Banks

Journal of International Development: Vol. 4, No. 6, 567-581 (1992)

THE APPROPRIATE ROLE OF
AGRICULTURAL INSURANCE IN
DEVELOPING COUNTRIES
PETER B. R.HAZELL*
Agriculture and Rural Development Department, The World Bank

Abstract: Multiple-risk crop insurance programmes have proven expensive to governments but have not lived up to their expectations. Many agricultural risks cannot be insured on a financially sound basis, but there is scope for increased insurance of farm assets, of the life and health of rural people, and of some specific perils that affect crop and livestock yields. Such insurance could be efficiently provided by the private sector if governments were to remove some of the important constraints impinging on commercial insurers. The greatest challenge is to find ways of insuring low-income rural households against natural hazards on a financially sound basis. Simple lottery schemes that provide insurance against catastrophic weather events (e.g. drought or flood) recorded at regional weather stations might prove effective.

1 INTRODUCTION

Agricultural production is inherently a risky business, and farmers face a variety of weather, pest, disease, input supply and market-related risks. Given an uncertain income each year, farmers must worry about their ability to repay debt, t o meet overhead costs (e.g. land rents and taxes) and, in many cases, their ability to meet essential living costs for their families. These same risks are also of concern to agricultural credit institutions. Confronted with risky borrowers, lenders must seek to reduce the possibility of poor loan recovery rates in unfavourable years, even if this means only modest levels of lending to agriculture.
The prevalence of risk in agriculture is not new and farmers, rural institutions and lenders have, over generations, developed ways of reducing and coping with risk. A key question is whether these traditional mechanisms of risk management are sufficient, or whether, given the highly covariate nature of many agricultural risks, public interventions, such as crop insurance, can provide a more efficient alternative.

* Principle Economist in the Agriculture and Rural Development Department, World Bank, Washington,
D.C. The views expressed in this paper are the author’s own, and should not be attributed to the World
Bank and its affiliated organizations.
0954-1748/92/060567-15$12.50
1992 by John Wiley & Sons, Ltd.

568 P . B. R. Hazell
2 HOW FARMERS MANAGE RISKS

The major risks confronting farmers can be grouped as follows: market risks, such as the prices of outputs and inputs, and interest rates; resource risks, such as uncertain supplies of labour, credit and irrigation water, or the timeliness of supply of seeds and fertilizers; production risks, which cover a whole gamut of pest, disease and weatherrelated risks; health risks (sickness, death and accident) of the farmer and his/her dependents; asset risks, such as theft or fire damage to buildings, machinery and livestock; and other risks, such as confiscation of land, war damage, and other ‘acts of god’. The dominant risks vary widely from one agricultural regime to another, but it is probably fair to claim that farmers in developing countries are exposed to most types of risk, and that low-income farmers, especially in semi-arid areas, are the most exposed. In order to cope with these risks, farmers and rural societies have developed a range of risk management measures. These can usefully be classified into risk-reducing and risk-coping strategies (Walker and Jodha, 1986).
Risk-reducing Strategies

These include crop diversification, intercropping, farm fragmentation and diversification into non-farm sources of income. Crop-sharing arrangements in land renting and labour hiring contracts can also provide an effective way of sharing risks between individuals, thereby reducing a farmer’s risk exposure.
Risk-reducing strategies can be quite effective in addressing many production and market risks (see, for example, Hazell et al., 1983; Low, 1974). But while they help to stabilize family incomes, they are typically costly for average income because they require that farmers forgo their most profitable alternatives. For example, crop diversification is usually less profitable on average than crop specialization, and land fragmentation imposes costs in the form of labour and transport inefficiencies.
Risk-coping Strategies

These are relevant for dealing with catastrophic income losses once they occur. In order to repay loans and to meet essential living costs in disastrous years, farmers may rely on new credit (especially consumer credit from local stores), the sale of assets, use of own food stocks, or temporary off-farm employment. In many rural societies, mutual aid or kin-support systems also provide an important safety net for member households. Risk-coping strategies can also be costly. For example, the sale of assets such as land or livestock can leave the household without adequate productive assets for the future. Debt also has to be repaid, and interest rates charged by informal lenders are rarely cheap. But a more fundamental problem with traditional risk-coping strategies is that they cannot deal effectively with the couariability problem that characterizes most agricultural risks. Production and price risks, for example, affect nearly all farmers simultaneously within a small rural community. As a result, many farmers seek consumer credit at the same time, thereby driving up local interest rates.

Appropriate Role of Agricultural Insurance in DCs

569

Similarly, local wages are driven down by a surge in the labour supply, and the value of farm assets declines as too many farmers try to sell at the same time. Moreover, once the crisis is over, farmers will find it difficult to replace assets as prices are bid up again through mutual competition.
The covariability problem does not exist for all risks; for example, most health risks are independently distributed across individuals. In these cases, local risk-coping strategies can be quite effective. But for covariate risks, local risk-coping strategies need to be reinforced by risk pooling arrangements that cut across small rural communities. Herein lies part of the promise of formal banking and insurance institutions, since their portfolio can span different regions and even different sectors within the national economy.

3 HOW BANKS MANAGE RISK
Formal banking institutions lending to agriculture face loan default risks arising from all the farm-level risks discussed in the preceding section. Additionally, they face uncertainties about interest rates, inflation, changes in government regulations affecting their banking practices, and risk in identifying problem borrowers who have little intention of repaying their loans. Again, many of these risks are covariate, and loan default rates can be expected to be high in regions affected by a major natural catastrophe, or when product prices decline sharply, or interest rates surge upwards.
To manage these risks, commercial banks typically diversify across different types of farming, across regions and across sectors of the national economy. They also maintain financial reserves, and establish contingent loan arrangements with other banks. They also build up personal relationships with their borrowers, providing a full range of banking services, and tailoring loan sizes and conditions to individual needs and debt-carrying capacity. In the event of financial distress, a good bank will also work with a borrower in establishing a rescue plan, rescheduling or rolling over debt to the extent possible. To protect their own capital, commercial banks adjust interest rates to reflect risk premiums and insist on collateral. Collateral not only helps to insure loans, but reduces the incentive for wilful default. These practices can be quite effective in managing risk but, because they increase the costs of administering each loan, they lead to lending portfolios that predominantly serve large commercial farms and neglect most of the smaller and high-risk farmers.
In contrast, government-sponsored agricultural development banks (ADBs) seek to serve a much wider clientele of farmers, but unfortunately they also seem to forgo most risk management strategies. Often they are highly specialized in agriculture, and in regions and types of farms that grow specifically targeted crops. They rarely offer any additional banking service beyond agricultural credit and, far from building up individual relationships with their borrowers, offer standard loan packages (e.g. credit for predetermined, per hectare amounts of seed, fertilizer, pesticides, etc. for specified crops) that often bear little relationship to the needs or debt-carrying capacity of individual borrowers (Von Pischke, 1986). They have limited capacity for helping borrowers deal with financial distress, are prohibited from adjusting interest rates to risk levels and, even if they require collateral, find it politically difficult to collect in the event of default. Given high-risk lending portfolios and limited ability to manage risks, ADBs almost invariably suffer from poor loan recovery rates and rely on direct access to government funds to maintain their financial viability.

570 P. B. R. Hazel1
4

THE PROMISE OF AGRICULTURAL INSURANCE

In principle, agricultural insurance offers important benefits to farmers and ADBs that have captured the imagination (and purse) of policy-makers in numerous developed and developing countries.
Since agricultural insurance institutions can overcome the covariability problem, they should be able to provide a more efficient risk-management tool than those traditionally available to farmers (Ahsan, 1985; Ray, 1967). If a farmer knows that he will be financially compensated when his income is catastrophically low for reasons beyond his control, then he will be more likely to allocate resources in profitmaximizing ways. For example, he will grow more of the most profitable crops even if they are more risky, and he will be more likely to adopt improved but uncertain technologies. The net effect could be an increase in value added from the agricultural sector, an increase in farm incomes, and a reduction in rural poverty.
At the same time, insurance should reduce the risk of loan default for given amounts of credit, enabling banks to increase their lending to agriculture and/or to improve their loan recovery rates. Smaller, high-risk farms should also become more attractive to commercial banks, improving their representation in the lending portfolio (Pomareda, 1986; Hazel1 et al., 1986a).
Whether these expectations translate into real benefits in practice depends on whether insurable risks are the major determinants of catastrophically low incomes, and whether the insurance can be provided at a cost that is lower than the benefits it bestows. An insurable risk has three characteristics. First, the likelihood of the event must be readily quantifiable. Second, the damage it causes must be easy to attribute and value.
Third, neither the occurrence of the event nor the damage it causes should be affected by the insured’s behaviour (i.e. absence of moral hazard). Typhoon damage is usually a good example of an insurable risk. Its likelihood of occurrence is obtainable from weather records; it is easy to monitor; the damage is often total; and the farmer cannot affect its occurrence or the damage it causes. On the other hand, crop damage due to pest attack is usually an uninsurable risk. Its likelihood of occurrence is hard to quantify; its actual occurrence and the damage caused can be difficult to verify and value, particularly when there are other causes of crop loss; and the farmer can, through negligent husbandry, increase both the likelihood of occurrence and the damage it causes.
If insurance is restricted to insurable risks, then only a subset of the total array of risks confronting farmers can be insured. These include some production risks (e.g. losses due to catastrophic weather events such as drought, pest, hail, and flood), most health risks, and some asset risks. But most market and resource risks, and a wide array of production risks (especially preventable damage from pests, disease, and excess humidity and temperature), are not strictly insurable.
Even some of the insurable risks cannot be viably insured in practice because they occur so frequently that farmers cannot realistically afford the required premiums (e.g. flood damage in Bangladesh). It is also difficult to viably insure small and mediumsized farms because of high administration costs.
In this context it seems clear that insurance is not a general panacea for farmers’ risk problems. Viable insurance cannot address many of the important reasons for severe shortfalls in family incomes, and as such it has a limited role to play in

Appropriate Role of Agricultural Insurance in DCs

571

protecting family consumption and debt repayment. There are situations where insurable risks are predominant, in which case insurance can make a useful contribution to farmers and banks. But these situations are likely to be exceptions rather than representative of the more general situation.
A good indication of the limited role for viable agricultural insurance is provided by the private sector. Private insurers provide health, life and asset insurance for farmers in many countries, but they have been very reluctant to move into crop and livestock insurance. Even where they have, they work almost exclusively with large-scale commercial farmers and only insure the kinds of insurable risks described above.
Policy-makers have not easily been persuaded by these limitations and, in order to broaden the reach of agricultural insurance, they have extended the coverage to include important but uninsurable risks, and have subsidized the insurance to reduce the cost to farmers. These policies are invariably implemented through the creation of publicly owned agricultural insurance institutions.

5 PUBLIC AGRICULTURAL INSURANCE

Most public agricultural insurers concentrate on crop insurance. With few exceptions this is often multiple-risk yield insurance that compensates a farmer for serious yield losses caused by a wide range of hazards. Typically, a normal yield and a compensation price are defined in the insurance contract, together with a rule for determining when an indemnity is due (e.g. if, because of an insured hazard, actual yield is less than 80 per cent of the normal yield). The compensation is then calculated as the difference between the normal and actual yield valued at the compensation price. Although the insured causes of yield loss are typically specified, these are so comprehensive that in practice it is virtually impossible to exclude damage due to uninsured hazards when assessing losses.
Public crop insurance is often tied to credit from ADBs. A borrower is required to purchase the insurance in order to obtain a loan. Moreover, the bank usually acts as an intermediary, collecting the premium from the farmer and receiving indemnities on his behalf. The farmer sees any money only in the unlikely event that the indemnity exceeds his/her debt to the bank.
Since many of the insured risks are essentially uninsurable, and/or because they occur so frequently that they are expensive to insure, multiple-risk crop insurance is inherently expensive and has to be heavily subsidized. The need for subsidies is also accentuated by the high administration costs associated with insuring small and medium-sized farms. Governments subsidize part or all of the administration costs, and additionally they reinsure the agency by providing direct transfers in high-loss years. Unlike commercial reinsurance, however, governments rarely require that the insurer reimburse these transfers in any way.
With few exceptions the financial performance of public crop insurers has been ruinous, both in developed and developing countries (Hazel1 et al., 1986b).
To be financially viable without government support, an insurer needs to keep the average value of its annual outgoings-indemnities plus administration costs-below the average annual value of the premiums it collects from farmers. Let A denote average administration costs, let I denote average indemnities, and let P denote the average amount of premiums collected from farmers, then, to be financially solvent, an

572 P.B. R. Hazel1 insurer must satisfy

Public crop insurers invariably fail to meet this condition and, as the seven insurance programmes in Table 1 illustrate, the 2 ratios are typically much larger than 1.0
Of the selected crop insurance programmes the Comprehensive Crop Insurance
Scheme (CCIS) of India has the worst financial record. Begun in 1985, the programme paid out over 5 rupees for each rupee of premium income collected from farmers in its first 5 years of operation (all payments in 1987 prices), and this does not even include administration costs. The Philippine programme fared little better, paying out 5.74 pesos for each peso of premium collected from farmers in 1981-89. The experience in
Brazil (1975-8 l), Mexico (1980-89) and in Japan (1985-89) has not been much better, and even the better-performing programmes have Z ratios that are larger than 2.4.
The Z ratios have also been disaggregated into their two components in Table 1;
ZIP-sometimes called the loss ratio-expresses the indemnities as a ratio of the total premium collected from farmers, and A / P expresses the administration costs as a proportion of the total premium collected from farmers. The high loss ratios indicate that most of the insurers ' losses occur because they charge premiums that are much too low in relation to the average indemnities paid. In the US, for example, farmers receive on average $1.87 for every dollar of premium they pay. Farmers in India,
Brazil, Mexico and the Philippines do even better, paying about 20 to 30 cents for

Table 1. Financial performance of seven agricultural insurance programmes.

+ I)/P

Country
(Insurer)

Period

IIP

AIP

(A

Brazil (PROAGRO)
Costa Rica (INS)
India (CCIS)
Japan (agriculture)

1975-81
1970-89
1985-89"
1947-77
1985-89b
1980-89d
1981-89
1980-90

4.29
2.26
5.11
1.48
0.99
3.18
3.94
1.87

0.28
0.54
n.a.
1.17
3.57 '
0.47
1.80
0.55

4.57
2.80
n.a.
2.60
4.56
3.65
5.74
2.42

Mexico (ANAGSA)
Philippines (PCIC)
USA (FCIC)

"Not including the 1989 Rabi season. bPaddy only.
Administration cost data based on 1989 only. dCrop insurance only.
Nore: The data were adjusted with the relevant national consumer price index to constant 1987 prices before performing the calculations reported in this table.
This adjustment was not possible for the Japan 1947-77 data, and the results are based on nominal prices.
Sources: Brazil (Lopes and Silva Dias, 1986); Japan 1947-77 (Ray, 1979); USA,

Philippines, Costa Rica, and Mexico (annual reports of the relevant insurers);
India (Gairola. 1991); and Japan 1985-89 (unpublished data).

Appropriate Role of Agricultural Insurance in DCs 573 every dollar of indemnities they receive. Only in Japan do farmers pay about as much in premium as they receive in average indemnities.
Administration costs are also a major financial drain, ranging from $0.28 (Brazil) to
$3.57 (Japan 1985-89) for every dollar of premium collected from farmers. Japan’s administration costs were not always so high, averaging only $1.17 per dollar of premium collected from farmers during 1947-77. But now they more than offset a relatively low loss ratio, leading to a 2 ratio of 4.56 during 1985-89.
To cope with these levels of financial loss all the programmes rely on government largesse. This may take the form of premium subsidies and periodic financial transfers
(e.g. Mexico and the USA), an initial capital endowment that earns investment income
( e g the Philippines), or the granting of the right to issue bonds with a government guarantee (e.g. Costa Rica).
In recent years the insurance programmes in India, Japan, Mexico and the USA have each cost their governments US$0.5 and 1.0 billion per year in total transfers.
Another measure of the public cost of crop insurance is the total government cost for each hectare of insured crop (Table 2). During the 1980s the US (FCIC) scheme cost the Federal government about US$lO/ha, and the Philippine PCIC scheme cost its government US$2l/ha. These costs were relatively low compared to US$55/ha in
Costa Rica during 1970-89, US$44/ha in Mexico during 1980-89, and $408/ha in
Japan (rice only) in 1989. Given the alternative uses to which these funds could be put, both in increasing agricultural productivity and in reducing risk (e.g. irrigation can achieve both these objectives), the rationale for public crop insurance is tenuous.
Indeed, social cost-benefit analyses of the Mexican and Japanese schemes show negligible social returns in relation to their high costs (Bassoco et al., 1986; Tsujii,
1986). Even when viewed as a way of transferring income to farmers in times of need, crop insurance programmes do not perform well. As shown in Table 2, it costs about
$1.50 to transfer one dollar of net benefit to farmers.

Table 2. Additional indicators of performance for selected crop insurance programmes (constant 1987 US dollars). cost to government per insured hectare USA (1980-90)
Mexico (1980-89)
Costa Rica (1970-89)
Philippines (1981-89)
Japan (1985-89)

cost to government to give $1 to farmersa

10
44
55
21
408

1.63
1.22
1.43
1.61

-b/

Premium rates (%)
Actual

Breakeven

4.8
8.2
7.6
3.3
1.6

11.9
29.9
21.3
16.5
7.4

“Total cost to government/(f - P ) .
*Japanese rice farmers paid $8.3 millions more than they received during 1985-89, and all government contributions to the insurance went towards administration costs.

574 P.B. R. Hazel1

6 WHY PUBLIC INSURANCE FAILS
As discussed earlier, the primary reasons for the high cost of public crop insurance schemes are that (a) they invariably attempt to insure uninsurable risks, (b) many insured hazards occur so frequently that the required premiums are too high for most farmers to afford, and (c) they insure many small farms and hence incur high administration costs. But these are not the only reasons for failure.
Another overwhelming factor is the incentive problem that arises once the government establishes a pattern of guaranteeing the financial viability of an insurer.
If the insurance staff know that any losses will automatically be covered by government, then they have little incentive to pursue sound insurance practices when setting premiums and assessing losses. In fact, they will find it profitable to collude with farmers in filing exaggerated or falsified claims. When the insurer underwrites the loans of an ADB, these incentive problems can easily infect the bank too, leading to a serious loss of discipline in banking practices. Why, for example, should ADB staff try to collect loans from tardy borrowers if they can more easily obtain repayment from the insurer?
The Mexican experience provides a good example of the destructive power of these incentive problems. Prior to its termination in February 1990 the Mexican insurer
(ANAGSA) had become so lax that claims were successfully filed on over threequarters of the insured crop area in each of the last 3 years. Farmers openly talked of bribery rates of about 30 per cent of the total indemnity, and the ADB (Banrural) had grown to depend on ANAGSA for recovery of about one-third of the value of its total loans. Another common reason for failure has been that governments undermine public insurers for political reasons. In Mexico the total indemnities paid bore a strong statistical relationship with the electoral cycle, increasing sharply immediately before and during election years, and falling off thereafter. In Costa Rica, Gudger and Avalos
(1986) report that, in one election year, the government declared a disaster and ordered INS to pay rice farmers despite late-season rains that saved a large part of the crop. In the US, the government had repeatedly undermined the national crop insurer
(FCIC) by providing direct assistance to producers in disaster areas. Why should farmers purchase crop insurance against major calamities (including drought) if they know that farm lobbies can usually apply the necessary political pressure to obtain direct assistance for them in times of need at no financial cost?
The uninsurability of many yield risks arises from moral hazard problems. Farmers simply have reduced incentive to pursue sound husbandry practices if they know that any serious yield loss will be compensated by the insurer. But the problem is aggravated by the practices of many public insurers. First, they rarely incorporate sizeable deductibles into insurance contracts, so farmers do not bear any significant part of yield losses themselves. Second, rather than insuring the actual crop damage, losses are assessed as the difference between actual and normal yields. The latter also tend to be set too high, particularly when insurance is tied to a credit scheme. The result is that farmers receive excessive indemnities in relation to their real yield losses, and it is often more profitable to ‘farm’ insurance than the crop itself. Third, premium rates are often set by government decree at unrealistically low levels (Table 2). This not only causes the insurer to lose money, but also prevents premium rates from being adjusted to reflect the losses claimed by individual farmers. Knowing that they will

Appropriate Role of Agricultural Insurance in DCs

575

not be penalized with higher premiums in future years, farmers have little incentive to curtail their claims.
Another reason for their high cost is that crop insurers tend to be too specialized, focusing on specific crops, regions and types of farmers, particularly when the insurance is tied to an ADB that has a mandate to serve particular target groups identified by the government. In Costa Rica, Japan and the Philippines, for example, the crop insurance programmes are heavily concentrated on rice farmers. Without a well-diversified insurance portfolio, crop insurers are susceptible to covariability problems, and face the prospect of sizeable losses in some years. Since public insurers are rarely able to obtain commercial reinsurance or contingent loan arrangements, this specialization increases their dependence on the government.
Public crop insurers also tend to have high administration costs. This is partly because they often insure small-scale farmers, but also because crop insurance work is very seasonal, and the absence of a well-diversified portfolio means that staff and field equipment are underemployed for significant parts of the year.
7 GUIDELINES FOR IMPROVING PUBLIC CROP INSURANCE

Publicly owned crop insurance programmes are probably impossible to design as sustainable entities without some level of financial support from the government. But the observed diversity in financial performance does suggest some scope for controlling costs if the right kinds of practices are followed. It is also noticeable that even some highly inefficient insurers (e.g. ANAGSA in Mexico) have made profits on some types of health and farm asset insurance. Thus, being a publicly owned insurer need not, by itself, spell disaster; the trick is again to identify the appropriate types of practice to follow. I offer nine principles for good crop insurance.
1. Make the insurer financially responsible for its own affairs; that is, no automatic access to government funds. Subsidies are not necessarily ruled out, but they should be set as some fixed percentage of the total premium. There should be no ex-post adjustment of the subsidy each year to reflect the insurers losses. If an insurer is to remain financially responsible, this also means that it should be free of government manipulation (especially in election years). The agency is more likely to succeed if it is an autonomous public institution with its own board of directors, and not a department within the Ministry of Agriculture.
2. To the maximum extent possible, the insurer should write coverage only on insurable risks; that is, perils that can easily be quantified, losses easily attributed and valued, and which are not subject to moral hazard problems. Most catastrophic weather risks are insurable (e.g. hurricane, typhoon, severe drought, and flood) but yield losses due to pests and diseases, or excess humidity or temperature, are not.
3. Farmers should be compensated only for actual crop damage or inputs lost, and not for their failure to achieve a normal or target yield.
4. Policies should include deductibles of at least 20 per cent of the coverage to ensure that farmers have an interest in minimizing losses.
5. Premiums should be based on sound, actuarial calculations, using available weather records and well-maintained records on insured farmers. The premium

576

6.

7.

8.

9.

P.B. R.Hazel1 should be set high enough to cover average indemnities, administration costs, and a contribution towards building up a financial reserve.
The insurer should develop a rational insurance portfolio to spread its risks and minimize the chance of large losses. This may require insuring a range of different crops, livestock and non-farm activities, spread across a number of regions. If the insurer is required to insure the loans of ADBs, then it should have some say in the structure of the lending portfolio, and in the loans it wishes to insure. Rational portfolio management also requires contingency arrangements for coping with large losses in any one year. Commercial reinsurance, particularly in the international market, is attractive because it requires the insurer to maintain high standards of insurance practice. If the government is to underwrite the scheme, then this should be set up as a contingent loan arrangement with the insurer paying back any loans on an agreed schedule.
The insurance should be voluntary and in competition with the private sector.
This will require that the insurer develop and market the types of insurance that farmers most want.
To avoid adverse selection problems, premiums and indemnities should be tailored to the risk levels of individual farmers, and not to regional averages.
Administration costs for small-scale farmers can be reduced by insuring groups of farmers as well as individuals. If group premium rates are sufficiently lower, farmers will be encouraged to organize themselves into insurance mutuals. The insurer can provide technical assistance as required.
The insurer needs to keep tight control of administrative costs. Staff and field costs can be contained either by diversifying the insurance portfolio to reduce seasonality in the work load, or by relying on seasonal inspectors (e.g. teachers in rural areas are often idle at harvest time, and could be employed to assess crop losses). Commercial insurers rigorously adhere to such principles, not least because they know that the government is unlikely to bail them out should they fall into financial distress. Even so, the volume of private crop insurance is not trivial. Gudger (1990) estimates total premium income at about $1 billion per year, and it seems to be growing quite rapidly. The drawback is that private insurance is almost exclusively confined to large-scale, commercial farms growing high-value crops. If public crop insurance were operated along the same principles, it would not be able to serve many of the medium to smaller-sized farms that are the traditional targets of government assistance programmes.

8 THE SCOPE FOR PRIVATE SECTOR INSURANCE
Private agricultural insurers have traditionally insured a range of health and asset risks in many developing and industrial countries, and there is also a long experience with hail and fire insurance for some crops. Recent years have also seen a surge in other forms of private crop insurance which now covers some 60 crops around the world. According to Gudger (1990), these include cereals, pulses, forage crops, fruits, aquaculture, tropical beverages, flowers, forest products, tree crops, vegetables and some highly specialized agricultural activities such as mushrooms, snails, crocodiles

Appropriate Role of Agricultural Insurance in DCs 577

and Parmesan cheese ageing. The perils insured against range from ‘the ordinary flood, freeze, fire, hurricane, and drought to the exotic such as foraging elephants, snakes, parrots and kangaroos, maritime algae blooms, and deer stampedes due to passing aircraft’ (Gudger, 1990, p. 26).
One of the most successful examples of private insurance is to be found in Chile, where a private insurance company responded to the insurance needs of a growing number of export-oriented farmers (Roberts and Dick, 1991). Despite heavy losses in its first year (1981), the programme has now successfully completed 10 years of operation without any financial support from the government. Its loss ratio for
1982-89 was remarkably low at 0.47.
Although confined to commercial farmers, private sector crop insurance may have considerable potential to expand in developing countries. There are several constraints that need to be overcome.
1. Private crop insurance is legally restricted in some countries, particularly where an existing publicly owned scheme is already operating.
2. Even where the private sector is allowed to operate, it may find it hard to compete with a heavily subsidized public scheme. While this has proved to be less of a constraint in insuring commercial producers of high-value crops, it probably discourages private insurers from attempting to insure many middle-sized farms, or from helping farmers organize into insurance mutuals.
3. In countries where insurance institutions have been nationalized, the absence of diversified private insurers with solid experience in automobile, home, life or health insurance offers a weak base for building up agricultural insurance.
4. There are infant-industry problems in establishing agricultural insurance, particularly where the initiative is taken by farmers rather than by an existing nonagricultural insurer. There are set-up costs in quantifying insurable risks and calculating premiums. More importantly, the insurance portfolio is likely to be small and specialized in the early years and, starting with a limited financial reserve, the insurer is likely to go under in the event of heavy losses. In principle, reinsurance can solve this problem, but this is difficult to obtain until the insurer has a proven record.
5. The potential amount of international reinsurance that can be obtained for agriculture may be limited (Gudger, 1990). At present, reinsurance can only be purchased in the European markets; the US and Japanese markets have been unwilling to participate.
These are constraints that governments can help overcome, and it is possible that private insurance could play a much-expanded role in many developing countries.
However, given the inherent difficulties and costs of insuring most agricultural production risks, the potential for viable crop insurance is limited. The scope for increased private insurance in developing countries of farm assets and life and health insurance for rural people is probably much greater.

9 INSURANCE FOR SMALL FARMS
A basic dilemma with agricultural insurance is that attempts to increase its financial viability also seem to reduce its applicability to the mass of small- and medium-sized

578 P. B. R.Hazel1 farms. Some of the smaller but commercially viable farms might still be reached by forming them into groups, or insurance mutuals, and the Japanese experience, notwithstanding its high administration cost, may offer useful lessons here. But what can be done to assist the risk management practices of the vast number of small, semicommercial or even purely subsistence farmers operating throughout much of the developing world?
In order to overcome the major problems associated with crop insurance, and to substantially reduce its administration costs, several authors have proposed areabased yield insurance (e.g. Halcrow, 1948; Dandekar, 1977; Miranda, 1991). Under this proposal, the crop yield for a homogeneous region is insured, and all insured farmers in the region pay the same premium and receive the same indemnity.
Indemnities are paid whenever the average yield for the region falls below some critical level irrespective of the actual yields obtained by individual farmers. Premiums are calculated on the basis of year-to-year variations in the average yield for the region, and would vary from one homogeneous region to another in accordance with differences in risk levels.
This approach reduces moral hazard problems, and hence broadens the range of yield risks that can be viably insured. Moreover, since the premiums and indemnities are identical for all insured farmers in a region, it avoids the adverse selection problem.
The latter refers to the situation in which farmers facing below-average risk tend to drop out of insurance programmes if they are charged premium rates based on average risk levels but are paid indemnities based on their own losses. Also, by eliminating the need for field inspections and loss assessments, the cost of administering an area-based scheme could be kept very low. Providing farmers pay their premium, it is not really necessary that they even grow the crop that they have insured. Despite its appeal and its potential scope for reaching small-scale farmers, there are problems with the proposal. First, the insurance will be attractive to individual farmers only if their yields are highly correlated with the average yield for the region.
Dandekar (1977) argues that homogeneous areas can be defined in India in which inter-farm yield correlations are positive, but there is a growing body of microevidence showing that yield correlations between plots within the same village, or even within the same farm, are surprisingly low (e.g. Walker and Jodha, 1986). Small differences in ground contours, slope, and wind and sun exposure can lead to substantial differences in the yield damage caused by unfavourable climatic, pest and disease events, as can a few days difference in planting dates or the crop varieties grown. Second, the scheme is subject to severe covariability problems. When the average yield is below the critical value in a region, all the insured farmers have to be compensated simultaneously. An unsubsidized insurer could hope to survive only if the scheme were to span a large number of regions with negative or positive but weakly correlated yields. The alternative of seeking commercial reinsurance seems unlikely given private insurers’ reluctance to insure yields against a wide array of perils. Another variant of the proposal is regional rainfall insurance (Industries Assistance
Commission, 1986). In this case the insurance pays out whenever the average rainfall for a region falls below some critical value. Rainfall is easier to measure than average yield, and with modern satellite imagery it is now possible to assess the soil moisture

Appropriate Role o Agricultural Insurance in DCs f 579

content for a region with the minimum of field inspection. This information could be used to confirm rainfall readings, or even used directly as the insured peril.
Unlike regional yield insurance, rainfall insurance is not tied to the performance of specific crops. Since most farm families rarely depend on a single crop for their total income, the attractiveness of rainfall insurance should depend more on how the total income for individual families correlates with regional rainfall. Moreover, if the insurance is limited to the occurrence of severe drought or floods, it is likely that in semi-arid or flood-prone areas the indemnities would coincide with catastrophic income outcomes for many rural families.
Drought or flood insurance could be marketed rather like lottery tickets, employing low-income people to sell tickets on a commission basis. Unlike standard lotteries, however, all ticket holders would win a prize in a disaster year, but no prize would be given in non-disaster years. There is no need to restrict the insurance to farm families, and all types of rural households might find it attractive. This is because a decline in farm incomes in drought or flood years usually leads to a sharp contraction in the rural non-farm economy, and the incomes of many workers and businessmen decline in tandem.
Drought or flood insurance faces the same covariability problem as regional yield insurance, but because it is limited to a specific weather peril, it might be much easier to obtain international reinsurance. This prospect could be enhanced if the scheme were run by a commercial bank or insurer within the country.
Schemes of this kind have yet to be tried, and without pilot projects it will be difficult to assess their potential value. Their cost-effectiveness would also need to be compared with alternative means of assisting vulnerable households, such as relief employment schemes, and targeted food rations or income transfers.
10 CONCLUSIONS
Despite the importance of risk to farmers and agricultural development banks
(ADBs), agricultural insurance would seem to have only a limited role to play in efficiently managing these risks. There is scope for increased insurance of some farm assets, of the life and health of rural people, and of some specific perils that affect crop and livestock yields. The private sector could do more to provide these kinds of insurance if governments were to address some of the important constraints impinging on commercial insurers.
The problem of assisting low-income rural households in coping more efficiently with catastrophic income losses remains to be resolved. It is conceivable that simple drought- or flood-based lotteries could provide a low-cost aid, but this remains to be demonstrated. The alternative is for governments to expand direct assistance programmes, such as relief employment or food rations schemes which, with careful targeting, might prove a more cost-effective form of aid than current crop insurance programmes. REFERENCES
Ahsan, S . M. (1985). Agricultural Insurance: A New Policyfor Developing Countries.Brookfield,
Vermont: Gower.

580

P.B. R. Hazell

Basocco, L. M., Cartas, C. and Norton, R. (1986). ‘Sectoral analysis of the benefits of subsidized insurance in Mexico’. In Hazell, P. B. R., Pomareda, C. and Valdes, A. (eds), Crop Insurance for Agricultural Development: Issues and Experience. Baltimore: Johns Hopkins University
Press, pp. 126- 142.
Dandekar, V. M. (1977). ‘Crop insurance for developing countries’. Teaching and Research
Forum Paper No. 10, New York: Agricultural Development Council.
Gairola, S. (1991). ‘Country papers: India’. In Agricultural Insurance in Asia: Planning and
Practices, Tokyo: Asian Productivity Organization.
Gudger, M. (1990). ‘Crop Insurance: failure of the public sector and the rise of the private sector alternative’. Paper presented at the World Bank’s Tenth Agricultural Sector Symposium,
Washington, DC.
Gudger, M. and Avalos, L. (1986), ‘Planning for the efficient operation of crop credit insurance schemes’. In Hazell, P. B. R., Pomareda, C. and Valdes, A. (eds), Crop Insurance for
Agriculrural Development: Issues and Experience. Baltimore: Johns Hopkins University
Press, pp. 263-280.
Halcrow, H. G. (1948). ‘The theory of crop insurance’. Ph.D. thesis, University of Chicago.
Hazell, P., Norton, R., Parthasarathy, M. and Pomareda, C. (1983). ‘The importance of risk in agricultural planning models’. In Norton, R. D. and Solis, L. (eds), The Book of C H A C :
Programming Studies for Mexican Agriculture. Baltimore: Johns Hopkins University Press, pp. 225-249.
Hazell, P., Bassoco, L. M. and Arcia, G. (1986a). ‘A model for evaluating farmers’ demand for insurance: applications in Mexico and Panama’. In Hazell, P. B. R., Pomareda, C. and
Valdes, A. (eds), Crop Insurance for Agricultural Development: Issues and Experience.
Baltimore: Johns Hopkins University Press, pp. 35-66.
Hazell, P., Pomareda, C. and Valdes, A. (eds) (1986b). Crop Insurance for Agricultural
Development: Issues and E.xperience. Baltimore: Johns Hopkins University Press.
Industries Assistance Commission (1986). Crop and Rainfall Insurance, Report No. 393.
Canberra: Australian Government Publishing Service.
Lopes, M. de Rezenda and da Silva Dias, G. L. (1986). ‘The Brazilian experience with crop insurance programs’. In Hazell, P. B. R., Pomareda, C. and Valdes, A. (eds), Crop Insurance for Agricultural Development: Issues and Experience. Baltimore: Johns Hopkins University
Press, pp. 240-262.
Low, A. R. C. (1974). ‘Decision taking under uncertainty: a linear programming model of peasant farmer behavior’, Journal of Agricultural Economics, 25( 15), pp. 31 1-320.
Miranda, M. J. (1991). ‘Area-yield crop insurance reconsidered’, American Journal of Agricultural Economics, 73(2), pp. 233-242.
Pomareda, C. (1986). ‘An evaluation of the impact of credit insurance on bank performance in
Panama’. In Hazell, P. B. R., Pomareda, C . and Valdes, A. (eds), Crop Insurance f o r
Agricultural Development: Issues and Experience. Baltimore: Johns Hopkins University
Press, pp. 101-116.
Ray, P. K. (1967). Agricultural Insurance: Theory and Practice and Application to Developing
Countries. Oxford: Pergamon.
Ray, P. K. (1979). Multiple-Risk Crop Insurance in Japan. Rome: Food and Agriculture
Organization of the United Nations (processed).
Roberts, R. A. J. and Dick, W. J. A. (1991). Strategies f o r Crop Insurance Planning. F A 0
Agricultural Services Bulletin 86. Rome: Food and Agriculture Organization of the United
Nations.

Appropriate Role o Agricultural Insurance in DCs 58 1 f Tsujii, H. (1986). ‘An economic analysis of rice insurance in Japan’. In Hazell, P. B. R.,
Pomareda, C. and Valdes, A. (eds), Crop Insurancefor Agricultural Development: Issues and
Experience. Baltimore: Johns Hopkins University Press, pp. 143-155.
Von Pischke, J. D.
(1986). ‘Can crop credit insurance address risks in agricultural lending?’ In
Hazell, P. B. R., Pomareda, C. and ValdCs, A. (eds), Crop Insurance for Agricultural
Development: Issues and Experience. Baltimore: Johns Hopkins University Press, pp. 87-100.
Walker, T. S.and Jodha, N. S.‘How small farm households adapt to risk’. In Hazell, P. B. R.,
Pomareda, C. and Valdb, A. (eds), Crop Insurance for Agricultural Development: Issues and
Experience. Baltimore: Johns Hopkins University Press, pp. 17-34.

References: Ahsan, S . M. (1985). Agricultural Insurance: A New Policyfor Developing Countries.Brookfield, Vermont: Gower. Basocco, L. M., Cartas, C. and Norton, R. (1986). ‘Sectoral analysis of the benefits of subsidized insurance in Mexico’ Dandekar, V. M. (1977). ‘Crop insurance for developing countries’. Teaching and Research Forum Paper No Gairola, S. (1991). ‘Country papers: India’. In Agricultural Insurance in Asia: Planning and Practices, Tokyo: Asian Productivity Organization. Gudger, M. (1990). ‘Crop Insurance: failure of the public sector and the rise of the private sector alternative’ Gudger, M. and Avalos, L. (1986), ‘Planning for the efficient operation of crop credit insurance schemes’ Halcrow, H. G. (1948). ‘The theory of crop insurance’. Ph.D. thesis, University of Chicago. Hazell, P., Norton, R., Parthasarathy, M. and Pomareda, C. (1983). ‘The importance of risk in agricultural planning models’ Hazell, P., Bassoco, L. M. and Arcia, G. (1986a). ‘A model for evaluating farmers’ demand for insurance: applications in Mexico and Panama’ Hazell, P., Pomareda, C. and Valdes, A. (eds) (1986b). Crop Insurance for Agricultural Development: Issues and E.xperience Industries Assistance Commission (1986). Crop and Rainfall Insurance, Report No. 393. Lopes, M. de Rezenda and da Silva Dias, G. L. (1986). ‘The Brazilian experience with crop insurance programs’ Low, A. R. C. (1974). ‘Decision taking under uncertainty: a linear programming model of peasant farmer behavior’, Journal of Agricultural Economics, 25( 15), pp Miranda, M. J. (1991). ‘Area-yield crop insurance reconsidered’, American Journal of Agricultural Economics, 73(2), pp. 233-242. Pomareda, C. (1986). ‘An evaluation of the impact of credit insurance on bank performance in Panama’ Ray, P. K. (1967). Agricultural Insurance: Theory and Practice and Application to Developing Countries Ray, P. K. (1979). Multiple-Risk Crop Insurance in Japan. Rome: Food and Agriculture Organization of the United Nations (processed). Roberts, R. A. J. and Dick, W. J. A. (1991). Strategies f o r Crop Insurance Planning. F A 0 Agricultural Services Bulletin 86 Tsujii, H. (1986). ‘An economic analysis of rice insurance in Japan’. In Hazell, P. B. R., Pomareda, C (1986). ‘Can crop credit insurance address risks in agricultural lending?’ In Hazell, P