Governing through the market and state: livestock insurance

As an alternative to standard indemnity-based insurance founded on assessments based on the statistical probabilities of loss from past experience, index-based parametric insurance is a stream-lined option whereby policyholders are compensated when an index related to expected losses falls below an agreed threshold. In this case, the index is a decline in forage levels that are expected to result in livestock mortalities (Taye et al. 2019, Chantarat et al. 2013, Mude et al. 2012). This can be assessed remotely by employing satellite imagery of grasslands and assessments are made during rainy periods through the Normalised Difference Vegetation Index (NDVI). When the predicted forage level in a particular index area goes below a certain level, pay-outs are triggered. The paying of insurance cash is, in turn, expected to result in pastoralists selling their animals and reducing pressure on the rangelands, later being able to repurchase animals once the drought has passed. Levels of pay-out are geared to expected frequency and severity of drought according to climate models, and the level of subscription to the insurance scheme (Zewdie et al. 2020, Johnson et al., 2019).

The model for index-based insurance systems is settled agriculture where fixed fields and pay-outs related to particular crops are more easily administered. Even here questions have been raised as to whether such insurance approaches crowd out local drought responses (Carter et al. 2017), although others claim this does not happen in livestock insurance systems, even if uptake remains low (Takahashi et al. 2019, 2014). Index-based livestock approaches have been promoted in Ethiopia as a private, commercial operation underwritten by Oromia Insurance SC and in Kenya as a state-funded social protection approach by the Kenyan government under the Hunger Safety Net Programme and also as a voluntary product for those who can afford it. Both programmes are supported technically by the International Livestock Research Institute (ILRI) and other development agencies.

Each insurance system has its particularities, but the principles are the same – an index-linked asset protection insurance product. In both settings, there are however questions about coverage, accuracy and leakage, notably where registrations in the Kenyan system can be abused2 https://rethink.earth/how-kenyas-herders-got-their-livestock-insured/. During the piloting phases of the subsidised scheme in Kenya, contradictory targeting of beneficiaries occurred. Despite the claim to support the most vulnerable pastoralists, a minimum of five cattle is required to be accepted into the programme. There are also questions raised as to whether a spatially distinct assessment of drought risk is appropriate given the high levels of mobility of pastoral populations in a drought period. Mobility in the pastoral system is highly dependent on the availability of resources (pasture and water) during seasonal droughts and technologies like index insurance are required to capture such patterns to minimise the unintended impact on policyholders.

In both cases, it is the insured that hold the ‘basis risk’, the difference between the measures from the models and the actual conditions on the ground. This can be quite large, as actual losses are not directly assessed (Johnson 2020, Johnson et al. 2023). The functioning of the system relies on trust between those insured and paying premiums and those insuring and offering pay-outs based on models with often quite major assumptions. Beyond the technical efficacy of the system (e.g. Jensen et al. 2019), wider questions are raised about how uncertain futures are co-constructed with politics through such an individualised market-based mechanism, and who wins and who loses.

For the originators of the system, the financialization of risk through insurance provides an arm’s length approach to responding to disasters, and providing protection for livestock keepers. Governing risk through the market creates new forms of politics and accountabilities, however. Market subjects are formed, linked to a complex apparatus through insurance forms and commitments through payment of premiums. The pay-outs are dependent on a remote system of satellite monitoring and modelling, requiring trust in those involved. When pay-outs do not occur because the NDVI did not drop below a threshold, then concerns are raised by pastoralists who may have experienced drought conditions. Very often concessions are made in order to ensure that those who adopt the insurance remain enlisted. Others drop out as they become dismayed at the results. Still others never take up insurance as they do not trust the system and prefer to use other mechanisms to offset risks. In other words, different people get enlisted in a financialised, market system for different reasons. For the commercial brokers, profit can be made from risk through selling insurance; for the state and development agencies the market resolves the challenges of delivery of social protection; and for pastoralists, insurance potentially allows the opportunity to extricate oneself from complex social negotiations around risk and uncertainty, operating independently as a market actor.

In Borana, Ethiopia, for example, GG, a 30 year-old pastoralist, who had in May 2020 adopted the insurance product commented on his rationale:

However, livestock insurance is not for everyone, as market relations at the centre of insurance systems must be sustained by trust. Another informant who initially purchased insurance and then decided to abandon it, observed:

Only certain pastoralists therefore take up the insurance offered and become ‘market subjects’ in a financialised system. Data show how it is those with larger herds (usually older men), and those who have a more diversified livelihood portfolio, and so can spread risks across livestock production, agriculture and trading, who adopt the insurance. Larger herd owners may take out insurance only for a proportion of the herd or flock, focusing on the more vulnerable animals (cattle rather than goats) and those whose loss would be most catastrophic (female breeding animals). One pastoralist explained:

Insurance coverage is therefore uneven, varying by wealth, gender, age, and location (Fisher et al. 2019, Matsuda et al. 2019, Bageant and Barrett 2017) and is usually seen as part of a wider set of responses to risk and uncertainty. Incorporation into a financialised system is therefore only partial and highly dependent on context.

As already noted, the commercial purchase of insurance is currently dominated by older, male, wealthy pastoralists. However, according to sales data from Oromia Insurance SC (2020) in Ethiopia, an increasing number of women are purchasing insurance, relating to their special rights over valuable lactating animals (Bageant and Barrett 2017). Meanwhile, richer men are increasing both the number of animals insured and the type, shifting from smaller to larger, high-value animals. Furthermore, areas that are characterised by high sales of insurance are frequently areas where there is a high prevalence of crop production, and insurance is seen a mechanism for diversification. By contrast, in those areas where extensive pastoralism dominates, insurance is directed to the protection of high-value animals in larger herds.

Thus, in the privatised system in Ethiopia, livestock insurance does not provide universal protection, and is largely focused on the already asset-rich or those with alternative options to spread risk. Financialising risk in this way potentially increases inequality as only certain groups are favoured, while others must make use of other forms of risk management. Insurance thus creates a particular local politics, allowing some to offset the risks of drought and so accumulate across diverse livelihood portfolios. This in turn results in an accentuation of difference within local communities, thereby pulling wealthier pastoralists away from local solidarity networks (Taye 2023, Johnson et al. 2023).

By contrast, in Kenya, with a government-subsidised system under the Kenya Livestock Insurance Programme (KLIP), there are different dynamics at play. Here the state pays the premium and everyone is effectively enlisted. Pay-outs therefore are supposed to be universal, although discrepancies arise when some get registered and others do not. Indeed, it is less engagement with an individualised market system as in Ethiopia, but more connections via patronage and other linkages that allow pay-outs to be received. The phone numbers registered in the scheme include many who are based in towns or in other parts of the country and may have little connection to pastoral areas affected (Johnson et al. 2019).3 https://www.rapidtransition.org/stories/tackling-drought-in-kenya-livestock-insurance-policy-to-help-pastoralists-beat-climate-change/ Bureaucratic failures in registration combined with corruption result again in a differentiated response (Janzen et al. 2016, Johnson et al. 2019, 2023). From the onset, KLIP was explicitly political, with the two major contending political parties in Kenya promising insurance for pastoralists during their election campaigns (Johnson et al. 2019). Those able to do so cash in on the opportunities offered by the state protection scheme, while others are excluded; again, as in the market-based system, very often poorer women and young people. One pastoralist from Isiolo in Kenya complained about the scheme:

The financialization of risk through insurance – delivered through the market or by the state – focuses on individuals subject to specified hazards, in this case drought. Broader uncertainties, including situations where risks interact (such as floods, a locust invasion, and a human pandemic), are excluded. The focus is on a singular, identifiable, and measurable risk (linked in this case to a generalizable index) where the future can be priced and managed. This excludes broader engagement with uncertainties, where the future is unknown and unmeasurable. As one pastoralist from Merti, Isiolo explained, multiple uncertainties impinge on herding practices:

The individualised risk-focused approach to insurance generates a local political economy with some included and others excluded. A focus on individual risks and benefits through insurance means that more collective ways of sharing risk and redistributing benefits are less visible and less promoted. The consequence is that insurance can act to accentuate already-existing differences in a pastoral setting, with consequences for gender, economic class and age-based differentiation, resulting in selective accumulation by some.