Urban digital infrastructure in China is raising skill premiums by complementing high-skilled labour, increasing demand for skilled workers across both skill-intensive and non-skill-intensive industries; deepening digital capital enables skilled workers to contribute more to output than material capital.

Main Finding

Urban digital infrastructure (UDI) is skill‑biased: its development raises the relative demand for high‑skilled workers and therefore increases the skill premium. This operates through complementarity between UDI-related digital capital and high‑skilled labor, with digital capital amplifying the productive contribution of high‑skill workers more than material capital does.

Key Points

• UDI exhibits skill‑biased technological attributes: digital infrastructure favors tasks and occupations requiring higher skills.
• UDI development increases demand for high‑skilled labor in both skill‑intensive and non‑skill‑intensive industries, changing the labor‑skill composition of cities.
• The rise in the skill premium is primarily driven by complementarity: digital capital and high‑skilled labor are complements, so marginal productivity (and wages) of high‑skill workers increase as UDI deepens.
• Digital capital has a stronger complementarities effect with high‑skilled labor than traditional material capital; deepening digital capital yields larger increases in high‑skilled labor’s contribution to output.
• The results imply that urban technology upgrades can be an important driver of wage inequality via changing returns to skills.

Data & Methods

• Theoretical framework: a model that distinguishes digital capital (linked to UDI) and material capital, and allows for differential complementarity with labor of different skill levels. The model predicts how UDI affects labor demand and skill premiums.
• Empirical data: individual‑level microdata from the China Labor Force Dynamic Survey combined with city‑level urban statistics on digital infrastructure (UDI) and other controls.
• Empirical strategy: estimation of the relationship between measures of UDI and skill premiums/wage gaps, assessing industry‑level heterogeneity (skill‑intensive vs non‑skill‑intensive) and testing complementarity between capital types and skill groups.
• Identification: leveraging cross‑city variation in UDI development and linking to individual wages and occupational skill composition (the summary does not specify instruments or quasi‑experimental sources; causal interpretation relies on model structure and observed variation).

Implications for AI Economics

• UDI as a general‑purpose digital capital: like AI and other digital technologies, urban digital infrastructure acts as a platform that disproportionately increases returns to high‑skill labor, so researchers and policymakers should treat digital infrastructure investment as a driver of labor‑market reallocation and wage inequality.
• Complementarity matters: models of AI adoption and digitalization need to incorporate heterogeneous complementarity between digital capital and worker skills (digital capital more complementary to high skills than material capital).
• Policy responses: to counteract rising skill premiums and potential inequality, complementary investments in human capital (education, retraining) and policies that expand access to high‑value digital tools across firms and workers are warranted.
• Urban and sectoral targeting: because UDI raises high‑skill demand across industries, urban infrastructure policy influences broader regional labor market outcomes—not only technology sectors—so local labor market planning should account for cross‑sectoral effects.
• Research extensions: quantify causal impacts using exogenous rollouts of UDI/AI infrastructure, examine heterogeneous effects by occupation/education, explore long‑run dynamics (task reallocation, automation vs augmentation), and compare the wage effects of digital capital (including AI) versus traditional physical capital.