January 15, 2014 by Alex Counts
Just as increasing use of a “gender lens” has transformed thinking about and the practice of international development in recent decades, so too can behavioral economics in the near future. In some cases, this discipline explains and reaffirms current practice. In other cases, the study of behavioral economics provides an alternative explanation of why some things work and others don’t. In still other cases, it suggests that current thinking and so-called “best practices” are wrong and counter-productive. Now and then, it prompts us to consider readopting a practice that has fallen out of favor. (For an introduction to the thesis put forward in the book Scarcity: Why Having So Little Means So Much, visit Part One of this blog.)
Now, I will comment on the specific insights and implications I see for microfinance and international development when looked at through a behavioral economics lens.
Group-based Lending Methodologies. The analysis of scarcity helps explain the power of, and the tendency to use, group mechanisms in microfinance. The frequent contact clients have with staff and other clients has the potential to “get inside the tunnel” poor people often experience, thereby mitigating some of the resulting destructive actions and behaviors. (Perhaps group methodologies should be re-thought through the behavioral lens to magnify this positive influence by refining how those contact points are used.)
It may be that the need to pay off one’s loan in public on an agreed-upon schedule incentivizes people to focus more on the medium-term considerations and strategies than would normally be the case for people who are “tunneling.” For instance, family members of loan clients may be willing to work longer hours or more productively, and less willing to give in to temptations, such as smoking or consuming alcohol, in order to meet their loan payment obligations. (The reduction in “sin expenses” in families taking a microloan has been confirmed in at least one randomized controlled trial.) Scarcity also gives a new explanation for the importance of frequent loan payments, as the authors say that having “frequent interim deadlines” is a proven way of penetrating the tunnel that a single deadline months off in the future usually cannot. It also raises questions about the wisdom of building in grace periods into loan contracts with low-income people.
The Role of Technology. This analysis has some interesting implications for the use of technology in microfinance. On the one hand, to the extent non-intrusive technology obviates the need in accessing financial products for human contact – contact that has the potential to pierce the tunnel and focus clients’ attention on so-called “important but not urgent” matters – it could be a negative influence. On the other hand, SMS reminders to do things that pay long-terms dividends but would otherwise be neglected while tunneling – like saving small amounts of money or taking ones medication on time – have shown promise. I find this a strong endorsement of Grameen Foundation’s rejection of simplistic technology solutions to meeting the diverse financial needs of the world’s poor, while strongly supporting technology solutions that involve human networks. (See below for some examples of that thinking in action.) The importance of human networks in making technology useful to the poor is something I delved into in my post on the Financial Inclusion 2020 conference held last October on the Center for Financial Inclusion blog. That conference, incidentally, is where I first heard of the book Scarcity, as one of its co-authors gave a very compelling presentation there.
The importance of savings. The wisdom of the recent emphasis of microfinance professionals on promoting savings is confirmed by behavioral economics. This is because accumulating liquid assets tends to reduce or eliminate the harmful impact of tunneling by creating what the authors term “slack.” However, it also suggests to me that compulsory savings, which recently has been deemphasized and dropped by many organizations, may deserve another look. (Or at least opt-out savings plans, where clients are required to save unless they choose not to participate.) This is because people exhibiting tunneling behaviors will tend to neglect savings unless it is automatic, which partially explains the phenomenon of dormant savings accounts littering the developing world.
Flexible Loans and Consumption Loans. Easy access to very small, short-term loans from microfinance institutions (MFIs) is another potential source of slack – they mitigate against the tendency of poor people to decapitalize their businesses or take very high interest informal loans to meet immediate needs, or deal with “shocks.” At one point the authors reflect on some of their own research in India and state, “[W]e were amazed by the high demand for loans that averaged less than $10. The product does not help build wealth [and] on the surface it does not look like the kind of sum that can transform a life. Yet it might do just that.” The authors correctly note that loans can be a tool for reinforcing the negative impacts of tunneling, or for alleviating them, depending on the context and especially how they relate to magnifying or reducing the negative impacts of scarcity and its taxation on bandwidth. This insight forces us to think differently about the stale and perhaps wrongheaded critiques of microloans centered around how some of the funds officially sanctioned for starting or expanding small businesses are in fact used for other purposes, usually defined as “consumption.” (In fact, some studies that have found that up to 40% of the funds clients receive in the form of microloans are not used for investment in business – statistics that are sometimes used to undermine the case for microcredit.) This critique loses some of its power when viewed through the behavioral economics lens. If microloan proceeds not directly invested in a business are used to provide helpful “slack,” reduce scarcity (and the negative behaviors it leads to), and/or pay off high-interest debts accumulated in the past, they may be as high-impact as the parts of the loans invested in income-generating ventures.
Microinsurance. The potentially critical role of well-designed micro-insurance products is reinforced, as they help people deal with shocks and as such, provide slack, and thereby reduce the need for periodic tunneling. However, getting people operating under scarcity to pay attention to and then agree to pay for long-term risk management is clearly more of a challenge than most have previously understood. Fortunately, behavioral economics gives us some tools to confront that challenge. It suggests that building insurance into loan contracts – something that is fairly widely practiced even though it is increasingly discouraged – deserves another look.
Over-indebtedness. Behavioral economics analysis of scarcity explains and warns against the insidiousness of over-indebtedness, which is a logical consequence of scarcity (particularly for the poorest), especially when access to loans is relatively easy. It also explains why many MFI clients continue to borrow from moneylenders. This in turn reinforces the importance of credit bureaus, self-restraint by MFIs in pushing loans beyond what can be usefully deployed by clients, and client protection principles. The magnitude of people’s tendency to focus on alleviating short-term scarcity (which is often through taking a loan regardless of the long-term consequences) suggests that these long-overdue measures may only prove to be partial solutions, especially for those suffering from the most acute scarcity. But clearly they move in the right direction.
The overall importance of microloans. The capacity of microloans to give poor people needed slack is directly related to their reliability. This affirms David Roodman’s point about the most important achievement of the microfinance movement being the delivery of loans on a mass scale in ways that are predictable. Reliability may even be more important than the cost of the loans, since even relatively high interest rates are likely to be much lower than the long-term costs of mortgaging one’s future by selling productive assets, neglecting children’s schooling, or not buying medication recommended by one’s doctor – behaviors that are commonly exhibited by all people (rich or poor) who are tunneling.
But perhaps even more important, reliable and appropriately used financial services decrease the need for tunneling through providing slack. In other words, they make it easier to juggle (akin to suddenly having fewer balls in the air) from a cognitive perspective. They can also dampen other negative impacts of tunneling, such as the inability to control temptations or temper, to be pro-active in taking care of one’s health, or to remember to vote – behaviors that are consistent with well-being in the long term. But as we saw above, financial services that are provided irresponsibly, or even with insufficient attention to behavioral economics, can also aggravate and worsen these same conditions.
The Concept of the Rational Borrower. Behavioral economics tends to reinforce an important implication of an idea underpinning the approaches many microfinance practitioners take, the one that assumes that in most instances, “borrowers know best.” The implication of this assumption is that extensive training of clients is usually not cost-effective, and possibly not useful at all. But behavioral economics arrives at this conclusion for a different reason. The advocates of “borrower knows best” often claim that the poor are best suited to leverage their existing “survival skills” and those can be pathways out of poverty if complemented by responsive financial services. Behavioral economics’ explanation for the ineffectiveness of most efforts to train microfinance clients is that their experience of scarcity gives them limited bandwidth to absorb and apply new information. On the other hand, behavioral economics challenges another aspect of the “borrower knows best” school – the part that argues that the poor can properly weigh short and long-term considerations, since people who are tunneling often fail to do this.
Optimizing Staff Performance. In a section dealing with how to apply behavioral economics to organizations, the authors describe and explain why workers, in many contexts, get more done in 40 hours a week than in 60. Basically, “working overtime” can pay short-term productivity dividends, but those dividends are usually short-lived. The cumulative impact on workers’ mental bandwidth is one way of explaining why this is so. For microfinance managers, the potential negative impacts of pushing field staff to consistently reach aggressive goals on long-term productivity and even ethical behavior are made manifest. And like many of the observed impacts in behavioral economics, it is not their existence but their magnitude which is most striking and unexpected. In addition, taking into account staff members’ “cognitive load” and adjusting their responsibilities accordingly makes good sense. For example, giving staff assigned to troubled, remote or otherwise difficult branches additional annual leave or smaller caseloads would be a wise human resource policy.
Enhancing Social Performance. In an important tangent that relates to the larger question of how scarcity and abundance relate to each other, the authors attempt to partially explain the financial crisis of 2008. They hypothesize that in the lead-up to the crisis, financial institutions were exhibiting the behavior of “tunneling” with an exaggerated focus on short-term profits. They argue that the growing importance of “chief risk officers” in financial institutions is a way to “penetrate the tunnel” and as such, better balance short and long-term considerations. In the microfinance context, some loan officers and their managers have shown a tendency to tunnel by focusing inordinately on short-term financial profits, to the detriment of long-term client relationships, consumer protection, honest and ethical behavior by staff, and organizational social performance (also known as client advancement). Having empowered “social performance” and “client protection” officers is one way to create a healthier balance, as is generating (and, importantly, making visible) real-time data on social performance through use of tools like the Progress out of Poverty Index – in the parlance of the book, they also have the potential to “get inside the tunnel” that the organization and its staff tend to inhabit.
Let me briefly look at a few issues which have broader application to poverty alleviation efforts beyond microfinance.
Client Training. While behavioral economics does not deny that the poor – like anyone engaged in business or even trying to manage their health – can benefit from learning new things, it concludes that the bandwidth that true learning requires is not something that the over-stressed poor can often afford to give. This certainly explains the phenomena of absenteeism and lack of impact on practice and even knowledge in many training programs. Said another way, training programs, especially traditional classroom approaches, usually fail to “penetrate the tunnel” and thus have minimal impact. Interestingly, research cited in the book shows that financial education training based on simplified “rules of thumb” does have an impact on applying knowledge and increasing micro business performance, whereas more complete classroom approaches do not. Behavioral economics tells us that this is because simplified approaches, while perhaps not ideal from a purely educational perspective, take up less mental bandwidth to absorb and, as a result, are more likely to penetrate the tunnel and lead to practical learning and behavior change.
The study of scarcity also challenges some premises behind many efforts to “educate” the poor. Lack of knowledge, the authors argue, is much less an issue than lack of bandwidth to absorb and apply knowledge. Mandatory training programs tend to take up bandwidth (cognitive capacity) in a variety of ways, and may be counter-productive. Furthermore, training programs are usually too linear and not “fault tolerant” enough, in that they make it hard to keep pace if one’s life, complicated by managing scarcity, causes you to miss a class or two (a behavior that is usually explained as a lack of motivation on the part of the poor). Woven throughout the book are the ideas that blaming the poor for their personal failings, lack of willpower and lack of motivation are misplaced and the wrong ideas around which to build anti-poverty programs. The authors note that poverty planners usually wrongly assume that unemployed people are not busy, when in fact they are often busier (especially mentally) than the employed non-poor.
Thinking about the Impact of Anti-Poverty Programs. As noted above, the authors were astounded by the demand for – and the potential impact of – very small loans. (They saw them as potentially positive to the extent they are effectively designed to dampen the negative impacts of tunneling by creating “slack” on demand.) One imagines that researchers looking at people using loans in this way often fail to appreciate the benefits involved with creating slack. The authors further hypothesize that the impacts of many anti-poverty programs are underestimated because they do not capture how those programs lower the “bandwidth tax” and the negative impacts of tunneling they usually lead to.
One obvious example is providing subsidized child care. Rather than having to juggle multiple options for having your children attended to while you work – your grandmother when possible, a friend, less than reliable babysitters, etc. – you have something consistently available. As a result, among other benefits, you relax. In others words, beyond saving the parent money and perhaps ensuring better supervision for the children, subsidized child care gives the poor “back all the mental bandwidth that [they] currently use to fret, worry, and juggle these [ad hoc] child care arrangements.” People stressed by scarcity typically do not sleep as well as those who are not, which further diminishes performance. Helping people manage scarcity and reduce the bandwidth tax presumably helps people get more and better sleep, which further enables their performance in other areas of their lives.
The authors persuasively argue that researchers often fail to measure these benefits. They conclude by stating, “Some of our pessimism about existing [anti-poverty] programs might come from a failure to appreciate and therefore measure such [positive] impact” on working memory, impulse control and self-control. This is both a call to action to researchers but also to program designers, who should think more deeply about how various program elements tax or free up mental bandwidth.
Conditional Cash Transfers. During Sendhil Mullainathan’s presentation at the Financial Inclusion 2020 conference, someone asked a question about the wisdom of conditional cash transfers – fast becoming a fad in international development – through the behavioral economics lens. Perhaps like the questioner, I thought it would get a positive endorsement, as it is a relatively simple and transparent way for people to access cash, which provides slack or at least alleviates immediate scarcity (thus lowering the “bandwidth tax”). Interestingly, Mullainathan seemed to be mixed to negative on CCTs, noting that all the mental bandwidth it takes to comply with the conditions built into these transfers may reinforce more than alleviate the phenomenon of tunneling and its tendency to cause people to neglect actions with long-term benefits.
Grameen Foundation and its partners have some experience applying the insights of behavioral economics. Sometimes those applications have been unconscious, while at other times they have been intentional and included advisers from organizations like ideas42 that was co-founded by one of the authors of Scarcity.
Grameen II. The Grameen Generalized System (GGS), better known as Grameen II, was launched by Grameen Bank in the early 2000s after several years of piloting. When it was announced, it was viewed skeptically by many microfinance experts. Most criticized were the provisions that made it easy for borrowers to reschedule their loans. Professor Yunus termed this element essential to making microcredit “tension-free.” In fact, in his long article about Grameen II, Professor Yunus addressed this directly: “There are many exciting features in GGS, but I think removing tension from microcredit … [is one of the] most important features of them all. Tension-free microcredit is a great gift of GGS. Now both sides in the micro-credit system, the lender and the borrowers, can enjoy microcredit, rather than having occasional nightmares created by one for the other.”
Through the behavioral economics lens, attempting to take the stress out of taking and repaying a microloan looks farsighted, even brilliant. It explicitly aims to free up much of a client’s mental bandwidth so it can be applied to long-term investments of their time and money. Grameen Foundation and Grameen Trust have jointly published a manual, the Grameen Guidelines, about how to implement Grameen II in any country context. It was co-authored by our senior technical expert Muhammad Nurul Alam who has helped advise many MFIs on how to apply these guidelines.
MOTECH. In our Mobile Technology for Community Health program that began in Ghana, we used voice and text messages about good practices during pregnancy to reinforce and amplify what the overstretched nurses of the Ghana Health Service could communicate. While the research on the impact of these nudges to alter behavior (presumably by “getting inside the tunnel” even briefly) is not yet complete, the usage patterns suggest that we are on to something important – and something that the arguments in Scarcity would predict. Another aspect of this program was to allow nurses to digitize information to streamline reporting and make it more efficient and accessible. One hopes that this frees up mental bandwidth and time scarcity of those nurses, thereby allowing them to provide more and better quality services.
Community Knowledge Worker. By the mid-2000s, Grameen Foundation saw its “village phone” programs in Uganda and elsewhere tapering off because shared access models were less viable as personal mobile phones proliferated. We partnered with Google and others to develop content related to health, agriculture and commerce that could be delivered through SMS to cheap phones to people throughout Uganda. We did this because there was so much valuable information in these domains that would benefit the poor if they knew about it and acted on it. When we launched these services, it turned out that we got enough of people’s attention – in other words, got inside their tunnels – to get millions of them to download information. We even measured that knowledge increased. But we did not penetrate their tunnels enough to apply the knowledge by changing behavior in the ways we had hoped. So in the end, we did not get inside people’s tunnels enough to really matter. With this insight, we created a network of more than 1,000 farmers throughout the country who became the champions of the agriculture tips, and called them “community knowledge workers” (CKWs). Having someone nearby who knows you know that you have information that can help you make better decisions that will help you six months later (at harvest time) turns out to be decisive, as a International Food Policy Research Institute study measured significant increases in adoption of good farming practices as well as better prices for crops sold (due to using information about market prices) in communities served by CKWs compared to those that were not. It turns out the human networks can turn an otherwise ineffective technology solution into one that has high impact, as the former can penetrate people’s tunnels and the latter often cannot.
The Impact of Loans in the Philippines. Clearly, all poor households are not alike. While there is evidence that for households with regular income streams (for example, from petty trade) access to a reliable source of liquidity provides security, we have also found that for poor farmers – who may have only one or two income generating “events” a year – credit can be a source of anxiety and can reduce subjective measures of well-being. For example, research conducted by Grameen Foundation in the Philippines found that smallholder farmers with debts were worried about how to pay those off – and reported lower well-being than those farmers who did not take out loans. This highlights the need for risk mitigating tools for farmers such as insurance and savings that gives them financial peace of mind, which that allows them to make better decisions in other parts of their lives. In other words, our research confirmed that for Filipino farmers, the terms of some loans amounted to a bandwidth tax that one can presume negatively impacted them in multiple ways.
Revamping Savings Offerings in the Philippines. Grameen Foundation, working with ideas42 and its enormously talented and energetic Vice-President Alex Fiorillo, recently used behavioral design and a randomized controlled trial to test the effect of applied behavioral economics to the design of savings accounts at CARD Bank in the Philippines as part of a larger project supported by the Bill and Melinda Gates Foundation. By way of background, after a couple of years working with CARD Bank to develop a new savings program, we were stymied by the fact that many clients weren’t saving as much as they said they wanted to. Nor were they building meaningful savings balances. Basically, an alarming number of clients were letting newly opened accounts go dormant, despite their expressed desire to save. It felt like all the effort and money that went into providing the new affordable, convenient and secure savings products and education on good savings habits was having limited impact. We used ideas42’s behavioral diagnosis and design methodology to pinpoint psychological and behavioral bottlenecks to account usage among both new account holders and individuals with existing accounts.
The interventions proposed by ideas42 were astoundingly simple and inexpensive to implement. The savings account opening form was made simpler – easier to complete and with a requirement to select how the first deposit will be made. In other words, the cognitive demands of account opening were systematically reduced. Clients were also encouraged to complete a savings plan with the help of CARD Bank’s staff. The plan, an effort to “penetrate their tunnel,” focused on clients stating in the presence of another human being not only their savings goals, but also committing in writing how much and how often they would save in order to meet that goal. This small but meaningful act of writing down a commitment turned out to go a long way towards ensuring follow through. To support this process, clients were given a calendar to write down how much they saved every day, allowing them to see tangible progress towards their goals. We found that clients in the treatment group were 60% more likely to use new savings accounts and made deposits that were 11% larger than those in the control group. These clients also maintained larger balances. Likewise, clients in the treatment group with existing accounts made larger deposits and withdrawals and significantly increased engagement with their savings accounts compared to those in the control group.
Fonkoze’s “Graduation” Program. Grameen Foundation partner Fonkoze was one of the first organizations in the world to adapt BRAC’s “ultra-poor” program, with influences from Grameen Bank’s “struggling borrowers” initiative and other approaches. It turns out that Chemen Lavi Mio (“Pathway to a Better Life”) drew on insights from behavioral economics even without having studied it. For example, teachings about behavior changes (“life skills”) are radically simplified. “In CLM, we use the concept of messaging where we identify what we think are the 10 most important messages for the clients and every week ONE of these ten is discussed in all interactions with the CLM members,” says longtime Fonkoze executive Anne Hastings. “They are things like ‘only drink the water you’ve filtered with the filter we gave you’ or ‘wash your hands before eating’ – very simple messages. These don’t overburden them as they are just struggling to survive.”
Years later, after learning about the scarcity thesis, Hastings reflected further: “The whole concept of the way CLM lowers clients’ bandwidth tax by removing some stresses from the clients’ lives, like the cash payments in the first few months so they don’t consume their new assets. Or making sure their house has a decent roof so the rain doesn’t pour in during the night, eliminating a chronic worry and also reducing sleeplessness.”
Scarcity challenges us to think differently about what causes poverty, what keeps many poor people in the deplorable conditions they live in, and why many living in poverty do things that not only keep them poor but also make them poorer in the future. It also prods us to get outside our sectoral silos. In one of the more memorable passages of the book, the authors argue that behavioral economics demands “…a radical reconceptualization of poverty policy [by forcing] us to recognize the many ways in which different behaviors are linked… [R]ather than looking at education, health, finance, and child care as separate problems, we must recognize that they all form part of a person’s bandwidth capacity.” They conclude by saying this about anti-poverty programs that liberate bandwidth: “[they] boost IQ, firm up self-control, enhance clarity of thinking, and even improve sleep. Far fetched? The data suggest not.”
What remains to be seen is how program designers, practitioners, front-line service delivery staff, and policy-makers take these insights into account in the years ahead. Grameen Foundation hopes to be one of many organizations that rethink what they do after looking at our pro-poor efforts through a behavioral economics lens.