Robustness Test Dataset for the Impact of Trade Secret Protection Innovation Pilot Policy on the Cultivation of Specialized, Refined, Distinctive, and Innovative (SRDI) Enterprises at the City Level

This dataset contains robustness test regression results examining the impact of the Trade Secret Protection Innovation Pilot Policy (launched by the State Administration for Market Regulation in 2022-2023) on the cultivation of Specialized, Refined, Distinctive, and Innovative (SRDI) enterprises at the city level. The data is sourced from the appendix tables of the research paper "Trade Secret Protection and the Cultivation of SRDI Enterprises."Research Background and Data Generation:Based on panel data covering 294 Chinese cities from 2019 to 2023, this study utilizes the quasi-natural experiment of the trade secret protection innovation pilots (implemented in 2 batches: July 2022 for the first batch, October 2023 for the second batch) and employs a multi-period Difference-in-Differences (DID) model to evaluate policy effects. To verify the reliability of baseline regression results, this dataset aggregates estimation results from seven robustness testing methods, including: Propensity Score Matching-DID (PSM-DID), replacement of dependent variables (absolute count vs. relative density of SRDI enterprises), exclusion of confounding policies (controlling for National Intellectual Property City Pilots and National Independent Innovation Demonstration Zones), inclusion of provincial-level omitted variables (tax burden, fiscal support intensity, industrial structure, urbanization rate), and adjustment of cluster-robust standard error levels (from city-level to city-year two-dimensional clustering).Data Content and File Structure:The data table contains 7 columns of robustness test results, with each column representing a specific model specification:Column (1): PSM-DID matched sample to control for selection biasColumns (2)-(3): Replacement of dependent variables using natural logarithm of SRDI enterprise count (y1) and SRDI enterprises per 100 residents (y2) to control for city scale differencesColumns (4)-(5): Confounding policy tests introducing dummy variables for "National Intellectual Property City Pilot" (policy) and "National Independent Innovation Demonstration Zone" (policy1) to exclude interference from concurrent policiesColumn (6): Considering provincial omitted variables by including provincial-level macroeconomic controls: tax burden, fiscal support, industrial structure, and urbanization rateColumn (7): Adjusting cluster level by changing standard error clustering from city-level to city×year level to address cross-temporal correlationVariable Definitions:Trade_Secret: Core explanatory variable representing the interaction term (Treat×Post) of the trade secret protection pilot policy treatment effect; coefficients reflect the average treatment effect on SRDI enterprise cultivationy/y1/y2: Dependent variables representing natural logarithm of city SRDI enterprise count, absolute count (unlogged), and SRDI enterprises per 100 residents, respectivelypolicy/policy1: Confounding policy variables representing intellectual property city pilot policy and independent innovation demonstration zone pilot policy, respectivelyTax Burden Level: Provincial-level tax burden indicator (general budget revenue/GDP)Fiscal Support Intensity: Provincial fiscal expenditure intensity indicatorIndustrial Structure: Ratio of tertiary to secondary industry value-added, reflecting industrial upgradingUrbanization Rate: Provincial urban population proportionControl Variables: Include economic development level (log of population density), industrial structure upgrading, human capital level (per capita university enrollment), public service level (public library books per 100 people), traditional infrastructure (log of railway freight volume), internet development level (log of telecom business volume), financial development level (loan balance/GDP), fiscal pressure level (expenditure/revenue ratio), and science expenditure level (log)Fixed Effects: City fixed effects and year fixed effectsSample Size: Valid observations for each model (ranging from 576 to 1,457)R-squared: Model goodness-of-fit statisticsData Value and Applications:This dataset provides multi-dimensional, multi-method robustness evidence for policy effects, which can be used to verify the causal effects of trade secret protection systems on SME innovation cultivation. It offers empirical evidence for intellectual property policy evaluation and regional innovation policy design, suitable for academic research and policy analysis in economics, management, and public policy fields.