Measuring the Quality of Agricultural Index Insurance: Concepts & Safe Minimum Standards Michael R Carter NBER & University of California, Davis BASIS I4 Index Insurance Innovation Initiative http://basis.ucdavis.edu Brown Bag Seminar, World Bank May 21, 2015 M.R. Carter Index Insurance Quality
22
Embed
Measuring the Quality of Agricultural Index Insurance: Concepts and Safe Minimum Standards
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Measuring the Quality of Agricultural Index Insurance:
Concepts & Safe Minimum Standards
Michael R Carter
NBER & University of California, DavisBASIS I4 Index Insurance Innovation Initiative
http://basis.ucdavis.edu
Brown Bag Seminar, World Bank
May 21, 2015
M.R. Carter Index Insurance Quality
Changing the Conversation ...
Disappointed (angry) farmers & what are sometimes called“Basis Risk Events” have punctuated the importance ofdesigning contracts that protect farmersThe problem is far from trivial as the following analysis of therelationship between average losses and indemnity paymentsunder rainfall insurance in India shows:
M.R. Carter Index Insurance Quality
... from Basis Risk to Contract Quality
We sometimes get caught up in discussions about what isbasis risk, or whether farmer disappointment reflects ’true’basis risk or farmer misunderstanding of contract limitsBut even if we define basis risk such that instances ofdisappointed farmers should not be called ’basis risk events,’we still need to answer the question whether a contract meritspurchase by the farmer (and public support to promote it)Rather than thrashing out the correct definition of basis risk,perhaps we should talk directly about what we mean by qualitycontracts that:
1 Adequately protect farmers against income fluctuations; and,2 Can achieve the objectives we seek in offering insurance to
developing country farmers (income & consumptionstabilization, & investment incentives)
M.R. Carter Index Insurance Quality
Where Index Insurance Can Achieve Development Objectives
Risk is not always the constraint, and even when it is, indexinsurance may not be the solution:
Source: Carter, Cheng & Sarris (2015). “Where and How Index Insurance CanBoost the Adoption of Improved Agricultural Technologies.”
M.R. Carter Index Insurance Quality
Back to Basics to Measure Contract Quality
At its simplest, insurance should stabilize income flows (and,or protect assets), underwriting:
More stable consumption and stable investment in humancapital (nutrition & education) that are subject to’irreversibilities’Improved investment incentives by reducing the risk of capitalloss
A few simple pictures can help frame our thinking:
M.R. Carter Index Insurance Quality
(Unavoidable) Quality Shortfalls of Index Insurance
M.R. Carter Index Insurance Quality
The “Will I get Paid if I Have a Loss” Measure
M.R. Carter Index Insurance Quality
The “How Much Will I get Paid” or “Value for Money”Measure
M.R. Carter Index Insurance Quality
Integral or Area Measures of Contract Quality
These two visual representations suggest simple integral orarea measures that tell us how far a contract i falls short of anidealized contract p that never fails:
For the probability measure:
Qprobi =
τ∫0
(Probi (y)−1)dy −∞∫
τ
Probi (y)dy
For the value for money measure:
Qvfmi =
∞∫0
[E (Ii |y)
πi− Ip(y)
πp
]dy
Comparing across contracts, the better measure might be thatwhich has the higher quality scoreBut let’s dig a little deeper as the above values all shortfallsfrom the perfect contract the same
M.R. Carter Index Insurance Quality
Comparing Alternative Contracts: Simple Case
Clear that contract B is closer to perfect contract than A, andB always does better for the farmer when there are losses (&transfers less money from bad to good states of the world)
M.R. Carter Index Insurance Quality
Comparing Alternative Contracts: Complex Case
Not obvious whether C is better than B–could go either wayIntegral measures will give an answer, but maybe not the rightone
M.R. Carter Index Insurance Quality
Comparing Alternative Contracts: Complex Case
Two aspects make the comparison of B and C difficult:
Contract C does better for the loss events that happen mostfrequently; but,Contract B does better when the farmer most needs help (i.e.,when the scarcity value of money is highest for the farmer)
Economics’ standard expected utility approach captures bothof these by providing a single measure that reflects:
The probability that an insurance failure happensThe ’marginal utility’ of money when failure happens
Expected utility allows calculation of a ’reservation’ pricedefined as the maximum amount an individual could pay for acontract without making herself worse offA safe minimum quality standard might be: Reservation Price> Market PriceLet’s look at a real example from Tanzanian rice cultivation
M.R. Carter Index Insurance Quality
Measuring Insurance Quality for Rice farmers in NorthernTanzania
M.R. Carter Index Insurance Quality
Village-level Area Yield vs. Optimized Satellite-basedContract
For each small area (“village”), we collected 10 years ofretrospective data on yieldsBest satellite predictor of village yields proved to be based on’Gross Primary Production’ (based on EVI, FPAR & LAI)Let’s compare this (cheap to administer) satellite based indexwith an (expensive) village-level area yield contract:
50
010
00
15
00
20
00
25
00
30
00
Pre
dic
ted
yie
ld (
kg/a
cre
)
500 1000 1500 2000 2500Actual yield (kg/acre)
Predicted vs. actual area yields in Makindube
M.R. Carter Index Insurance Quality
The “Will I get Paid” Probability Measure
Consider a contract that pays anytime either measured orsatellite predicted village yields fall below average:
Probability of receiving a payout by zone-level yields
M.R. Carter Index Insurance Quality
The Value for Money Measure
M.R. Carter Index Insurance Quality
Reservation Price Quality Measure
Actuarially fair prices for these contracts are 130 kg of riceper-hectare insuredUnrealistically, assuming no local risk sharingMinimalist Quality Standard: Reservation Price > MarketPrice of Contract
M.R. Carter Index Insurance Quality
Can We Do Better with an Audit Rule?
Can see that the satellite does well separating good from bad,but has trouble distinguishing quite bad from slightly bad
What if we followed Ruth Hill’s model & proposed an auditscheme with:
Agreed upon crop cut methodology at village levelIncentive compatible penalties to prevent unnecessary audits
M.R. Carter Index Insurance Quality
Can We Do Better with an Audit Rule?
Assume that audits only requested when predicted yields are5% below actual village area yields17% of the time audits will take place
Shadow price will be close to pure area yield insuranceData collection costs will only be 17% of those under areayield!
M.R. Carter Index Insurance Quality
Doing Better in other Agro-ecological Environments
The Tanzanian Analysis is for an irrigated crop where expect alarge proportion of risk to be idiosyncratic and uninsurable byany index insuranceWhat if we look at well designed contracts for rain-fedagriculture where covariant risk is a larger fraction of all risk?We can look at preliminary work on cotton in West Africawhere we have implemented a dual-scale trigger contract:
Village level area yield triggerBut payment only made if surrounding 3-4 villages corroboratebad conditions
M.R. Carter Index Insurance Quality
Doing Better in other Agro-ecological Environments
Let’s visually compare probability and value for moneymeasures of dual-scale contract with village-level area yieldcontract
Still need to estimate individual variation from village averageto get a shadow price (assuming no sharing)
M.R. Carter Index Insurance Quality
Conclusion: Quality Standards & Their Importance
Work here suggests a safe minimum quality standard:
Reservation Price > Market PriceNote that even if insurance is subsidized, beneficiaries wouldbe better if simply given the subsidy rather than given cheapinsurance that approximates a lottery ticket
Issues of quality assurance are very important:
Clarke & Wren-Lewis (2014) make a very compelling argumentthat without quality certification standards, markets will reachan equilibrium in which low quality contracts predominate
Much to learn about how to implement those standards, butalso how to design contracts that meet those standardsHope that the novel contract designs we have shared todaystimulate some ideas!