Answers are grounded in this paper and cite their sources.
Digitalization can disrupt production before complementary organizational changes begin to pay off. Using panel data on 30 Chinese provincial-level units for 2013–2024, we estimate a first-difference distributed-lag model with year fixed effects and province-clustered inference. Growth in the Digital Economy Index (DEI) is associated with an immediate decline in industrial labour productivity and a substantial but incomplete offset one year later. The immediate decline recurs across alternative lag lengths, polynomial and autoregressive distributed lags, local projections and controls for differential provincial trends; the offset is clearest when lag weights are unrestricted, and the data reject the smoothness restrictions that weaken it. Coefficients beyond the first lag are unstable. The cumulative association is negative through the second lag and imprecise thereafter, and no specification yields a significantly positive cumulative effect within the observed horizon. A one-standard-deviation rise in annual DEI growth is associated with an immediate productivity loss of about 1.1%, roughly one-fifth of average annual productivity growth. The evidence describes an early adjustment-cost phase rather than a completed productivity J-curve. The larger per-unit association in initially less digitalized provinces reflects diminishing returns to the level of the index. Instrumental-variable, procurement and innovation-participation estimates are reported as supplementary associations.
No takes yet. Share an insight, caveat, or question.
Weiran Zhang (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: