“…Reflecting this situation, highly developed statistical models are often utilized ( [2], [11], [12]). As for Japan, some use aggregate data ( [13], [14]), others use micro data ( [15], [16], [17]) to find a statistically significant negative price elasticity As Gilleski [5] describes, many studies, which consist of a Probit equation for the probability that people consult doctors or linear equations for medical expenditures, implicitly assume that people decide to consult a doctor or not before they get sick. This might not be realistic, and dynamic decision making should be incorporated.…”