Trade in green patents: How do green technologies flow in China?

Abstract

Green technology transfer can help narrow the regional differences in green innovation, thereby contributing to a more coordinated green development. This study uses the social network analysis approach to explore the characteristics and regional differences in green technology diffusion in mainland China from 1985 to 2021. A unique dataset of patent transactions was used to construct green technology transfer networks. The findings demonstrate that energy-related technologies had the highest demand in the market. In addition, private sectors, especially multinational enterprises, were the most active network entities, whereas universities played a limited role in diffusion. Moreover, green patent transactions were highly localized and presented regional disparities. Resource-based regions with high pollution had lower green technology flows and formed a path dependence, whereas a few developed regions served as influential spreaders. The network exhibited a core-periphery pattern and dis-assortativity, thus creating a Matthew effect and widening the regional gaps. The absorbers and beginners tended to form connections with bilateral spillover, while peripheral provinces faced delinked risks. These findings help us understand the regional disparities and diffusion patterns of green technologies in China, thus accelerating the diffusion of green technologies.

Appendix: The formulars of the metrics

• (1) Network Density: $D=\frac{m}{n\left(n-1\right)}$,

  where m refers to existing links, while $n\left(n-1\right)$ refers to the maximal number of links in the network.

• (2) Network transitivity. $C = \frac{{\left( {3 \times number\,{ }of{ }\,triangles{ }\,in{ }\,the{ }\,network} \right)}}{{number\,{ }of\,{ }connected\,{ }triples\,{ }of\,{ }vertices}}$

• (3) Clustering coefficient. ${C}_{i}=\frac{2{E}_{i}}{{k}_{i}\left({k}_{i}-1\right)}$ ,

  where ${E}_{i}$ is the number of triangles through node i, and k_i is the degree of node i.

• (4) Average degree: $k=\frac{1}{N}\sum_{i=1}^{n}degree({V}_{i})$,

  where N is the number of nodes in the network.

• (5) Average path length: $L=\frac{1}{n(n-1)}{\sum }_{i,j=1.i\ne j}^{n}{d}_{ij}$,

  where ${d}_{ij}$ is the distance between node i and node j. It is obtained by counting the number of links that compose the shortest path connecting two nodes.

• (6) Node in-strength: ${s}_{i}^{in}=\sum_{j=1}^{n}{w}_{ji}$

  Node in-strength: ${s}_{i}^{out}=\sum_{j=1}^{n}{w}_{ij}$

• (7) Centrality: ${k}_{i}\times \frac{1}{{cc}_{i}+\frac{1}{{k}_{i}}}+{\sum }_{j\in {N}_{2}}{cc}_{j}$ ,

  where ${k}_{i}$ and ${cc}_{i}$ respectively refer to the degree and clustering coefficient of node i. $\frac{1}{{k}_{i}}$ underlines the inverse clustering coefficient of the node, and ${\sum }_{j\in {N}_{2}}{cc}_{j}$ represents the sum of clustering coefficients of the second-level neighbors for node i.

• (8) Intra-block spillover index: ${P}_{{B}_{i}}^{intra}=\frac{{t}_{ii}}{\sum_{j=1}^{M}{t}_{ij}} ,$

  where ${B}_{i}\mathrm{ means the block }{B}_{i}$, $M\mathrm{ is the total number of blocks},\mathrm{ and }{t}_{ii}$ is the number of patents transferred within block ${B}_{i}$. $\sum_{j=1}^{M}{t}_{ij}$ means the total number of green patents sent by block ${B}_{i}$.

• (9) Inter-block inflow index: ${P}_{{B}_{i}}^{inter}=\frac{{{\sum }_{j=1,j\ne i}^{M}t}_{ji}-\sum_{j=1,j\ne i}^{M}{t}_{ij}}{Total \, number\, of \, transferred \, green \, patents}$,

  where ${{\sum }_{j=1,j\ne i}^{M}t}_{ji}-\sum_{j=1,j\ne i}^{M}{t}_{ij}$ represents the net inflow green patens in block ${B}_{i}$.