Causal Inference: Effect of Time, Dual Roles of the Unobservables

In the first chapter of my dissertation, I developed a new method to identify and estimate the effect of time exposure to a social context.

Take for example the effect of grandparent’s life-course overlap on children’s cognitive development. The effect of grandparent overlap is really the effect of extending a grandchild’s exposure to all of her grandparent’s observed and unobserved, fixed and time-varying characteristics. I propose the cumulative fixed effects to address the causal question where unobserved factors play the double duty as confounders and as moderators of the treatment and to recover the entirety of the effect of grandparent’s time.

Multigenerational Inequality under Population Aging

How does aging shape multigenerational inequality?

In the second chapter of my dissertation, I investigated how grandparent overlap affects the children’s standardized test scores using the Danish register data. I found positive grandparent overlap effects on children’s cognitive ability with a much larger effect size than those estimated using conventional methods. Grandparent overlap effects vary remarkably across lineages, SES, health, residential distance, and availability. The secular rise of life expectancy may shape future inequality through both the unequal increase in lifespan and the unequal benefit associated with each year’s increase for children of different backgrounds.

Family complexity, relationship and Aging

Divorce, remarriage, stepkin. In the US, the generations first experienced family complexity have aged into their seniorhood.

In the third chapter of my dissertation, I focus on modes of intergenerational transfers among adult children and their senior parents in stepfamilies. Departing from past studies which show a weak norm of support among stepkin, I find that stepkin respond more sensitively to each other’s past signals of helpfulness with more financial and emotional support despite their lower transfers when no signals of past help are present. Stepkin relationships are more “on the fence of a family” rather than “not in a family”.

Multigenerational Policy, Coresidence and Child Development

Population aging underscores important policy changes.  Grandparent coresidence became increasingly prevalent in liberal market countries in recent years, driven by the deteriorating house affordability, delayed transition to adulthood of the younger generations, increased work-life stress, and the rising cost of childcare.

In a paper published at Child Indicators Research, I examined the effect of three-generational household arrangements on children’s cognitive and behavioral outcomes as children grow up. I find that grandparent coresidence affects the children’s development negatively in early childhood, while the effects approach zero at children’s school age. The study stresses the importance of thinking about policies of work, childcare, and housing from a multigenerational perspective.

Aging, Migration and Politics

Immigration is an important issue for aging societies, but what are some social consequences of migration and what are the policy implications?

In spite of the myth of high criminality of undocumented immigrants in the US, there was no empirical evidence of scientific scrutiny. We used the most complete administrative data from Texas and a rigorous quantitative approach to find considerably lower felony arrest rates among undocumented immigrants compared to legal immigrants and natives (second-author, published at PNAS, reported by Scientific American).