As we discussed in an earlier entry, a composite scheme that combines an NDVI and a Weather Index represents the latest development, and possibly the future, of Agricultural Insurance. The reason this new scheme is considered so promising is that it achieves greater yield accuracy alongside a similar ease of implementation. This can be seen in the comparative chart below.

Source: Patankar (2009)

(more…)

In the face of the shortcomings of weather index and yield based insurance, a new solution based on remote sensing applications has been proposed to provide an accurate and effective Insurance program. Currently being researched by the CIRM, remote sensing refers to the use of Imagery from Satellites orbiting space to estimate production yields in landholdings through a variety of methods. These include estimating the rainfall in a given area by measuring the temperature of storm clouds, a technique known as Thermal Infra Red (TIR), and measuring the foliage, or ‘greeness’, of a specific area through measuring the wavelength of radiations absorbed by leaves, creating what is known as the Normalized Difference Vegetative Index (NDVI). The NDVI and TIR help solve the issue of a lack of ground-based weather collection infrastructure, as they can be remotely calculated from Satellites. Thus, they bypass one of the biggest drawbacks of a traditional Weather Index based scheme. Moreover, an NDVI based production yield has been shown to share a far greater correlation to actual production yields, leading to more accurate estimates and therefore a more effective insurance product. Finally, NDVI schemes are also extremely advantageous because they are considered faster to estimate, easier to scale, and cheaper to implement than traditional weather index schemes.

An example of satellite imagery measuring a plant’s ‘greeness’

Source: Upadhyay Gargi, Ray S S, Panigrahy Sushma (2008)

Yet, just as with all the other Insurance solutions detailed before, the NDVI based schemes also suffer from certain issues. One of the primary concerns with an NDVI solution is that a large amount of agricultural land mapping must be carried out for the NDVI scheme to be put into place. This is because each landholding must be assigned into a specific grid, based on latitude and longitude, as the estimates of production within each production grid are used as proxies for the yields of the landholdings within the grid. Such information will most likely not be known by the policyholder, and will have to be attained through comprehensive land mapping, which may be a time consuming process. Another important concern is that an NDVI estimate may be prone to moral hazards, with policyholders potentially damaging their landholdings to lower production estimates and avail higher payouts. However, this problem can seemingly be overcome by combining the NDVI with the Weather Index to form a new composite index, seeing as moral hazard does not affect Weather Indexes as much. We shall discuss this new form of insurance in a future entry.

NDVI Insurance schemes have already been implemented throughout the world, as shown in the below table.

Source: Patankar (2009)

In regards to developing countries, India and nations in Western Africa have led the way in the implementation of NDVI schemes. In fact, a new set of NDVI insurance schemes in Western Africa is currently being researched and proposed by a team led by Michael Carter and Rachid Laajaj. This is not entirely surprising as India and Western Africa are particularly well suited to the requirements of NDVI. Both areas do not have extensive cloud cover, an impediment to NDVI measurement, as they have large, vast plains. Moreover, both India and Western Africa tend to grow the same crop, usually paddy, across large land areas, making NDVI measurement all the more accurate. In contrast, countries such as Sri Lanka, with small plot areas and high levels of cloud cover, are unsuited to NDVI Insurance schemes due to the difficulty of NDVI measurement in the given areas. This then represents yet another shortcoming of NDVI Insurance – it can only be implemented in locations that meet a very specific set of agricultural and geographical criteria. Outside of these areas, it is an ostensibly impractical insurance solution. However, the fact remains that in suitable locations, NDVI Insurance schemes carry many advantage and can possibly be used in conjunction with older schemes to create viable Insurance solutions. We shall explore this idea further in the next entry.

One of the biggest problems that farmers face in crop production is the volatility of weather patterns. A sudden drought, be it due to climate change or the whims and fancies of the gods, can lead to huge losses in crop yield and therefore income. As such, farmers require protection against the sheer unpredictability of weather – the so-called random acts of God that we are yet to fully predict or understand. The solution to this, many feel, lies in Weather Index Insurance.

Weather Index based Insurance is one of the new forms of Agricultural Insurance in India that holds much promise for overcoming the flaws of previous yield based schemes. This particular form of Insurance uses a specific parameter of weather in a given area, usually rainfall, as a proxy for the crop yield of the farmer’s landholding. Taking rainfall as the parameter, if net rainfall in a given season varies from an established level, the farmers will be compensated based on variance of the present rainfall from the established level of required rainfall.

This method holds many important advantages to area yield based schemes. First of all, it is far easier to calculate the indemnity payout in weather index based schemes, as it doesn’t necessitate actual assessment of yields in multiple tiny plots in remote rural areas. All that is required is to make a payouts based on the variation of the rainfall in a given area with an accepted standard level. This accepted standard level is calculated by comparing historical rainfall levels with crop yields and assessing the correlation between the two variables. Now, the advantage behind this form of insurance is that it is extremely transparent as the information required can be easily publicized and the method for calculating payouts is simple to understand. Moreover, as this index doesn’t require actual field based evaluation, the insurer is subject to far less information asymmetry leading to a reduction of moral hazard and adverse selection. These are particularly important considerations as they allow the insurers to dramatically reduce the premiums on the insurance products.

Finally, Weather Index Based Insurance has a much lower administrative cost leading to greater efficiency for both the insurer and the insured. The insured particularly benefits from this as the lower administrative costs and lower time lag allow for quicker claims payouts, which can be hugely important in times of severe income shocks. Thus, Weather Index Insurance has largely been seen as a more effective form of insurance than what has been provided in the past.

However, even this form of Insurance suffers from many important drawbacks that we must consider. To begin with, the weather in a given area may not always be representative of the weather in the entire area due to fluctuations in weather over small areas. This could lead to policyholders receiving inadequate compensations for their losses. Another important consideration is that the weather collection infrastructure in India is currently quite poor, leading to a lack of reliable data in remote rural areas. This can cause significant inaccuracies in production estimates due to the potentially significant proximity between a weather station and a given landholding.

An example of a simple weather station

Furthermore, historical weather data, needed to create models to estimate production yields, are also hard to come across, and when they are present, the costs of acquiring them are very high. Such data collection issues lead to imprecise payouts, with a variation of up to 40% between payouts and actual losses. In addition, the uncertainties caused by such data force insurers to higher premiums to compensate for what is known as ‘uncertainty loading’. Therefore, we find that a Weather Index based Insurance solution may not be the ideal solution either for providing Agricultural Insurance for farmers in rural areas. In the face of such problems, another new solution has been proposed which we shall proceed to see in the next blog entry.

Despite the many advancements made in different sectors of the economy in recent years, the field of Agriculture remains immensely important, especially in regards to developing nations. Even today, Agriculture provides employment for a staggering 60% of the Indian labor force. Yet, the field in itself is by no means a particularly desirable one to work in, especially for small and marginal farmers. Farmers are plagued by numerous, unpredictable risks that can lead to sudden income shocks, which, if unplanned for, can lead to disastrous consequences. Risks pertaining to the income from agriculture practice can be  largely categorized into:

•    Productions risks – Risks caused by variations in the yield or output

•    Price risks – Risks caused by sudden fluctuations in the price level of crops or inputs

A variety of risk management and reduction strategies are practiced by farmers to tackle these risks, including fragmenting plots, using lower cost inputs, and investing in lower cost, lower yield crops. However, these very strategies can often come in the way of escaping poverty, as they are inherently low-risk, low-return activities. This can be particularly detrimental to lower income families, as most of the income will be spent towards a minimum requirement of food, leaving little in the way for capital required for future investment. As such, large risks force farmers into lifestyles that are largely insufficient for making the transition out of poverty.

The obvious solution for this is a sustainable and effective agricultural insurance program, and the Government of India has provided such solutions since 1972. Starting with the CCIS (Comprehensive Crop Insurance Scheme), the state has striven to create new insurance solutions, creating the NAIS (National Agricultural Insurance Scheme) in 1999, and then the FIIS (Farm Income Insurance Scheme) in 2003. However, these programs have drawn large criticism for being unsustainably expensive and inefficient due in no small measure to the fact that they are based on the ‘area yield index’ method. In ‘area yield index’ insurance contracts, insurers make indemnity payouts to individual farmers based on the crop yield performance of a specific plot of land through a process known as a Crop-Cutting Experiment (CCE), which is typically the responsibility of a Government Department. This given plot of land essentially serves as a representative proxy for measuring the agricultural yield of a region for the specific crop variety. However, this method of assessment suffers from many inherent problems. First of all, the chosen area of land is not always entirely representative, due to the fact that crop yields can easily fluctuate within nearby areas due to differences in land fertility as well as localized calamities. In such cases, farmers may not receive adequate compensation for their losses leading to a basic failure of the insurance mechanism. Moreover, the method by which these yields are estimated is inherently inefficient, as it requires substantial time and manpower to undertake CCE’s in remote rural locations.

An employee from agriculture department making sample plots with the help of measuring tape in a wheat field in Nagpur District, of Maharashtra, India, 2009(copyright, CIRM, 2009)

Due to this, farmers are, at times, forced to wait up to an entire year before they receive the payouts for their losses. Furthermore, area yield measurements are also prone to Moral Hazards(MH) and Adverse Selections (AS) as a large asymmetry of information exists since the farmer always knows much more about the composition of his land(therefore, probability of  yield loss) than the insurer. This leads to higher premiums in order to compensate for these effects, which in turn leads to large subsidies in order to ensure the affordability of such insurance schemes. However, even with these subsidies, the schemes suffer from huge losses due to the variety of problems already detailed. Below, we can see the high loss ratios of various similar yield based agricultural insurance schemes implemented in different nations.

Source: Skees et al (1999)

Such large losses, coupled with extensive subsidies, make yield based schemes largely unsustainable. Clearly, new, adaptive, solutions are needed in order to ensure the viability and profitability of Agricultural Insurance in the long run. We shall look at one such solution in the next entry in the CIRM blog.