Challenges in Reshaping the Sectoral Innovation System of the Chinese Automobile Industry

Before China’s national developmentstrategy turned to indigenous innovation, its original sectoral system of automobiles was mainly supported by two institutional pillars. The first pillar was strict regulation of entry permission. Through collaboration among the National Development and Reform Commission (NDRC); Ministry of Industry and Information Technology (MIIT); Ministry of Environmental Protection; General Administration of Quality Supervision, Inspection and Quarantine; and other ministries, the government has maintained substantive control over entry into the car-manufacturing sector. The core means is a regulative system based on the “product catalog management,” according to which all car products formally launched into the Chinese domestic market for sale must have prior permission from the MIIT to be listed in a regularly updated product catalog. This policy is nominally based on product safety (e.g., collision safety), environmental protection (such as mandatory emission regulations), quality qualification (such as production consistency), and other conditions. Though some critics regard the “catalog-based regulation” a rigid legacy of the planned system, it continues even after the transition to indigenous innovation. At the same time, since the state plays a critical role in resource allocation in China, especially given its influence on the financial sector, local car producers seldom challenge the fundamental regulatory arrangements.

The second pillar was the national strategy of TMFT since the 1980s, which was originally developed to promote the expansion of local industrial capacity. This strategy literally involves the trading of market access for technology with the multinationals possessing advanced technology, as acquisition of intangible IP assets on technologies would be less productive without the benefit of associated operational learning and expertise. So, TMFT in practice encourages JVs between the backbone SOEs and multinationals. The approach began with a JV between Beijing Auto and American Motors Corporation (AMC) in 1984.¹³ Then rapidly, all backbone SOEs, both those managed by the central government (such as First Automobile Work (FAW), Dongfeng Auto, and Shanghai Auto) and those managed by provincial governments (such as Guangzhou Auto, Nanjing Auto, and others), have been included in this wave. By 2000, a total of 71 such automobile-producing JVs had been established, in addition to five other cooperative projects involving multinationals. In order to promote the localization rate of manufacturing, another 557 JVs had been set up to produce components and parts by 1998.¹⁴ By 2005, nearly all multinationals active in the international market had entered China with car-producing JVs with Chinese SOEs.¹⁵

As the backboneSOEs focused their strategy on the operation of China-foreign JVs (except in rare cases),¹⁶ they officially abandoned their own efforts to develop self-owned car brands.¹⁷ Instead, they stressed production based on imported blueprints with the brands of their foreign JV partners. JV production was often dependent on the introduction of complete sets of equipment, technical licensing, knockdown components, and personnel training from the foreign value network. Before the policy transition to indigenous innovation in 2006, most backbone SOEs had only retained the functions of investment and administration while relying on cooperation with multinationals for product development, manufacturing, and branding.

After the policy transition to indigenous innovation in 2006, backbone SOEs were also required to innovate. However, they did not have sufficient motivation or technical capacity. They had gained great benefits through JV production, while doing indigenous innovation would have created huge risks, without providing a clear prospect of decent returns. Therefore, backbone SOEs generally adopted three strategies during the second half of the 2000s under political pressure and social appeals.

First, they acquired foreign assets. Some multinationals were weakened after the fierce global competition during the 1970s–1990s, which created good conditions for the Chinese to acquire them as a shortcut for gaining technological assets. Not only did the backbone SOEs have such a desire, but so did some new entrants. The well-known case of acquisition of Volvo in 2010 was by the new entrant Geely. However, as the backbone SOEs had not built up their absorptive capacity for assessing and assimilating technologies, they did not perform well in identifying quality assets and negotiating better prices. Beijing Auto bought most of the IPs of SAAB in 2009. However, as revealed by the media afterward, many high-value IPs had already been peeled off before the deal, as previously SAAB had been traded for several rounds on the market. So what Beijing Auto obtained were just two relatively aging platforms, the Saab 9–3 and Saab 9–5, the latest update of which was in the late 1990s.

The Chinese acquisition of Rover was more dramatic. In 2005, two Chinese SOEs, Shanghai Auto and Nanjing Auto, announced acquisition of Rover’s core assets. In fact, Rover successfully sold its assets to these two separate buyers, namely, the soft technologies (IP, drawings, etc.) to Shanghai Auto and the hard assets (production equipment and factory buildings) to Nanjing Auto. The chaos about the Rover technologies was not finally resolved until 2007, when Shanghai Auto acquired Nanjing Auto, which allowed it to combine the soft technologies and the hard assets from Rover. Of course, this approach can provide an initial shortcut for backbone SOEs to build their own technological capabilities. However, these firms still need to earn experience through indigenous innovation to become a genuinely competitive car developer for the long term. According to our observations, this has not happened yet.

As a second method to respond to the indigenous innovation drive, backbone SOEs established new brands through their JVs with multinationals. Looking for shortcuts, they often purchased the IPs of some existing or even outdated products (for more affordable cost) from abroad. Adjustments were done on the existing platforms, which were often very marginal, such as redesigning bodywork, installing new auto electronics, improving interiors, etc. However, most products of this nature were less satisfactory to local consumers, who have become picky as they have witnessed superior products entirely developed for them. Many of these self-built brands were soon marginalized on the market, such as the Everus brand launched by the Guangzhou Auto-Honda JV and the Ciimo brand developed by the Dongfeng Auto-Honda JV. Only those really based on localized R&D were able to survive through tough competition, such as the Baojun brand of the SAIC-GM-Wuling JV.

The third way was to buy technology from local developers. Some backbone SOEs have come to ally with new local entrants or directly procured technological packages from them. For example, in 2012 Guangzhou Auto signed an agreement with Chery¹⁸ to obtain the basic technology of the A-class platforms from Chery, as well as the full package for a G3 model. With this agreement, Guangzhou Auto also outsourced the engineering service provided by Chery concerning the engines and the continuously variable transmission gearboxes adopted in the relevant car models. In fact, these inputs aided Guangzhou Auto in developing its own Trumpchi brand.

It is only after 2013, witnessing how the new administration of Xi continues to stress indigenous innovation, that the backbone SOEs have begun to take innovation seriously. Within these firms, the units running self-built brands get promoted in terms of resource support. Baojun, the brand of the Shanghai Auto-GM-Wuling JV, has become a competitive brand in the low-end market. The Trumpchi brand of Guangzhou Auto, as mentioned, also made its name in the mid-range market segment. The continual investment in R&D and relevant experience accumulation after the policy shift have forged their rise. But as newcomers to indigenous innovation, their trajectories of capability growth are not different from those of the forerunning local innovators such as Chery, Geely, BYD, and Great Wall. Longer periods of investment and experience accumulation are needed in order to develop sustainable capability building. All these facts will be clearly demonstrated by our patent analysis later on.

Since 1997, new local entrants such as Chery and Geely began to produce cars based on different strategies, suggesting the possibility of strategies other than TMFT.¹⁹ But the entry of these manufacturers was “illegal” to some extent, because they had not been listed in the official catalog yet. These new entrants were mainly supported by local governments. As the spin-off effects of car-making on related industries were attractive, some local governments, in their bid to promote local GDP growth, risked supporting local firms even without official permission from the central regulator.²⁰ The support was vital for new entrants, as the local governments not only provided funds but also allowed them to sell their products in local areas.²¹ It is important to note that these new entrants are not the industrial players directly managed by the central government, and represent only local economic power, which is critical for understanding the subsequent policy changes. At least in the car industry, the policy transition from TMFT to the indigenous innovation strategy has not happened in a completely radical manner. This is partly so because the leading cadres of the ministries of industrial administration, especially the MIIT, had often been the former heads of large SOEs. This explains why policy-makers acquiesce in the slow response of backbone SOEs to this strategic transition, especially in their continued emphasis on JV production.

The influx of new local entrants has fundamentally changed the rules of the game in China’s car industry. They offer more cost-effective products and are more inclined to listen to the needs of local consumers by continually launching new products, which makes Sino-foreign JVs unable to sustain high profits by producing a handful of outdated models. Thus, since 2001, as the new entrants gradually obtained regulatory approval, the entire production scale and the amounts of new products launched annually in China’s car industry have skyrocketed, due to the leveraging effects created by new firms (see Fig. 1).

As new local entrants often lack technology accumulation, their first-generation products were developed mainly by mimicking existing products. Thus, after 2001, when their threat to incumbents was recognized, multinationals sued China’s new local car-makers for IP violation in a series of cases. These disputes affected public opinion and reinforced the notion that new local entrants were technologically backward.

Therefore, the new entrants who survived the initial phase of market competition demonstrated obvious commitment to developing self-owned technologies to solve their sustainability problems. Their efforts, though diverse, can be categorized into three types. Firstly, they invested heavily in in-house R&D. Chery, for example, was founded in 1997, and by 2005 it had an R&D center of more than 2000 professionals. In addition to the development of car models, Chery also has specialized in engines, gearboxes, new materials, electronic control systems, and so on. The cultivation of this R&D force was supported by high resource density: in 2005, on average, each of Chery’s R&D personnel was involved in more than 12 projects at the same time. By 2012, Chery’s R&D team reached its peak, with more than 6000 people and more than 120 ongoing projects for developing car models and important assemblies.²²

The second type of activity that the surviving new entrants undertook was to establish their own core supply system to provide key technical modules, such as air-conditioning, automotive electronics, stamping molds, and so on. At the beginning, they had to rely on existing local suppliers. However, under TMFT, most car component suppliers, though not all, were attached to a Sino-foreign JV value network. As those incumbents, especially the multinational part, had tight control over core suppliers, i.e., the tier-1 suppliers, the new entrants could only reach tier-2 or even more marginal suppliers. However, the technologically inferior tier-2 suppliers²³ were not willing to provide engineering R&D services to new enterprises, to avoid potential pressure from their car-making partners. All these constrained the scope and quality of components the new entrants could outsource.²⁴ To break through the roof, the new entrants had to set up their own tier-1 core supplier system. The concrete approach was to invest in cooperation with talented engineers who have already accumulated considerable experience through their prior work at giant international firms. For these engineers, the plans of Chery and Geely offered them a chance to start their own businesses. After witnessing their rapid growth, some established suppliers gradually became willing to set up JVs or specialized factories for these new players. Chery, for example, by 2012, had established a network of more than 200 tier-1 suppliers and already had several tier-1 suppliers to choose from in some areas. These achievements enable the new entrants to independently develop and produce cars and key modules, including a large number of patented technologies.

The third type of activity that the surviving new entrants underwent was a wide range of technical cooperation with international professional technology firms. After the 1990s, the growth of the global mainstream market had slowed, and some professional engineering technology providers began to develop strong incentives to open up the Chinese market, giving the new entrants opportunities to outsource engineering services and learn through cooperation with these service providers. As the earliest case of this kind, the Italy-based design house, Pininfarina, in 1991, took the initiative to approach Hafei Auto (a Chinese minibus manufacturer) to inquire into whether the latter was willing to employ its services.²⁵ This pattern quickly spread among local newcomers, who sought specialized services on the global market in different fields, such as systemic configuration, chassis engineering, engine calibration, and so on.

These kinds of cooperation were not just simple turn-key projects, but involved personnel training. Young and most elite Chinese engineers participated in these projects. The foreign partners had to cooperate in such learning schemes to win a long-term place in the emerging Chinese car market.²⁶ In such schemes, the Chinese participants played a more dominant role in cooperation, which helped them better identify technologies at the systemic level and explore more frontier issues.²⁷

New local entrants kept their focus on technological progress for sustainable survival, which also changed the behaviors of multinationals. Before they were challenged by new local entrants, multinationals rarely launched new products on the Chinese market, despite the fact that a large number of multinationals had actually established their Sino-foreign JVs before 2001. In 2001, there were only 13 new car models in China by 12 vehicle-producing JVs set up by international giants (see Table 1). Only after the new local entrants were allowed to enter the industry was product-based competition triggered. After that, the rules of the game have changed, and the amount of new products on the market every year has increased rapidly, above 100 for most of the years after 2006, many with investment by multinationals.

The MOST encouraged local firms to participate in projects it commissioned and funded. Backbone SOEs, such as FAW and Dongfeng Auto,³² were those that the the MOST needed to persuade to join in. However, some new entrants signed up for these programs and also invested their own money. A typical example is Chery, which took part in the 863 EV program in 2001 and infused its strategic resources into it.³³ Therefore, in the second half of the 863 EV program, which started in 2003, Chery has become the most important participant, and its automotive engineering institute was accredited by the MOST as “national engineering and technology center for energy-saving and environment-protection vehicles.”

During this phase, Chery invested heavily in HEV, partly out of its eagerness to be recognized by the central government as a competent car-maker. It even invested in a series of collaborative developmental projects with Ricardo, a renowned British technological firm, starting in 2004. With intensive learning through a 4-year period, Chery mastered various mainstream HEV technologies, including a range of light, medium and full modes of electric motor-engine hybridization.

However, Chery’s technological choice was problematic. As the development proceeded, Chinese developers found that Japanese manufacturers, particularly Toyota, have obvious first-mover advantage in this area. Leading Japanese firms had already patented a large number of high-value inventions, and many of these patented technologies were fundamental to this domain and difficult for latecomers to circumvent. Therefore, not only Chery but also giant firms from Europe and North America were hardly able to proceed with commercialization of their technologies. Technological ecology in the fuel cell EV was also similar, and some core patents were in the hands of a limited number of international giant firms.

The rise of BYD in EV is a story highly similar to that of the new entrants’ rise in the FFV sector. BYD was originally below the radars of the government and traditional Chinese automotive community. BYD had built strong capacity in battery development by the early 2000s and became globally competitive in providing lithium and Ni-MH batteries for laptops and cell phones.

In 2005, a research team at MIT released their findings about lithium iron phosphate as the cathode material for batteries. BYD reacted quickly, launched its industrialization program in 2006, and started to produce large volumes of lithium iron batteries in 2009. The reason for BYD’s efficient reaction lies not only in its long-term accumulation in battery technology but also in its research-oriented core organization. BYD was created by a group of former scientists of China’s General Research Institute for Nonferrous Metals and maintained the tradition of carrying out wide-ranging developmental activities in-house, from researching metal materials to producing battery packaging and developing cars. After BYD successfully launched its EV models named F3DM and Qin, most of the EV manufacturers in China turned to PEV or PHEV. Chinese developers created a niche market and gave up their previous mistaken trajectory, the HEV.

China’s EV industry outpaced the USA’s to become the world’s No.1 in 2015 and accelerated the technological development in China. Relevant suppliers of components and engineering service have been quickly established, including batteries, charging facilities, electrical and electronic control, etc. Taking the battery as an example, in addition to BYD, there are more than ten large-scale battery providers in China, including CATL, Fortunate Solar, Guoxuan High-Tech, Optimum Nano, and LiShen.³⁴

New local entrants are the major drivers to China’s EV rise, including BYD, Chery, Zotye, Geely (in cooperation with the Kandi Group), and Great Wall.³⁵ Attacking the previous industrial ecology, they actively embraced the disruptive change of technology to gain greater market share and greater government recognition.

In China, patents are divided into three categories: inventions, utility models, and appearance designs. Among these, the invention patent requires the highest novelty and undergoes a more rigorous examination procedure, which gives it longer protection and higher value. Therefore, the patent portfolio of a firm can partly indicate its technical orientation. In the car-making industry, from 1999 to 2017, multinationals have submitted a total number of 87,089 patent applications in China, while new local entrants filed 66,043 and backbone SOEs and JVs filed 32,227. As for portfolios in the three categories of patent, multinationals have the highest ratio of inventions, at 79.53% (69,265), and a low ratio of utility models, merely 3.06%. The respective ratios for new local entrants are 32.84% and 50.24%, and for backbone SOEs and JVs 20.55% and 57.72% (see Fig. 2). The group of backbone SOEs and JVs exhibits obvious weakness in patent applications, as 58% of their patents fall in the category of utility model (Yang and Mu 2009).

The patent grant rate is another index to look into. According to the report released by National Intellectual Property Administration in 2016, the rate for the JVs is just 35.35%, markedly below that of multinationals and new local entrants, which are 57.01% and 50.28% respectively.

Multinationals are the first-movers at implementing patenting strategy in China, accounting for the absolute majority of patents in automobiles prior to 2004 in China. However, due to the absence of strong rivalry in China, the patent portfolios of multinationals remained thin early on. Only after being challenged by new local entrants did their quantity of patent applications in China rise quickly after 2005. Toyota, GM, Honda, Hyundai, and Ford are the top five applicants for Chinese car-making invention patents, accounting for 56.3% in total among the 25 selected firms. Most patent applications of multinationals are submitted by entities from their home countries, which indicate that their collaboration with their JV partners in China remains highly irrelevant to their patent applications. For example, among the 16,708 patents applied for Toyota in China, there are only 5 patents with Chinese co-applicants. Toyota subsidiaries in China, including the Toyota (China) Research Center, FAW-Toyota R&D Center, and Guangzhou Auto-Toyota R&D Center, have only submitted 19 invention patent applications in the past 40 years. That is to say, the offshore R&D may have helped the multinationals to localize their technological capability to better fit the local market or forward local customers’ feedback to their global headquarters to enhance R&D performance.³⁹ Anyhow, it indicates that multinationals such as Toyota do not allow their subsidiaries in China to cultivate real technological capability.

In 2003, GM claimed that the QQ model, a minicar product of Chery, violated its IP of the Matiz model, which was originally developed by Daewoo. In response to the lawsuit, Chery urgently submitted more than 20 patent applications to SIPO, which could be considered the point at which local car-makers began to emphasize patenting. After that, we have witnessed the rise of new local entrants, which gradually become the second largest group of patent applicants, only after the multinationals. In 2016, the number of invention patent applications submitted by seven Chinese new local entrants has reached 3386, equivalent to a half of that submitted by the 13 multinationals. The leading group of new local firms, namely, BYD, Chery, Geely, and Great Wall, even exceed some multinationals, such as Volkswagen, Kia, and BMW, in patent applications (see Fig. 3).

We are not arguing that the technological expertise of these Chinese firms has already overtaken that of multinationals. However, at least data indicate a shift in their IP strategies in the Chinese market. In fact, the total number of invention patent applications submitted by the new local entrants has already surpassed the sum submitted by Toyota and Honda after 2012. This is remarkable, as the leading Japanese firms are always very active in patent applications. The growth rate of invention patent applications filed by the new local entrants as a whole is almost equal to that of the multinationals. As for Sino-foreign JVs, although the sum of their patent applications has also increased after 2010, the growth rate is still significantly behind that of the multinationals and the new local entrants.

Multinationals started their patenting in China much earlier with much better performance. Over the past 20 years, multinationals have already submitted at least 7695 applications on engine technologies, and this has grown steadily, from less than 100 applications in 2003 to above 600 every year after 2011. US and Japanese car manufacturers have striking advantages in this field, as Toyota has accumulated 2285 applications on engines, GM 1167, Ford 1084, and Honda 1049 (see Fig. 4). In recent years, although people are arguing about the American deindustrialization or the decline of Detroit’s automotive industry, the American car-makers still maintain a higher innovation rate on engines than their overall general level through strengthening specialized R&D and utilizing links with European and Japanese collaborators.⁴⁰ Gradually, Japanese and Korean actors have already transferred their focus to the EV technologies rather than the ICE. For example, the invention patent applications of Honda for engine technologies in 2016 have declined by more than 50% from 10 years ago.

Compared with the leading group of multinationals, the new local entrants and backbone SOEs with their Sino-foreign JVs exhibit obvious disadvantages in engine technologies, with a total of only 1434 and 704 cumulative invention patents, respectively. Before 2002, only some backbone SOEs had the competence to patent engine technologies. However, the number of their applications was very limited, just one or two a year. Only with the rise of the new entrants did local applications in engine patents start to increase. For example, Chery’s invention application started to top the local Chinese applicants in 2005, with about 40 annual applications on average.

In general, the patenting performance of new local entrants is far better than the backbone SOEs and their JVs with multinationals. Among the backbone SOEs, Shanghai Auto has the best performing Sino-foreign JV, Shanghai Auto-GM. However, Shanghai Auto and its JVs have only 208 invention patent applications, clearly not comparable to the leading new entrants such as Chery or Great Wall. Out of the 208 applications, only 38 are co-submitted with its multinational partners, which again indicates the inefficiency of technological learning under the TMFT scheme.

Compared with the FFVs, the power of EVs mainly depends on new energy supplies such as batteries or fuel cells, which leads to remarkable changes in core modules and the overall design. The huge potential of the Chinese EV market has attracted multinationals to apply for patents in China. Thanks to the governmental incentive programs, local applicants have made rapid growth in EV patenting.^41,42 Prima facie, the group of multinationals still has an absolute advantage with over 16,758 invention applications compared to 938 from backbone SOEs and JVs and 4303 from new local entrants. Yet a closer look shows that among those multinationals, 6316 applications are submitted by Toyota alone. The gap between local Chinese applicants and multinationals is not as formidable as in the ICE sector. Especially BYD, together with GM, having 2188 and 2537 invention patent applications, respectively, has become the second-tier applicants, with visible advantages over the third-tier applicants, including Hyundai, Ford, Nissan, Honda, and Chery. Even a handful of backbone SOEs, such as Beijing Auto, Shanghai Auto, and ChangAn Auto, are incentivized by governmental policies to develop their own technologies and file for patents (see Fig. 5).

BYD, as the largest EV producer in China today, cooperates with Daimler-Benz in a JV, namely, the Denza, to produce PEV. BYD and Daimler-Benz collaborate on the bodywork design. And BYD leads most of the rest product development tasks in Denza. Indeed, the technological capability of BYD in EV, embodied in 2188 invention patent applications, is the reason why Daimler-Benz chose to collaborate with the young Chinese car-maker.⁴³

In fact, only Toyota and GM can maintain their leading positions ahead of BYD, although the advantages of GM have begun to fade. Further, given that Toyota has long been more focused on HEV rather than PEV or PHEV, which are the domains that BYD stresses, more detailed analyses need to be done in the future to judge the leadership contest in each market segment.