Working Papers

• Spousal Labor Response to Primary Income: Identification and Heterogeneity (with  Elin Halvorsen  and Marios Karabarbounis). 

We present a new estimate for the elasticity of spousal labor supply in response to changes in the primary worker's income, the so-called ``added worker effect." By leveraging firm-side information of the primary worker as an instrument, we isolate income changes that are uncorrelated with the spouse's productivity, addressing endogeneity bias.  We find an economically meaningful role for the spousal labor supply, especially among young households with limited financial assets. We construct a heterogeneous agent model consistent with the estimated spousal employment response to design a government transfer program that effectively mitigates the negative income shock. Richemond FED's Economic Brief 

• Two Decades of Earnings Inequality and Dynamics in South Korea (with  Jong-suk Han, Hoyeon Jang, Donghoon Kim, Eun-Mi Kim, Jaiyong Kim, Minji Koo, Jungmin Lee, Jaeyong Seo). 

Using administrative data from the National Health Insurance Service and the National Tax Service, we document the evolution of individual earnings inequality in South Korea from 2002 to 2022. We find that South Korea experienced a sustained decline in income inequality throughout this period. While this trend was broad-based across genders and age groups, it was primarily driven by compression in the lower half of the distribution. To investigate the underlying drivers, we employ a ``beta'' approach to estimate heterogeneous responses to changes in the minimum wage and economic growth. Our results suggest that sustained increases in minimum wage are associated with a reduced inequality in the lower half of the distribution. Although economic growth benefited all groups, its impact on reducing earnings dispersion was modest. 

• Asymmetric Frisch Elasticity of Labor Supply (with  Sun-Bin Kim  and Junseok Lee). 

We propose an economic mechanism that generates asymmetric Frisch labor supply elasticities. Based on the Australian panel data (HILDA), we document two stylized facts on labor supply: (i) asymmetric distribution of hours gap between preferred and actual hours of work: a pronounced concentration of workers whose actual hours are larger than preferred ones (workers are on average over-employed), and (ii) a persistently declining age profile of preferred working hours (relative to actual hours) over the life cycle . We build a heterogeneous agent model that matches these facts and, as a result, generates asymmetric responses of hours at the aggregate level.  Our results justify a use of sizeable aggregate Frisch elasticities for the business cycle analysis and, at the same time, such labor supply schedule is still compatible with small elasticities found from the studies based on tax holidays.