Agricultural Risk Management in India

How to cite

Cole, Shawn; Tobacman, Jeremy. 2015. "Agricultural Risk Management in India." AEA RCT Registry. AEARCTR-0000677. https://doi.org/10.1257/rct.677-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Shawn Cole* , Jeremy Tobacman

* Primary investigator

Registry ID
AEARCTR-0000677
DOI
https://doi.org/10.1257/rct.677-2.0
First registered
Last updated
Status
On going
Trial dates
2006-05-01 – 2015-12-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/677

Abstract

Many rural poor in the developing world engage in agricultural work, either as cultivators or laborers. These individuals often face serious risks from weather fluctuations – above all, erratic, excess or inadequate rainfall. Households attempt to informally insure themselves against these risks by diversifying the crops that they grow and by relying on friends and family in times of distress. However, informal mechanisms do not work as well when everybody in a given area is affected by the same adverse weather event. Historically, formal yield-based crop insurance has suffered from monitoring, selection, and incentive problems as well as high administrative costs.

Rainfall index insurance is a product that could help farmers hedge against agricultural losses due to excess or deficit rainfall, which is often cited as the largest risk to agricultural households. In a rainfall index product, policies pay out when rainfall exceeds or falls below pre-determined triggers based on historical rainfall data. Payouts are easy to calculate and index-based policies are simple to price since rainfall data are readily available. Another advantage of rainfall insurance is that the transaction costs can be low since insurance companies do not need to verify claims. Further, since farmers have no control over the level of rainfall, their choices to adopt the insurance or modify other behavior do not affect the sustainability of the insurance, in contrast to many yield-based crop insurance programs.

This randomized evaluation will help shed light on decision-making about weather risk and evaluate the potential of rainfall insurance to improve the livelihood and sustainability of rural poor households.

Intervention

Intervention is the index-based rainfall insurance product marketed to treatment households, the product has been marketed and distributed by SEWA, the Self Employed Women’s Association.

Experimental design

The study has been conducting annual household surveys since the beginning of May 2006, collecting data on agricultural production, income, consumption and well-being. In addition to the village-level random variation in access to the insurance, incentives and information relevant for purchase choices are randomized at the individual level.

The first incentive manipulation implemented was a price discount, to estimate the elasticity of demand. In addition, willingness to pay for the rainfall insurance has been elicited repeatedly via the Becker-Degroot-Marschak (BDM) methodology. The random offer prices induce random variation in coverage, supporting the intent-to-treat impact evaluation design. The study utilized paper flyers and handheld video players, among other means, to randomize information relevant to purchase decisions. Variations of both flyers and videos were presented to randomly selected subsets of treatment households. Also, in 2010 a short financial literacy module was conducted, to which invitations were randomized at the household level. The incentive and information manipulations administered in the project are useful as instruments for take-up and also are of independent interest.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Village Level
Clusters
108 Villages (Initial plan was 100 villages, 8 villages added)
Observations
2041 Households
Arms
The ongoing study includes 108 villages in the Ahmedabad, Anand and Patan districts of Gujarat. 48 villages belong to the control group, while weather insurance was rolled out to the 60 treatment villages in three waves.
Primary outcomes
Insurance purchase, Household level agricultural outcomes (ex-post and ex-ante) i.e. revenues, costs, area cultivated, cash crop production etc., determinants of household demand for rainfall insurance.

Cited by

No citation data available for this registration yet.

Keywords