As behavioral scientists, we know that context matters. But in India, it matters in new and sometimes surprising ways. As a country of nearly 1.3 billion people across 28 states, speaking more than 20 officially recognized languages, India is not just associated with diversity, it is diversity.
This reality means translating the model of a behavioral science team or nudge unit in, say, the United Kingdom or Australia to India is not so straightforward. The challenge for Indian behavioral experts at the forefront of this translation is not only to design for the context of the population being nudged, but to consider the diverse contexts of those who are doing the nudging, too.
To appreciate this last challenge, it’s helpful to know a little more about India’s union government, which is housed in the national capital, New Delhi, and its 28 state governments. The union is responsible for developing foreign, economic, and defense policy (to name a few areas). State-level administrative officers implement policies, collect taxes, and oversee development within smaller constituencies in each state. But the national and state-level governments often work together.
Many of India’s policies are implemented by bureaucrats at the union or state (or lower) level, meaning that it’s possible that behavioral interventions can be implemented at multiple levels and by multiple stakeholders. It’s therefore possible, and even likely, that individuals working in national offices won’t have sufficient local context to understand how a nudge might be translated and adapted—and won’t have the foresight to know which parts of an intervention could be misguided.
The challenge for Indian behavioral experts is not only to design for the context of the population being nudged, but to consider the diverse contexts of those who are doing the nudging, too.
If national-level policymakers lack context, local-level bureaucrats can lack time and resources. Let’s take the case of our experience with the Mumbai Metropolitan Region (MMR), where we implemented a recycling intervention to encourage residents to reduce plastic waste going into landfills. Mumbai has a population of nearly 25 million people (the combined population of the 10 most populous cities in the United States of America, or the entire Australian continent). This means improving waste management is a big task, but also one with the potential to make a huge positive impact. Much urban management is conducted by staff of the local government, or a development authority (e.g., MMRDA). These organizations are generally comprised of a group of local actors with differing levels of motivation, context, and bandwidth to accommodate a behavioral intervention like ours.
High-visibility infrastructure projects (e.g., Metro Rail) or pressing agrarian issues (such as access to irrigation for farmers) often consume their time; they’re expected to juggle larger-scale, critical projects at the same time as monitoring smaller behavioral interventions. We worked with one department, the Chief Minister’s Fellows, who are particularly bandwidth-strapped: in one typical “war room” meeting for this group, they examined nearly 30 different government projects in the span of two hours.
This lack of time and resources had a clear effect on our intervention. After we provided feedback to households on their recycling, and intended to start measuring the results, we discovered that households didn’t have enough plastic-only recyclable bags (which they needed to properly sort their waste). This is such a simple issue. And it could have been easily avoided with timely troubleshooting and closer monitoring.
Another contextual feature of national and state policymakers in India is the desire to establish the merit of using behavioral science to inform policy. At the outset, this seemed fantastic—a government that wants to apply science to improve outcomes across a number of policy areas. However, whether behavioral science is being used is often given more importance than how it is being used, leading to stumbles in implementation.
We experienced one of these stumbles ourselves during a training for the first part of our recycling intervention. Our plan was to train residents to separate “waste”—plastics, paper, dry and wet waste—and then give them feedback on how well they were doing.
Whether behavioral science is being used is often given more importance than how it is being used, leading to stumbles in implementation.
About midway through the training, as we were explaining what should go in the plastic bin, one resident turned to a woman beside her and said, almost as a realization, “Are you listening?” This woman was her housemaid. At that moment, we realized that we’d been targeting the wrong people. Handling waste is typically delegated entirely to house maids (typically women from lower castes) thanks in large part to notions of purity and pollution; people from lower castes are still often perceived as less pure, and therefore perceived to be better positioned to handle refuse. (We touch briefly on how behavioral science and castes can collide below.)
We’d also been speaking the wrong languages—in a mix of English and Gujarati that the residents would comprehend. Luckily, a local policymaker was there to help translate our training into Marathi, so the domestic workers gathered could understand the instructions.
At first glance, the solution to the problem we ran into seemed simple: adapt the processes and training guides to speak to a different audience. But this solution didn’t address an underlying question, which took root long before our visit: Why didn’t policy stakeholders and implementing agencies see this coming? As our work in Mumbai continued and similar problems sprang up, we sensed a pattern: while behavioral science practitioners repeat the mantra that “context matters” in reference to the population they are targeting, what is less discussed is that context matters for the nudgers too.
So, how do we better address the unique context of eager Indian behavioral science units, which, depending on their position, may be context blind, resource strapped, incentivized to focus on high-visibility political projects, or all of the above?
First, we could experiment with a new system of policy decision-making and implementation, creating a model that makes knowing the local context a need-to-have standard, rather than a nice-to-have, for implementing a behavioral intervention. One way to do this: let local and subnational governments set up smaller, dedicated nudging teams to tackle micro problems and share their lessons with a national team that can then build capacity for setting up similar teams in other cities or states. The role of the national behavioral science team would therefore be to translate lessons from a target city into the context for a similar one, whereas the city-or-state-based teams have the autonomy to tailor interventions to their specific areas.
Take for example the smog problem in Delhi, which originates in neighboring farming states of Punjab and Haryana. If a union government team tackled it (e.g., a behavioral team from Ministry of Environment, Forest, and Climate Change), that could mean losing important context from state or local government leaders in Punjab/Haryana, who might be able to not only design better but also implement better. In India, a decentralized approach to applying behavioral science is particularly important given the powers devolved to local governments and the sheer diversity within the country.
While behavioral science practitioners repeat the mantra that “context matters” in reference to the population they are targeting, what is less discussed is that context matters for the nudgers too.
Second, to circumvent the resource problems that often plague local actors, the national government could identify pairs of state teams to resolve a common issue, distributing the task and encouraging states to share resources and knowledge with each other. For example, there are high rates of child deaths due to malnutrition in two states: Rajasthan and Uttar Pradesh. Rajasthan, in western India, has average income levels that are nearly double in that of Uttar Pradesh. While the underlying reason for malnutrition in these states may be different, the administrative resources may differ too. A public health official in Rajasthan could have the resources to hire a behavioral science consultant to help design a behavioral intervention, which could be out of reach to her counterpart in UP. Why not encourage these states to coordinate on a shared problem, exchanging ideas, time, and resources?
Third, it is entirely possible that caste differences intersect with the implementation of nudges (as it did in our recycling intervention), making it an important and unique factor to address. As the World Bank co-head of the Mind, Behavior and Development Unit Varun Gauri suggests, policymakers must keep a close watch on such inequities that may be perpetuated by behavioral interventions, especially in the Indian context. The Indian caste system and its associated discrimination goes far beyond traditional occupational roles, making it vital to consider when designing nudges in seemingly innocuous contexts (like waste management).
Perhaps one of these approaches could have helped the success of our recycling intervention, which found no discernible effect on behavior. Although that was disheartening, we couldn’t help but think about the context within which that intervention was implemented (with a team of chief minister’s fellows who were definitely motivated to establish proof of concept for implementing nudges, but were likely ill-prepared to see the intervention through). In the end, households were left with some half-baked ideas about recycling, a little experience, and not much else to show for it. Barely registering the null results, most of the policymakers were excited to move on to the next idea. They were still only all too eager to suggest that nudging had a lot of promise for India, even if they, too, had not much to show for it.
It seems that enthusiasm to implement behavioral science in policy work is a necessary but not sufficient condition—and it is important to ask what constitutes the latter. In the large and diverse Indian context, the willingness and ability to learn from smaller pilots such as these may be critical in realizing the potential for nudging India forward.