In Dragon Innovation: Shenzhen, Grace Yuehan Wang explains Shenzhen’s transformation (from an agricultural county to China’s Silicon Valley) through the combination of firm-led R&D, efficient local networks and knowledge-sharing between stakeholders. This nuanced, ambitious book illuminates Shenzhen’s formula for sucess, though whether this would be replicable elsewhere remains unclear, writes Mohammed Afsal KH.
What explains Shenzhen’s success?
Shenzhen occupies a prominent position in global technological competition. Once an agricultural county bordering Hong Kong, it has evolved into a leading global science and technology cluster. While scholars such as Ezra F. Vogel emphasise the central state’s role in creating and protecting Shenzhen’s Special Economic Zone, Learning from Shenzhen, edited by Mary Ann O’Donnell, Winnie Wong and Jonathan Bach, highlights the city’s decentralised experimentation and policy exceptions, Grace Yuehan Wang’s Dragon Innovation: Shenzhen (2026) challenges this monolithic portrayal. A Visiting Fellow in LSE’s Department of Media and Communications, Wang studies technology and innovation in the Asia-Pacific and approaches Shenzhen through historical, political, institutional, business and social perspectives, drawing on interviews with officials, business leaders, investors, academics and entrepreneurs.
Shenzhen is home to major technology companies such as Huawei, Tencent, BYD and DJI, spanning telecommunications, internet services, electric vehicles and drones. Wang demonstrates that Shenzhen’s ascent cannot be explained by pure state direction or market autonomy. Instead, she presents a system in which firms conduct much of the R&D, local authorities build platforms to support them, and manufacturers, suppliers and entrepreneurs share knowledge and resources.
Wang is most convincing in demystifying the bottom-up, enterprise-led dynamics of Shenzhen’s R&D ecosystem and the tight coupling of design and assembly.
In Wang’s account, public authority remains central, but it is reconfigured: rather than directing industrial outcomes from above, the local state acts as a platform facilitator, using Special Economic Zone (SEZ) privileges, such as preferential tax rates and preferential land-use terms, to pilot local policy experiments, while private-sector enterprises drive R&D. Wang is most convincing in demystifying the bottom-up, enterprise-led dynamics of Shenzhen’s R&D ecosystem and the tight coupling of design and assembly. Her broader discussion of how Shenzhen’s developmental lessons can be adapted to other Global South contexts remains less fully demonstrated given the city’s unique, historically specific enabling conditions.
Wang shows that China’s regional innovation landscapes are heterogeneous; while Beijing’s R&D spending was roughly evenly split between government and corporate sources, with public funding supporting national technology projects such as Red Flag Linux, Shenzhen’s ecosystem was overwhelmingly enterprise-led. This enterprise-led structure is reflected in four municipal indicators: over 90 per cent of Shenzhen’s R&D institutions, researchers, funding, and patent applications are corporate-driven.
Wang highlights local cadres as “bureaucratic entrepreneurs” who acted as platform builders rather than administrative regulators, drawing on networks of private investors to facilitate innovation. These officials experimented with regulatory openings, including a 1987 document permitting private technology enterprises. A prime example is the “Small- to Medium-Size Technology Enterprise Rating Index,” co-developed with Peking University’s Shenzhen Research Institute, which uses a credit rating mechanism to help SMEs obtain bank financing, addressing information and credit constraints while preserving corporate autonomy.
Reconfigured networks and the commons
The physical anchor of this ecosystem is Huaqiangbei, a major electronics marketplace in Futian District, Shenzhen. Wang adapts Henry Jenkins’s concept of “participatory culture” into “grassroots participatory innovation” to explain the informal sharing of circuit designs, components and technical knowledge among Shenzhen’s merchants and makers. Through cases such as Seeed Studio, an open-source hardware company, and the HAX accelerator, a hardware-focused startup programme, she shows how Shenzhen’s dense hardware ecosystem brings entrepreneurs into close proximity with suppliers and manufacturers, supporting rapid product development.
Wang’s account of the close integration between design, prototyping and physical production can be productively interpreted through Gary Pisano and Willy Shih’s “industrial commons” framework. Pisano and Shih argue that offshoring physical production can erode a region’s shared web of technological know-how, supplier networks and operational skills. By preserving its manufacturing base, Shenzhen has maintained the physical infrastructure and engineering capabilities that enable rapid product iteration, illustrating how the erosion of the “commons” can be mitigated when design and assembly are co-located.
This spatial integration also shaped how Shenzhen reconfigured its knowledge networks. Rather than relying on a conventional research-university base to anchor R&D on-site, Shenzhen established alternative knowledge linkages through external academic connections and specialised outposts. These included the Research Institute of Tsinghua University in Shenzhen (RITS) and the Shenzhen Institutes of Advanced Technology (SIAT) for technology transfer, alongside vocational institutions such as Shenzhen Polytechnic University, which supplied technical talent. These institutions did not simply replace physical universities. Instead, they reconfigured university-industry linkages. This configuration illustrates Ron Boschma’s proximity theory, which argues that spatial proximity is neither necessary nor sufficient for interactive learning, as other dimensions, such as cognitive, organisational and institutional proximity, can facilitate coordination.
Geopolitics and the limits of transferability
This applied-heavy model, however, faces significant constraints under geopolitical strain. Shenzhen’s R&D expenditure has been historically concentrated in experimental development, with basic scientific research receiving less than three per cent of total R&D funding, specifically 0.2 per cent in 2009 and 2.9 per cent in 2016, according to Wang’s analysis of Shenzhen Statistics Bureau data. This low historical investment in basic science represents a potential limitation in foundational scientific capacity. Under geopolitical strain, restrictions on advanced technologies expose the importance of external technological capabilities. Rather than proving a direct causal chain, the supply chain pressures affecting Huawei illustrate this vulnerability, demonstrating how even highly efficient bottom-up co-creation networks can face significant constraints when access to external technologies is restricted.
The municipal state is attempting to address these bottlenecks through targeted talent programmes and policies encouraging reverse “brain circulation”, the return of highly skilled professionals from abroad to contribute knowledge and expertise to their home countries. Initiatives such as the Peacock Talent programme and research academies, including the Shenzhen Medical Academy of Research and Translation under structural biologist Yan Ning, are intended to strengthen the cluster’s frontier scientific capabilities.
Wholesale replication is less feasible than adapting individual policy instruments to recipient contexts.
The more ambitious claim of Wang’s work is that Shenzhen offers lessons for the Global South. Evaluated against David Dolowitz and David Marsh’s policy transfer framework, this claim faces clear limits. Wang’s evidence shows that Shenzhen’s trajectory was contingent on historically specific enabling conditions, including its proximity to Hong Kong, China’s domestic market scale, and SEZ political privilege that cannot simply be assumed elsewhere. Replicating this trajectory in other developing nations risks an inappropriate transfer. This suggests that wholesale replication is less feasible than adapting individual policy instruments, such as vocational curricula or local policy experiments that allow new rules to be tested on a limited scale before wider adoption, to recipient contexts.
Dragon Innovation complicates simplified narratives of both Silicon Valley exceptionalism and top-down state determinism.
Dragon Innovation complicates simplified narratives of both Silicon Valley exceptionalism and top-down state determinism. Its strongest contribution is to show how Shenzhen’s technological capabilities emerged from the interaction of enterprise-led R&D, active local governance, manufacturing networks, and reconfigured knowledge institutions. By demonstrating how design and manufacturing remain closely connected, it also offers a cautionary lesson to post-industrial economies that have offshored physical production under the assumption that they could cleanly retain R&D. For development scholars and industrial policymakers, the book provides a detailed account of the institutional arrangements and networks that built China’s leading technopole. Its discussion of transferability raises a broader question: how far can Shenzhen’s experience travel beyond the conditions that made it possible?
Note: This review gives the views of the interviewer and author, not the position of the LSE Review of Books blog nor of the London School of Economics and Political Science.
Image: kylauf on Shutterstock.
Enjoyed this post? Subscribe to our newsletter for a round-up of the latest reviews sent straight to your inbox every other Tuesday.
