With or without him? Experimental evidence on cash grants and gender-sensitive trainings in Tunisia

Read the paper


Should male partners be included in programs designed to empower women and encourage their involvement in more income-generating activities? There is mixed evidence on how household dynamics influence programs aimed at increasing female agency. Traditional thinking assumes that a household has only one set of preferences and that information flows freely, so including the male partner should not impact outcomes.

More recent research, however, has found that male partners often have different preferences and intentionally hide income, savings, or behavior from their female partners in order to control more resources. If this is true, excluding husbands might increase a woman’s privacy and agency, especially in contexts where women face a high risk of expropriation and strong gender norms that limit the roles they can play. Alternatively, male involvement may minimize resentment or backlash in response to women’s empowerment and, ultimately, maximize the positive impact of these programs.

Bottomline: Overall, cash transfers plus training improved household standards of living but improved women’s labor participation and business ownership only when male partners were excluded.

Methods and Results

The research centered around a randomized control trial (RCT) targeted at women in Jendouba, Tunisia. 1,000 women received an unrestricted cash grant of four times their median monthly income. Before receiving the grants, women had to participate in a one-day, gender-sensitive financial training designed to encourage women to invest their money productively in physical capital (e.g., starting a business) or human capital (e.g., vocational training of their choice). The training covered three main modules: i) financial planning and budgeting, ii) savings, and (iii) debt management. Each module included a series of videos, exercises, and guided discussions, all specifically designed to increase women’s agency. Half the participants were encouraged to bring their male partners to the training, and half were told to attend alone.

After two years, in terms of labor market outcomes, the program had positive impacts on women’s likelihood of having an IGA (+3.3 p.p.) and on their income (+60%), but only for women who had to attend alone. The involvement of partners in the training appears to have backfired: compared to women who had to attend the training alone, women who could invite their male partner were significantly less likely to own a business, earned lower incomes, and worked fewer hours.

More generally, both groups saw improvements in overall household living standards (measured by an increase in food consumption and ownership of assets such as motorcycles, televisions, and generators), and women in both groups reported higher satisfaction with their lives, better mental health, and better access to finance. Interestingly, there was no effect on women’s agency in either group, meaning that improvements in employment status and financial gains are not enough to increase feelings of empowerment. 

Implications for Policy Makers

Overall, this research provides additional evidence that large cash grants can have positive effects and be highly cost-effective (with benefits exceeding program costs after 1.2 years). Unfortunately, it also highlights the difficulty of stimulating women’s agency in traditional societies and the need to carefully account for intra-household dynamics when aiming to improve women’s involvement in income-generating activities.



Jules Gazeaud (The American University; Université Clermont Auvergne), Nausheen Khan (World Bank), Eric Mvukiyehe (Duke University), Olivier Sterck (University of Oxford; University of Antwerp)

Related Content

Related Links


Human Development, Education, Health, Economic Development, Impact Evaluation, MENA