Location Is Everything: The Promises and Limits of China’s Development Zones

Trustee Chair in Chinese Business and Economics  >  Trustee China Hand   

As described in the recent Trustee Chair report The Power of Innovation, a core part of China’s innovation strategy in the post-reform era has been place-based economic policies. These include three levels of localities: (1) defined industrial parks and zones; (2) more loosely defined greater municipalities, such as the greater Guangzhou area; and (3) large macro-regions that incorporate multiple provinces, such as the Jing-Jin-Ji region (meant to integrate Beijing, Tianjin, and Hebei) and the Yangtze River Economic Belt. Regardless of scale, regionally oriented innovation policies are meant to improve the overall business climate, develop a dense concentration of specific capabilities, and bring together multiple parts of supply chains. The goal is for these efforts to strengthen hard and soft infrastructure, spur innovation, raise economies of scale, and generate spillovers to other parts of the economy.

These efforts do not always pay off, but China is focused heavily on each component. This blog post first summarizes trends regarding defined economic development zones. It then provides information about the relative performance of greater municipalities. And finally, it introduces China’s early efforts to promote regional clusters for artificial intelligence.

Economic Development Zones

China did not create the idea of defined economic zones; it borrowed extensively from the experiences of Japan, South Korea, and Taiwan, and, further afield, from those of the United States and Europe.

The Reform Era began with selective opening up in particular regions, most notably the establishment of Special Economic Zones (SEZs) 1980–81, which were geared toward attracting technology and investment from foreign investors and businesses. Shenzhen is perhaps the most well-known of these SEZs, having gone from a minor settlement bordering Hong Kong to a center of China’s burgeoning tech sector today. SEZs continue to operate today, and China has since established a number of other kinds of zones, such as science parks, economic and development zones, and high-tech zones. These are larger geographic areas that typically contain smaller industrial parks (工业园) that cater to specific clusters of companies.

The number of these special zones has proliferated in recent years. According to the central government’s last full-fledged audit of development zones in 2018, there were 2,543 development zones, including 552 central-level and 1,991 provincial-level zones. And, according to the Ministry of Industry and Information Technology (MIIT), as of 2025 there were 179 high-tech zones (高新技术产业开发区), which cumulatively accounted for 14.5 percent of China’s national GDP.

Data from private Chinese research group Qianzhan shows that when looking at the full spectrum of economic development zones, their number has continued to grow even further since the government’s last official count in 2018. Qianzhan identified over 77,339 parks (including logistics, cultural, and e-commerce parks) throughout the country, with 13,904 industrial parks specifically (see Figure 1). While national policies have increasingly focused on creating economic opportunities in central and western provinces, the majority of such zones remain concentrated in the more developed, eastern provinces, such as Guangdong, Jiangsu, Shandong, and Zhejiang. This is at least in part driven by the provinces themselves, which have been eager to establish their own provincial-level zones to complement and reinforce national policies.

Of course, while the number of such zones tells us about the general use of the tool, there is also the question of the underlying goals and effectiveness of these policies. For example, industrial parks have historically served as an important engine for China’s tech sector and a key way to attract and absorb foreign investment and expertise. One of the most notable examples is Suzhou Industrial Park, formed in 1994 as a cooperative venture between the Suzhou municipal government and Singapore. While the zone came under considerable pressure in the aftermath of the Asian Financial Crisis, it still came to be an important driver of the city’s growth. In 2014, while the zone only accounted for 3.4 percent of the total land, it still contributed around 15 percent of the city’s GDP, foreshadowing the similar numbers we see today in zones nationwide.

At the same time, industrial parks are by no means silver bullets. Research from MIT found that, among 110 industrial parks surveyed near major cities, 70 percent were associated with increases in productivity in surrounding areas, largely driven by human capital investment. However, spillovers failed to materialize, or were actually somewhat negative, for the other 30 percent of parks. One potential reason for the mixed record is China’s political economy; researchers found that city officials with connections to key provincial decisionmakers were more likely to obtain industrial parks, as well as that parks chosen for political reasons rather than based on their economic fundamentals generated lower economic benefits.

Regional Innovation Hubs

While China’s experience demonstrates that state-driven regional policy can be a potent tool, urbanization and regional clusters in and of themselves are powerful drivers of innovation globally. Because of this, another way to evaluate region-oriented policies is to focus on large metropolises rather than specifically on defined development zones. The World Intellectual Property Organization is known for its research on innovation clusters. More recently, Tsinghua University’s Center for Industrial Development and Economic Governance has started issuing its own “Global Innovation Hubs Index,” which shows significant variation in the performance of Chinese cities when compared globally using common benchmarks.

The index evaluates hubs based on three criteria: research innovation, innovation economy, and innovation ecosystem. Research innovation looks at indicators like the number of world-leading universities, the number of highly cited papers, and total citations. Innovation economy looks at the commercialization of innovation through metrics like the total number of valid patents, the number of “leading innovative companies,” and fundamentals like GDP growth and labor productivity. Finally, the innovation ecosystem category looks at openness and collaboration, as well as the enabling environment, through indicators like venture capital (VC) and private equity (PE) funding, the provision of related public services, and the availability of professional talent.

Their research found that Chinese innovation hubs tended to perform well across measures in the “research innovation” and “innovation economy” categories. While the driving variables differ between hubs in China, generally China’s innovation hubs had strong knowledge production in the form of both patent filings and highly cited papers and were strong across measures of commercialization, such as market value of high-tech manufacturing and revenue from new-economy industries. Compared to U.S. and European cities, Chinese hubs tended to have lower marks for their “innovation ecosystem”—largely due to comparatively low levels of VC and PE funding and lower levels of professional talent inflow and average years of schooling.

While the performance of China’s innovation hubs is impressive—particularly for the top three cities shown in Figure 2—the global index also shows the power of knowledge clusters around the world in places without the same intense level of support for localities, such as in the United States, Japan, and Europe.

AI Pilot Zones

A growing area of focus for economic development zones in China is artificial intelligence (AI). China launched its “National AI Innovation Pilot Zones” in 2019, and 18 cities have been designated for such zones so far, many overlapping with the previously established high-tech zones.

According to China’s Ministry of Science and Technology, AI pilot zones are intended to reinforce innovation in areas that are already centers of AI development, with a focus on applications that MIIT sees as important for broader economic and social development, including finance, logistics, and robotics. They offer direct financial support, access to enabling energy and data infrastructure, and a favorable regulatory environment.

A potential challenge for AI parks is conceptual—while place-based economic policy has to a certain degree always relied on the benefits that come from bringing talented people into close proximity to each other, this is even more the case in an innovation-driven economy, which Beijing is eager to have replace its previous investment-driven growth model. While the government is familiar with what it takes to support the physical infrastructure of AI, the design of the zones, with prescribed focus areas, is reflective of a top-down approach that may not produce the kind of results that might emerge from a more organic, market-oriented approach.

Conclusion

China has spent almost half a century learning from both global best practices and its own experiences with place-based industrial policy. Whether or not place-based initiatives can continue to deliver across the country and how much growth such an approach can generate remains to be seen. It is also possible that China’s urban clusters will continue to become powerful incubators of technology and engines of growth in spite of top-down initiatives.

Related Trustee Chair Activities

Scott Kennedy, The Power of Innovation: The Strategic Value of China’s High-Tech Drive (Washington, DC: CSIS, March 2026), https://www.csis.org/analysis/power-innovation-strategic-value-chinas-high-tech-drive.

“China’s Great Tech Leap Forward and the Implications for the United States” (public event, CSIS, Washington, DC, March 2, 2026), https://www.csis.org/events/chinas-great-tech-leap-forward-and-implications-united-states.

Ilaria Mazzocco and Ryan Featherston, “Wins and Losses: Chinese Industrial Policy’s Uneven Success,” Big Data China, November 19, 2024, https://bigdatachina.csis.org/wins-and-losses-chinese-industrial-policys-uneven-success/.

Ilaria Mazzocco, “Unpacking Linkages Between the Chinese State and Private Firms,” Big Data China, March 21, 2024, https://bigdatachina.csis.org/unpacking-linkages-between-the-chinese-state-and-private-firms/.

Qin (Maya) Mei and Matthew Barocas, “Visit to a Shenzhen Tech Market: Imitation Before Innovation,” CSIS, Trustee China Hand, October 23, 2023, https://www.csis.org/blogs/trustee-china-hand/visit-shenzhen-tech-market-imitation-innovation.