A moving window path analysis has been developed to explore the spatially varying NEET (Not in Education, Employment, or Training) factors in the Tokyo Metropolitan Area. Unemployment among young people has increased since the 1990s. With the lack of job opportunities for these people, the NEET problem is a growing concern in Japan. Previous research on NEET have highlighted that NEET factors take into account academic records, unemployment records, and whether or not young people are living with their parents. Since the Tokyo Metropolitan Area comprises sectors that have different characteristics, as in the inner city and the suburbs, a moving window regression is helpful in analyzing the spatially varying NEET factors. In order to explore the complicated relationships between these factors, moving window regression and path analysis are integrated to develop the moving window path analysis. A Monte Carlo test for spatial non-stationarity suggests that NEET factors spatially vary in the Tokyo Metropolitan Area. The results of the moving window path analysis reveal that junior high school graduates have a direct effect on NEET, mainly in the suburbs of the Tokyo Metropolitan Area. It is also pointed out that those NEET who do not express a willingness to work occupy the majority in the suburban areas. In contrast, mainly in the inner city of the Tokyo Metropolitan Area, junior high school graduates have an indirect effect on NEET due to unemployment. This implies that those NEET who express a willingness to work occupy the majority in the inner city. Based on these results, it can be stated that local governments need to implement substantive measures for NEET according to spatially varying factors.
