Does gender affect medical decisions? Results from a behavioral experiment with physicians and medical students

Abstract

It is rarely the case that differences in behaviors of females and males are described under a ceteris paribus condition, and behaviors can potentially be influenced by the environment in which decisions are made. In the case of medical decisions, physicians are expected to account for patient characteristics as well as observed and unobserved contextual factors, such as whether the patient has a healthy lifestyle. Since one usually do not randomize physicians to context, reported gender differences in medical practice can have several alternative interpretations. A key question is whether the medical treatment of a given patient is expected to depend on the gender of the physician. To address this question, we quantify gender effects using data from an incentivized laboratory experiment, where Chinese medical doctors and Chinese medical students choose medical treatment under different payment schemes. We estimate preference parameters of females and males assuming decision-makers have patient-regarding preferences. We cannot reject the hypothesis that gender differences in treatment choices are absent. Preference parameters of females and males are not statistically different in a log-likelihood ratio test, and there is no evidence that the degree of randomness in choices differs between genders. The absence of gender effects in the laboratory, where choice context is fixed, provides nuance to previous findings on gender differences, and highlights the general difficulty of separating individuals’ behavior from the context they are in.