Professor Pranab Bardhan recently visited LSE’s Economic History Department to speak about research gaps in the interface between history and development. While he was on campus, Sonali Campion from LSE South Asia interviewed him about the value of interdisciplinary approaches to social science questions, and the insights that his research sheds on key debates in India today.
You’ve been a real champion for interdisciplinary approaches. Could you talk about your frustrations with “pure economics” and what you think other disciplines like social anthropology and history bring?
What attracted me about economics was that it is a social science that gives you the habit of precision of thought. But then precision to what purpose? That is an important issue. Quite often I find – and this is increasingly so – that while economics is very useful in giving you the habit of rigorous thought, it often doesn’t challenge the assumptions which economists automatically make. Also, in order to think about which questions are important, and which are not that important, some knowledge of the other disciplines is very necessary. Some of the great questions (in my judgement) are raised elsewhere, while economics gives you better tools and concepts to analyse them.
I edited two volumes on conversations between economists and anthropologists. This came to my mind when I started doing village surveys. I very soon found out that the way statisticians go about it, and the way anthropologists go to a village to study it was often methodologically quite different. Sometimes they come up with different results, then they question each other’s methods. So I put together some economic statisticians on one side and anthropologists on the other side as I wanted to get a conversation going. It’s a difficult thing to do because they tend to talk along parallel lines and they don’t meet.
That made me particularly aware of what the methodological differences are, and what one can gain from the other. Suppose economists collect data and then they analyse those data. But the data are about certain outcomes, because you don’t collect the data when things are happening. You don’t see the process. Something happened, and then you go and ask questions about outcome variables. Anthropologists, on the other hand, give you a very rich, insightful story about the process.
What you really need is both and that’s partly what I’ve tried to do. I’ll give you an example. Anthropologists quite often will just say something about their village, and because they go so deep they can’t do too many villages. Economists work with large statistical samples. They claim that way their statements can be representative of whole regions. So a statistician colleague and I were looking to understand a landlord-sharecropper relationship. We found economists don’t do a very good job when it comes to processes, relations and dynamics. So we wanted to capture those. Economists ask, “in this village, what percentage of land is sharecropped?” – that information is important, but you don’t get an idea of the actual nuances in the relationship. How the landlord tries to control the sharecropper, what is the sharecropper’s reaction etc. You get those stories from anthropologists but not from economists.
At the same time, we didn’t want to be confined to one village. Our study looked at the whole of West Bengal, which now has a population of more than 90 million people. You want to be able to generalise about the landlord-sharecropper relation for that large region. So we designed a sample survey and then drew villages from a random sample. Then we went to the villages, we asked sharecroppers a relational question. Not “what is your relation with the landlord?” because that villager, who is one of a village of perhaps 1000 people, may not be representative. We asked “what do you think is the typical relation in this village?” Of course he may not know, but in each village we asked six villagers the same question about the typical relationship. We found there was by and large harmony in these answers, because they were thinking not about themselves but what is typical. If there were discrepancies we went back and crosschecked as to why he or she had given us a different story. Then we had the typical relations in a village out of a random sample, so we could say something about relations which economists otherwise don’t capture for a whole area. That was an example of compromise between the two methods, a meeting ground where both disciplines can give us insights.
You start one of your papers with “all around the world in matters of governance, decentralisation is the rage”. Based on your research, in what ways do you think the devolution of power in India can be designed to counter growing inequality?
Policies to reduce inequality quite often are taken from the centre. Even if it is from the regional capitals rather than Delhi, many of the states are larger than most countries in the world so decisions taken are highly centralised. For example, land is a basic source of inequality in India. So you try to redistribute it by taking a measure – from Patna, or Lucknow, or Chennai – to reform land. But then if you don’t take care of how it’s implemented at the village level, quite often what you say in the capital cities will be completely negated at local level. Unless something is done that comes from the ground up rather than top down, even with all the best intensions in the capitals, nothing really happens.
Now having said that there is a downside to most cases of decentralisation. Local institutions should ideally be the units of local democracy. But suppose you leave everything to the village, Let them take action on inequality. The danger there is the village power politics are dominated by the same people whose power you’re trying to reduce. In our work we call it ‘local capture’. In the land question, for example, local landlords and big farmers dominate the decisions that are made. In fact, the advantage of higher-level decision-making is that it is much harder to capture at the central level. When you go to the district level, already the vested interests of different villages will neutralise one another because they are competing at district level. In that sense it is more difficult to capture (although it does still happen quite often). Still I believe in decentralisation. Power should be devolved but there should be checks and balances, and village democracy is the best way to install those checks and balances.
Also there is the matter of technology and capacity, in that there are some things which the villages do not have the capacity to do. There are also two sides here: in certain matters some local indigenous technology may be better for a given village than some technology from a manual that a bureaucrat writes in Delhi. But in some other matters – today you want connectivity or engineering technologies to build a water supply system or a dam for irrigation – they may not have the expertise. It depends on the problem. There is also the issue of what I call the largest brain drain in the world. We usually think about doctors, lawyers, other professional migrating from poor countries to rich countries. But the largest brain drain is from villages to cities, particularly in developing countries. So on certain matters the villages could have had the expertise, but those with knowledge have migrated.
Finally there is the funding situation: if the village panchayat had enough money they could hire experts, and say come and work with us for the next two years and help with these things we need. However, in India the village panchayats are starved of finance. This is the other major problem with Indian centralisation: we have panchayat elections and everything but village councils have to look up to the upper levels of government with folded hands for money. And it comes late; it comes in too small driblets and so on.
We’re currently seeing quite serious water conflicts at play – are there policy lessons from your research into collective action over resources at a local level that could be used to inform higher level policymaking?
Water is always a source of disputes, not just between Tamil Nadu and Karnataka but between farmers in a given place. Those farmers whose land is a little higher can always block the water to those below them and so on. I wanted to understand why local commons management problems get solved in some areas more easily that in others, so I studied 48 irrigation communities in Tamil Nadu. They used ‘tank irrigation’ where there are ponds and lakes – which are part of the local commons – from which they draw water to the farmers’ fields.
I collected data from the 48 communities and then did a statistical analysis where I had some dependent variables where I measured cooperation – simple questions like how do they resolve disputes? How quickly do they resolve disputes? The statistical analysis highlighted significant explanatory variables. One was inequality, which I had predicted. Villages where land was more unequally distributed had more problems in resolving disputes in general. This is what our hunch will tell us, but this is also what we found with a sample large enough so people wouldn’t say this is just a fluke. It did come out: inequality in general harms cooperative efforts.
There was another variable which I didn’t expect, but once I found it I wondered why I hadn’t thought of it before: villages which are closer to urban areas had more difficulty resolving disputes. It’s because cooperation involves give and take, but when you are proximate to the city your exit option is more accessible. There are certain obligations in the village cooperative arrangement which are unpleasant or costly for you, and when the crunch comes you can opt out. We called it the exit option variable, and that turned out to be quite significant. So these are just examples of things which are important in local commons management.
I originally started working on this issue on a grand all-India level because it is a problem which has bothered me since my childhood. There are things, such as cleaning the neighbourhood, that could be done and everybody agrees it should be done. Yet it doesn’t get done. I call it a collective action problem: we all know what is to be done, but I’d rather that you do it. I’ll enjoy the benefit but I’ll not share in the costs.
In my judgement, the more unequal a society, the more difficult it is to address the collective action problem. India does not just have economic inequality, it is socially one of the most unequal countries in the world. So this plays out at state and national level – even on big issues, people sometimes know the solution but it doesn’t happen. Litter should be cleaned by the untouchable, not you and me – that’s not my job! That’s the attitude. Ultimately inequality plays a negative role in collective action. That’s what I then carried to the local commons issue.
This post originally appeared on the SouthAsia@LSE blog.
This article gives the views of the author, and not the position of the South Asia @ LSE blog, nor of the London School of Economics.