Rajeev Arora & John Floretta
The Royal Swedish Academy of Science awarded Abhijit Banerjee, Esther Duflo and Michael Kremer the Nobel Memorial Prize for their “experimental approach to alleviating global poverty” in Stockholm.
Fundamental to these economists’ methods is working in communities to rigorously evaluate promising anti-poverty policy ideas to learn what works and why in the real world. Testing innovations within the government requires robust partnerships between policymakers and researchers.
The department of Health, Government of Haryana, has been partnering with Banerjee, Duflo, and Abdul Latif Jameel Poverty Action Lab (J-PAL) staff since 2013 to apply experimental approach to improve immunisation coverage. In a large-scale randomised evaluation, involving nearly 3,00,000 households, we learnt how to build demand at scale for primary healthcare services. The project led to large gains in immunisation coverage and has the potential to inform health policy in the state and across India.
Immunisation cost-effective way
Immunisation is one of the most cost-effective ways to improve child health and survival. Yet in India, approximately 10 million children are not fully vaccinated. The collaboration, anchored by the Haryana National Health Mission and State Health Systems Resource Centre, complemented the government’s broader efforts to improve access to immunisation by testing a range of interventions designed to boost demand in the Aspirational District Nuh and six other high priority districts. It was inspired by previous J-PAL research in Rajasthan, which demonstrated how the provision of non-cash incentives to families positively impacted full immunisation rates among children. However, could a programme informed by these insights be implemented at scale and through the government? How effective and cost-effective would the programme prove in another context? These questions together were explored in Haryana over the past six years.
To better understand the context and immunisation coverage in these seven districts, a baseline survey was conducted in 15,000 households. Then, 1,400 frontline nurses were trained on how to use handheld tablets to generate real-time data on implementation. To identify how to build demand for immunisation, 140 Primary Health Centres and 755 sub-centres were randomly assigned across seven districts to test several different combinations of small incentives (in the form of mobile credit to caregivers), monthly text message reminders, and leveraging the village social networks to disseminate information compared to “business as usual” services being implemented.
The first insight attests to the powerful influence of social networks to shape health seeking behaviour. Simply asking people in the village, who in the community is good at spreading information, and then sending monthly text messages about vaccination to both these nominated community members and caregivers increased full immunisation rates by 25 per cent.
The result is particularly interesting given the fact that sending SMS messages about immunisation to caregivers alone did not have an impact, nor did mobilising the community members, who were nominated as trusted sources of information. It was specifically identifying and mobilising community members, who are socially well-connected and loquacious that was effective. These people in turn increased awareness in the village of the benefits of vaccination as well as awareness of when the next monthly vaccination camp would be held.
The simplicity and power of this model is that to identify who is well-connected, all we needed to do is ask the community. To mobilise these nominated community members as “ambassadors for immunisation” we just needed to send them a monthly SMS message. Given the relative ease of implementation and large magnitude of impact, the intervention proved highly cost-effective and scaleable. Second, we learned that adding to this model, small mobile credit incentives to caregivers each time they vaccinate their child, structured in increasing amounts, increased full immunisation rates by 45 per cent compared to “business as usual”. An incentive model can be particularly effective in low immunisation coverage areas, where an extra push is needed.
Why the collaboration
The Health Department and J-PAL are partnering to translate the results of this study into action to inform the state’s routine immunisation programme across all 22 districts. These efforts will substantively improve full immunisation rates. The project also demonstrates large gains in health coverage that can be generated by increasing demand for health services from the public. It can be inexpensive to increase demand through smart, innovative approaches, which complement ongoing investments to increase healthcare access. J-PAL is interested in leveraging this learning from Haryana to work with other states in India to inform their health programmes.
The collaboration has demonstrated how a curious government department, open to innovation and rigorous research, can improve social policy. Before the project, neither the government of Haryana, nor J-PAL knew which combinations of programmes would prove most effective. Our partnership shows the benefits of designing policies informed by theory and research and then rigorously testing these innovations at scale. Making this type of policymaker-researcher collaboration work at multiple levels in the state (from villages, to Primary Health Centres, to the Secretariat in Chandigarh) required mutual trust and adaptability. The government needed to be willing to incorporate research protocols into the rollout of new programmes and academics had to adapt to the government’s administrative requirements. This case shows the magic that can happen when both sides learn how to work together: better anti-poverty programmes, which improve lives and contribute to global knowledge.
In September at the Massachusetts Institute of Technology (MIT), the Haryana Health Department and J-PAL signed an MoU committing to another five years of partnership, together with IFMR, J-PAL’s host institution in South Asia.
High-quality health data
There is tremendous promise to leverage the high-quality health data the government is now digitally collecting at the village level through tablets and m-health apps. This administrative data will be analysed and used to inform the design of the next generation of health programmes. These progammes would be rigorously evaluated in the field, as was done for immunisation.
In turn, the findings would be integrated into the state’s health policy. We are optimistic that investing in this iterative cycle of innovation, testing and learning, will accelerate progress towards a healthier Haryana.
(Writers are Additional Chief Secretary, Health, Govt of Haryana, and Director of Policy, J-PAL)
A Bench headed by Justice Arun Mishra will hear arguments on...
Born to a Jamaican father and an Indian mother, 55-year-old ...
Death toll crosses 48,000