This paper examines how changes in wage structure and divorce laws in the United States influenced labor and marriage market outcomes from 1968 to 2018. We develop a general equilibrium overlapping generations model that incorporates female labor supply, education, and marital choices to analyze the impacts of rising college premiums, narrowing gender wage gaps, increasing wage volatility, and the shift from mutual consent to unilateral divorce regimes. We model both endogenous labor and marriage markets and allow for rich interactions between the two markets during the transition. To exploit both the rich empirical dynamics and the variation in state-level divorce law changes, we propose an innovative methodology and estimate the model by fitting connected equilibrium transitional paths to time-series data. This approach allows us to disentangle the effects of sequential exogenous driving forces. Our decomposition analysis reveals that the increasing college premium was the primary driver of delayed marriage and increasing high educational attainment, while divorce law reforms partially offset this trend. Before the 1990s, the returns from the marriage market predominantly incentivized higher education, while labor market returns became the key motivator for educational advancement thereafter. 

This paper estimates a dynamic discrete choice model to analyze how cash transfers affect schooling and marriage decisions among teenage girls in low-income countries. We focus on a two-year duration cash transfer program implemented in Malawi among school-age girls, and the model is estimated to fit policy effects on encouraging school attendance and delaying marriage. We use data of control group and unconditional cash transfers group for estimation, and reserve the conditional cash transfers group for out-of-sample validation. With the aid of the estimated structural model, we address the dynamic selection issue stemming from recruitment eligibility requirements (unmarried, in school girls) in the presence of unobservable heterogeneity in household preferences for girls’ schooling, and unveil the full policy effects by accounting for the impacts on the ineligible girls who would have chosen to stay unmarried and in-school with the cash transfers. We also use the model to simulate policy effects with the hypothetical longer duration cash transfers. The synergies between RCTs and fully specified, estimable structural models provide us convincing model identification and validation, as well as more reliable extrapolation of important policy effects. 

We examine the effects of marriage market competition when these markets allow participants from outside their traditional borders. Increased globalization has led to increased linkages across marriage markets that were autarkic for most human history. We examine the case of Taiwan, which in 1992 relaxed its immigration proceedings leading to an unprecedented surge of marriages between local Taiwanese men and mostly Asian, non-Taiwanese women who moved to Taiwan for the purpose of marriage. To understand the impacts of such a shock, we estimate a dynamic equilibrium model of the marriage market an investigate the impact of increased competition by non-local brides in a marriage market that opens up beyond its traditional boundaries. 

This paper builds a life cycle model with a limited-size endogenous marriage marketwhere meeting probabilities depend on searching efforts and the availability of singles inthe market. Individuals form rational expectations of single population distribution inthe future marriage market and make optimal searching and marital decisions accord-ingly. These optimal decisions conversely determine the evolution of populations, andtherefore establish marriage market outcomes as an equilibrium result. We argue thatthe model with this endogenous marriage market setting is able to match the hump-shaped marriage trend we observe in the Japanese Census and Vital Statistics data.Moreover, the model implies that marriage delay caused by any exogenous driving forceis self-reinforcing, as more available potential partners to meet in the future increasethe value of being single today. Marriage delay will be further enhanced as individualsreduce their searching efforts when everyone else is waiting longer. The counterfactualanalyses suggest that, in response to a negative shock to the average match quality, theupdates in beliefs regarding the future market will largely reinforce marriage delay, andthe adjustments in searching efforts play an auxiliary role. Similar results are obtainedby analyzing the impact of a wage decline, while now the searching efforts adjustment isquantitatively more important as the oppurtunity cost of search changes. On the otherhand, the boost in birth rates caused by monetary child support policy will be amplifiedby the endogenous marriage market effects.

This paper builds a life cycle model with human capital accumulation and frictionalmarriage search. The wage growth rates depend on the past working hours and areeducational specific. The marriage markets are separated by cohorts and have lim-ited sizes. The marriage search is costly, in the sense that the probabilities of meetingwith potential partners are increasing in the number of hours invested in socializing bysingles. Individuals make optimal time allocation decisions of working, relaxing, andsocializing to maximize the discounted sum of utilities from consumption, leisure, andpotential marrital surplus. Individuals form rational expectations of single populationdistribution in the future marriage market and make optimal search and marital de-cisions accordingly, and these optimal decisions conversely determine the evolution ofpopulations, which consequently establish marriage market outcomes as an equilibriumresult. The model is estimated using the American CPS and time use datasets. Usingthe estimated model, we recover the lifespan returns to each hour spent by singles fromlabor or marriage markets, and discuss how the wage dynamics play a role in determin-ing the optimal marriage search behaviors. By comparing the career paths and marriagepatterns of two cohorts, we shed light on how the changes in wage structures have shiftedoptimal marriage timings and adjusted individuals’ search strategies quantitatively.

This paper explore methodologically the computation and estimation of overlapping generations models for explanining empirical dynamics using connected equilibirum transitional paths. Aside of the application in my job market paper, this empirical strategy can be expanded to explaining many other important economic outcomes whose changes are driven by multiple sequential shocks. This project aims at providing rigorous identification and statistical inference analysis, and making clear constrast to the traditional methodologies such as two-steady-state comparison and single transitional path. To address the intensive computational demand, I discuss the employment of machine learning techniques in addressing the curse of dimensionality.

This paper focus on nonparametric identification of model primitives when agents make both discrete and continuous choices simultaneously, such as labor supply and consumption decisions. The model allows for unobserved marginal utility shifters related with the continuous choice, and simultaneous decision making raises identification issues. I discuss how exploiting panel data retrieves the identification in two cases, where the unobserved individual-specific marginal utility shifters are either persistent or transitory.