Document of
The World Bank
FOR OFFICIAL USE ONLY
Confidential
Report No. 22289-ME
MEXICO
Crop Insurance
Main Report
May 2001
Colombia, Mexico & Venezuela Country Management Unit
Environmentally and Socially Sustainable Development Sector Management Unit
Latin America and the Caribbean Region
This document has a restricted distribution and may be used by recipients only in the
performance of their official duties. Its contents may not otherwise be disclosed without
Bank authorization.
CURRENCY EQUIVALENTS
Currency Unit
US$1
US$1
=
=
=
Mexican New Peso (N$)
N$9.5 (March 2001)
N$9.3 (September 2000)
FISCAL YEAR
January 1 – December 31
WEIGHTS AND MEASURES
1 meter (m)
1 kilometer (km)
1 hectare (ha)
1 metric ton (ton)
=
=
=
=
3.28 feet (ft)
0.62 mile (mi)
10,000 m2 = 2.47 acres
2,205 pounds
ABBREVIATIONS AND ACRONYMS
AGROASEMEX
ANAGSA
ANFA
ASERCA
BANRURAL
CNA
CNSF
DDR
FIRA
FONDEN
FONDOS
GOM
PET
PROCAMPO
PROGRESA
RE
SAGAR
SEDESOL
SHCP
Agricultural Insurance Company
National Agricultural Insurance Company
National Association of Insurance Funds
Support Services for Agricultural Marketing
National Rural Bank
National Water Commission
National Commission for Insurance
Rural Development District
Trust Fund for Agriculture
National Fund for Natural Disasters
Mutual Insurance Funds
Government of Mexico
Temporary Employment Program
Farm Direct Payment Support Program
Education, Health and Nutrition Program.
Rural Entrepreneurs
Secretariat of Agriculture
Secretariat for Social Development
Secretariat of the Treasury
Vice President:
Country Director:
Sector Director:
Sector Leader:
Task Manager:
David de Ferranti
Olivier Lafourcade
John Redwood
Adolfo Brizzi
Panos Varangis
ii
TABLE OF CONTENTS
EXECUTIVE SUMMARY________________________________________________ vi
Introduction ______________________________________________________________ vi
The Current Insurance Market ______________________________________________ vi
The Role of the Fondos ____________________________________________________ viii
Crop Insurance and Small Farmers ___________________________________________ ix
FONDEN and Agricultural Insurance __________________________________________ x
Insurance and Credit ________________________________________________________ x
Weather-based Index Insurance ______________________________________________ xi
Conclusions _______________________________________________________________ xi
Recommendations ________________________________________________________ xiv
1. BACKGROUND______________________________________________________ 1
Introduction _______________________________________________________________ 1
Methodology Used in the Study________________________________________________ 2
Rationale for Bank involvement _______________________________________________ 2
Organization of the Study ____________________________________________________ 2
2. THE CURRENT NATIONAL INSURANCE SYSTEM FOR THE RURAL SECTOR
IN MEXICO ___________________________________________________________ 3
Brief History of Agricultural Insurance in Mexico ________________________________ 3
The Agricultural Insurance Market ____________________________________________ 3
The Livestock Insurance Sector _______________________________________________ 9
Evaluating the Performance of AGROASEMEX _________________________________ 9
Subsidies for the Insurance Premium__________________________________________ 11
Dealing with Subsidies in Crop Insurance ______________________________________ 14
Regulatory Issues __________________________________________________________ 15
Who should regulate the Fondos?_____________________________________________ 16
3. THE EXPERIENCE OF FONDOS DE ASEGURAMIENTO IN PROVIDING
INSURANCE TO FARMERS ____________________________________________ 18
Background_______________________________________________________________ 18
The Role of AGROASEMEX in Promoting Fondos ______________________________ 19
Evaluating the Performance of the Fondos _____________________________________ 20
The Financial Viability of the Fondos _________________________________________ 23
Explaining Insurance and Re-insurance Pay-outs to and by the Fondos _____________ 27
Challenges for the Fondos ___________________________________________________ 29
iii
4. RISK MANAGEMENT AND CROP INSURANCE FOR LOW INCOME
FARMERS ___________________________________________________________ 31
Characterizing Small and Resource Poor Farmers_______________________________ 31
Sources of Income Risks Faced by Small/Resource Poor Farmers __________________ 34
How Small/Resource Poor Farmers Manage Risks to Their Income ________________ 35
Demand for Crop Insurance by Small Farmers _________________________________ 37
Alternative Programs for Risk Management by Small Farmers ____________________ 38
5. FONDEN AS A PROVIDER OF AGRICULTURAL INSURANCE ___________ 44
Background_______________________________________________________________ 44
How well does FONDEN protect small holders?_________________________________ 44
What is the market value of FONDEN coverage?________________________________ 47
Comparisons with parametric insurance _______________________________________ 51
Are there benefits to parametric triggers for FONDEN?__________________________ 52
6. INSURANCE AND CREDIT __________________________________________ 55
Background_______________________________________________________________ 55
Does crop insurance facilitate credit?__________________________________________ 57
Can insurance allow farmers to adopt riskier but more productive technologies? _____ 58
7. DEVELOPING WEATHER-BASED INDEX INSURANCE IN MEXICO ______ 59
Background_______________________________________________________________ 59
The International Experience ________________________________________________ 59
Preliminary Feasibility Study ________________________________________________ 60
Potential Application of Index or Parametric Insurance in Mexico’s Rural Sector ____ 62
Does Weather-based Index Insurance Require Subsidies? ________________________ 63
8. CONCLUSIONS AND RECOMMENDATIONS___________________________ 64
Conclusions _______________________________________________________________ 64
Recommendations _________________________________________________________ 66
REFERENCES ____________________________________________________________ 69
Appendix I: Fonden’s expected payouts from damages to crop due to drought. ____ 72
This report was prepared by Panos Varangis (Task Manager, DECRG), Don Larson
(DECRG), and Paul Siegel (HDNSP and FAO/CP). For its preparation, significant inputs
were provided by Reyes Altamirano (consultant), Chris Barham (LCSEN), Fabrizio
Bresciani (consultant), Juan Manuel Galarza (consultant), Vijay Kalavakonda (FSD),
Rodney Lester (FSD), and Jerry Skees (consultant). The report has also benefited greatly
from the comments of Adolfo Brizzi, Carlos Cuevas, Marcelo Giugale, John Pollner, and
Isabelle Tsakok.
iv
LIST OF ANNEXES AVAILABLE UPON REQUEST
A.
Report on the Preliminary Results of a Study for Mexico Rainfall Insurance, by Jerry
Skees, November 2000
B.
The Experience of Insurance Funds in Providing Crop Insurance to Farmers, by Reyes
Altamirano, March 2001
C.
Crop Insurance for Resource Poor Small Farmers, by Paul Siegel, March 2001
D.
Commodity Price Risk Management in Mexico: A Short Update on ASERCA’s Hedging
Program, by Panos Varangis and Paul Siegel, February 2001
v
EXECUTIVE SUMMARY
Introduction
1.
Rural households face a variety of risks to their livelihood. Chief among these are
risks related to agricultural production. In turn, the strategies that households employ to
manage risks affect how rural households allocate their time and resources. The
strategies affect technology choices and the productivity of household resources as well.
Crop insurance is market instrument that can supplement other ways of managing risk. It
is subsidized by the Government of Mexico and delivered, in part, by a governmentowned enterprise.
2.
This study describes the current market for crop insurance products. It reviews the
how crop insurance is delivered to farmers, how it is subsidized, what it costs and how it
is regulated. The report examines the whether crop insurance can facilitate credit and can
result in the adoption of productivity-enhancing technologies. The sustainability of
farmer-based organization, Fondos de Aseguramiento (fondos), that mutually insure
farmers is addressed.
3.
Crop insurance is not the only way of managing production risks and the subsidy on
crop insurance is not the only way the Government of Mexico helps farmers cope with
production risks. In this context, the study also examines the significance of subsidized
crop insurance for the rural poor relative to other programs that supplement rural incomes
and provide safety nets. FONDEN, a program that, among other features, provides free
insurance-like coverage to rural families against natural disasters is also reviewed.
4.
The report concludes with recommendations intended to limit current incentives
that encourage some types of risk-taking over others. Other recommendations presented
are intended to make better use of government resources by targeting the use of subsidies.
Ways in which the government can improve the institutions that provide oversight to the
industry and promote mutual insurance associations among farmers are presented. And
finally, the report recommends pilot tests of parametric insurance products that could
potentially lower the cost of insuring against weather-related risks.
The Current Insurance Market
5.
AGROASEMEX, the state crop insurance company, the farmer mutual insurance
funds (fondos) and four private companies comprise the national insurance system for the
rural sector in Mexico. From 1991 to 2000, crop insurance coverage has increased from
about 650,000 hectares in 1991 to about 2 million hectares. However, insurance
coverage is concentrated in the north of the country. Four states, namely Sinaloa, Sonora,
Chihuahua and Tamaulipas account for nearly 60% of areas insured.
6.
Generally, the private companies cater to larger clients. In terms of insurance
coverage for small and medium farmers, the share of insured areas of less than ten
hectares is around 50% for AGROASEMEX and the fondos and only 5% for the private
insurance companies. However, there are significant differences amongst the four private
companies and at least one private company appears to have a portfolio that is closer to
that of AGROASEMEX. Moreover, the provision of insurance to small and medium
vi
farmers through fondos is concentrated in the Northwest and Tamaulipas and the
coverage is mostly for irrigated lands.
7.
Broadly, the analysis of insurance markets gives some support to the hypothesis that
AGROASEMEX crowds out private companies and the fondos in providing direct
insurance to farmers. However, there are important exceptions such as the insurance
AGROASEMEX provides to small and medium size farmers of basic crops (like beans
and maize) in rainfed areas. Particularly in Central and Southern Mexico,
AGROASEMEX plays a significant role and shares the market with only one private
company. Administrative costs are higher than average in these areas and farmers
currently insured in Central and Southern Mexico are most likely to be adversely affected
as AGROASEMEX withdraws from offering direct insurance products.
8.
The situation in the livestock sector is dramatically different. AGROASEMEX
dominates the market, fondo participation is negligible and there is only one private
insurance company with measurable participation.
9.
All crop insurance premiums in Mexico are directly subsidized regardless of the
provider. In addition, the industry is indirectly subsidized through capital transfers to the
state-owned insurance company, AGROASEMEX. Smallholder farmers are also eligible
for publicly financed disaster insurance through FONDEN.
10. In absolute terms, premium subsidies favor larger farmers with high technology and
high value crops, particularly in irrigated areas. The current system of providing
subsidies as a percentage of the premium tends to proportionally favor geographical areas
and agricultural activities that are riskier. Current proposals to increase the percentage of
subsidized premium for less favorable areas will only make matters worse, by favoring
even more these riskier areas and activities. A less distortionary system of providing
subsidies should adhere to the following principles: 1) to the extent possible farmers
should be required to pay the pure premium risk of crop insurance; 2) if subsidies beyond
the pure premium are desired they should be introduced in a risk adjusted fashion so as to
not favor the higher risk farmers or regions; and 3) subsidies should foster the greatest
involvement from private markets. When farmers pay something that is equivalent to a
pure premium, the program will not favor the high-risk farmers or regions.
11. Historically, AGROASEMEX’s mandate included developmental objectives – for
example, fostering the development and regulation of fondos and targeting disadvantaged
farmers. However, AGROASEMEX was also expected to operate as a private profitable
company. The contradictory objectives and lack of clear focus are currently the subject of
debate and review in Mexico. The results of these deliberations will have fundamental
implications for the pricing of and coverage provided by the agricultural insurance
system in the future.
12. As the Government of Mexico considers relying more heavily on private insurers,
the industry has the decided advantage that the current regulatory system for crop
insurance is, on the whole, good. Additional proposed regulations, discussed in this
report, have the objective to strengthen the financial conditions of the fondos. While
regulations are of course needed to safeguard the financial viability of the fondos, they
should not be overly restrictive as to stifle the creation and growth of fondos. An issue to
vii
address is changes in the regulations to allow for certain types of financial insurance and
re-insurance products.
The Role of the Fondos
13. Fondos are democratically organized non-profit civil associations. They provide
mutual insurance benefits by collecting premiums from their members and purchase reinsurance on behalf of the members. When premiums collected, net of operational costs,
exceed indemnities paid out, the difference goes to a reserve fund to be paid out against
future liabilities.
14. A review of the experience and performance of the fondos shows that it is possible
to develop a mutual insurance scheme at the farmers’ level that is both financially viable
(even without subsidies) and efficient. The loss ratio of fondos is very good (around 56%
for 1990-2000), particularly for agricultural insurance, and comparable to the lowest of
the private companies. Nevertheless, the fondos have certain weaknesses and challenges
that they need to face in order to remain viable and expand further.
15. Fondos have a very high dependency on AGROASEMEX in terms of insurance
product innovation, reinsurance capacity and technical assistance necessary to operate.
The fondos need to start developing new products that better meet the needs of their
members otherwise they will keep loosing market share to private companies.
16. Members of the fondos are mainly small and medium size farmers with average size
of insured area of about 8 hectares. But, fondos have a very high geographic and product
concentration. Their concentration is particularly high in basic crops, especially grains,
that are experiencing low prices and reduced margins. Also, two-third of the area insured
with fondos is irrigated. Thus, the expansion of fondos may be limited to newly irrigated
areas.
17. Fondos could consider to diversify by moving to other insurance products, such as
life insurance, property insurance (machinery, building, equipment, etc.) that are
uncorrelated with crop losses and also have a lower and more stable loss ratios compared
to crop insurance. This however, will require the authorization of SHCP.
18. Even though the number of funds has been increasing and they have been
accumulating reserves, efforts to create a second level organization of fondos have not
yielded any results. The National Association of Insurance Funds (ANFA) is mainly
focusing its activities in lobbying with AGROASEMEX and the government.
19. The fondos could work towards harmonizing their administrative and accounting
systems. This harmonization will make easier to supervise the fondos and will facilitate
their integration in some regional federation or association.
20. To integrate the fondos in a national structure is very difficult because of the
diversity of conditions amongst the fondos and their performance. Thus, it can be
proposed that first try to create regional associations of fondos in relatively homogeneous
regions to obtain enough of geographic risk spread. The regional federation of the fondos
can then obtain re-insurance. The federation could develop technical services to provide
to the fondos and assist in the development of new products. The association could also
play a role in the fondo supervision (delegated supervision).
viii
Crop Insurance and Small Farmers
21. Small landholding is a somewhat imprecise characterization of small farmers.
Small producers can be defined as resource poor small farmers (basic crops, selfsufficiency with some small surplus, low technology, low asset base) and small
commercially-oriented farmers (marketable surplus, high technology, often irrigation,
good access to markets and institutions/associations).
22. Small farmers, particularly resource poor farmers, receive only a relatively small
part of their income, around 20%, from own farm activities (crop or livestock). And
although it could be argued that non-farm income is indirectly connected to broad farm
activities in the community or region, traditional crop insurance (which insures losses at
own farm) is not the most appropriate instrument. Small farmers are exposed to many
risks that can lead to income shortfalls. Surveys in Mexico indicate that risks such as
illness of the farmer and household members are no less important than either production
(yield) or price risks. Thus traditional crop insurance can provide only a very partial
coverage of the risks by resource poor small farmers. In addition, low commodity prices
and low margins for small producers of basic crops in rainfed areas make traditional crop
insurance relatively expensive.
23. If traditional crop insurance is neither the most appropriate nor an affordable means
for resource poor small farmers, what else can be done to assist them manage income
risks?
24. First, there are several government programs that assist small growers to increase
their income and expand their asset base (broadly defined). These programs (e.g.
PROCAMPO, PROGRESA, PET, Alianza para el campo, etc.) have important
implications for risk management by resource poor farmers and might even help some of
them become more commercially oriented over time.
25. Second, there is the potential to develop new parametric insurance instruments,
such as weather based index insurance, that could actually provide group risk coverage to
small farmers in a cost efficient way and without the moral hazard and adverse selection
problems of the traditional crop insurance. Alternatively, these index products could be
used by resource poor small farmers to insure against systemic risks off their farms - in
the broader rural economy (because of their dependence on off-farm agricultural and nonagricultural activities for income and employment).
26. Third, there are various safety net programs in Mexico that can assist resource poor
small farmers in the event of weather and/or economic shocks. In particular, there is the
disaster relief program - FONDEN - that provides insurance-like relief to small farmers
for catastrophic systemic risks, and also provides income supports through public works
projects.
27. Fourth, it is important for GOM to view existing (and future) programs described
above in the context of risk management for resource poor small farmers and a holistic
approach to rural risk management. For example, proposed changes in SAGAR's
PROCAMPO program for farmers with less than 5 hectares should help in their risk
management strategies. Also, there is a need to coordinate activities both within SAGAR
and between SAGAR and other GOM ministries and agencies. Moreover, there is a need
ix
for more transparent, objective, and well specified triggers and targeting mechanisms that
can provide resource poor more certainty and security with respect to the criteria and
benefits that they are entitled to receive.
FONDEN and Agricultural Insurance
28. FONDEN is a disaster relief program rather than an insurance program; however
the agricultural component of the program has characteristics similar to privately
provided crop insurance. For example, FONDEN makes payments in response to many
of the same perils covered by private insurers – for example, wind, drought and frost. In
addition, claims of loss trigger field inspections as do claims to crop insurers in Mexico.
Analysis shows that payments under FONDEN rules cover lost investments, but not the
value of the crop for most small farmers in Mexico.
29. There are differences between FONDEN coverage and crop insurance as well. The
program does not pay out based on damages accruing to individual farmers, but rather
when most farmers in a region suffer losses. In addition, while FONDEN rules provide
clear meteorological definitions of certain hazards covered by FONDEN, the process has
additional discretionary and political aspects.
30. Analysis also shows that expected benefits vary greatly among farmers. This
results from climatic differences among producing regions and from differences in crop
characteristics. Consequently, the rules provide differing incentives for farmers to raise
different crops. This distorting aspect of the program can be corrected by choosing
payout triggers so that payouts are the same for all producers.
31. Historic payouts related to crop damage average about N$330 million per year from
1997-1999. This however is only a portion of the payout related to rural incomes. For
example, payouts by SAGAR averaged about half of all drought-related payments from
FONDEN in 1997-1999.
32. Potentially, using parametric triggers – for example, linking payouts to weather
station data – could provide benefits to FONDEN. These benefits include equal
protection to all smallholder farmers; payouts can be made more promptly and at reduced
transaction costs; budgetary uncertain, which has been evident at the state level, can be
eliminated through reinsurance.
Insurance and Credit
33. Most of crop insurance is linked to credit. From the about 5.5 million hectares
receiving credit, at least 1.8 million are also covered by crop insurance. About 2.8
million hectares use PROCAMPO as a loan guarantee and thus do not need to obtain crop
insurance. For small farmers crop insurance is an important form of guarantee in order to
obtain credit. Increasingly BANRURAL and FIRA provide incentives for farmers to
obtain crop insurance together with credit. For farmers the benefit of linking credit to
crop insurance is that the latter could contribute in reducing the size of the collateral.
34. In principle, insurance should enhance the ability of borrowers to repay loans; this
is the motivation for current BANRURAL policy. However, empirical evidence is
lacking. Data at BANRURAL and FIRA that could possibly be used to measure the
x
relationship between production risks, insurance and loan default rates should be used to
further investigate the relationships between insurance and credit.
Weather-based Index Insurance
35. Weather-based index insurance is a relatively new insurance instrument whose
payouts are based on the occurrence of a weather event, rather than on actual crop losses.
The key advantage to this kind of insurance is that the weather or “trigger” event (e.g. a
rainfall shortage) can be independently verified, and therefore not subject to the
possibilities of manipulation which are present when insurance pay-outs are linked to
actual farm losses. And since the contracts and indemnity payments are the same for all
buyers per unit of insurance, the usual problems of moral hazard and adverse selection
associated with public crop insurance are lessened. Besides, the insurance would be easy
to administer, since there are no individual contracts to write, no on-farm inspections and
no individual loss assessments. This can help make the insurance affordable to a broad
range of people, including agricultural traders, shopkeepers and landless workers whose
incomes are also affected by the insured events.
36. Results from a feasibility analysis indicate that there is a good potential to develop
such insurance for certain regions in Mexico. The analysis was conducted in four states:
Durango, Jalisco, Tamaulipas and Zacatecas. The correlation between yields and rainfall
levels for 40% of the planted areas in these four states ranges between 60-80%. The
analysis also showed that rainfall contracts could reduce yield risk by at least 23% (9%
in Tamaulipas) for 40% of the planted area in these four states. A pilot project should be
considered to test whether such savings can be achieved in practice.
Conclusions
37. The market for crop insurance in Mexico is a market distorted by subsidies and
fiscal support to the government insurance company. The government has a legitimate
objective to provide resources to protect smaller and poorer farmers from severe income
losses due to disruptions in their production caused by weather events or other calamities.
At the same time, the government has to protect its fiscal resources dedicated to protect
farmers, particularly from catastrophic events.
38. The government is in the process of redefining its role in crop insurance. Proposed
legislation would preclude AGROASEMEX, the state agricultural insurance company,
from directly providing insurance to farmers and would convert the company to a reinsurer. The government also wishes to encourage the development of new private
insurance schemes and products that would deliver insurance at lower cost and thus
require less government support compared to the current system.
39. The crop insurance sector consists of AGROASEMEX, about 200 farmer mutual
insurance funds (fondos de aseguramiento or fondos) and four private insurance
companies. Crop insurance is often linked to credit and is a lending requirement for
loans from BANRURAL. Currently crop insurance is geographically concentrated in the
North and insures primarily irrigated areas. With the exception of one company, private
companies insure medium and large commercial farmers. AGROASEMEX, the fondos,
and one private company, cater to smaller farmers in addition to medium and larger
farmers.
xi
40. Insurance is subsidized both in terms of premium (30%) and in terms of delivery
costs. The premium subsidy applies to all crop insurance policies, while the latter is the
budgetary support of the government to AGROASEMEX. Despite acceptable loss ratios
(i.e., the ratio of indemnities paid over premiums collected), AGROASEMEX has very
high administrative costs. Administrative costs are high for two main reasons. First,
AGROASEMEX caters relatively more to smaller farmers of basic crops and in rainfed
areas. Second, AGROASEMEX has a dual role of a commercial entity and a
development organization. This latter role is for the support of the fondos in terms of
supervision and technical assistance, and the development of new insurance products.
41. Larger and medium size commercial farmers, can afford and do currently purchase
crop insurance if they see they need it. Many of these farmers are already clients of
private insurance firms. These farmers benefit most from current premium subsidies.
42. Delivering affordable crop insurance to smaller farmers, particularly for nonirrigated areas/crops, outside the North of the country is a key challenge, particularly as
AGROASEMEX is terminating the provision of direct insurance. But before addressing
this issue, one needs to examine the insurance needs of small farmers first. This will
provide a better idea of what are the appropriate instruments and delivery mechanisms for
these farmers.
43. It can be argued that land size alone is a rather poor criterion to classify farmers in
Mexico. There are commercially-oriented small farmers located in productive areas, with
high production technology, often using irrigation, and have other tangible and intangible
assets such as available infrastructure, existence of organization, etc. On the other hand
there are small farmers, in marginal areas, growing basic crops and using low technology.
Further, these small farmers have a low asset base, including weak producer
organizations. We call these farmers, resource poor farmers.
44. Traditional crop insurance may not be the most appropriate risk management
instrument for resource poor farmers. First, a large part of their income (over 50%)
comes from non-farm activities including off-farm labor. Second, most of the risks that
these farmers face are due to low demand for their labor services and because of illnesses.
Third, traditional crop insurance protects for lost assets and farm income, that small
resource poor farmers have little of. Given, that the traditional crop insurance may be of
limited help for these farmers, what are the appropriate insurance instruments for them?
45. The approach to provide insurance to resource poor farmers needs to be a holistic
one. The government of Mexico has several programs that have implications for risk
management for resource poor small farmers. For example, the PET (temporary
employment), PROGRESA (health, nutrition, education), Alianza para el Campo
(improve technology and productivity). Furthermore, resource poor farmers are impacted
by systemic risks, such as droughts and floods, that affect the economic livelihood of
their region. Thus, systemic, catastrophic insurance coverage is important to them, such
as FONDEN. At present, the criteria to trigger payments by FONDEN are rather ad hoc.
Making these criteria objective, that is based on the occurrence and the intensity of
certain weather events, will make the process more transparent and will increase the
certainty of receiving payments when a catastrophe occurs. In addition, if FONDEN
xii
payments are triggered using objective criteria, the government could obtain re-insurance
coverage and thus protect its fiscal resources dedicated to catastrophic insurance.
46. For the commercially-oriented small farmers, insurance is more appropriate. These
farmers are able to pay for such insurance, particularly because its linkage to credit, and
the adoption of improved production technologies. The challenge is how to deliver
insurance to many small farmers without incurring prohibitive administrative costs.
Traditional crop insurance requires individual farm inspection that can be costly and is
subject to adverse selection and moral hazard problems. The high administrative costs of
AGROASEMEX are partly due to their client base of small producers.
47. A key issue of delivering insurance to small farmers is the existence of producer
organizations. The farmer mutual insurance funds or fondos, mentioned earlier, provide
a good example. Fondos provide insurance to their members by pulling together
resources (premiums) to pay for future indemnities. Because indemnities can be high,
due to some systemic risk, fondos obtain re-insurance, currently through
AGROASEMEX. The latter also supervises the fondos and provides technical
assistance. The experience of the fondos so far has been very good. They have relatively
low loss ratios, making most of them viable even without government subsidies.
However, fondos have their limitations. They tend to concentrate amongst small farmers
in irrigated areas, and in regions with pre-existing good farmer organizations. As such,
fondos tend to concentrate in the North of the country (Sonora, Tamaulipas, Sinaloa).
Being mutuals, fondos tend to be very selective regarding their membership. Currently
there are around 70,000 farmers belonging to fondos covering an area of around 600,000
ha. Overall, one can conclude that while there is a scope for supporting the creation of
new funds and expanding existing ones, the fondo model is only a very partial solution
for insuring small farmers.
48. Catastrophic insurance coverage, through FONDEN, is useful for small farmers,
particularly those in rainfed areas. However, the coverage is for catastrophic events
while commercially oriented small farmers may wish to obtain coverage for less
catastrophic events; that is the events occurring with higher frequency. Weather
insurance could provide a useful coverage for such farmers. Weather insurance pays out
based on the frequency and intensity of a specific weather event. It can cover
catastrophic events but also less catastrophic ones. Because weather insurance depends
on the occurrence and objective measure of intensity of a specific event, it does not
require individual farm inspection, that can be very costly for small farms. Weather
contracts can be sold in standard units and appeal to a broader customer basis than just
farmers. Agribusinesses, rural financial institutions, input suppliers, farm workers could
purchase such insurance as long as the specific weather event triggers economic losses
for them. However, weather insurance relies on the correlation between farm yields and
weather events and can vary from farm to farm, thus exposing farmers to what is called
basis risk; that is the probability of incurring losses without the weather trigger. This
would limit the appeal of weather contracts. This implies that one should select areas to
offer weather insurance contract that are more or less homogeneous in terms of farm and
climatic conditions.
xiii
Recommendations
49. Current subsidies on premiums primarily benefit better-off farmers while resourcepoor farmers depend more on income-support and targeted safety-net programs.
Moreover, the current system of providing subsidies as a percentage of the premium
tends to proportionally favor geographical areas and agricultural activities that are riskier.
Some risks however are rare, catastrophic and expensive to insure. Such risks can affect
entire communities and the government may want to provide citizens with protection
against catastrophic risks. Consequently, this report recommends a less distortionary
system of providing subsidies guided the following principles: 1) to the extent possible
farmers should be required to pay the pure premium risk of crop insurance; 2) if subsidies
beyond the pure premium are desired they should be introduced in a risk adjusted fashion
so as to not favor the higher risk farmers or regions; and 3) subsidies should foster the
greatest involvement from private markets.
50. If, as a transitional measure, some continued subsidy is desired, a flat rate off the
pure premium should be used for all. For example, 1 or 2 points off the premium rate
would give everyone the same subsidy – that is, if the subsidy is 1 point, a farmer facing
a pure premium of $5 per $100 of liability would pay $4 and one facing a pure premium
of $10 would pay $9.
51. FONDEN could provide more effective coverage for farmers and rural people. The
government should consider using parametric triggers—for example, linking pay-outs to
weather station observations. Doing would result in similar protection for farmers in the
regions; pay-outs could be made promptly and at reduced transaction costs with greater
certainty for the beneficiaries as to whether and how much money they will be receiving
after a disaster. Furthermore, the government could benefit as the budgetary uncertainty,
which has been evident at the state level, can be significantly reduced through parametric
weather re-insurance.
52. The government could also have a role in the development of new private insurance
products, particularly for commercially-oriented small holders. As a next step, it is
recommended that the government considers a pilot project to test the concept of weather
insurance in Mexico. The pilot would require the cooperation between AGROASEMEX
and private companies interested in retailing weather insurance products to farmers
(again directly or indirectly). The pilot will require: a) evaluation of the infrastructure of
weather stations and weather data in selected regions; b) design of weather contracts; c)
demand assessment through interviews with potential clients; d) capacity building and
educational programs; and e) design of a monitoring and evaluation system to assess the
experience and recommend its replication.
53. Another area where the government could assist commercially oriented small
farmers is to support existing fondos and help form new ones. Further, the government
should investigate whether fondos could be strengthened by creating regional
associations of fondos. The creation of regional associations would provide greater
diversification through a greater pooling of geographic risk spread. Regional association
could potentially provide more effective supervision, allow for greater bargaining
strength for reinsurance and would allow fondos to otherwise pool their financial
resources for mutual benefit. The regional association of fondos could develop technical
xiv
services to provide to the fondos and assist in the development of new products. The
association could also play a role in the fondo supervision (delegated supervision).
54. Finally, the report recommends that BANRURAL and FIRA investigate the
relationship between crop insurance, production risk and default risk for agricultural
loans.
xv
MEXICO
CROP INSURANCE
1. BACKGROUND
Introduction
1. The Mexican national insurance system for the rural sector is undergoing important
changes. Proposals call for the removal of AGROSEMEX, the state agricultural
insurance company, as a direct provider of insurance to farmers and converting it into a
re-insurance company. The government also wishes to develop new insurance products,
expand the insurance coverage, particularly to small farmers, and develop ways to reinsure catastrophic risk in the agricultural sector.
2. Given the plans for reforms, the main objective of the present sector study is to
analyze the existing system of crop insurance, particularly focusing on small farmers, in
order to provide policy and operational recommendations for the insurance industry, to
develop viable and effective crop insurance schemes. The need for alternatives comes
from the high cost of the current insurance program and in particular the desirability of
finding insurance schemes that could lower the delivery costs to the farmers, reduce the
problem of moral hazard and increase the number of insured smaller and medium sized
farmers. It is worth noting from international experience that all traditional crop
insurance programs are subsidized. Consequently, an important issue is whether lessexpensive alternatives to traditional crop insurance are feasible in Mexico.
3. A specific objective of the study is the access of small farmers to crop insurance. The
study examines the limitations of the existing system to expand coverage to these
farmers. At the same time the study analyzes the risks small farmers face, the relative
importance of production (yield) risks, and risk coping mechanism farmers currently use.
Thus, the study would question the appropriateness (and affordability) of traditional crop
insurance for small farmers, particularly the poorer ones. The study then examines
various mechanisms to deliver risk management tools to small farmers by: a) analyzing
the experience of fondos as a potential channel and highlights its limitations; b)
examining the role of catastrophic insurance (FONDEN) as an effective mechanism to
compensate farmers for losses due to systemic events; c) presenting the implications of
other government programs (such as PROCAMPO, Alianza, PET, etc.) on small farmer
risk management. Finally, the study assesses the development of weather-based index
insurance and looks into its application for Mexico. This insurance can be a mechanism
for providing a low cost and effective risk management to rural people and institutions
for coping with weather events that have a severe impact on agricultural production and
in general rural incomes.
1
Methodology Used in the Study
4. The study uses various data sources, mainly form AGROASEMEX and SHCP, but
also has relied on information from the ejido survey, the 1995 survey of rural financial
markets, and sample crop budgets. The methodology for assessing the feasibility for
developing rainfall insurance is based on establishing the correlation between yields and
rainfall measures in municipalities within selected states and designing/rating prototype
rainfall insurance contracts for these municipalities. Finally, the study uses econometric
analysis (Tobit analysis) to explain insurance and re-insurance payments by the fondos.
Rationale for Bank involvement
5. The rationale for the Bank involvement is threefold. First, the Bank can bring
technical expertise based on worldwide experience on insurance issues, to analyze the
existing crop insurance system in Mexico and make recommendations for improving it.
Second, in early 2000, the Bank has initiated a project study to look into the feasibility of
developing weather-based index insurance as a viable and efficient alternative to many
crop insurance schemes in developing countries. The project study, which is financed by
the Development Marketplace and the Italian Trust Fund, indicates that such index-based
insurance holds some promising results. The Bank can bring this experience to analyze
the feasibility for developing weather-based index insurance in Mexico. Third, the Bank
completed a study on options for financing of catastrophic losses in Mexico under
FONDEN, in preparation of the recent National Disaster Management Project. Since,
under some circumstances disaster insurance covers crop losses, the two studies are
complementary.
Organization of the Study
6. This study is organized in seven chapters. Following the introduction, the second
chapter describes the existing system for agricultural insurance in Mexico. This chapter:
a) analyzes the type of insurance markets and clients of AGROASEMEX, the fondos and
the private companies for both the crop and livestock insurance; b) provides indicators
about the performance of AGROASEMEX; c) discuses the issue of subsidies for the
insurance premium including ways to make subsidies less distortionary; and d) presents
the regulatory issues. Chapter three provides a detailed analysis of the fondos, their role
in the agricultural insurance system and the challenges they face. Chapter four discusses
the risk profile of small farmers, shows existing mechanisms to cope with risks, and
describes government programs that assist in risk mitigation and coping. Chapter five
analyzes the role of FONDEN in agricultural insurance. It also indicates mechanism
based on parametric triggers that could make it more efficient and perhaps less costly for
the government. Chapter six briefly outlines the link between insurance and credit.
Chapter seven presents the concept of the weather-based index insurance and presents the
results from a feasibility study to develop rainfall-based index insurance for certain states
in Mexico.
2
2. THE CURRENT NATIONAL INSURANCE SYSTEM FOR THE RURAL
SECTOR IN MEXICO
Brief History of Agricultural Insurance in Mexico
7. Crop insurance in Mexico started in 1940, although it was not until 1961 that the
government created a government insurance agricultural company called ANAGSA in
order to provide insurance to all farmers that received credit from the official
(development) bank. The provision of insurance to farmers was the monopoly of
ANAGSA. The need to link insurance to credit was mainly because 80% of the financed
areas were rainfed, with high probability of losses due to climatic events. In addition,
crop insurance was considered a way to extend credit to more than 1.5 million farmers,
mainly ejidatarios.
8. During the period 1983-1988 almost 90% of the agricultural areas financed by the
development bank were also insured. The loss levels were quite high. Registered partial
losses accounted for 42% and total losses for 20% of the insured areas, respectively. This
high level of losses are explained by the relatively low levels of non-recoverable loans
that were on average around 10% over the same period. During 1983-88, about one third
of the recovered loan amounts to agriculture were from indemnities paid by crop
insurance. This percentage even reached 51% during the spring-summer 1988 crop
season. Thus, a significant part of the funds lent for agriculture through the development
bank was recovered by the government insurance company. This was an expensive and
non-transparent way to transfer funds to the farmers.
9. The agricultural insurance system under ANAGSA proved to be very expensive and
fiscally unsustainable. Between 1983 and 1988, insurance paid indemnities totaling
around US$4 billion while farmers paid a total of US$1.3 billion in premiums. This
created losses of US$2.7 billion for ANAGSA and the of course ultimately for the
government. And this figure does not include the operating costs of ANAGSA. In 1990
the government decided to eliminate ANAGSA and replace it with a new government
company, AGROASEMEX. The objective of the new national insurance system for the
rural sector in 1990 was to reduce the scope of insurance coverage, to focus more on
irrigated and good rainfed areas, and to allow other agents to provide crop insurance in
addition to the government company.
The Agricultural Insurance Market
10. The current national insurance system for the rural sector consists of
AGROASEMEX, the state-owned insurance company established in 1990 to succeed
ANAGSA, the Fondos de Aseguramiento, or Insurance Funds, and private insurance
companies. AGROASEMEX provides direct insurance to farmers and reinsurance to the
Fondos and in very limited cases to private companies. A Fondo is a group of farmers in
a more or less homogeneous area that provides mutual insurance to its members.
Because of the limited capacity to absorb systemic risk, the fondo reinsures itself through
AGROASEMEX. In addition, AGROASEMEX supervises the fondos and provides
technical assistance, for example in management, technical expertise, training, etc. There
are 295 fondos of which in 1999 there were 195 in operation and 200 in 2000. The
experience of the fondos in Mexico will be presented and analyzed in a separate section
3
of this report. In addition to AGROASEMEX and the fondos, there are about four main
private insurance companies providing crop insurance in Mexico. These are Proagro ,
General de Seguros, Comercial America, and ACE Seguros. Although there are over
ninety private insurance companies in Mexico, only these four offer crop insurance.
Private insurance companies usually reinsure themselves in the international market.
11. The market shares of AGROASEMEX, the fondos and the private companies are
presented below. Table 2.1 shows the market shares in terms of hectares and number of
animals insured, while table 2.2 presents the same results in terms of shares in the total
insurance premiums. Table 2.1 shows a steady increase in the insured hectares and
animals since 1991. It also shows a steady growth of the share of private insurance
companies until 1999. However, AGROASEMEX and the fondos account for a large
share of the market and since the fondos reinsure through AGROASEMEX, this latter
accounts for a significant portion of Mexico’s agricultural insurance/re-insurance
market.1 Table 2.1 also shows that to a large extent the expansion of agricultural
insurance is coming from the private companies, although part of the expansion of the
private companies comes at the expense of AGROASEMEX and perhaps the fondos to a
lesser extent. Comparing table 2.1 and 2.2 shows that the private companies have a much
larger share in terms of premium collected compared to hectares insured. This indicates
that the private companies concentrate to insuring higher value crops compared to
AGROASEMEX and the fondos.
Table 2.1: Insured crops and livestock and share by type of firm, 1991-2000
Agriculture
AGROASEMEX
Fondos
Private
Livestock
AGROASEMEX
Fondos
Private
1991
1992
1993
1994
643.6
1,014
975
1,104
54.9
45.1
-
45.2
54.8
-
47.6
52.4
-
45.4
54.1
0.5
575.7
769
710
922
64.2
35.8
-
75.5
24.5
-
94.6
5.4
-
99.2
0.8
-
1995
1996
Thousand hectares
906
1,353
percentage
43.4
50.0
54.0
45.2
2.6
4.8
1997
1998
1999
2000
1,586
1,605
1,697
1,948
39.8
41.3
18.9
37.6
39.2
23.2
31.1
36.1
32.8
40.4
30.9
28.7
3,835
5,168
9,725
96.8
0.1
3.1
71.3
0.0
28.7
72.0
1.4
26.6
Thousand heads
847
2,459
3,082
percentage
99.5
99.9
97.0
0.5
0.1
0.1
2.8
Source: AGROASEMEX
1
Fondos cede about 22-25% of collected premiums towards reinsurance from AGROASEMEX.
4
Table 2.2: Collected premiums and share by class of firms, 1991-2000
1991
1992
1993
1994
131.3
191.4
207.2
233.2
AGROASEMEX
Fondos
Private
68.8
31.2
-
52.0
48.0
-
50.2
49.8
-
49.3
50.5
0.2
Livestock
28.7
24.0
19.0
22.8
AGROASEMEX
Fondos
Private
90.6
9.4
-
86.5
13.5
-
95.5
4.5
-
98.6
1.4
-
Agriculture
1995
1996
Million pesos
260.3
451.3
Percentage
42.7
41.6
54.9
53.0
2.4
5.4
Million pesos
28.7
67.6
Percentage
99.5
99.8
0.5
0.2
-
1997
1998
1999
2000
681.7
840.9
811.6
754.2
33.1
43.3
23.6
29.8
38.3
31.9
22.9
33.6
43.5
28.6
34.3
37.1
111.2
156.4
187.9
343.7
96.8
0.2
3.0
80.3
0.2
19.5
71.4
0.2
28.4
59.3
2.8
37.9
Source: AGROASEMEX
12. The present crop insurance system in Mexico focuses on agricultural regions with
productive potential and financial viability. Subsistence and poor non-commercial
farmers are supposed to be covered through the government’s national disaster scheme
called FONDEN. Thus, from the 21.9 million hectares of agricultural land in Mexico, the
commercially insurable area is 8.8 million ha. From these, in 2000, 1.95 million hectares
were insured. Thus, insurance covers about 22% of the total insurable agricultural area or
9% of the total agricultural area. In comparison, the share of insured area is 45.9% in the
US, 54.7% in Canada, 42.5% in Spain, and 79.3% in Japan. This indicates a low
penetration of insurance in Mexico. Insurable area has increased from 643,500 hectares
in 1991 to 1.7 million in 1999 and 1.95 million in 2000. About 55.7% of the insurance
(in terms of area coverage) goes to four states namely Sinaloa, Sonora, Chihuahua and
Tamaulipas. These are states with relatively large share of commercial agriculture. To a
very large extent insurance is linked to credit.
13. AGROASEMEX, the fondos and the private insurance companies have a somewhat
different clientele. Table 2.3 shows the size of the insurance market in 2000 by
technological level (irrigated versus non-irrigated), groups of crops (basic versus other)2
per insurance entity and source of insurance: AGROASEMEX, fondos and private
companies. In terms of hectares insured, AGROASEMEX has a higher share in rainfed
areas (74%) and only 26% in irrigated areas. Two-thirds of the areas insured by the
fondos are irrigated while the private companies operate 45% in irrigated areas and 55%
in rainfed. Within the private companies there are some notable differences.
Commercial America mainly insures irrigated areas (73%), while Proago has a higher
share in rainfed areas (64%). Overall, irrigated areas account for 44% of the total area
insured while rainfed accounts for the reminder 56% where insurance is mostly needed.
However, in terms of premiums collected 72% are in irrigated areas (higher investments
and value of crops) while 28% are for rainfed areas. For all insurance entities,
AGROASEMEX, fondos and private companies, the focus is on basic products (grains,
oilseeds, etc.).
2
The following eleven crops are considered to be basic crops in Mexico: maize, beans, rice, cartamo (an
oilseed), sorghum, soybeans, wheat, sesame, cotton, barley, and coffee.
5
Table 2.3: Insurance by Technological Level and Group of Crops in 2000
Technology
Irrigated Rainfed
Type of crop
Basic
Other
Total
Area Insured (hectares)
AGROASEMEX
786,710 205,114 581,596 712,146 74,564
Fondos
602,665 398,679 203,986 529,100 73,565
Privates
558,969 253,314 305,655 462,654 96,315
Proagro
385,364 137,822 247,542 321,407 63,957
Gral. Seguros
62,520 35,377
27,143
57,308
5,212
Com. America
108,115 78,565
29,550
81,487 26,628
ACE Seguros
2,971
1,550
1,421
2,453
518
Total
1,948,344 857,107 1,091,237 1,703,900 244,444
Premiums Collected (millions of Pesos)
215,651 125,491
90,160 129,364 86,287
AGROASEMEX
Fondos
258,680 201,722
56,958 212,395 46,285
Privates
279,915 215,258
64,657 140,588 139,327
Proagro
129,080 83,048
46,032
81,431 47,649
Gral. Seguros
23,589 19,311
4,278
18,624
4,965
Com. America
124,686 110,670
14,016
39,507 85,179
ACE Seguros
2,559
2,230
329
1,026
1,533
Total
754,246 542,471 211,775
82,347 71,899
Source: AGROASEMEX
Shares of total
Irrigated Rainfed Basic Other
26%
66%
45%
36%
57%
73%
52%
44%
74%
34%
55%
64%
43%
27%
48%
56%
91%
88%
83%
83%
92%
75%
83%
88%
10%
12%
17%
17%
8%
25%
17%
13%
58%
78%
77%
64%
82%
89%
87%
72%
42%
22%
23%
36%
18%
11%
13%
28%
60%
82%
50%
63%
79%
32%
40%
64%
40%
18%
50%
37%
21%
68%
60%
36%
14. AGROASEMEX claims to cater relatively more to smaller and medium income
farmers that are commercially oriented and thus have a productive potential. The fondos
require a group of well-organized farmers that can be of any size, but most often are
small-to-medium income farmers, more often located in irrigated lands. The average size
of insured area through a fondo is around 8 hectares. The private companies mainly cater
to larger commercial farmers, although at least one private company claims to cater to
associations of small/medium size producers. There is no specific information on the
client base of each individual private company.
15. Table 2.4 below presents data for 2000 on the size of the area that purchase direct
insurance from AGROASEMEX, the private companies and the fondos. The average
insured area by AGROASEMEX and the fondos is similar, 7.3 hectares and 8.1 hectares
respectively. The average insured area by the private companies is much larger at 59.1
ha. The number of farmers insuring less than 10 hectares as a share of the total number
of farmers is 88% for AGROASEMEX and 52% for the private companies. There are no
such data on the fondos, but it is assumed that the share of farmers with fewer than 10
hectares insured is not that different from AGROASEMEX. In terms of total land
insured, 45.4% falls in the category of 0-10 hectares for AGROASEMEX versus only
4.7% for the private companies.
6
Table 2.4: Size of Insured Area and Number of Farmers Insured
Size of holding
Total Area
hectares
AGROASEMEX (OI 99/00 and PV 2000/00)
0-10 ha
349,951.81
10-10,147 ha
420,063
Total (ha)
770,014.81
Private Companies (FY 2000)
0-10 ha
27,896
10-14,856 ha
567,419
Total (ha)
595,315
Fondos (FY 2000)
Total (ha)
567,700
Source: AGROASEMEX
Producers
Number
Average Area
hectares
Producers
number
Area
% of total
92,905
12,654
105,559
3.8
33.2
7.3
88.0%
12.0%
100%
45.4%
54.6%
100%
5,208
4,864
10,072
5.4
116.7
59.1
51.7%
48.3%
100%
4.7%
95.3%
100%
69,700
8.1
NA
NA
16. In terms of geographical distribution (see table 2.5), AGROASEMEX has a large
percentage of insured areas in the North and South of the Mexico. Fondos are dominant
in the Northwest while the private companies have a high share of insured hectares in
Central Mexico. However, the private companies have a large share of premiums
collected in all areas except the Northwest, an indication that they tend to concentrate
their activities on high value crops and agricultural activities. States where the fondos
dominate (about 50% or more of the market) are Sinaloa (72%), Sonora (54%),
Tamaulipas (49%) and also Nayarit (42%). These four states and Baja California account
for most of the areas insured with Fondos. Private companies dominate (more than 50%)
in San Luis Potosi (98%), Campeche (96%), Aguascalinetes (85%), Nuevo Leon (81%),
Puebla (61%), Guerero (58%), Guanajuato (55%), Tabasco (55%), and Veracruz (52%).
AGROASEMEX has a dominant share (more than 50%) in Txlaxcala (99%), Durango
(96%), Estado de Mexico (94%), Morelos (93%), Queretaro (91%), Coahuila (87%),
Hidalgo (81%), Oaxaca (77%), Zacatecas (73%), Chiapas (52%), Yucatan (52%) and
Quintana Roo (52%).
7
Table 2.5: Market Distribution of Agricultural Insurance by Region in 2000
Regions
Firms
Center
AGROASEMEX
Fondos
Private
Total (million N$)
AGROASEMEX
Fondos
Private
Total (million N$)
AGROASEMEX
Fondos
Private
Total (million N$)
AGROASEMEX
Fondos
Private
Total (million N$)
AGROASEMEX
Fondos
Private
Total (million N$)
Northwest
North
South
National
Participation (%)
Area
Premiums
Percent of Total
Percent of Total
38.5
42.0
18.4
14.3
43.1
43.8
335.8
159.6
22.4
21.1
58.8
51.5
18.8
27.4
565.5
327.5
47.3
26.3
25.5
28.7
27.2
45.0
835.7
197.3
58.5
47.6
4.3
5.8
37.2
46.6
211.4
69.7
28.6
40.4
34.3
30.9
37.1
28.7
1,948.3
754.2
Source: AGROASEMEX
17. In terms of crops, the fondos have a significant market share (around 50% or more of
insured area) in chickpeas (garbanzo, 69%), wheat (52%), cartamo (an oilseed, 49%),
coffee (49%) and sorghum (46%). The private companies concentrate in apples (92%),
tomatoes (75%), broccoli (74%), sugar (73%), soybeans (63%), rice (55%), cotton (54%)
and bananas (53%). These are high value crops (fruits and horticulture) or
industrial/plantation crops. AGROASEMEX has significant activities in sesame (86%),
beans (74%), tobacco (72%), barley (71%), mango (71%), lemon (69%), potato (59%)
and oranges (52%). Thus, AGROASEMEX portfolio in terms of insured crops mixes
high value commercial crops, such as fruits and horticulture, industrial crops (tobacco)
with some basic crops (e.g. beans). In terms of crop categories the portfolio of
AGROASEMEX does not appear to be very different compared to the privates and the
fondos.
18. In conclusion. Although fondos have small and medium farmers as clients they very
much concentrate their insurance coverage in terms of geographic locations (Northwest
and Tamaulipas) and in terms of irrigated areas. At least some private companies appear
to have a portfolio that is closer to that of AGROASEMEX, but there are significant
differences amongst the four private companies. Of the four private companies, it is
Proagro that has a portfolio more similar to that of AGROASEMEX.
19. The data presented above give some support to the hypothesis that AGROASEMEX
crowds out, at least partly, private companies and the fondos in providing direct insurance
to farmers. However, there are certain market segments, such as small and medium size
farmers of basic crops (like beans and maize) in rainfed areas, particularly in Central and
Southern Mexico, where AGROASEMEX plays a significant role and where perhaps
8
only one private company, if that, would be able to service these clients if
AGROASEMEX stops offering direct insurance. The main reason that certain market
segments such as the one mentioned above, are not likely to be attractive to private
companies is mainly due to the high administrative costs.
The Livestock Insurance Sector
20. The situation in the livestock insurance sector is dramatically different compared to
agricultural (crop) insurance described above. As table 2.6 below indicates
AGROASEMEX has the largest market share in both number of animals insured but also
in premiums collected. The private companies focus more on high value livestock
activities such as insuring animals in stables, feedlots, corrals, etc. thus they have a larger
share in premiums collected compared to number of animals insured. Unlike crop
insurance, the fondos have an insignificant share in the livestock insurance market.
Table 2.6: Livestock Insurance during 2000
Firms
AGROASEMEX
Fondos
Private Companies
Total
Number of animals
7,005,172
135,280
2,584,694
9,725,146
Market share
72.0%
1.4%
26.6%
Premiums
203,653
9,776
130,270
343,699
Premium Share
59.3%
2.8%
37.9%
100%
Note: premiums in thousand pesos
Source: AGROASEMEX
21. The private companies are also very concentrated geographically in the North of
Mexico accounting for 75% of the livestock insurance market. In the center, northwest
and the south, AGROASEMEX accounts for 85%, 93% and 97% respectively of the
livestock insurance market. Furthermore in states such as Chiapas, Guanajuato,
Guerrero, Morelos, Oaxaca, Sinaloa, and Tlaxcala, the market share of AGROASEMEX
is almost 100%. Fondos for livestock insurance concentrate in very few states namely in
Chihuahua and Queretaro and have negligible activities in livestock insurance outside
these two states. Thus, unlike crop insurance, the departure of AGROASEMEX from a
direct provider of livestock insurance is likely to leave a large market segment without
insurance, at least in the near term.
Evaluating the Performance of AGROASEMEX
22. AGROASEMEX offers a wide range of insurance products that cover agriculture,
livestock, life and casualty, although most of the activity is concentrated in agriculture
and livestock. With respect to crops, insurance covers climatic risks (frost, excess
humidity, flood, hurricane, cyclone, tornado and wind storms, hail, drought, fire, heat
wave), biological risks (diseases) and pre-germination risks (riesgos relacionados con la
nacencia) that make planting impossible. Insurance provided by AGROASEMEX covers
actual loss or damage to the crops caused by the risks the producer has purchased
insurance for. AGROASEMEX has a very high technical capability in assessing losses
and damages. However, the system of individual crop inspection, particularly dealing
9
with smaller and medium size farmers, imposes significant administrative costs and also
exposes the company to the moral hazard and adverse selection problems.
23. AGROASEMEX has a very expensive overhead structure having about 1000 staff,
including 8 regional offices and 74 local offices, and a significant part of this, is allocated
to field inspections. For reasons that are not entirely clear every insured field has to be
inspected each year. AGROASEMEX has local staff and offices even in locations with
relatively little insured area. The operational costs of AGROASEMEX are thus
significantly higher compared to private companies for three main reasons. First,
AGROASEMEX deals relatively more with smaller and medium size farmers and
individual field inspection imposes high costs. Second, it supervises and provides
technical assistance to the fondos. Third, it bears the costs for developing new products
and markets for agricultural insurance that the private sector would eventually enter.
24. The government subsidizes the premium cost of the insurance by 30% (see below for
details). In addition to the 30% premium subsidy, the government provides financial
support to AGROASEMEX for its operational costs. The support to operational
expenses has been reduced from over 50% in mid-1990s to 13% in 1999 and to 11% by
2000. During the period 1990 to 2000 the cumulative cost of crop insurance to the
government has been 4.89 billion pesos (in 2000 terms) of which 3.27 billion went to
subsidize the premiums and 1.62 billion went to support the operations of
AGROASEMEX. Despite the government cash infusions to support the operations of
AGROASEMEX, the company has registered losses during most of the years 1990-2000.
25. For the period 1991-2000, AGROASEMEX has a loss ratio3 for agricultural
insurance of 81.4% and 63.8% for livestock insurance. Overall, the loss ratio is 75.7%.
The loss ratio is higher compared to the fondos for the same period estimated at 56.5%.
The overall loss ratio of AGROASEMEX is within a reasonable range for agricultural
insurance. However, given the high administration costs of AGROASEMEX this loss
ratio leads to financial losses over time and requires additional funds from the GOM.
26. Calculations of the financial performance of crop insurance in Mexico indicate that
the ratio of the average administrative costs plus the average indemnities paid over the
average premiums paid is approximately 2.2. This means that for every dollar of
premiums received the cost (admin cost plus damages paid) is $2.2. By comparison,
similar calculations for other countries show that the ratio of financial performance are
for Brazil 4.6, for Costa Rica 2.8 and 2.4 for the US (see Skees, Hazell and Miranda,
1999). It is of interest to point out that this ratio was 3.7 for Mexico during the period
1980-89, that is prior to the creation of AGROASEMEX (same source as above). This
indicates that while the cost of crop insurance system in Mexico is still expensive, it has
been coming down particularly in comparison to the pre-AGROASEMEX period.
27. AGROASEMEX has a difficult joint role as both a development organization and
reinsurer (mainly in support of the fondos and new product development) and as a
commercial direct insurer. AGROASEMEX does not keep separate accounts to
3
A simple definition of loss ratio is the ratio of indemnities paid over premiums collected. A loss ratio
above 100% means that the insurance company is loosing money. The loss ratio presented here does not
include administrative costs. Thus, even with a loss ratio below 100% the insurance company may be
loosing money if it has high administrative costs.
10
distinguish the cost of operating as a development institution and the cost of the
insurance/re-insurance company4. This dual role of AGROASEMEX, developing the
market for the fondos and the private sector but also competing, creates a conflict of
interest. In particular, by supporting the fondos, AGROASEMEX is strengthening
competitors to its direct insurance business. Thus, there is no incentive to create more
fondos. While AGROASEMEX is supported in both capital and expenses by the
government it is required to minimize fiscal drain and thus has an incentive to retain its
market position. This has resulted in AGROASEMEX encroaching into market segments
that the private insurance companies or the fondos could serve. But, if AGROASEMEX
had not moved into these more lucrative market segments its financial results would have
been worse. Thus, avoiding crowding out the private companies and fondos causes a
larger fiscal drain for the government. There have been allegations that AGROASEMEX
has been crowding out the private insurance companies and that at times it reduced
premium levels significantly in order to retain or increase its market position in a number
of states. In addition, there is some indication that at least some of the fondos are now in
a position to deal directly with the world reinsurance markets, and have in fact been
having discussions with international reinsurers.
28. A further important factor is that agricultural policies create some moral hazard
problems in the crop insurance market in Mexico. This is because farmers are able to
still buy insurance even when it is clear that rainfall will be poor (especially in El Nino
years) and cereal farmers under the PROCAMPO scheme are virtually required to plant
under the current incentive system. There is also anecdotal evidence that
AGROASEMEX has extended the insurance deadline for social welfare purposes on a
number of occasions. While reinsurers adjust pricing in anticipation of rainfall patterns,
it appears that they are looking for a return over a full (approximately 5-year) rainfall
cycle, rather than every year.
Subsidies for the Insurance Premium
29. Subsidies for the insurance premium started in 1991 supporting 20% of the premium
cost and were exclusively for the direct crop insurance sold only by AGROASEMEX. In
1992 the subsidy increased to 30% and was also extended to insurance by the fondos. In
1994 the private insurance companies became eligible for the subsidy but only if they
were re-insured through AGROASEMEX. In 1995, the subsidy also applied to livestock
insurance. In 1996, the subsidy to the private insurance companies became independent
from the re-insurance through AGROASEMEX. In that year, rules were established for
channeling of subsidies to the private companies.
30. At present, AGROASEMEX, the fondos and the private companies have equal access
to the premium subsidy for their clients but AGROASEMEX distribute the government
subsidy. In other words, the private companies ask AGROASEMEX to pay the amount
of subsidy corresponding to the number of contracts/coverage they have offered. For the
fondos, AGROASEMEX gives the difference between the total amount of premium
4
For example, during the period 1990-2000, AGROASEMEX has conducted a total of 353 training
seminars, benefiting 1,724 fondos and trained 8,864 people. It is preliminary estimated that the total cost
for 29 courses in 2000 was around 2 million pesos without including the cost of AGROASEMEX staff
presenting the courses. Including this latter, could easily double the total cost figure of 2 million pesos.
11
subsidy and the cost of re-insurance (the fondos only re-insure with AGROASEMEX as
of this writing). The private companies need to provide to AGROASEMEX monthly
records of all their paid policies in order to receive the subsidy payment corresponding to
the premiums collected.
31. The rules about the subsidies for the premium indicate that at least 50% of the total
amount given should be for basic crops and again at least 50% should be for protecting
investments. For 2000, basic crops received 64% of the total premium subsidies and
insurance to protect investments received 91% of the subsidies. Starting from the 200001 fall-winter (otono-invierno or OI) crop cycle the amount of subsidy was capped at a
maximum 2,000 pesos per hectare.
32. During 2000, 225 million pesos (around $22.5 million) were spent to subsidize
insurance premiums for agriculture. Of these, 28.5% were for the direct insurance by
AGROASEMEX, 34.3% for the insurance by the fondos and 37.3% for the insurance
sold directly by private companies. Table 2.7 shows the application of subsidies per
entity (AGROASEMEX, fondos, private companies) during 2000. By examining this
table, the following observations can be made5. In terms of subsidies applied per insured
hectare, the largest beneficiaries are the other crops (fruits, horticultural, industrial crops),
and irrigated crops. The private companies and the fondos receive much higher subsidies
per hectare they insure compared to AGROASEMEX. This is a reflection of the lower
peso value per hectare that AGROASEMEX insures in comparison to the fondos and the
private companies. Because AGROASEMEX also uses less subsidies per hectare, it can
insure more hectares per 1,000 pesos of subsidies. The lower values per hectare insured
show that 1,000 pesos of subsidies can cover significantly more hectares for basic, nonirrigated crops compared to irrigated, higher value crops.
33. Looking at the share of the subsidy over the insured value per hectare shows that
while there is not much difference between basic and other crops, this share is
significantly lower for irrigated areas compared to rainfed areas. This is because
insurance premiums for irrigated areas are lower compared to rainfed areas. For 2000,
the loss ratio (indemnities over premiums) was 38% for irrigated areas and 62% for
rainfed indicating that the latter are indeed riskier to insure.
5
These results are mostly indicative as they are based on one year of observations (2000) for which data
was available.
12
Table 2.7: Application of Subsidies for Insurance Premium during 2000
Subsidies Applied per Insured Hectare (pesos/ha.)
Total
Group of Crops
Level of Technology
Other
Irrigated
Rainfed
Basic
53.9
343.6
183.5
46.5
AGROASEMEX
81.4
Fondos
128
120.4
182.2
151.8
83.8
Private
150
91.2
432.6
254.9
63.5
115.5
84.7
330.1
189.9
58.2
Total
Insured Area per 1000 pesos of Subsidy to the Premium (ha.)
AGROASEMEX
12.3
18.5
2.9
5.4
21.5
Fondos
7.8
8.3
5.5
6.6
11.9
Private
6.7
11
2.3
3.9
15.8
8.7
11.8
3
5.3
17.2
Total
Insured Value per Hectare (pesos/ha.)
AGROASEMEX
3,511.3
2,312.6
14,959.6
9,523.3
1,390.6
Fondos
5,155.9
4,862.9
7,263.7
6,639.8
2,255.9
Private
5,317.1
3,429.3
14,385.6
9,428.9
1,909.4
4,538.1
3,407.7
12,417.4
8,154.2
1,697.7
Total
Subsidy as a Share of the Insured Value per Hectare
AGROASEMEX
2.3%
2.3%
2.3%
1.9%
3.3%
Fondos
2.5%
2.5%
2.5%
2.3%
3.7%
Private
2.8%
2.7%
3.0%
2.7%
3.3%
2.5%
2.5%
2.7%
2.3%
3.4%
Total
Source: AGROASEMEX and authors’ calculations
34. Examining the share of indemnities paid to subsidies received for the year 2000, it
shows that $1 worth of subsidies corresponded to $1.48 of indemnities paid. But there
are differences here. AGROASEMEX and the fondos paid more indemnities per $1 of
subsidies received compared to the private companies. This shows that AGROASEMEX
and the fondos are involved in riskier activities compared to private insurance companies
as a total (there may be differences between these companies). Also, rainfed areas pay
out more indemnities per $1 of subsidies received compared to irrigated areas.
35. Table 2.8 below calculates the loss cost for 2000. This reveals which group has the
largest relative risk in insurance. Loss cost is defined as the sum of indemnities paid plus
subsidies over the overall amount of liabilities (or insured value). The result confirms
again that rainfed areas are by far the most risky areas in terms of crop insurance. Also,
AGROASEMEX and the fondos are in riskier activities/areas compared to the private
companies. Finally, there is a small difference in terms of riskiness between basic crops
and other (mainly horticultural), with the former being less risky than the latter.
13
Table 2.8:Estimates of Loss Costs for 2000
All
Basic
Other Irrigated
AGROASEMEX 6.9% 6.8%
7.0%
5.3%
Fondos
6.4% 6.3%
6.9%
5.4%
Private
5.7% 5.0%
6.6%
4.9%
Total
6.3% 6.1%
6.8%
5.2%
Source: AGROASEMEX and authors’ calculations
Rainfed
10.6%
12.3%
8.9%
10.5%
36. In absolute terms, the premium subsidy applies to all types of producers, small,
medium and larger. However, as crop insurance is more appropriate for commercial
agriculture, is mostly linked to credit and as commercial farmers are those who likely
receive such credit, it could be said that premium subsidies benefit more the
commercially oriented farmers in Mexico and benefit very little if at all, the less
commercial and poorer farmers.
Dealing with Subsidies in Crop Insurance
37. The analysis of the premium subsidies presented above indicates that the present
system of subsidies as a percent of the premium tends to favor proportionally more
agricultural activities and geographical areas that are riskier. The new administration in
Mexico has begun to reconsider the issue of subsidies in crop insurance. What follows
presents a framework for rethinking subsidies in crop insurance with the aim of reducing
distortions and fiscal costs.
38. Several aspects of subsidies should concern any government. When the cost of risks
are not internalized, decision makers will always take on more risk and move to a higher
threshold of risk once a subsidy is introduced. In the case of farmers who move to a
higher risk threshold in riskier areas, society ends up paying for the higher losses.
Further, as with any explicit subsidy, crop insurance subsidies will ultimately get bid into
assets, favoring the land owners, so that subsidies create an over investment in the
subsidized enterprises.
39. Unlike price subsidies that are related to a direct per unit value of production,
subsidies for crop insurance that are fixed as a percent of premium, directly favor higher
risk producers, crops and regions. This is a very important difference that has received
far too little attention in countries such as the U.S. and Canada. In short, every producer
faces similar price risk within the same country. Price movements are basically very
highly correlated. While yield movements may also be highly correlated within regions,
not every producer faces the same yield risk. Yield risks are a function of soils,
management and weather. Consider one producer who would be charged $20 per $100
value (or liability) for an unsubsidized crop insurance product that pays for losses below
70 percent of his average versus another who pay $10 per $100 of value for the same
policy. If the insurance is subsidized at 30 percent of the premium costs, the high risk
producer obtains a $6 per $100 value subsidy versus $3 per $100 value for the low risk
producer.
40. Subsidies result in rent seeking in a political economy context. Insurance products
are complex. Complexity means that the private sector reinsurance and insurance
companies who understand insurance will be in a favored position in advising
14
government about how to structure the subsidies. Designing of the insurance products
and the structure of the government relationship may be primary beneficiaries. This has
been the U.S. experience. Given recent legislative changes, private companies will
receive one third of the subsidies that are put into the risk management programs in the
US. In the U.S., the farmer now pays about 25 percent of the total cost of these
programs. Each time crop insurance reform is past the government does so with an
explicit goal of eliminating ad hoc disaster aid that has become so common when there is
a crop failure. Still, ad hoc disaster aid has not stopped.
41. Despite the many concerns discussed above, subsidies can still be introduced in crop
yield insurance in a fashion that avoids many of these problems. Three principles should
be followed: 1) to the extent possible farmers should be required to pay the pure premium
risk of crop insurance; 2) if subsidies beyond the pure premium are desired they should
be introduced in a risk adjusted fashion so as to not favor the higher risk farmers or
regions; and 3) subsidies should foster the greatest involvement from private markets.
When farmers pay something that is equivalent to a pure premium, the program will not
favor the high-risk farmers or regions. If more subsidy is desired, a flat rate off the pure
premium should be used for all. For example, 1 or 2 points off the premium rate would
give everyone the same subsidy. Thus, if the subsidy is 1 point, a farmer facing a pure
premium of $5 per $100 of liability would pay $4 and one facing a pure premium of $10
would pay $9.
42. There are numerous ways to implement the principle of fostering the greatest
involvement from the private sector. Each of the principles outlined above is motivated
by the overriding objective of using limited government dollars to reach the highest
development of insurance markets without introducing undue inefficiencies. This is no
simple task. Further, with careful thought, it may also be possible to also introduce
government involvement at some level for disaster aid in a way that also facilitates
private insurance development. This will require careful analysis of alternatives in close
concert with international re-insurers but with a careful eye toward designs that reduce
rent seeking. Clear and transparent rules offer one key to controlling rent seeking.
Regulatory Issues
43. La Comisión Nacional de Seguros y Fianzas (CNSF) regulates insurance, including
crop insurance. Insurers underwriting crop risks are subject to the same technical
provisions and/or reserving rules that apply to general insurance or P&C insurance.
CNSF regulates the fondos indirectly, by regulating the reinsurers who provide
reinsurance to the fondos. Reinsurers also are regulated by CNSF and are subject to same
solvency requirements as those that apply to the general insurers or P&C insurers.
44. CNSF is working on a new set of guidelines to address such issues as technical
reserves, contingency funds and/or surplus and capital that each fondo will be required to
adhere. The objective of the proposed set of regulations is to strengthen the financial
conditions of Fondos and also to provide incentives to Fondos that are well managed.
The regulation will also explicitly address the issue of risk retention by both the Fondos
and private insurance companies with the objective of avoiding any potential insolvency
problems. While regulations are needed to safeguard the financial viability of the
Fondos, they should not be overly restrictive as to stifle the growth and creation of new
15
Fondos. CNSF is also looking into the way technical reserves are determined by the crop
insurers, and there is a proposal to move away from a purely formula based approach to a
risk based one keeping in mind the highly volatile nature of the risk.
45. Under the existing regulation insurance companies can obtain credit for reinsurance
only if they reinsure with a re-insurer whose credit rating is BBB and above by S & P
and/or of similar levels by other internationally accredited rating agencies. In case
reinsurance is obtained from a non-(re)insurance company, such as weather derivatives
companies, the primary insurer will not be given credit for the reinsurance and hence as
result may have to put up the necessary capital for solvency purposes.
46. Insurance products based on parametric triggers or indexes may be considered as
insurance products for crop insurance purpose provided sufficient information is
available which will enable pricing of the product and also help in determining the
required reserves for such type of products. Similarly, reinsurance based on parametric
index can also be considered a insurance product. CNSF is working on changes in the
regulation that will permit certain types of financial reinsurance, on a case-by-case basis.
On the issue of (re)insurance to the Fondos the existing regulation states that the Fondos
can only obtain reinsurance from AGROASEMEX and cannot access international reinsurers directly. Changes are being proposed for other re-insurance companies to
provide re-insurance to the Fondos.
47. CNSF should assess the solvency regulations for reinsurance companies; these are
currently similar to those governing P&C companies. Future regulations should
recognize differences in the nature of risk between the two types of companies and
differences in the type of coverage offered. Reinsurers typically accept risks from
insurers who should have done the due diligence of the insured before ceding the risks to
the reinsurer. Second, the reinsurers either share the risk and subsequent payout, with the
direct insurers, in the form of proportional reinsurance treaty or else provide an excessof-loss cover. In either scenario, the insurer is better-off minimizing the losses and, as a
result, the risk to the reinsurer is also minimized. And, finally the risk to the reinsurer is
further minimized due to the diversification of risks at the global level and also because
the reinsurers retrocedes some the risk.
48. Regulating AGROASEMEX under the changing environment is going to be a
challenge to CNSF, since, potentially, at least in the near term, AGROASEMEX will be
acting both as a reinsurer for crop insurance and as a direct insurer of livestock. Even as
a reinsurer AGROASEMEX has limited choice in the type of risk it will be accepting.
CNSF needs to develop a strategy to ensure that AGROASEMEX remains financially
viable.
Who should regulate the Fondos?
49. OPTION-1: Currently, AGROASEMEX regulates the Fondos and hence may be
provided with the mandate to do so in the future. The benefit of such an approach is that
fondos will be regulated by a company that has an understanding and knowledge of how
they works and also by trained staff who could monitor fondos more effectively. The
downside is the increased administration cost for AGROASEMEX to regulate these
fondos. Given that the plans of the government are to let the reinsurers compete for the
16
fondos business, there will not be sufficient margin for AGROASEMEX to sustain
expenses incurred towards regulating the fondos. Also, by exclusively letting
AGROASEMEX regulate fondos could provide a advantage for AGROASEMEX in
terms of providing reinsurance covers to fondos and also would provide AGROASEMEX
access to reinsurance treaties signed by fondos with various reinsurance companies.
50. OPTION-2: Recently it is suggested that the regulation of fondos should be by the
companies offering re-insurance to fondos, based on criteria provided by the CNSF. The
benefit of such an indirect approach to regulating fondos is to minimize the cost of
regulation, given that there are close to 200-odd fondos currently operating. Also, given
that the reinsurers would like to limit the payout from the reinsurance treaty there is an
incentive for the reinsurers to ensure that sufficient technical reserves and other
contingent funds are established and also that the funds are technically solvent. The
downside is that the reinsurers will typically operate only in favorable areas namely
fondos in the irrigated areas, hence an alternative mechanism may have to be put in place
for the regulating the fondos in rainfed and/or less favorable areas. Also, there is the risk
of lack of regulatory control on the international reinsurance companies, since the
regulation as it is today only stipulates that reinsurance needs to be purchased for a
reinsurer with a rating of BBB and above, i.e. it is not necessary for the reinsurance
company to be a local company.
51. OPTION-3: Federation of Fondos could be created either at State level and/or
Regional level. Benefit of creation of such a federation of fondos will assist greatly in the
supervision of the fondos. The federation will regulate its members, more like a SRA
(self regulatory authority e.g. the stock exchanges being regulated by its board who then
are regulated by the SEC) and itself will be regulated by those offering re-insurance
coverage again based on norms by CNSF. The Fondos within the federation could
contribute a small annual fee, which among other things could be used to supervise and
regulate the fondos.
52. OPTION-4: In addition, the government may wish to establish a system of rating to
rate the fondos by a third party. This will allow the better fondos obtain better terms for
their re-insurance thus providing an incentive to perform better. The downside is that
those fondos operating in the unfavorable areas may be left out of the system since, the
reinsurance covers will in any case be high given the frequency of losses.
17
3. THE EXPERIENCE OF FONDOS DE ASEGURAMIENTO IN PROVIDING
INSURANCE TO FARMERS
Background
53. The idea for the creation of the mutual insurance funds, Fondos de Aseguramiento or
simply fondos, came from the farmers. The coalition of ejidos colectivos de los Valles
del Yaqui y Mayo in Sonora, located in an irrigated area with low losses, discovered that
they had consistently been paying significantly more in premiums than what they had
collected in indemnities.6 This, along with the low level of service received from
ANAGSA encouraged these farmers to propose to BANRURAL, their main source of
credit, that they deposit the annual amount of premiums that they would be paying for
crop insurance into a special account to guarantee potential crop losses. In 1988, the
government redefined its role as a provider of insurance and decided to include other
agents, such as farmers’ organizations and private companies as direct providers of
insurance. Sonora was the first state where farmer organizations started offering
insurance by constituting fondos, an experience that was expanded with the support from
AGROASEMEX.
54.
It is very important to emphasize that unlike other experiences with farmer
organizations established from top down, the fondos are created by the farmers’ own
initiative, encouraged and supported by the government but not trying to use them or
control them for political or other purposes. For most fondos, their members already had
experience in working together. As such, fondos are very selective in choosing their
members. They include farmers that have an interest in joining others to develop
common projects and to negotiate in groups. The mutual confidence and common
interest are the basis and farmers are very selective in accepting new members in their
organization. Many farmers closely observe each new fondo. A good experience of a
new fondo often creates interest amongst farmers to create a new one or to integrate into
the first one.
55.
Fondos are non-profit, civil associations. They operate in such way that the
collected premiums cover the operational costs and create reserves to pay indemnities.
The fondos have a general assembly that chooses an administrative council composed of
a president, a secretary and a treasurer to manage the everyday work.7 The administrative
council is elected for a period of three years. Each fund has two areas (departments): a
technical area to conduct the technical insurance work and an accounting area to manage
the financial resources. Each fondo also elects its own self-supervisory council.
6
Every year the Coalicion ejidos colectivos de los Valles del Yaqui y Mayo was paying on average N$21
million to ANAGSA for crop insurance while it was receiving on average around N$2 million in indemnity
payments.
7
There is an ordinary and an extraordinary general assembly. The main functions of the ordinary general
assembly are to elect the administrative council, and to approve the budget, the insurance and re-insurance
programs, the results of the evaluation, and financial statements. The extraordinary general assembly is to
approve the bylaws of the fondo, and to admit or exclude members (farmers).
18
56.
In 1988 SHCP published the basic framework for the creation and operation of
fondos which, in 1992, became the rules for operating and creating “peasant funds for
crop and life insurance” and other activities in the agricultural sector. The rules for the
fondos include: the process that farmer organizations must follow to get authorization to
operate as insurance funds, requirements for accepting new members, the basic
administrative structure of the fund, procedures to elect the board (assembly) and the
roles that the organization much have.
57.
To cover its administrative costs, fondos can use a maximum of 25% of the
premiums collected after paying for their re-insurance and the remaining 75% should be
set as a reserve for current risks (Figure 3.1.) At the end of the insurance cycle (crop
season), after indemnities have been paid from the reserve account for current risk, and
there is a surplus, 30% of this goes to a special reserve for contingencies. This reserve is
to be used to reduce the cost of re-insurance and to cover possible losses in the future that
cannot be covered by the reserve account for current risks. The remaining 70% of the
surplus goes to a “social fund” and can be used to provide services to the members of the
fondo and/or can be invested in joint capital equipment. These could include,
contributing to premium payments for the next crop cycle, as well as, the provision of
services and creation of new enterprises and activities benefiting all members.
Figure 3.1: Use of Fondo funds.
Total premium (%)
Reinsurance
premium
Premium
after
reinsurance
payment
Administrative
costs (25%)
Losses
payment
Current
risk
reserve
( 75% )
Surplus
Special reserve of
contingency (30%)
Surplus
to the Fund
( 70% )
Source: Own elaboration.
The Role of AGROASEMEX in Promoting Fondos
58.
Most fondos were promoted by AGROASEMEX. In fact, AGROASEMEX has a
special department to support the development of fondos consisting of 38 staff who have
as their main task to promote the creation of new funds, provide technical assistance to
existing ones (training, education, etc.), and supervise them. AGROASEMEX supervises
the fondos by reviewing their financial statements and performing random field
inspections and random visits to areas that announce losses. In addition,
AGROASEMEX provides reinsurance capacity to help spread the risk in cases of
catastrophic loses that are not possible to cover with the fund reserves. Fondos pay
19
AGROASEMEX a re-insurance premium to provide them with an excess loss insurance.
AGROASEMEX is also the channels to provide the 30% subsidy for the premium to
farmers insured through the fondos.
59.
There are currently 200 fondos in operation and during 2000 they accounted for
30.9% of the total insured area and 34.3% of the total premiums collected. The market
share of fondos increased until 1994/95 and after that has declined as private companies
were allowed access to premium subsidies in addition to the fondos and
AGROASEMEX.
Figure 3.2:Market share by firm type
120
0
0
0
100
0.5
2.6
4.8
18.9
80
45.1
54.8
52.4
54.1
54
23.2
32.8
28.7
45.2
Private
41.3
60
39.2
36.1
30.9
Fondos
AGROASEMEX
40
54.9
20
45.2
47.6
45.4
43.4
50
39.8
37.6
31.1
40.4
0
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Source: AGROASEMEX
Evaluating the Performance of the Fondos
60.
Most of the fondos were created during the period 1991 and 1992, but the creation
of new fondos continued after that, although at a significantly reduced rate. This may
reflect the conflict of interest that AGROASEMEX has as supporting the creation of
fondos and competing with them for providing direct insurance to farmers. The number
of new fondos has increased in 2000 perhaps reflecting the upcoming changes for the
elimination of AGROASEMEX from the role of direct provider of insurance (figure 3.3.)
20
Figure 3.3:Number of Funds per year of creation (include only
Funds that operated in the 2000 year)
Number of funds
45
39
37
36
24
27
18
14
14
10
10
15
14
11
8
9
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Source: Elaborated with data from Agroasemex.
61.
In 2000 fondos insured a total area of 567.7 thousand hectare and reported 69.7
thousand members. The insured value was N$2.89 billion. Thus, the average farmer in a
fondo has around 8 hectares and insures a value of N$41,463 (or about N$5,100 on
average per hectare insured8). These figures indicate that fondos members are small and
medium farmers but with commercial orientation and using intensive, higher cost
production technologies that lead also to higher yields. The main insurance product the
fondos offer to their members is investment insurance. This protects farmers against a
drop in the yields that would prevent them from covering their initial investment at the
beginning of the crop season. Investment insurance is mainly linked to the amount of
credit obtained at the beginning of the crop season.
62.
Overall, the fondos have a relatively low loss ratio. For the period 1991 to 2000
this loss ratio is estimated to be 56% which is lower compared to AGROASEMEX and
around the same level with the best loss ratios of private insurance companies. One of
the reasons as to why the fondos have a low loss ratio, particularly for agricultural
insurance, is that the fondos include proportionally more irrigated areas in their portfolio.
Usually, about three-quarters of fondos operations are in the fall-winter (OI) crop cycle
that is mostly irrigated. Overall two-thirds of the insured areas by fondos are in irrigated
areas.9 Another, important reason for low loss ratios is that the fondos, being a mutual
insurance organization, have incentives to minimize losses in order to retain a higher
percentage of premiums that would benefit their members. Also, fondos are very
selective of their membership and have good technical skills in processing the claims and
perform loss assessments. The technical competence in insurance issues of the fondos is
to a large extent due to the technical assistance they receive from AGROASEMEX.
Table 3.1 below ranks the fondos based on their loss ratios over the period 1990-2000.
Most of the fondos have low to medium loss ratios. These results are robust to whether
8
To put it into perspective, the input costs for gravity irrigated, improved seed, and fertilized maize or
sorghum are around N$5,000 to N$6,500 per ha.
9
The growth of fondos in non-irrigated areas has been slow and less successful.
21
loss ratios are calculated by including or not the premium subsidies. Thus, even without
subsidy most fondos are still financially viable.10
Table 3.1: Fondos classified by average loss ration for 1990-2000
Loss ratio
>1
0.75 < 1
0.5 < 0.75
< 0.5
Situation with subsidy Situation without subsidy
Number of funds
17
18
13
16
23
22
47
44
Source: R. Altamirano (2001) calculations based on data provided by AGROASEMEX
63.
Fondos are very much geographically concentrated. For the period 1990-2000,
Sinaloa, Tamaulipas and Sonora accounted for 77.8% of the total insured areas by the
fondos and 73.1% of the total premiums collected. It is also the case that Sinaloa and
Sonora have relatively higher loss ratios: around 60% with premium subsidies and over
80% if premium subsidies are not included. Table 3.2 below shows the location of
fondos per state. The first column shows the area insured by the fondos in each state as a
share of the total area insured by the fondos in all states. The second column shows the
share of each state in total premiums collected and column three shows the number of
fondos in operation in each state in 1999.
Table 3.2: Location of Fondos per State: 1990-2000
State
Sinaloa
Sonora
Tamaulipas
Guanajuato
Jalisco
Baja California
Chihuahua
Nayarit
Share of agricultural Area
31%
23%
23%
3%
3%
4%
3%
3%
Share of Total Premium
40%
21%
13%
4%
5%
3%
4%
4%
Number of Fondos
49
31
39
14
15
6
9
9
64.
Fondos are also very much concentrated in terms of product coverage. Again for
the period 1990-2000, maize, wheat, sorghum, and beans account for 79.1% of the total
areas insured by the fondos and for 73.5% of the premiums collected. Soybeans and
beans have very high loss ratios: over 80% with premium subsidies and above 120%
when premium subsidies are excluded. It is worth noting that fondos for coffee have a
loss ratio of almost zero during 1990-2000, indicating that there are negligible losses
reported over that period.
65.
Two key variables explaining the regional concentration of the fondos are the
amount of credit and the irrigated areas available per state. Using 1998 data, the
correlation between these two variables and the amount of premiums collected is 0.88 for
10
For the period 1990-2000 the overall loss ratios of the fondos without the 30% premium subsidy is 77%.
22
credit and 0.62 for irrigated areas.
The Financial Viability of the Fondos
66.
Are fondos financially viable? Looking at their experience so far the simple
answer is yes. Overall, most fondos have loss ratios low by agricultural insurance
standards and comparable to the best of private insurance companies (table 3.3). As
explained, when premiums collected exceed payouts fondos accumulate special reserves
for future contingencies. Taking into account these reserves, the ratio of fondos payouts
over the sum of premiums collected plus special reserves averaged 44% for the period
1990-2000, and 47% for 1999. However, there are variations in the financial
performance of fondos. Payouts have been very low in Veracruz, Colima and Hidalgo
but relatively high in Oaxaca, Chiapas and Estado de Mexico.
Table 3.3: Measures of Fondo viability, by state
Loss ratios
Indemnity over
all reserves
State
Reinsurer Fondos
Fondos
Baja California
0.31
0.41
0.25
Baja California Sur
0.00
0.31
0.18
Campeche
0.00
0.67
0.79
Chiapas
1.29
0.62
0.75
Chihuahua
0.45
0.54
0.42
Coahuila
0.63
0.37
0.40
Colima
0.00
0.13
0.08
Durango
0.00
0.37
0.40
Estado de Mexico
0.00
0.78
0.94
Guanajuato
0.17
0.35
0.25
Hidalgo
0.00
0.14
0.18
Jalisco
0.57
0.56
0.59
Michoacan
0.23
0.37
0.34
Nayarit
0.20
0.51
0.47
Nuevo Leon
0.49
0.44
0.20
Oaxaca
0.48
1.31
1.54
Puebla
0.14
0.48
0.48
Queretaro
0.31
0.40
0.38
San Luis
0.00
0.47
0.54
Sinaloa
0.60
0.60
0.51
Sonora
0.74
0.58
0.53
Tabasco
0.41
0.36
0.38
Tamaulipas
0.80
0.46
0.39
Veracruz
0.00
0.00
0.00
Zacatecas
0.00
0.36
0.43
Unweighted average
0.56
0.51
0.44
Source: authors’ calculations based on data provided by AGROASEMEX
Note: The Reinsurer loss ratio reflects the payments (over reinsurance premiums) of the re-insurer
(AGROASEMEX) to the fondos for excess losses. The fondos loss ratio reflects the payments (over
insurance premium) that the fondo had to make to its farmers. Data are for the period 1990-2000.
23
67.
When reserves have been exhausted, reinsurance (through AGROASEMEX) has
provided additional indemnity payments. On average, the loss ratios for the reinsurance
have been comparable to the fondo insurance loss ratios – averaging around 56% for the
1990-2000 periods. However, the range of experience across states differs dramatically.
Reinsurance has never paid out in eight states, Campeche, Colima, Durango, Estado de
Mexico, Hidalgo, San Luis Potosi, Veracruz and Zacatecas, and reinsurance payouts have
been nominal in Baja California Sur. In contrast, reinsurance payouts have been
concentrated in Sonora, Sinaloa, Tamaulipas and especially in Chiapas. This latter has
recorded the highest reinsurance payouts.
68.
Table 3.4 presents average area insured by crop along with calculated crop loss
ratios and average premiums over value insured. Again, the experience among crops
varies greatly. Losses paid out to sugar cane and coffee are rare, while fondos frequently
pay out for tomatoes, mango and other horticultural crops. Moreover, the premiums
charged for different crops – which are dictated by reinsurance rates – do not seem to
fully reflect the inherent risks of different crops. For example, on average, the premium
charged for mango and tomatoes are 5% and 6% respectively which are lower compared
to that for maize (8%) even though mango has a loss ratio of 123%, tomato 147%, and
maize only 59%. This is most clearly demonstrated by looking more closely at some of
the largest payouts – where payouts were ten times larger than collected premiums.
These cases, reported in table 3.5, show a concentration among horticultural
commodities, and premiums relative to value insured similar to the rates charge for coffee
or sugar where payouts are rare.
24
Table 3.4: Area insured, Fondo loss ratios and premium over insured value for
selected crops, average 1996-2000
crop
Maiz
Trigo
Sorgo
Frijol
Algodon
Soya
Garbanzo
Cartamo
Café
Vid
Arroz
Cebada
Chile
Caña
Tomate
Naranjo
Cempasuchil
Cacahuate
Okra
Maiz Dulce
Maiz Palomero
Calabaza
Mango
Melon
Cebolla
Ajonjoli
Olivo
Cebada Malta
Sandia
Average annual Average fondo Average fondo premium
insured area
Loss ratio
over insured value
165,612
0.59
0.08
111,210
0.48
0.05
87,422
0.54
0.09
39,540
0.92
0.13
29,701
0.88
0.08
28,931
1.31
0.12
19,149
0.88
0.08
0.82
0.07
6,005
3,160
0.00
0.04
3,109
0.54
0.05
2,396
0.69
0.08
2,380
0.48
0.08
1,369
0.99
0.07
1,215
0.03
1,108
1.47
0.06
944
0.04
0.05
594
0.86
0.05
517
0.81
0.07
366
0.22
0.04
257
0.90
0.06
252
0.45
0.05
217
2.32
0.07
208
1.23
0.05
198
0.95
0.12
187
0.48
0.07
179
1.87
0.15
149
2.45
0.06
140
0.34
0.07
132
1.28
0.08
Source: AGROASEMEX and authors’ calculations
25
Table 3.5: Large Fondo losses by crop and region
Region
Noroeste
Noroeste
Noroeste
Norte
Pacific Norte
Noroeste
Noreste
Pacific Norte
Noroeste
Noreste
Noreste
Centro
Pacific Norte
Noroeste
Pacific Norte
Noreste
Noreste
Noreste
Noreste
Noreste
Noreste
Noreste
crop
Frijol
Maiz
Sandia
Sandia
Garbanzo
Maiz
Chile
Maiz
Maiz
Sorgo
Chile
Maiz
Garbanzo
Maiz
Calabaza
Sorgo
Cartamo
Chile Jal.
Chile
Tomate
Chile Pim.
Chile S.
year
2000
1998
1991
1997
1992
1998
1998
1999
1990
1997
2000
1994
1992
1998
1994
1991
1997
2000
2000
2000
2000
2000
Loss ratio
10.01
10.04
10.14
10.16
10.18
10.34
10.49
10.90
11.45
11.98
12.21
12.30
12.73
13.33
14.40
14.81
18.03
19.64
19.64
31.49
35.23
39.63
Premium as share Insured
of value
value
payout
Per hectare
0.07
6,300
4,655
0.08
5,000
4,250
0.07
4,237
3,222
0.06
8,000
5,210
0.07
1,878
1,289
0.08
4,600
3,921
0.02
31,400
6,619
0.06
6,849
4,841
0.08
1,220
1,172
0.05
1,300
718
0.04
8,500
3,715
0.07
1,296
1,046
0.05
1,794
1,169
0.06
6,000
4,398
0.05
4,120
2,747
0.02
1,273
443
0.04
1,600
1,159
0.04
31,783 24,901
0.02
44,071 17,400
0.02
42,623 30,202
0.02
78,868 55,847
0.02
48,213 38,403
Source: AGROASEMEX and authors’ calculations
69.
Fondos provide multi-peril insurance; and when payouts are made, the reason for
the payout is recorded. Table 3.6 reports the hazards that prompted Fondo payouts from
1993 to 2000. Excess humidity is a peril frequently triggering payments by fondo
insurance; however several hazards prompt repeated payouts. Other important perils
prompting insurance payments by fondos during that period are droughts, winds, frost
and pests. Surprisingly floods have a relatively low share in the payouts by the fondos.
Table 3.6: Share of Fondo payout by peril, 1993-2000
Peril
1993 1994 1995 1996 1997 1998 1999 2000
drought
2.1% 3.1% 3.0% 11.3% 6.4% 6.3% 10.0% 9.0%
excess humidity
28.7% 15.1% 11.3% 6.5% 20.6% 13.9% 7.4% 3.7%
frost
4.2% 0.8% 2.7% 2.9% 6.7% 8.2% 8.3% 2.3%
low temperature
1.0% 2.4% 1.5% 3.1% 3.3% 4.5% 9.4% 9.0%
flood
1.5% 1.0% 1.4% 0.8% 0.6% 2.2% 0.3% 0.5%
hail
4.4% 5.4% 2.6% 1.6% 5.2% 1.9% 2.9% 1.9%
fire
0.0% 1.2% 0.3% 0.3% 0.3% 0.0% 0.4% 0.1%
wind
5.6% 8.4% 6.4% 4.3% 6.6% 4.9% 4.1% 7.3%
high temperature
5.7% 9.1% 4.1% 3.6% 3.7% 3.7% 5.1% 3.9%
pests
3.6% 12.7% 9.6% 11.6% 6.8% 4.4% 6.2% 8.1%
disease
7.8% 7.4% 4.6% 4.5% 3.3% 3.8% 5.1% 4.8%
muddy soil
2.1% 2.2% 0.4% 0.3% 0.6% 0.6% 0.6% 0.6%
hard soil
6.7% 4.4% 4.8% 2.2% 3.4% 9.2% 1.8% 1.5%
can't plant
0.0% 0.0% 0.0% 0.0% 0.3% 0.1% 1.1% 0.5%
didn't emerge
0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.5%
Source: AGROASEMEX and authors’ calculations.
26
70.
Over the period 1990-2000, and in constant 2000 N$, the government has
provided the fondos with N$962 million in premium subsidies, fondos have paid for
reinsurance to AGROASEMEX a total of N$771 million and received N$450 million as
reinsurance pay outs. Thus, the net outflow of public resources to the fondos over the
period 1990-2000 (subsidies plus re-insurance payouts minus reinsurance premium costs)
has been N$640 million. AGROASEMEX has paid 58 cents for every N$1 of reinsurance premiums collected by the fondos, and the premiums subsidies by the
government had been enough for the fondos to pay for the re-insurance and also
contributed to a surplus.
71.
Over the period 1990-2000 it is calculated that the fondos should have
accumulated significant reserves. This calculation is as follows (all figures in constant
2000 pesos). It is estimated that the fondos have collected N$3,400 million in premiums
from farmers and subsidies from the government. They have paid N$771 million in
reinsurance and they have had N$657 million in administrative costs (calculated as the
25% of the premiums collected minus reinsurance payments). Fondos have paid farmers
N$1,446 million and have N$79 million in special reserves. This gives an estimated
cumulative surplus of N$446 million over the entire period 1990-2000. If we divide this
by the total insured value for 2000 (N$2.89 billion) we obtain a ratio of cumulative
surplus to total insured value of about 15%.
Explaining Insurance and Re-insurance Pay-outs to and by the Fondos
72.
As discussed earlier, fondos pay to farmers in case of losses. And fondos receive
re-insurance when total losses exceed the premiums and reserves that the fondos have
collected. This section tries to explain the loss ratio of the fondos, and the loss ratio that
the re-insurance has vis a vis the fondo.
73.
The method used to explain the Fondo’s and reinsurance loss ratios is based on
the Tobit analysis.11 The key findings of the analysis are as follows (table 3.7).
74.
If correctly priced to account for the risks, the higher the premium charged per
peso of insured value, the lower the loss ratio. However, the Tobit analysis shows the
opposite – that is, the coefficient on the premium over value variable is positive. This
suggests that in many cases, high losses were associated with high prices – that is, that
losses have occurred because risks were under-priced. This is consistent with the table
on high payouts reported earlier. The analysis also shows that technology choices and
regional differences matter as well. Not all technology choices result in the same loss
ratios. For example, separate analysis showed that irrigation technologies reduced loss
ratios. Not all hazards had the same effect on loss ratios. The analysis shows that most
11
A Tobit analysis is similar to a regular ordinary-least-squares regression; however adjustments are made
to the estimation process to account for dependent variables that are truncated. In the problem at hand, loss
ratios, which are always greater or equal to zero, constitute the dependent variables.
27
hazards are significant, however drought, frost, high temperature and excess humidity
appear to be quantitatively more important.12
Table 3.7: Determinants of Fondo loss ratios
variables
price
hazards
unspecified
drought
excess humidity
frost
low temperatures
flooding
hail
fire
wind
high temperatures
pests
disease
Muddy soil
hail
unable to plant
plants fail to emerge
Region effects
coefficient t-score
1.72
2.30
0.92
1.45
1.33
1.39
1.00
0.01
0.67
0.37
0.36
2.21
0.60
0.36
0.59
-0.14
0.91
0.46
14.41
14.36
21.28
15.14
10.51
0.07
6.11
1.16
4.55
16.69
8.35
4.28
2.90
-1.72
3.09
0.63
Centro
Noreste
Noroeste
Norte
Occidente
Oriente
Pacific Norte
-0.91
-0.69
-0.59
-0.78
-1.21
-0.61
-0.68
-3.08
-2.43
-2.03
-2.61
-4.21
-1.26
-2.34
Technology
irrigated
rainfall
-0.27
-0.41
-0.92
-1.41
Source: AGROASEMEX and authors’ calculations
75.
A similar analysis was also applied to reinsurance loss ratios. Unlike the Fondos,
increased premiums relative to insured value reduces loss ratios as expected. Moreover,
higher special reserves also reduce the loss ratios. This is logical, since reinsurance only
applies once all Fondo reserves have been exhausted. As with the Fondos, most perils
are significant for reinsurers. The main perils for re-insurance payouts indicated in the
analysis are drought, frost, high/low temperatures , wind and pests. The implication here
is that a simple parametric insurance – for example on frost or rainfall alone – would not
provide adequate coverage on aggregate for the fondos. However, on a case by case
basis, if a certain fondo is particularly exposed to one or two risks, a parametric insurance
could be more appropriate.
76.
The inclusion of several regional dummies indicates regional differences. This is
significant since it implies that regions experience systemic risks and regional federations
12
For the Fall-Winter (OI) crop cycle, which is mostly irrigated, the main reasons for payouts are low
temperatures, frost, excess humidity and pests. In the Spring-Summer crop cycle, the main reasons for
payouts are excess humidity, drought, pests and strong winds.
28
(or associations) of fondos that provide a measure of reinsurance would, themselves,
require reinsurance. In other words, by grouping fondos at some regional level, the
degree of geographic diversification will be limited, since the regional fondo group is
likely to be subject to similar system risks as the individual fondos. Finally, some year
dummies were significant – probably indicating dominant weather trends.
Table 3.8: Determinants of reinsurance loss ratios
Coefficient t-score
s
reserve fund
-0.02 -10.08
premium over insured value
-60.31 -2.92
Hazard
Drought
0.02 12.69
excess hum.
0.01 11.82
Frost
0.02 14.83
low temp
0.01 10.93
Flood
0.01
5.70
Hail
0.01
9.76
Fire
0.04
1.17
Wind
0.01 10.96
hi temp
0.02 10.44
pests
0.02 11.53
disease
0.01
3.38
muddy soil
0.01
3.57
hard soil
0.01
2.12
unable to plant
0.08
0.94
plants fail to emerge
-0.04 -1.35
Region effects
Centro
-3.48 -2.43
Noreste
-1.16 -0.90
Noroeste
-3.33 -2.50
Norte
-2.82 -1.98
Occidente
-3.34 -2.49
Oriente
-1.12 -0.46
Pacific Norte
-2.36 -1.86
Year effects
1992
6.49
7.26
1993
6.01
7.26
1994
3.12
3.34
1995
3.45
3.78
1996
4.07
4.73
1997
1.62
1.59
1998
3.00
3.53
1999
3.02
3.65
2000
1.92
2.30
constant
-5.83 -3.66
Source: AGROASEMEX and authors’ calculations
Challenges for the Fondos
77.
The review of the experience and performance of the fondos shows that it is
possible to develop a mutual insurance scheme at the farmers’ level that is financially
29
viable (even without subsidies) and that it is efficient. Fondo, is a mechanism to deliver
crop insurance to small farmers. Nevertheless, the fondos have certain weaknesses and
challenges that they need to face in order to remain viable and perhaps expand further.
78.
Fondos have a very high dependency on AGROASEMEX in terms of insurance
product innovation, reinsurance capacity and technical assistance necessary to operate.
The fondos need to start developing new products that better meet the needs of their
members otherwise they will keep loosing market share to private companies.
79.
Fondos have a very high geographic and product concentration. Their
concentration is particularly high in basic crops, especially grains, that are experiencing
low prices and reduced margins. Also, two-thirds of the areas insured with fondos are
irrigated. Availability of irrigation might limit the expansion of fondos. So far, the
expansion of fondos in rainfed areas has been slow and less successful.
80.
Fondos could consider to diversify by moving to other insurance products, such as
life insurance, property insurance (machinery, building, equipment, etc.) that are
uncorrelated with crop losses and also have a lower and more stable loss ratios compared
to crop insurance. This however, will require the authorization of SHCP.
81.
Even though the number of funds has been increasing and they have been
accumulating reserves, efforts to create a second level organization of fondos have not
yielded any results yet. The National Association of Insurance Funds (ANFA) is mainly
focusing its activities in lobbying with AGROASEMEX and the government.
82.
The fondos could work towards harmonizing their administrative and accounting
systems. This harmonization will make easier to supervise the fondos and will facilitate
their integration in some regional association or federation.
83.
To integrate the fondos in a national structure is very difficult because of the
diversity of conditions amongst the fondos and variations in their performance. For
example, fondos in Chiapas and Oaxaca have very high loss ratios (and lower reserves)
compared to low ratios (and higher reserves) in Baja California, Puebla and Veracruz.
The fondos in these latter areas may not wish to participate in a national scheme together
with the fondos in Chiapas and Oaxaca. Thus, it can be proposed that first try to create
regional associations of fondos in relatively homogeneous technologically regions to
obtain enough of geographic risk spread. The regional association of the fondos can then
obtain re-insurance. The association could develop technical services to provide to the
fondos and assist in the development of new products.
30
4. RISK MANAGEMENT AND CROP INSURANCE FOR LOW INCOME
FARMERS
Characterizing Small and Resource Poor Farmers
84. Usually, farmers in Mexico are characterized in terms of a single asset - land. That is,
small farmers are usually defined in terms of landholding size – usually, less than 5-20ha,
with farmers having less than 5 hectares being considered “resource poor small farmers.”
This delineation is quite imprecise because other factors such as soil quality and agroecological conditions, access to irrigation, physical infrastructure (e.g. roads), markets,
and local institutions, have a great impact on agricultural potential and incomes. These
factors vary greatly in Mexico13 - even within a state.
85. However, it should be emphasized that - ceteris paribus – small landholdings do,
indeed limit the level of income that small farmers might be able to obtain from
agriculture, and thus can limit their ability to manage income fluctuations (the variance of
income. With respect to landholding size, a large proportion of farmers have less than 5
hectares. A representative survey of rural households in 1995 (World Bank, 1995)
indicated that 65% of households involved in farming had less than 5 hectares (and 85%
less than 10ha). In the nationally representative 1997 Ejido Survey, about half of
ejidatario households had less than 5 hectares.
86. In this section we will focus attention on resource poor small farmers who primarily
use their on-farm agricultural production for household consumption (e.g., food staples
such as maize and beans, and livestock products14), use little if any improved
technologies (e.g., little or no purchased inputs such as improved seeds or fertilizer), and
usually depend on a single growing cycle using rainfed agriculture.15 In contrast, there
are small farmers that might also have small landholdings, but who use improved
technologies, are more commercially oriented, and many of whom have access to
irrigation and can grow more than one crop per year (and in most cases have other assets,
such as good social capital location/infrastructure and institutions). Some of these
commercially oriented small farmers are, members of the fondos and clients of
13
In fact, a farmer with 10+ rainfed hectares in low rainfall areas might still be considered a small farmer
(see the section on FONDEN for their classification of “small farms”). And, a “small farmer” with 3
hectares of irrigated land used for horticultural crops exported to the US can not really be compared to a
small farmer with 3 hectares of rainfed grain crops
14
As can be observed in Table 4.2, income from livestock comprises a larger share of imputed farm income
from farmers with under 5 hectares, than farmers with larger landholdings. It is also important to note that
following land, livestock are one of the most important assets held by small farmers – for consumption and
for savings and self-insurance purposes.
15
A survey of representative rural households in 1995 (World Bank, 1995) found that about ¼ of farmers
did not use purchased inputs for their agricultural production. On the other hand, about ½ of the farmers
used chemical fertilizers, insecticides and/or herbicides, and only about ¼ used improved seed varieties.
Also, about ¼ of the sample used irrigation. Likewise, the 1997 survey of the ejido sector found only
about ¼ of the respondents used irrigation (World Bank, 1999).
31
AGROASEMEX and the private insurance companies – as discussed elsewhere in this
report.16
87. Another important characteristic of many resource poor small farmers, is the
relatively low share of on-farm agricultural production in total household income (World
Bank, 1996). For small farmers, in general, a significant proportion of total household
income comes from off-farm activities. Table 4.1 below summarizes sources of income
for farmers by farm size.
Table 4.1:Sources of Income by Farm Size in Hectares, 1995
Agriculture
Non-Agriculture
Remittances
Total
50%
36%
14%
0-2
22%
58%
20%
2-5
35%
50%
16%
5-10
47%
36%
17%
10-18
55%
29%
16%
>18
72%
24%
3%
Source: Casco and Rosensweig (2000).
88. From table 4.1, it can be observed that, on average, for farmers with less than 5
hectares, income from agricultural production accounts for 22-35% of household income,
with non-agricultural income and remittances accounting, for the remaining 65-78% of
household income. In contrast, farmers with over 10 hectares tend to depend on
agricultural income for 55-72% of their household income. Results from the 1997 Ejido
Survey given in table 4.2 tell a similar story. On average, agricultural income accounts
for 34% of household income for ejidatarios with less than 5 hectares, with the share of
agricultural income in total income increasing with landholding size to about 46% for 15
hectares and above.
16
Of course, we do not want to conclude that all resource poor small farmers will not adopt improved
technologies (possibly with irrigation) and transform into more commercially oriented farmers. In fact, as
will be discussed later, there are several SAGAR programs aimed at facilitating this transition.
32
Table 4.2: Ejido land holdings and agricultural income, 1997
Size of holding (hectares)
0
0.1 to 5
5.1 to 15
Averages
Number of household members
4.65
5.79
5.36
Years of school for household head
3.32
2.98
3.29
Age of household head
54.52
50.18
54.64
Hectares of land owned, adjusted
0.00
2.71
9.12
Hectares of land used, adjusted
0.67
2.54
7.94
Herd size
1.17
2.95
6.72
Per capita income, pesos
5,450
3,108
6,432
Access to irrigation
4%
17%
35%
Share of income from farming
11%
34%
43%
Share of farm income from livestock
77%
80%
62%
Share of farm income from maize
8%
7%
11%
Per capita government transfers
684
316
847
Households receiving PROCAMPO
17%
83%
86%
Households receiving Alianza
0%
6%
11%
Transfers as share of income
13%
10%
13%
Source: Data from the 1997 Ejido Survey and authors’ calculations.
>15
5.58
3.34
52.85
30.16
24.72
13.63
7,415
26%
46%
65%
11%
842
81%
17%
11%
All
5.57
3.17
52.45
10.75
9.12
6.53
5,266
25%
40%
69%
10%
627
83%
10%
12%
89. It is important to point out that in most cases, own-farm agricultural production is
referred to as agricultural production and all other economic activities referred to as nonagricultural activities. In reality, a major source of income for many small farmers is offfarm agricultural employment (as wage labor for other, usually larger farmers). In
addition, there are many economic activities in the rural economy that are closely linked
to agricultural production. Thus, small resource poor farmers are indirectly dependent on
the ups and downs of agricultural incomes (and thus, yield and prices risks) faced by
medium and large farmers and risks faced by others in the rural economy.
90. For small resource poor farmers, who use little purchased inputs and self-finance their
production, their direct exposure to agriculture (yield or price risk) is relatively low.
These small farmers tend to use low risk-low return technologies and cropping patterns to
minimize yield losses. And as their marketed surplus tends to be small, their exposure to
price risks is rather limited. In fact, it might even be more important for resource poor
small farmers to obtain livestock insurance to protect their major non-land own-farm
asset and/or insurance that reflects agricultural production off their farm – that is – for
commercial agricultural production activities that dominate the rural economy.17 As is
discussed elsewhere in this paper, one of the advantages of weather based index
insurance is that it can be designed to compensate for yield shortfalls for both producers
and non-producers in the rural economy.
91. This leads us to the following subject, the sources of income risks faced by resource
poor small farmers.
17
For example, in the state of Tamaulipas, sorghum is very important for the rural economy. A resource
poor small farmer who grows maize and beans primarily for home consumption might be better able to
manage his/her income risk by insuring against shortfalls in sorghum yields or prices. Similarly, it might
be more important for the small farmer that larger, more commercially oriented farmers have yield and/or
price insurance to manage their income risks.
33
Sources of Income Risks Faced by Small/Resource Poor Farmers
92. A survey of rural entrepreneurs (REs), which was carried out 1994, indicated that
about 60% of the respondents experienced an “economic crisis” between 1989 to 1994,
with about half of all of these events occurring in 1993 and 1994 (World Bank, 1995).
The reasons given for experiencing an economic crisis are presented in table 4.3 below.
Table 4.3: Reasons for Experiencing an Economic Crisis
Percent of respondents
that faced an economic
crises
Because of
Low Income Due To:
Low Yields
Low Prices
Low Sales
Weak Demand for
Services
Illness of Entrepreneur
Other
Subtotal
High Expenditures
Due To::
Illness of Entrepreneur
or HH Member
Other
Subtotal
TOTAL
Economic sector
Non-Farm
Farm &
Non-Farm
56
60
Lowest 2
Quartiles
51
Wealth
Third
Quartile
66
Fourth
Quartile
66
22
4
33
13
11
4
20
18
27
4
21
8
27
11
22
7
3
9
72
2
5
79
3
12
68
7
6
73
2
9
78
18
18
17
20
22
15
2
20
100
10
28
100
4
21
100
12
32
100
5
27
100
7
22
100
All
Res
59
Farm
23
6
21
12
48
11
1
3
5
3
34
18
4
9
75
6
11
80
18
7
25
100
63
Note: The survey, conducted in July-August 1994 covered 1,944 rural households in Guanajauto, Puebla,
Taumalipas and Veracruz.
Source: World Bank (1995).
93. With respect to the sources of risk, farm REs attributed almost half of the economic
crises to low yields, with 11% attributing the crisis to low prices, 3% to weak demand for
services and 1% to low sales. Illness of the RE or members of the household were the
source of about 20%-25% of the reported economic crises – mostly due to unexpected
increases in expenditures (as opposed to income losses). For non-farm or mixed
farm/non-farm REs, economic crises related to low yields were considerably lower. For
these REs, about half of the economic crises were attributed to low sales and weak
demand for services.
94. With respect to the sources of risk for poorer REs, low sales and weak demand for
services accounted for about 40% of the economic crises, with only 11% from low yields
and about 4% from low prices. This is consistent with the fact that the poorest REs
obtain most of their income from non-farm sources and also that they use low yielding,
yet hardy, traditional varieties primarily for home consumption.18 It is important to note
18
In fact, there is some evidence that some of the new maize high-yielding varieties (HYVs) are actually
more sensitive to weather conditions than widely used “traditional” varieties (World Bank, 1996).
34
that economic crises attributed to illness of the RE and/or family members was the source
of 23% of the crises (more than the total from low yields and low prices). REs in the
third and fourth quartiles attributed 27% of the crises to low yields. REs in the fourth
quartile reported about 12% of the crises could be attributed to low prices. This is
consistent with the profile of wealthier farmers being more exposed to price risks. Thus,
traditional crop insurance – even when assuming that it could effectively mitigate income
losses from yield shortfalls – could only possibly help small farmers manage one (and
perhaps not as significant), of several, sources of income risks.
How Small/Resource Poor Farmers Manage Risks to Their Income
95. Besides looking at the various sources of income risk, it is also important to assess the
risk management strategies and capabilities of small farmers. Unfortunately, the 1994
survey (World Bank, 1995) did not report on risk reduction (e.g., crop or field
diversification) and/or risk mitigation strategies (e.g., crop insurance, savings), but only
on risk coping strategies. However, the reported risk coping strategies do shed some light
on alternative risk management strategies available to small farmers, as presented below
in table 4.4.
96. The major risk coping strategies for small farmers were increased labor market
participation by the farmer and/or household members (38%), reduced consumption
(22%), interest-free loans and/or donations from friends and relatives (15%), sales of
assets – notably livestock19 - (10%) and interest-bearing loans from formal or informal
sources. Interestingly, only about 1% of the respondents reported that they delayed
repayment of loans (although, only a small proportion had access to formal or informal
loans). It was also noted (World Bank, 1995) that loans for consumption smoothing were
usually received from informal lenders (since formal lenders usually restrict loans to
productive purposes), and that interest rates for these informal loans were considerably
higher than for formal sector loans, although repayment rates were more flexible and
based on the ability to repay. Another, more recent study of rural financial markets
(World Bank, 2000c) indicates that many small farmers resort to a variety of selfinsurance strategies that include precautionary savings – primarily using livestock and/or
food stocks20 – along with social capital (e.g., assistance through social networks,
including remittances from relatives living outside the area) and, to a much lesser extent,
financial assets (e.g., savings). Clearly, it is difficult to assess the short- or long-term
efficiency of these coping strategies, without knowing more details about the benefits and
costs of the different options.
19
In many cases, the asset sales were livestock, which are often held by small farmers for home
consumption, cash sales, and for their value as a form of precautionary savings.
20
A major problem with holding livestock as a form of savings to be used in times of economic crisis is
that many farmers might also be experiencing such a crisis and attempting to sell their livestock – thereby
lowering the value of the livestock for risk management. Use of financial savings could improve the ability
of small farmers to utilize their assets to manage risk.
35
Table 4.4: Coping with an Economic Crisis
ALL
Percent of respondents
59
that faced an economic
crises
Percentage of which received
credit in the form of
Loans with positive
7
interest rates
formal/informal
Loans from friends,
5
relatives
Delayed Loan
1
Repayment
Subtotal
13
Percentage of which that
received donations
From friends, relatives
12
From GOM, NGOs
1
Subtotal
13
Percentage of which balanced
income through
Sales of assets
8
Increased labor market
34
participation
Reduced Consumption
25
Economic sector
Non-farm
Farm &
Non-Farm
63
56
60
Farm
Wealth
Third
Quartile
Lowest 2
Quartiles
Fourth
Quartile
51
66
66
7
7
9
4
8
11
5
5
9
5
6
3
1
1
3
1
1
2
13
13
20
10
15
16
10
1
11
16
1
16
6
2
7
13
1
14
14
2
15
11
0
11
10
38
7
31
4
42
6
40
4
37
15
26
22
28
19
22
24
29
TOTAL
100
100
100
100
100
100
100
Note: The survey, conducted in July-August 1994 covered 1,944 rural households in parts of Guanajauto,
Puebla, Taumalipas and Veracruz.
Source: World Bank (1995).
97. The 1994 survey indicated that only 1% of respondents coped with income shortfalls
through Government safety nets. However, to counter risk coping strategies that have
serious negative short- and long term affects, in recent years, the GOM has been devoting
more budget resources to temporary employment programs (PET) and
health/education/nutrition programs such as PROGRESA (more details on PET and
PROGRESA are provided in a later section).
98. Importantly, since 1995, budget allocations for rural safety nets in Mexico have, in
fact been increasing (World Bank, 1999b; World Bank, 2000d; World Bank, 2001).
Also, SAGAR’s budget for rural poverty reduction has increased by about 17% from
1997-99 and accounted for about 1/3 of the total SAGAR budget, with investments in
physical capital accounting for 6.7% and for human capital development 26.7% (see
Table 4.5). These programs have short-term income benefits along with longer-term
asset-building impacts – all of which can help in the management of income risks for
small resource poor farmers and the rural poor in general. In addition, FONDEN which
provides disaster relief assistance (e.g., via transfers and public works projects) has, in
some cases, been an important source of risk coping for weather-related events that cause
widespread yield losses.
36
Table 4.5: Share of government expenditures for rural development, by program,
1997-1999
PROGRAMS FOR PRODUCERS
Of which:
PROCAMPO
APOYOS A LA COMMERCIALIZATION (ASERCA)
ALIANZA PARA EL CAMPO
INFRAESTRUCTURA HIDROAGRICOLA
PROCEDE (land titling)
FORESTAL
OTHERS
50.90%
16.70%
3.70%
3.90%
4.40%
1.80%
0.40%
20.00%
PROGRAMS FOR FARMERS IN TRANSITION
Of which:
PROGRAMA DE EMPLEO TEMPORAL (PET)
APOYOS PRODUCTIVOS EN ZONAS MARGINADAS
INFRA. HIDROAGRICOLA EN ZONAS MARGINADAS 0.5%
INFRAESTRUCTURA PRODUCTIVA
PROGRAMA NACIONAL DE REFORESTACION
OTHERS
15.70%
5.50%
3.10%
0.05%
3.00%
0.50%
3.10%
PROGRAMS TO COMBAT RURAL POVERTY
33.40%
Of which:
PHYSICAL CAPITAL DEVELOPMENT
6.70%
HUMAN CAPITAL DEVELOPMENT
26.70%
NOTE: From 1997 to 1999 the following changes took place in outlays in real terms:
PROGRAMS FOR PRODUCERS
+1.0%
PROGRAMS FOR FARMERS IN TRANSITION
+1.1%
PROGRAMS TO COMBAT RURAL POVERTY
+17.2%
TOTAL SAGAR BUDGET
+5.1%
SOURCE: Casco and Rosenzweig, 2000, p.223.
Demand for Crop Insurance by Small Farmers
99. There is a long history of linking credit and to crop insurance in Mexico (Hogan and
Aubey, 1987), so that it is often difficult to separate the demand for crop insurance from
the demand for credit, since in many cases the purchase of crop insurance is a precondition for receiving a bank loan. However, the vast majority of small farmers (under
5ha) do not use formal credit for their agricultural production (World Bank, 1995; World
Bank, 2000c). In the 1997 Ejido Survey, less about 10% of farmers with less than 5
hectares received production credit from BANRURAL and about 2% had crop insurance.
100. In recent years, PROCAMPO payments have served, to some extent, as a substitute
for formal credit for many small farmers (Casco and Rosenzweig, 2000). PROCAMPO
payments are pledged to banks (or other FIs) in return for cash payments - often
discounted at rates of 5-7% per month - but they do not require any crop insurance since
the PROCAMPO payment is guaranteed by GOM.21
101. With respect to resource poor small farmers attempting to make the transition to
small commercial farmers, the low profitability of agricultural production – notably basic
commodities (e.g., maize, beans, sorghum, wheat) – has been threatened by declining real
21
Actually, there was some demand for crop insurance for seed emergence, as opposed to insuring yields.
This is because PROCAMPO payments were contingent on farmers planting basic commodity crop, and
when monitored, seed emergence is proof of planting.
37
producer prices and increasing input costs. Clearly, with small profit margins, and credit
and insurance adding another 10-20% to input costs, this affects the demand for, and
willingness to pay for credit and crop insurance by small farmers. As can be observed
from some sample crop budgets (see table 4.6), most crop insurance policies cost about 510% of expected input costs.22 Thus, even for more commercially oriented small farmers,
it is questionable whether they would demand traditional crop insurance if it were not
linked to credit. In addition, small farmers might have other means of managing their
income risks.
Alternative Programs for Risk Management by Small Farmers
102. As has been emphasized the major challenge facing resource poor small farmers is
their low expected income, which results from their very limited asset base (broadly
defined to include land, other tangible and non-tangible assets). In addition, because of
their low levels of income and low asset bases, they are more vulnerable to various
sources of risks – because they have limited capabilities to manage the risks. Traditional
crop insurance, which is based on on-farm agricultural production, can, only – at best –
provide some compensation for one, of several, risks that resource poor small farmers
face. Thus, in addition to designing crop insurance products (e.g., index based insurance)
and organizational structures (e.g., insurance mutuals, like FONDOS) that might be more
appropriate to resource poor small farmers (and small commercial farmers), it is
important to consider alternative approaches to risk management. In practice, SAGAR,
and other government agencies have designed and implemented several programs that
can help resource poor small farmers (but also small commercial farmers) better manage
risks – from different sources – by increasing household incomes, improving the
household (and community’s) asset base, and providing safety nets in times of economic
crisis (resulting from weather and/or economic factors). What is really needed is greater
recognition of how these programs can improve risk management and how the programs
can be better coordinated in a more holistic framework for rural risk management.
22
From table 4.6, a small maize grower on rainfed land with low level of production technology would
expect a net revenue of N$235 per ha., while crop insurance had cost him/her N$386 per ha. Thus, had
he/she not paid for insurance his/her net income would have more than doubled. The same farmer but in an
irrigated area using high production technology would expect a revenue of N$1,949 and had to pay N$432
for crop insurance. Of course in either case crop insurance may have been necessary to obtain credit.
38
Table 4.6: Representative Crop Budgets per Hectare for Selected Rainfed and
Irrigated Crops in Mexico D.F., Guanajuato, Queretaro, PV Cycle 2000
Mexico, D.F
Maize
TCF
2.5 ton/ha
1,500/ton
3,750
Maize
TMF
3.5 ton/ha
1,500/ton
5,250
Input Costs N$,
of which:
Land Preparation
Seeds and Sowing
Fertilizer and Application
Cultivation (e.g., weeding)
Irrigation
Disease & Pest Control
Harvest
2,830
Guanajuato and Queretaro
Maize GMF
Bean GMF
7 ton
1,350/ton
9,450
2 ton
5,000/ton
10,000
Sorghum
GMF
9 ton
1,100/ton
9,900
3,849
6,643
6,306
6,101
600
325
645
400
600
750
954
400
340
520
465
680
1,225
900
1,213
690
600
1,385
630
1,200
1,100
803
760
600
783
1,060
1,225
800
1,213
440
600
968
855
Insurance & Finance Costs
N$, of which:
Crop Insurance
“Life Insurance”
Finance Costs
685
818
858
803
952
386
8
291
386
12
420
432
15
411
313
17
473
504
15
433
Total Costs N$
% Crop Insurance Costs
% Insurance/Finance Costs
3,515
11%
19%
4,667
8%
18%
7,501
6%
11%
7,109
4%
11%
7,053
7%
13%
Expected Net Revenues N$
235
583
1,949
2,891
2,847
Expected Yield
Expected Price N$
Expected Gross Revenue
N$
Net Rev/Gross Rev
6%
11%
21%
29%
29%
Source: Banco de Credito Rural del Centro, S.N.C., 2000.
NOTES: TCF: rainfed, criollo seed, fertilized; TMF: rainfed, improved seed, fertilized; GMF: surface
irrigation by gravity, improved seed, fertilized
103. Below is a brief description of some of the major programs and initiatives in rural
areas, with a focus on their role in risk management for resource poor small farmers.
Income Support Programs
104. PROCAMPO is an income support program for agricultural producers of basic
commodity crops. PROGRESA and PET (Temporary Employment Program) are income
support programs aimed at poor households in rural areas.
PROCAMPO
105. The PROCAMPO program was designed to compensate farmers for anticipated
declines in commodity prices as a result of market liberalization (Casco and Rosezweig,
2000; Claridades Agropecuarias, 2000). PROCAMPO payments are lump-sum payments
per hectare that are not linked to output. In recent years, PROCAMPO payments are
around $70-80/hectare per crop cycle range. PROCAMPO payments are, still an
39
important component of household income for many small farmers. For example, over
83% of ejidatarios surveyed in 1997 received PROCAMPO payments (see Table 4.2).
PROCAMPO payments provided about 10% of household income in the ejido sector in
1997, but reached up to about 40% for some low-income households (World Bank,
1999a).23 In 1999, PROCAMPO payments were made to 3.2 million farmers, about 85%
of them with less than 5 hectares.24
106. Changes in the PROCAMPO program could increase its real value for small
farmers and possibly have some important risk management implications. First, the
timing of PROCAMPO payments will be changed so that small farmers will receive them
before planting time. For small farmers that have been pledging their PROCAMPO
payments for discounted cash advances, this will increase the real value of their
PROCAMPO payments and increase their net income. Also guarantee them a higher
minimum income that will help small farmers manage risks. Second, small farmers with
fewer than 5 hectares will no longer be required to plant to qualify for payments. This
will have several potential risk management benefits for small farmers by allowing them
to diversify their cropping patterns and/or sources of income. Thus, there is potential to
increase their expected income and stabilize their total household income.
PROGRESA,
107. PROGRESA, which was introduced in 1997, is an income support program
administered by the Ministry of Health (and by a branch of the Mexican Social Security
Institution), which provides monetary and in-kind payments to poor households in rural
areas. It is targeted to help reduce current and future poverty by providing assistance for
education, health and nutrition. In 1999 PROGRESA covered about 2.3 million
households – about 90% in rural areas – including about 40% of all rural households in
Mexico, with an average monthly transfer worth about $22/month or about $264/year
(the equivalent of about 3.5 hectares of PROCAMPO payments), and about 20% of the
mean value of consumption of poor rural households. Although individual households
receive the PROGRESA payments, it is communities that qualify. And to qualify, a
community must have a health clinic, and primary and secondary schools.
PET
108. The Temporary Employment Program (PET), which began in 1995, is an important
source of income and risk management for many rural households. Originally supported
by the Secretariat for Social Development (SEDOP), with an urban focus, it has
increasingly been expanded to rural areas, and SAGAR – through the Sub-secretariat of
Rural Development, with FIRCO as the technical agent - has also begun devoting
budgetary resources . PET jobs include work related to improving basic infrastructure,
roads, highways, water resource development and irrigation, and reforesting and soil
conservation projects.
23
For small farmers obtaining a yield of about 1 ton/hectare of maize (the average yield/hectare in the ejido
sector), a $70-80 payment PROCAMPO per hectare represents about 50% of the market value of maize
produced (assuming a price of about $150-160/ton).
24
Despite the fact that the majority of the total value of PROCAMPO payments goes to farmers with more
than 5 hectares(Sadoulet, de Janvry, and Davis, 1999).
40
109. Given the small sizes of their landholdings and low-technology cropping practices,
farming (especially rainfed one-cycle per year) does not provide full-time employment
for small farmers and their households. Thus, seasonal employment can offer
opportunities to increase household income and lower the relative importance of income
risks related to on-farm production. Also, in times of natural (or economic) disasters,
temporary employment can be an important safety net25 that helps households cope with
risk (as was noted previously, seeking off-farm employment is the major risk
management strategy for small farmers).
ASERCA’s Marketing Support Program and Price Risk Management Program
110. As part of reforms in commodity markets in Mexico, GOM supports the
agricultural sector via a program of direct income support – PROCAMPO- a program to
improve the marketing of agricultural products, and a program to provide market-based
price commodity risk management. These programs are administered through ASERCA
(Apoyos y Servicios a la Comercializacion Agropecuaria).
111. On price risk management, ASERCA provides technical support to educate farmers
about commodity markets and subsidizes (about 50%) the cost of various hedging
options. Although ASERCA’s subsidized hedging operations have increased
significantly since 1996, they still account for a relatively small share of the national
production of agricultural commodities, with little participation by small farmers (see
Volume of Annexes on ASERCA’s price risk management program). Small farmers’
participation in the price risk management is mainly through producer organizations and
is concentrated in the Northern part of the country. On supporting agricultural marketing
ASERCA helps to improve agricultural marketing institutions and infrastructure.
Improvements are needed, for example, in quality, grading and standards; trade
regulation and enforcement; information services; financing for marketing; storage and
transport infrastructure; research, technical assistance and extension for small farmers;
and small farmer associations (see World Bank, 2000a).
Asset Development and Enhancement and Technology Adoption Programs
112. Besides income support programs, it is also important to develop and enhance the
asset base (broadly defined) of small farmers and to help them adopt improved
technologies. Asset development and enhancement, and technology adoption programs
can, in turn, lead to longer-term sustained increases in incomes – from both agricultural
and non-agricultural activities. In particular, to counter the declining profitability of the
production of basic commodity crops, it is important for farmers to adopt improved
technologies that are more productive, and/or diversify into higher value agricultural
enterprises. There are several programs aimed at these objectives. It should be noted,
however, that adoption of new technologies and higher returns might be associated with
higher risks.
25
PET is used in conjunction with FONDEN’s disaster assistance programs (as discussed later).
41
ALIANZA PARA EL CAMPO
113. The Alianza para el campo (ALIANZA) program was started in 1996. It helps
small-medium farmers, organized in groups, to invest in new technologies that increase
productivity and promotes a more commercial orientation in their crop production
patterns (e.g., higher-value crops). About 10% of all respondents in the 1997 Ejido
Survey reported that they participated in the ALIANZA program (Table 4.2). Because it
requires group participation for investment in community-level productive infrastructure,
there is a natural bias towards more socially cohesive communities. Also, there is a
tendency towards higher participation by farmers with more than 5 hectares. There is
also an associated program, Apoyos Productivos en Zonas Marginadas (rural
development program for marginal areas), that targets poorer small farmers26 in
communities located in more marginal agro-ecological regions (many of them indigenous
communities.)
114. It is important to note that, in addition to the federal SAGAR budget, ALIANZA
receives funding from state governments under a matching-grant scheme. States also
supply effective support services (research, extension, information, technology) for a
wide range of sub-programs. Therefore, small farmers in states and municipalities with
well-functioning governments dedicated to agricultural and rural development, and to
rural risk management, are at a distinct advantage.
115. ALIANZA clearly aims at strengthening the asset bases (e.g., physical, human,
financial, social, infrastructure, and political/institutional) of small farmers to provide
them with opportunities to attain higher and more sustainable incomes. In addition, there
is an important environmental component of the programs to enhance and maintain small
farmers’ limited natural asset – land.
116. Small farmers who participate in ALIANZA are potential clients for formal credit
and crop insurance (and price insurance) – especially if they are abandoning low-return
low risk-activities for high-return high-risk activities. It is important for these programs
to help farmers improve their organizational skills and institutional structures so that they
can access credit and insurance as an association (or group) – thereby improving the
access and reducing the costs of credit and insurance, and improving access to markets
for inputs and outputs.
26
Defined as farmers with less than 20 hectares of rainfed land or 5 hectares of irrigated land.
42
Table 4.7: Possible Options for Rural Households to Manage Risk
Risk reduction
Risk Mitigation
Asset portfolio
Management
Insurance
Household Level
Investment to protect, maintain
and enhance assets
Adopt new technology
Adjust asset portfolio and
income-generating activities
Permanent migration
Community Level
Investments in physical and
social infrastructure
Social ties and network
Participation in community
institutions and decisionmaking
Rules and regulations
Rights and security
Extra-Community Level
Information on risk and risk
reduction
Rules and regulations
Guaranteed rights and security
Stable macro-economy, policy
regime, and political system
Open and free markets
Responsive institutions
Investments in public goods,
physical and social
infrastructure
Adjust asset portfolio and
income-generating activities
Hold financial or non-financial
assets (e.g., livestock, food
stocks) for precautionary savings
Seasonal migration
Formal insurance
Informal insurance based on intrahousehold social capital claims
Inter-linked contracts
Markets for household assets
Physical and social
infrastructure
Markets for household assets
Market information
Investments in physical and
social infrastructure
Informal insurance based on
community social capital
claims
Formal community insurance
pooling associations
Community credit unions and
savings clubs, and “banks”
for other asset stocks
Formal insurance, private and
public sector, and international
organizations (e.g., crop
insurance, health insurance)
Disaster aid funds
Financial systems, national and
international
Inter-community credit
associations and “banks” for
other stocks
Targeted safety nets (transfers,
public works)
Social investment projects
(e.g., social funds)
Depend on charity or aid from
national or international
organizations
International food aid
Donor assistance
Finance
Formal and informal credit
Inter-linked contracts
Risk coping
Draw down assets (e.g., skip
meals, mine soil, not pay school
fees)
Use underemployed assets (e.g.,
off-farm employment, child labor)
Sell assets
Encroach on assets of others
Illegal activities
Formal and informal credit
Depend on charity
Draw down community
assets (e.g., reduce
maintenance, harvest or mine
natural resources)
Depend on charity or aid
from outside community
Source: Siegel and Alwant (1999.).
43
5. FONDEN AS A PROVIDER OF AGRICULTURAL INSURANCE
Background
117. The Fund for Natural Disasters (FONDEN) was established in 1996 within the
Ministry of Finance as a last resort source of post-disaster financing inter alia for
reconstruction of public infrastructure, and compensation to low income producers for
crop and livestock losses arising from natural disasters. In terms of expenditures,
Fonden’s coverage of public infrastructure is much larger than its coverage of crop and
livestock losses. Recent estimates put the total value at risk of public sector assets at
N$10 trillion, of which N$4.5 trillion represents Fonden exposure27.
118. To date, average losses per year to all public sector assets (insured and uninsured)
due to earthquake, hurricane and flood are N$6.6 billion, N$3 billion and N$3 billion,
respectively. Losses to agriculture are included in those amounts; additional agricultural
losses result from drought and frost. Still, program rules limit Fonden’s exposure to crop
and livestock losses and for 1997-1999, total FONDEN payments to agriculture have
averaged about N $333 million annually.
119. FONDEN is a disaster relief program rather than an insurance program; however
the agricultural component of the program has characteristics similar to privately
provided crop insurance. For example, FONDEN makes payments in response to many
of the same perils covered by private insurers – for example, wind, drought and frost. In
addition, claims of loss trigger field inspections as do claims to crop insurers in Mexico.
120. There are key differences between FONDEN and private insurance, however.
FONDEN payments are triggered only when droughts, frost or other perils affect most
producers in a region – that is, FONDEN only pays out against systemic risks. Thus
individual farmers that lose their crops due to idiosyncratic risks are unprotected.
Moreover, payments from FONDEN are made only after the declaration of a disaster by
SEGOB following a formal request from the State Governor or head of a federal
Ministry, after a potentially conflictive loss adjustment process has been completed and
agreed to by an inter-secretarial committee, and after state or municipal co-participation
payments have been made to state trust funds. This is a time-consuming and potentially
conflictive process, despite the strong guidance provided by Fonden’s rules. Indeed,
there are reports that it may take 5-6 months for FONDEN payments to be actually made
to state trust funds, thereby engendering liquidity problems and complaints at the state
level.
How well does FONDEN protect small holders?
121. Only smallholders are eligible to receive FONDEN payments; however, the
definition of smallholder varies according to regional and agronomic differences.
Farmers who own less then twenty hectares in Baja California, Baja California Sur,
Campeche, Coahuila, Colima, Chihuahua, Durango, Jalisco, Nuevo Leon, Sinaloa,
Sonora, Tabasco, Tamaulipas, Veracruz and Zacatecas are considered smallholders. In
27
These preliminary estimates are reported in Managing the Financial Impacts of Natural Disaster Losses
in Mexico: Government Options for Risk Financing and Risk Transfer, World Bank, draft.
44
Aguascalientes, Chiapas, Guanajuato, Michoacan, Nayarit, Quintana Roo, San Luis
Potosi and Yucatan, the definition restricts owners to ten hectares or less. In Distrito
Federal, Guerrero, Hidalgo, Mexico, Morelos, Oaxaca, Puebla, Queretaro and Tlaxcala
smallholders own five hectares or less. FONDEN also restricts the number of hectares
eligible for payments in order to limit payments to any one farmer. In addition, irrigated
land and insured lands are not eligible for FONDEN payments. Payments vary with type
of crop. Maximum payments and area limits are reported in Table 5.1.
Table 5.1: Limits on FONDEN payouts
Crop
Annual crops
Coffee
Fruit and other perennial crops
Nopal (cactus) plantations
Tropical forest
Temperate forest
Livestock
Cattle
Source: SHCP
Maximum payment per hectare (pesos)
281
315
782
281
731
731
Maximum payment per head
197
Maximum area
5 hectares
3 hectares
2 hectares
3 hectares
5 hectares
10 hectares
Maximum number
25
122. Table 5.2 reports payments for agricultural losses from FONDEN from 1997-99;
these totaled nearly N$ 1 billion for the three year period. Generally, the FONDEN
payments were spread out among many states, with no single state accumulating more
than 6% of total pay outs. The exceptions are Puebla with 9% of total payouts and
Oaxaca with 14%.
123. When maximum Fondo payments are made, analysis shows that they are likely to
cover a significant portion of the investment smallholders have made in the lost crop.
Using sample data from the 1996 ejido survey, average production costs were calculated
for producers owning five or less hectares. Average revenue was calculated as well;
these are reported in Table 5.3. The costs and revenue were then adjusted for inflation
and compared to the maximum FONDEN payment of 281 pesos per hectare for annual
crops. The table shows that, on average, FONDEN payments cover nearly 80% of the
irreversible investments farmers make in inputs and labor when planting their crops.
There are wide regional differences however depending upon the level of inputs
commonly used – that is, the crop-investments of farmers that use low-input techniques
are better protected by FONDEN than the crop-investments of farmers using more costly
intensive techniques.
45
Table 5.2: FONDEN payouts managed by SAGAR, 1997-1999
1997
Aguascalientes
Baja California
Baja California Sur
Chiapas
Chihuahua
Coahuila
Colima
Durango
Guanajuato
Guerrero
Hidalgo
Jalisco
Mexico
Michoacan
Nayanl
Nuevo Leon
Oaxaca
Puebla
Queretaro
San Luis Potosi
Sinaloa
Sonora
Tabasco
Tamaulipas
Tlaxcala
Veracruz
Zacatecas
Total
Source: SHCP
1998 1999
Total
Million pesos
2.0
2.8
1.5
6.3
4.0
4.0
1.9
2.9
4.8
18.0
9.1
6.3
33.4
5.0 20.2 21.0
46.2
4.5
4.6
4.1
13.2
1.4
1.4
5.0 38.6 21.1
64.7
25.0
25.0
33.0 29.0
62.0
6.0 23.7 21.3
51.0
20.0
1.9
21.9
30.0
30.0
15.0
5.9
6.2
27.1
4.9
4.9
10.0 10.1
7.7
27.8
58.0 37.8 49.4
145.2
19.0 27.1 40.0
86.1
1.2
2.8
1.4
5.4
34.2 15.1
8.1
57.4
9.2 13.6
22.8
16.0
3.2 11.2
30.4
4.5 43.4
47.9
5.5 30.5 16.5
52.5
3.5 10.0
13.5
3.0 18.4 41.6
63.0
17.0 10.5 33.5
61.0
304.9 349.9 354.2 1,009.0
124. Looking at the revenue side, regional differences emerge as well. However the
analysis shows that in no region do FONDEN payments compensate farmers for lost
income – covering on average only 8% of expected revenue.
Table 5.3: Variable field costs, average revenue and FONDEN payments for
producers owning less than five hectares of land.
Region
North
North Pacific
Center
Gulf
South Pacific
Average
Variable input costs
1999 pesos per
FONDEN payment
hectare
as share of costs
75%
376
1,172
24%
979
29%
248
114%
754
37%
357
79%
Average revenue
1999 pesos per
FONDEN payment
hectare
as share of value
1,675
17%
2,644
11%
7,668
4%
1,424
20%
7,935
4%
3,658
8%
Source: 1997 ejido survey.
Note: Variable costs include: seeds, labor, fertilizer, insecticide, herbicide, technical assistance, irrigation
46
What is the market value of FONDEN coverage?
125. As already mentioned, FONDEN payments are triggered through a discretionary
process, and for this reason it is difficult to determine the likelihood of a FONDEN
payout. However, setting aside political uncertainties, FONDEN guidelines provide strict
definitions of certain types of perils. These guidelines are summarized in table 5.4.
126. For perils such as drought and frost, the guidelines are similar to the types of
triggers found in parametric insurance. Unlike regular crop insurance, parametric
insurance does not directly compensate for assessed losses, but rather pays out when an
agreed upon indicator meets an agreed upon condition – for example when the
temperature recorded at a defined weather station falls below a certain level.
Consequently, transaction costs associated with the insurance are lower since field
assessments of damage are not required. Parametric insurance is also easier to price,
since the expected pay outs from the insurance can be estimated by calculating from
historic data the probability of the trigger condition being met.
Table 5.4: Guidelines for selected operations of FONDEN.
Hazard
Activity
Trigger
Drought
Livestock
Two consecutive months of average rainfall below either 50% of its
historical average or its historical minimum.
Crops
As above, but for any period of time during which damages may occur.
Horticulture
Minimum temperatures. Beans 3. Potato –2, Tomato 0, Watermelon 3,
Cabbage 0, Melon –1.
Fruits
Minimum temperatures. Apple –34, Orange –2, Peach –26.
Grains
Minimum temperatures. Corn 0-5, Sorghum –6, Wheat –9, Oats –8.
Fibers
Minimum temperatures. Cotton –1.
Any
Not specified. In general, an unusual occurrence is defined as event that
is below one standard deviation of its historical average.
Frost
Hail and snow
Source: Diario Oficial de la Federación, February 29, 2000.
127. This approach is used to calculate the expected FONDEN payouts for drought
reported in table 5.5. FONDEN rules state that livestock owners are eligible for drought
payouts when cumulative rainfall is below either 50% of its historical average or
historical minimum for two consecutive months (any combination of the two events is
allowed); small crop producers are eligible when rainfall is below these thresholds for a
period of time judged to be sufficient to have generated permanent damages to the crops.
Although this common rule would apply to all states, the variance in rainfall differs
among states and consequently the probability of drought, as defined by FONDEN rules,
will differ among states as well. These probabilities, based on weather data from IMTA
are reported in the first two columns of table 5.528. In the next two columns, the expected
28
The probabilities of rainfall were calculated by fitting a standard gamma distribution to average monthly
State-level measurements for the months May to November. Probabilities of a drought being declared were
then computed on the assumption that the level of monthly average rainfall is distributed independently
47
probabilities are multiplied by the FONDEN payment levels of 281 pesos per hectare for
crops and 197 pesos per head for cattle to arrive at expected payments. The analysis
shows that the drought coverage offered by FONDEN is worth about 190 pesos per
hectare in Colima and worth only 33 pesos in Chiapas. Livestock owners in Chiapas can
expect negligible payments from FONDEN for livestock affected by drought, while small
produces in Baja California can expect about 35 pesos per head of cattle.
128. FONDEN does not only pay out for drought – a coverage important for Baja – but
also for flooding, for excess rains and for other calamities. (Figure 1 shows the areas of
Mexico most subject to flooding.) When calculating the value of FONDEN to
smallholders, each peril covered adds to the probability that FONDEN payments will be
triggered. Calculating this number precisely is impossible; since not all perils are defined
parametrically by FONDEN rules. However this general point is illustrated for two
parametrically triggered insurance: drought and frost. As with drought, frost is strictly
defined by FONDEN rules. A frost is said to occur when temperatures fall below a
defined temperature (recall table 5.1) during critical growing months. In turn, critical
growing months will differ by crop. Consequently, probabilities associated with the frost
trigger will differ by state and by crop. Moreover, payment levels differ among field
crops and horticultural crops. Table 5.7 reports the probable payment from FONDEN for
either drought or frost29. For these two perils, producers in Colima, Morelos and Sonora
can expect to receive payments with high frequency.
Figure 5.1: Risk of flooding
High
Medium
Low
across months. The distribution of average State-level temperatures was assumed to be distributed as a
normal and monthly realizations as independent from each other.
29
The joint probability is 1 minus the probability of neither frost nor drought occurring.
48
Table 5.5: Expected FONDEN payment for drought, by state
Probability of drought
State
Chiapas
Puebla
Guerrero
Veracruz
Yucatan
Q.Roo
Oaxaca
S.L.Potosi
Tlaxcala
Michoacan
Campeche
Jalisco
Queretaro
Chihuahua
Tabasco
N.Leon
Taumalipas
Mexico
Zacatecas
Guanajuato
Coahuila
Aguascalientes
Hidalgo
Sinaloa
Durango
BajaCalifornia
BC.Sur
Nayarit
Sonora
Morelos
Colima
Average
Livestock
Crops
0.3%
1.5%
2.3%
1.2%
1.0%
3.0%
3.7%
5.9%
4.3%
4.8%
0.9%
2.7%
8.6%
9.7%
4.0%
7.9%
7.3%
5.1%
5.5%
8.7%
11.4%
9.1%
7.4%
13.8%
10.3%
18.1%
15.0%
10.3%
17.8%
9.1%
16.9%
N/A.
11.7%
12.8%
16.6%
17.4%
21.3%
26.6%
27.5%
28.5%
29.9%
30.1%
30.4%
30.5%
30.9%
32.4%
32.7%
33.3%
33.4%
33.7%
35.5%
37.6%
38.4%
39.1%
40.3%
42.9%
45.4%
46.9%
47.6%
49.3%
49.9%
50.3%
67.8%
N/A.
Expected pay out
Maximum payout
(pesos)
(pesos)
Per head of
Per
25 head of 5 hectares
livestock
hectare
cattle
0.6
32.8
15.10
164.2
3.0
36.0
74.1
179.9
4.5
46.6
112.4
233.2
2.3
49.0
58.2
245.1
1.97
59.9
49.25
299.5
5.8
74.7
145.5
373.5
7.4
77.4
184.1
386.8
11.7
80.1
292.3
400.7
8.4
84.0
210.8
419.8
9.4
84.7
234.0
423.4
1.8
85.6
43.8
427.8
5.3
85.6
133.7
428.0
17.0
86.8
424.1
433.8
19.0
90.9
476.0
454.7
8.0
92.0
199.0
459.8
15.6
93.5
390.2
467.4
14.3
93.9
358.0
469.3
10.0
94.8
250.6
474.1
10.8
99.8
270.9
498.8
17.2
105.6
430.8
528.0
22.5
107.8
563.5
539.0
18.0
110.0
449.0
550.1
14.6
113.2
364.3
566.0
27.3
120.5
681.5
602.5
20.3
127.5
506.9
637.4
35.6
131.8
889.6
659.2
29.6
133.6
740.1
668.1
20.3
138.4
508.6
692.0
35.0
140.3
875.1
701.3
17.9
141.3
448.4
706.3
33.2
190.6
830.0
952.8
15.0
98.2
375.5
491.2
Source: authors’ calculations based on data from SHCP
Table 5.6: Drought-related expenditures by agency, 1997-1999
1997
SAGAR
SEDESOL
CNA
Total
1998
Million pesos
263
349
200
66
217
462
632
1999
140
258
398
Source: World Bank (2001)
49
50
State
corn sorghum wheat barley Cotton beans peach orange apple melon cabbage watermelon tomato potato
Aguascalientes
110
110
110
110
110
1
123
123 123
123
123
124
123
123
Baja California
132
132
132
132
132
1
148
148 148
148
148
149
148
148
Baja California Sur 134
134
134
134
134
0
150
150 150
150
150
150
150
150
Campeche
86
86
86
86
86
0
96
96
96
96
96
96
96
96
Coahuila
108
108
108
108
108
0
121
121 121
121
121
121
121
121
Colima
191
191
191
191
191
0
214
214 214
214
214
214
214
214
Chiapas
116
33
33
33
48
87
37
47
37
54
66
148
66
47
Chihuahua
199
91
91
91
92
155
102
102 102
103
110
253
110
102
Durango
135
127
127
127
127
20
143
143 143
143
143
159
143
143
Guanajuato
106
106
106
106
106
0
118
118 118
118
118
118
118
118
Guerrero
47
47
47
47
47
0
52
52
52
52
52
52
52
52
Hidalgo
114
113
113
113
113
2
127
127 127
127
127
128
127
127
Jalisco
86
86
86
86
86
0
96
96
96
96
96
96
96
96
Mexico
96
95
95
95
95
3
106
106 106
106
106
109
106
106
Michoacan
85
85
85
85
85
0
95
95
95
95
95
95
95
95
Morelos
141
141
141
141
141
0
158
158 158
158
158
158
158
158
Nayarit
93
93
93
93
93
0
105
105 105
105
105
105
105
105
N. Leon
138
138
138
138
138
0
155
155 155
155
155
155
155
155
Oaxaca
77
77
77
77
77
0
87
87
87
87
87
87
87
87
Puebla
36
36
36
36
36
0
40
40
40
40
40
40
40
40
Queretero
96
75
75
75
81
26
84
89
84
91
94
111
94
89
Q.Roo
87
87
87
87
87
97
97
97
97
97
97
97
97
S.L.Potosi
80
80
80
80
80
0
90
90
90
90
90
90
90
90
Sinaloa
121
121
121
121
121
0
135
135 135
135
135
135
135
135
Sonora
140
140
140
140
140
0
157
157 157
157
157
157
157
157
Tabasco
92
92
92
92
92
103
103 103
103
103
103
103
103
Taumalipas
94
94
94
94
94
0
105
105 105
105
105
105
105
105
Tlaxcala
89
84
84
84
84
11
94
94
94
94
94
106
94
94
Veracruz
49
49
49
49
49
0
55
55
55
55
55
55
55
55
Source: SHCP and authors’ calculations.
Table 5.7: Expected pay out from FONDEN by state and by crop due to either frost or drought
Implications for smallholders
129. Although the rules for drought and frost are reasonable and technically well
defined, the analysis above shows that the rules result in differing levels of coverage for
different regions. Drought protection is greater in areas where the variance of rainfall is
greater and frost coverage is greater for colder climates and for crops whose growing
seasons makes them most susceptible to frost. Consequently, the FONDEN rules
unintentionally reward risky behavior. Such perverse incentives can be easily changed
by rewriting FONDEN rules so that payouts are given equal probability across regions.
For example, using historic weather data, define drought as occurring when the rainfall
for two consecutive months falls below a trigger defined as having a 10% probability of
occurring. In terms of centimeters of rain, this trigger will be closer to average rainfall
levels in Chiapas than it will be in Baja.
130. The rule that makes smallholders who purchase private insurance ineligible for
FONDEN payments also unintentionally encourages smallholders to take on more risk.
For example, methods of crop financing through parafinancieras are growing in Mexico
– due in part to incentives provided by the government through FIRA guarantees. Often
crop insurance is a precondition for participation. In addition, price insurance is
frequently required as well. The practice facilitates credit, and also helps protect
variability in farm income. (These schemes are described in more detail later in the
report.) However, smallholders who participate lose FONDEN eligibility – a loss that
may make alternative methods of risk management, including participation in
parafinanciera, less attractive. Consequently, FONDEN and FIRA inadvertently work at
cross-purposes.
Comparisons with parametric insurance
131. As already discussed, the parameters for drought is well defined within FONDEN
rule, but there is room for discretion in the application of the rules. Having a measure of
the cost of discretion can provide insights into how FONDEN functions. Unfortunately,
the task of calculating the discretionary component of FONDEN is hampered by a
number of difficulties. Nonetheless, we provide below a partial comparison, based on
estimated and recent historic payments for drought.
132. In making the comparison, it is worth noting that SAGAR is only one of three
agencies that make payments related to drought. Both CNA and SEDESOL make
payments as can be seen in Table 5.6. In this section we are only able to compare the
amount payable by SAGAR for crop damage. We find that the historic FONDEN crop
related payouts for drought are in line, but slightly below, rough estimates based on
probabilities. This may mean that current FONDEN criteria are triggered too frequently
and that the discretionary component of FONDEN has worked to limit expenses.
However, it also shows that parametric triggers can generate similar payments. This
finding is relevant for all programs, since parametric triggers could be used for social
protection expenditures by CNA and SEDESOL that are additional to payments for croprelated losses.
51
Are there benefits to parametric triggers for FONDEN?
133. If suitable triggers can be found, there are several good technical reasons for
converting to a scheme that pays out based solely on parametric triggers. First, triggers
can be derived that provide equal protection to all smallholders. Second, reinsurance
becomes straightforward, since payouts are based on historic probabilities that reinsurers
can easily understand and price. Third, the process is quick and low cost since field visits
are not required.
134. As discussed in the previous section, current rules provide varying levels of
protection – as measured by the probability of a payout – among farmers growing
different products and among farmers growing the same products in different states.
Consequently, the program protects some farmers more than others. Designing triggers
that have an equal probability of payout regardless of crop type or location would
eliminate this distortion. In addition, because compensation depends on crop damaged –
as measured by field inspectors – farmers who take preventative action to limit crop
losses are penalized. In contrast, payouts based solely on weather conditions would
restore normal incentives to take preventative action. Of course, any free insurance
distorts by providing marginal compensation without cost. This can also be reduced by
designing the triggers so that the pay out is only for rare events. For example, farmers
are likely to discount a one-in-fifty-years drought.
135. If short-term budget issues can be solved, there are often advantages for
governments who self-insure. However, in the case of Mexico state governments
especially have been hard pressed to meet recent requirements for matching (70/30)
funds. Purchasing reinsurance from private firms converts large lump-sum expenditures
into annual premiums that can be more easily budgeted. Doing so however will come at
a price, which in turn, is related to the probability of payouts. For parametric insurance
these costs can be calculated in a straightforward way. In contrast, when payouts include
a measure of discretion, reinsurers will have difficulty pricing the probability of payouts
and may be reluctant to reinsure.
136. Because field visits are not required, the administrative costs to FONDEN should
fall if parametric insurance is adopted. For this report, we did not have administration
costs for FONDEN and this perhaps should be the subject of a further investigation.
137. Breaking the link between payouts and field inspections has a cost as well. The
most significant drawback of parametric triggers is the potential mismatch between
conditions at weather stations and effects in the field. This basis risk has an analogy with
price hedging where local prices may differ from price observed in Chicago or other
hedging centers. Whether the basis risk for parametric insurance overwhelms the
significant benefits of parametric insurance is an empirical question. Consequently, field
testing of parametric triggers is recommended before Fonden’s present approval and
safeguard systems are changed.
138. The following is a set of recommendation for FONDEN
•
Remove the restriction that prevents Fonden’s recipients from purchasing private
insurance.
52
•
Set FONDEN rules so that the probability of FONDEN payouts are constant across
all regions and crops. Avoid coverage for commonly (frequently)30 occurring events.
•
Assess whether FONDEN agricultural liabilities present budgetary risks, either at the
federal or state levels. If so consider reinsuring a portion of FONDEN risks.
•
On a pilot basis, restate FONDEN rules in terms of parametric triggers and compare
experiences with existing rules.
30
Being a disaster insurance, FONDEN should try to set its parametric triggers in such levels as to reduce
the probability of pay-outs for events that happen with frequency of more than 1 in 6 or 7 years.
53
# UPR
Total land
used (Ha)
Total
Total Has
number of of eligible
eligible
private
private
rainfed
farmers with
land
some
rainfed land
Estimated
Total
Total
Total
Coefficient Estimated Total eligible
farmers
eligible
eligible
eligible land payout per
parceled
of
(Private and (private and hectare
rainfed
farmers in eligible land cultivated
social
social
social
in social land, social cultivated
sector
sector)
land, social
sector
sector- Has)
sector
sector (Has)
under Fonden’s rule of operation
54
Group A: Up to 20 hectares.
Baja California
5.5
2,176
20,029
594
1,810
5,652
26,105
55.7
794
6,246
2,604
131.8
Baja California Sur
0.4
248
777
107
229
3,046
9,405
34.0
13
3,153
242
133.6
Campeche
92.0
2,814
6,206
2,643
4,959
4,963
20,404
81.0
15,210
7,606
20,169
85.6
Coahuila
50.0
2,752
16,974
1,754
6,273
32,833
138,759
77.9
54,066
34,587
60,339
107.8
Colima
48.1
1,051
6,949
823
2,787
8,915
40,214
35.3
6,822
9,738
9,609
190.6
Chihuahua
67.4 11,563
93,959
6,537
27,996
33,081
153,013
95.6
98,566
39,618
126,563
90.9
Durango
81.3
4,997
36,307
3,931
16,543
58,539
261,496
94.0
199,928
62,470
216,471
127.5
Jalisco
87.6 29,198
155,988
27,243
98,795
58,514
234,020
84.2
172,577
85,757
271,372
85.6
Nuevo Leon
67.7 10,262
53,213
8,274
28,089
24,007
102,731
69.0
47,999
32,281
76,089
93.5
Sinaloa
53.7
4,878
41,198
2,658
9,280
80,712
384,240
87.5
180,678
83,370
189,959
120.5
Sonora
8.4
7,492
54,071
2,174
4,355
19,865
91,600
67.9
5,209
22,039
9,564
140.3
Tabasco
98.9 32,297
114,288
32,049
76,820
67,440
175,623
57.7
100,242
99,489
177,062
92
Taumalipas
71.0
8,338
80,927
5,502
24,041
47,220
224,562
89.6
142,891
52,722
166,932
93.9
Veracruz
93.8 101,632
317,839
100,033
240,128
169,290
637,708
85.7
512,436
269,323
752,564
49
Zacatecas
88.2 21,701
138,381
18,934
78,583
70,887
326,273
86.1
247,816
89,821
326,398
99.8
Group B: up to 10 hectares
Aguascalientes
65.6
3,044
8,915
1,473
3,986
9,484
41,611
94.8
25,884
10,957
29,870
110
Chiapas
98.0 31,447
88,839
31,001
82,402
35,056
121,055
84.0
99,689
66,057
182,091
32.8
Guanajuato
60.9 26,694
90,190
22,535
67,825
70,612
254,835
96.4
149,645
93,147
217,470
105.6
Michoacan
70.2 33,139
84,711
29,076
67,912
59,196
212,610
90.0
134,404
88,272
202,316
84.7
Nayarit
89.6
1,682
5,204
1,501
3,810
22,230
84,673
74.2
56,299
23,731
60,109
138.4
Quintana Roo
96.8
1,758
4,373
1,751
4,392
134
438
92.0
390
1,885
4,782
74.7
San Luis Potosi
87.9 13,025
34,159
11,997
30,614
59,060
207,399
90.7
165,419
71,057
196,033
80.1
Yucatan
97.0
3,574
7,971
3,180
7,356
8,872
23,588
92.7
21,217
12,052
28,573
59.9
Group C: up to 5 hectares
Guerrero
95.8 17,766
35,455
16,543
32,713
42,340
103,380
86.1
85,248
58,883
117,961
46.6
Hidalgo
78.3 47,537
65,952
34,589
48,288
57,524
131,672
96.4
99,443
92,113
147,732
113.2
Mexico
82.9 68,156
77,440
55,985
62,571
173,525
230,087
97.8
186,538
229,510
249,109
94.8
Morelos1/
70.2
4,224
6,364
3,353
4,763
22,898
50,275
93.7
33,056
26,251
37,819
141.3
Oaxaca
95.1 55,620
88,956
50,903
77,538
29,174
60,764
89.4
51,635
80,077
129,173
77.4
Puebla
88.7 153,099
205,307
140,383
185,042
103,672
239,682
96.2
204,610
244,055
389,652
36
Queretaro
71.3
7,814
12,001
6,581
10,081
19,484
51,952
95.0
35,194
26,065
45,275
86.8
Tlaxcala
89.6 30,610
25,261
27,669
22,919
33,413
78,332
98.4
69,027
61,082
91,946
84
Total estimated payout
Source: SAGARPA, Diacon information system, 2000. Procuraduria Agraria, SRA and staff computations.
1. Information on percentage of parceled land cultivated not available for Morelos. Instead, the average of the coefficients of cultivation of Mexico, Guerreo and Puebla was used as an estimate.
% rainfed
Table 5.8: Expected outlays for drought-related damages to crops
5,496,988
16,723,223
23,615,542
5,343,793
9,998,003
14,027,481
3,929,851
7,723,487
367,840,623
3,285,710
5,972,573
22,964,840
17,136,168
8,319,094
357,245
15,702,280
1,711,536
343,146
32,297
1,726,444
6,504,508
1,831,392
11,504,543
27,600,080
23,229,421
7,114,279
22,890,019
1,341,851
16,289,728
15,674,876
36,875,653
32,574,569
Total
estimated
annual
payout
6. INSURANCE AND CREDIT
Background
139. Overall, AGROASEMEX, the Fondos and private insurers insured about 2 million
hectares of agriculture land in 2000. During the same year, about 5.5 million hectares
received credit for crop production. From these, ASERCA reported that about 2.8 million
hectares of PROCAMPO payments was endorsed to financial institutions and thus
insurance is not required in order to receive credit. This only leaves about 700,000
hectares with credit but no insurance.31
140. According to estimates from the development bank, most medium and small
farmers that receive credit purchase insurance. This is because BANRURAL requires the
purchase of crop insurance as a precondition of crop production loans. Moreover, FIRA
offers the strong incentive of improved guarantees to encourage rural credit institutions
and commercial firms to require crop insurance. AGROASEMEX staff estimate that most
insured hectares – at least 1.8 million-- are covered under the FIRA and BANRURAL
programs.
141. An analysis of the 1996 Ejido data survey suggests however, that most smallholder
farmers do not obtain formal credit nor purchase formal insurance. (Table 6.1). Overall,
about 3% of the ejido farmers surveyed in 1996 indicate that they purchase crop
insurance with the higher percentage of those surveyed have 5 hectares of land or more.
On the other hand, 19% of the farmers surveyed report that they receive official credit of
which 11% is from BANRURAL. Aggregate data also shows that the link between
insurance and credit was not as binding at the time of the ejido survey in 1996 but may
have increased as a result of the more recent policies of FIRA and BANRURAL that
encourage risk management at the farm level. Moreover, few ejidatarios participate in the
types of associations that might facilitate credit. According to the survey results, only 2%
of those ejido farmers surveyed reported that they belonged to a mutual insurance fund
(fondo de aseguramiento). And a higher percentage of these farmers had more than 15
hectares of land. Table 6.2 provides aggregate data on insurance and credit by crop for
1993 to 1998. For most crops, the total area receiving credit exceeds the insured area.
There are only few cases where the converse holds (e.g. wheat and sorghum in 1998, rice
in 1996 and horticultural products in 1993).
31
The 700,000 hectares financed without insurance is mainly for larger farmers that can post adequate
collateral and thus, do not need to buy insurance.
55
Table 6.1: Credit, insurance and household characteristics
Landholdings in hectares
0
0.1 to 5 5.1 to 15 >15
All
Organizational participation
Sociedad cooperativa
0.0%
1.0%
3.0% 4.0% 2.0%
Sociedad productores
0.0%
4.0%
4.0% 3.0% 4.0%
Union de credito
0.0%
0.0%
0.0% 0.0% 0.0%
Fondo aseguramiento
0.0%
1.0%
2.0% 5.0% 2.0%
Credit
Recibio credito oficial
0.0% 16.0%
20.0% 25.0% 19.0%
Credito comercial
0.0%
1.0%
1.0% 1.0% 1.0%
Credito Banrurarl
0.0%
7.0%
12.0% 19.0% 11.0%
Pronasol
0.0%
4.0%
4.0% 4.0% 4.0%
Credito organizaciones
0.0%
1.0%
0.0% 1.0% 1.0%
Otra fuente credito
0.0%
4.0%
2.0% 3.0% 3.0%
Solicito credito formal
9.0% 19.0%
24.0% 31.0% 23.0%
Recibio credito informal
17.0% 14.0%
17.0% 18.0% 16.0%
Tiene cartera vencida
0.0%
9.0%
14.0% 17.0% 12.0%
Insurance
Seguro de cosecha
0.0%
1.0%
5.0% 6.0% 3.0%
Seguro para maquinaria
0.0%
0.0%
0.0% 0.0% 0.0%
Seguro de vida
0.0%
2.0%
5.0% 2.0% 3.0%
Seguro medico
17.0%
5.0%
12.0% 13.0% 9.0%
Otro tipo de seguro
0.0%
0.0%
0.0% 0.0% 0.0%
Algun tipo de seguro
17.0%
6.0%
16.0% 16.0% 12.0%
Number of observations
23
511
457
263 1,254
Source: 1996 ejido survey.
56
Table 6.2: Insurance and credit by selected crops
Year
Total
Cotton Rice Cartamo Oats Beans
Corn Sorghum Soybeans Wheat
Other
19981/ Insured hectares
594,985 34,912 13,022
7,385 4,533 186,145 105,955 101,849
12,033 46,352 82,799
Credited
1,381,500 71,500 20,500
327,900 717,100
79,400
19,200
2,100 141,400
hect.PV
(%)
43.1
48.8
63.5
56.8
14.8
128.3
62.7 2,207.2
58.6
1997 Insured hectares
635,542 37,944 18,733
Credited
1,435,100 73,100 37,200
hect.PV
Credited hect.OI 730,500 3,000 1,300
Totalcredit hect. 2,165,600 76,100 38,500
(%)
29.3
49.9
48.7
10,563 4,326 158,666 138,355
400
292,800 686,100
90,548
142,000
8,379
32,200
25,400
25,800
40.9
122,800 203,900
415,600 890,000
38.2
15.5
163,700
305,700
29.6
3,200 167,000 40,200
35,400 173,800 204,500
23.7
35.5
52.0
1996 Insured hectares
653,595 51,779 16,375
Credited
930,000 80,100 14,500
hect.PV
Credited hect.OI 736,600 3,100 1,600
Total credit
1,666,600 83,200 16,100
hect.
(%)
39.2
62.2 101.7
4,633 4,584 182,889 137,047
345,600 244,700
113,803
69,700
7,032
12,200
197,300
267,000
100 191,700 85,300
12,300 213,000 226,200
1995 Insured hectares
405,246
Credited
590,700
hect.PV
Credited hect.OI 920,500
Totalcredit hect. 1,511,200
(%)
26.8
23,000
900
83,100 14,900
52.1
84.3
703 4,413
2,900 10,300
2,600
600
5,500 10,900
12.8
50,100 192,000
395,700 436,700
30.1
43,264 12,561
60,100 14,000
1994 Insured hectares
493,321 13,554 11,481
Credited
509,100 17,900 14,200
hect.PV
Credited hect.OI 666,500 9,700
900
Total credit
1,175,600 27,600 15,100
hect.
(%)
42.0
49.1
76.0
19932/ Insured hectares
580,796
Credited
523,800
hect.PV
Credited hect.OI 619,600
Total credit
1,143,400
hect.
(%)
50.8
15,400
15,400
46.2
64,040 71,413
21,300 140,900
31.4
42.6
57.2
30.1
31.6
974 3,905 31,710 116,194
197,400 132,600
30,583
68,500
13,212
35,900
74,026
5,700
78,817
75,900
18,600
18,600
5.2
30,700 135,700
- 228,100 268,300
13.9
43.3
184,100
252,600
12.1
9,161 2,670
100
39,556 191,927
44,900 228,000
42,443
31,300
9,200
9,300
51,200 200,200
96,100 428,200
165,400
196,700
-
98.5
41.2
6,445 3,776
23,500
23,500
40.5
61,709 106,319
6,800 164,300
-
27.4
- 145,900 381,700
35,900 151,600 457,600
36.8
48.8
17.2
36,781
70,900
21.6
51.9
50,465 178,723
40,100 285,700
49,465
17,200
20,044
55,700
45,400 206,500
85,500 492,200
115,500
132,700
36.3
37.3
65,302
91800
- 188,400 41,600
70,900 197,900 133,400
44.8
59.0
80,446
9,500
40.6
49.0
86,022 180,740
4,200 106,800
- 188,000 37,500
55,700 192,200 144,300
36.0
44.8
125.3
Source: BANRURAL
Does crop insurance facilitate credit?
142. Banks extend loans based on their perception of the borrowers ability and
willingness to repay. Potentially, crop insurance can augment the later. Of course loss of
crop is only one threat to loan, and may or may not be a significant one. In such instances
the banks would prefer to do away with the requirement for crop insurance. Banks
typically need a collateral of 1.5 to 2 times the loan amount. For loans to smaller and
poorer farmers, BANRURAL staff feels that crop insurance provides at least some
measure of safety since losses are covered, but also because the insurance provides
progress reports on the status of the crop in the field.
143. Promising data set on crop loans at FIRA and BANRURAL may provide an
empirical test of the relationship between credit, insurance and other forms of collateral.
The Government of Mexico should consider further investigating this link. However, at
this point evidence concerning the empirical link is anecdotal.
57
Can insurance allow farmers to adopt riskier but more productive technologies?
144. Insurance alone is unlikely to lead directly to changes in technology. But two
integrated systems of credit and insurance, the Fondos and parafinanciera schemes result
in the application of improved technologies. The Fondos do so through mutual
arrangements and these were discussed in more detail earlier. A new innovation is the
role of the parafinancieras.
145. Under a typical parafinancing arrangement, an agribusiness company -- usually a
processing (e.g. feed or flour milling) company, poultry producer, or beer company -will work with farmer organizations to provide inputs in kind as well as crop insurance
and often price insurance -- via ASERCA or brokers. The firm will pledge to purchase
the final crop at market value. Farmers pledge their PROCAMPO payments as collateral
and are mutually obligated to provide either the crop or repayment of the value of in-kind
loans. In turn, the agribusiness company will take out a loan to pay for inputs advanced to
the farmers. FIRA provides guarantees to commercial banks to encourage their
participation for programs. The participation of agribusiness is motivated by a need to
acquire local supplies of agricultural goods that meet their standards.
146. Since 1999, the number of farmers participating in this type of parafinancing has
grown from 22,600 to 181,900 and the amount of credit has increased from roughly 1.1
million pesos to 2.5 million pesos.32 In 1999 FIRA provided credit through 320
parafinancieras. Most of this credit is linked to crop insurance. A main objective of the
parafinancieras is to spread risk, so crop insurance is seen as an instrument to transfer risk
to insurance companies.33
147. Overall, we find that crop insurance is almost always linked to credit, but not
necessarily the other way around. However, very recently, both the development bank
and rural financial institutions provide strong incentives for farmers to purchase crop
insurance when they borrow.
32
Parafinancieras financed 1.285 million ha. in 1999, about half of them were for sugar cane, 300,000 were
for malt barley and the rest for maize, cotton, tobacco and vegetables.
33
The club de maiz, or maize club, a type of parafinanciera for maize organized by MASECA (a major
maize flour producing company), requires the purchase of crop insurance. Crop insurance covers 100% of
the farmers’ cost of production.
58
7. DEVELOPING WEATHER-BASED INDEX INSURANCE IN MEXICO
Background
148. An important issue of the existing crop insurance program in Mexico is the
relatively high operation, administrative, cost. Weather-index insurance could present a
viable alternative for reducing such costs. Furthermore, such insurance could even
complement existing insurance products. This section of the study presents the results
from a preliminary investigation into assessing the feasibility for developing weatherbased index insurance in Mexico. A complete report can be found in the Annex volume.
149. Weather-based index insurance is a relatively new insurance (or risk management)
instrument whose payouts are based on the occurrence of a weather event, rather than on
actual crop losses. Thus, for instance, in the case of drought insurance contracts would be
written against severe rainfall shortfalls (say 30 percent or more below the norm)
measured at a regional weather station. The insurance would be sold in standard units (for
example US$10 or $100), and all buyers would pay the same premium and would receive
the same indemnity payment per unit of insurance if the pre-defined rain shortfall
happens.
150. The key advantage to this kind of insurance is that the weather or “trigger” event
(e.g. a rainfall shortage) can be independently verified, and therefore not subject to the
possibilities of manipulation which are present when insurance pay-outs are linked to
actual farm losses. And since the contracts and indemnity payments are the same for all
buyers per unit of insurance, the usual problems of moral hazard and adverse selection
associated with public crop insurance are lessened. Besides, the insurance would be easy
to administer, since there are no individual contracts to write no on-farm inspections and
no individual loss assessments. This can help make the insurance affordable to a broad
range of people, including agricultural traders, shopkeepers and landless workers whose
incomes are also affected by the insured events. This type of insurance would be also
easy to market. For example, it could be sold through banks, farm cooperatives, input
suppliers and micro-finance organizations, as well as being sold directly to farmers,
perhaps in the same way that lottery tickets are sold in village stores. Weather insurance
is not only for producers and rural people. Banks and rural finance institutions could
purchase such insurance to protect their portfolios against defaults caused by severe
weather events. Similarly, input suppliers could be the purchasers of such insurance.
Once financial institutions can offset the risk with this type of index insurance contracts,
they would be in a better position to expand credit to farmers, at perhaps improved terms.
The International Experience
151. The development of weather-based index insurance has its origin in two
international developments. First, the deregulation of the US power sector that started in
1996 created the need for power companies to hedge their revenue volatility. This results
because extreme low or high temperatures create peak load problems for the electricity
industry. When local companies cannot generate enough electricity, they must go on the
open market to meet additional demand at a higher cost. By using temperature-based
index contracts electricity companies can hedge against these added costs.
59
152. A second development is the proliferation of catastrophe or “cat” bonds—bonds
whose coupon and principle payments depend on the performance of an index or pool of
natural catastrophe risk. There are already successful examples of using cat bonds in
Japan and the United States to spread the risks of earthquake insurance, and expansion of
this approach offers a unique opportunity to link world financiers and poor people in a
partnership that is mutually beneficial. Although cat bonds and weather derivative
instruments in 1997 accounted for$2 billion in worldwide insurance premiums out of a
total of $125 billion, the reinsurance industry considers their potential substantial.
153. Although the application of weather-based index insurance and cat bonds is quite
advanced in the energy sector and for earthquake insurance respectively, its applications
to the agricultural sector are still very limited. For one, this type of insurance is very new.
Secondly, it has to compete with highly subsidized agricultural insurance schemes in
developed countries. In other words, high subsidies for traditional crop insurance crowd
out the development of weather-based index products. And third, farmers in several
developed countries, such as the US, Canada, and Australia, are large commercial
farmers demanding individual loss assessment.
154. There are yet only few applications of weather-based index insurance in the world.
There are two insurance plans in Canada in the province of Ontario and Saskatchewan
that use rainfall index and a third one in Alberta for corn that uses temperature—heat
units.34 Also, a private insurance company in Argentina is offering a rainfall insurance
contract to a milk producing cooperative (there is correlation between rainfall and milk
yields).35 Perhaps there are few more initiatives but the overall number of applications is
still relatively very small. At the same time, there are several applications of index
insurance in agriculture not based on rainfall (or temperature) but on average area yields.
Instead of rainfall, the index that triggers the insurance payments is based on the
assessment of the average yield, on a sample basis, over a predetermined area. Areabased yield insurance has similar benefits as weather-based index insurance as long as
there is a reliable assessment of area yields. Some of the countries that have developed
agricultural insurance products based on area yields are the US (the so-called Group Risk
Management program), Canada, Brazil, and Morocco, the latter still on a pilot basis.
Preliminary Feasibility Study
155. The Mexican government has expressed an interest in assessing the feasibility for
the development of weather-based index insurance in Mexican agriculture. For this, a
preliminary study was conducted and its methodology and main results are summarized
below.36
156. The study examines the development of rainfall contracts to insure against drought
during the critical crop growing seasons. The study focuses on four states: Durango,
Jalisco, Tamaulipas and Zacatecas. The feasibility study has two main parts. First it
examines the correlation between rainfall and yields in order to determine the loss due to
34
The program in Ontario started in 2000 and the program in Saskatchewan in 2001. Both programs cover
forage and pasture. The program in Alberta started in 2000.
35
Brazil, India, Morocco and Nicaragua are some other countries actively considering to develop weatherbased index insurance products.
36
Based on the “Report on the Preliminary Results of a Study for Mexico Rainfall Insurance” prepared for
the World Bank and the Government of Mexico by Prof. Jerry Skees in November 2000.
60
lack of rain. Second, it designs a prototype rainfall contract and examines how this
contract affects the variance of revenues from the crops.
157. To perform the correlation between rainfall and yields daily rainfall data are
collected for stations scattered across these states. However, these data are not complete
for each station and there is significant number of missing rainfall observation. The study
at this phase uses estimates of average rainfall observations in nearby stations to replace
missing data. However, a more sophisticated procedure that uses geographic correlations
to estimate a surface of rainfall is now being completed and will be used for the next
phase of the feasibility study. Regarding the yield data, the analysis used production data
for each municipality. These data used some subjective procedures to allocate both
production and plantings to each municipal for the period 1980-99.37 It is likely that the
estimates are of higher quality in the recent years. To mitigate some potential errors, the
decision was made to aggregate the yield data to the DDR level. Yields were developed
on a planted hectare basis and trend adjusted by municipality.
158.
Rainfall contracts were designed to reduce the relative risk of the 19 years of
trend adjusted yields per hectare for each DDR.38 Periods for rainfall were selected based
on the highest correlation between yields and cumulative rainfall using the time periods
for growth to determine most the appropriate periods as well. Durango and Zacatecas
beans use summer rain. For Durango is the cumulative rain between June and October
and for Zacatecas between May and October that critically affect the bean crop. In
Tamaulipas sorghum is planted in January and rainfall from December to May is most
strongly correlated with yields. In Jalisco maize uses summer rain between June and
October. More work is needed to investigate these periods.
159. The feasibility study focuses on proportional rainfall contracts that simply pay in
percentage terms for levels of rainfall below a well-specified critical threshold. With
rainfall contracts for drought, the payment is based solely on the rainfall event. If crops
suffer a serious problem due to freeze, hail or even excess rain, there may be no
payments. To make an assessment of how well the rainfall contracts will work, the
feasibility study simply assumes that an insured (e.g. a farmer) would purchase a value
that would equal the mean yield value within a given DDR. The study then develops an
estimate of gross yield for the DDR with no insurance and with rainfall insurance.
160. Results from this preliminary feasibility study show that for about 40% of the
planted area in these four states, rainfall contracts could reduce relative yield risk
between 9% and 29%. More specifically, in Durango the risk reduction in yields is 23%,
in Jalisco 28.6%, in Tamaulipas 9% and in Zacatecas 29%. These preliminary results
suggest that rainfall contracts have potential in Mexico. Detailed description of the
methodology and results are presented in the Annex volume.
161. An important challenge for continuing with the development and field-testing (pilot
scheme) of rainfall insurance is the quality of the rainfall data. Rainfall data are of mixed
quality and significant care must be taken to target any pilot testing and to shore up
rainfall data collection systems. In the end, rainfall insurance contracts will be preferably
37
See Juan Manuel Galarza, “Estimación de series históricas de rendimiento a nivel Municipal”, note
prepared in October 2000 and available upon request.
38
DDR is the Spanish acronym for Districto de Desarrollo Rural (Rural Development District). DDR is a
homogenous rural region consisting of several municipalities or parts of them.
61
designed where rainfall data are credible and with very few gaps. Related to this, there
needs to be an investigation of the quality of the equipment to measure rainfall data in the
weather stations where rainfall contracts can be designed. In particular, the potential for
tampering with rainfall observations needs to be assessed and if needed, to indicate
whether upgrading the equipment in these stations should be considered.
Potential Application of Index or Parametric Insurance in Mexico’s Rural Sector
162. As discussed earlier, rainfall insurance or more generally an index (parametric)
insurance could have several interesting applications in Mexico’s rural sector. Because
index insurance is not linked directly to crop loses at an individual farm it can appeal to a
broader set of rural clients that may not necessarily be farmers. Some ideas about
potential applications are briefly presented below.
163. Insurance by state governments against catastrophic weather events. The new
administration is been shifting part of the responsibility for catastrophic events to state
governments. These governments could use weather index insurance to re-insure
themselves against weather events that have a catastrophic impact on agricultural
production and rural incomes in general. For example, a state government could specify
the weather events it seeks to insure and the triggers for such events. If the event(s) occur
then the re-insurance would pay to the state government who will then distribute funds to
affected individuals based on a set of rules and institutional arrangements set for this
purpose.
164. Use parametric triggers for FONDEN. Similarly, the federal government could
use weather indexes to a) re-insure FONDEN against severe weather events; and/or b)
trigger payments by FONDEN on objective criteria to rural people. This issue is further
explored in another section of this study.
165. Re-insurance for the Fondos. The re-insurance for the fondos usually is triggered
when a severe weather event occurs: most often drought, frost, excess rain, etc. Funds
could seek re-insurance based on parametric triggers. This may provide a less expensive
alternative to the traditional re-insurance arrangements.
166. Development of Group Insurance Programs for small farmers or rural people.
Small farmers are expensive to insure due to the cost of individual farm inspection. By
not requiring individual inspection, a parametric insurance could provide effective risk
management to groups of farmers in a more or less homogeneous areas. Insurance
policies could be distributed like coupons whose payout depends on the occurrence of a
specific event or a set of events. As discussed earlier, this insurance may also appeal to
rural people whose income is negatively affected by the occurrence of certain weather
events.
167. AGROASEMEX could consider providing parametric re-insurance.
AGROASEMEX could use index contracts to cover major systemic risks. They would
provide this as re-insurance to the private insurance companies and these companies
would then have to design insurance products that they could underwrite properly. The
'basis risk' on these products would then be covered in the traditional reinsurance market.
This is perhaps a good way to reduce rent seeking by the private companies as
AGROASEMEX becomes a re-insurance company.
62
168. Rural financial institutions could purchase index insurance to insure their
portfolio. In this case, the banks become the direct clients of insurance to insure the part
of their portfolio that is likely to be in arrears if a severe weather event occurs. The banks
still will seek to collect the loans from the individual farmers, but they will have the
flexibility to reschedule loans and perhaps even consider forgiving some part of a loan.
Similarly, input suppliers could be the purchasers of such insurance. For example, input
suppliers that provide inputs on credit could state that if a certain weather event occurs,
they will be forgiving a certain part or all of the re-payment of their inputs by farmers.
169. Rainfall index insurance could even be considered for irrigated agriculture. In many
parts of Mexico the availability of water in reservoirs depends to a large extent on
rainfall. If there is no rain, there is less water in the reservoirs and farmers may not plant.
Index insurance could have a payout if there is no sufficient rain to fill the local reservoir.
This way the farmers will receive some compensation since they cannot plant (or reduce
their plantings).
Does Weather-based Index Insurance Require Subsidies?
170. There is really no compelling reason to indicate that weather-based index insurance
requires subsidies. Nevertheless, in a world where crop insurance is already heavily
subsidized (see US, Canada, the EU, etc.), it is likely that some subsidy might be
considered by governments. Earlier, in chapter 2, we discussed some key principles for
providing subsidies on crop insurance that are less distortionary. These same principles
could apply to weather insurance. .
171. For example, the government could consider as a subsidy the provision of a more
catastrophic insurance coverage (based on the frequency of weather events) with some
backing from the international re-insurance industry. More specifically, the government
could consider taking the risk out of the tail of the distribution where market ambiguity
would be likely to create high rates for insurance premiums. With the government
providing the more catastrophic insurance coverage, the private insurance companies and
the fondos could provide a less expensive insurance to farmers.
63
8. CONCLUSIONS AND RECOMMENDATIONS
Conclusions
172. The market for crop insurance in Mexico is a market distorted by subsidies and
fiscal support to the government insurance company. The government has a legitimate
objective to provide resources to protect smaller and poorer farmers from severe income
losses due to disruptions in their production caused by weather events or other calamities.
At the same time, the government has to protect its fiscal resources dedicated to protect
farmers, particularly from catastrophic events.
173. The government is in the process of redefining its role in crop insurance. Proposed
legislation would preclude AGROASEMEX, the state agricultural insurance company,
from directly providing insurance and would convert the company to a re-insurer. The
government also wishes to encourage the development of new private insurance schemes
and products that would deliver insurance at lower cost and thus require less government
support compared to the current system.
174. The crop insurance sector consists of AGROASEMEX, about 200 farmer mutual
insurance funds (fondos de aseguramiento or fondos) and four private insurance
companies. Crop insurance is often linked to credit and is a lending requirement for
loans from BANRURAL. Currently crop insurance is geographically concentrated in the
North and insures primarily irrigated areas. With the exception of one company, private
companies insure medium and large commercial farmers. AGROASEMEX, the fondos,
and one private company, cater to smaller farmers in addition to medium and larger
farmers.
175. Insurance is subsidized both in terms of premium (30%) and in terms of delivery
costs. The premium subsidy applies to all crop insurance policies, while the latter is the
budgetary support of the government to AGROASEMEX. Despite acceptable loss ratios
(i.e., the ratio of indemnities paid over premiums collected), AGROASEMEX has very
high administrative costs. Administrative costs are high for two main reasons. First,
AGROASEMEX caters relatively more to smaller farmers of basic crops and in rainfed
areas. Second, AGROASEMEX has a dual role of a commercial entity and a
development organization. This latter role is for the support of the fondos in terms of
supervision and technical assistance, and the development of new insurance products.
176. Larger and medium size commercial farmers, can afford and do currently purchase
crop insurance if they see they need it. Many of these farmers are already clients of
private insurance firms. These farmers benefit most from current premium subsidies.
177. Delivering affordable crop insurance to smaller farmers, particularly for nonirrigated areas/crops, outside the North of the country is a key challenge, particularly as
AGROASEMEX is terminating the provision of direct insurance. But before addressing
this issue, one needs to examine the insurance needs of small farmers first. This will
provide a better idea of what are the appropriate instruments and delivery mechanisms for
these farmers.
178. It can be argued that land size alone is a rather poor criterion to classify farmers in
Mexico. There are commercially-oriented small farmers located in productive areas, with
high production technology, often using irrigation, and have other tangible and intangible
64
assets such as available infrastructure, existence of organization, etc. On the other hand
there are small farmers, in marginal areas, growing basic crops and using low technology.
Further, these small farmers have a low asset base, including weak producer
organizations. We cal these farmers, resource poor farmers.
179. Traditional crop insurance may not be the most appropriate risk management
instrument for resource poor farmers. First, a large part of their income (over 50%)
comes from no-farm activities including off-farm labor. Second, most of the risks that
these farmers face are due to low demand for their labor services and because of illnesses.
Third, traditional crop insurance protects for lost assets and farm income, that small
resource poor farmers have little off. Given, that the traditional crop insurance may be of
limited help for these farmers, what are the appropriate insurance instruments for them?
180. The approach to provide insurance to resource poor farmers needs to be a holistic
one. The government of Mexico has several programs that have implications for risk
management for resource poor small farmers. For example, the PET (temporary
employment), PROGRESA (health, nutrition, education), Alianza para el Campo
(improve technology and productivity). Furthermore, resource poor farmers are impacted
by systemic risks, such as droughts and floods, that affect the economic livelihood of
their region. Thus, systemic, catastrophic insurance coverage is important to them, such
as FONDEN. At present, the criteria to trigger payments by FONDEN are rather ad hoc.
Making these criteria objective, that is based on the occurrence and the intensity of
certain weather events, will make the process more transparent and will increase the
certainty of receiving payments when a catastrophe occurs. In addition, if FONDEN
payments are triggered using objective criteria, the government could obtain re-insurance
coverage and thus protect its fiscal resources dedicated to catastrophic insurance.
181. For the commercially-oriented small farmers, insurance is more appropriate. These
farmers are able to pay for such insurance, particularly because its linkage to credit, and
the adoption of improved production technologies. The challenge is how to deliver
insurance to many small farmers without incurring prohibitive administrative costs.
Traditional crop insurance requires individual farm inspection that can be costly and is
subject to adverse selection and moral hazard problems. The high administrative costs of
AGROASEMEX are partly due to their client base of small producers.
182. A key issue of delivering insurance to small farmers is the existence of producer
organizations. The farmer mutual insurance funds or fondos, mentioned earlier, provide a
good example. Fondos provide insurance to their members by pulling together resources
(premiums) to pay for future indemnities. Because indemnities can be high, due to some
systemic risk, fondos obtain re-insurance, currently through AGROASEMEX. The latter
also supervises the fondos and provides technical assistance. The experience of the
fondos so far has been very good. They have relatively low loss ratios, making most of
them viable even without government subsidies. However, fondos have their limitations.
They tend to concentrate amongst small farmers in irrigated areas, and in regions with
pre-existing good farmer organizations. As such, fondos tend to concentrate in the North
of the country (Sonora, Tamaulipas, Sinaloa). Being mutuals, fondos tend to be very
selective regarding their membership. Currently there are around 70,000 farmers
belonging to fondos covering an area of around 600,000 ha. Overall, one can conclude
that while there is a scope for supporting the creation of new funds and expanding
existing ones, the fondo model is only a very partial solution for insuring small farmers.
65
183. Catastrophic insurance coverage, through FONDEN, is useful for small farmers,
particularly those in rainfed areas. However, the coverage is for catastrophic events
while commercially oriented small farmers may wish to obtain coverage for less
catastrophic events; that is the events occurring with higher frequency. Weather
insurance could provide a useful coverage for such farmers. Weather insurance pays out
based on the frequency and intensity of a specific weather event. It can cover
catastrophic events but also less catastrophic ones. Because weather insurance depends
on the occurrence and objective measure of intensity of a specific event, it does not
require individual farm inspection, that can be very costly for small farms. Weather
contracts can be sold in standard units and appeal to a broader customer basis than just
farmers. Agribusinesses, rural financial institutions, input suppliers, farm workers could
purchase such insurance as long as the specific weather event triggers economic losses
for them. However, weather insurance relies on the correlation between farm yields and
weather events and can vary from farm to farm, thus exposing farmers to what is called
basis risk; that is the probability of incurring losses without the weather trigger. This
would limit the appeal of weather contracts. This implies that one should select areas to
offer weather insurance contract that are more or less homogeneous in terms of farm and
climatic conditions.
Recommendations
184. In insurance, the main role of the government should be to provide effective
protection for small holders and the rural poor in general. The government has already
several programs that have broad implications for insurance and risk management for
small farmers and rural poor (such as. Alianza para el Campo, PET, PROGRESA, etc.).
These programs have been discussed in length in the main report.
185. With regard to specific insurance products, the main focus of the government
should be to provide the most cost effective form of government aid that will help small
farmers and the rural poor after a catastrophic event. This aid should be provided in an
objective fashion with clear rules for when and how much assistance to provide. Further,
the assistance should not unduly distort economic incentives. In particular, careful care
should be taken not to provide so much aid that new economic activity begins in areas
that are more risky or vulnerable to natural disaster risk. Doing so will likely result in
more losses and suffering when the next disaster strikes. Finally, structuring the aid in a
fashion that it facilitates risk sharing markets rather than crowding out such markets,
should be a goal.
186. FONDEN could provide more effective coverage for farmers and rural people. The
government should consider using parametric triggers—for example, linking pay-outs to
weather station observations. The advantages of doing so include similar protection for
farmers in the regions; pay-outs can be made promptly and at reduced transaction costs;
and lower uncertainty for the beneficiaries as to whether and how much money they will
be receiving after a disaster. Furthermore, the government could benefit as the budgetary
uncertainty, which has been evident at the state level, can be significantly reduced
through parametric weather re-insurance.
187. The government could also have a role in the development of new private insurance
products, particularly for commercially-oriented small holders. The major problems in
providing insurance to these farmers are the high administrative costs along with the
66
moral hazard and adverse selection problems. This can make traditional crop insurance
very expensive. Using AGROASEMEX as a re-insurer, the government could develop
weather-based coverage for systemic risks (floods, droughts, excess humidity, low
temperatures, etc.) offered to private insurance companies and the fondos. These
companies then can pass on this coverage to farmers (directly or indirectly through agribusinesses, rural financial institutions, etc.) or package it along with traditional insurance
(“wrap-around” insurance contracts) and sell it to farmers. By providing re-insurance
coverage through AGROASEMEX on systemic risks, the government could greatly
encourage the development of new insurance products that would benefit smaller but
commercially-oriented farmers.
188. As a next step, it is recommended that the government considers a pilot project to
test the concept of weather insurance in Mexico. The pilot would require the cooperation
between AGROASEMEX and private companies interested in retailing weather
insurance products to farmers (again directly or indirectly). The pilot will require: a)
evaluation of the infrastructure of weather stations and weather data in selected regions;
b) design of weather contracts; c) demand assessment through interviews with potential
clients; d) capacity building and educational programs; and e) design of a monitoring and
evaluation system to assess the experience and recommend its replication.
189. Another area where the government could assist commercially oriented small
farmers is in supporting the fondos. There is a scope to support existing fondos and help
form new ones. Up to now, AGROASEMEX had a conflict of interest in setting up new
fondos because it was creating competitions for its direct insurance. Now that
AGROASEMEX does provide direct insurance this problem does not exist anymore.
Thus, AGROASEMEX should continue its technical assistance for existing fondos and
actively support the creation of new ones.
190. Further, the government could strengthen the fondos by creating regional
associations of fondos. To integrate the fondos in a national structure is very difficult
because of the diversity of conditions amongst the fondos and their performance. Thus, it
can be proposed that first try to create regional associations of fondos in relatively
homogeneous regions to obtain enough of geographic risk spread. The creation of
regional associations would allow for a more effective supervision, allow fondos in a
certain region to spread their risks over a larger area (diversification) and to pull their
financial resources together creating a stronger overall organization. The regional
association of fondos could develop technical services to provide to the fondos and assist
in the development of new products. The association could also play a role in the fondo
supervision (delegated supervision).
191. With regard to government financial support for insurance, it is recommended that
if the objective is to alleviate rural poverty, government resources should target mainly
catastrophic insurance and other programs that have implications for risk management for
small holders and the rural poor. Spending government resources on premium subsidies
for commercial farmers is a questionable issue. Premium subsidies create distortions in
the crop insurance market. In absolute terms, premium subsidies favor larger farmers
with high technology and high value crops, particularly in irrigated areas. The current
system of providing subsidies as a percentage of the premium tends to proportionally
favor geographical areas and agricultural activities that are riskier. Thus a proposal to
increase the percentage of subsidized premium for less favorable areas will only make
67
matters worse, by favoring even more these riskier areas and activities. A less
distortionary system of providing subsidies should adhere to the following principles: 1)
to the extent possible farmers should be required to pay the pure premium risk of crop
insurance; 2) if subsidies beyond the pure premium are desired they should be introduced
in a risk adjusted fashion so as to not favor the higher risk farmers or regions; and 3)
subsidies should foster the greatest involvement from private markets. When farmers pay
something that is equivalent to a pure premium, the program will not favor the high-risk
farmers or regions. If more subsidy is desired, a flat rate off the pure premium should be
used for all. For example, 1 or 2 points off the premium rate would give everyone the
same subsidy. Thus, if the subsidy is 1 point, a farmer facing a pure premium of $5 per
$100 of liability would pay $4 and one facing a pure premium of $10 would pay $9.
68
REFERENCES
AGROASEMEX. 2000. Diez años de seguro agropecuario en México: Memoria 1990-2000.
AGROASEMEX. Santiago de Querétaro, México.
AGROASEMEX. 2000. Normas y procedimientos de operación que deberan observar los fondos
de aseguramiento que contraten el reaseguro ganadero con AGROASEMEX, S.A.
AGROASEMEX. Santiago de Querétaro.
Altamirano Cárdenas, J.R. 2001. “El seguro en la administración de riesgos de la actividad
agropecuaria”. Tesis doctoral para obtener el grado de Doctor en Problemas Agroindustriales.
Universidad Autónoma Chapingo/CIESTAAM. México
ASERCA. 2001. Subprograma de Cobertura de Precios de Productos Agricolas: Resultas 19962001. SAGAR: Mexico.
Casco, A. and A. Rosenzweig. 2000. La Politica Sectoral Agropecuaria En Mexico: Balance de
una Decada. IICA: Mexico.
Claridades Agropecuarias. 2000. Published by ASERCA, SAGAR: Mexico. Issue 85. September.
Cord, L. 2001. “Rural Safety Nets Lessons from Experience”. The World Bank: Washington,
D.C. Draft. February.
Cunningham, W. and W. Maloney. 1999. “Measuring Vulnerability: Who Suffered in the 1995
Mexican Crisis.” World Bank: Washington, D.C. Draft. March 15.
De Janvry, A. and E. Sadoulet. 2001. “Concepts for an Approach to Rural Development in
México and Central America: Regional Development and Economic Institutions.” Paper
prepared for the SAGARPA-FAO Foro Nacional “El Desarrollo Rural Integral: Una
Oportunidad para México”, México City, March 2, 2001.
FIRA. 2001a. “Oportunidades para el Desarrollo Agropecucuario Sustenabale.” Presentation
prepared for World Bank Mission, February. FIRA: México City.
FIRA. 2001b.”Estrategia de FIRA para Fomentar el Acceso al Credito en el Sector Rural.”
Presentation prepared for World Bank Mission, February. FIRA: México City.
FIRCO. 2000. Programma Integral de Agricultura Sostenible y Reconversion Productiva 2000,
En Su Molalidad de Reconversion Productiva para Regiones Con Sequia Recurrente. Mexico.
November.
Hogan, A.J. and Robert T. Aubey. 1987. “Crop Credit Insurance in the Agricultural Credit
System: The Mexican Case.” Socio-Economic Planning Science. 21(1):37-47.
Holzmann, R, and Jorgensen, S. (2000) “Social Risk Management: A New Conceptual
Framework for Social Protection, and Beyond”. Social Protection Discussion Paper No. 0006.
World Bank, Washington, D.C. see www.worldbank.org/sp
69
Olinto, P. 1999. “Background Paper 3: The Impact of the Ejido Reforms on Land Markets in
Mexico”. In Economic Adjustment and Institutional Reform: Mexico’s Ejido Sector
Responds. Volume II. World Bank: Washington, D.C. Report No. 18897-ME.
Presidencia de la República. 1999. Ley General de Instituciones y Sociedades Mutualistas de
Seguros. Editorial Porrúa. México.
Presidencia de la República. 1999. Ley sobre el contrato de seguro. Editorial Porrúa. México.
Mazarr, M.J. 1999. Mexico 2005: The Challenges of the New Millennium. Center for Strategic
and International Studies: Washington, D.C.
Sadoulet, E., A. de Janvry, and B. Davis. 1999. “Cash Transfer Programs with Income
Multipliers: PROCAMPO in Mexico.” Draft. October 20.
Secretaría de Hacienda y Crédito Público. 2000. Reglas generales para la constitución, operación
y funcionamiento de los fondos de aseguramiento agropecuario, de vida campesino y conexos
a la actividad agropecuaria. SHCP: México, D.F.
Siegel, P.B. and J. Alwang. 1999. An Asset Based Approach to Social Risk Management: A
Conceptual Approach. Social Protection Discussion Paper 9926. The World Bank:
Washington, D.C. see www.worldbank.org/sp
Skees, J. P. Hazell, and M. Miranda. 1999. “New Approaches to Crop Insurance in Developing
Countries”. EPTD Discussion Paper No. 55. International Food Policy Research Institute,
Washington, D.C. November.
Skoufias, E. 2001. “PROGRESA and its Impacts on the Welfare and Human Capital of Adults
and Children in Rural Mexico: A Synthesis of the Results of an Evaluation by the
International Food Policy Research Institute. IFPRI: Washington, D.C. Draft. February.
USDA. 1999. Managing Risk in Farming: Concepts, Research, and Analysis. Economic Research
Service. Agricultural Economic Report 774. Washington, D.C.
Villalpando-Benitez, M. and C. Gonzalez-Vega. 1998. “Financial Apex Organizations:
Experiences from Mexico.” Rural Finance Program. Economics and Sociology Occasional
Paper 2503. Ohio State University: Columbus.
Wiggins, K. Preibisch, and S. Proctor. 1999. “The Impact of Agricultural Policy Liberalization on
Rural Communities in Mexico.” Journal of International Development. 11:1029-1042.
World Bank. 1995. Mexico: Rural Financial Markets. The World Bank: Washington, D.C. Report
No. 14599-ME.
World Bank. 1996. Mexico Rural Poverty. The World Bank: Washington, D.C. Draft September
30.
World Bank. 1998. Mexico: Agricultural Productivity Improvement Project. Project Appraisal
Document. The World Bank: Washington, D.C. Report 18640-ME
World Bank. 1999a. Economic Adjustment and Institutional Reform: Mexico’s Ejido Sector
Responds. Volume I and II. World Bank: Washington, D.C. Report No. 18897-ME.
70
World Bank. 1999b. Government Programs and Poverty in Mexico. Volumes I and II. World
Bank: Washington, D.C. Report 19214-ME.
World Bank. 2000a. Mexico Food Marketing: Grains and Perishables. Volumes I and II. World
Bank: Washington, D.C. July 27.
World Bank. 2000b. Mexico Rural Development and Agriculture. Policy Note B5. World Bank:
Washington, D.C. July.
World Bank. 2000c. Mexico Rural Finance: Savings Mobilization Potential and Deposit
Instruments in Marginal Areas. World Bank: Washington, D.C. August 31.
World Bank. 2000d. Mexico Social Protection. Policy Note C4. World Bank: Washington, D.C.
December.
World Bank. 2000e. Mexico Water. Policy Note D1. World Bank: Washington, D.C. September.
71
APPENDIX I: FONDEN’S EXPECTED PAYOUTS FROM DAMAGES TO CROP
DUE TO DROUGHT.
192. The present analysis attempts to provide an estimate of the costs that FONDEN is
likely to sustain in the long run due to drought related weather adversity. Due to the lack
of data on livestock ownership by herd size, we limit the assessment to crop damages.
The estimates provide information about Fonden’s budgetary impact if it were decided to
restructure the disaster assistance towards a parametric type of insurance. A more
detailed study would require a breakdown of cultivated land covered by FONDEN by
farm size and these figures were not available at the time of the study. It is hoped that the
agricultural census presently underway will provide fresh and more suitable information
for this kind of exercise. In that sense, the estimates presented here could prove as a first
step towards an appropriate methodology for the budgetary expected costs represented by
a disaster assistance program such as FONDEN, possibly allowing for the possibility of
reinsuring its risk.
193. The first step in the estimation exercise was to determine the number of farmers
whose landholdings are below the thresholds contemplated in the rules of FONDEN and
the total agricultural area considered to be used for cultivation. In the case of private
farmers, the data comes from the 1991 Agricultural Census. In the case of the ‘social’
sector (ejidos, communities, etc..) data from the PROCEDE registration was used. This is
likely to bias slightly downwards the estimate of cultivated land, as nearly 80% of the
ejidos have been certified and measured as of 2001. Moreover, the majority of the ejidos
that still need to be certified are located in arid and semi-arid areas of the country and
possess great extensions of common lands that are ill suited for cultivation. Thus,
agricultural land registered by the PROCEDE is likely to be around 90% of the available
ejido agricultural land state-wise and the present computations can therefore be
considered as a lower bound on the maximum expected cost of providing assistance when
drought-related damages to crops occur. An added advantage of using PROCEDE data is
that it allows to determine exactly the number of farmers with less than 10 hectares of
agricultural land. As the 1991 census divides farmers into land-ownership or landusufruct brackets with limits set at 2, 5, 20, 50, 100, 200 hectares, it forces to estimate
very roughly the number of farmers with less than 10 but more than 5 hectares. While use
of the PROCEDE data allows to avoid this problem, paying the price of sacrificing a
portion of the ejido land as explained above, in the case of private farmers, we estimated
that one half of the private farmers in the class of farms with more than 5 but less than 20
hectares had between 5 and 10 hectares and that they owned one third of the total land
owned by farmers into this class. In the case of private farmers, computations were based
on land used for agricultural purposes. Moreover, the data allowed to separate farmers
that used only irrigated land from the rest. In the case of farmers that used both irrigated
and rainfed land, the available data allowed to distinguish between the total amount of
land with and without access to irrigation. Consequently, given the 5 hectares ceiling
imposed on the amount of land that could receive compensation from FONDEN for the
great majority of crops, it was assumed that the insurable land is equal to the minimum
between 5 hectares times the number of eligible farmers and the total amount of rainfed
cultivated land these farmers possessed. In the case of ejidatarios we did not have
information on the amount of land actually used for cultivation and with access to
irrigation by farm size categories. It was therefore assumed that all farmers cultivated the
72
same share of land under possession and used state-level aggregate data on the total
amount of land used for cultivation on individually owned plots. Finally, we assumed that
of the total amount of land under individually owned plots used for cultivation, the share
with access to irrigation is equal to the average share of land with access to irrigation at
the state level. The sum of land farmed by eligible private farmers and ejidatarios
represents the estimate of eligible farm land under the actual FONDEN rules of
operation.
194. We then multiplied the total amount of estimated eligible land by the per hectare
expected payouts previously computed on the basis of the historical rainfall data. This
amounts to assume that in case a drought will occur, the maximum amount of
compensation will be disbursed on the maximum amount of eligible land.
73