The effect of employer tenure on wages in Japan

We estimate the effects of employer tenure on wages using the instrumental variable (IV) methodology by analyzing data from the Japan Household Panel Survey and Keio Household Panel Survey from 2004 to 2014. We find that the estimates based on the IV method, which corrects for omitted variable bias due to individuals' unobserved abilities and unobserved matching qualities, are significantly lower than those found using the ordinary least squares methodology. These results are robust across subsamples and specifications and cannot be rejected using an alternative estimation method. Our findings suggest that the recent returns on employer tenure in the Japanese labor market are likely to be lower than those discussed in previous studies.