Beyond the confines of one’s private space, the geographical unit that is real and lived is a village. A village is not only a platform for social engagement, it is also the unit where public policies and programmes come to fruition. Yet, villages are not part of the mainstream public policy or academic discourse in India. Meanwhile, the geographical unit that dominates the policy discourse in India is the district.

Recent efforts such as identifying “aspirational districts" (i.e., districts that are lagging on health and human development indicators and can be prioritized for intervention) or developing “district rankings" are prime examples of an exclusive focus on districts in policymaking. But focusing solely on districts can be grossly misleading.

Consider Haridwar, an aspirational district in Uttarakhand. In Haridwar, 39% of children under the age of 5 are stunted (an indicator of chronic undernutrition). It has 1.2 million inhabitants across 518 villages. According to our estimates from the 2016 National Family Health Survey (NFHS), stunting prevalence in Haridwar varies between 0% and 85% across its villages, rendering the district average of 39% practically meaningless. More generally, in the presence of large village-level variation, in what sense is it useful for a particular district administration to know that it is an “aspirational" district? In all likelihood, the district administration already knows that.

As per the 2011 Indian census, a district, on average, has about 1,000 villages with a rural population of about 1.3 million. If all the 1,000 villages within a particular district are identical, then the district administration can randomly select any village and implement whatever is considered to be an effective intervention. Unfortunately, the reality is that villages within a district vary a lot.

Indeed, a map showing, say, the prevalence of stunting among children (one of the key indicators for the recently launched Poshan Abhiyan) across districts will always appear to vary quite substantially. For instance, the prevalence of stunting among children varies between 13% and 65% across districts. But simply because districts differ on a set of indicators, it does not automatically mean that those differences are uniquely attributable to districts. Yet, interpreting “district differences" as being the same as “the difference districts make" is widespread in policy as well as academic discourse. Such conflation needs to be avoided.

A recent study conclusively showed that the net contribution of districts to explaining variation in household poverty was insignificant. Meanwhile, the units of action with regard to variability in poverty were villages and states. This observation is likely to hold for health and nutrition indicators, given their strong connections to poverty. Thus, without knowing which villages to target within a district, identification alone serves little purpose.

Seventy per cent of India’s rural population resides in 645,856 villages. It is time to empower the village population and its governments by collecting and sharing important health, nutrition and other socio-economic indicators. The decennial census on villages on selected demographic and socio-economic variables is too little, and too late.

The Sansad Adarsh Gram Yojana (SAGY) presents an effective and accountable policy framework for governance based on data. Initiatives such as supporting parliamentarians on analysis and research in constituency (Sparc) can be extended to village-level governments and linked to elected representatives at the state and Central levels. The Maharashtra Village Social Transformation Foundation, a collaborative effort involving multiple stakeholders, including the government of Maharashtra, is an important step in the right direction to realizing SAGY’s goal of every village having a tailor-made village development plan.

Other data monitoring efforts, such as the Health Management Information System, Mother and Child Tracking System and Integrated Child Development Services, can be linked to village-level data initiatives.

For a district to reduce its prevalence of stunting, some of its villages must reduce their stunting prevalence, and for the villages to see a reduction in stunting prevalence, children residing in these villages must see significant improvements in their nutritional status. Put simply, precision in public policy is critical. And, precision public policy requires precision data.

The 115 aspirational districts prioritized by the Union government for action have approximately 120,000 villages. Public policy needs to develop surgical precision to identify which of these villages needs be prioritized. Importantly, if there are large variations between villages within a district, that implies that there are successful villages that can be leveraged to serve as a possible model for the lagging ones.

In 1931, after traversing the length and breadth of India, Mahatma Gandhi observed, “India does not live in its towns but in its villages." Over 85 years later, Prime Minister Narendra Modi reiterated Gandhi’s insight, stating, “If we have to build the nation we have to start from the villages." It is time for mainstream policy discourse to meaningfully integrate the adarsh gram and “aspirational district" perspectives into policies aimed at improving the health and well-being of the country.

S.V. Subramanian, William Joe and R. Venkataramanan are, respectively, professor of population health and geography at Harvard University, assistant professor at the Institute of Economic Growth, and managing trustee at Tata Trusts.