I study how our closest relationships shape our health — and build data to support the science.
My work is organized around three questions: how relationships get under the skin, who does the work of family life, and how unions are changing over time. To answer these questions, I had to create something new — contemporary population-representative data on couples and families that other researchers can now use too.
Relationships & health
According to our research, for most couples, the happiest you’ll ever be in your marriage is right at the start. My early work with Miles Taylor established what’s now called the honeymoon-as-ceiling effect: most spouses report their highest marital quality at baseline, and even the most satisfied experience little improvement from there. That finding has been replicated and extended many times since, and it reframed how relationship scientists think about change in marital quality over time.
This is important because relationships and health are deeply connected. Moving in together or marrying eases emotional distress — even the second time around — while depressive symptoms and alcohol problems shape who forms unions and who marries at all. Today, with the NCHAT-BIO study, Lisa Christian and I are tracing these dynamics all the way down to inflammation and epigenetic aging, asking how even who you love and are partnered with can leave marks in the body itself – because of interpersonal, community, and structural stress that results from being minoritized.
Gender & the division of labor
My colleague Sarah Schoppe-Sullivan and I asked new parents who were primarily highly educated, dual-earner, white couples having their first child, to estimate their workloads including paid work, housework, and childcare. With our former graduate student Jill Yavorsky, we found that men and women reported the same thing: everybody’s had more hours of work after the baby was born. But time diaries — which require participants to account for each minute of the day — told a different story. Couples were largely egalitarian before the baby arrived, splitting housework equally. But by nine months postpartum, women had added fifteen hours of weekly work; men had added five.
Then we built the first minute-by-minute couple time diaries in the US to answer the important question: what exactly were fathers doing while mothers did that extra childcare and housework? On their days off, the most common answer was leisure. When men did report housework and childcare, women were usually doing the same things with them. The stalled gender revolution, it turns out, might stall in the first months of parenthood.
Life course trends in unions
How we partner is changing and marriage itself may be changing. Comparing national cohorts of young adults born twenty years apart using the NLSY97 and NLSY79, we found that the age at first union had changed little, but what had changed, was what kind of union it was. The 1997 cohort entered unions earlier — but married later and cohabitated earlier — with socioeconomic disadvantage consistently undermining union stability no matter how it shaped union timing. And partnering patterns echoed forward: mothers with more partners had children who, decades later, had more partners too, even after accounting for economic instability.
A major question in this line of work has been: is the institution of marriage itself changing? With the Work and Family Life Study, we fielded the same survey married Americans answered in 1980 and 2000 to a new cohort in 2022-2023. Everything about who married has shifted in ways that should have made marriage better. What actually happened is more complicated, and it’s the subject of a major new paper that is currently under review.
Building the data
The data I needed didn’t exist — no population-representative study of contemporary US couples that included all couples, no rich harmonized historical and contemporary data on American marriages, no central home for the state and county contextual measures health researchers kept rebuilding from scratch. So I created it. These studies and tools are open to researchers everywhere, and each has its own unique strengths.