Indexing the soil 35
that index insurance operates in isolation from other financial products
and, as explained above, only one-year time frames are considered for the
payout models. The aim is to predict payout ratios accurately, and this ultimately constitutes the socioeconomic logic of the product. By transforming
environmental data into a standardised assemblage of relations, the models construct a form of risk that is approximated by simulating the model’s
various dimensions as stochastic elements (RM: 154–55). In other words,
‘risk’ here can be claimed to be ‘abstract,’ as it has nothing to do with, say,
the concrete loss of the harvest; rather, it concerns an abstract value derived
from the model.
Even if RM stipulates payout ratio simulation as the most suitable
method for designing index insurance products, it does consider two other
approaches for modelling environmental risks, perhaps for didactic reasons.
The first is to model the index so that environmental data, such as rainfall,
not only serves as a submodel for payout ratio simulations but is also used to
simulate dynamic changes in the environment. This approach would make it
possible to model sequential relationships between different years and areas
and therefore include weather-related changes in the process (RM: 270).
The second alternative is to model the weather itself. This approach would
require a more holistic weather system model, where the product’s trigger
levels would be formulated on the basis of simulated hazard data instead of
historical hazard data. This signifies a much more comprehensive alternative
where even the inclusion of multi-year weather cycles, such as El Niño, could
be used in the design of an index insurance product (RM: 271). Considering
these alternative approaches, the authors of the guide weigh better understanding of the world’s complexity in relation to instrumental needs:
In many cases, analysts start off thinking that they need very ‘realistic’ models to capture the behavior of the real world. However, in our
experience it is best to start with the simplest model that fulfills all the
needed functions and uses valid assumptions. Only then should analysts add more complexity as necessity dictates.
(RM: 271–22)
Hence, the alternative methods are not recommended for index insurance
product development. With this, the guide draws a conclusion for the mapping process, basing its ecological risk modelling recommendations on
financial performance. The perception of the soil and the weather as static
systems is deemed essential for achieving precision and coherence in the
pricing of risks. Thus, index insurance, as advanced by RM, disregards both
real property damages and environmental changes in its technical definition
of risks. The most important consequence of this move is that climate change
is pushed outside of the range of objects that the models can recognise.
In transforming the soil into an object of governance, index insurance is
treating the policy’s underlying environmental uncertainties as analogous
that index insurance operates in isolation from other financial products
and, as explained above, only one-year time frames are considered for the
payout models. The aim is to predict payout ratios accurately, and this ultimately constitutes the socioeconomic logic of the product. By transforming
environmental data into a standardised assemblage of relations, the models construct a form of risk that is approximated by simulating the model’s
various dimensions as stochastic elements (RM: 154–55). In other words,
‘risk’ here can be claimed to be ‘abstract,’ as it has nothing to do with, say,
the concrete loss of the harvest; rather, it concerns an abstract value derived
from the model.
Even if RM stipulates payout ratio simulation as the most suitable
method for designing index insurance products, it does consider two other
approaches for modelling environmental risks, perhaps for didactic reasons.
The first is to model the index so that environmental data, such as rainfall,
not only serves as a submodel for payout ratio simulations but is also used to
simulate dynamic changes in the environment. This approach would make it
possible to model sequential relationships between different years and areas
and therefore include weather-related changes in the process (RM: 270).
The second alternative is to model the weather itself. This approach would
require a more holistic weather system model, where the product’s trigger
levels would be formulated on the basis of simulated hazard data instead of
historical hazard data. This signifies a much more comprehensive alternative
where even the inclusion of multi-year weather cycles, such as El Niño, could
be used in the design of an index insurance product (RM: 271). Considering
these alternative approaches, the authors of the guide weigh better understanding of the world’s complexity in relation to instrumental needs:
In many cases, analysts start off thinking that they need very ‘realistic’ models to capture the behavior of the real world. However, in our
experience it is best to start with the simplest model that fulfills all the
needed functions and uses valid assumptions. Only then should analysts add more complexity as necessity dictates.
(RM: 271–22)
Hence, the alternative methods are not recommended for index insurance
product development. With this, the guide draws a conclusion for the mapping process, basing its ecological risk modelling recommendations on
financial performance. The perception of the soil and the weather as static
systems is deemed essential for achieving precision and coherence in the
pricing of risks. Thus, index insurance, as advanced by RM, disregards both
real property damages and environmental changes in its technical definition
of risks. The most important consequence of this move is that climate change
is pushed outside of the range of objects that the models can recognise.
In transforming the soil into an object of governance, index insurance is
treating the policy’s underlying environmental uncertainties as analogous