Impact of artificial intelligence pilot policy on household medical burden: A quasi-experimental study based on the China Family Panel Studies
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Artificial intelligence policy is usually judged on innovation and output. What it does to the way households pay for medical care is a different question, and a far less examined one. This paper takes that question up. China’s National New Generation Artificial Intelligence Innovation and Development Pilot Zones were designated in three approval cohorts covering 18 cities, and that staggered rollout, paired with four waves of the China Family Panel Studies (2016–2022), gives us a quasi-natural experiment that can be taken to a staggered difference-in-differences design. Designation cuts real household medical expenditure by roughly 37% in the fully adjusted specification (\(\:\beta\:\) = −0.470 on log(1 + expenditure), \(\:\rho\:\) < 0.01; −0.372 with fixed effects alone). Catastrophic health expenditure (CHE) becomes about 6.0 percentage points less likely (\(\:\beta\:\) = −0.060, \(\:\rho\:\) < 0.01; −0.055 with fixed effects alone). Both estimates survive. Alternative clustering, winsorised controls, restricted samples, and placebo assignment of treatment status leave them where they stand. Insurance enrolment rises, household health status improves, and Internet use increases; each carries part of the effect. Gains in financial protection are not confined to particular institutional settings; formal interaction tests reject equality across subsamples only for long-term care insurance pilot status, and the effect is smaller where other health policies are already in force. Industrial policy aimed at AI may therefore double as social protection. How much of it households keep depends on the institutions around them.
Source: Research Square
Authors: Yilong Wang, Aihong He
Language: English
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7226029469
