Learning from a publicly subsidised agricultural insurance: evidence from Bolivia

3ie evidence programme:  Agricultural Insurance Evidence Programme
Author(s): Alex Armand, Sergio Daga, Carlos Gustavo Machicado,  Ricardo Nogales, Boris Branisa, Joseph Flavian Gomes,  Giulia Montresor, Ivan Kim Taveras
Institutional affiliation(s): INESAD, University of Navarra and NCID, University of Navarra and NCID, University of Navarra and NCID, University of Essex, Universidad Privada Boliviana 
Grant-holding institution: Navarra Center for International Development, University of Navarra
Main implementing agency: Navarra Center for International Development, University of Navarra
Sex disaggregation: No
Gender analysis: No
Equity focus: Yes
Study type: Process evaluation

Context

Bolivia is characterised by the prominence of agriculture in its economy and by growing climatic risk. Approximately 30 per cent of its labour force works in agriculture and is prone to extreme weather (with a mean altitude of 1,192 meters). The agricultural sector is also the main source of income for 72.4 per cent of the rural population (UDAPE, 2015). At the same time, Bolivia faces high poverty rates, with a GDP per capita of 2,867 USD (World Bank), the lowest in the South American continent. To mitigate the risk that farmers face in developing countries, micro-insurance has become an increasingly popular solution. In Bolivia, the government pioneered this strategy to support smallholder farmers in deprived areas in the form of the PIRWA crop insurance programme.  In 2011, the Government of Bolivia enacted the Law of the Productive Community Agricultural and Livestock Revolution (Law No. 144), which led to the creation of the National Institute of Agricultural Insurance (INSA, for its acronym in Spanish) as a decentralised public institution with its own assets and autonomy of management under the patronage of the Ministry of Rural Development and Lands.

Intervention design

This study assessed the feasibility and acceptability of the PIRWA programme. It is primarily targeted at subsistence farmers in the poorest municipalities of Bolivia. The zero-fee insurance was designed for farmers who own up to the 3 hectares of land and guarantees participating households a payout of 1,000 Bs or 146 USD per hectare against damages caused by floods, droughts, hailstorm and frost.

The farmers enrolled in the programme have to meet four targeting criteria: (1) smallholder farmers with less than 3 hectares of productive land, (2) located in municipalities with higher rates of poverty, (3) located in municipalities with a high dependence on agriculture and (4) produce crops for local consumption, rather than exporting. Specific crops covered include potatoes, corn, wheat, barley, quinoa, beans, oats, alfalfa and frijoles.

Evaluation design and methodology

The evaluation uses a mixed methods design.  Quantitative data came from a range of datasets, consisting of the Bolivian household survey data for the years 2005–2014 available annually with municipality identifiers, the 2013 Agricultural Census, and the 2008 and 2015 Agricultural Surveys. The researchers used the Difference-in-differences (DID) methodology to estimate the impact of the phased implementation of the PIRWA programme to new municipalities over time. Qualitative methods were used to help interpret the causal effects computed in the quantitative analysis; it included focus group discussions (FGDs) with 141 farmers.

Primary evaluation questions

The evaluation answers the following questions:

1. How do weather shocks affect the productivity, behaviour and welfare of farmers in Bolivia?
2. What determines the take-up of micro-insurance once we remove the liquidity constraints that smallholder farmers face?
3. What are the welfare and productivity effects of the PIRWA programme?

Primary findings

Unfavorable weather shocks affect individuals negatively by reducing yields and inputs but not expenditures. An extra day with favorable temperature during the growing season leads to an increase in yields of around 25 per cent, but an extra day with extremely high temperatures leads to a reduction in yields of around 20 per cent and a reduction in expenditure on agricultural inputs by roughly 60 per cent.

The main determinants of participation in the PIRWA programme are the demographic characteristics targeted by the programme. Higher share of agricultural land that is farmed with crops targeted by PIRWA is associated with increased take-up of the programme. Higher poverty rates also increase take-up, but extreme poverty reduce take-up. Weather conditions were also important determinants of take-up.

Being exposed to the PIRWA programme has significant and robust effects on agricultural productivity in the treated municipalities. No significant effects of PIRWA were detected on either total household expenditure or food expenditure. There was a significant positive effect on extreme poverty. Ensuring all the agricultural land in the municipality leads to a 7 per cent reduction in the extreme poverty incidence.

Results from focus group studies indicate that farmers have limited knowledge of insurance products and limited financial education. However, farmers recognise that there are many requirements to access insurance.

Implications

Authors conclude that removing the lack of trust and liquidity constraints does not ensure a 100 per cent take-up of agriculture insurance.  Even after repeated exposure and insurance being offered at zero cost, take-up was hindred by lack of understanding about insurance.

Sensitisation activities at the municipality level, with local leaders, and then with communities were unsuccessful for improving knowledge and understanding about insurance. More research and action is needed to understand effective strategies to increase insurance knowledge and understanding.