III. A Balance Sheet of Local Government Involvement in the Solar and Electric Vehicles Sectors

In the solar industry, the photovoltaic solar cell producer Suntech, in part due to support by the Wuxi government, has often been pointed to as a model of business and innovation success. (He 2006) (These successes –including in innovation²⁸ - were most notable before 2006, i.e., prior to the firm’s financial problems and the 2012 embroilment of the whole Chinese solar panel industry in international trade disputes.) As is well-known, the Wuxi municipal and Jiangsu provincial governments helped lure the firm and its entrepreneurial founder, Shi Zhengrong, an overseas Chinese engineer living in Australia. The Wuxi government provided sizable initial start-up grants and subsidies, taking a major equity stake in the firm by providing US$ 6 million in return for a 75% equity stake. The city later helped Suntech search for additional funding from national and local sources, for example, from the provincial Science and Technology department. Some of these loans did not need to be repaid. Wuxi officials further helped organize a sizable package of loans from banks and local venture capital groups; notably, former official turned board chair Li Yanren helped arrange for 5 billion yuan in low-interest loans. Subsequently, and in anticipation of Suntech’s 2006 listing on the NY Stock Exchange, the Wuxi government offloaded its shares to other investors – including local state backed “venture capital.” (Ahrens 2013, 2) In addition, the government was instrumental in setting up a regional cluster for the PVC industry in Jiangsu, completing a relatively full value chain in a few years. It also set up an R&D center that has led to manufacturing and efficiency improvements – though the commoditization of solar cells in recent years have rendered it more price and quality-driven than innovation driven. Nevertheless, at least in part as a result of these activities, higher-tech silicon slicing technology could be developed and commercialized in the region. (Liu and Chen 2012) Wuxi government officials also sought to help on the demand side, finding projects that Suntech could supply. In short, the “Wuxi Model” of financial support from the city government has been touted as a major reason for Suntech’s success, with the implication that Suntech was an innovative firm. Suntech was, furthermore, the prototype for Wuxi’s 2006 “530 Plan,” designed to supply between 1 and 3 million RMB per approved project, the major criterion for approval being that the project is technologically promising. The number of registered companies under this program has reached 876, with a total registered capital of 2.5 billion RMB.²⁹

The role of the Baoding municipal officials (Hebei Province) was similarly important for the emergence of the private Yingli Solar (英利太阳能). Key supports are in the provision of land, tax reductions, and aid in obtaining building permits. Local officials also have supported the firm in its applications for central government funding. (Local government contributions are generally required to be awarded central funds.) Local government approval of bank loans remains necessary, and also is considered a de facto government guarantee for the repayment of loans. Local government help has not been limited to that supplied by Baoding. Yingli Solar began to diversify away from PV production toward downstream power generation, and has developed a provincial government-based organizational strategy.³⁰ Yingli planned to substantially decentralize operations away from its Baoding headquarters, to build independent generating stations for solar power, especially in western and southwestern provinces (e.g., Xinjiang, Yunnan, Guangxi, and Shaanxi). Building new branches in other provinces meant courting close ties with those provincial governments, as Yingli needed substantial funds for these generating stations come from local governments. The firm also wished for these governments to contribute land and facilitate loans from other sources including provincial branches of the China Development Bank and Bank of Transportation. Yingli also expanded to Hainan, where it was necessary to work with local government as well. Representatives attribute the ability to build the production facility within three months to positive coordination between the firm’s CEO and the Hainan provincial government.

In similar fashion, the founder of LDK Solar (江西赛维 Jiangxi saiwei), Peng Xiaofeng, shopped around to find the most favorable city government, one that would provide financial support and generally mobilize around the industry. Peng found his answer in the small city of Xinyu (Jiangxi Province). (Wang 2007) yet although LDK gained the support of the Xinyu government, and despite LDK’s ability claim itself part of a strategic emerging industry, there is little apparent innovation involved in the company – a problem discussed below.

Turning to the second case, note that China’s auto industry structure has been characterized by the emergence of a few national firms (many a mix of SOE and private-foreign joint ventures. Beijing’s EV policy seems to have assumed these large firms would lead the foray into EV sedans. At the same time, many provinces have home-grown local auto industries, many of which have been beneficiaries of barriers to entry via local (provincial) protectionism. These local firms often produce lower quality and price vehicles This dual structure illustrates keenly how the two core variables – industry structure and policy signals from Beijing – play out.

In the EV sector, Shenzhen’s BYD Auto (比亚迪), has been at the forefront of China’s electric vehicles industry. BYD exemplifies how local government behavior has intersected the emergence of this sector. Subsidies on the supply and demand sides alike have been important.³¹ BYD’s income statements show government grants of 400-500 million RMB. BYD also enjoys large revolving lines of bank credit, for example, a 10 bil RMB line of credit from the China Development Bank. The 2011 interim financial report by the Shenzhen Development and Reform Commission shows total subsidies of 1 bil. RMB.³²

The Shenzhen municipal government also provided coordination between the firm, banks and other SOEs, while the Guangdong provincial government issued official documents showing clear support for BYD electric car projects. (Chen Zhijie 2008) Perhaps most important, the Shenzhen municipal government agreed to purchase BYD electric vehicles for its municipal taxi and public security bureau fleets. So did other governments where BYD agreed to invest, including Tianjin, Xian, Kunming, and Chengdu. This “demand pull” support became a central pillar of BYD’s business plan. Shenzhen municipality also subsidized the purchase of BYD electric vehicles for individual consumers,³³ and facilitated the establishment of a number of charging stations for electric cars, though as we shall see these have been insufficient. (A similar story can be told for Xiangfan New Energy Vehicles in Xiangfan, Hubei. Li 2010.)

Discretion by local governments – to tailor policies to local characteristics or opportunities – is often characterized as “experimentation,” and as noted is typically viewed as a positive, flexible aspect of Chinese policy-making. In the EV arena, Beijing has authorized local experimentation as well through the 2009 “10 Cities, 1000 Vehicles” program.³⁴ Cities such as Guangzhou, Shanghai and Hangzhou were granted leeway to decide how to best promote the expansion of electric vehicle usage in their cities, and tended to develop strategies that centered on their ‘local’ industry (for example, BYD in Shenzhen).³⁵ Hangzhou, for example, built on its past positive experiences with bicycle rentals to build electric battery rental and mini-bus rental models, and to promote “battery switching” as a remedy for the lack of a charging infrastructure. Kandi Automotive (康迪汽车), traditionally a manufacturer of small vehicles such as ATVs, go-karts, and golf carts, was at the forefront of these efforts. Kandi worked closely with the Hangzhou and Zhejiang Province governments³⁶ and, subsequently, Shanghai governments. These major municipal governments have participated extensively by providing land for rental locations and charging stations operated by a Kandi JV, ZZY (左中右)³⁷. Kandi also was greatly benefited by forming a joint venture with the large private Shanghai firm Geely. This JV was important for Kandi to be able to switch from the use of lead acid batteries to lithium iron phosphate batteries. Geely also facilitated Kandi’s attempts in 2013 to gain approval from the central NDRC of its low-speed models, resulting in these models being listed in the MIIT directory as approved and qualifying for consumer EV subsidies from the national government, not just from provinces. The combination of business model innovation, cooperation with other major market actors, and positive support from local governments allowed Kandi to lead sales of all EVs in China in the 2012-14 period, and facilitated the firm’s entrance to major markets outside of Zhejiang Province, such as Shanghai and Nanjing.³⁸ These sorts of experiments, when aligned with Beijing’s intentions, are therefore seen as a relatively positive and unique part of China’s policy process. (Marquis et al, 2013).³⁹ As will be discussed below, however, there is a significant downside to these local experiments insofar as they can promote protectionism and hinder the adoption of nationwide standards that might better facilitate EV adoption.

Beijing views other local experimentation to meet local conditions as more problematic, although the impact may be better for innovation in the sector and, ultimately, EV adoption in the local market. The promotion of low-speed electric vehicles in Shandong Province is a case in point. Local economic development agents and auto firms in Shandong - notably Shifeng Automotive Group (山东时风汽车集团), traditionally a producer of trucks and tractors - lobbied hard beginning in 2009 for the provincial government to approve a pilot for producing low-speed, light weight vehicles particularly suited for rural areas.⁴⁰ Local (county and municipal) support included not only lobbying but also provision of investment funds and land for industrial parks. These vehicles would be upgrades from bicycles, electric scooters, and three wheel motorcycles, and could be offered affordably to rural businesses and families.⁴¹ Part of the low cost was due to use of lead acid batteries as opposed to the lithium batteries preferred by Beijing. Low-speed EVs were even more affordable because, as until recently the vehicles were not classified as “automobiles,” consumers required no license and therefore could avoid licensing fees. However, as the vehicles are very lightweight, and because their low speed capacity could disrupt regular traffic, safety concerns persisted. NDRC officials in Beijing publicly called such vehicles “junk technology” (垃圾技术), saying they failed to meet national standards.⁴² Despite the absence of central approval, and with the backing of the provincial government, this local pilot has been relatively successful at putting affordable electric vehicles on the roads, especially in rural areas; whereas China has underperformed its target for putting 500,000 hybrids and EVs in use by the end of 2015, in 2012 Shifeng delivered about 30,000 low speed vehicles to dealers.⁴³ Beijing eventually relented in its opposition, and in the fall of 2014 allowed subsidies to be applied to these vehicles, suggesting such innovation is no longer “illegal.”⁴⁴
Distortive Influences of Local Government
Local governments clearly have substantial opportunities to take on coordination roles in support of the development of clean energy technology as a public good. Beyond the cases discussed briefly in the previous section, there are myriad examples where, on the surface, local governments have pursued coordination and financing mechanisms to support prospective innovators. Yet local government participation often has failed to create a positive ecosystem for new energy firms. This section lays out several patterns of local government behavior that hinder and at times even overshadow what otherwise might be their positive catalytic function.
The Interplay between Central “Signals” and Tangible Local Results

Local governments in China must be conduits of central government policy. Despite the shift from directive planning to “guidance” planning in which central statements serve more as a guide, localities are still expected to respond to the center, including to both broad five-year plans and sector-specific policies. First and foremost, the central guidance, such as in the SEI initiative in general, and new energy industry promotion specifically, are signals of the center’s policy preferences, to which local officials must show some degree of responsiveness, as if they comply. Not only is this to be expected as a function of the unitary political system, but also the promotion system for local officials is tied to responsiveness to signals. The top-down cadre management system produces career incentives for local officials to show responsiveness. This occurs through the regular performance evaluation, rotation, and turnover.⁴⁵ Furthermore, the cadre management system – in addition to and supportive of a norm of local “experimentation” – helps protect local cadres’ promotion ambitions as they creatively experiment with how to align their concrete governance interests with the signals of the center. Officials’ creativity was aided by 1984 reforms according to which the central party organization only manages the careers of officials of the immediate lower (provincial) level. This change helped insulate local leaders if their experiments come to be deemed contrary to central directives.⁴⁶

If we accept, as most of the literature on Chinese local governance does, that local cadres are much concerned with how their performance affects their prospects for promotion, and understand the local flexibility to develop specific plans for innovation, then it becomes clear how well-known cracks in the alignment between central guidance and local interests might affect local governments’ innovation incentives. Local officials in the state’s nomenklatura system typically have quite short time horizons.⁴⁷ As their terms in office often are shorter than the official 5 years, their promotion prospects – the ability to show results from their leadership –are quite short. The incentive to make an impression on immediate superiors in the short-term (2 or 3 years) is strong. This short time horizon in and of itself is a problem when technological innovations may require a longer term to come to fruition, especially when commitments to innovation projects are being used by a particular leader to demonstrate political results. ⁴⁸

How might local officials demonstrate short-term “innovation results” to superiors? The establishment of a HDTZ or the launch (and early success) of an innovative industry – particularly one that is the focus of national level policy, such as an SEI – can bring positive attention to officials who can claim responsibility. For example, snagging one of the 54 “national” label high-tech zones (as of 2011) would be a coup. A second marker is the ability for local firms to snag the designation of “National Key Lab,” suggesting that this firm has been singled out by the center for its innovation. Third, the potential to create jobs and local growth through the attraction of industries that can be labeled “innovative” also is a powerful incentive for local officials. (Dai 2015)

Local government officials are well trained to jump onto this bandwagon, and even be seen as “ahead of the curve.” Jiangsu Province attempted to get a head start on the competition in terms of its reputation as an innovative region. Whereas the central government set the goal that China should be an “innovative country” by 2020, Jiangsu set the provincial goal at 2015. Similarly, in response to the central government’s emphasis in 2009 on SEIs, governments from some cities proposed they would carry out “100 major projects” related to SEIs during each year of the 12^th FYP. (Liu and Chen 2012). Such goal-setting compels governments to meet the numeric goal regardless of the quality of projects.

These dynamics were evident in Wuxi, where the ability of entrepreneurial officials to attract and support Suntech, and create a local value chain, led it to promote the “Wuxi” model. The multiplier effect of a solar industry value chain centered around Suntech held tremendous value not only for Wuxi but for Jiangsu province as a whole. According to Suntech’s founder Shi Zhengrong, the Jiangsu PV industry can be valued at 200 billion RMB, and employs 170,000 local workers. Such enticements would be attractive to any local leader in any country.⁴⁹ In addition, the Wuxi mayor, prior to the collapse of the PV export market, was able to leverage this and other “pro-innovation” activities into a promotion to Jiangsu party secretary. Even if not achieving the notoriety of Wuxi, this dynamic is repeated in localities across China. By showing support for an innovation project, a local government increases its prospects of gaining coveted central governmental “support” – as in the case of Shenzhen’s (BYD) and Hangzhou (Kandi), noted above – or gaining attention as an experimental site (shidian). This dynamic may be even more accentuated outside of major cities, where officials strive to compete with more developed locales. In pursuit of a chunk of the low-speed EV industry, officials in small cities with access to rural markets have set up many “new energy automotive industry parks,” and contributed millions of yuan in funds to support investment, and many acres of land.⁵⁰

Being designated a model city by the center is itself a signal that may encourage other firms to invest in its jurisdiction – it helps the city to create an innovation “brand.” Baoding City (Hebei) leveraged early successes in promoting solar technology firms into its label as a “Green City,” despite having horrendous air quality. Indeed, these green energy successes were likely a result of earlier efforts by key leaders in Baoding, including the mayor, in leveraging central funds as well as local human capital to build “Baoding Electronics Valley.”⁵¹ And Baoding has in fact been home to successful “green” industries, notably Yingli Solar.

Model status can allow local leaders to gain access to additional central government resources (such as from NDRC and the Ministry of Finance, or for low interest loans from the China Development Bank), as well as foreign investment. At the same time, to the extent that leaders need quick payoffs from an industry or firm they support, that can lead them to construct financing mechanisms that require quick pay-outs – something also potentially distortive of a firm’s rational growth. This criticism has been made of the implementation of the Torch program and establishment of HTDZs by local governments; funding went to mature technologies that could be rolled out to market quickly, but were not often very “high-tech” or very “innovative.” (Breznitz and Murphree 2011, 77-8) Jiangxi’s LDK Solar faced this issue with the financial support it got from the local (city) government: if the firm did not pay out profits and reimburse the government within a short period of time, then the cost of the capital would be increased.⁵²

While these behaviors have led to some positive results, in other cases they have created political failures, including wasted and duplicative investment (with attendant opportunity costs of capital), shady land deals, corruption, and other ills. The desirability of being seen as responsive to central signals and help build the local brand often leads to a shallow or even a “false” response. And the competition to gain the “doorplate” label of, e.g., “Green City,” has often led local officials to act with great alacrity. This is especially true with the contributions of land, which in capital-short areas in particular have been a major form of contribution by local governments to industrial innovation. In the wake of the 2008 financial crisis, funds flowed quickly from the central government to localities. Indeed, such funds created the first real push in favor of SEIs since the program was announced in 2006. Localities that suddenly had reason to hope for a flood of central funds had to find where to quickly spend them, and too often the use of SEI categories such as the “internet of things” was the excuse to set up new development parks that supposedly focused on such high-tech industries. Instead, however, the result was simply new, often empty, office parks and other industrial real estate projects. This problem was frequently observed in Wuxi; one example is the “Internet of Things Office Park,” a real estate development project, sponsored by Wuxi government.⁵³

BYD provides a related example of wasted effort. Local governments in which BYD invested outside Shenzhen attempted to create additional favorable conditions for electric car sales in the form of setting up charging stations. However, once a few stations had been set up – enough to demonstrate political support of electric vehicles but not enough to spur consumer demand in any meaningful way – local leaders are said to have lost their interest and moved on to line up behind other projects.⁵⁴ The provision of charging stations is a major hurdle for EV firms, especially in space-deprived first-tier cities (a problem not just in China). A contribution by government is to convert city-controlled land to charging stations.⁵⁵ The point is not that local governments have created the charging hurdle but, rather, that they have been unable – despite major efforts – to ameliorate it. A similar dynamic surrounds “National Key Labs.” In firm visits, managers touted they had received the designation, but there was not evidence of significant activity going on in the lab. This was the case, for example, in the wind turbine sector, at Guodian Wind Turbine (Baoding). A tour of the key lab – well-advertised on plates but kept behind locked doors - revealed broken equipment, and virtually to no activity. Similarly, Yingli Solar in Baoding showed that it had designations for a “State Key Lab of Photovoltaic Materials and Technology” (2010) and “State Key Lab of National Energy Photovoltaic Technology” (2011), yet little obvious evidence of outputs ongoing from these labs.⁵⁶

Two issues should be emphasized. First, the dynamic in which local officials are incentivized to respond to central signals and the resulting effort to create a “brand” for a locality – in order to bring in other cognate industries – has not always served to create a negative environment for firms in industries picked by the center for growth and innovation. Baoding is often pointed to as a positive example. However, that local government involvement is dominated by these incentives, and evidence of wasted efforts and establishment of firms that are not successful, much less innovative, is problematic. Hence, the longstanding criticism (within China) that the Chinese system’s propensity to incentivize wasteful investment is repeated in the clean energy sector. Second, it is clear that local governments are keen for market entry of new firms, as symbols of and perhaps the reality of growth. Yet, as examples below will further illustrate, local governments are not keen to allow market exit of local firms, especially those that they have supported with funds in the past.
Protectionism, Fragmentation, and Stovepiping

While we can observe successful examples of local government generated coordination (e.g., Suntech in its heyday), local politics in China is often characterized by three interrelated factors that run counter to coordination: protectionism, fragmentation and stovepiping.⁵⁷ Competition between regions to attract firms for local development has been intense through the reform era, and continues to result in efforts to gain advantages.⁵⁸ Indeed, China’s nascent post-Mao market economy was marked by significant local protectionism (Wedeman 2003; Cai 2004; Bai et al 2004). To this day, local government officials remain willing and able to try to sustain local advantage. Despite much progress in the development of national markets, and central government efforts to rein in local protectionism, local officials often continue to erect internal market barriers. Firms are often complicit.

BYD’s reliance on local government jurisdictions to purchase electric vehicles provides a good illustration of the impact of local protectionism. Local governments can often be relied upon to provide a demand pull for new technologies. To this end, several subnational governments, including Shenzhen (the BYD headquarters), Tianjin, Xian, Yunnan, Changsha, and Chengdu, agreed to buy BYD vehicles for their municipal fleets. But these local officials would only purchase if from a “local” company. BYD therefore had to invest in local outlets in order to make these sales as a political rather than a business strategy decision. Local officials could then claim these investments as their own “pro-innovation” industries, demonstrating their responsiveness to Beijing’s signal.

More broadly, governments in the “Ten Cities” EV pilot often have endeavored to protect their own local auto companies. Most of these cities allowed just two auto companies to sell EVs, with one of these firms being a local company.⁵⁹ Even when the NDRC directed that these cities needed to allow 30% of sales from other provinces, the province seeking protection would make a deal with another province that the whole of that outside 30% would come only from that other locality. Local governments often subsidize more generously purchases of EVs from local auto firms.

Fragmentation between actors within a jurisdiction also can have deleterious effects on coordination. Multiple funding sources at the central and provincial levels may be used for local innovation projects – such as from NDRC, Ministry of Finance, Chinese Academy of Sciences, and MOST and their local branches – yet these sources often are poorly coordinated. Thus, local efforts to snag resources and then show results from the investment often mean not just competition between local projects, but also wasteful duplicative investment sponsored by different agencies within a location. Particularly in an environment where market exit is discouraged by local governments as a sign of political failure, duplicative investment can be presumed to be especially harmful. Recent efforts to criticize and eliminate overexpansion of technology parks, particularly those related to the SEI industry of cloud computing, reflect recognition of this problem by the center.⁶⁰ Yet as long as local bureaus of national ministries are more likely to serve horizontal interests than their vertical masters, the problem will remain. (OECD 2008, p. 363-4)

Related to the stovepiping of the local bureaucracy is that each bureaucracy within a local jurisdiction may feel compelled to respond to the center’s signal, exacerbating wasteful duplication. Indeed, even SASAC has taken on the mantle of promoting S & T at the local level. (Naughton 2012) Duplication of efforts is not the only problem to grow out of fragmentation and stovepiping; such a structure creates barriers to diffusion of technology that is seen as crucial to creating effective innovative “clusters.” (OECD 2007, 41) Also detrimental is the stifling of development of national standards for new industries. As local governments focus on developing standards that benefit their own specific location, and local companies, rather than working toward national (or even international) standards. Similarly, while the development of locally-oriented technologies is rational from one perspective, it can hinder the development of a more robust national market.⁶¹ Chongqing municipality’s deployment of a fast-charging battery model - the only pilot city to do so – is a prime example. This model was based on the unique characteristics of proximity to the Three Gorges dam and relatively reliable power grid. A battery-swapping model a la Hangzhou and Shenzhen relies on relatively flat geography, and hence is not feasible in hilly Chongqing, whereas the energy supply intensity and reliability needed to make a fast-charge model succeed is not available elsewhere. (Marquis et al 2013)

While government supports for the EV industry have led some companies to jump on the bandwagon, others have assessed the fact that the industry is extremely reliant on government support rather than market demand as reason to wait to enter the sector. Representatives of the private firm Chang Cheng (长城) Motor (Baoding) indicated that while they had explored entering the market (including with a failed cooperation scheme with a California battery company), they at least temporarily decided to hold back. They indicate that the market is not mature, that the push is completely led by the government, but with insufficient and unreliable financial supports. They reiterated the common complaint that there is poor interconnection with the grid and that there are insufficient charging stations. Chang Cheng has a stronger interest in hybrid technology, and its 200-person “new energy team” has participated in studies with the central Ministry of Finance (though the firm ultimately withdrew).

Problems with interconnection also have hindered China’s solar industry. Solar cell manufacturers large and small have been frustrated by efforts to sell electricity on the domestic grid. Local governments have contributed land and housing to huge but solar farms projects, many of which sit idle for reasons already discussed. But in addition, local governments have been unable to guarantee connectivity to the grid. Consumers have faced similar frustrations. Some consumers have invested in solar cells, partly spurred by the national “Golden Sun” and “Golden Rooftop” program subsides for installation of solar cells, and partly spurred by prospects of not only providing a source of power for their own use but also of selling the electricity they generate at home to the grid. Yet state power grid monopolies have been unwilling, despite directives from the regulator (China Electricity Regulatory Commission), to facilitate interconnection for households or solar farms without burdensome forecasts of generation and at a reasonable price.⁶⁴
Local Governments as Financiers

Successful entrepreneurs in solar and EV industries often cite the importance of local government financing. Yet two main problems with China’s model of local government support can be identified: the questionable ability of local governments to make wise investments in the absence of market based mechanisms, and the creation of firms dependent on government funding. Criticism of local governments’ failures to effectively “pick winners” is common in the Chinese press. Local government is on the front line of choosing which firms to reward, and yet there often is a structural expertise gap. Or where expertise does exist, such as in local S & T bureaus, these expert opinions must compete with other considerations, such as the need for quick returns by local officials who are seeking ways to respond to the center’s signals, as discussed previously. Political decisions often must substitute for market-based decisions as to worthy funding recipients. Well-functioning venture capital would act as a funnel, withdrawing funding from ventures that have failed to show promise (or live up to promise), while continuing funding for viable ventures.

But there remains a gap in such market-based institutions in China. As noted, there has been a push to establish locally-capitalized venture capital firms to (help) fund innovative ideas. More market-based VC firms are generally uninterested in clean energy because there is not enough money to be made. Local VCs that have invested in clean energy have come under increased scrutiny for making poor decisions, being too closely tied to the government, and even helping to create too much government risk when they gain implicit government guarantees. Wuxi Venture Capital Fund, an integral part of the Wuxi “Model” has met criticism for failing to assemble a “professional team to carry out project selection, due diligence, post-monitoring, intellectual property investigations, business partner surveys, communications platform, and other procedures.” (He[a] 2006) Other “risk-reward” loan systems also exist, as in the case of the “gamble” (duì dǔ, 赌) agreement between solar firm Jiangxi LDK and “state-background” lenders, in which the loans had to be repaid very quickly or else face extremely high interest rates. (“PV Impulse” 2012⁶⁵) At the same time, when local governments are faced with the prospect of a failure by a firm around which they have tried to create a “brand,” particularly when huge amounts of resources have been poured into not just the firm but also supportive supply industries, they may be tempted to throw good money after bad. Such is the case with Xinyu government in Jiangxi Province, the backer of LDK Solar. As with Wuxi’s Suntech adventure (though for different reasons), local officials have felt compelled to use government funds and help ensure the availability of funds from elsewhere (such as local banks) in order to bail out the firms.⁶⁶

Thus, in light of this gap in expertise and market institutions to pick worthy “winners,” it is not surprising that local officials’ choices might then be made based on other considerations, such as political connections. Altogether, there remain a host of moral hazard problems. And, as w/ Xinyu LDK and Suntech, local governments have been compelled to bail out firms so as not to shoulder a black mark to its brand.

A second problem with the model of financial support is its impact on industry. Rather than spurring innovative industries such support seems to promote waste. The solar panel industry is perhaps the most well-known example. Many local governments saw export of solar panels as a lucrative business, and so – responding to international market demand and drawing on national policy incentives – too many localities invested in solar cell production, creating a glut on the international market, and helping precipitate a collapse of that market.⁶⁷ Yet extensive government supports also have created a tendency for firms to depend on government supports, and to use this support to go in business directions they might better have avoided. LDK Solar, given extensive support from the entrepreneurial Xinyu government, made aggressive forays into ancillary businesses, though as a family-run company it lacked expertise to do so. Similarly, BYD Auto expanded into the solar industry largely based on government supports. In the EV industry, two of the “10 Cities” pilot cities with little existing infrastructure in the auto industry (Xiangfan and Nantong) have been accused of strategies that focus on “receiving preferential policies and financial resources, as opposed to developing their EV adoption capability.” (Marquis et al, 2013) Reports of similar dynamics have appeared for clean energy sectors of wind, LEDs, and bioimass. (“PV Industry Incubators”) Interviewees have also suggested there is a class of smaller firms that have been established and remain in existence due solely to government

IV. Conclusion
The broad literature on innovation suggests that local governments in China have a potentially positive role to play in creating a favorable ecosystem for innovative firms through mechanisms such as coordination and funding. The influential position of local governments in the Chinese political system, including as a conduit for central “pro-innovation” policy, reinforces that expectation. This role could be expected especially in clean energy industries, which have been given a high profile role in China’s planning process.

It is evident from the previous discussion that innovation by firms does occur. Cost and process innovation results from deep competition among solar panel producers, many of whom survived the export industry shake-out. When firms have been able to draw on existing infrastructure and expertise in their location, they have moved forward, as has been the case with EV manufacturers Kandi in Zhejiang, Shifeng in Shandong, and to a lesser extent BYD in Guangdong. It also can be argued that local governments have contributed to some successes, at least initially, through use of traditional industrial policy tools. Subsidies (direct funds, tax breaks, preferential access to loans from state banks, contributions of land, etc.) appear to have helped some solar producers, such as Suntech and Yingli, get off the ground. More interestingly, local governments have protected firms that have chosen to innovate in ways to which Beijing has been unfriendly. Shifeng low speed autos benefited massively from the advocacy of the Shandong provincial government. As a result, innovation occurred at the low end of the market, and this segment is the most successful portion of China’s entire EV market. Kandi (also benefiting from its partnership with Geely) sells more EVs in China than any other manufacturer. Perhaps one reason for the slow takeoff of new energy sedans was the focus on pure electric vehicles, and the desire to avoid hybrids (due to the need to purchase foreign IP). This choice was made by the central government, but local governments followed. (Tillemann 2014) In 2014, however, some EV firms have gone their own way on this issue as well; similar to the situation with low-speed vehicles, local governments have supported the development and sale of hybrids, particularly in Shanghai. (Wang Tao 2015) Moreover, local governments have continued to encourage firms continue to engage deeply with the global value chain as a consumer of foreign technology, instead of making serious attempts to reinvent an “indigenous” wheel.

On the other side of the ledger, however, while local governments have created experimental sites and pushed the development of industries along, there are myriad examples in which local governments have failed to contribute to vibrant local innovative ecosystems. This paper has pointed to a number categories of governance deficiencies. The incentives for local officials to respond to central signals, but to do so “creatively” in a way to support their political needs for quick payoffs in terms of branding and employment, has been a major contributor to wasted, duplicative, and sometimes “false” investment. The problem of waste is exacerbated by incentives for local governments to prevent exit of unsuccessful firms if it means harm to employment. Moreover, local governments in China face difficulty providing effective coordination due to longstanding problems of protectionism, stovepiping and fragmentation. Protectionism in turn harms efforts at standardization for a national market, such as when EV pilots have been encouraged to suit local conditions.⁶⁸ Reliance on government demand (for EV sedans) has been unable to ameliorate the basic problem of market failure the efforts were intended to address. From the financial side, reliance on local governments to make sound investment decisions when private financing does not exist means officials have often lacked adequate signals about the quality of projects. Thus, incentives for local governments – a mixture of incentives to align with the center but also create their own image as innovative locations – too often block the creation of an environment in which state-supported innovation can happen. Waste and the intolerance of firm failure and the consequences for market exit are particularly pernicious problems.

In this context, some firms, as in the solar panel sector, have simply never become viable market actors apparently in part due to their expectations that government supports will continue, a phenomenon that cannot help the center’s innovation visions. In the EV sector, despite some obvious innovation at the low end of the market, and facing many of the same technical barriers this industry confronts around the world, all the supports and subsidies has not met the goals for sales set by Beijing’s industrial policy. Beijing acknowledged that, as of September 2014, China had met only 12% of its target for alternative-energy vehicles to be introduced by 2015.⁶⁹ One study calculates that of all the cars produced in 2013 (the world’s largest auto producer), only .008% were high-way ready passenger vehicles. (Most of these were purchases by taxi and bus companies.) (Altenberg et al, 2015) Outside of the low-speed category (such as produced by Kandi and Shifeng), most sales were to local governments; 80% of new energy vehicles sales are in the public fleet, indicating failure thus far in the transformation from state support to a consumer market. (Howell et al, 2014) Although in the past few months the picture on sales has appeared brighter due to greater toleran of hybrids (Wang 2015), in general Chinese industrial policy on EVs has been unable to move the market for EVs forward on a par achieved by, for example, many European countries and California. (Mock and Yang, 2014)

What can we conclude about the role of local government in these sectors? First, it is important to reiterate that it remains incumbent on local officials in China to play a role in potential economic development schemes, such as SEI industries. Especially when the signals from Beijing that industrial policy should be pursued, local officials cannot sit idly by and allow markets to work things out (even if private firms such as Chang Cheng Auto can). And yet the ability of local governments to respond to Beijing’s signals in a positive and conforming way has been shown here to be largely a function of industry and market conditions. In short, the interaction of the two categories of variables denoted in the introduction – industry characteristics and the strength and nature of Beijing’s signals - has affected the development of the two industries considered in this paper. A simple taxonomy helps to illustrate this relationship.

	Industry structure (high barriers to entry, strong incumbents, large-scale cognate industries)			Need to respond to central signals (high priority of center)
	yes	no
strong	EV sedans (pure electric)
Moder-ate		Solar panels (lower priority signaled than EV sedans)
weak (locals pursue despite signals)	EV hybrids (post 2015)	Low-speed EV, electric bikes

This taxonomy perhaps contains an (intuitive) endogeneity, namely, that large incumbents (e.g., auto producers) are likely to be favored in Beijing’s policy signals – as makes sense if we conceive of new energy policy as being largely about industrial policy. Local governments are likely to be more creative, as in the case of low-speed EVs, or to sponsor rapid market entry (as in EVs) where the signal from Beijing is not particularly strong. This may be create a context in which cost-innovation, and serving the low-end of the market, is most vibrant. On the other hand, we see a less innovative environment where barriers to market access are high and Beijing’s signals are strong. We cannot say this is the result primarily of local government; in the case of EV sedans, the problems China faces are the same as elsewhere in the world. However, local government efforts, particularly to create market demand through government purchases or to compete to produce standards that will not become national standards, do not help overcome these problems. Note that the taxonomy ignores (perhaps mistakenly) the role of firm and industry ties to global markets.

In conclusion, then, when has local government involvement proven most helpful in the specific cases considered in this paper? When local market infrastructure and conditions are favorable, especially where existing technology and value chains permit firms to serve the low end segment of the market, local governments have been instrumental. With low barriers to entry and extant industrial infrastructure, provincial and municipal officials have supported market entry by local firms, and presumably have gained some political benefits – employment, growth – for themselves. This dynamic was evident in the low-speed EV industry. Kandi, under the rubric of the 10 Cities pilot program and the help of Zhejiang and Hangzhou (and eventually Shanghai), and Shifeng, under the protection of Shandong government. Kandi in particular has engaged in business process innovation, with battery switching and rental models that grew out of Hangzhou’s previous experiences with bicycles. (Though not explored in this paper, the story may be similar in EV bicycles.)

Early on in its development, China’s solar panel industry – also facing low barriers to entry and the absence of massive incumbents – in some cases appears to have benefited from local government help, especially in the form of subsidies, government facilitation of financing schemes, and contributions of land. When government behavior in the solar panel industry (as with low speed EVs) could support an existing market, i.e., be market conforming, local officials appeared helpful. Yet once the export market collapsed, and NDRC supported an end to many supports of solar firms (by limiting bank loans), local officials in some cases were reluctant to allow firms they had supported to fail, i.e., to allow market exit.

In the higher end segment of the EV market, for sedans and buses, we have seen strong signals from Beijing – as in the 10 Cities pilot program - and efforts by local governments to respond. Yet efforts by local officials have been much less successful. In this segment, the technological barriers to entry by firms are high, and the sector is dominated by incumbents (such as BYD). Consumer demand is low, due to high costs and relatively poor quality compared to combustion engine sedans. Governments have not only provided supply side support through subsidies and contributions of land and loans, but also have made extensive attempts to stimulate the market through industrial policy tools on the demand side, in the form of major consumer subsidies from central and local governments, government procurement, and government support of infrastructure (in terms of charging stations). Despite these efforts to conform to industrial policy signals – efforts that merely by virtue of being made can have political payoffs for local officials - the market has remained moribund.⁷⁰ Indeed, local government behaviors have exacerbated problems long criticized by Beijing (even if caused by some of its policies): duplication of investment and local protectionism that hinders the growth of a more integrated national market. In the EV case, efforts by local governments to create a market, and thereby ameliorate market failure, could not succeed due to industry and market conditions (lack of access to technology, high cost relative to where market demand lies) not found uniquely in China but perhaps more difficult to overcome there.

Some observers point out that local entrepreneurs are complicit in these failures. Local entrepreneurs, even if not parasites on government funding, often seek to exploit the funds available only through local governments (available only through local governments because that is how central support funding flows, and due to the absence of private capital). Indeed, some critics of Suntech’s failure do not single out Wuxi’s government as the problem but, rather, its founder Shi Zhengrong. Similarly there is equal criticism handed out to LDK’s founder Peng Xiaofeng and the Xinyu government. It is clear there is a symbiotic relationship between entrepreneurs and local governments. Others have suggested that the quality of local leadership is extremely important; Baoding possessed quite impressive leadership that created a better ecosystem, whereas Xinyu (in its treatment of LDK) did not.

Moreover, many problems– such as the difficulty in generating consumer-driven markets for new energy – are far from unique to China, though in some respects (e.g., problems related to grid interconnection and the existence or private financing to substitute for local government financing) may be felt more intensively there. There are, then, many fingers to be pointed when looking to problems in the local ecosystem for innovation and upgrading (or even basic firm success) in clean technologies. Perhaps the most pertinent issue is that some sectors remain dominated by industrial policy concerns –seen especially in the EV sector – making Beijing just as complicit as the local government officials who respond to its signals. The previous discussion indicates that China’s industrial policy approach to the promotion of new energy and sustainability, and the intensive role played by local governments in this institutional structure, does not solve the problems of market failure. Local government’s deep involvement has not been able to correct for market failures in solar (in which failure is largely due to exogenous demand forces that forced local producers to try to create domestic demand) and EVs (where other barriers such as price and grid problems create little organic market demand). At the same time, local governments have participated in the political failures that have in turn have to tremendous waste (the pressure to create “false brands” and difficulties tolerating market exit) and to problems of developing a national market (local protectionism and an insufficient grid network).
References
Ahlers, Anna L. and Gunter Schubert. 2009. “’Building a New Socialist Countryside’: Only a Political Slogan?” Journal of Current Chinese Affairs 36: 4.
Ahrens, Nathaniel. 2013. “Case Study: Suntech.” China’s Competitiveness: Myth, Reality, and Lessons for the United States and Japan. Washington, DC: CSIS.
Altenburg, Tilman, Eike W. Schamp, and Ankur Chaudhary. 2015. “The emergence of electromobility: Comparing technological pathways in France, Germany, China and India.” Science and Public Policy No. 10. doi: 10.1093/scipol/scv054.
Amsden, Alice H., Asia’s Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, 1989).
Arrow, Kenneth J. 1962. “Economic Welfare and the Allocation of Resources for Invention.”

in The Rate and Direction of Inventive Activity: Economic and Social Factors. National Bureau of Economic Research, ed. Princeton: Princeton University Press pp. 609–626.