By Sergio Torres Rueda (London School of Hygiene & Tropical Medicine)
The scale of the problem
At the end of 2018 we presented findings of our cost analysis of programmes to prevent violence against women and girls (VAWG) in a meeting in Kathmandu. Professor Charlotte Watts, one of the pioneers of research on VAWG, chaired our session. In her opening remarks she mentioned that in some areas of VAWG research, such as the study of the effectiveness of prevention strategies, the scientific community was already halfway up the mountain. However, when it came to understanding the economics of VAWG, we were still very much at the foothills.
Nearly a third of women worldwide have experienced physical and/or sexual violence during their lifetime. That is hundreds of millions of women. Yet, to date, only two economic evaluations of VAWG-prevention programmes have been published (IMAGE and SASA!). Momentum has been building for states to take greater action in preventing violence and we now have a better understanding of the kinds of prevention interventions that work. However, governments seriously considering investing and scaling up violence-prevention initiatives have scant evidence to understand the resources needed to do so. The body of research in this area is simply not commensurate with the scale of the problem.
What will it cost?
We present the costs involved in developing, adapting, setting up, and implementing six violence-prevention interventions across six countries in sub-Saharan Africa and South Asia. The interventions vary in terms of their delivery platforms, settings, and target populations. The cost data collection was carried out alongside randomised controlled trials and was part of the What Works to Prevent Violence against Women and Children (‘What Works’) programme, a DFID-funded multi-country effort to build the evidence base on violence prevention.
We found that the total costs of violence prevention programmes vary greatly depending on the types of activities involved, platform of delivery and intensity. The cost per beneficiary reached varied between US$4 in a community-based intervention in Ghana and approximately US$1320 for a counselling intervention in Zambia. VAWG-prevention interventions are staff intensive: staff costs made up the greatest share of costs across all interventions, and total costs were highly sensitive to staff salaries. Substantial costs had to be spent before intervention could be rolled out. Training of staff was a large investment; ensuring retention of staff is therefore important. We found that adapting interventions to the local context could be resource-intensive and costly, yet necessary, when implementing behaviour-change interventions originally developed elsewhere.
We used cost data at the pilot level to model costs of scaling up five interventions at the national level. We found that the cost per beneficiary would change when interventions are scaled up. These unit costs would likely decrease for community-based and workshop-based interventions as fixed costs got spread over a larger number of units (in other words, economies of scale). However, the cost per beneficiary may increase in interventions with fixed platforms (such as those rolled out in schools) when the average ratio of students per school is lower at the national level than at the pilot level, due to high fixed costs at the delivery site.
Our findings help fill some of the evidence gap needed to make policy decisions, but they also raise further questions, particularly when thinking about costs at scale and over time. While the programmes costed had different delivery sites within one country, this variation will not capture the full degree of heterogeneity that could be expected should the programmes be expanded to the entire country. When expanding to the national level, programmes may need to overcome constraints that were not encountered in the pilot. For example, potential workers and volunteers may have different skill levels and may therefore require greater training or more remuneration. Beneficiaries may be located in more remote areas and thus increased transport costs may be incurred. Cultural and linguistic variation may require re-adapting curricula. Overcoming these types of constraints will require additional resources. We advocate that, as programmes expand, implementers build cost data collection into their monitoring and evaluations in order to shed light into some of these key question. Partly with this in mind we developed a set of publically-available cost data collection tools that could be useful to implementers.
Another area of uncertainty relates to how costs may also change over time. Several of the interventions costed rely largely on volunteer labour. Training costs are high and total programme costs are highly sensitive to the useful life of training, in other words to how often training needs to take place. In order to keep costs low, it will be important to ensure that volunteers are retained in service in the long term. To do so, we need to understand what kinds of financial and non-financial incentives have the greatest impact on volunteer motivation.
It is, of course, important to mention that understanding costs only gets us half-way there. Costing studies are helpful to understand the resources needed to invest in programmes. But on their own they do not tell us whether a programme presents good ‘value for money’ compared to another violence-prevention programme, or compared to other possible investments within the health or other sectors. The next logical step for our team has been to combine the cost data presented here with the effectiveness data that emerged from the six trials to understand the cost-effectiveness of the different interventions. We look forward to published our findings very soon.
Conclusion
VAWG is an enormous problem. Everywhere. But we know that VAWG is preventable—and now we have a better idea of how much it is going to cost to prevent it. We hope that our work takes us one step further up the mountain and helps inform policy decisions that have positive impacts on women’s lives.
Image credit: Indashyikirwa intervention in Rwanda. The photo shows women making baskets in a ‘Women’s Space’ in a rural area, a component of the intervention. These were safe spaces where women could participate in group activities and receive sensitisation and support.
