Research at a distance
Multinational firms have not yet farmed out all their research and development to China and India but they are taking a much more global approach to this high-value activity.
Research and development has gone global. Two-thirds of companies already offshore at least some of their R&D, says the Economist Intelligence Unit (EIU). In three years' time that proportion is expected to rise to 84 per cent. In 1994 the Asian developing countries held a 3 per cent share of overseas R&D investment by American multinationals, says Torbjörn Fredriksson, senior economist for the UN trade and development agency Unctad. By 2002 that had risen to 10 per cent. "The increase was particularly noticeable for China, Singapore, Hong Kong and Malaysia," he observes. The examples of globalised R&D among US firms stack up. General Motors has about 1,000 'technical' and 500 process-engineering employees at its R&D centre in Brazil. Microsoft runs R&D centres in Shanghai and Beijing. China is its fifth-largest centre of research investment. General Electric has invested $80m in its 1,600-employee John F Welsh technology centre in India. Computer chip giant Intel, meanwhile, has conducted R&D in India and China since the 1980s, says the EIU, and its "major research centres" in both countries "are now responsible for leading global projects". Intel also conducts R&D activities in Ireland, Russia, Malaysia and the Philippines. Among non-US multinationals, Nokia employs over 500 workers in centres in Beijing, Shanghai and Hangzhou, China. AstraZeneca's 100-employee Bangalore research centre is working on treatments for TB and other diseases. Dutch firm ST Microelectronics designs chips in Rabat, Morocco. Siemens is investing $119m in South Korea over five years. Philips's largest non-European R&D centre, with 1,000 engineers, is in Singapore. And Toyota's R&D centre in Thailand "has a global mandate", says Unctad. So, why the migration? The EIU says that greater efficiency, speed to market and access to talent are key drivers. But other surveys point also to the lower costs of talent, materials and facilities. "The fact that some emerging economies are now perceived as attractive locations even for highly complex R&D [such as chip design] opens the door for the transfer not only of technology created elsewhere, but also of the technology creation process itself," Fredriksson says. IT has made dispersal of such knowledge easier, says Professor Marek Szwejczewski, senior research fellow in operations management at Cranfield School of Management. "It allows companies to share information much more easily, so co-location isn't so necessary or helpful." "IT helps a lot," agrees Dr Trevor Cross, research director of electronic components maker e2v. The company runs development work for its Californian operation in the UK, says Cross: "It doesn't matter where those people sit." IT allows companies to pass the development baton round the globe - eight hours in the UK, eight in Singapore, eight in the US - so that someone is working on a project round the clock. A quarter of corporations now do this for important projects, according to one survey. Moreover, some offshore markets are now attractive as R&D centres in their own right. Says Szwejczewski: "You need to put R&D where the biggest market is." Being in a market gives firms a better understanding of working requirements, so you do development there as well as make the goods. Consultant Dr Colin Mynott points to Nissan, with whom he works: "Nissan feels it's essential to have a design and development centre in the culture of the market. They opened a design studio in London to deal with European matters, one in Korea, because it's significantly different from Japan, and they've had one in America very successfully for some years." Most of the products being designed and developed in China and other low-cost economies are destined for western markets, says Dr Tim Moore, head of the Hong Kong office of UK contract-research company Sagentia. However, to succeed in local markets, the multinationals will have to design products there too, he argues. The cultural divides are too big, and the language differences too great, for the R&D to be done outside of local markets.
'R' versus 'D'
However, research is not necessarily the same as development. Professor Rajneesh Narula of Reading University business school draws careful distinctions between 'R' and 'D'. Development, says Narula, is inevitably closer to the market. "It's about local modifications and minor issues - what colour is a chair or a mobile phone? Research, on the other hand, is about big ideas - what replaces the automobile in 20 years? It's about fundamental changes." Development is following manufacturing offshore, says Moore: "I can't see that staying here [in the UK]. Design and development, electronic and mechanical creation, needs to be close to the manufacturing. It's just the back end to that. I have seen mobile phone development companies in Shenzhen with 400 developers creating a new phone every six months." Bob John, chief executive of TWI, formerly the Welding Institute, says it's important for the UK not to export all its product development. But global companies will put 3D modelling and testing where the cost is lowest, "and we have lost most of that". He adds: "Most of the manufacturing R&D that can migrate has migrated." What's left, however, is well worth having. John's example is Rolls-Royce and companies like it, which have "very high design intensity and high-value products". Pharmaceutical and biotechnology companies provide further examples. The outsourcing of R&D by multinationals can vary according to their different product strategies. Computer and phone makers use external manufacturers and developers for non-strategic products, but keep strategic product development inhouse. Motorola, for example, keeps development of its high-end Razr mobile in house. Hewlett-Packard's approach is to have most of its laptops and other equipment designed by external developers. Electronics industry 'board stuffers' have transmuted into electronic manufacturing services (EMS) companies and moved up the value chain into design. Phone giants like Motorola and Nokia used to go to EMSs with drawings. Now market leading Indian EMS providers such as Wipro, Infosys, Satyam and HCL are providing their own low-cost phone 'platform' designs for these giants.
Competition for R&D talent has forced rivals to form strategic technology partnerships (STPs). Dr Michele Cincera of the Brussels University Department of Applied Economics points to the way Mercedes, BMW and General Motors jointly developed hybrid drives to catch up with Toyota. Daimler, Volkswagen and Renault are collaborating with fuel companies Sasol Chevron and Royal Dutch Shell on synthetic fuels. Texas Instruments has about 1,000 partnerships, says Narula: "They're hedging their bets." STPs mean they don't have to buy a laboratory. But they do have to decide how much technology know-how they want to risk sharing. In other words, they need to cling on to their intellectual property (IP). "I don't foresee e2v, with its niche products in aerospace and electronics, outsourcing IC [integrated circuit] design or sensor design," says Cross. "If you go outside to work with less-established economies like China, you need to control and protect your intellectual property. China has come a little way on that but I'm always very very cautious."
Legal contracts between STP partners are a valuable IP protection, says Narula: "Patenting doesn't really work. Thirty per cent of all innovations are patented, the rest rely on secrecy. The advantage of an STP is that you are sharing information within the legal framework of a document. You keep your friends close, and your enemies closer still."
Szwejczewski compares the collaborative development trend to the computer industry's 'open systems' model. "The outstanding 'open innovation' example is Procter & Gamble [the US household goods maker]. The argument is that you won't have all the knowledge to develop your business, so you use outside agencies, universities and even other companies and involve them in innovation. "The 'parent' acts as a 'driver'. Instead of keeping control of development, you draw a boundary round it. That makes you comfortable with using external resources. You keep the blue-sky research internal; for the other stuff, you work with other people."
In 1993, says Narula, 2 per cent of STPs involved Chinese companies. By 2005 that proportion had quadrupled to 8 per cent. The proportion of companies with a US partnership is between 40 and 50 per cent, he estimates, mostly in high tech. But Narula also warns that much of what we read about Chinese and Indian expertise is "hype". He says his brother-in-law, who runs a business in India, believes only 10,000 of India's 125,000 graduates a year have degrees worth the paper they are written on. India doesn't have a good infrastructure. Setting on one side the difficulties with supplies of electrical power, water and telecommunications, many of its graduates have never seen a laboratory. The quality of offshore research may reflect this. The journal Science reported one investigation which showed that, of 249 R&D-intensive companies headquartered in the United States and western Europe, 50 per cent of R&D effort in developed-economy sites is for new science, compared with 22 per cent in emerging-economy sites. Cincera says the early adopters of outsourced R&D overdid it and found themselves in an organisational nightmare of duplication and inefficiency. Though R&D "remains globally dispersed", he says, "multinationals are now reintegrating their R&D activities in a more hierarchical system to improve their corporate control." Szwejczewski detects a move back to centralised R&D: "Companies that used to have five R&D centres round the world may now want to reduce to two or three, one each in the US, Asia and Europe."
So, just how far can the Chinese travel from development towards research? "Three years ago people thought, 'these guys are beginning to do fundamental technological development'. But it's still in its infancy in China," says Moore. The same is true of the other 'bric' countries - Brazil, Russia and India. There's no need for western R&D workers to despair just yet, then. While US corporations plan huge investments in India and China, Europe is their third-favourite destination. More than half plan to expand their R&D in either eastern or western Europe. And in the UK, while industrial production is declining, R&D has risen 9 per cent between 1994 and 2006, according to engineering employers' body the EEF. Over 70 per cent of companies have research in the UK, and 80 per cent have design and development here. The picture in five years will be much the same, EEF research shows. In addition, the UK's tax regime is now benign, says TWI's Bob John. Recent changes to tax credits, allowing R&D-spending companies to claim exemption from corporation tax, are a sign, he says, "of genuine attempts by the UK to bring the regime into line with other countries". Cross adds: "The UK is a brilliant place to do R&D. We have world-leading universities." It's important for the government to continue to help, he says, as it has with tax credits and consultation about innovation. Fredriksson advises governments to keep their nerve: "They should consider ways to improve the performance of their innovation systems. This may involve shifting into higher-level R&D activities, fostering world-class centres of excellence and helping displaced research staff to find new employment."