🚨 Call for abstracts for #SHESG2025 is OPEN! Abstracts: June 15. Conference: September 19-20 in Chapel Hill. Your esteemed local hosts @UNC_econ are Donna Gilleskie, Qing Gong, @JustinTrogdon, and Tara Templin. JOIN US, FRIENDS! https://t.co/seyYBqHTvq

BUT there is a lag in the effect (almost two years) and later treated localities are very different from earlier ones (less urban and greater number of non-IRL dating options).

For the #SHESG paper, they study the impact of CP on births. They use a CS DiD with one control (the RUCC urbanicity code) because of the way Craigslist rolled out. They find that the birth rate for 15-44 year old women rises after CP arrives.

Craigslist Personals (CP) developed granular lists for people looking for very specific kinds of matches - and importantly it rolled out gradually over time to different locations. They collect data on Craigslist local activity from 2000 to 2010, to support a DiD from 1995-2007.

One difficulty with studying “swipe based” app effects on dating markets is that they all hit everyone nationally at the same time. In addition, there’s a decades-long secular decline in births that confound usual suspect DiD models. This is where Craigslist Personals helps.

Before the 1990s few people met their partners online; by the mid-2010s the majority of couples first meet online (more than 40% today). Christine and her colleagues are going to study the effect of online dating by going back to the beginning: Craigslist personals in 2000.

We wrap up #SHESG2025 on a high note with Christine Durrance presenting work with her colleagues on online dating, dating markets, and family formation. The internet caused a major shock in the way people meet - especially with the introduction of dating-specific apps.

They use the CS diff-in-diff and exclude not-yet-treated PFL states from the treatment group. They find very significant (economically and statistically) reductions in child maltreatment after PFL benefits start. Benefits start immediately and are persistent over time.

He uses NCANDS data to understand how PFL in three states causally affects cases of child maltreatment reported to authorities over the time frame of 2002-2022. This obviously spans the COVID years. He aggregates data to the county level.

As is commonly found in the literature, women more commonly take advantage of PFL opportunities, though male take-up is rising. There’s already substantial evidence that mothers and infants benefit from PFL; Wei is making a contribution by studying child maltreatment.

But, 13 states and DC have enacted PFL to compensate for federal failures. The question is: what are the broad spectrum benefits of this? State programs vary in terms of their benefit coverage, of course, and Wei chooses his diff-in-diff method carefully to account for this.

For our third presentation of the morning at #SHESG2025 we have Wei Fu presenting work on the benefit of paid family leave in reducing child maltreatment. Obviously, the US stands out among developed countries as not having national paid family leave.

Did the down side (labor frustration and disruption) materialize? There’s suggestive - though not statistically significant - indications that there may have been small reductions in part time CNA labor. These small effects mean that the quality of patient care is unaffected.

The mandates do seem to translate into lower active COVID rates of illness among the staff. Helpfully, this improvement in staff outcomes benefits residents who see associated reductions in COVID infections and deaths.

They’re able to leverage staggered mandate adoption and enforcement by nursing homes across the country. Using standard DinD models they find that employer mandates do have a large effect: adoption increases vaccination by 21%; enforcement has similar effects, just faster.

Katherine et al. study these questions in nursing home labor markets. Nursing homes are a great case study because of the salience of vaccination among staff - and because (weirdly) vaccination rates among certified nurse assistants were surprisingly low.

This set up a trade off for employers: enforce a mandate and lose workers who object but the workers who remain will be healthier and more productive. So, in the end, did the mandates work to increase vaccination rates and improve worker productivity?

Next up at @SHESG2025, Katherine Wen presents work on employment and vaccine mandates. Katherine and her colleagues study the impact of vaccinations mandates surrounding COVID. You may remember that there was significant push-back from workers about the mandates at the time.

When the simulate counterfactual life courses they also find that the impact of transitory mental health shocks later in life on labor supply are much larger than for transitory physical health shocks. Interestingly, labor market negative shocks are smaller but more persistent.

Their model recovers the persistence of health states. But they can now determine that full time work has a positive direct effect on mental and physical health. The combination of good mental health and full time employment raises utility more than other combinations.

They move on to examine the role of large life shocks (e.g., the death of a friend) on health and labor. Negative life shock drive down both mental and physical health. They find people rebound, but often over periods of years. They build a structural model around these facts.