this weekend after reading Dr. Krishna’s papers (Duke University) i feel even more confident that large scale rural poverty in India can be solved, but that it is certainly not going to be solved the way we have been going about it historically. if we want to make a dent in the numbers, the survey data clearly indicate that poor villagers simply need make more money to afford basic needs like food and health care sustainably year after year. information services catering to mainstream agricultural needs are not enough to do this. we need new information services that can *connect* villagers to new local and regional economic opportunities/markets beyond agriculture. roughly, the total market for agriculture is limited to perhaps $100B or $200B annually and crop yields are at the mercy of weather, droughts, and so on. they need to somehow find 10x that. see my earlier post “India’s 450 million.” that got me thinking about all of this.
these papers show several commonly held assumptions are wrong. the survey teams talked to many thousands of families in 36 villages in each of Rajasthan, Andhra Pradesh, and Gujurat in which poverty rates are 25-65% and vary a lot from village to village (http://www.pubpol.duke.edu/krishna/communities.htm). they also conducted parallel studies in Kenya/Uganda and in all cases found several similar patterns across india and Africa.
take education. it is a commonly held belief that villagers don’t want to send their kids to school. but that doesn’t seem to hold up against the data. even then, nor has primary education played significant role in which large numbers of families have risen or fallen out of poverty.
in the surveys they asked what villagers consider to be the poverty threshold, in order of priority. there is some variation among the states, but within each state they found universal agreement on villagers’ priorities which define the point at which they consider a family to be poor — only after this cutoff, villagers opinions diverged in each state.
Rajasthan — food, primary education, clothes, paying off debt
Gujurat — food, clothes, primary education, paying off debt, patching leaky roofs, farming a small plot of land (sharecropping)
Andhra Pradesh — food, patching leaky roofs, paying off debt, clothes
with the exception of AP, primary education was high up on the list only after food. why not in AP? the reason in AP is “not because parents care less about educating children, but because primary education is almost universally provided here and it is no longer something that is out of reach of even very poor households. All but a miniscule number of primary school-age children attend primary schools in these 36 villages, and annual fees for government-run primary schools are a pittance, easily within reach of even very poor households.”
across all three states, there were similar patterns in how families fell into poverty or rose out of poverty over 25 years. The number of families rising out of poverty (11% in Rajasthan, 10% in Gujurat, and 14% in Andhra) has been almost counteracted by the rate of falling into poverty (8% in Rajasthan, 6% in Gujurat, 12% in Andhra) for a net poverty reduction of only 2-4%.
in all three states, the most common pattern for 60-80% of villagers who fell into poverty was nearly identical — inability to afford the heavy costs for health care in chronic and life threatening illnesses, costs of social functions like death ceremonies and marriage functions, which in turn leads these families to take on high interest rate debt from local loan sharks (~3% per month). factors like laziness/drunkenness were identified in single digit percentages and not a major factor.
in the other direction, in almost all cases the only way families have risen and managed to stay out of poverty is simple: making more money by getting out of agriculture and diversifying their income sources so they can afford health care, support a family/social life, get a roof over their head, and repay their debt. they have done this by getting a better job in a city or accessing a new urban market through a contact/friend, farming very different more profitable crops, or getting a government job. it turns out primary education played a relatively minor role compared to other factors — so why push education on them if they will send their kids to school anyways, once they make enough money to eat?
migration to cities and government jobs can work for a limited number of families, and can only go so far and is not scalable. there needs to be a way to enable small-scale entrepreneurs with specialized skills to gain access to local markets to build up their own villages/regions by supplying higher value-add services/goods where the need is greatest and so the price is highest (long tail) — just like how AdWords/AdSense, EBay, and Amazon are enabling all sorts of small businesses in the US to gain access to markets they couldn’t otherwise reach before.
how can we do this in india? Internet access via kiosks and rugged laptops for kids (OLPC) is only part of the equation. perhaps one way to get started would be to create a service for villagers to input classified ads via the web and browse/search them the same way (like craig’s list) — keeps it simple to setup/operate, and easier to focus on getting the content right with categories tailored to rural economy’s needs such as electrical repair work, ads for doctors/hospitals/specialists, tractor repair needs, water purification, dramas/skits/performances, new shop/business openings, bicycles for sale, animal care, and so forth. the list is endless, unpredictable, and long-tail. the important thing is should be derived directly from demand/supply in the regional rural economy and can be extended to incorporate literally anything if a user needs to include it. if the site enables broader distribution not otherwise possible, the early users would be entrepreneurial buyers who want to get lower prices and entrepreneurial suppliers who will “do what it takes” to gain a competitive edge by connecting supply to demand.
for the foreseeable future, my understanding is that Internet/web terminal penetration will remain much lower than cell phones. Internet kiosks are mostly limited to larger population centers, but still within a day’s travel reach of most villagers. cell phones can reach a larger audience anywhere in the villages on a daily basis, and this can be combined with less frequent access to kiosks in larger centers. as site usage grows, it could justify supporting other access methods — which requires additional software development. a rural user could be given the capability to subscribe to “alerts” on a cell phone (txt message, voice mail), even via an interactive voice interface on a cell phone, or via the web itself at an Internet terminal. for instance, a worker who is skilled in electrical work may want to be notified by cell phone when an ad shows up within a radius of 20 miles of his village — once he gets a notification, he can request more details on his cell phone or otherwise go check at the nearest Internet terminal to see more details of the job requirements. replace “electrical work” with “bharata natyam” dance performance and you have a local cultural network.