Xiangchen Liu
Fertility Intentions of China's One-Child Generation: A Comparative Analysis with the US, Journal of Family Studies, 1-34.
Genevieve Gorst Herfurth Award for Outstanding Research - Honorable Mention, UW-Madison,
2025
In 1979, China introduced the unprecedented one-child policy, penalizing households exceeding birth quotas to slow rapid population growth. The policy's implementation experienced localized variations and evolutions over time. In this study, I examine the relationship between only child status and fertility intentions in China. Using data from the 2017 Chinese General Social Survey, OLS and multinomial logistic regression models are estimated. Results demonstrate that only children have significantly lower fertility desires, even after controlling for demographic and socioeconomic characteristics. Only children have a 20% higher probability of wanting 0-1 children compared to those with siblings. An OLS model interacting only child status with birth cohort and urban/rural origin provides nuanced evidence on heterogeneity in ideals. Those born under peak 1979-1983 enforcement period show reduced fertility intentions, potentially reflecting heightened internalization of small family ideals. Although later birth cohorts show declining magnitudes in the association between only child status and lowered fertility intentions, suggesting some softening of policy impact over time, the association between being an only child and reduced fertility desires persists. I also conduct a comparative analysis using data from the 2016 and 2018 US General Social Survey (GSS). The results show that the patterns observed in China regarding place of birth and birth cohort effects are specific to China's only children, highlighting the unique influence of the one-child policy on fertility intentions. The findings imply China’s sustained low fertility may be partially driven by enduringly low fertility mindsets among generations shaped by the one-child policy.
Under Review
Presentations: APPAM 2022 Fall Research Conference, PAA 2023 Annual Meeting, Consumer Science Graduate Workshop (at UW-Madison)
Utilizing data from the National Longitudinal Survey of Youth 1979 (NLSY79) and employing group-based trajectory modeling, this paper aims to delineate the career pathways of men and women in the U.S., with a particular focus on the role of parenthood. A key contribution of this study is the introduction of a newly defined and validated measure of promotion, which relies on thresholds of wages and working hours, addressing a notable gap in labor surveys. Within each gender, the research uncovers distinct promotion trajectories and reveals that individual characteristics, such as educational level, are correlated with different career outcomes. Specifically, for mothers, more interruptions due to transition to motherhood are associated with worse career outcomes, including lower earnings. Across genders, the study finds that mothers face more career interruptions and delays compared to fathers, thereby exacerbating existing gender disparities in the workplace. These findings have notable policy implications: the observed intragender inequalities tied to educational level call for targeted interventions, while the intergender disparities spotlight the need for support systems that assist mothers - and parents more broadly - in balancing family and work commitments early in their careers.
Childcare costs, public support, and maternal employment in the U.S.
Presentations: APPAM 2023 Fall Research Conference, Consumer Science Graduate Workshop (at UW-Madison), ACCI Annual Conference 2024
The study employs a multi-level dataset, combining the National Longitudinal Survey of Youth 1997 (NLSY97) with restricted geocode files, the National Database of Childcare Prices (NDCP, 2008-2018), and state-level childcare policy data, to examine how variations in childcare contexts influence maternal employment in the U.S. Unlike prior research, which has largely focused on statelevel childcare prices, this study accounts for local variations by linking individuals’ locations with county-level childcare prices and state-level childcare policies. Utilizing an event history approach with individual fixed effects, the study identifies changes in individual women’s employment status around the "event" of transitioning to motherhood. Results demonstrate that local childcare prices and state-level policies have significant and varying impacts on maternal employment, particularly across races/ethnicities and educational levels. The research provides a nuanced, geographically informed analysis, aiming to pinpoint county-level variations that contribute to inequalities in mothers’ labor supply. The findings intend to inform childcare and family policies with the ultimate objectives of encouraging maternal employment, reducing gender disparities in the labor market, and enhancing the well-being of women from various sociodemographic backgrounds.
Young Scholar Award, the Asian Meeting of the Econometrics Society 2023 (NTU)
Presentation: 2023 Asian Meeting of the Econometric Society (NTU)
This paper studies the anchoring effect and loss aversion in the Non-Fungible Token market with complete user-auction level data from Crypto Punks. With the estimated market price and return indexes of the NFT market from hedonic regression, we measure the anchoring effect and loss aversion of bidders and sellers separately. We confirm that both the anchoring effect and loss aversion appear in the NFT market. Both sellers and bidders deviate toward the last period transaction price, as the reference point, and more so when they expect potential loss. We interpret these results as indicating the presence of loss aversion among sellers and the emergence of endogenously generated loss aversion among bidders. The results are robust when all behavioral biases are measured in Ether. We also test whether experience moderates reference dependence and find evidence that experience reinforces agents’ reference dependence in the NFT market. Finally, we do not find evidence in the NFT market showing sellers and bidders act strategically toward their counterparty.
Payday Lenders as Utility Payment Centers: Can Paying Down Bills Lead to Debt? (with Yiwei Zhang)
Presentations: APPAM 2023 Fall Research Conference, MEA 2024 Annual Meeting, ACCI Annual Conference 2024
This paper examines whether vulnerable households are induced to take out payday loans when given the opportunity to pay utility bills at payday lending locations. In 2016, Missouri banned utilities companies from using payday lenders as bill payment centers. Using a quasi-experimental synthetic control approach, we find payday loan usage decreased 14.2% in Missouri relative to other states following the regulation. This suggests paying bills at payday lenders causally increases payday loan uptake among low-income households. Ongoing work examines spillovers to other financial services. Our results have implications for policies aimed at improving financial wellbeing through decoupling bill payment and payday loan access.
Health Insurance and High-Cost Borrowing: State Heterogeneity and the Effects of Medicaid on Pawn Loans, Payday Loans, and other Fringe Bank Products. (with Anne Fitzpatrick and Katie Fitzpatrick)
Richard L.D. Morse Consumer Economics Award, American Council on Consumer Interests Annual
Conference, 2025
Presentation: ACCI Annual Conference 2025
Linking 2009 through 2019 March and FDIC-sponsored CPS Supplements, we estimate average treatment effects, as well as by state and time, of the Medicaid expansion’s effect on insurance coverage and financial security with a synthetic control approach. We measure financial security by use of “fringe banks”, non-banks offering financial products (e.g., payday lenders, check cashers, etc.). Medicaid coverage increased 6 percentage points (21%) reducing demand for fringe bank credit by 2 percentage points (21%). With treatment effect variation, machine learning identified the policy environment of state Medicaid expansions affected enrollment and financial outcomes differentially across states.