prime minister without first pulling a stint at MITI. There was also a very close nexus between MITI and the corporate sector. In 2001, MITI morphed into METI or Ministry of Economy,

Trade and Industry. So what has been their record?

One of MITI?s early directives was to Sony, in the 1950s, to stay away from the transistor industry. It decided that since Sony had no background in the newly emerging sector, it would be a waste of precious foreign exchange for it to license the technology to make transistors from Western Electric. It took two years for Sony to persuade MITI to let them produce transistors. Sony used this technology to achieve global prominence by producing cheaper, smaller and more durable transistor radios as compared to the valve radios then in vogue.

Another directive was to the nascent Japanese automobile sector to stay away from the export market, and instead consolidate the 10 firms into two and produce a domestic ?people?s car?. Wisely, the Japanese automobile sector resisted every one of these requests. The various automobile firms maintained their separate identities and prospered by expanding rapidly overseas.

Subsequently, Japan tried to follow a similar strategy for supercomputers and space exploration, with similar results. South Korea and the other Asian Tigers that looked to Japan for inspiration fared no better. Korea poured in huge amounts of resources in getting their shipbuilding sector off the ground. While it is the second largest in the world, it is hardly competitive and requires frequent bailouts.

However, anecdotes are never enough. Perhaps the largest and most rigorous study of MITI was published in 1996 by two American economists, Beason and Weinstein of Alberta and Harvard, respectively. They found that the sectors favoured by MITI (mining, textiles and rice production) had among the lowest growth rates of all industries. Further, productivity growth in the supported industries was also the lowest.

So, if the MITI model does not work, what is the alternative? The alternative is the National Science Foundation (NSF) model. NSF was created by the US government in 1950. In practice, it works very closely with its allied institutions, the National Institutes of Health and the Defence Advanced Projects Agency or DARPA. (Incidentally, NSF is currently headed by an Indian American, Subra Suresh). So, how does NSF operate? Its working may be illustrated by how a world-changing technology was incubated.

In 1993, there were fewer than 100 Websites. But their numbers started to grow exponentially. The growth was so rapid that Yahoo!, Lycos and Altavista were created to index the Web. Yet, as anyone who used the technology at that time would remember, they were anything but user-friendly. When confronted with a search term, they searched primarily on the basis on how often a Website used that term. The results were quantity heavy but light on quality, largely because companies figured out that they could game the system by adding dummy words to their Websites. As a result, usable results were few and far between. It was not uncommon for individuals to have to plough through hundreds of links before finding one that was truly useful. It was as if a vast library had been created but the index to it was largely a hit and miss affair.

To address the issue, NSF created the Digital Library Initiative and requested proposals from all US-based research institutions to come up with a better index to the Internet. It funded six proposals to the tune of $4,516,573. One of them was to three professors at Stanford?Terry Winograd, Hector Garcia-Molina and Rajiv Motwani. Along with their two graduate students, Larry Page and Sergey Brin, they came up with an innovative algorithm called the Page Rank Algorithm (also a play on the name of the principal inventor) that used principles of citation developed in library science but with several key mathematical twists to deal with large numbers of pages. Later on, Page and Brin founded Google based on this technology. Stanford obtained the patent on the technology. Since the research funding was assigned to it and the researchers did the pioneering work there, the technology was owned by Stanford University, which also owned the patent that was filed on the technology. Google licensed the technology back from Stanford in exchange for 1.8 million shares, which were sold by Stanford for $336 million in 2005. Google today is worth over $150 billion and is one of the most valuable and recognised brands on the planet.

The final leg of the story requires elaboration. How was it that Stanford could own the patent on the government-funded research? This is where an arcane piece of legislation passed in the US in 1980, piloted by two prominent US senators, Evan Bayh and Bob Dole, came in. Until then, there was little incentive for any company to develop technology funded by government research since they did not own the title to the intellectual property, which lay with the government. The $30 billion that the US government spent on scientific research in 1979, for example, resulted in precious little by way of commercially useful technology. That changed when the Bayh-Dole Act became law. Under it, non-profit entities (such as universities or federally-funded research labs) automatically got rights to all Intellectual Property Rights (IPR) created by research routed through them. Almost overnight, all major universities sprouted Offices of Technology Commercialisation or OTCs that aggressively commercialised their IPR portfolio as Stanford demonstrated in the case of Google.

Today, it is almost impossible to come across a new and innovative technology in pharmaceuticals, healthcare, computing, nanotechnology, genetics, alternative energy or indeed any other sector where the US government funding of basic research has no role to play. Yet the US government has wisely refrained from picking winners and losers among firms in these areas.

What are the lessons for India and China? It is clear that if even highly homogeneous and collectivist but relatively open societies such as Japan and Korea have not been very successful in picking winners among firms, it is very unlikely that a collectivist but closed society such as China would be more successful. And certainly, a highly diverse and argumentative society such as India is likely to find such an exercise as going deeply against its grain. Combine that with the opportunities for corruption inherent to this exercise and it becomes clear that India should stay away from any attempts at emulating MITI or the Chinese.

Clearly, a decentralised model such as that of the NSF and Bayh-Dole would be relevant for India. Yet the abilities of the Indian equivalents of NIH and NSF?the Indian Council of Medical Research and the Department of Science and Technology?leave much to be desired. Recently, their combined budgets have been in the R100 crore range. Further, mechanisms for the ownership of the IPR resulting from this funding are not in place. The equivalents of a Bayh-Dole Act and setting up OTCs would go a long way towards filling the research-based innovation vacuum in India.

The author is academic director of Duke Corporate Education and visiting professor of business policy at IIM-A