India Transformed

by Rakesh Mohan

  Second, within industrial sectors, both South Korea and China have invested more in R & D semiconductors, especially in South Korea. This reflects a deepening of technical capability within sectors. This shows particularly in the growth of R & D spending at a few giant R & D spending firms, as Table 6 below shows. If we consider the top ten R & D spending firms in South Korea, China and India, the emergence of firms such as Samsung (at $15.3 billion, close to India’s total investment as a country in R & D) and Huawei (at $6.6 billion higher than India’s total industrial investment in R & D) illustrates the impact of a few large firms. It brings us back to our earlier point of R & D being highly concentrated; a few giant firms invest giant amounts in R & D. South Korea and China have seen their emergence as they have deepened their technical capability; India still needs to. Some of this gap is to India’s entire advantage: the single big cost in R & D is people. Costs in India are still between one-fourth and one-half that of an equivalent engineer or scientist in South Korea or China, presenting an opportunity I will return to later.

  3.1. The Trade Regime

  A separate chapter in this book covers our trade policy and export achievements; I provide only a few brief comments on the impact on technical capability. A critical component of the success of South Korea (and also Taiwan, Singapore, and China) in moving up the value chain and deepening investments in technical capability was their export orientation. Domestic protection was often combined with export promotion—indeed exports were often required as a condition for continued domestic protection. This had a dual benefit: first, export success forced firms to be competitive, and this had efficiency benefits for domestic production too. But there was also a direct impact on technical capability. Exporting provided for a substantial flow of technology from demanding buyers. Studies of how South Korea and Taiwan built their technical capacity identified technical flows from overseas buyers as the most important source of technology for firms, ahead of in-house R & D, technology licensing and domestic R & D institutes. This reflected the industries involved—textiles and garments, consumer goods, and OEM suppliers of appliances and electronics assembly. None of these required substantial investment in R & D. As the industrial structure changed towards automobiles, semiconductors and IT hardware, greater investment in in-house R & D supplemented, and eventually replaced, these flows of technology from buyers.

  Anecdotal comments from the two industries in India that are strongly export oriented—software services and pharmaceuticals—says technical flows from demanding buyers have played a similarly important role. But the absence of export orientation for most of Indian industry has deprived it of this flow of technology.

  3.2. Imports and Exports of Technology and In-house R & D

  Tables 7A and 7B show the growth of technology imports into India compared to other countries. Note that until the 1990s, import of technology into India was severely restricted. In recent years ($5 billion in 2014), payments for technology are comparable to total in-house investment in R & D ($6 billion in 2014). Exports of technology—in the form of contract research—have also been growing, if not as strongly. We often have a mental construct of technology import as bad, and exports as good. This is wrong. Growing technology imports together with growing in-house R & D reflects strong investment in technology by Indian firms. In the same way, I have long viewed the export of technology as benefiting the buyer more than the producer. In short, technology is not the product: firms and countries get rich by turning technology into products and services, not by selling technology.18 As Table 7B shows, the most successful periods of rapid industrialization across countries—Japan in the 1950s and 1960s, South Korea and Taiwan in the 1970s and 1980s, China since 1990—have been accompanied by significant imports of technology, considerably higher levels than in India until the noughties. Much innovation happens without recourse to formal R & D. R & D started to contribute significantly to Korean and Taiwanese industrialization only in the 1980s, and to China’s only in the 2010s. As we saw earlier, industrial development must precede the choice of investing in R & D.

  Table 7A: Charges for the Use of Intellectual Property (2014)

  Country Name

  Payments

  ($ million)

  Receipts

  ($ million)

  Deficit/Surplus

  ($ million)

  United States

  42,124

  1,30,361

  88,237

  China

  22,614

  676

  –21,937

  Japan

  20,942

  37,336

  16,395

  United Kingdom

  11,225

  19,826

  8601

  Korea

  10,546

  5167

  –5379

  Germany

  9311

  14,993

  5681

  Russia

  8021

  666

  –7356

  Brazil

  5923

  375

  –5548

  India

  4849

  659

  –4190

  Source: World Development Indicators (2014), http://data.worldbank.org/.

  3.3. Searching for an Explanation: Why Does Indian Industry Not Invest More in R & D?

  The share of manufacturing in GDP: China’s manufacturing sector is today over ten times India’s ($3.7 trillion to India’s $330 billion). But Chinese firms today invest over twenty-five times what Indian firms do ($164 billion in R & D as opposed to Indian firms’ $6 billion). And the comparison extends beyond manufacturing.

  The scale and profitability of firms: Are Indian firms just too small to invest in R & D? The general understanding from the field is that there is a threshold level beyond which scale economies do not apply to R & D, and Indian firms in many sectors are today well beyond that threshold. Our ten largest pharmaceutical, IT services, chemical and engineering firms all have a turnover of $500 million or more, which most would consider to be beyond this threshold level. Are Indian firms not profitable enough to invest in R & D? Even after five years of slow industrial growth, average corporate profitability in India (a return on sales of 10 per cent) compares well with China or South Korea.

  Table 3B showed that Indian firms invested somewhat less in R & D as a percentage of sales than their global counterparts. This is true in both our most R & D-intensive sectors of pharmaceuticals and auto, where our firms invest roughly half as much as a percentage of sales as the global leaders. But this is particularly true in the case of software. Compare the software industry in China and India (see Table 8). The top ten companies in China invest 8 per cent of turnover in R & D; in India, the top ten companies invest 1 per cent in R & D. An obvious explanation is that India’s software companies are software-service firms, not product firms. But most Indian software companies are worried about how long the existing model of labor arbitrage combined with excellence in project execution can continue to drive growth. No one would consider TCS, Infosys or Wipro to be either small or unprofitable. They just invest little in R & D.

  4.   From Half-formation to Transformation: What Should Our Priorities for Reform Be?

  The purpose of an NIS is to build competitiveness in the long run. As wages rise, as natural resources are consumed, as the easier catch-up options are exhausted, it is the NIS that should enable an economy to keep growing over decades. This is particularly vital for India. It is entirely possible for an economy focused on labor-intensive manufacturing— textiles and apparel, footwear, food processing—to grow rapidly for years, as long as wages remain low; that is the source of comparative advantage. The East Asian NICs did so for years until rising wages forced a move to higher value-added sectors, and with it, forced investment in innovation. But as Table 5A shows, Indian manufacturing is not concentrated in these labor-intensive sectors: for various historical reasons going back to the 1950s,
and unchanged even through the reforms of the 1990s, Indian manufacturing has not focused on labor-intensive manufacturing. Kochhar and her impressive roster of co-authors made just this point:

  … it is striking that India’s share in skill-intensive manufacturing, which was already high in 1980 despite its lower level of per capita income, has been increasing and is at levels reached by Malaysia or Korea at much higher levels of per capita income. There is also a striking contrast with China. China’s share of output in skill-intensive industries is lower than India’s and has been virtually flat whereas India’s level has been higher and rising. The move towards skill-intensive goods is also reflected in India’s exports: the share of exports of skill-intensive goods has risen sharply from about 25 percent in 1970 to about 65 percent in 2004.19

  In Section 3 above, I have argued that India must transform its NIS in the next twenty years, just as Korea did between 1970 and 1990, and China did over the last twenty years. It is instructive that near the beginning of this transformation, India already has a lower share of manufacturing coming from these labor-intensive sectors of textiles, apparel and food processing (19 per cent in 2010) than Korea (25 per cent in 1990) or China (22 per cent in 2010) near the end of theirs. In other words, for various reasons (hopefully explained elsewhere in this book20) we have for long had a manufacturing structure out of synchronization with our comparative advantage. We have a manufacturing structure focused on skill-intensive and capital-intensive sectors, sectors which require constant innovation, and constant and substantial investment in innovation, to be competitive over time. It is time our NIS matched our industrial structure.

  So what must we do?

  4.1. Adding Dimensions to Structural Change: An Aside and a Point

  A key driver of economic growth as countries catch up is the movement of people from lower value-added activities into higher value-added activities. Given our poor employment data,21 it is difficult to say specifically how many are employed where and how this is changing. Despite adding 10 million people annually to the workforce, there does not seem to be a drastic increase in those looking for work. In the absence of real data, simple observation says that we have been creating employment by the million quite successfully, though the bulk of the jobs created are in the service sector and support jobs. My favourite example is drivers in Delhi: Delhi accounts for a quarter of all cars sold in India, one-third of which are chauffeur-driven. That says Delhi has added over 1 million drivers in the last decade or so, but no employment data captures this. Regarding addition of jobs in the areas of hospitality, retail, delivery and construction, many Indian cities report reasonably tight labor markets: people are available for jobs, but they need to be attracted to better jobs from the ones they already have. So the problem is not jobs but good jobs. What is a good job? One which has higher productivity, and higher potential for improving value over time. This is why manufacturing matters. As Arvind Subramanian pointed out in a recent Economic Survey, ‘If the entire Indian economy were employed in registered manufacturing, India would be as rich as say South Korea.’22 Large-scale, labor-intensive manufacturing matters for precisely this reason.

  So, although an aside to an analysis of our NIS, fixing employment in India requires making labor-intensive manufacturing dramatically more attractive. A second aside is that given that most employment creation has been in services, we need to understand how to grow productivity in services over the long run. Productivity growth in manufacturing is well understood internationally, less so in services.

  Coming back to the NIS, we need to see a much stronger presence in those sectors that are the most technologically dynamic worldwide. Table 3A listed the top ten R & D-intensive sectors: pharmaceuticals, auto, technology hardware and equipment, software and services, electronic and electrical equipment, industrial engineering, chemicals, aerospace and defence, general industrials, and oil and gas. We have a small—or even non-existent—manufacturing base in sectors such as technology hardware, electronic equipment, and aerospace and defence. So our priority must be to establish a strong industrial presence in these sectors. The government’s initiative to attract investment by firms such as Foxconn and Flextronics, the world’s largest technology hardware manufacturers, is very welcome. Their growing presence will create a pull on component manufacturers to establish a local presence. As the component manufacturers in turn grow, the competitiveness of other (and smaller, higher value-adding) downstream assemblers will also be enhanced.23

  4.2. Use the Availability of Skilled People to Build a Competitive Position Based on R & D

  India has long had a distinct advantage over every other country in the availability of skilled technical people at relatively low cost. This is the source of our software industry and the growth in IT-enabled services. The IT industry has drawn on India’s massive production of 1.5 million engineers annually—compare that with the annual US production of 1,40,000 engineering (including computer science) and 1,60,000 natural science undergraduate degrees.24 At the height of the Indian software boom in 2000, when that single industry was recruiting over 1,00,000 engineers a year, it is striking that no Indian firm in any field had problems recruiting fresh engineers. They were still available in abundance, if not in quality.

  Indian industry has long had this luxury of an abundance of low–cost qualified people. But while engineers have been cheap, they have also been treated cheaply: recruited to perform jobs with undemanding technical content. It is only as economic reform has created a demand for product innovation and as engineer remuneration has risen sharply (in 2016, a good graduate engineer with five to ten years of experience would earn about five times in real terms what he or she earned in 1991) that firms expect their average engineer to do work with more demanding technical content. Few Indian firms, however, recognize the huge advantage they have in terms of low-cost qualified people, in reducing R & D cost and thus building competitive positions based on R & D. The pharmaceutical industry is a major exception: firms such as DRL, Sun Pharma and Cipla are betting on India’s lower R & D costs as a basis for competing long term in a research-intensive industry. So too are auto firms such as Tata Motors and Mahindra and Mahindra. But they are still exceptions. Multinational investment in R & D in India has been much more widespread. Eighty-three of the top 100 global R & D spenders are reported to have an R & D presence in India.25 For example, Astra Zeneca has one of their largest R & D labs outside Sweden in Bengaluru, and all their work on tropical diseases is now done there. Cummins has a design centre in Pune, which became its second largest (after the US) in three years. Emerson employs 4000 engineers at their R & D centre, also in Pune. Bosch employs 13,500 R & D engineers in India, their second-largest facility worldwide, and has announced a major expansion of their R & D presence in India. And most prominently, GE set up the Jack Welch Research Centre in Bengaluru in 2000, which has since become GE’s largest R & D facility worldwide. It employs 5300 people, two-thirds with master’s or PhD degrees, and today has the largest chemistry and chemical engineering PhD concentration in the country.

  Therefore, industry must recruit the best talent available and aim it at pushing forward product designs in their own industries.26 This talent should aim to learn from the best firms and research worldwide. As firms invest more in research, the demand for graduate engineers and PhDs will also grow. This will tie in nicely with the focus on graduate programmes and research in the IITs and progressively in other leading educational institutes.

  Our major software-service firms have a long record of success with rapid growth and high profitability sustained over decades. As we saw earlier, this success has not been accompanied by substantial investment in R & D. As the opportunity of further growth built around labor arbitrage27 decreases, the pressure to grow through innovation in both product and service will rise.

  All this will amount to an increased investment in R & D by Indian firms. The obvious powerful incentive that drives firms worldwide to inve
st in R & D is that they will otherwise be put out of business by competitors. When Tata Motors and DRL become more typical of their industries, as sectors such as software start investing a share of turnover in R & D comparable internationally, and as new R & D-intensive sectors such as technology and electronic equipment and defence grow, the investment in R & D by firms will meet the 1 per cent of GDP set as a national target.

  4.3. An Invigorated Higher-Education System and an Invigorated Scientific-Research System: Two Sides of the Same Coin

  The advantages of doing research in universities are manifold. First is the apprentice–journeyman benefit: the graduates that industry hires will come trained in doing research. Second, the industry–research linkage issue is immediately drastically reduced: every university has an automatic, costless and strong linkage with industry through students. Each time industry employs a graduate a new link is formed. Students know their professors and vice versa. Third, not just does teaching benefit from the research–teaching combination, research benefits too.

  Our end objective is clear: we should have a few really world-class research-teaching centres. The 2016 Budget announcement of developing ten public and ten private universities into world-class institutions is most welcome, but where will we get our highly qualified research faculty from? I would suggest we bifurcate CSIR and make it part ‘unexciting-but-useful Technology Assistance Institution’ and part ‘large pool of professors’. The 4000 CSIR scientists would be a huge impetus to the national higher-education system, particularly graduate education. The IITs are best off, but even they have faculty vacancies of over 20 per cent. Where are they all going to come from except through such an initiative?