Our work aimed to build a reasonable proxy for unmet medical demands of China's urban residents. We combined health demand modeling and stochastic frontier analysis to produce a frontier medical demand function, which allowed us to disentangle unmet medical demands from the unobservable effects. We estimated unmet medical demands by using China's provincial dataset that covered 2005-2018. Our estimates showed that unmet medical demand at the national level was 12.6% in 2018, and regions with high medical prices confronted more unmet medical demands than regions with moderate or low medical prices during 2005-2018. Furthermore, medical prices and education were the main factors that affected unmet medical demand; therefore, policy making should pay more attention to reducing medical costs and promoting health education.
Keywords: Mundlak specification; frontier demand function; medical demand; the unmet medical demand.