Tax Incentives and Firm Skill Composition: Evidence from China's Employee Training Deduction Reform
China & World Economy · 2026-03 · 34 (2) · 32-66
DOI: 10.1111/cwe.70016 ↗At a glance
Journal article in China & World Economy, published online on 2026-03-16. Written by Ce Huang and 3 co-authors. 19 references.
Abstract (original language)
This study investigated how changes in China's tax policy regarding firms' employee training expenses affected the proportion of high‐skilled employees. Focusing on a 2018 tax reform that raised the pretax deduction limit for employee training expenses, it analyzed data from listed companies using a difference‐in‐differences approach. The reform increased the share of high‐skilled employees in affected firms. The proportion of technical personnel rose by 1.22 percentage points, with stronger effects in firms that had higher pre‐reform training expenses, lower financial constraints, and greater capital intensity. Internal training and employee upward mobility were the main drivers, with a 24 percent rise in average training expenses per employee, primarily for front‐line staff. There was a modest increase in research and development intensity and no reduction in average wages. Targeted tax incentives can thus encourage human capital investment and enhance workforce skill composition without adverse wage effects.
Cite
APA
Huang, C., Jin, X., Liu, R., & Wang, R. (2026). Tax Incentives and Firm Skill Composition: Evidence from China's Employee Training Deduction Reform. China & World Economy, 34(2), 32-66. https://doi.org/10.1111/cwe.70016
Vancouver
Huang C, Jin X, Liu R, Wang R. Tax Incentives and Firm Skill Composition: Evidence from China's Employee Training Deduction Reform. China & World Economy. 2026;34(2):32-66. doi:10.1111/cwe.70016
BibTeX
@article{Huang2026,
author = {Huang, Ce and Jin, Xiaoshu and Liu, Rong and Wang, Ruiting},
title = {Tax Incentives and Firm Skill Composition: Evidence from China's Employee Training Deduction Reform},
journal = {China & World Economy},
year = {2026},
volume = {34},
number = {2},
pages = {32--66},
doi = {10.1111/cwe.70016}
}
EUVIMED