This paper measures the effect of sorting on wage disparity. We extract worker skills by estimating the wage equation using Japanese individual data. We show that the skills of workers in metropolitan areas are 9.68% higher than the skills of workers in nonmetropolitan areas. In the counterfactual situation in which there is no migration (i.e., workers stay in their hometowns), the skill difference decreases to 5.56%. These results indicate that migration increases wage disparity by 22.0%. Thus, both hometown differences and migration behaviors are significant for wage disparities. Furthermore, features of migration sorting vary according to various worker characteristics

Dual agency, representing both the buyer and the seller in the same transaction, has recently garnered the public’s attention in Japan with an increasing doubt on its fairness and efficiency in the market. Although a number of empirical studies have examined the relationship between different types of agency and transaction prices, only few theoretical studies assess the pros and cons of the agency system exist. This paper aims to offer a fundamental model to analyze the role of dual agency as determinants of price and probability of a successful transaction. This model predicts that the price under dual agency transaction is higher than the price under cross-agency transaction, in which a buyer and a seller have different agents. The probability of transaction under the dual agency transaction is shown to be higher (lower) than the probability under the cross-agency transaction when the variation of sellers’ reservation prices is small (large). It is also shown that an agent, who has attempted to close a deal under dual agency transaction but failed to do so, has an incentive to significantly lower the price, instead of switching to a cross-agency transaction

Some retail gasoline markets exhibit Edgeworth cycles where a sharp price increase is followed by a series of gradual price decreases. We empirically analyzed retail gasoline prices in New South Wales (Australia) and showed what factors affect periods of Edgeworth cycles. First, we found that (A) an EC in an area tends to become longer as the number of stores per population in the area is larger. Next, to explain the generation mechanism of (A), we estimated stores’ behavior in pricing for each area. The estimation result showed that (B1) as the number of stores per population in an area is larger, stores in the area give more weight in pricing to “state average price” than “area average price” and (B2) as stores in an area give more weight in pricing to “state average price”, the EC in the area tends to become longer. Furthermore, we showed that (B1) and (B2) can explain (A)