Index-based insurance to control vector-borne disease

.

......

Index-based insurance: a new tool to controlvector borne disease under climate change ?

Ahmadou H. Dicko@dickoah

PhD candidate in climate change economics

Ahmadou H. Dicko @dickoah Index-based insurance: a new tool to control vector borne disease under climate change ? 1 / 25

Introduction

The demand for livestock products will double in the 2050horizon (Herrero et al. 2009)

Most of the livestock products are made by smallholders indeveloping countries

These households are vulnerable to socio-economic and climaticshocksThey face many challenges on their daily activities:

▶ Climate variability and drought (Arid area)▶ Livestock diseases and mainly animal trypanosomiasis

(Sub-humid to humid zone)

(Thornton et al. 2007)


Introduction

Figure: Study area in light green (Guinea SavannahZone)


Outlines

...1 Livestock health and climate change

...2 Index-based insurance for livestock disease

...3 Preliminary results

...4 Conclusion and perspective


Livestock vector-borne disease

The risk of livestock diseases is high on sub-humid zone of WestAfrica

Most of the the livestock diseases are transmied by arthropod(vector)

Vector-borne disease are highly sensitive to climate change andvariability (Moore et al. 2012)


African Animal Trypanosomosis

African animal Trypanosomiasis (AAT) is a fatal vector bornedisease that causes serious economic losses in livestock.

Most trypasonomoses are transmied by tsetse flies.There are two ways to control the disease:

▶ control tsetse flies (insecticide, bush and game clearing, etc.)▶ treat the infected host (treat cale with drugs)

(Bouyer et al. 2006)



Figure: A cow suffering from AAT (Image from ILRI)



Figure: Tsetse flies Figure: Traps



..Trypanosomosis risk.

Tsetse density

.

Habitat suitability

.

Tsetse infection rates

.

Climate



..pastoralactivities

.....

Climatechange

.

Trypanosomosis


Index insurance in a nutshell

Based on a index (e.g rainfall) correlated to the losses

Payout are made on a agreed-upon threshold of the index

Advantages : Low transaction cost and no moral hazard andadverse selection

Disadvantages : Imperfect correlation between chosen index andlosses ( Basis risk )

Basis risk have to be as low as possible in an optimal insurancedesign


Demand for index-based insurance

Demand for index insurance is lower than expected specially from themost risk averse :

credit constraint (Giné, Townsend, and Vickery 2008; Cole et al.2012)

competition with cheaper self-insurance mechanism(Binswanger-Mkhize 2012)

Basis risk and uncertainty (Clarke 2011a)


Some known index-based livestock insurance

IBLI in Mongolia (Mahul and Skees 2007)

Combine three hedging products to cover from small tocatastophic losses

Public-private partnership (backed by World Bank)

IBLI contracts in Northen Kenya and Ethiopia (Chantaratet al. 2012) :

NDVI based index (estimated mortality) built using predictivemodeling on rich herd history data

Seasonal contracts with private insurance companies


Challenge in scaling up IBLI

Index based on remote sensing data :▶ Solve data scarcity problem▶ Build a response function that map satellite images into average

loss (morbidity, mortality, etc.)▶ Predictive modeling to build the index −→ low basis risk

Demand driven product design :▶ Insurance contract that meet an existing (rational) demand

(Clarke 2011b)▶ In some countries compliance with islamic finance


Index-based animal disease insurance

Traditional animal disease insurance are usually costly :▶ moral hazard and adverse selection▶ administrative and loss assessment cost

Traditional insurance diminish the effort of livestock owner toadopt beer management practices (bad incentive)


Index-based animal disease insurance

Our hypothesis is that only index-based insurance can (at thesame time) :

▶ hedge against the risk of livestock (vector-borne) diseases▶ encourage herders to beer manage their livestock and control

the disease

However the design of an effective index-based animal diseaseinsurance (IBADI) remains a complex task.


Steps (we choose) to design an IBADI

Risk assessment:▶ A thorough understanding of the epidemiology of the disease▶ A quantification of the risk of disease (prevalence, factors, etc.)▶ An analysis of livestock keepers risks perceptions

Design of the index:▶ Directly related to the disease (e.g predicted prevalence in

space/time)▶ Indirectly related to the disease (e.g predicted density of the

vector in space/time)

Demand driven contract based on field experiments:▶ Strike points (multiple triggers ?)▶ (Fair) Premium price▶ Delivery channel (microfinance institute, veterinary offices, etc.)▶ Create a risk management portfolio (with early warning system )


Risk assessment framework

The chosen index is known by epidemiologist as entomologicalinnoculation rate or tsetse challenge. It is the product of:

Suitability index▶ Using species distribution models▶ 10 years average of monthly MODIS and precipation data (NOAA

RFE)

Tsetse density▶ Spatio-temporal model of tsetse density (vector abundance)▶ Time series of monthly satellite images (temperature and NDVI)

Tsetse infection rates▶ Generalized linear mixed models▶ Time series of monthly satellite images (mainly temperature)


Remote sensing data

Cum. prec

100 120 140 160 180 200

NDVI

−0.2 0.0 0.2 0.4 0.6 0.8

MIR

0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50

LST Day

24 26 28 30 32 34 36 38 40 42

LST night

20 21 22 23 24 25

Vector presence

●●●●●●●●●●●●●●●●●●●●●●● ●●● ●●● ● ●●● ●●●●●● ●●● ● ●●● ● ●● ●● ●●●●● ●●●●●● ●● ●●●●●● ●●●●●●●● ●●●●●●●● ●●●● ●●●● ●●●●●●● ●● ● ●●●●●●●●●●●●●●●●●●● ●● ●●●●●●●●●●●●● ●●●●●● ●●● ●●●●●●● ●●●●● ●●● ● ●●●●●●●● ●●●● ●● ●● ●●●●● ● ●●● ●●●● ●● ●● ●●●● ●●● ●●● ● ● ●● ●●●● ●●●●●● ●●● ●●●● ●●● ●●● ●●●●● ●● ●●●●● ●● ●●●●● ●●●●●●●● ●●● ●●● ●● ●● ● ●●●●● ●●● ●●●● ●● ●●● ●● ●●●●

●●●●●

● ●●●●●

●●●●●●●●●●●●●● ●●● ●●●●●●●●●●●●●●●●●●●● ●

●●●●●●

0.6 0.8 1.0 1.2 1.4


Habitat Suitability

0.0 0.2 0.4 0.6 0.8 1.0

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●

●●●●●●●

●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●

●●●●●●●●●●●●●●●●

●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●

●●●●●●

●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●

●●●●●●●●●

●●●●●

●●●●●●●●●●●

●●●●●●●●●●●●●

●●●●●

●●●●●●●●●●

●●●●●●●●

●●●●●●

●●●●●●

●●●●●●●●●●●●

●●●●

●●●●●●

●●●●●●●●●●●●●●●●●

●●●●●

●●●●●●

●●●●●●●

●●●●●●●●●

●●●●●●

●●●●●

●●●●●●●

●●●●●●

●●●●●

●●●●●●●●

●●●●●●●

●●●●●●●●

●●●●

●●●●

●●●●●

●●●●●●●●●●●

●●●●

●●●●

●●●●●

●●●●

●●●●●●●●

●●●●

●●●●●●

●●●

●●●●

●●●●●●

●●●●●●●●●●●●

●●●

●●

●●●●●●●

●●●

●●●

●●●●●●●●

●●●●●●●●

●●●●●●

●●●●●●

●●●●●

●●●●●●

●●●●●●

●●●

●●●●

●●●

●●●●●●●

●●●

●●●●

●●●●●●●●●

●●●●●●●●

●●●●

●●●

●●●●

●●●●●●

●●●

●●●●●●●●●

●●

●●●

●●●

●●●●●

●●●●

●●●●●●●●●●●●●●●

●●●●●●●●

●●●●●●●●●●●●●●

●●●●●●●●●●●●●●

●●●●

●●●●

●●●●●●

●●●●●●●●

●●●●

●●●●●●●

●●●●●●●●●●●

●●●●●●●●●●

●●●●●

●●

●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●

●●●●●●●●●●●●

●●●●●●●

●●●●●●●●

●●●●●●●●●

●●●●

●●●●

●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●

●●●●

●●●●●●●●●●●●●●●●●●

●●

●●●●●●

●●●●●●●●●●●●●●●●

●●●

●●●●●●●●●●●●●●●●●●●●●

●●●●●●●

●●●●●●

●●●●●●●●

●●●●●●●●●●●●●●●●●●●

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

False positive rate

True

pos

itive

rat

e

AUC = 0.91


Tsetse (apparent) densityJan 2010 prediction

−0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0

2.2

2.3

2.4

2.5

2002 2004 2006 2008 2010 2012 2014date


Predicted indexMarch 2013

0 10 20 30 40 50

●

● ●

● ●

●●

●●

●

●

● ●●

●

● ●●

● ●

● ●●

●

●

●

●● ●

● ● ●

●

●

●●

●

● ●● ●

● ●●

● ● ●

●●

● ●

●● ● ●

●

●

●

●●

● ●●

●● ● ●

●

●

●

● ●●

●●

● ● ●●

●

● ●●

●

●● ●

●

●●

●

●

●

●

●●

●

●

●●

● ● ●

● ●●

●

●● ●

●

●●

●●

●

●

●

●●

●

●

●

●● ● ●

●

● ●

●

●

0

50

100

150

200

2004 2006 2008 2010 2012 2014dateP

redi

cted

tset

se c

halle

nge


Conclusion and perspective

Targeting other major climate sensitive diseases (create a singleproduct)

Coupling index-based product to a disease early warning systemA need to integrate a feeding component to the index

▶ Animal stressed by hunger are more sensitive to diseases

Climate change simulations▶ MIROC C5 and CSIRO MK3 will be used to simulate the built

index▶ Spot area with high risk variability (in space and time)

Field experiments and pricing in Burkina Faso and Senegal


References..Binswanger-Mkhize, Hans P (2012). “Is there too much hype about index-based agricultural insurance?” In: Journal ofDevelopment Studies 48.2 (2012), pp. 187–200.

Bouyer, J. et al. (2006). “Mapping African Animal Trypanosomosis risk from the sky”. In: Veterinary research 37.5 (2006),pp. 633–645.

Chantarat, Sommarat et al. (2012). “Designing Index-Based Livestock Insurance for Managing Asset Risk in Northern Kenya”.In: Journal of Risk and Insurance (2012).

Clarke, Daniel J (2011a). A theory of rational demand for index insurance. Department of Economics, University of Oxford, 2011.– (2011b). “Insurance design for developing countries”. PhD thesis. Oxford University, 2011.Cole, Shawn et al. (2012). “Barriers to household risk management: evidence from India”. In: Harvard Business School FinanceWorking Paper 09-116 (2012), pp. 104–35.

Giné, Xavier, Robert Townsend, and James Vickery (2008). “Paerns of rainfall insurance participation in rural India”. In: TheWorld Bank Economic Review 22.3 (2008), pp. 539–566.

Herrero, Mario et al. (2009). “Livestock, livelihoods and the environment: understanding the trade-offs”. In: Current Opinion inEnvironmental Sustainability 1.2 (2009), pp. 111–120.

Mahul, Olivier and Jerry Skees (2007). “Managing agricultural risk at the country level: The case of index-based livestockinsurance in Mongolia”. In: World Bank Policy Research Working Paper 4325 (2007).

Moore, S. et al. (2012). “Predicting the effect of climate change on African trypanosomiasis: integrating epidemiology withparasite and vector biology”. In: Journal of The Royal Society Interface 9.70 (2012), pp. 817–830.

Thornton, P.K. et al. (2007). “Coping strategies in livestock-dependent households in East and southern Africa: a synthesis offour case studies”. In: Human Ecology 35.4 (2007), pp. 461–476.


Thank you for your aention


Index-based insurance to control vector-borne disease

Education