This paper replicates models developed by previous research to study the effects of graduate education on new graduates' initial employment in the Japanese labor market. If education is the best investment for an individual's economic success, then graduate degrees are expected to provide an individual with higher-earning job opportunities. Despite this reasonable economic premise, previous research showed that master's degrees in the humanities or social sciences in Japan have, in fact, a negative impact on obtaining initial employment compared to those with only a bachelor's degree in the humanities or social sciences. This previous research, however, could not overcome omitted variable bias because of data limitations. Omitted variable bias is a key problem for research on education; therefore, this study uses new longitudinal data to overcome omitted variable bias and clearly demonstrate the robustness of these earlier findings. The empirical results of this study corroborate earlier work, showing that master's degrees in the humanities or social sciences do not provide graduate students with an advantage in obtaining initial employment, after controlling for potential bias. At the same time, this study also confirms that natural science majors have a higher probability of obtaining initial employment in comparison with humanities or social science majors. In other words, this paper offers a valid replication of existing research. This shows that the Japanese labor market structure for graduate students has, in essence, remained the same since previous research was completed.