CESR Events

Series resumes Monday, January 26, 2026

Monday, January 26
12pm – 1pm
VPD 203 and

Abstract:

Background: Family care resources have profound implications for care use among those with functional limitations, especially when coupled with dementia. A better understanding of family availability among those at high risk of dementia is important to predict care transitions at the onset of dementia and subsequent years.

Methods: Using Health and Retirement Study (HRS) data, we developed a holistic measure of family care availability (FCA) for adults 55+ who will have dementia onset within the next two years and have disability. We incorporated multiple family contextual factors, considering presence, disability status, geographic proximity, and working status. We conducted cluster analysis to identify groups based on FCA measures specific to relationship type (spouse, adult child, sibling, grandchild). We then examined variation in FCA across population groups and assessed nursing facility utilization at the onset and over the course of dementia.

Results: Of the 2,119sample persons, 54% did not have a spouse, and 6% did not have an adult child at baseline (i.e., 2 years prior to the onset of dementia). The overall FCA index ranges from 0 to 100, with a mean of 32.0 (SD = 20.6). The distribution of the FCA index was primarily shaped by the spousal and child availability. FCA is significantly lower at ages 85+ compared to 65-84 (26.0 vs. 33.4-36.7; p<0.001). Males’ FCA is substantially greater compared to females (38.9 vs. 27.2; p<0.001). Hispanics have significantly higher FCA than Non-Hispanic Whites and Blacks (37.02 vs. 30.0-31.6; p<0.001). Adults with some college education and above have a greater FCA index compared to adults with a high school education (35.7 vs. 29.4; p<0.001). Cluster analysis identified three distinctive groups: high spousal availability (N=552); high adult-child-availability (N=497); and low spouse- and low adult-child-availability  (N=1,292). Those in the high adult-child-availability group are significantly less likely to use nursing facilities than those in the low spouse- and low child-availability group at dementia onset and subsequent years.

Conclusion: The FCA index, a holistic measure of family availability to care, will help identify vulnerable populations lacking family care resources and predict care transitions at the onset and over the course of dementia.

Bio:

Dr. HwaJung Choi is an Associate Professor at Internal Medicine, School of Medicine, and Health Management and Policy, School of Public Health, University of Michigan. She is also a Faculty Associate at the Institute for Social Research (ISR), University of Michigan. She was a Fulbright scholar and received Ph.D. in Economics at the University of Michigan.

Dr. Choi is an economist and family demographer whose research focuses on family and community contextual factors and their implications for health and care use among older adults with physical and cognitive limitations. Dr. Choi also examines trends and cross-national differences in health and disability outcomes. She is awarded large research grants from the National Institute on Aging as PI, which support the following studies: to assess the impacts of Covid19 pandemic on care use among adults with dementia (R01AG075002); to examine the implications of residential location for disability and cognitive function outcomes (R01AG080491); to develop and evaluate new measures of family care availability (RF1AG083037). Dr. Choi is a Co-lead of the Research and Education Core at the Michigan Center for Contextual Factors in Alzheimer’s Disease and supports early-career scientists in advancing their research and training on contextual factors related to dementia risk and care.

Monday, February 2
12pm – 1pm
VPD 203 and Zoom

Abstract

We consider a generalized method of moments framework in which a part of the data vector is missing for some units in a completely unrestricted, potentially endogenous way. In this setup, the parameters of interest are usually only partially identified. We characterize the identified set for such parameters using the support function of the convex set of moment predictions consistent with the data. This identified set is sharp, valid for both continuous and discrete data, and straightforward to estimate. We also propose a statistic for testing hypotheses and constructing confidence regions for the true parameter, show that standard nonparametric bootstrap may not be valid, and suggest a fix using the bootstrap for directionally differentiable functionals of Fang and Santos (2019). A set of Monte Carlo simulations demonstrates that both our estimator and the confidence region perform well when samples are moderately large and the data have bounded supports.

Monday, February 9
12pm – 1pm
VPD 203 and Zoom

Monday, February 23
12pm – 1pm
Zoom

Abstract

Psychosocial development interventions often carry implicit cultural assumptions that may conflict with local culture and norms, potentially reducing their effectiveness. Yet, the costs of such cultural incongruence likely vary by context and fade as individuals become familiar with different cultural frames. We study the causal impact of two versions of a psychosocial intervention in a randomized trial with female microfinance members in Vietnam, a setting where different cultural models of agency coexist. Participants received a training emphasizing either independent or interdependent agency, or no training. Both arms improve economic outcomes, and limited heterogeneity by baseline cultural orientation suggests limited costs of cultural incongruence in this setting.

Monday, March 2
12pm – 1pm
VPD 203 and Zoom

Abstract

We study the spillover effects of cognitive decline in one member of a coupled household on the financial outcomes of their partner and assess how “own” and spillover effects are moderated by the structure of household financial decision making. We use a large, nationally representative longitudinal dataset spanning 2000-2017 that includes credit report data merged at the individual level with Medicare claims and enrollment data. We find the own adverse financial consequences of cognitive decline depend on household financial integration and other characteristics associated with household financial management, and find significant, albeit smaller (vs own), adverse financial spillover effects on partners

Bio

Wilbert van der Klaauw is an economic research advisor in Microeconomics. He is also the director of the Center for Microeconomic Data and chair of the Research Group’s Workforce and Recruiting Committee. He is a labor economist and applied econometrician whose research interests include the study of life cycle labor supply and occupational choice decisions, household financial behavior and expectations, the economic determinants of household formation and dissolution, educational investment and productivity, and econometric approaches to program evaluation. Prior to joining the New York Fed, Dr. van der Klaauw was a Professor at UNC-Chapel Hill and Assistant Professor at New York University. He holds a Ph.D. from Brown University.

Monday, March 9
12pm – 1pm
VPD 203 and Zoom

Monday, March 23
12pm – 1pm
VPD 203 and Zoom

Monday, March 30
12pm – 1pm
VPD 203 and Zoom

Monday, April 6
12pm – 1pm
VPD 203 and Zoom

Monday, April 13
12pm – 1pm
VPD 203 and Zoom

Monday, April 20
12pm – 1pm
VPD 203 and Zoom

Monday, April 27
12pm – 1pm
VPD 203 and Zoom

Monday, May 4
12pm – 1pm
VPD 203 and Zoom

Thursday, January 15
9am – 10am
Zoom (See email for Zoom link)

Abstract: Much is known about how parental resources shape the human capital of children, but far less is known about the role of parental traits and behaviors. Using unique data from the Lifelines Cohort Study linked to administrative records, we study the intergenerational effects of a random increase in parental genetic predisposition for externalizing behavior. There are three main findings. First, a stronger predisposition for externalizing behavior is a strong predictor of hostile behavior, attention deficit disorder, and substance abuse, but is less strongly related to education or income. Second, this predisposition leads to sizeable reductions in the education of the next generation. Third, these intergenerational effects cannot be explained by the genetic endowments of children. Our results thus present a unique case in which intergenerational effects on education exceed the effects on these outcomes in the first generation.

This project is joint work with Sander de Vries

Bio: Sjoerd van Alten is an economist researching the genetic and environmental determinants that shape inequalities in human capital, health, and labor market outcomes. He investigates these topics as a Postdoc Researcher at the School of Business Economics (SBE) at the Vrije Universiteit Amsterdam. Additionally, some of his work focuses on the effects of selection bias in large-scale Biobanks that are widely used in medicine, epidemiology, statistical genetics, and the social sciences.

He has authored and published various papers on these topics in outlets such as Nature Communications, Nature Genetics, and Intergenerational Journal of Epidemiology. In 2024, he completed his PhD-thesis “Genetics, Human Capital Formation and the Intergenerational Transmission of Socioeconomic Status” at Vrije Universiteit Amsterdam and Tinbergen Institute, under supervision of Titus Galama, Maarten Lindeboom, and Kevin Thom (UW-Milwaukee).

Thursday, February 5
9am – 10am
Zoom (See email for Zoom link)

Thursday, March 5
9am – 10am
Zoom (See email for Zoom link)

Thursday, April 2
9am – 10am
Zoom (See email for Zoom link)

Thursday, May 7
9am – 10am
Zoom (See email for Zoom link)

Thursday, June 4
9am – 10am
Zoom (See email for Zoom link)

In its eighth installment, the Current Innovations in Probability-Based Household Internet Panel Research (CIPHER) Conference expands its scope to include artificial intelligence (AI) as a new area of focus. As always, CIPHER builds on a rich legacy of methodological innovation, international collaboration, and emerging data modalities. Bringing together researchers, technologists, and policymakers, this year’s conference will explore how AI can enhance panel design, data quality, respondent engagement, and ethical governance. Join us as we chart the future of probability-based internet panels at the nexus of artificial intelligence and survey science.