Combating malnutrition

4

OCTOBER 2019

Can group procurement be equitable?Results fRom a food secuRity pRogRam in RuRal BihaR

This note is based on the paper The Distributional Consequences of Group Procurement: Evidence from a Randomized Trial of a Food Security Program in Rural India, authored by P. Christian.1

CONTEXT

Bihar is the third most populous state in India with 8.1% of the country’s population or 104 million people, 89 % of whom live and work in rural areas. Rural Bihar is characterized by suboptimal service delivery, a complex political and social fabric, limited inclusion in institutions, limited economic opportunities, poor development infrastructure and limited availability of financial services for the poor. Bihar lags behind the national average in all indices of nutritional status – 50% children aged five and under are stunted, 21% have muscle atrophy and more than 44% are underweight.

The Public Distribution System (PDS), India’s primary food security program is plagued by inefficiencies, leakages and wastage related to systemic factors of procurement, transportation, storage and targeting. Notably, the costs of the system are higher in administratively weaker states that account for a large share of the poor.

Landless subsistence workers2 are most at risk since they depend on local markets to purchase seasonally available goods. As a consequence of male out-migration, women headed households are dependent on infrequent

HIGHLIGHTS

India has among the highest rates of child •malnutrition worldwide, with Bihar one of the worst affected states.

With the intent of combating food •insecurity, the Government of India (GoI) and the state Government of Bihar (GoB) offer a Food Security Fund (FSF) via the Bihar Rural Livelihoods Promotion Society.

The program has proved successful in •giving rural women agency to ensure food security for themselves and their families.

The scheme is most effective in reaching •the poorest households when local committees procure the most affordable grains.

_______________1. The original paper is available at http://documents.worldbank.org/curated/

en/438651467041713497/pdf/106448-WP-PUBLIC.pdf2. Comprising 70% households in Bihar.

and insufficient remittances to purchase food which is typically neither nutritionally diverse nor enough in quantity to meet their nutritional requirements.

In order to avoid these inefficiencies, programs have begun decentralizing procurement

South Asia Agriculture and Rural Growth Impact Note SeriesP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

ed

impaCt note 4: oCtober 2019

to allow communities to choose the goods themselves and procure them locally. An important concern in Bihar is whether decentralized procurement will improve or worsen targeting outcomes. If communities choose to buy the lowest cost, but also lowest quality foods available, the intervention could be self-targeting since poor households will find it worthwhile to participate. If communities choose to buy higher cost foods that may be higher quality, the program will not maximize the degree to which benefits accrue to the poor.

_______________

3. These objectives were also attuned to the World Bank’s Country Strategy (CAS-August 2004) for India, which highlighted support for lagging states including Bihar through sustained growth, poverty alleviation and improved access to quality social services.

BOX 1: OBjecTIveS (MDGS)

The stated objectives of this intervention directly supported the following Millennium Development Goals:

(i) Goal 1, Target 1: improving incomes of the poorest - reducing the proportion of the people living on less than $ 1 a day.

(ii) Goal 3, Target 4: promoting gender equality and empowering women.

(iii) Goal 8, Target 45: development and implementation of strategies for decent and productive work for youth.

Cluster-level Federation

42

102

225

Project CompletionTarget Revised Target

5,90,000

15,00,000

Poor Households

18,36,505

Households

25,000

Village Organizations

770

10,000

10,445

1,46,291

Community members

60,000

62,460

Self-Help Groups

1,50,000

1,54,626

55,000

BOX 2: BRLPS TaRGeTS anD acHIeveMenTS

INTERVENTION

The Bihar Rural Livelihoods Promotion Society (BRLPS, colloquially known as JEEViKA) was established by the GoB to help develop scalable and sustainable models for livelihoods enhancement of rural poor households. The objective was to reduce the Below Poverty Line (BPL) population to 22% by 2015 or about 1.5% per year, in order to achieve the Millennium Development Goals (MDG) target and the GoI’s Eleventh Five Year Plan strategy which emphasized investment in women’s socio-economic empowerment, enhancing livelihood opportunities, and participation of local level institutions to make service delivery more accountable.3

JEEViKA piloted in 6 blocks in 6 districts of Bihar – Gaya, Khagria, Madhubani, Muazzafarpur, Nalanda and Purnea. With the NRLM program it expanded to all 534 blocks in 38 districts by 2011.

The BRLPS’s primary role is to organize women into Self-Help Groups (SHGs) that conduct microfinance activities and participate in other BRLPS programs. SHGs are further grouped into Village Organizations (VOs) consisting of 100-150 women, all of whom live within the same village. The project design targets poor rural women by considering their existing social vulnerability and spatial distribution. From 2008-2011 the project mobilized more than 1.8 million poor rural women from rural households into 1,54,626 Self-Help Groups (SHGs) and 10,445 village organizations.

While the BRLPS was not specifically mandated to reduce malnutrition, since most households incurred debts due to health and nutritional expenses, their SHG and VO network became an ideal vehicle for policy makers to deliver targeted public goods in order to combat the issue.

The Food Security Fund (FSF) is a direct livelihoods intervention designed to target the most credit-constrained

impaCt note 4: oCtober 2019

members of the VOs, without exposing them to high interest rates. In response to the issues of leakage and wastage in existing programs, and to improve community participation in the process, the program incentivizes group procurement in local markets. Although the rice is not directly subsidized, the program encourages bulk purchasing - thereby reducing the cost of rice at the vendor level. Pooling the demands of the entire VO together also improves bargaining power, and calls for more competitive bids. Most importantly, purchases through the food security funds avoids interest costs for (the poorest) households that would otherwise rely on credit to buy rice.

baseline, 1,518 households were re-surveyed in at least one of the two follow-up rounds with balanced attrition (p=0.76).

Importantly, in all the clusters being evaluated, rice was the food item on which the FSF subsidy was targeted4. The empirical data thus collected was correlated with a pre-existing model to evaluate the impact of the program.

KEY FINDINGS

Since rice is a non-homogenous grain with several different qualities available in the same shop5 it is problematic to

FSF`1,00,000 VO

VO solicitsbids from

nearby grain vendors

Grain distributedto the VO

members on the basis of individual demand

3 Monthsto pay

individualshares

BOX 3: FOOD SecuRITy FunD

_______________

4. In Bihar rice is one of the three staple food, others being wheat or maize, which together account for 33% of the food expenditure of the households surveyed. Including home produced goods, rice alone accounts for 25% of all expenditures of food in the state.

5. The quality and price of rice are positively correlated, with better quality rice costing more.

EVALUATION DESIGN

A cluster-randomized control trial was conducted in six administrative blocks, which had a significant numbers of VOs and had not yet availed of the FSF. The blocks were divided into three clusters each, the program was implemented in 18 clusters comprising 90 VOs, with 45 VOs each designated as treatment and control. Baseline data was collected from five VOs in each of the 18 evaluation clusters in July 2012 prior to initiation of the lending program. The lending program began in September 2012 and a post-intervention survey was completed in February-March 2013 covering detailed food expenditure, weekly consumption, experiences of food insecurity, agricultural production, income and child health indices. Out of the original 1,529 households surveyed at

target the subsidy at the desired beneficiaries. The poorest households that are typically most at risk from food insecurity benefit the most when the VOs choose the lowest quality of rice for procurement. Empirical results show significant increases in consumption for the wealthiest in the average community. Impacts for the poorest households are sensitive to which variety of rice is procured by the community Increase in consumption for the poor can be seen when enough villages choose to opt for the lowest quality grain.

Moderately poor households, who may typically consume better quality rice, adjusted their preferences to select into the program when the lowest quality of rice is procured. However, higher income households chose to opt out of the program in this scenario.

Figure 1: Correlation between household wealth and FSF participation

Lowest cost rice Highest cost rice

Calculations from jeevika FSF IE ending survey. Household wealth is the natural log of household expenditure inthe last month in Rupees.Lowest and highest cost are the bottom and top quartile of price of purchased rice

60

40

307.5 8 8.5 9 9.5 10%

of

Ho

use

ho

lds

Par

tici

pat

ing

50

When better or the best quality of rice is chosen, take up is concentrated amongst the wealthier households. The poorest households may still take up the program to increase the quality of rice they consume, but it leads to a reduction in their consumption, because they shift to purchasing higher cost food than they did in absence of the program. This reduction of food grain consumption among the poor and moderately poor was an unintended effect of the program.

POLICY LESSONS

Understanding the processes by which some of the VOs successfully selected the lowest quality and price of rice could reveal mechanisms that make the group procurement approach of the program better targeted towards the poorest households. Decentralizing the choice of good(s) to subsidize may seem like a good way to incorporate local preferences, but in the absence of adequate controls, the process may exacerbate local inequalities. The quality of rice selected reflects the VO committee members’ desirability

for that rice, and not necessarily the needs of the poorest households in the cluster.

An analysis of VO committees reveals that wealthier members are disproportionately represented therein and are more likely to participate in the program than non-committee members. To reinforce the potential for program participation to be well targeted, the program design should be adjusted to ensure that the choices of the poorest households are given more influence in procurement decisions.

CONCLUSIONS

Besides the well-documented challenges of elite capture, limited information about markets and preferences continue to plague even the most well-intentioned development efforts. Improving both efficiency and equity in programs involving group procurement will require policies that combine in-kind transfers with mechanisms that address beneficiary preferences, including for the poorest sections of society.

ABOUT THE IMPACT NOTE SERIESThis note is part of the South Asia Agriculture and Rural Growth Impact Note Series, that seeks to disseminate research and analysis focused on World Bank financed rural, agriculture and food systems programs in India. Series editors: Abhishek Gupta and Gayatri Acharya. Photographer: Ritayan Mukherjee.

We are grateful for funding from/the support of The South Asia Food and Nutrition Security Initiative (SAFANSI), Gates Foundation, Ministry of Rural Development and various state rural development departments.

This note was prepared by Deepti Kakkar (Consultant, World Bank) and published by the Food & Agriculture Global Practice at The World Bank, 1818 H Street, NW Washington, DC 20433 USA. Website: www.worldbank.org. For further information or copies please email: galex2@worldbank.org.

Disclaimer: The findings, interpretations, and conclusions expressed in this note are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent.

REFERENCESK. Das. Research Gaps and Policy Priorites: Evidence from Bihar. What is available and what is required? International Growth Centre, 2018

Note 1. Jim Hancock, Mrinalini Penumaka and Mio Takada. Main Findings of Implementation Completion Review Report—Lessons from a Decade of Implementation of BRLP, and Note 8. Anna-Lisa Noack, Padma Buggineni, Apolenarius Purty, and Somya Shalini. Improving Nutrition through Livelihoods in M. Takada and M. R. Choudhury. Ed. A Decade of Rural Transformation, Lessons Learnt from the Bihar Rural Livelihoods Project – JEEViKA, JEEViKA Learning Note Series, No.1, World Bank Publications, 2017

Note 8. Anna-Lisa Noack, Padma Buggineni, Apolenarius Purty, and Somya Shalini. Improving Nutrition through Livelihoods in M. Takada and M. R. Choudhury. Ed. A Decade of Rural Transformation, Lessons Learnt from the Bihar Rural Livelihoods Project – JEEViKA, JEEViKA Learning Note Series, No.1, World Bank Publications, 2017

P. Christian. The Distributional Consequences of Group Procurement: Evidence from a Randomized Trial of a Food Security Program in Rural India, The Social Observatory – Development Research Group, World Bank, July 1, 2015

P. Sanyal, V. Rao, and S. Majumdar. Recasting culture to undo gender: A sociological analysis of Jeevika in Rural Bihar, India. World Bank Policy Research Working Paper, (7411), 2015

R. Radhakrishna and K. Subbarao, India’s Public Distribution System, A National and International Perspective. World Bank Discussion Paper No. 380, 1997

U. Datta. Socio-Economic Impacts of JEEViKA: A Large-Scale Self-Help Group Project in Bihar, India. World Development, Volume 68, pp. 1-18, 2015

U. Datta, Retrospective Impact Evaluation of BRLP (2007-2010), Socio-Economic Impacts of JEEViKA, a project by the Bihar Rural Livelihood Promotion Society: A Retrospective Study, The Social Observatory – Development Research Group, World Bank, 2016

V. Hoffmann et al. Poverty and empowerment impacts of the Bihar Rural Livelihoods Project, 3ie Grantee Final Report, New Delhi: International Initiative for Impact Evaluation (3ie), 2017

World Bank Document – Working Paper. A decade of rural transformation : lessons learnt from the Bihar Rural Livelihoods Project - JEEViKA (English), January 1, 2018, Report No. 122548

World Bank Document - International Development Association Project Appraisal Document on a Proposed Credit in the Amount of US$ 290 million to the Republic of India for a Bihar Transformative Development Project “JEEViKA II” June 1, 2016, Report No. PAD1892

World Bank Document – Implementation Completion Report Review. India - Bihar Rural Livelihoods Project (English), July 5, 2017, Report No. ICRR0020734

impaCt note 4: oCtober 2019