Principals, Agents and Bailouts

For anyone paying any attention to any kind of press, there seems to be an unending stream of thoughts analyzing the financial crisis in the US. This has made skipping the finance page and heading straight for the cricket news that much more difficult, and, as such, I find myself reading about the crisis and comparing the issues involved to analogous ones in development.

While explaining the cause of this crisis seems complicated to even the best of minds, it seems a proximate cause for bankruptcies like AIG was excessive investment in instruments like credit default swaps that insured the debt of other institutions like Goldman and then sent AIG down-under once the housing bubble decided to burst.

The evolving consensus seems to be that taxpayers are now in the inextricable situation of having to bailout Wall Street for its own excesses (perhaps with the exception of Lehman), so as to prevent or soften the blow to the ‘real’ economy. Doing otherwise would presumably result in financial contagion leading to reduced investment and job creation, a reduction in lending leading to difficulties in financing mortgages and a whole host of other problems that result from the latter.

On the downside, the very people who caused this crisis are getting bailed out by tax-payers and only have to endure some bad press and corporate restructuring; the proverbial slap on the wrist. While a perhaps radical juxtaposition, these arguments remind me of the debate on the Indian Government's recent loan waiver program.

At some level of specificity, I’m sure it’s unfair to compare the following two situations, but narrowing the scope of our analysis to government (or any third-party) intervention in a principal-agent relationship, I can’t see much of a difference. To wit, the farmer who has used her loans unwisely is in a similar position to the investment banker who has imprudently invested in CDS’s or what have you. In each case, the agent (the farmer or the banker) has done a bad job of satisfying whatever contract that exists with the principal and is therefore unable to make good on the promised return, be it to an MFI or the shareholders at Lehman.

In the case of the farmer, we acknowledge the desperation of the situation but still make the argument that a loan waiver could affect long-term lending and ‘credit discipline’ more broadly. (See V. Mahajan in CMF Policy brief.) As for the banker, we must similarly acknowledge that by bailing him out, we run the risk of a further moral-hazard type problem by essentially insuring him against his own incompetence, but at the same time to not do so would result in even worse consequences.

Two questions occur to me from this exercise. Firstly, what criteria do we use to decide on the appropriateness of government intervention in financial markets? Do we simply base our decisions on historical contingency, or theoretical scenarios which – perhaps by virtue of their scale – have little or no ‘hard’ evidence, free of statistical qualms.

Secondly, if there is good empirical evidence to support non-intervention – perhaps the IRDP experience is an example – but a democracy wills that the humanitarian consequences are even more dire, perhaps what we should be focusing on is learning how to design institutions which are capable of responding to such a situation, together with capacity-building and financial-literacy efforts to prevent such problems in the future?

Microfinance and the Oldest Profession

Revolving-credit associations, informal insurance arrangements, women lacking credit histories and bank accounts... Sudhir Venkatesh's article, "But What Does it Mean for Prostitutes? How the Financial Crisis Affects the Oldest Profession," left my head spinning.

New Microfinance Map: See which MFIs operate in your city or town

At a Sa-Dhan conference yesterday on increasing financial resources flows to the microfinance sector, Sa-Dhan Executive Director Matthew Titus mentioned a new resource that details which MFIs operate in different cities throughout India. Upon review, it’s a pretty useful, interactive map. If one wants to learn about microfinance, or contact an institution for another reason, in a given city or state, this map let’s you know which MFIs you could reach out to.

More on the conference later (the part on private equity was very interesting), but for now, here’s a link to Sa-Dhan’s mapping resource.

Communication Among Stakeholders in Microfinance

Several weeks ago I was attempting to explain what “irony” to one of my colleagues, one of the most notoriously difficult to define words in English language. My lack of success was due mostly to poor examples (those I did remember were evoked by an Alanis Morisette song). If only I had read Carissa Paige’s 2006 CMF working paper “Reputation and Communication in Microfinance” just a few weeks earlier. I would have immediately quoted my colleague the following anecdote from a section of her paper on government-MFI relations:

SBI [State Bank of India] is frustrated because most MFIs in Orissa are not interested in availing their wholesale loans. Indeed, SBI offers MFIs bulk lending at a generous interest rate of 7%, but then requires that its MFI borrowers on-lend at no greater than 11%, a condition to which only one MFI in the State to date has agreed. However, there are four MFIs in the State that are currently choosing to borrow at around 10% from SIDBI, which, ironically, receives it financing from SBI. SBI is therefore aggravated that MFIs are opting to borrow at a higher rate in order circumvent SBI’s interest rate cap.

Paige’s excellent paper focuses on the lack of public relations and communications efforts by MFIs in Orissa, and was written in large part as a response to the Andhra Pradesh Crisis. The best parts of her paper are in my opinion when she describes the various government-run, subsidized microfinance programs and how they might affect adversely affect the effectiveness MFIs trying to collect repayment. She suggests that, at that time, government officials and MFI practitioners often viewed each other as competitors, and that they had little regard for each others work.

I’d like to finish this post with a question to readers. As the Indian MF industry matures, it seems to me, decreasingly likely that we will see another similar crisis to what happened in Andhra Pradesh in 2006. Is this a sentiment that other readers share? Is it only my perception, or have microfinance practitioners, through groups like Sa-Dhan, begun to work more intimately with government?

Random Discontents: III

I appreciate Emmerich and Dan for their spirited defence of RCTs and having spent some time with CMF and J-PAL and the brilliant academics we worked with, my impressionable mind has been reasonably impressed. The C-GAP blog, Rodrik's paper and the responses to the same, in my opinion, have been largely academic, evaluating the cost-benefit of running these experiments. I will take a slightly more lay-man stance here.

On innovation, yes – RCTs are amazing! Imagine using a scientific trial to evaluate social programs - breaking the myth that in social programs, all we could rely on are our gut feel and years of experience and may be, qualitative studies. The studies take time – that’s no issue. Any good study will and should take time. The studies are costly – no problem, as long as there are donors with research funding. RCTs exclude some people from benefits for some time – no problem, no organization can serve everyone at the same time and tricky field issues can be resolved with some clever designing.

Where I have a problem is the way in which RCTs researchers promote it as the ultimate tool. Social experiments, which we call development interventions, when tried in a particular context, lets say Orissa, and succeed are not these days acknowledged as successful enough, if they cannot be proven to be replicable. The romance with scale, replication and cost-effectiveness have brought damning judgements on many development projects and many promising interventions have lost out on funding because of their seemingly non-replicable nature. Academic investigations into such studies seem to escape this scrutiny. And rightly, this is not just with RCTs, this also holds for other kinds of studies which attempt to judge social programs. Why would an MFI, for instance, in Bihar modify its operations because a study in Hyderabad puts out a certain result?

Secondly, researchers need to internalise that most practitioners are a different breed of people, unlike themselves and some academics-turned-policy makers. These people have their eyes and ears on the ground and are also exposed to a large amount of information that comes their way regarding different organisations and their programs in different parts of the country. From what I have seen, RCTs almost seem to assume that their study examines a problem from scratch and in isolation. Thus, when RCTs are promoted using results that the initial rounds of the study throw up, one needs to exercise more restraint. For example, in the evaluation of an intervention where a south-Indian MFI bundled a micro-insurance product with their micro-credit product, the researchers have said there was no evidence that introducing the insurance product in any way adversely affected the composition of the MFI’s clients. In another study with repayment schedules in a state in east India, researchers found that monthly repayment schedules did not seem to increase default rates. Both these studies have been widely publicized and I can imagine, must have amazed (and horrified) many practitioners who attended conferences and seminars where these studies were presented.

However, the forcefulness of these conclusions tend to gloss over the fact that most of these are only preliminary findings; worse still, they look at only one small component of a program over a pre-defined period of time. It is obvious that human relationships change with time – my relationship with my banker will change as soon as I discover there are hidden charges that were not explained to me previously and the same will dramatically improve as soon if I am told that as a reward for my excellent credit history, my subsequent loans will cost me less; my relationship with my insurance company will change if it takes them 3 weeks to process my claim, after having hassled me for over a week about proper documentation in presenting my claim – meaning, my behaviour with an organization/program with which we have a transactional relationship is dynamic. My social networks impact the nature and the extent to which I influence others regarding the particular organization/program.

Also, the artificial separation of the organization from the program simply refuses to convince me. Especially in contexts where programs are made or marred by those implementing it, this separation is quite inexplicable. Unless we standardize implementers all over the world, we cannot study them the way RCTs propose. These studies spend a considerable amount of time and expertise determining sample sizes, emphasizing on the law of large numbers; how then would the same law work when it came to the number of experiments? How many experiments would I need before I can say that I have covered all types of organizations in India and I have an answer for the standard prototype? I cannot even hope to achieve it for a state, let alone a country or a continent.

I firmly believe one must temper down such ‘conclusions’ that these studies yield. It is true governments and practitioners indulge in rhetoric and without some good rhetoric, they wouldn’t survive. However, academics, in my opinion, ought to desist from such strong posturing – even if all this is being done in an attempt to legitimize a particular methodology and out of the conviction that this is “the way”. These results, promoted as methodologically sound to a bunch of practitioners and policy makers, which later, could be revised, corrected or retracted (on studying the same program location for a longer duration or on studying multiple locations and realizing that much of the previous results are actually attributable to non-replicable factors), erode the credibility of the researcher; and in this case, since the methodology is promoted as the infallible hero, it is RCT that is likely to take a beating. A far more modest study would have looked at a bunch of MFIs at the same time and remarked that in a majority of cases, insurance schemes did seem to be doing badly and that it is possible that one can find reasons for the same if one looked into how programs are being managed and implemented by each organization. By looking at anthropological and sociological accounts of the same population over time, some more inferences can be drawn as to why these events happen as they do. Not the most scientifically accurate study, but one that is possibly a better representative of reality than what one RCT will reveal.

Having proven that RCTs can study a snapshot better than any other technique can, we need to hear from its proponents, how they can be made less costly and easier to handle – so that we can have multiple rounds of the same experiment and in multiple locations at the same time. With this, we will also need to know how to integrate the existing knowledge of an area, its people and interventions. Also, if RCTs are like incisions into a program/organisation’s body, we have to see how it can be made as painless and non-disruptive as possible. This will give it some further credence and then, the non-standardization issue can be approached (and solved, I am sure, using some complicated equation). As I see it, converting socio-economic impacts into solve-able mathematical equations is not the final frontier; being able to answer the ‘why’ and ‘how’ and attaching the same to the estimated impact probably is…

Finally, as Dan points out, the debate on how good RCTs are, are probably premature. I got caught up in this since I was an ‘insider’. If I were not, there is little chance I would worry too much about these questions, especially, which study was better than the other? I would just worry in general about any evaluation and its impact on my work in the field…

Randomization and its Discontents (Part 2)

Before we get started, I want to mention that this follow up to Emmerich’s post was composed prior to having seen what he wrote. I suggest reading his post before before mine, so to check our Part 1, click here or just scroll down for a bit).

Because I manage a long term randomized evaluation (RE), the debate about the merits of this type of study is close to my heart, and I enjoy reading eminent economists and development experts go back and forth attacking and defending their value. It is fun to get caught up in intellectual disputes where the adversaries have ground to stand on. (some might say that I am making too much of the disagreement, but it seems to me that Dani Rodrik and Abhijit Banerjee and Esther Duflo have strongly contrasting views).

But when I take a step back from the discussion, it all seems quite premature. It has only been in the last ten years that a large number of randomized evaluations on development programs have been conducted. And among those, only a handful that have statistical power and major policy implications. The impact of randomized evaluations will take years to understand, both in how greatly the results from these studies will affect policy and whether these results can lead to a new paradigm in how we think about development.

I like to use the analogy of international economic development as a large corporation like the old RJR Nabisco (a cigarette and snack company). Randomized evaluation is a new product that the company is putting a fair amount of investment in, but understands that it is a calculated risk. It might turn out to be a grand failure like RJR’s smokeless cigarette (a grand failure) or it may well turn out to be the Animal Cracker (a massive success). I tend to believe that randomized evaluations will turn out to be quite fruitful in developing and promoting effective programs, but either way, the intellectual foundation for conducting these types of evaluations is sound, so it seems a sensible risk.

As I have discussed on this blog, my personal hope for randomized evaluations is that they demonstrate the importance of public funding for public goods, rather than the wasteful attempts to create “sustainable” markets for goods that households perceive as having little value. But I have no idea whether the variety of randomized evaluations that have convinced me to take this position will also convince policy makers and philanthropists (many of these papers are highlighted in this Kremer and Holla paper).

Will Jessica Cohen’s fascinating and oft quoted study on bednet usage affect the way bednet distribution is conducted in the future? Will the J-PAL study on the impact of de-worming drugs on schooling actually lead to huge de-worming projects funded by government or major philanthropic institutions?

In his post, Emmerich highlights some reasons for optimism on this front.

I believe/hope it will, but I am not positive about how open microfinance policy makers are to conclusions that contradict some of there most closely held beliefs on the importance of weekly meetings and joint liability for microfinance. Policy makers may simply not believe in the external validity of the study documenting a certain effect (and they may be right).

Of course the impact of REs is not simply up to chance. How the results are promoted and explained to stakeholders will make a large difference in their influence. In the recently published Banerjee and Duflo article that I link to in the first paragraph, the authors speak of the power of the relationship created during REs between the experimenters and the institution having their program evaluated, and how this relationship can develop into intimate cooperation. From my experience, developing this type of bond between researcher and institution is the exception and not the rule, but perhaps this can change if researchers begin to put more emphasis on this relationship.

They conclude the paper by highlighting the importance of developing research in concert with policy makers so that they are invested in the results. This is a promising tactic, and I think would go a long way towards dealing with some of the criticism targeted at REs.

This blog was originally suppose to be a response to Evelyn Stark’s list of issues that “elicited the most concern with randomized control trials.” And my job was to respond to her last three comments (I apologize to Emmerich for digressing). So here are my responses:

Stark: “Time – results are measured over a relatively long periods (at least 1 year) and therefore the inability to make changes to the product/service in that time”

Randomized evaluations do generally take a great deal of time to produce statistics. It is not just the interventions that take time but also collecting data and cleaning that data. It seems the suggestion from this complaint is that if one uses simple observation and operational data one could come more quickly to conclusions. But from my experience, I actually do not perceive randomized evaluations prohibiting researchers from using observation and operational data. To give one example, conducting an evaluation of SKS’s health insurance program did not stop the researchers from publishing a toolkit or stop SKS from further rolling out the program once they believed it was successful.

Stark: “Relative merit (time, cost & results) to good market research and product piloting seemed unclear, or negative.”

The majority of this post was spent responding to this issue. Put simply, I don’t think proponents or detractors can say with any certainty whether randomized evaluations do have more merit than “market research and product piloting.” It will take years of analyzing the impact of REs to make any conclusions.

Stark: “Limited by the specificity of the research question; unable to probe nuance and context or make results replicable across programs in different contexts.”

I would be interested in an expansion on what Stark’s colleagues meant by this, because I don't see why this is true. Why can’t researchers conducting REs delve into nuance? I believe that we certainly do this in the study that I am currently involved in, and I believe all my colleagues at the Centre for Micro Finance do the same. And how can other types of evaluators “make results replicable across program in different contexts” any better than randomized evaluators can? I hope I am not beginning to sound overly defensive, but these issues sound more like problems with the concept of structured evaluation itself, rather than problems with randomized evaluations specifically.

We are eager to hear others reactions to our opinions and the issues Stark raises, even if it is too early to discuss. J

THE DAY

19^th September

Oberoi hotel: roof –top, Mumbai

Purpose: Round-table conference on “Indian Micro-insurance: What works?”

Full of market players in insurance sector: There were Insurers, MFIs, NGOs, government, researchers, Funders…..you name it…and the rooftop of the Oberoi Hotel had it.

The event was organized by Center for Insurance and Risk Management & Micro-finance Insights with due support from IFMR foundation with intended focus on “Technology, regulation and trends in Micro-insurance sector”.

To have a meaningful discussion a well thought-out plan was charted whereby three different panels were formed, which discussed on:

1. Growing Role of Technology in Micro-insurance: Enabling Efficiency and Outreach

The Panel discussed technology as a driving factor for the micro-insurance sector. High transaction costs, poor databases and time taken to settle claims- are considered to be the most challenging factors that hinder the outreach of the micro-insurance. And so technology is often hailed as one major vehicle that can perk up the micro-insurance sector. But there is poor understanding of the use and applicability of technology by Micro-insurance providers in India. So the discussion was majorly focused on the micro-insurance delivery to the poor and how micro-insurance can be a viable business. The Panel discussed -mobile services as an upcoming technology to handle Indian micro-insurance and to everyone’s surprise, it was told by the industry leaders sitting on the dais that in the coming years we will be able to use mobile for the same. The panel highlighted that before its use for commercial purpose problems like security, viability and connectivity need to be sorted.

2. The Need to Lead: The Role of Policy in Creating an Enabling and Facilitating Environment

Having discussed the operational issues in the first panel, the focus shifted to the major enabling factor- that is government policy and regulations was discussed diligently by the panel. Inadequate regulations, inadequate understanding of client needs and expectations, lack of proper insurance literacy and awareness; and whether regulation is such a good thing for a nascent sector like MI was discussed in detail. The Panel agreed that overhauling the administrative and regulatory system will enable penetration of micro insurance, especially in rural areas. Panel discussed on the role that regulatory policies can play to channel the right ecosystem needed for reaching micro-insurance scale and outreach. During the discussion “health mutual” came up as a good option to reduce the dependence on the insurers and to inculcate the habit of self-insuring to a certain possible level while after a specific level risk carriers and government should come to rescue for the communities.

3. Current Trends and Challenges: India’s Micro-insurance Movers & Shakers

During the session on trends in present market product and process innovations were the major focus. Discussion went on product design and models of delivering risk hedging products. Innovative products were discussed and inputs taken from the audience as well as the panelists on the way ahead and the future vision for the entire MI sector.

The conference had a very new and innovative format, with none of the panelists making any presentations, but conversing openly and without scripts. This made the conference more participatory and lively and not just another discourse session. Everyone took part in the debate and the some of the questions posed by the crowd were answered from among the audience itself. Even the lunch was a novel idea, as each Lunch table had a theme that was pre-decided and lively discussions were seen happening alongside lunch. Such tables were hosted by NABARD, MANN-DESHI BANK, etc.

The success of the conference was obvious when the seats in the auditorium fell short of accommodating all those who had turned up. The audience contained top leaders of various insurance companies who had come just to attend the session. The media strength was thrice what was expected and even till the last session the attendance was the same as the start.

Conference ended with an optimistic note, with everyone gearing up to do their bit to provide safety nets to their society. Without a doubt it was “The Day” with the big forces of the sector coming together and charting out the plan for the future of the Indian micro-insurance sector with enthusiasm and a spirit of cooperation that is so rare among competitors.

New Capacity Development Programme for Indian Researchers

The Centre for Micro Finance (CMF), in collaboration with the Ford Foundation, is launching a 3-year intensive research capacity programme for Indian researchers. The programme targets early-to-mid career candidates with an interest in microfinance and background in research (e.g., PhD, lecturer at a university, experience with field research).

The programme provides a range of support to selected researchers, including:
1) Access to research methodology workshops conducted by CMF and others;
2) Exposure visits to CMF projects;
3) Opportunities to present their ongoing research at various seminars;
4) Opportunities to work on existing CMF projects with leading researchers; and
5) Support in finding project funding.

Applications are due October 31st, and if you are interested in learning more please visit the website for this programme here, or check out the programme description linked here.

We are really excited about this Microfinance Researchers Alliance Program (MRAP) and its potential to equip researchers to effectively evaluate microfinance practices and lead product innovation. If you are interested, please apply!

India's Golden Quadrilateral

Picture Source: National Geographic

My colleague Amy Mowl points me to this National Geographic article on India's Golden Quadrilateral Highway, India's "largest and most ambitious public infrastructure project in [her] history." Excellent reporting and vivid photos...check it out.

Randomisation and its Discontents Part I

I recently discovered CGAP's excellent new blog on micro finance. Bloggers in the micro finance space are always welcome.

One of the blogs first entries touches upon something that is near and dear to us at the Centre for Micro Finance: Randomised Evaluations (REs). In a post on randomised evaluations in micro finance, Evelyn Stark discusses the difficulties in running a randomised evaluation of micro finance activities. Ms. Stark raises six points that she considers to be problematic when using REs to evaluate the impact of micro finance. They are:

1. Difficulty in setting up a randomized trial (client selection, staffing, follow-up);
2. Expense - cited as equal to the cost of the micro finance intervention being tested;
3. Ethical challenge - the control group is “excluded” from service for the period of the test;
4. Time – results are measured over a relatively long periods (at least 1 year) and therefore the inability to make changes to the product/service in that time;
5. Relative merit (time, cost & results) to good market research and product piloting seemed unclear, or negative;
6. Limited by the specificity of the research question; unable to probe nuance and context or make results replicable across programs in different contexts.
   

The debate over REs has been raging in the academic world recently (for a broad overview, see this Economist article), and it is only appropriate that the same concerns are raised with regards to micro finance. My colleague Dan Kopf and I will try to address Ms. Stark's concerns point by point.

Stark: "Difficulty in setting up a randomized trial (client selection, staffing, follow-up)"

What we try to do at the CMF is to integrate our evaluation within the framework of the MFIs normal operations: we use the MFIs loan officers, they are in charge of client selection, we merely provide the technical knowledge and assistance. The very RE that Ms. Stark cites is evaluating a product designed and implemented by a Filipino MFI, not by the external researchers. Granted, there is an additional cost in running surveys, but this cost (financial, manpower, monitoring) is not borne by the MFI but by the researchers, and doesn't detract from the appeal to the MFI of an evaluation of their product.

Stark: "Expense - cited as equal to the cost of the microfinance intervention being tested"

The problem is that Ms. Stark has framed this concern the wrong way. We should be asking what are the costs of the intervention if they have no impact? Or, worse, a negative impact? Although in the short-run (1-2 years) randomised evaluations might be as expensive as the interventions they are trying to evaluate, the long-run (any period after the evaluation is complete) savings could be tremendous. To give several examples, the seminal Worms study by Miguel and Kremer, although probably costly in the short-run, has helped focus the Kenyan government's deworming and U.S. international aid efforts in the long-run. On an even grander scale, Mexico's PROGRESA/Oportunidades (Wikipedia article; World Bank summary; and the Mexican governments exhaustive website in Spanish) has been rolled out countrywide and has motivated copycat programs from Brazil to New York city. Less encouraging for practitioners, but equally as valuable, are those interventions that REs find to be counter-productive or ineffective. The Computer-Assisted Learning Project in Rajasthan was found to be only marginally beneficial, and with the evaluation, Pratham did not roll-out a very costly and largely ineffective intervention en-masse. Although these were not evaluations of micro finance, the general point still holds: the high short run costs can prove to be incredibly beneficial in the long-run.

Stark: "Ethical challenge - the control group is “excluded” from service for the period of the test"

Not all evaluations involve excluding clients the service. The variable repayment schedule project in Kolkata, for example, uses the traditional micro finance contract as its control group, comparing it with a new experimental repayment schedule as a treatment group. The beauty and power of randomised evaluations is that it allows us to compare two products against each other and ask very specific questions of them. The goal in these type of evaluations is experimentation - a goal that gets back to the roots of the micro finance movement: trying to improve on existing ideas and products. If exclusion is involved, this does not have to be as ruthless as Ms. Stark makes it seem. Although MFIs often have ambitious expansion plans, they cannot expand everywhere simultaneously. Researchers can use as a control group a region in which an MFI plans to expand but does not have the resources to as yet. Like the Worms study and Oportunidades/PROGRESA, the roll-out of the service can be staggered across time: The control group can be "excluded" from the program in the first year, and then given the program in the second. With careful coordination and planning between the MFI and evaluator, this can often work in conjunction with the MFI's own expansion plans.

Although Ms. Stark raises some valid concerns with REs, these are issues that those in the field have given serious thought too. In some cases (points 1 and 3) researchers have tried to address these concerns through intervention design. Other concerns (point 2), are possibly misguided: the point of REs is to save money, not impose an additional financial burden. My colleague Dan Kopf will be back later to address points 4-6.

Regardless of Ms. Stark's and CGAP's take on REs, the CGAP blog is one that we at CMF will be following intently!

Caste and India's New Economic Order

His imperial majesty Gary Becker recently posted an interesting blog suggesting that India’s economic liberalization might be the main reason that caste-based discrimination has decreased over the past couple of decades (link: http://www.becker-posner-blog.com/archives/2008/09/competitive_mar.html). Becker himself is the author of economics’ seminal literature on discrimination – in his 1973 paper “The Economics of Discrimination” he demonstrated that employers in competitive markets who discriminate on the basis of race, caste, or other arbitrary factors actually suffer. The intuition behind this paper is remarkably simple – if employers are unwilling to hire minorities who are willing to work for equal or lesser pay than the general population simply because they are minorities, the employer’s costs increase and thus their profits decrease. Discrimination hurts the discriminator because it places them at a competitive disadvantage relative to firms who make hiring decisions solely on the basis of profit maximization.

Becker discusses an August 29 NYTIMES article (link: http://www.nytimes.com/2008/08/30/world/asia/30caste.html?pagewanted=1&_r=1) which examines the views of Chandra Bhan Prasad, a dalit activist (and former maoist naxalite) who has fervently espoused the view that India’s economic liberalization is the main factor that has enabled millions of low-caste Indians to transcend the rigid social boundaries that previously trapped them in degrading and low-income work. Prasad himself is coordinating a massive survey of Dalit households that has revealed, perhaps unsurprisingly, that Dalits are far less likely today to be employed in their traditional caste-defined line of work than before liberalization. Given that India’s economic reforms were also paralleled by the ascendance of low-caste politicians, it is difficult to determine whether this newfound upward social mobility is actually the result of economic reforms rather than increased political representation. Becker compellingly argues that simple economic logic implies that there is no doubt such reforms have indeed played a role.

India’s economic boom, driven primarily by service-sector growth, has resulted in skyrocketing demand for skilled workers. As economic liberalization has forced Indian firms to contend with increased foreign competition, it becomes increasingly costly for such firms to discriminate based on arbitrary factors such as caste – the cutthroat global marketplace has no mercy for those who consider anything other than profits in the way they operate. And in order to maximize profits, firms will necessarily hire the most productive workers willing to work for the lowest pay. Firms that have to bear the cost of discrimination will be incapable of competing with firms that do not.

This is the theoretical mechanism by which economic liberalization reduces caste discrimination, and there is anecdotal evidence of it everywhere. Anyone who has spent time in India has seen firsthand that casteism is far less prevalent in urban than rural areas, where it still tends to be quite pervasive. India’s urbanization, which has facilitated the emergence of truly competitive labor markets, has also provided an escape route for some lower-caste Indians who are willing and able to compete.

To be sure, economic reform is hardly a panacea for casteism in India. In order for liberalization to yield any tangible benefits for the lower castes, members of these castes must at the very least be capable of taking advantage of these emerging opportunities. Sadly and perhaps unsurprisingly, the pervasive lack of education and human capital in Dalit communities has prevented the vast majority from even competing for new jobs in the first place. Prasad himself has recognized this, and repeatedly stresses the importance of education and skill development in empowering his Dalit brethren. It’s no secret that India’s public education system is horribly broken, something that has and will continue to inhibit the prospects of lower castes seeking upward mobility (not to mention economic growth nationwide). In spite of this, it’s quite heartening to see that some real progress is being made.

A "rooted" view

Mihir Shah's article in The Hindu seeks to bring some grassroots logic to economics - a must read!

Biofuels in India

The national government released its first ever national biofuel policy last week, which was perhaps most notable for its mandate that 20% of all diesel demand should be met using plant-based rather than fossil-based diesel by 2017. The policy also stipulated that 10% of all gasoline demand should be met using ethanol, compared with the current mandate of 5%.

The debate over biofuels is particularly interesting in the Indian context, where food security has consistently been one of the government’s major concerns. Finance minister P. Chidambaram memorably called the conversion of food crops into biofuels a “crime against humanity.” In light of the rapid inflation that has driven up food prices nationally, it is easy to understand why the idea of using potential food stocks to create fuel is politically controversial. But when we examine the full spectrum of costs and benefits of biofuels to the Indian economy, it seems that this controversy may actually be misguided.

One of the major factors contributing to recent global inflation has been the dramatic and sustained surge in oil prices. In addition to driving up the cost of production of various goods, it has also dramatically increased the cost of transporting these goods, resulting in higher consumer prices. It therefore seems logical that measures that attempt to reduce dependence on (mostly imported) oil and diesel might also serve as an insurance measure against future oil price volatility and therefore help stabilize price inflation of more basic consumer goods such as food.

The assumption that biofuel production will crowd out food production is also misplaced. Datropha, the plant which is the main source of biodiesel in India, is a hardy plant that can grow on land where traditional food crops would typically not grow, which means that arable land need not be sacrificed for fuel production. In addition, datropha production is estimated to use less than 1% of the water required for traditional agricultural crops, and most other crops used to produce biodiesel have similarly low inputs, suggesting that even large-scale production of biodiesel will not result in a significant diversion of scare water resources from food production.

It is a bit harder to make an unqualified endorsement of ethanol production in India. Since almost all Indian ethanol is produced from sugarcane, which is a water-intensive crop, concerns about food security appear to be a bit more founded in this context. The high variability of sugarcane yields from year to year also makes it difficult to draw generalizations regarding “appropriate” use of cane crops. However, recent scientific innovations suggest that the adoption of new production technologies, such as drip irrigation, might be able to dramatically increase sugarcane output per unit of water. Cellulosic ethanol, which is produced from “scrap” agricultural products such as wood chips, rice husk, straw, is currently not an efficient option but future technological development might make worthwhile.

The imperative for India to reduce its dependence on imported fuels is clear. Lalu Prasad Yadav, our beloved railway minister, is perhaps more familiar than anyone with the real costs of rising fuel pricies: Indian Railways alone consumes one third of India’s imported diesel fuel. In response to the clearly unsustainable price trajectory of imported diesel, Lalu has initiated a jatropha-growing progam that uses the millions of acres of wasteland held by the Railways to grow a renewable source of biodiesel. In my humble opinion, if Lalu is doing it then it’s probably worth paying attention to.

In spite of the recent gains in momentum for biofuels, there are considerable obstacles that must be overcome before they make a significant contribution to Indian energy independence. One of the most glaring such obstacles is the massive subsidy that the government provides for petrol, diesel, and LPG fuels. This massive subsidy costs roughly $18 billion USD, or approximately 2% of total Indian GDP. Although the stated goal of the subsidy is to insulate ordinary consumers from fuel price shocks, there is considerable evidence that in practice it amounts to a regressive subsidy of fuel consumption for the rich (the ISI recently estimated that 75% of the LPG subsidy went to the richer half of urban households. Given that vehicle ownership is densely concentrated among the rich, one expects the situation to be similar for petrol and diesel subsidies). The artificially low price of fuel in India, apart from costing the government an exorbitant amount of money, also dampens financial incentives to invest in biofuel innovation and production, as well public transportation and other public goods. (The government’s fuel subsidy is a massively complex issue that I hope to address with more completeness in future posts). In addition, it will be difficult to incentivize farmers to initially start producing biofuel crops, many of which do not yield returns for 5-10 years. The pervasive skepticism of Indian farmers towards change must be overcome before wide-scale production is possible.

Regardless of these obstacles, I think that it is only natural that investment in biofuels will increase as international fuel prices continue to fluctuate wildly. The Indian government’s mandate seems far more sensible than American biofuel policy, which amounts to a blatant subsidy of corn farmers at the expense of domestic consumers, international sugar-cane farmers, and the environment. One can only hope that the high political stakes and competing interests surrounding biofuels in India will not ultimately yield a policy as senseless as America’s.

Financial Fables on Rotating Savings

It is the end of a long night and five friends are sitting at a bar when the bill arrives. Everyone goes for their wallets but nobody seems to have the correct amount of change. One of the friends, the responsible one who always collects the money, grumbles about the weekly nuisance of spending ten minutes attempting to get the correct of amount of money from each member of the group. The least risk averse friend makes a novel suggestion, “The five of us spend a lot of time together, why don’t we just have one of us pay the bill each time? It will even out in the end.” The others give the suggestion some thought, and one chimes in, “Not a terrible idea, but how do decide who pays on each occasion?” Our daring friend who made the original suggestion, the Mohamed Younus of our clan, responds, “Why don’t we all just put our credit cards in a bag and let the waiter randomly choose? The laws of chance will make sure no one of us will end up taking to big a hit in the long run.” Amused but intrigued, the other 4 friends consider it and after deliberation and some pressure on the most cautious of the companions, the five friends agree to give it a try and call the soon to be bewildered waiter over to their table.

The five friends have just decided to play credit card roulette. Financial innovation at its finest (or at least its most crass and bourgeois). Time and hassle is saved, eventually it evens out and the gambling man gets a little bit of a thrill. On the negative side, I would note that this game could be harmful if one of the friends is afraid to admit that he is broke, and the game leads him to overdraft.

It is the end of a long day and five friends sit around playing cards in a village in Tamil Nadu. As they play, one of the men despondently speaks of how badly he needs to fix his roof, but how difficult it is to get a loan at a reasonable interest rate. For a busy man of his means, the idea of getting a low-interest loan from a state bank is a pipe dream. The high costs of using a moneylender is unappealing, as is spending the time in trainings and meetings one has to participate in to be a member of an MFI. His friends and family tell him that they wished they could lend him money, but the 3000 rupees he needs is more than they can spare.

Pondering this man’s problem, one of the other card players offers a suggestion, “My cousin told me that in his village, when one man needed money to pay for his daughters marriage he started a revolving fund. Basically, a group of 10 people each agree to put 500 rupees a month into a pot, they hold a lottery, and the winner gets to take home the money. But once you win, your name can’t be in the lottery next time. Also, the person who organizes the fund gets to take the first pot.” The others are intrigued, all of them could definitely use a lump sum of money for some purpose, and this might be a good way to force them to save. Skeptical but interested, one of the players asks how you make sure the winner does not just run away with the money. The answer comes that there are some risks, but if you don’t trust the other members, you don’t have to participate, and anyhow the amount of money in the pot is not so high that people would be able to run away and live off it. After further discussing the details, the five men agree to start a fund like this among themselves for 600 rupees a month, with the first 3,000 rupee pot going to man who needs to fix his roof.

The five friends have started a chit fund, as it is called in India or ROSCA as it is known internationally. Financial innovation at is finest.

As pointed out by Steven Klonner in his paper, “Understanding Chit Funds: Price Determination and the Role of Auction Formats in Rotating Savings and Credit Associations,” chit funds are a major part of the financial landscape in the developing world, but compared to microlending, they receive very little interest from academics or the media. To give an idea of the size of the market, Klonner highlights that in Tamil Nadu “the turnover in formal Roscas has been estimated at 100 billion Rupees, about 2.5 billion US dollars, in 2001.” Klonner also mentions that in 1980 in Central Africa, it was estimated that about 20% of all household savings were in informal ROSCAs. In the 1950s, famed anthropologist Clifford Geertz even wrote a paper on their importance in Indonesia.

You may have noticed that in the quotes above, I mentioned both informal and formal chit funds. This is because after originating informally (informal chit funds are still ubiquitous), chit funds have also become a large regulated business with chit companies taking a cut of each month’s pot, but also providing the service of collecting payment.

Another evolution that has occurred over the years will upset all of you clever arbitragers who read the Indian Development Blog (perhaps some of the out of work Lehman Brothers and Bear Sterns dealmakers now have time to check in) and must be drooling over the thought of starting chit funds in villages across India and living off of the interest. These days the majority of formal chit funds, and some informal ones, do not use a lottery to decide the month winner, but an auction instead. The person willing to pay the most for the month’s pot wins. Lets say the winner of a 10 person, 1,000 rupee a month chit fund offer bid 1,800 rupees to take the first months pot of 10,000 rupees. In this case she/he ends up taking home 8,200 rupees, but has to give 200 rupees back to each of the 9 other members. It is a bit complicated, but in the end those who take the earlier pots end up paying interest to those who wait for the last ones. In Klonner’s study the median “recipient of the first chit lends at an interest rate of 1.6% while the recipient of the last chit earns an interest of 0.7% per month.” See the paper for more details.

I don’t really know exactly what to make of chit funds and their role as a development tool, but since “Rosca members are mainly poor individuals who have little access to formal savings and credit markets because of high transaction costs and incomplete markets” it is clear to me that any financial product that is this successful and prevalent across countries must have lessons to teach us in what kind of products poor consumers want. Perhaps this tells us something about the importance of commitment saving or maybe it could be the basis of a tool that could be taught to SHGs to increase internal lending. Less likely, maybe it means that, like our credit card roulette playing friends, people simply enjoy a little risk and want something a bit more fun and exciting than a boring old joint liability loan.

For more info read this.

The Importance of Financial Literacy Education: What we know and what we don't

Harvard Professor Shawn Cole and CMF Research Associate Nilesh Fernando recently wrote a very interesting paper that reviews how levels of financial literacy education effect financial behavior. The paper uses several assessments, both from the developed and developing world, to discuss the consequences of low levels of financial literacy, financial literacy's effect on savings and investments, and current efforts to improve financial literacy.

The review was written for the Asian Development Bank's quarterly microfinance newsletter, and can be found here, and below I review some key points discussed by Professor Cole and Mr. Fernando.

What we know
According to Cole and Fernando, a significant body of evidence demonstrates a strong link between financial literacy and household well-being. Accordingly, survey findings consistently show that households with low levels of financial literacy:
1) Borrow at high interest rates
2) Do not tend to plan for retirement
3) Acquire fewer assets

Partly because of this body of evidence, policymakers in wealthy and poorer nations alike advocate increased expenditures on financial literacy education. But, will this increased spending have its intended effect of increasing household savings and participation in financial markets, helping to reach the goal of reducing poverty?

What we don't know
While many organizations provide evidence that suggests financial literacy education is effective, Cole and Fernando could not identify a completed study in emerging markets that tests the value of financial literacy education. One ongoing study in Indonesia evaluates a literacy program that teaches unbanked individuals how to open an account. Preliminary results suggest that the program did not increase the use of banking services among households.

What does the lack of evidence on financial literacy education's impact on financial behavior mean for policymakers and for interested parties? We don't know whether financial literacy education dramatically impacts savings and investment decisions. Other factors (e.g., household income, education levels) may more directly effect individual behavior, or it may not. We are left with little evidence to guide policy or investment behavior that seeks to improve financial behavior of poorer households.

That said, the authors applaud the firms, governments, and organizations (e.g., SEWA Bank in India) that provide financial literacy education. Teaching low-income clients about financial products, and how to use them, will help clients understand how to use financial products (which is a definite plus). But, having a clearer idea of the impact of different types of educational initiative will help us make informed decisions about how to invest limited resources to improve the lives of the poor.

Once again, here is the link. I look forward to any comments or thoughts.

Development's Double Standards

My former colleague writes about double standards in development. This is true worldwide, not only in India.

The Links

I will save you the energy of having to read my opinions and this time only direct you to the places they come from:

- The Times of India's Swaminathan S Anklesaria Aiyar argues for the poverty reducing effects of high food prices for India's rural poor.

- Esther Duflo puts on a "virtuoso" performance in a Q & A session for the International Herald Tribune. Perhaps her most interesting answer involves her suggesting that "the extent to which [randomized evaluation] results generalize is actually surprisingly high."

- Ben Olken's paper "Monitoring Corruption: Evidence from a Field Experiment in Indonesia" is more fodder for people like me that believe the power of promoting grass roots participation in the provision of public goods is wildly overestimated.

- A nice brief explanation on why you don't want a "flat" interest rate (the form in which many MFIs present their loans) from Joydeep Ghosh.

Index Insurance

When it comes to insurance, we have the notion of 'direct' insurance. You insure something, and that insured object gets damaged somehow. Then you go to the insurer, ask for the compensation. The insurer comes, inspects the damage and repays some reasonable (hopefully) amount to you... Great... No hassles? Probably not if you are thinking of insuring your Hyundai car!

Now imagine that you are a farmer and you insure your crop... You have half an acre of land on the west side of the village, some acres in the north, some in south and one more acre right in the middle of the village... Oops! So much hassle for the insurer to take a trip to your not so picturesque village and arrive at the right amount! The insurance guy is surely going to charge you heavily for this picnic through the premium.

Solution: Opt for an index based insurance. There are different types of indices based on which you get your compensation (Indirectly). Currently there are two kinds of schemes prevalent in India. First is the 'area' yield index scheme. This method requires to calculate an approximate yield for a specified area after harvesting for a particular season. Then the payouts will be based on the 'level' of the index which corresponds to the 'average' yield for a particular crop in the area. What is good in it? Now your insurer does not need to take trip of your farmland. All s/he needs to do is to have a small 'sample' farm in the area growing the particular crop. Harvesting of this farm will 'represent' your production level (hopefully). Wait... Is it mathematical? Unfortunately 'yes'. Is it transparent? Can be! Does it require hassles? Surely not much as compared to the earlier method...

Second type is weather index insurance. As you know - your produce is pretty much dependent on the weather - be it rain, frost or temperature. The insurance guy can set some levels of these weather parameters and pay you accordingly. Suppose your region receives only 5 mm of rainfall in monsoon. That clearly means that if you are growing paddy crop - you are probably on the verge of bankruptcy. What to do? Buy a weather index insurance. It is like betting on the rain (or any other parameter). Wait... Isn't it much better than the earlier method? You don't need to wait for the harvesting. You don't need to even harvest at all- all you need is a rain gauge. But what if the rain gauge data is not available (It requires around 20 to 30 years of past data to build a good pricing model for weather index insurance)? And imagine if the rain gauge specified by the insurer is available at Ahmedabad whereas your field is near Mumbai? Simple! Get one more rain gauge at Mumbai! Then find the correlation (what is meant by that?) with the Ahmedabad data and boast that you have the arrived at a good model. Hmm… Does that mean that the insurer has to go on setting up those so called ‘automated’ rain gauge machines in every village of the country?

Well certainly not! As it is clear, weather index is a ‘point based’ method – where the compensation is based on the weather parameter measured at a point. Plus the correlation method may not work properly given the erratic nature of weather in our country. Instead, if I can somehow get the right data about the production in your village directly, I may not pay heed to such weather parameter at all. Normalized Difference Vegetative Index (NDVI) is such a ‘parameter’ which is based on the greenness of the fields in a given region. Against the point based weather index, it is a spatial index.All it requires is…………… A satellite image of the region! Are you crazy? We were talking about expenses on the simple rain gauge. Do you want a poor insurer to buy a satellite for this? Not really! One can buy such images for the whole country. The only problem here is how good is the image resolution? If the resolution is not very good, one can even build even a combined index on the weather and NDVI figures. The most challenging thing in achieving all this is ascertaining the levels for the index and associated payouts (which is called pricing or modeling).

Is it available in India? Not really (only one small pilot project exception)! :)

Is it mathematical? Oh yes :((((((((((