Behavioral approaches to development economics and policy have gained momentum in recent years. A growing number of papers studying behavior of people in poor countries have been published in top journals, accompanied by the rise of randomized controlled trials (RCTs). In 2015, the World Development Report was dedicated to behavioral and cognitive research and policy. Papers studying how to nudge farmers to use fertilizers or increase savings have become classics in the field. Lots of hope has been placed into social experiments and behavioral policies to fight global poverty.
Behavioral policies are of course not reserved for policy-making in poor countries. In fact, nudges became famous with a US-American savings plan. Many behavioral instruments have been discussed and tested in and for rich countries. But there has been an important difference as compared to the debates in development economics: when debating behavioral policies in rich countries, scholars have also devoted lots of time to consider normative and ethical concerns. For example, following Thaler and Sunstein’s exposition of Libertarian Paternalism (see also here), a debate unfolded on whether nudges could be anti-libertarian (here, here, here, or here). Implications of the use of nudges as a new form of government policy have been analyzed, for example, from a Foucauldian perspective, or with a focus on institutional change. Books have been written about ethical concerns. The debate has reached a great level of differentiation, e.g. when authors argue that so-called social nudges (these are nudges that seek to stimulate voluntary cooperation in social dilemma situations) may be justified for different reasons than those targeting individual welfare. Overall, the debate has become really sophisticated, and the autonomy, welfare, and dignity of citizens in rich countries as well as consequences of the use of behavioral policies for these countries’ modes of government have received lots of careful scrutiny (recently again here).
Double standards for the North and South
Much less so in developing countries: here, effectiveness rules. The bulk of behavioral development economics is first and foremost concerned with questions of how most effectively to design behavioral policies. In their 2008 review of behavioral development economics, for example, Cardenas and Carpenter focus on lessons to be learned from field experiments, and the design of those, but they do not mention normative considerations. Likewise, in Duflo and Banerjee’s Poor Economics, the chapter that comes closest to welfare implications is concerned with how to correctly implement policies in the face of political dynamics. This recent paper by Christian Berndt and Marc Boeckler summarizes and critically discusses this rise of behavioral policies (see also here for a comparison of Global South and North). Of course, you may say, Thaler and Sunstein do not debate the normative implications of their policy proposals, either – but in rich countries, lots of other people have taken over this job. All these papers do not explicitly exclude the relevance of their arguments for poor countries. But in behavioral development economics, this literature is not referred to. This opinion piece by Guy Standing is an exception in that it discusses RCTs and behavioral interventions in poor and rich countries together.
There is, thus, the risk that development economists are too quick in implementing their RCTs, without sufficiently considering the impacts and side-effects on the lives of the people they ‘treat’. Yes, it may well be that “informal savings technologies can substantially increase investment in preventative health and reduce vulnerability to health shocks”, as Dupas and Robinson summarize their 2013 paper. But we would be on the safe side if we knew that the money that is being saved is not missing elsewhere, that the savings technologies do not create social tensions, or that the intervention does not influence behavior in different ways than those intended. By the way, if we want to improve health care provision for poor people, public policies are often much more straightforward than behavioral interventions on poor individuals. We should not forget that originally, behavioral policies were advocated as compliments and not substitutes.
Let’s not take for granted that behavioral changes improve people’s lives
Or consider the following intervention by Bernard, Dercon, Orkin and Seyoum Taffesse that was conducted in Ethiopia. The authors were concerned with prevailing low levels of aspirations. Low and downward-biased aspirations may lead to psychological poverty traps when people forego opportunities to improve their lives (see here and here). It is understandable that researchers or development practitioners want to avoid situations in which poor people do not make use of their full potential. A life in material hardship is difficult enough, and many living conditions are difficult to change – so why not go for apparently low-hanging fruits? The researchers developed and tested an intervention aimed at influencing presumably fatalistic aspirations of poor Ethiopians. In an experiment, they implemented a one-day treatment. Participants in the treatment group watched movies about people from similar backgrounds who had successfully set up a small business or agricultural project. Success, as reported in these stories, was achieved through goal-setting, hard work, and perseverance. The researchers argue that this intervention has had positive effects on future orientation and optimism: in the treatment group, reported aspirations, as well as savings, the use of credit, and investment in education had all increased six months after the intervention. The World Development Report 2015 praises this intervention, saying that it illustrates the ability of an intervention to change a mental model — one’s belief in what is possible in the future” (p. 4).
Unfortunately, things are often not so easy. How do we know that increased goal-setting and perseverance will improve the living conditions of the people in the treatment group? For example, interventions that promote optimism run the risk of provoking frustration. Particularly under uncertain living conditions, being ambitious and hopeful might be rewarded on some occasions, but not on others. Not expecting too much may thus be a strategy that reduces psychological costs. Of course, you may object, this strategy may make people forego chances that may have improved their living conditions. Yes, sure: when poor people have adjusted to their poor living condition, this is certainly no situation worth being conserved. But then, behavioral policies should not be our first instrument of choice. As the authors of the Ethiopian experiment write in their last paragraph: “Are we giving false hope? We cannot judge this. But we did not suggest to individuals – rightly or wrongly – what path would lead them out of poverty, unlike most interventions that offer ‘solutions’ in microcredit, health or education. We only invited our treatment group to listen to stories told by their peers from similar backgrounds. The extent and nature of their response has surprised us.”
We must take ethics seriously in the Global South, too
This clearly invites us to pay close attention to what behavioral interventions actually do with the people they target. Systematic considerations of both intended and unintended effects and their normative evaluations can help improve the design of behavioral policies. For example, there has been some normative discussion on the policy implications of aspiration traps (here, here, and here) from which we can derive some concrete directions and constraints for the design of policies. (In a nutshell, policy interventions must in most cases tackle aspiration traps in a policy mix alongside traditional policies, and they should never push people into specific directions.) Normative evaluations, broadly conceived, should be undertaken across the board – to design better policies and experiments, and because the autonomy and dignity of people in the Global South should be of no less value to development researchers and practitioners.
Svenja Flechtner is a Post-doc research assistant at Freie Universität Berlin.
Photo by Nick Youngson.