|Author Name||ASAI Yukiko (University of Chicago / Waseda University) / Dmitri K. KOUSTAS (University of Chicago)|
|Creation Date/NO.||August 2021 21-E-071|
|Research Project||Productivity Effect of HRM Policies and Changing Employment System|
|Download / Links|
How does initial placement in a temporary work contract affect workers' subsequent labor market outcomes? To provide insight on this question, we study a unique set of natural experiments: In the mid-1990s, the major firms in the Japanese airline industry changed personnel policy so that all newly-hired flight attendants were employed on temporary employment contracts, while incumbent workers remained on permanent employment contracts. Workers who were hired on temporary contracts and remained at the firms were transitioned to permanent contracts after three years; however, their salaries were approximately half of those permanent workers in their first three years, and the policy change made it difficult for workers to take parental leave for the first four years. The reverse policy change occurred in the mid-2010s: the major firms transitioned back to hiring their flight attendants on permanent contracts. The firms also transitioned all incumbent flight attendants on temporary contracts to permanent contracts. These industry-level hiring practices provide natural experiments that can be used to study whether starting on temporary contracts matters for long-run career and family outcomes for young workers. Our research design uses the universe of employment records from one of Japan's major airlines as well as government surveys to compare outcomes for cohorts of flight attendants hired just before to those hired just after these changes in industry policy. The first key outcome we study is job separation: we find that workers starting on temporary contracts were less likely to remain with the firm over time. These separations appear to have been initiated by workers, not by firms. We are also able to study the effects these policies had on family formation. We find that workers starting on temporary contracts are significantly less likely to have children within 10 years after starting the job. These findings do not appear to be the result of selection on observables.