LSE - Small Logo
LSE - Small Logo

Editor

March 21st, 2017

How nudge units can aid development

2 comments | 4 shares

Estimated reading time: 5 minutes

Editor

March 21st, 2017

How nudge units can aid development

2 comments | 4 shares

Estimated reading time: 5 minutes

As the Government of India moves forward with plans to establish a nudge unit, Feroza Sanjana outlines what this behavioural approach to policymaking actually entails. She writes that while nudging can provide a powerful tool for academia to generate real-world impact, there are some strong caveats against unbridled enthusiasm.

The news that the Modi government is setting up its very own Nudge unit comes as a significant evidence-based turn for policy-making in the country. Through the several years in India’s development story, there has been no dearth of detailed and well-intentioned policies. However, its translation into intended outcomes has often been stymied by inadequate implementation and/or a poor understanding of realities at the ground level. Administrative corruption, lack of bureaucratic will and low uptake of welfare policies by citizens themselves are but a few well-known issues.

Behavioural public policy proposes an innovative solution. From the appointment of nudge pioneers Cass Sunstein and Richard Thaler to the White House, to the setting up of “nudge units” across the United Kingdom, France and Singapore, leaders are increasingly recognising the utility of applying insights from cognitive and social psychology to policymaking. Indeed, the move holds much promise to improve the design and implementation of public service delivery and government outreach programmes. But what does “nudging” really mean? And what will it imply for development practice in South Asia?

Harm to self

Nudging is motivated by two core insights into human nature. The first is that we do not always make decisions in our own interest. One reason is because even though we are intendedly rational, we are hard-wired with decision-making biases including but not limited to: confirmation biases (seeking information that confirms our existing opinions), hyperbolic discounting (preferring immediate gains to future gains, even if future gains are larger) and status quo bias (preferring stasis to change). Another reason why people act against their self-interest is because they lack the support to necessary to make choices that benefit them.

Nudge tries to improve people’s choices through simple tweaks in the way choices themselves are presented. Take for instance, an experiment run by the UK’s Behavioural Insights Team to improve the low rate of road tax payments even though non-payment could result in the owners’ car being clamped or towed away. The team found that while sending a reminder letter to one group garnered an 11 percent response, adding the headline “Pay your tax or lose your car” to the letter doubled the payment rate in comparison to the first group. Similarly, adding a photograph of the person’s car to the third group tripled the results.

Nudging also can involve offering social support to enable better choices. For example, the Abdul Jameel Latif Poverty Action Lab (J-PAL) under Professors Elisabeth Duflo and Abhijit Banerjee, which has been the pioneer in running several hundred randomised control trials (RCTs) to improve the quality of evidence informing programme design in the developing world, found that they could increase attendance rates in primary schools in Kenya by offering children de-worming tablets and offering iron tablets in anemia-stricken areas in India.

Harm to society

The second insight that informs nudging is that people do not always behave in ways that benefits society.  This can be not only because of self-interest but also because people may not have the requisite information to make better choices. A recent study by the International Growth Centre to increase motivation of civil servants to collect property tax in the Punjab province compared the impact of three different types of performance-based incentives on revenue collection. It showed that increasing the salary of tax collectors increased the growth rate of tax collection by 46 percent in treatment areas as opposed to 28 percent increase in comparison areas while at the same time, not reducing tax-payer satisfaction.

A simple behaviourally-informed intervention can thus demonstrate actionable results. It is not surprising then that nudging has garnered strong interest in recent times, appealing to a very intuitive need to understand how people behave as well as focussing on what actually works. As such, it can provide a powerful tool for academia to generate real-world impact.

However, there are some strong caveats against unbridled enthusiasm. Firstly, designing an intervention requires an intricate and fine-grained understanding of the culture and history of the setting within which an intervention is planned. Secondly and conversely, this implies problems of achieving external validity.  In other words, an intervention that is found to work in Tamil Nadu need not work in Rajasthan due to differing culture and context. In addition, the same intervention may not work to address problems which at first glance, appear to belong to the same family or genus. Revisiting the example of provision of tablets for intestinal worms, J-PAL found that when they realised that girls did not attend school on the days that they were menstruating, providing them with menstrual cups  did not lead to the same success. In a related point, not all nudges work  due to counter-nudges emanating from other actors or obstacles in the participant’s environment which might defuse the choice architect’s efforts.

Thirdly, testing and scaling up projects needs time and therefore political will to keep the unit alive. However, the increasing commitments from some governments, particularly in Andhra Pradesh and Tamil Nadu in India is promising. Fourth, we need to continuously improve our monitoring and evaluation methodologies to really understand what works and why. Currently, RCTs are the main method which take a leaf out of  drug-trialling practices in medicine, where one or more group(s) are randomly selected to receive a treatment while the control group does not receive treatment to observe the impact of an intervention. Indeed, aside from issues of external validity, as Professor Vivek Dehejia in his article points out, RCTs cannot pinpoint the causal mechanism by which the intervention actually produced the desired outcome. This is not insurmountable however, requiring researchers to spend more time articulating a theory of change. A more pressing challenge is that RCTs only provide the average effects which may mean that researchers overlook the distribution of impact of an intervention, particularly if an intervention affected some participants adversely. But as the UK’s Behavioural Insights Team suggests more qualitative research like surveys, in-depth interviews and focus groups can help understand better how individual participants react to an intervention. On a practical note then, training government officials and field researchers in data collection methods for monitoring and evaluation will be crucial. It also means that strong government-academia linkages will be important to continually improve, learn and grow.

Finally, we need to think carefully about the normative questions surrounding the use of nudges. Choice architecture runs the risk of governments appearing paternalistic, by claiming to know what’s good for people – more than citizens themselves. It will also be important for governments, citizens and academics to engage in a dialectic over questions of what we mean exactly by words like welfare and development, particularly with a focus on sustainability and the environment. But by being driven by a commitment to understand the challenges the underprivileged face in their everyday lives and to provide targeted solutions, nudging certainly provides the tools for academics to effect social change.

Cover image credit: Anna Tanczos. Wellcome Images CC BY-NC-ND 4.0

This article gives the views of the author, and not the position of the South Asia @ LSE blog, nor of the London School of Economics. Please read our comments policy before posting.

About the Author

Feroza Sanjana completed her MSc in Philosophy and Public Policy at the LSE and is currently a PhD candidate at the Berlin Graduate School for Transnational Studies.

Print Friendly, PDF & Email

About the author

Editor

Posted In: Development | Featured | LSE

2 Comments

Comments are closed.

Jaipur Palace

CONTRIBUTE

South Asia @ LSE welcomes contributions from LSE faculty, fellows, students, alumni and visitors to the school. Please write to southasia@lse.ac.uk with ideas for posts on south Asia-related topics.

Bad Behavior has blocked 729 access attempts in the last 7 days.