The Marginalized Mind

Anandi Mani on the intersections of psychology, poverty, and agency

In its relatively short history, behavioral economics has revolutionized our understanding of how human psychology and economic conditions intersect to influence behavior and decision-making. And while the bulk of this scholarship has focused on the lives of people living in affluent countries, largely due to the availability of resources and research subjects, an important and growing literature also sheds light on the lives of people experiencing poverty in low- and middle-income countries.

Anandi Mani, a Professor of Behavioural Economics and Public Policy at the Blavatnik School of Government at University of Oxford, has played an integral role in advancing this field. Her work on the behavioral economics of poverty and social exclusion, gender issues, and public good provision underscores the intersection of economics and psychology in developing contexts. One of her most well-known studies, published in Science, presented evidence that financial preoccupations due to poverty can significantly reduce certain measures of cognitive performance – comparable to losing a full night's sleep or experiencing a 13-point dip in IQ. More recently, she has explored the deeper psychological dimensions of poverty and how psychological strengths like self-affirmation, aspirations, and agency can often be powerful tools for addressing poverty.

In a recent interview, Mani spoke about her path to this field of study, how her research has evolved, and how the insights generated by behavioral economics offer important lessons for making development programs and policies more effective and responsive to the needs and limitations of those they aim to assist. 

What drew you to behavioral economics?

I came to economics after completing an undergraduate degree in accounting, and I came to it out of my interest in economic development and the lives of people on the margins. When you live in Mumbai City, it's very hard not to see poverty – and it feels like a truly first-order concern. You want to see if there are things you can do to help fix it.

I also came to economics shortly before the early-2000s ‘empirical revolution,’ when empirical research from developing countries showed that the evidence did not always line up with what the rational choice framework would predict. This started pushing the profession to look more systematically at deviations from rational choice theory. As a graduate student who had just arrived in the US, I recall being struck by how even everyday situations involved so much choice, and hence so many decisions. Ordering a sandwich was not just picking what you wanted, it was choosing the type of bread, choosing the cheese, deciding on extras, etc. – which frankly was mentally exhausting! And I found it odd that economics assumed scarcity of the material resources used to make choices, but not scarcity of mental resources. So, as someone who had not quite drunk the “rational choice kool-aid,” as I like to say, I was naturally drawn to studying these questions through a broader conceptual lens. 

What are the key strands of your research on the psychological dimensions of global poverty, and how do they depart from traditional economic approaches?

In the standard economics paradigm, we think about the decisions people make in terms of their preferences and basic material resource constraints, like time or money. But when you bring in the behavioral dimension, there are also what I call ‘internal’ or psychological constraints, which manifest themselves in a lot of different ways. My work has explored a few different forms of such internal constraints, but two key strands stand out.

One strand is mental bandwidth or attention resources – the idea that making decisions, making choices, and dealing with trade-offs can be really cognitively challenging and fatiguing. This is true for all of us, of course; I certainly experienced it in graduate school, as I referred to earlier. But our research finds that the worries of poverty can often capture people’s attention resources, hence impairing their performance on certain basic cognitive (logic or reasoning) tasks. In India, for example, we studied sugar cane farmers both in their ‘lean’ period right before harvest and after harvest, when they are financially better off. We found that the same sugar cane farmer has much higher cognitive function after harvest than during the lean period right before, when they were more stressed about their economic circumstances. And given that people experiencing poverty typically lack a savings buffer, we also find that they have to cope with much higher levels of uncertainty and anxiety. What if the harvest is bad? What if I don’t get a paycheck this month? What if I can’t pay the rent? These can all be a big drain on mental bandwidth. 

The other strand is focused on the deeper psychological and social channels that can disadvantage people who are poor or otherwise marginalized due to their limited life opportunities – such as their (lack of) aspirations or (negative) self-image. The central idea is that, if you feel bad about yourself, it's very hard to motivate yourself or believe that you can do something you can be proud of. I remember reading the chapter in Poor Economics on education, where parents in a poor Indian community say “Oh, you know our kids, they’re just not smart enough to make it through school.” That really stuck with me: as children, we’re so sensitive to how our parents and teachers treat us, and those narratives can have a strong influence on how we view ourselves. Your self-image – the mental model of who you think you are – can end up really shaping how you interpret and respond to every single stimulus in your environment, positive or negative. 

How is your work informed by the concept of agency?

Agency is a good word, in terms of all the ideas it encapsulates. It overlaps with a lot of earlier existing ideas – like some of Sen’s capabilities approach, and certainly the work on aspirations. For instance, how do we set our aspirations? Setting high targets is easy for people from high-achieving environments, but many people are not even aware of the goals they can aspire to reach. I think agency is that ability to introspect, reflect, and visualize what you want, then articulate it and identify a path from where you are to where you want to be.

But when you define what you want, will you define it as something small or something big? That’s a major difference in the sense of agency for people from less or more privileged backgrounds, just by virtue of their exposure. This also gets into the recursive nature of our aspirations; imagine you and I start off with the same level of aspirations, but I am rich and you are poor. Even though we have the same aspirations, as a poorer person you will always find more things hindering your progress – maybe you don't have the right networks, the right resources, or the right information. Maybe you live in an environment with very traditional social norms. Maybe you have to work two jobs and have less time. Over time, our levels of achievement may diverge and you may end up feeling that you are not smart and able enough to succeed. This is the core idea in a paper I wrote with Patricio Dalton and Sayantan Ghosal. Agency is very much hindered and constricted by poverty and limited opportunity, in terms of someone’s ability to set the right goals and then achieve them.

What about your work’s cross-cutting themes, in terms of how these psychological constraints interact with other issues?

Most of my cross-cutting work on these topics has been around gender, in terms of mental bandwidth and psycho-social aspects for marginalized groups. Gender is not the only dimension in which these intersectionalities occur – but around the world, women face many social norms that make us beat ourselves up internally if we don't conform to them. There are often strong standards of whether you should be married by a certain age, or have kids, or how you should keep your house, or just what the idea of success is for you as a woman – even beauty standards, which you can see very palpably just in terms of how people behave. Many of these norms are self-imposed, which are often the most challenging ones to get rid of.

Growing up in India, I was lucky to never personally feel held back for being a girl – but even so, there were certain deeply ingrained norms in terms of what I knew women ‘should do.’ So all of these things are on a continuum, but as Akerlof and Kranton describe in their wonderful book on identity, there’s often just no getting around them. All of these norms we grow up with create a lot of internal constraints, unless you learn to re-examine or challenge them.

What are some examples of how your research plays out in practice?

For our work on mental bandwidth, a good example is cash transfers. By now, almost all low- and middle-income countries have cash transfer programs, and most have the same structure: small, monthly transfers. But if you look at recent evidence on what actually helps people break out of poverty traps – like the ultra-poor program evidence from Banerjee, Duflo, and their co-authors – you find that lumpier investments are often the most impactful. My recent work in Kenya with Paul Niehaus and Carolina Kansikas also found that most people don't want small, monthly transfers – they’d rather have that same amount in bigger chunks, since it helps them make bigger investments.

The other dynamic with cash transfers, which hasn’t been explored much at all, is timing and seasonality. Basically, the idea is that poor people are not equally poor throughout the year. Research by Joshua Merfeld and Jonathan Morduch shows how much variability there is – it’s not just about having a low level of income, on average, but uncertainty and variability around that low average — both in the amount earned in a given month or year and also the timing of that income.. This gets back to the uncertainty of poverty, the anxiety of the lean season for farmers – it's that juggling which causes a strain on the mental bandwidth. So cash transfers can actually help deal with these irregular income flows by trying to inject them at the right time – during a lean season, for instance, or whenever the ebb may be in a given community. The evidence suggests that this could be quite beneficial, for individual households as well as through broader general equilibrium effects, even without increasing overall program budgets.

In the US too, there’s a company that learned about our paper in Science and started working on how to align individual income and expenditure flows better to ease the mental stress from juggling the two. They partner with companies to modify how salaries get paid: not the same fixed amount each month, but variable amounts tailored to the needs of individual employees that add up to the same annual salary. Something so sophisticated as tailoring payments to individual needs might not be possible in developing countries, but even varying transfers seasonally at a regional level could be an improvement, since people’s incomes vary a lot across the local seasonal cycle, especially in rural areas. One could even make the amount of the transfer contingent on how the harvest is during a given season – similar, in many ways, to the concept of automatic stabilizers in macroeconomics. And most importantly, this could be done without demanding much additional administrative resources, or any financial resources from governments. 

What about practical examples of your work on psycho-social impacts?

My colleague from Warwick and I had the opportunity to work with a foundation that was doing pioneering work to rehabilitate sex workers in India. Not only do these women face social stigma, they often tend to internalize this stigma and feel really bad about themselves. Worse, this can become a public health issue, since they often don't look after their sexual health when they feel bad about themselves, with negative externalities for the wider community. The foundation’s programming focuses on empowering sex workers through collective efforts, so this was an opportunity for us to study in a systematic, rigorous way the impact of efforts to reshape self-image and self-narratives for people enduring a great deal of internalized stigma.

Our study included training to reexamine their self-image – asking them, for instance, to evaluate themselves for how they earn their money as compared to a thief. These relatively modest interventions enabled the women to perceive themselves much more positively and led to improved preventive health and savings behaviors. We saw really sustained effects too, in both domains; in the case of preventive health behavior, the positive effects continued for 21 months after the program. Our analysis of long-term effects used administrative data rather than data we collected ourselves, so we were confident that what we saw in the data was not driven by what study respondents thought we wanted to hear. As an economist, I found the results quite remarkable – but based on personal experience I can appreciate how they arise: when there is a shift in one’s self-narrative, it can really energize us, make us perceive our interactions with others differently and change how we navigate our lives.

What are some of your work’s implications for policy and programming?

I think we are seeing a gradual shift in how we talk about these issues, especially around gender and race, but it needs to happen at a much more programmatic level, both in design and in implementation. It would help to have greater awareness of these mental bandwidth and psycho-social issues that people face in situations of relative adversity – times when people can really lack agency, especially if they are marginalized – and to consciously factor these concerns into program design. Also, at a higher level, we need to be mindful of the power of narratives in motivating action for change. A recent book by the economist Sam Bowles, for instance, The Moral Economy, argues that rather than only using conventional economic policy tools like prices, taxes, and subsidies, combining them with meaningful narratives for why particular policies are being put in place can often make all the difference.

Equally, I think harnessing the power of a positive collective narrative can be really important, especially for people who are severely marginalized. I saw this firsthand with the sex workers we worked with, and I’ve seen it in terms of political voice for women in India – where having women in political office seems to change the comfort level that women have for talking about and reporting gender violence. There’s a sense of agency you feel from having people in office who have some commonality with you, who can understand your predicament, and who will support you ‘against the system,’ so to speak. When there is a woman leader, women go to meetings more, and they speak more about what’s happening to them. There’s also strength in numbers, and it can create a positive feedback loop by reinforcing a different self-narrative.

This collective dimension can be really powerful for agency, and we can harness it with program design – for gender and all sorts of issues faced by other marginalized groups. 

Another important reason to harness the power of the collective is people’s increasing sense of social isolation and lack of any sense of belonging — something that has been documented both by academic research and other journalistic work. Loneliness was declared a global epidemic by the World Health Organization last year. I think a sense of belonging is a very basic human need, and it helps fuel people’s sense of agency. So we need to do more research, have more dialogues, and think a lot more deeply about how a collective sense of agency can be fostered as part of program design. We may, for instance, need to do more programs where there is greater local involvement – to bring the ‘social’ dimension into social protection.

Ultimately, we need to leverage this collective dimension. Rather than making people feel like some program is being shoved down their throats, we need to give people more agency – so they have a real voice and a real say in how these things are designed and implemented in ways that cater to their actual needs. 

Written by Greg Larson. 

This blog is part of a series on Leveraging Personal Agency for International Development, a convening co-organized by The Agency Fund and the SEE Change Initiative at Johns Hopkins University.