Around the world, one in three women experiences domestic violence. How can it be reduced? New research is bringing unexpected insights into this problem — and its potential solutions.

A study accepted for publication this month by the Review of Economics and Statistics found that, in Bangladesh, improving the economic status of women can decrease domestic violence if the women also took part in an educational program that helped elevate their social standing in the community.

Women who received both economic assistance and education for two years reported a 26 percent decrease in domestic violence — even several months after the program stopped.

But here's the surprising part: Violence against women was never mentioned in the training.

In fact, the project, run by the (IFPRI), started out as a study on nutrition. Did women who were given food (or cash to buy food) improve their household's health when they were also educated on healthy diets?

The researchers observed three groups of 5,000 very poor women in Bangladesh — a control group given no resources or training; a second group of women who were given food or cash to buy food; and a third group of women who were given cash for food along with intensive training sessions on nutrition.

Shalini Roy, a research fellow in IFPRI's poverty, health and nutrition division, and her colleagues began to notice that the third group — the women who received both cash for food and education — were "noticeably more confident, outspoken," Roy told NPR.

To investigate further, the researchers began a new study. Six to ten months after the nutrition program ended, they interviewed all of the women about whether violence in their homes had changed.

The first two groups saw no lasting changes. But the third group saw a one-quarter decrease in spousal abuse.

There are a couple of possible reasons for this, the researchers say.

Having more cash or food in the household can reduce stress and conflict that leads to abuse. So even temporary cash and food can have some effect.

But clearly it was the training program that made the difference. The researchers suggest that perhaps the women in that part of the study developed stronger social ties, an improved status in the community and more confidence – and thus have a greater ability to negotiate improvements in a violent home situation or are more willing to walk out.

It's possible focusing more specifically on abuse in the training sessions would help address the root problems of domestic violence — but it's also possible the husbands might have objected if their wives were to attend such sessions, Roy pointed out.

The program challenged gender norms by bringing the women out of their homes and giving them assistance and training, Roy said. "But the training itself was focused on a domain that's accepted to be within a woman's purview" — food.

The women made friends through the nutrition program, and they also became local experts on eating better — so much so that their neighbors, who had previously looked down on them for being very poor, began asking for their advice on nutrition.

"Suddenly they had a lot more social interaction," Roy said. In fact, it's possible, that the social interaction and improved social status were the keys to reducing domestic violence, she added.

Another new study, published in PLOS One, offers a similar finding: Economic progress in and of itself may not have an effect on acceptance of domestic violence.

As women have made social and economic gains, "gender expectations and stereotypes haven't changed in tandem," LynneMarie Sardinha, lead author of the study, told NPR.

The researchers looked at responses to USAID's demographic and health surveys conducted between 2005 and 2017 in 49 low- and middle-income countries. A total of 1.17 million men and women were asked about domestic violence.

They found that one-third (36 percent) of men and women in these countries said violence against women was justified in some cases — if she burns the food, neglects the children, argues with her husband, goes out without telling him or refuses to have sex with him. Those numbers, though high, are in keeping with previous research.

Even in countries where women had made strides in gender equality – a high percentage of girls in school and of women in the workforce and a significant number of seats held by women in the national parliament – there were still relatively high rates of acceptance by women of domestic violence.

The researchers hypothesize that rapid changes in some areas — in this case, in politics — don't necessarily mean the rest of the country has changed as quickly. In Rwanda, for example, only 17 percent of the seats in parliament were held by women in 1995, while the 2016 figure is 56 percent. Yet in the USAID survey, 56 percent of women and 25 percent of men see violence as an acceptable response.

Sardinha, a fellow at the University of Bristol's Economic and Social Research Council, has a hypothesis: simply improving the lives of women is not enough to decrease violence against women.

"This narrow definition of what empowerment is, and how gender equality is measured needs to take into account that we need prevention and intervention that specifically targets the attitudes of both men and women," she says.

Economic empowerment programs, coupled with social change programs, can help, says Julia Weber, a professor of domestic violence law at Golden Gate School of Law. "Yet we still have not developed enough programs that reflect this understanding."

That's the challenge facing researchers, says Weber – and a topic that the authors of these two reports plan to continue investigating.

Melody Schreiber (@m_scribe on Twitter) is a freelance journalist in Washington, D.C.

Copyright 2021 NPR. To see more, visit https://www.npr.org.