财经问题研究

深刻理解产业政策的运行特征是促进产业政策有效转型的重要前提。本文从政府工作报告中提取关键词，以识别2003—2022年省、市级政府的产业政策，全方位、多角度刻画中国地方产业政策的特征事实。研究发现：产业政策数量与经济发展阶段密切相关，产业政策数量随着人均国内生产总值和市场化指数的增加而减少；产业政策实施年数普遍较短且广泛存在实施间断情况；产业政策行业分布差异显著，产业政策实施数量和产业政策实施年数在行业间、行业内都分布不均；产业政策在同级政府间与上下级政府间均具有相似性；产业政策网络连接较为松散且聚类系数不断下降。本文的研究为中国产业政策研究奠定了数据基础。

A comprehensive understanding of the operational characteristics of the industrial policy is a crucial prerequisite for promoting its appropriate and effective transformation. Existing literature has primarily focused on evaluating the effectiveness of the industrial policy. However, the empirical findings vary widely. The long-standing controversy over policy effectiveness stems mainly from the difficulty in accurately measuring its effects. This paper attempts to summarize the stylized facts of local industrial policy by leveraging rich textual data from government reports. First, this paper extracts and measures industrial policies from provincial and municipal government work reports from 2003 to 2022, yielding 1 257 provincial-level and 14 309 municipal-level policy entries, respectively. Then, based on this extensive dataset, this paper summarizes the basic characteristics of China's local industrial policy in the dimensions of time, industry, government level, and region, providing policy implications to promote industrial policy transformation. This paper summarizes the stylized facts of China's local industrial policy into five aspects. First, the quantity of industrial policies is closely associated with the stage of economic development. For both provincial and municipal governments, the number of industrial policies decreases as GDP per capita and the marketization index increase. However, the number of policies for strategic emerging industries(SEIs) grows with these same indicators. Second, the implementation duration of industrial policies is generally short. The average implementation lengths for provincial and municipal policies are 2.290 years and 2.280 years, respectively. Concurrently, policy implementation is characterized by widespread discontinuity. Third, the quantity and duration of policies vary significantly both across and within industries. Notably, the number of policies targeting high-end manufacturing, SEIs, and technology-intensive industries is increasing annually. Fourth, industrial policies exhibit meaningful similarity among peer governments as well as across hierarchical levels. Similarity among governments at the same administrative level is lower when policies are classified at the four-digit industry level than at the two-digit level. The industrial policy of cities within the same province displays slightly higher similarity than that of cities overall. Vertical similarity between upper-lower-level governments follows a U-shaped pattern. Fifth, industrial policies exhibit a core-periphery network structure. Core-node cities are few but highly connected, while peripheral cities are numerous but sparsely connected.The policy implications of this paper are as follows. Industrial policies should be timely in both their implementation and cessation; long-term perspectives and coherence should be emphasized; local governments should adhere to a path of specialization and differentiation, while rationally aligning with the developmental intentions of higher-level governments.