Reducing Economic Violence against Women and Improving Family Financial Wellbeing: A Cluster Randomized Controlled Trial in India

The couples-based intervention, titled “Let us Grow Together: Economic Wellbeing for Families”, consists of six community group sessions targeted at a group of fifteen spouses. Rather than addressing individuals separately, the intervention is designed for both husbands and wives, treating them as a family unit to foster collaboration, encourage joint decision-making, and promote more gender-equitable household dynamics. Complementing the focus on spouses, the curriculum also includes gender-segregated components to provide participants with a safe space for personal reflection and critical discussions on sensitive topics such as violence, gender norms, and financial stress.

The six sessions of the curriculum integrate gender-transformative approaches with financial literacy and economic empowerment training. Drawing on the design of previous financial literacy programmes—including those tailored for IPV survivors—the intervention aims to enhance couples’ financial knowledge and self-efficacy, foster healthy household financial discussions, and strengthen family financial resilience. Core modules cover:

(i) drafting a monthly household budget,
(ii) tracking of monthly expenses,
(iii) understanding the importance of emergency savings,
(iv) learning about interest rates in the context of both debt and savings,
(v) evaluating different saving and investment strategies, and
(vi) devising a family savings goal and saving plan.

Beyond financial literacy, the curriculum incorporates gender-transformative content to challenge traditional gender norms that position men as financial providers and women as caregivers and homemakers. It promotes non-violent communication for de-escalating conflicts, encourages women’s engagement in financial matters, and supports positive attitudes towards female employment. Here, core modules cover:

(i) examining the distribution of paid and unpaid labour between spouses,
(ii) deconstructing societal expectations of the “ideal man”, and “ideal woman”,
(iii) introducing conflict resolution skills and non-violent communication skills,
(iv) raising awareness about economic control and abuse.

The intervention incorporates additional design features proven to enhance the effectiveness of financial literacy programmes in low-resource settings. First, to accommodate potentially low literacy and numeracy levels, financial training modules use simple, easy-to-remember rules of thumb and heuristics. Second the intervention prioritises an “active learning” approach over traditional lecturing, fostering engagement through illustrated case studies, role plays, group exercises, and visual and tactile elements to reduce cognitive load. Third, the intervention promotes active practice and reinforcement of new skills and behaviours through integrated homework exercises, which participants receive at the end of each session and later reflect on—sharing experiences and challenges—at the start of the next session. Additionally, in line with gender-transformative programme design, the intervention actively involves husbands as changemakers rather than positioning them as obstacles to women’s financial autonomy. It emphasises the benefits of teamwork, challenges the male breadwinner norm, and promotes an equal partnership model where both spouses contribute their unique strengths.

Lastly, the intervention will be designed to be low-cost and thus possibly scalable, ensuring its relevance for broader programming and policymaking across India and potentially in other LMIC settings. This is achieved through a "train-the-trainer" approach, which leverages existing community resources and leaders. By fostering local ownership, this approach enhances the programme’s ability to adapt to local circumstances, ensuring high acceptability and cultural appropriateness within the target community. Additionally, the intervention incorporates a digital component in the form of SMS reminders, reinforcing key lessons from the programme curriculum. This aligns with a growing body of literature highlighting the effectiveness of technological interventions—an area that has gained particular momentum during the COVID-19 pandemic—in the prevention of and response to intimate partner violence. Specifically, spouses will receive two SMS reminders following each community session. The SMS reminders will be sent out five and ten days after each session These messages will reiterate key takeaways, provide concise rules of thumb, and remind participants of small joint activities (e.g., creating a monthly savings plan) to complete at home, promoting sustained engagement and practical application of the programme’s principles.

2026-01-01

2026-10-01

Economic intimate partner violence, including experiences (wife report) and perpetration (husband report)

The primary outcome variable will consist of self-reported victimisation (IPV) (for women) or perpetration (for men) of economic IPV. Here, we will use a culturally adapted and extended version of the Scale of Economic Abuse. The item battery will capture four key dimensions of economic IPV, namely (i) economic control (6 items), (ii) economic exploitation (10 items), (iii) employment sabotage (5 items), and (iv) refusal to contribute economically (2 items). The reference period for the measures will be the past six months to align the measure with the time to follow-up from the intervention’s implementation.

Further, to mitigate the risk of non-disclosure due to perceived stigmatisation and social desirability, economic IPV will be elicited using a list experiment designed to increase anonymity and disclosure. Here, respondents will be randomly assigned to one of two lists of statements. One list will consist of four innocuous statements (reference group) and the other list will include these same four statements and an additional sensitive item on economic IPV, namely “I am not allowed to decide how money is spent in our home.”/”My wife is not allowed to decide how money is spent in our home” (experimental group). Respondents will then specify how many of the presented statements apply to them (e.g., “3 out of 5”) without having to indicate which statements they are referring to. The prevalence of economic IPV will then be determined by subtracting the average number documented in the experimental group from the average number in the reference group.

In addition, we will make use of the implicit association test (IAT) technique to measure potential subconscious attitudes towards economic IPV, using an adapted version of the Gender Violence IAT. Following previous research, the IAT will consist of a series of stimuli in the form of words, which respondents are asked to sort into two categories by touching a left or right button on the mobile tablet. Specifically, the IAT will be composed of different blocks in which respondents need to sort male and female names into the categories (a) woman and (b) man (target concept) and words such as “employment”, “career”, “care”, “cleaning” into the categories (c) work and (d) home (attribute concept). In a subsequent block, words corresponding to our attribute concept (work vs. home) will be sorted to the target concept “women” vs. “man” in a stereotypical manner, i.e. words belonging to the attribute work will be assigned to “man” and words belonging to the attribute home will be assigned to “woman”. In a last block, words corresponding to the attribute concept will be placed into the categories that are not stereotypically associated (i.e., “career” – “women”). The key assumption of the IAT approach is thereby that respondents’ reaction time is faster if a perceived association between concepts is stronger. The final IAT score will be defined as the difference between the mean response times for incongruent category pairs (e.g., women and work) and congruent category pairs (e.g., women and home). Considering that the IAT takes approximately 15-20 minutes to complete, we will only include the IAT at endline.

1. Financial literacy and knowledge (husband and wife report)
2. Household savings balances (husband and wife report)
3. Past-month income (husband and wife report)
4. Financial distress (husband and wife report)
5. Women's employment status (wife report only)
6. Women's empowerment and agency (wife report only)
7. Physical, sexual, emotional intimate partner violence and partner's controlling/coercive behaviours (wife report only)

A first group of secondary outcomes will seek to capture the possible impacts of the intervention on aspects of spouses’ economic wellbeing. Specifically, we will include:
(i) Financial literacy and knowledge. We will measure financial literacy and knowledge based on a five-item scale in which participants rate their financial knowledge and competencies on a 5-point Likert scale (e.g. “I know how I can deposit and withdraw cash from a bank account”), and two additional factual questions drawn from a previous study conducted in India to assess respondents’ financial literacy.

(ii) Household savings balance. We will capture participants’ self-reported savings based on a survey module used in a previous RCT in Pune, Maharashtra, which will capture the total amount of savings kept at home, in a bank, post office, national savings centre or mobile phone account, in a savings group, money kept with relatives or friends, and the estimated amount of in-kind savings in jewellery/gold/silver.

(iii) Past-month income. We will capture participants’ self-reported earnings in the past month, including payments made in-kind to account for sectors in which wages are more typically paid in-kind.

(iv) Financial distress. We will measure participants’ self-reported financial distress based on an 8-item scale adapted from a previous RCT in Tanzania, where participants indicate how often in the past six months they, for example, “had trouble buying food or other necessities for their family” or were “very worried/stressed about their general financial situation”.

Second, we will assess women’s employment status (including formal and informal work) and female empowerment and agency based solely on wives’ reports. The former measure will be a binary indicator, coded as 1 if the woman reports engaging in income generating activities, including formal and informal work. The latter measure will capture multiple dimensions of women’s empowerment, including: (i) freedom of movement, evaluated using a five-item scale developed by Richardson et al. (2019) to assess women’s agency in India (e.g., the ability to visit the market or one’s natal family unaccompanied), (ii) self-efficacy, measured using a four-item adaptation of Rotter’s locus of control scale, (iii) community participation, assessed through items from the Indian Human Development Survey, which capture women’s membership in various groups (e.g., self-help groups, savings groups) and their attendance at village panchayat committee meetings, and (iv) self-rated empowerment, measured using a 10-step ladder scale previously employed with rural women in Nepal, where respondents position themselves relative to the perceived status of women in their community.

Third, we will evaluate treatment effects on other forms of IPV based solely on wives’ reports by combining items from the World Health Organisation’s multi-country study on domestic violence against women and the Indian Family Violence and Control Scale. The measure will capture four sub-forms of IPV, namely (i) physical IPV (4 items), (ii) sexual IPV (2 items), (iii) emotional IPV (3 items), and (iv) coercive behaviours (3 items). Like economic IPV, the reference period for the measures will be the past six months.

The impact of the couples-based intervention will be evaluated by means of a cluster randomised controlled trial (RCT). After completion of the baseline data collection, clusters will be randomly allocated to treatment or control arm by a 1:1 ratio, stratified by state, rural/urban location, and household wealth quartiles using data on participants’ asset ownership and living standards collected during baseline that we will aggregate to the cluster level.

Not available

Within each strata, clusters will be randomly assigned to the treatment and control arm based on simple random sampling. To ensure allocation concealment, the randomisation will be performed by the appointed trial statistician using the “sample” command in R.

The units of randomisation are villages in rural areas and administrative wards in urban areas. A cluster RCT is deemed preferable to individual randomisation because it will reduce the risk of spillover to the control arm.

Yes

150 clusters

2,250 spouses, i.e. 2,250 wifes and 2,250 husbands

Treatment: 1,125 spouses, i.e. 1,125 wifes and 1,125 husbands across 75 clusters
Control: 1,125 spouses, i.e. 1,125 wifes and 1,125 husbands across 75 clusters

We have conducted power calculations using Optimal Design Software, inferring the assumed effect size from two previous RCTs conducted in South Africa and Rwanda that assessed women’s financial decision-making power in interventions similar to the one proposed here. We used a lower-bound minimum detectable effect size of 0.25 and a relatively high intra-cluster correlation (ICC) of 0.2, showing that a sample size of 130 clusters with 15 spouses per cluster would be necessary to ensure 80% power at a significance level of α=0.05%. We will oversample by 15% to account for the risk of loss to follow-up, thus yielding a final cluster size of 150 communities (75 per trial arm) and a total sample size of 2.250 husband-wife pairs (1125 per trial arm). The true intra-cluster correlation in the trial’s target community will be determined after completion of the baseline survey, which will allow for possible adjustments should it exceed the assumed ICC of 0.2.