2026-10-08T20:10:25Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34010784 2026-10-07T08:26:48Z category_25966 item_type_3 month_year_10_2026

Data and Code：Digital proximity and cross-boundary digital innovation in Chinese traditional manufacturing firms: Evidence from graph-assisted double machine learning Wenjing Li (21652226) Experimental economics digital proximity cross-boundary digital innovation traditional manufacturing firms structure learning double machine learning <p dir="ltr">This study examines the relationship between digital proximity and cross-boundary digital innovation using an unbalanced panel of Chinese A-share listed traditional manufacturing firms observed from 2010 to 2024. The unit of analysis is the firm-year. The objective is to estimate the covariate-adjusted association between firm-level digital proximity and cross-boundary digital innovation while flexibly accounting for observed high-dimensional and nonlinear confounding. The empirical framework consists of four stages. First, firm-level financial, patent, geographic, and regional data are collected and integrated. Second, a theory-guided candidate covariate pool is evaluated using structure learning together with firm-cluster bootstrap stability assessment. Third, the selected adjustment set is incorporated into a partially linear double machine learning framework implemented with firm-grouped cross-fitting. Finally, a series of prespecified robustness analyses, placebo tests, supplementary instrumental-variable analyses, channel-consistency analyses, and exploratory boundary-condition analyses are conducted.</p> 2026-10-07T08:26:48Z Dataset Dataset 10.6084/m9.figshare.34010784.v2 https://figshare.com/articles/dataset/Data_and_Code_Digital_proximity_and_cross-boundary_digital_innovation_in_Chinese_traditional_manufacturing_firms_Evidence_from_graph-assisted_double_machine_learning/34010784 CC BY 4.0