Tata-Cornell Agriculture and Nutrition Initiative

Agricultural pathways to better nutrition and poverty reduction

Reflecting on this summer’s MNDA data collection: our indispensable enumerators

September 9, 2014 by · No Comments · Dietary Diversity, ICRISAT, MNDA, TCi Field Notes

Andrew Pike, a TCi intern and Cornell University student, reflects on his participation during the final phase of this summer’s Minimum Nutrition Dataset for Agriculture (MNDA), dietary diversity module pilot test.  Below, Andrew shares what it was like to return from the field, start the analysis phase and begin to digest what the challenges are for operationalizing the MNDA dietary diversity module. Read a synthesis of TCi’s work in creating a minimum set of nutrition metrics for use in agriculture surveys here, or catch-up with two previous MNDA intern posts about the first few weeks in India and the data collection process during their village-stays


Andrew Pike (white hat) and his local enumerator, Vaijanteemala, speak to woman and her child in Kanzara, Maharashtra.

AUGUST 2014: It’s an exciting time for the TCI intern team – we’ve just completed data collection and are now analyzing the results to see how the MNDA household dietary diversity scores compare with ICRISAT’s extensive nutrition survey. Professor Pingali arrives on Sunday and we present our findings to him and the whole of ICRISAT on Monday. It’s go time, my sweet mango lassis!

Now, while writing our final report, I’ve been reflecting over the generalizability of the data collection process across different countries and cultures. What strikes me most, and what I want to blog about, is that the field investigators (enumerators) are indispensable. We relied on them to accurately and impartially relay questions and responses, but also to provide context and facilitate relationships with respondents. This importance of this role means that investigators with certain skills and qualities are needed.

The interviewer and investigators must be able to communicate and so it is crucial that the investigator can read and speak English well (or another common language). The onus is on the interviewer to develop a rapport to improve communication. We found that role-playing in training was very helpful in this.

Investigators must also understand the survey methodology, specifically how and when to probe for foods. Again role playing in training helped because it highlighted potential miscommunications and mistakes that might occur, such as other household members trying to answer questions.

We were fortunate to work with the experienced field investigators for the ICRISAT Village Level Surveys (VLS), an intensive nutritional survey that captures all ingredients, no matter how small the quantity (as opposed to just major food groups with the MNDA).  However, we still needed to explain our instrument’s purpose and what information we were looking for due to confusion with the purpose of the ICRISAT instrument. It was also necessary to ask investigators to relay all information that might help contextualize dietary diversity, and not just the meal ingredients and their sources. For example, one respondent bought pigeonpea from the market, but this was unusual and only because the PDS was out of stock.

The investigators also had or were pursuing masters in agricultural economics or nutrition, which was useful but not necessary because the MNDA is designed so that it can be administered without a background in nutrition.

It is critical that investigators understand local customs and cultures and can brief the interviewer on how to act appropriately. It was useful that our investigators lived in the village because they had relationships with the households, who were therefore comfortable with the interview. This could be seen in the way they laughed and willingly answering questions. Also, through living in the village, investigators knew about household occupations, government welfare programs, and the ingredients in local dishes.

Researchers have a strong role to play in enabling investigators (enumerators) to do this kind of quality work. Productive  communication between the  the interviewer and the researcher is essential. Before this experience I did not realize how important my relationship, including my communication style and training, would affect the data collection process. As a result, I realize now that for the MNDA to succeed, we are going to need to ensure that rigrours trainings are held prior-to and during any data collection effort. Those in charge of implementation should select critically and train comprehensively, with emphasis on productive communication.  The interviewer should also take the time to talk with and ask questions of the investigator so that asking questions in the interview becomes habitual, and also so that investigators feel confident clarifying concepts.
All in all, it was a life-changing summer!


Agricultural Policy Priorities for the new Indian Government

August 11, 2014 by · No Comments · Agricultural Policy, Biofortification, Dietary Diversity, Food Subsidy, India Rural Poor, Public Distribution System, Research Agenda, Women's Empowerment

Despite overall economic growth that the country has been experiencing over the past few decades, rural India has largely been left behind.  Poverty and malnutrition are becoming increasingly concentrated in rural areas, particularly in the lagging regions of Central and Eastern India.  Even the more favorable states that benefited from the Green Revolution led growth have been exhibiting signs of productivity stagnation and a slow down in income growth.

Rural India has stalled in the process of structural transformation.  Despite large productivity and income differentials between urban and rural areas, and despite the declining share of agriculture in GDP, labor exit out of agriculture has been slow.  The lack of growth in labor-intensive manufacturing jobs in the urban sector, coupled with overly restrictive labor market regulations that lead to a dampening of growth in formal employment, have been the primary reasons for the slow withdrawal of unskilled and semi-skilled labor from rural areas. The stickiness of labor in low productive rural jobs has resulted in slow progress in poverty reduction in rural India.

Jumpstarting rural growth ought to be the highest priority for the new government.  Agricultural productivity growth continues to be the primary driver of income growth, but the approaches and priorities have to be different for the progressive states relative to the lagging states. For the progressive states, such as Punjab, Haryana, etc., growth will come from meeting urban middle class food needs.  The phenomenal growth in urban middle class population with rising incomes is leading to a dramatic transformation in diets.  The share of staple cereals, such as rice and wheat, is declining and is being replaced by an increasingly diverse set of vegetables, fruit, dairy, eggs, and meat products.  Traditional systems for supplying food to urban populations are gradually being replaced by the modern supermarkets.

Linking small farms to the growing urban food retail system is an enormous opportunity for enhancing productivity and incomes.  In the traditional Green Revolution areas – the irrigated, high productivity lands such as in Punjab, Haryana, etc., agricultural diversification from a predominant emphasis on staple crops to high value horticulture targeted towards urban markets ought to be the top priority.  National and State governments ought to enable the transition by promoting contract farming and liberalizing land tenancy laws.  Investments in cold storage systems, communications and transport infrastructure would be essential components of the strategy.  The current political ambivalence about private sector participation in food retail has discouraged such investments.  The participation of large and small private sector is crucial for building the food value chains that can provision the cities while at the same time enhance rural employment and incomes.

Agriculture is also the engine of growth for the lagging regions of central and Eastern India, states such as Bihar, Jarkhand, and Odisha, fall into this category. The lagging regions are still in the early stages of the agricultural transformation process and need high levels of infrastructure and R&D investments for enhancing productivity, improving food security and reducing poverty.  Unlike in the more progressive regions, the emphasis for the lagging states ought to be on staple food crops, including traditional crops, such as millets, and smallholder livestock systems.  Improving the ability of the agricultural system to tolerate frequent droughts, flash flooding, and poor soil conditions ought to be high priorities for public, as well as, private sector research and technology development.  The focus on weather variability will also help smallholder agriculture adapt, to increased incidence of extreme events and higher temperatures resulting from climate change.  The important role of biotechnology for agricultural technology development ought to be examined carefully and transparently.

NFHS Data 2005-2006 compiled by TCi (20014)

NFHS Data 2005-2006 compiled by TCi (20014)

The new government has a unique opportunity to reset agricultural policies, specifically moving away from the almost unilateral focus on staple crops, especially rice and wheat.  Diversification of food systems ought to be a specific policy goal, not just for meeting the needs of the urban middle class, but also for improving the poor’s access to nutrient rich food. Enhancing the supply of diversified food requires the government to level the playing field in terms of the incentives provided to farmers, such as price support and procurement policies.  Scaling back fertilizer and power subsidies that are disproportionately focused on the favorable, irrigated lands, especially in Punjab and Haryana, would significantly reduce the policy bias towards rice and wheat, and also help address the problems of environmental sustainability.

Finally, it is time to completely revamp safety net programs targeted towards the rural poor.  The current system of subsidized food grain provision to the poor is fraught with enormous problems – leakages and corruption are better known problems.  However, it is important to realize that the current system, even where it gets through to the poor, does not address the problem of malnutrition in India.  The exclusive emphasis of the PDS on rice and wheat keeps the program focused on adequacy of calories rather than address the need for a diet that is balanced and rich in micronutrients, such as iron, zinc and vitamins.

BMI of rural women (15-49 years) with BMI less than 18.5 (ICRISAT VDSA Data, 2011, compiled by TCi)

BMI of rural women (15-49 years) with BMI less than 18.5 (ICRISAT VDSA Data, 2011, compiled by TCi)

There are significant opportunities for testing new methods for providing the safety cushion that the poor need, cash transfer directly to the household woman head, in lieu of food, ought to tested as a mechanism to reduce graft. The incredible progress that India has made in unique biometric identification and in electronic banking can be major assets in this regard.  The effectiveness of the food transfer system itself could be significantly improved through better targeting of the poor and the food that they need.  From a nutrition point of view, the need is for greater diversity of food provided through the PDS system, especially, the provision of pulses and millets that are rich in micronutrients.  Expanding the food procurement system to include millets and pulses and sourcing them from Eastern India would achieve the dual objectives of enhancing smallholder incomes and addressing rural malnutrition through expanding food diversity.

In summary, creating true rural transformation will require the government to pursue a three-pronged approach.  First, building efficient value chains that connect rural smallholders to urban food retail systems. Second, addressing low agricultural productivity in the lagging regions of central and eastern India, with an emphasis on improved technologies for stress tolerance and nutrition improvement for the primary staples, such as rice and wheat, as well as, traditional millets and pulses.  Third, enhancing the food security of the ultra poor by implementing innovative approaches for better targeting of beneficiaries and improving micro-nutrient access through food diversification and bio-fortification.

Communicating TCi work through digital shorts: the MNDA story

July 29, 2014 by · 1 Comment · Community Participation, Dietary Diversity, ICRISAT, MNDA, Rural Diets

“To solve the world’s most pressing problems, people do not need more volumes of information and knowledge, they need to acquire the capacity to talk to each other across the boundaries of culture, religion and language…{that kind of} dialogue is an extremely difficult form of speech.”

-Hamlink, C (2002)

Development communications is a critical component of international development work. The use of communication processes, techniques and media can help people toward a full awareness of their situation and enable them to identify options for change. It can aid in conflict resolution, consensus-building, and knowledge acquisition that can improve livelihoods and make institutions more effective (Fraser and Restrepro-Estrada 1998).

Beyond the quantitative and qualitative skills necessary to conduct research (say, data collection, analysis and research design), TCi works to facilitate opportunities for staff and scholars to communicate their research and findings more effectively–and to learn how to listen and facilitate the kind of dialogue that can help make change.

TCi runs or collaborates to provide a number of “short courses” and seminars throughout the semester that can help individuals gain experience in science communication. In the spring of 2014, TCi joined with Homelands Productions to offer a short course on video storytelling for TCi scholars and research staff. The purpose was to equip researchers with skills to create compelling 2-4 minute short videos that can help them convey research and fieldwork findings to larger audiences. The TCi video storytelling workshops and projects are continuing in the fall of 2014 and beyond.

During the June and July, our TCi interns experimented with creating these 2-4 minute videos (see link below). In the video below, Christian Delrado-Owens shoots and narrates a short video on the MNDA work that’s currently underway in Maharashtra and Andhra Pradesh. We’re looking forward to a few more videos directed by Christian and other TCi staff/students later this fall!

TCi Video: Pilot Testing the MNDA in Andhra Pradesh


Hamelink, C. (2002). ‘Social Development, Information and Knowledge: Whatever Happened to Communication?’ Development, Journal of the Society for International Development 45(4): 5-9.

Fraser, C., Restrepo-Estrada, S. (1998). Communicating for Development: Human Change for Survival. London and New York: I.B. Tauris.


The first round of MNDA dietary diversity data collection concludes: reflections from Dokur village (Andhra Pradesh)

July 15, 2014 by · 1 Comment · Community Participation, data collection training, Dietary Diversity, India Rural Poor, MNDA, Rural Diets, TCi Field Notes

Christian DiRado-Owens, a TCi intern and Cornell University student, reflects on the first round of data collection in the field. Christian, in addition to four other interns and staff, are hard at work pilot testing the dietary diversity module of the Minimum Nutrition Dataset for Agriculture (MNDA). Read a synthesis of TCi’s work in creating a minimum set of nutrition metrics for use in agriculture surveys here or previous field-update from TCi intern Katy Merkel… 

Christian DiRado-Owens (left, TCi) and Swathi (right, ICRISAT) speak with women from Dokur, in Andhra Pradesh. (Photo credit: DiRado-Owens, TCi)

This week the TCi team left the ICRISAT headquarters to officially begin the first of two rounds of data collection. The team traveled to two villages in Andhra Pradesh: Aurepalle and Dokur, located just over two hours from Hyderabad. Our purpose is to evaluate and validate the use of a module within the Minimum Nutrition Dataset for Agriculture (MNDA) tool that attempts to capture dietary diversity. Using a simplified methodology and focusing on women (ages 18-45) responsible for cooking, the MNDA attempts to capture a household-level and individual level dietary diversity score. Over the last week two teams completed a total of 66 dietary diversity surveys with women from various income levels, castes and classes. Amrita Rao, another TCi intern, and I were stationed in Dokur while Andy Pike, Katy Merkel, and Alex Cordova (the other three TCi interns) were stationed out of Aurepalle.

Arriving in Dokur and facilitating group discussions about diet

Upon arrival in Dokur, Amrita and I met with our investigators — Vidya and Swathi — who would be translating back and forth between English and Telegu during our household interviews and focus group discussions. Along with another ICRISAT intern named Asha, we held our focus group discussions on a multipurpose purple, green, and orange mat. On this mat, we slept, organized our materials, took our meals from our sweet home-stay cook Yadamma, and spent the majority of our down time when the village women were working and not available to meet. Together, Amrita and I held three focus group discussions and conducted 28 household surveys from Monday through Friday.

The first focus group discussion consisted of both men and women for a total of ten villagers, while the second consisted of only women for a count of eight from the forward and backward castes. The purpose of our focus group discussions (FGDs) was to validate our food groups, familiarize ourselves with local dishes to inform what we should probe for in the household surveys, and to validate the assumptions our survey implicitly made regarding meal frequencies and times, cooking responsibilities, and foods cooked on special days, etc. The FGDs were helpful in familiarizing ourselves with local dishes, though we did not find much variation in the responses between focus groups. The biggest variation was in response to food preference and who ate first: the mixed group reported that children have first priority, followed by men, then women. In the female-only group, some participants reported that children have first priority and others reported that men do. The female-only group indicated that women always ate last. In the FGD with villagers from the scheduled caste, however, there was consensus that men have first priority, followed by children, then women.

Migration makes things messy: especially when identifying a household dietary proxy

The MNDA is  organized around the assumption that the women who cooks for the household (someone between 18-45) could plausibly be a proxy for the household diet, given that she herself was a member of the household and that she cooked the food eaten by the majority of the members of the household. We, however, found a few exceptions. For example, one women in our sample population had just returned from an extended period of time (the greater part of a year) living and working in Hyderabad to send remittances home and earn supplemental income for the family. Although this woman met all of the selection requirements, she clearly could not be a proxy for the household as she would neither be eating nor cooking the same foods as the rest of the family who remained in the village while she was away. While others in our individual survey respondents did not fall into this category, it was clear that finding the household cook did not mean that she was always the ideal respondent. We asked the first question during our focus group discussion with the women from the scheduled caste, and learned that a  portion of the female population in Dokur migrated for extended periods of time and came back based on seasonal opportunities (in agriculture). Interestingly, when we asked who assumes the household cooking responsibilities when the women (and the men) are away, most informants said that the children take over.


Amrita Rao (left corner, TCi) listens during a mixed, male-female focus group discussion about diets in Dokur, Andhra Pradesh. (Photo credit: DiRado-Owens, TCi)

Dokur, however, exists in some unique circumstances. Due to the impacts of persisting drought over the last thirty (and specifically 12 years) with low investment power and low capital, agricultural production has declined greatly and families are looking elsewhere for supplementary income. Coupled with Dokur’s relatively isolated location (in relation to Hyderabad) in the district of Mahbubnagar, Dokur’s out-migration rates have increased dramatically. Due to the distance and expense of traveling to Hyderabad, many migrants opt (or are forced) to stay for longer periods of time in or around Hyderabad. This is especially the case for the poorer, lower castes and classes who are forced to migrate for income but cannot afford the regular commute. A six to eight hour roundtrip bus ride, for example, costs an estimated 108 rupees.

Spending time in the village: getting context, gaining insight, and making friends 

Over the course of a week, I can confidently say that we established strong working and personal relationships with both Vidya and Swathi (and Asha), and I am grateful to have had the opportunity to do so. We also were able to do some filming this week for a TCi video project about  the challenges of pilot testing the MNDA and about what fieldwork is like for a Cornell student. We captured footage of our household interviews, focus group discussions, agricultural landscapes, market transactions, and a woman named Buchemma who taught us how to make traditional raghi roti.

Our last night before returning to ICRISAT, we all met with some of the women and children in the village, exchanged words, sang songs, and danced, coalescing into a great time and a great memory. A special thank you to our ICRISAT investigator colleagues who are currently stationed in Dokur: Vidya and Swathi!

Tags: ···