On August 20, 2015 a strong hailstorm hit Georgia, devastating crops and infrastructure in eastern Kakheti. In Kvareli alone, the hailstorm destroyed about 1,300ha of Saperavi and 1,000ha of Rkatsiteli grapes, affecting more than 500 families. This was only one in a string of natural disasters striking Georgian farmers in recent years. One of the worst calamities occurred in July 2012, when heavy rain, strong winds, hail and floods damaged thousands of hectares of arable land in Kakheti, ripping roofs and destroying vital infrastructure.
While natural disasters cannot be prevented, farmers can use technological solutions, such as hail nets or irrigation, to minimize crop losses. Not to put all eggs in one basket, they may choose to hedge against localized weather-related losses by diversifying crops and growing locations. While coming at a cost, such risk mitigation measures would help reduce income variability. Finally, crop insurance could be used as a complementary risk-mitigation tool in case of extreme weather events such as early freezes, floods, droughts and hurricanes.
Yet, until 2014, agricultural insurance was almost unheard of in Georgia. Instead of buying insurance, the vast majority of Georgian farmers preferred to play Russian roulette with Mother Nature. They did so being chronically short on liquidity, underestimating risks and … counting on a government bailout.
In 2014, the Georgian government started piloting a heavily subsidized crop insurance scheme in the hope of shifting the market to a new “equilibrium”. Almost 21,000 policies were “sold” in that year at a symbolic price equal to about 6% of the actual cost (the government covered the remaining 94%, on average). Yet, one year later the number of policies sold plummeted to less than 3,500, as soon as the government reduced the amount of subsidy to 55% (see Table 1 below).
Table 1. Agricultural Insurance Pilots, 2014 vs. 2015
There were also other reasons for the dramatic drop in the number of policies sold in 2015. Not only was the subsidy reduced from 94 to 55%, but also it was made available only to smallholders (this also explains the even larger drop in the amount of land insured). And, of course, some more insurance policies may be sold until the end of 2015. Nevertheless, it is already clear that the pilot has failed to produce sustainable results.
Two questions must be asked: why more than 10mln GEL in taxpayers’ money spent on subsidies failed to jump-start the agricultural insurance market? And, can the Georgian government do any better?
FARMERS AS BAREFOOT HEDGE-FUND MANAGERS
In their 2011 bestseller Poor Economics, two MIT professors Esther Duflo and Abhijit Banerjee argue that managing a poor rural household is akin to the responsibility of running a hedge fund, yet doing so in the absence of modern risk mitigation instruments, such as insurance, options, futures, etc. Moreover, while hedge fund managers (or people in the middle of the income ladder) can afford small negative income shocks, for poor households even a small drop in income would require cuts in basic food, housing, health and education expenditures.
According to Duflo and Banerjee, to manage risk, the poor engage in primitive hedging strategies that prevent them from specializing and achieving productivity gains. Examples of such strategies are: mutual help within extended families, plot and crop diversification (thus forfeiting any gains from economies of scale in production), and abstention from costly investment in (risky) innovations such as higher yield hybrid seeds. While essential in the presence of severe downside risks (e.g. related to weather events), these strategies are costly and keep the farmers entrapped in poverty.
Agricultural insurance could relieve poor farmers from the need to engage in primitive hedging and in this way help them to specialize in more productive activities. This being the case, the poor could be expected to flock to agricultural insurance, whenever available. Yet, as Banerjee and Duflo point out, the opposite is true. According to Robert Townsend and coauthors, when given the opportunity, only 5 to 10% of low-income Indian farmers insure themselves against drought, even though they identify precipitation as a major source of risk. According to Dean Karlan, farmers are often not willing to purchase insurance even when its price is close to zero and much lower than actuarially fair price. The natural question to ask is why there is such a low demand for insurance among poor farmers?
Banerjee and Duflo identify several factors. The first among these is what the authors call a “demand-wallah” argument, which very well applies to the Georgian situation. When hit by hailstorm and winds in July 2012 (just three months before the fateful parliamentary elections), Kakhetian farmers received around GEL 150 million in compensation, including GEL50 million in government funds and another GEL100 million from ex-Prime Minister Ivanishvili’s foundation. In other words, Kakhetian farmers may have been playing Russian Roulette with natural disasters, yet they did so with a gun loaded with blanks!
Agricultural insurance is not a cheap proposition in Georgia. At the going rates, a Georgian hazelnut grower owning 1.5ha would be asked to pay an insurance premium equal to 2.4% of his/her income, if subsidized by the government, and 6.1%, if not. It is, therefore, not surprising that given an implicit bailout guarantee by the Georgian government, farmers have no strong incentives to purchase insurance, and, even more detrimentally, mitigate risk in their farming practices. On the other side of the coin, given the meager size of the agricultural insurance pie, insurance providers have little incentive to invest in research and data analysis or come up with innovative products that are a better fit for the Georgian market.
There are, of course, many factors standing in the way of an orderly insurance market. For example, farmers may not trust insurance providers and lack a proper understanding of the insurance concept. Moreover, they may be reluctant to commit their scarce resources given their experience in dealing with unskilled sales agents and loss adjusters. But, above everything else, the Georgian government’s efforts to roll out agricultural insurance have thus far been undermined by its own (implicit) commitment to bail out uninsured farmers. Certainly in an election year!
WAY FORWARD: NUDGING GEORGIAN FARMERS TOWARDS THE USE OF INSURANCE
Despite their general unwillingness to purchase insurance, farmers might be nudged towards insuring themselves with the help of simple behavioral “tricks”. For example, in an experiment conducted in the Philippines, randomly selected participants were asked to fill in a questionnaire about their health status. When subsequently offered health insurance, those who answered the survey, were significantly more likely to buy insurance.
Farmer cooperatives may also be an important instrument of nudging farmers to insure against risks. Miles Kimball was the first in 1988 to acknowledge and model farmers’ cooperative as a self-enforcing body able to provide insurance to its members. In Georgia, the first steps have been already taken towards this end by the Georgian government and EU’s ENPARD initiative. However, more can be done.
In particular, as has been shown by Karlan et al (2013), the best results could be achieved by combining grant incentives with the requirement to acquire insurance. While true in case of individual farmers, the same policy could be particularly effective in case of farmer cooperatives seeking to specialize and innovate in a competitive market environment. At present, the vast majority of ENPARD-supported cooperatives are not insured. By providing relevant training and grant incentives, ENPARD could complement the Georgian government’s efforts to prevent the hollow Russian Roulette practice from stifling development in Georgian agriculture.
 |
The article was produced with the assistance of the European Union through its European Neighbourhood Programme for Agriculture and Rural Development, Austrian Development Cooperation, CARE Austria or CARE International in the Caucasus. The contents are the sole responsibility of the authors and can in no way be taken to reflect the views of the European Union, Austrian Development Cooperation, CARE Austria or CARE International in the Caucasus.
|
Comments
Very good piece, indeed!
One fact I would like to underline is that (it is mentioned in the blog also), on the one hand, the ministry of agriculture is supporting development of the agricultural cooperatives (the main idea of this, among many others, is to enlarge the farm sizes), but on the other hand, the pilot agricultural insurance project has changed this year, and the farmers (or farmer groups) with more than 5 ha land cannot participate in this subsidized program. Thus, these two policies (coop development and agro insurance) are not lined-up to each other, which partly might explain the absence of agricultural insurance among already functional cooperatives...
Dear Rati,
Thank you for your comment.
I agree that all the projects implemented by APMA should be in line with each other in order to maximize the overall impact created by government interventions. However given that government resource are limited and any subsidy implies diversion of those resources from one sector to another, one should be very careful while subsidizing agriculture. Subsidizing large farmers (including cooperatives) in my view might not be the best option, because large farmers have (or should have) enough resources to purchase insurance without subsidy, unlike small farmers. Plus, cooperatives already have quite strong support from both government and donor organizations. So do we need subsidize cooperatives even more given our limits?
Going back to the projects' alignment, one good example is making crop insurance compulsory for those who get cheap loans. This is done in many countries and Georgia can follow suit. This kind of combination is likely to generate better outcomes for both projects subsidized by government as well as better value for government money.
An excellent point, Salome! Both grants and cheap loans provide an opportunity to "nudge" people towards insurance.
The issue is much more complex than our comments. What I meant, is that, when MoA is implementing the different projects, it would be better if those projects are lined-up to each other (in this case, agro cooperatives run by ACDA and agro insurance run by APMA).
The agricultural cooperatives are not large, unfortunately. So far, their size is still small or medium, and while government already started the creation of “greenhouse” conditions for them, I think, it should support them more (e.g. involve them in agro insurance program), in order to avoid - again and again - the misallocation and waste of resources (especially, at the end of the day).
I agree the point that it might be even compulsory, but at the current stage, as new born cooperatives, they are not eligible to buy the insurance if it is not subsidized, because it is expensive and not affordable with the incomes derived from this sector. As for the subsidizing the agriculture, I have talked a lot about that, and to say shortly, I don’t agree that policy in general, mainly because of high alternative costs and the wrong signals (expectations) what it sends towards the stakeholders (e.g. farmers).
I agree that government should not be cutting the rope it is walking on.
However, when you advocate that government should 'support them more' how do we know that more support would not lead to more waste? And more importantly, what should more support entail?
What government can do however, is to support creation of competitive market for insurance so that vulnerabilities of small scale farmers will not be exploited by large insurance sharks. This way, Georgian farmers will not be forced to play a game of chance with Mother Nature, but with few more follow up initiatives they can be 'nudged' towards purchasing insurance and consequently feel safe and focus their effort and energy on building successful enterprises.
More concretely: a very hot topic in development economics is microinsurance: http://www.a4id.org/sites/default/files/files/Short%20guide%20to%20Microinsurance.pdf
The idea is to offer affordable insurance to the world's poorest - a market that is believed to have billions of dollars of potential.
And here's an excerpt from NYU's Jonathan Morduch's article "Microinsurance: the next revolution?" regarding what concrete steps government can take to promote the creation of microinsurance market in the country:
"...the first is the need for reinsurance, the second is having data on which to base premiums, and the third is the ability to cut the costs of dealing with many small transactions. There is a potential role for public action with regard to the first two at least, with the first being most pressing. The more people that create insurance schemes, the thicker (and thus cheaper and more effective) will be the reinsurance market. But coordination failure may keep the market from getting that far. Public action to encourage reinsurers to develop products and protocols to deal with micro-products could be an important step toward expanding insurance access broadly. 12 collecting reliable data on health, demographic, and agricultural trends will also aid the development of a well-functioning market. The hope is that with those basic elements in place, innovators can be found to deal with both information asymmetries and transaction costs. The microfinance parallel offers cause for encouragement, but establishing widespread insurance will require more detailed regulatory architecture that the mircofinance pioneers needed. Finding ways to cut costs will be best left to entrepreneurs, although there is scope for supporting pilot studies".
And...
"Practitioners have worked hard, sometimes against the odds, to get micro-insurance. A micro- insurance revolution could be a major step toward improving a well-being of the world's poor, but, it is important to design products with a full picture of how the products will fit into clients' lives".
Great insights! Thanks for bringing this to our attention. Do we know if reinsurance is available for Georgian insurance companies? If not, this should be the first order of business for the government... The Georgian market is certainly anything but "thick", making it very expensive for both insurers and the insured. The challenge is to get the ball rolling in the right way.
Majority of Georgian companies do not have reinsurance most probably because of small portfolios and reinsurance costs. Since portfolios of different companies are not pooled each of them deals with reinsurance agents separately and tries to get good conditions (high coverage, low reinsurance rate).
This (reinsurance program) might be the best way how to UPDATE the project (instead of quit, but remember one-size-does-not-fit-all, and one should consider the country specific issues).
What do you suggest instead? To quit many projects every other year (or every year) because admit that you failed? Maybe that is the best solution sometimes, in most cases it is unrealistic, especially because of the political losses.
I dream the market driven economy with minimum intervention from the government, but it is almost impossible in like the country Georgia. I do agree that the government do have major role to create the proper environment, but I also agree that it should sometimes intervene when market is failure (but NOT permanently), or take the first steps in creating the absent market (like the agro insurance project in Georgia, where government tried to create a precedent, but with many bags and NOT proper plan how to update or make it sustainable in a long-run).
The decision itself, whether intervene or not, should be gone through the process, where one should have the precise plan how to make it work with (1) properly identifying the problem, (2) find the ways to solve it, (3) choose the best alternative(s) and (4) plan very carefully to avoid any bags on the way of implementation (plan the worst and hope the best!). And more importantly, plan to UPDATE any program after the monitoring it, according to the lessons learned from the implementation. But once you intervene, then it is very difficult to quit or pull out, especially when donors are also involved, and the elections are getting closer...
One does not need to admit anything - look at Table 1 and the numbers are self explanatory. It's better just to move on and think about novel approaches (as opposed to quitting) rather than constantly being in the Ostrich Position https://waterman99.files.wordpress.com/2010/01/us-ostrich-head-in-sand1.jpg
Very good points Rati! Agricultural insurance as many agriculture related projects in most cases require government support. Even in the most developed countries like Switzerland, farmers are still subsidized. This is a common practice in Europe and US as well.
As experience of other countries shows, you need a lot of years (10, 15 or even more) to develop insurance market in the country. This is a very slow process and I cannot think of a country which developed this market without government intervention at least at the initial stage.
I like this post, but I am not entirely sold that moral hazard (the assurance of a bailout) played such a central role. Bailouts were primarily for Kakheti anyway, thus not so relevant for farmers in the West.
Here is my take:
- the policy was indeed complex, and in my understanding did not cover you up to 100% of your harvest; thus you paid a good chunk to cover a relatively small chunk;
- context matters: the risks differ really in their seasonality; hail risk is very region-specific;
- persistence matters: adoption of such products takes a long time, everywhere;
As mentioned, I think your post is great -- but that you can look to relatively obvious reasons why this isn't yet working, plus insufficient persistence.
You are right when saying that bailouts were mainly for Kakhetians and they played an important role in Kakheti, but not in Western Georgia. Main issues there were (and are) related to insurance service quality and low awareness of farmers about the benefits of agricultural insurance.
Under quality I mean the following:
Loss adjustment and indemnification procedures require significant improvements. That's because loss adjusters are not certified, insurance companies do not have branches in all the regions, sales agents in most cases (if not in all cases) do not have any background in agriculture and cannot answer the questions of farmers. This affected not only West Georgia, but other regions as well.
- As to insurance coverage, it was pretty high (90%) meaning that almost total loss is covered. There is always a deductible as far as I know, that's why insurance is complementary to other risk mitigation techniques. Maybe the scheme was a bit complex, but it is obvious that sales agents in many cases did not explain well to farmers how insurance works, which again speaks about the quality of service.
- Context matters, that's right, but context specific products were offered and coverage was high for all types of products. One could insure against hail, frost, wind, excessive rainfall and fire.
- As to adoptions, indeed it takes many years to reach relatively high penetration rates and this is related to awareness issues.
I agree that moral hazard (related to possible disaster relief payments) did not play a central role here. Nevertheless, this year the government taught a very good lesson to Georgian farmers when decided not to make ad hoc payments. To my knowledge, only a symbolic assistance (some input vouchers amounted to 65GEL per family) was offered. Let’s see if the government will stick to this decision also next year (when approaching election). Very doubtful!
Moral hazard (related to farmers’ incentives to be involved in risky activities, knowing that insurers will compensate them for their losses) will stay a problem also in developed agricultural insurance market. This is a common problem of indemnity-based insurance products. This could be eliminated by developing and offering index-based insurance products (e.g., weather index).
In general, agricultural insurance is quite complex and is affected by several factors, both on the demand side and the supply side of the market. Farmers are characterized by cognitive errors involved in analyzing extremely low probability events, and they often underestimate their risk exposure. They also have limited information and cognitive problems that makes it harder to understand the nature and the details of the contract. On the other hand, insurers add so called ambiguity load (due to data limitation) and administrative cost load (which is particularly high in the case of agriculture) to the expected losses, ending up with high risk premium estimates. This causes insurance markets to operate at less than a socially-optimal level of risk transfer, justifying governmental intervention in this market.
Pati - is there not a critical typo in your first sentence "I agree that moral hazard (related to possible disaster relief payments) did NOT play a central role here". Judging by your second sentence it seems like you want to say the opposite.
Eric, I meant it is not a central issue but an important one and I welcome government's decision not to make disaster payments this year.
There are many other problems contributing to the absence of agricultural insurance market in Georgia. There are no sale agents dedicated to agro insurance, loss adjustments and indemnifications are not timely and transparent, etc. All this leads to farmers' distrust in the system.