Intimate Partner Violence (IPV) against women is a major public health and inequality concern. It's not clear what causes it or how to reduce it. Gaby Deschamps' JMP considers that IPV is positively correlated with motherhood, and asks if we can understand why. www.gabrieladeschamps.com/research - ThreadSky

economeager.bsky.social • 93 days ago

Intimate Partner Violence (IPV) against women is a major public health and inequality concern. It's not clear what causes it or how to reduce it. Gaby Deschamps' JMP considers that IPV is positively correlated with motherhood, and asks if we can understand why. https://www.gabrieladeschamps.com/research

Comments

jimc5.bsky.social•92 days ago

This is great work

I think this also highlights the dangers of making it harder for women to divorce or otherwise leave relationships

I wonder why some people disagreed with the hypothesis, and want them to explain their ideas why on making it harder for women to get contraception or leave is good

jimc5.bsky.social•92 days ago

This also confirms my priors, as a dad. In my experience, Having kids does not mark people violent, but it does tie people to relationships.

ahumbleman.bsky.social•92 days ago

Thank you for sharing this! I wouldn’t have stumbled across this otherwise. What a poignant and thought provoking paper! Looking forward to actually reading it when I have time.

mousebsky.bsky.social•93 days ago

Another reason why people don't want to have kids anymore.

yarine.bsky.social•92 days ago

@planasrodriguez.bsky.social I meant this one, sorry (although Michelle’s is very relevant to all of us as well)

economeager.bsky.social•93 days ago

The main problem for this question is that there are competing causal stories that look observationally equivalent in the cross section. In this case, perhaps having a baby (stressful) triggers men’s violent behavior; or motherhood makes women more likely to stay with a violent partner; or both.

economeager.bsky.social•93 days ago

To disentangle the stories, one needs at minimum data on both partnered and separated women that provides (proxies for) 3 key outcomes: whether women currently experience violence, whether their (ex-)partner has exerted violence, and whether they are separated.

economeager.bsky.social•93 days ago

Basically we need, as much as possible, the complete history of violence, or at least SOME information on both current and past IPV, so we can identify if women who are now separated but who previously experienced violence when they were partnered. This kind of data is rare and difficult to collect.

economeager.bsky.social•93 days ago

Gaby has access to such data: a detailed time series on women's relationship experiences, partner status histories and fertility from multiple waves of the Mexican survey Encuesta Nacional de Dinámica de las Relaciones en los Hogares (ENDIREH).

economeager.bsky.social•93 days ago

With this data, and a new theoretical model, Gaby is able to temporally disentangle the stories. As an additional exercise, she also exploits a fertility shock induced by the opening of pharmacies throughout the survey period using a DID design.

economeager.bsky.social•93 days ago

First, there is much Gaby can do just with the survey data. For example, under the "direct violence" channel, mothers will be more likely to have experienced IPV with their first partner, as having a baby causes violence. But in fact, the coefficient is a precise zero.