Two ideas in development – activating agency of citizens and using “nudges” to change their behavior – seem diametrically opposed in spirit: activating latent agency at the ground level versus  top-down designs that exploit people’s behavioral responses. Yet both start from a psychological focus and a belief that changes in people’s behavior can lead to “better” outcomes, for the individuals involved and for society.  So how should we think of these contrasting sets of ideas? When should each approach be used?

By way of context, in May 2018, I joined a conference and workshop in Cerrito, Paraguay devoted to reducing poverty, hosted by the Fundación Paraguaya, a social enterprise with a mission to reduce poverty through microfinance, education and social change.  The centerpiece of the workshop was the approach to tackling poverty embodied in the Poverty Stoplight, a multi-dimensional, participatory poverty diagnostic tool, developed by the Fundación, now used by many nonprofits, and some private firms, around the world.  The essence of the Stoplight is activation of the agency of the poor themselves. This activation happens through participatory diagnosis in itself, the selection of specific goals for change, and ongoing interactions with various forms of coaches, such as microfinance agents or other front-line workers.  (We at IMAGO are working with the Poverty Stoplight.)

In one session of the conference, Oscar Calvo-González, of the World Bank, gave a comprehensive summary of thinking and practice in behavioral economics. He offered examples of a range of experiments in which policies have been designed to explore how specific design tweaks can lead to different individual decisions.  A famous example is whether people are given the choice of opting in or opting out in a decision—from pension plans to organ donation.  When the active choice involves opting out, substantially larger fractions of participants choose to take the option (i.e. join a pension plan or an organ donation scheme).  People aren’t rationally calculating actors, but work with simplified mental models, or go with the perceived social herd. Some governments have set up “nudge” units, to explore how to more effectively get citizens to change behavior and achieve better social outcomes. More notoriously, related approaches have been used to nudge voters in favored directions—as alleged in the UK’s Brexit referendum and the 2016 U.S. presidential election.

Let’s compare the two approaches with respect to diagnostic frame, practice and ethics.

Diagnostic frame.  

The common ground is recognition that people use short-cuts for decision-making, in ways that can hurt their own interests.  In both approaches, there is an emphasis that decision-making is particularly tough for poor people, given the sheer weight of daily problem-solving.  In behavioral economics one core idea is that we have limited mental “bandwidth” and this form of scarcity hampers decision-making. However, in the “agency” tradition, there is much more emphasis on unearthing and working with the origins of the prevailing mental models, with respect to social exclusion, stigmatization, and the typically unequal economic and cultural relations with respect to more powerful groups in a society.  One approach works more with symptoms, the other with root causes.

Implications for practice.  

The two approaches on display in Cerrito both concern social gains, and both involve a role for an external actor.  But here the contrast is sharp. In the “nudge” approach the external actor is a beneficent technocrat, trying out alternative offers to poor (or non-poor) people to improve outcomes.  A vivid example is alternative messages to tax payers in Guatemala, that induce varying improvements in tax payments.  In the “agency” approach the essence of the interaction is between a front-line worker and an individual or family, with a co-created diagnosis and plan, designed around goals and specific actions that the poor person chooses.  This is akin to what anthropologist Arjun Appadurai termed increasing the “capacity to aspire,” and can extend to greater engagement in civic and political life.

Ethics.

In both approaches, ethics is central.  As implicated in the “nudging for social good as opposed to electoral gain,” some form of ethical regulation is surely needed. In “action to activate agency,” the central ethical issue is of maintaining equality in design between activist and citizen, and explicit owning of any decisions.

What does this imply?

To some degree this is a question of domain of action.  Nudging is most appropriate in a program for which there is a fully supported political and social program, and the issue is how to make it work (as in paying taxes).  The agency approach has a broader ambition, but starts from domains that are potentially within an individual’s control once the sources of “ineffective” or inhibited behavior are tackled, including via front-line interactions with public or private actors.

In IMAGO we have firmly chosen to work in the domain of activating agency. We believe we need to turn the way we think about development upside down — shifting the approach to one in which the poor get to define and design. That is the difference between nudging (recognizing the behavioral) and activating latent agency (disrupting the system to effect transformation). Finally, it is noteworthy that behavioral economics has built an impressive range of specific evidence from experiments in the past decade or so.  While there is a wealth of material on inequalities or suppression of agency, more evidence is needed on what kind of public action is effective.