POLICY IMPLICATIONS OF PRICE DYNAMICS AND THE WAY FORWARD - Food Security, Food Prices and Climate Variability - page 186

Index insurance cannot bring the needed widespread improvements in yields and ultimately
food production without a scalable, accurate, weather-related index that can be used to
estimate the probability that a region will experience a drought and to trigger payouts across
multiple regions and agro-ecosystems. Currently, weather insurance programs are based on a
single remote sensing dataset, such as the NOAA Rainfall Estimate product, that has not been
validated for the agricultural regions in which it is working (Helmuth et al ., 2009). Reliable
and accurate remote sensing observations can ensure that index insurance is based on scalable
datasets, to enable their implementation across large areas without weather stations. Using
satellite-derived information on moisture availability from satellites promises to enable a much
wider expansion of index insurance beyond Ethiopia and Kenya where the programs have
developed (Helmuth et al ., 2009).
Estimates of moisture conditions can inform, validate and assist in pricing of index insur-
ance. The probability of a drought will be estimated using precipitation and evapotranspira-
tion information for each community, and this information will be integrated into index
insurance pilots. There are currently nearly 40 pilot projects introducing index insurance to
over a dozen developing countries (Hazell et al ., 2010). The economists at Columbia Univer-
sity's International Research Institute for Climate and Society (IRI) have been working to
scale-up some of these pilot projects to allow for implementation in many more communities
at once (e.g., see Oxfam, 2011). There is a new pilot project in Senegal that will help develop
replicable methodology using previously successful projects that are sustainable and have the
potential to expand into new areas much more rapidly and efficiently than previous pilots.
Improving the characterization of the nature, magnitude and probability of drought
return frequency for Africa, and by helping to identify trends in the occurrence of these
events, remote sensing datasets can provide the quantitative underpinnings for improved,
more reliable index insurance pricing (Brown et al ., 2011). The final price (or premium) of
an index insurance product typically reflects the cost of both: (1) insurance protection for
climate risk; and (2) insurance protection for information uncertainties about climate risks.
When insurance providers lack sufficient or reliable climate data, they are unable to accu-
rately price their products, forcing them to price contracts conservatively (meaning higher
premiums) to ensure that they can be honored. The longer climate records and better char-
acterizations of drought recurrence have the potential to improve the accuracy of index
insurance pricing and therefore improve access to, and the sustainability of, climate insur-
ance for poor farmers.
Climate data for index insurance
Climate data that are useful for triggering index insurance must be highly correlated to actual
yields in farmers' fields. In order to verify that a particular environmental or satellite data
index can capture accurately production, we need to have accurate, high resolution and crop
specific yield information. Yield information can be difficult to collect and is often not trans-
ferable across countries or regions due to differences in the type of crop or the markets for the
food produced. This kind of information must be geolocated to a specific location and have
explicit information about the area where the production occurred. For many countries that
are food insecure, calculating yield and even tracking the total amount of food produce is
enormously challenging given the large number of very small producers, many of whom
never market their produce. They grow food and consume it themselves or sell only to their