Guidelines for agrarian systems diagnosis


Material Information

Guidelines for agrarian systems diagnosis
Physical Description:
iii, 70 : ill., ; 28 cm.
Groppo, Paolo
Tanner, Christopher
Merlet, Michael
Food and Agriculture Organization of the United Nations -- Rural Development Division
Food and Agriculture Organization of the United Nations, Rural Development Division, Land Tenure Service
Place of Publication:
Rome, Italy
Publication Date:


Subjects / Keywords:
Rural development   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Statement of Responsibility:
prepared by Paolo Groppo, Christopher Tanner, and Michael Merlet.
General Note:
Cover title.
General Note:
"August, 1999."

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 692186477
System ID:

Full Text





Land Tenure Service
Rural Development Division
Sustainable Development Department
Food and Agriculture Organization

Rome, Italy

August 1999


This publication on Agrarian Systems Diagnosis' represents the result of an effort done by
the Land Tenure Service (SDAA) of the Rural Development Division of FAO in cooperation
with several institutions, and field projects.

It describes and summarizes SDAA's experiences in developing and applying a holistic
approach to land tenure dynamics in rural areas. It also presents both the conceptual elements
as well as practical methodological proposals for operationalizing agrarian systems diagnosis.
The examples presented are drawn from concrete experiences in which the Service has been
and still is presently involved.

This publication was prepared by Paolo Groppo, Christopher Tanner, and Michel Merlet,
who are, respectively, Land Tenure Systems Analysis Officer at the Land Tenure Service
(SDAA), Land Tenure Consultant at Cambridge SEPR Associates, 28 Houghton Road, St
Ives, Cambridge, England and Agrarian Reform Consultant at IRAM, 39 rue de la Glaciere,
75013 Paris, France.

We are grateful to many colleagues from FAO as well as to institutes and individuals from
outside FAO who have participated in the process of developing the philosophy that
underlines this handbook. Special thanks go to Prof Marcel Mazoyer and Prof. Marc
Dufumier from the Institut National Agronomique Paris-Grignon, France
[] who have been developing the main theory
on the evolution of, and difference between agrarian systems. Thanks to their work, future
decisions for technical cooperation activities involving land tenure in developing countries
will be better informed, more effective and more helpful to those making their living from
food production.

A short brief on ASD is presented in


This handbook is based on worldwide experience and uses knowledge obtained from both
failures and successes. Its main objectives are:

1. To demonstrate the interest of a system approach for the formulation of land
regularization policies.

2. To contribute towards the improvement of land tenure policies through

better understanding of rural dynamics which will allow us to anticipate any
secondary effects and plan appropriate additional actions or policy changes
more effective regulation of tenure and land management policies

3. To help national experts involved in land tenure issues to define their own
methodological guidelines.

This Handbook is not a comparative analysis of different systems methods, nor is it a
theoretical investigation on agrarian system approaches. Many rapid appraisal methods share
similar global objectives and principles, and different methodological frameworks can be
used. The Handbook does not intend to provide you with an overall view of these methods.

Instead, the Handbook is first and foremost an educational instrument for readers looking for
new, efficient and adapted methods and tools. It aims to obtain immediate results by offering
a tried and tested methodology for immediate field use. The Handbook offers practical tools
developed all over the world in FAO projects and used by other development agencies during
the last 15 years. It should also contribute however to improved investigation and
development skills amongst those carrying out field studies. This is even more important
because it is also a self-training process for those carrying out the project.

The target audience includes development technicians working in national Institutions in
charge of agrarian reform and land settlement, NGO field experts, and development
managers. It also aims to help technicians and consultants who have been working on
development issues, to carry out land tenure studies and propose policies to improve land

How to Use the Handbook

These guidelines are written as a handbook containing brief explanations on methods in
everyday language. Technical words and concepts are used only where absolutely necessary.
With boxes on specific topics, the handbook offers distinctive illustrations of those methods
and tools, in concrete situations where land studies have been based on Agrarian Systems

The handbook also serves as a reference work. A list of main publications related to the
different schools of system approach is presented. This list provides complementary sources
of information for each of the specific topic illustrated in the handbook.

Table of Contents


1. Land Tenure and Agricultural Production: A functional Definition of Agriculture
2. Historical and Geographical Differences in Land Access Systems
3. A Bottom-up Approach: Opting for Family Farming
4. A Systems Based Approach
5. Different Levels of Analysis and Inter-Relationships: Corresponding Concepts


1.1 Defining the FHS
1.2 Gender issues and social categories

2.1 Risks management at the farm level
2.2 The Invisibility of Women
2.3 The Family Reproduction Cycle
2.4 Relationships between individuals, farm-households, and the rural community


4.1 The theory of the ordinary farm
4.2 Designing an economic model of a household
4.3 Reproduction and accumulation thresholds
4.4 Modeling: a proposal

5.1 Zoning
5.2 Selecting the Sample
5.3 Carrying out an In-Depth Farm Household Survey
Different Phases of a Detailed Farm Household Survey
Phase A: Preparation
Phase B: Informing the Farmers
Phase C: Life history and inventory of household members
and resources
Phase D: Initial Consolidation of Data
Phase E: Second Farm Visit and Visits to Different Fields
Phase F: Second Data Consolidation, Production System
Coherence Assessment, and Preliminary Data Analysis
Phases G and H: Validation and consensus with the
community and fine-tuning of the proposals


Additional bibliography













Land reform programs are complex, and planned and carried out in very different social and
cultural environments. They aim to change the ownership and management of land, rationalize
farming, and establish a sound balance between agriculture and the rest of the economy. While a
focus on farm management is important, it is too limited for the requirements of sustainable
development. New approaches to land reform are instead increasingly multi-dimensional.

New approaches are needed, paying new
attention to old words such as diversity, [ The development that is needed must be
qualitatively different [...]. Development will
participation, and bottom-up approach. This bring food security only if it is people-
implies a different style of development, which centered, if it is environmentally sound, if it is
is endogenous and self-reliant, at the heart of participatory, and if it builds local and national
each society and having full meaning only if capacity for self-reliance. I...I At every tun,
rooted at local level in the praxis of each the lesson keeps hitting us m the face that
involving rural people actively in the defining,
community (Hamrell and Nordberg in designing and decision-making stages of
Development Dialogue). agricultural development is not optional but
essential. We see this missing requirement in
This new vision of development is gaining many projects that failed because they adopted
ground, even within the UN System (see box traditional top-down approaches, because they
were based on narrow technical specializations.
1). The President of the World Bank, James D. Sustainable agricultural development will not
Wolfensohn also recently spoke of "the new merely come from introducing better crops,
direction the WB is taking in its support of new cattle breeds, more credit or rural
participation, by recognizing that there is a cooperatives, as important as these may be.
diversity of stakeholders for every activity we Rather, it is achieved by farmers working in
very specific farm-household systems. It must
undertake, and that those people affected by be based on the tasks, needs and aspirations of
development interventions must be included in the farmers themselves and on the dynamics
the decision making-process". He added: "I and constraints they face, not only in their
personally believe in the relevance of farming but also in their domestic and non-
farm activities. It must take account of their
participatory approaches and partnerships whole rural life situation, including real-world
development and am committed to making factors beyond the control of the household -
them a way of doing business in the Bank" the ecology and natural resources of the zone,
(Participation Sourcebook http://www. the social-cultural environment in the community, and the policies, prices, services
and infrastructure that affect rural
prospects".(J. Speth. UNDP)
Unfortunately, recognizing the complexity of (
land reform is just the starting point.
Answering the myriad of questions that follow
on from this recognition is far more difficult. Systems analysis offers one way forward. FAO
has made considerable progress in this respect, with the Farming Systems Development
training kit (, Land-
Use Planning approach (
HTM#aglds and Community Forest Participation

The Sustainable Development Department has worked hard to integrate communication tools
into participatory rapid appraisal methodologies, including visualized analysis. With this
book, Agrarian Systems Diagnosis, FAO offers a further contribution from the Land
Tenure viewpoint to the realization of genuine Sustainable Development.


1. Land Tenure and Agricultural Production: A Functional Definition of Agriculture

Every traveler will notice that agricultural practices substantially vary from region to region,
from state to state. Everybody knows that agricultural practices also change with the times,
with the local, regional, international economic situation. These changes will not follow the
same patterns from one place to another.

Agriculture is a human productive activity, but it is quite different from others, such as
industry or services. It consists in the transformation of the environment in a determined
social context.

It is therefore highly unlikely that we will find the same agricultural activities within
distinctive ecological situations. Environmental variables have to be taken into account as a
key issue for understanding agricultural practices. Human societies have more or less deeply
transformed the former natural environment. Each transformed and cultivated environment
has its own production capacity, located within specific ecological limits. If some or all of
these are surpassed, sustainability will be lost, crises will occur and societies will have to
move to another place or to develop new practices for survival. The management of fertility
is therefore of paramount importance in agriculture. Fertility is primarily a product of the
ecological environment (soil, climate, natural vegetation, etc.), but effective and often
intensive human management of the environment is also necessary if fertility is to be
maintained or improved.

The agricultural use of natural resources is undertaken with human means (manpower, know-
how) using implements tools; and with cultivated plants, fertilisers, and domesticated
animals. Different kinds of social organizations, cultural patterns, economic relationships
between the different actors enable varied and distinctive ways of exploiting the ecosystems.

We can therefore easily understand how important the access to natural resources is for rural
people. In every society, land and other natural resources have value, and control over them
can fundamentally determine the wellbeing of specific populations or sub-groups within
populations. Access to land and other natural resources is an economic and social factor, not
an agricultural one. It is only once the rules of access have been determined that people can
begin cultivating, and sustaining their lives. There are probably as many versions of access
rules as there are human societies, but basically these come down to ownership (either as
individuals or some other collective group); or rental arrangements of one form or another.

How these rules are worked out and subsequently managed is perhaps as important for
sustainable agriculture as the production techniques used. Indeed changes in these rules can
render effective techniques inappropriate, and lead to the same type of crisis that might occur
if inappropriate techniques had been used and ecological limits surpassed. This simple fact
has been of great importance in agricultural development all over the world and through

It would therefore be impossible to carry out studies on agrarian issues without taking into
consideration the relationships between distinctive actors when it comes to control over land
access and use. By the same token, understanding these rules and how they operate and
how they might be changed or improved requires an equally deep understanding of the
development of agricultural societies. Here we refer not only to the technology used, the
crops planted or animals raised, but also the way in which power relations have evolved, the
way in which people gain access to land, and the systems that have evolved to manage land
access and use. Once again, we are talking fundamentally about social processes, which
underlie and enable the sustainable use of land and other resources for the greater (or
lesser) social good.

2. Historical and Geographical Differences in Land Access Systems

When population density is very low, and land is still abundant, access to labour rather than
capital is the basic bottleneck. Strong private rights over specific areas will be weak or even
non-existent. It would be mistake however to assume that no rights exist at all. Early
European settlers in North America or Southern Africa saw huge tracts apparently
'unoccupied' and therefore 'free for the taking'. Land rights held or exercised by larger
collective social entities could be identified upon closer scrutiny. In other words, it is
important to understand the history of land occupation and the manner in which local people
exploit their environment before deciding whether or not rights exist over an 'unoccupied' or
'empty' area.

These observations are important not only for understanding how land rights are established,
but also for determining how land policy should be developed. One key area of policy in the
modern world is how to manage access to land and natural resources by 'outsiders' who may
or may not view land rights and land management in the same way as local people. Invaders
or colonial settlers rarely carried out social and historical surveys before they occupied the
land of others. They had a quite different agenda, and either ignored existing rights, or
sincerely believed that land was free for taking. Such 'outsiders' naturally tended to use their
own experience as a major reference point: overcrowded, privatised farm sectors where most
land was already occupied and cultivated, or a farm sector from which they had been
expelled, thus negating any misgivings they might have had about doing the same to others.
These attitudes unfortunately still live on in the minds of many of those responsible for
agricultural development, and critically influence policies for land and other natural

Private rights to land do emerge gradually when population density increases. Yet the degree
of privatization is rarely a simple reflection of population density. And where stronger rights
over land and other natural resources do emerge, they do not necessarily have to be 'freehold'
or private property rights. Secure longer term rental agreements such as leasehold systems
can offer an adequate basis for economic investment and the allocation of scarce land
resources to those who really know how to use them.

'Abundant land' too is a relative concept. Not all land is equally fertile or close to water. Nor
is all land close to markets. And most importantly, land is rarely equally divided up and
shared out to all according to a simple arithmetic calculation. Private rights can emerge very
suddenly and create land scarcity for the majority where only a short time ago there was
plenty when an outsider invades or a new plantation project expels the local population.

Population density in relation to available land then increases dramatically, pushing
populations into crisis or forcing changes upon their own land access and use systems.

Many different situations have emerged over time. Where abundant land is exploited
extensively in a well-worked out balance with nature, numerous systems have evolved which
focus more on use than on ownership per se. In Africa today, many of these systems still
exist under the label of 'traditional land rights', and are often mistakenly seen as some form
of primitive communism where land and natural resources are collectively owned. Even here
however, gaining access to, and the right to use, specific resources, has thrown up many
complex rules and conventions between different socio-economic groups and sub-groups.
These can either be quite separate ethnic groups who use different parts of the same
landscape Balanta rice growers and Manjaco upland cereals and palm oil producers in
Guinea Bissau for example or they can be sub-groups of the same society with unequal or
controlled forms of access to the resources around them. Other ecological and historical
contexts give rise to longstanding agreements over shared use long distance cattle grazing
through land belonging to neighboring communities for example or the communal use of
specific resources such as lakes and forests.

In fact empirical evidence all over the world supports the view that there is hardly a
landscape anywhere that is not covered by access or use rights of some sort. As power
relations between distinct social groups have evolved, less visible social rights have been
transformed into more private or exclusive forms of land access, conditioning the relative
wealth and wellbeing of those on either side of the fence. Elites with new economic
ambitions may exert their authority over land once used communally the Enclosures in 17t
Century England for example transforming peasant or sharecropper agriculture into more
extensive, commercially focused enterprises. Landlord estates may subsequently continue to
have their entire area cultivated by other users tenants but under contracts with specific
commercial terms. Note that tenants may not necessarily suffer in the process, provided that
their contract with the landlord is legally enforceable and allows for some autonomy (for
example, present day tenant farmers in England).

In contrast, Latin American "haciendas" parcel out land to workers who cultivate family plots
for subsistence and provide paid or unpaid labor services to the owner. The relations
governing access to land are highly exploitative and allow few opportunities to experiment
with new crops or accumulate wealth. Large plantations also function on the basis of grossly
exploitative conditions, such as slavery or later forms of forced or indentured labor. In India,
particularly since the advent of the 'Green Revolution', many thousands of smallholders and
sharecroppers have lost individual land rights granted by private landlords. Weak legal
protection and a vastly inferior socio-economic position left them defenceless against
landlords wanting to consolidate tiny production units into larger and more capital intensive
commercial enterprises.

Even where strong private rights exist, it is possible to find several distinctive types of land
access. In general, once private appropriation of land has been established as the norm,
landless rural producers have still been able to get access to land through a range of contracts
with land owners (including the State). These include share contracts or sharecropping
(paying a certain percentage of production, or crop share, to the owner) and tenancy contracts
(paying a fixed amount in cash or in kind or in labor). There are also innumerable formal or
informal arrangements which involve some form of rent or the exchange of services and

goods. In addition, all contracts can cover either short or long periods, and can include
restrictions on what is done (planting trees for example is often prohibited in Africa, where
trees are seen as a sign of permanent occupation and defacto ownership).

Private holdings and other individualised land rights are not the end of the process however.
Owner-operated family farms have become common across Western Europe and still
predominate in France and most Mediterranean countries. Yet in England, the small family
farm and farm tenancies are becoming less common, as large corporations buy up land and
contract professional, salaried farm managers to run them. Private and communal forms of
land use also still exist side by side, with areas of common land used for grazing and public
leisure activities. Membership of service cooperatives for processing and marketing farm
goods offers small farms some of the economies of scale enjoyed by larger units, while the
remaining large landlord estates are farmed either directly by their owners, or by tenant
farmers. There are also large public or State land holdings for a variety of purposes, National
Parks, and other publicly-controlled land holding institutions which own and manage land
assets in the name of conservation or some other 'public good' (the non-State National Trust
is an excellent example).
National land reform programs, or the more
In many countries where the population has general agrarian reform programs, are not matters
apparently suffered through exploitative to be undertaken lightly in any country. Such
private land rights and grossly unequal reforms are evoked only [..] by a profound
access to land and other natural resources, poverty and by inequality of opportunity and
the State has intervened through agrarian power in rural societies. An emphasis upon land
m p c. w f o l a reform merely recognizes that man's relationship
reform policies. New forms of land access land use and occupncy is not only
to land use and occupancy is not only
have appeared: state farms, collective farms fndamenal for decent survival in an agrarian
(cooperatives), specific agrarian reform titles society, but that controls over his relationship to
for individual farmers. In many cases the use and occupancy of land provide lewrage
however, access to land through agrarian for the transformnaon of rural society and
economy. If society is to be transformed,
reform distribution has led to restricted administrators mus have ways to influence
property rights, even when individual land social practices at strategic places. Land policies
access has been allowed (e.g. the sale of land and controls are strategic in the early stages of
is prohibited for a long period after rights are economic growth for they can be used to give
attributed). direction to the transformation of less developed
countries, which are handicapped by gross
inequalities in wealth, power, and opportunities
The state can also impose restrictions on use (Parson in FAO, 1984).
(for example, in conservation areas), or it can
control or tax the inheritance process. Such
measures are often advocated by politicians who believe that they are in the public interest:
nationalising land and creating state farms will end exploitation and ensure that production
meets national needs; taxing or controlling land inheritance will create a more egalitarian
society; conservation will maintain biodiversity.

In this context land reform can be a double-edged sword, wielded to meet the ideological
objectives of essentially urban thinkers concerned more with social engineering than with
issues of rights and productivity. This was the case in Mozambique after Independence, when
colonial plantations and other holdings were nationalised and reformed into large state-farms.
Quite apart from the management and other technical problems subsequently encountered, it
is now increasingly clear that local people felt dispossessed, at precisely the time they
expected to have their pre-colonial land rights restored.

While different land access systems often coexist at the same time in the same area and are
sometimes even complementary, the same can be said for production systems. Together,
they meet a wide range of needs that one system alone could not satisfy. "Historically, the
emergence of new farming methods in response to the need of growing populations for more
food did not supplant existing systems of food production. Thus, the earliest hunter and
shepherd stage survives in the desert nomads and modem range farmers of today. The
succeeding planting stage, with tree-protected shifting cultivation, persists unchanged in
many tropical rainfed areas; it has also developed into horticulture and market gardening in
semiarid and temperate zones. The last stage, field crop farming, developed ultimately as
highly specialized, uniform crop farming on large areas with powered machinery and
chemical methods and has not displaced either of the earlier methods of production. Each
stage came into existence to meet new demands. At the same time the older ones continued to
develop. Agriculture is so diversified and so flexible by nature that abundant production can
be secured from a variety of systems" (FAO, 1969).

This brief discussion raises important questions. Firstly, land rights are deeply rooted and
evoke strong feelings. People who feel wronged can easily and, in their eyes justly, make a
strong case for restoring land rights that existed many decades, even centuries ago. These
rights may have been superseded or negated by some political or social process beyond their
control., but are still as real to the present day claimants as they were to the original
occupants. This question is at the root of many of the problems now being faced by African
governments seeking to modernise so-called 'traditional' land access and land use systems.
Mozambique is an excellent example, and one where serious efforts have been made in recent
years to recognize historically acquired rights while still allowing space for 'outsiders' to
come in and share land resources with local people.

The issue of 'empty land' is also still very much alive today. Where there are very low
population densities and no evidence of intensive exploitation, it is easy for outsiders (who
are not necessarily foreigners) to argue that they can occupy land freely without due regard to
pre-existing or already acquired rights. They either maintain that such rights 'obviously' do
not exist, or that there is so much free land that new areas can easily be found for the few
who must leave to make way for their new schemes or projects. This modem face of the old
European attitude is found across the world today, where powerful national elites seek to gain
access to land resources that are under-used or not actively exploited by local people.

The European, 'developed country' case is also of great interest. In truth development never
ends. The strength of corporate capital in England is transforming farming, and at the same
time changing the cultural context of rural England. Today it is almost impossible for
somebody without land either inherited or in a secure long-term tenancy to 'go into
farming'. Land prices are too high, competition from large corporate units is too intense, and
farmers are increasingly becoming outgrowers for supermarket chains that impose demanding
and difficult to achieve cost and quality criteria. On the other side of the coin, in parts of
Scotland descendants of crofterss' (a traditional form of sharecropper or tenant farmer)
expelled during the Enclosures are now reclaiming and winning back land rights lost by
their expelled forefathers.

What emerges from this picture is that land access systems are varied, and dynamic. They
change not only in line with climate and ecological change, but also in response to a wide
range of human inspired or imposed conditions. New technologies come on line, more
powerful groups appear on the scene with new ideas and resources, new governments
introduce measures to tax producers, control the factors of production, and manage prices in
favour or one group or another (the famous 'urban bias' scenario). And where there is
change, there will always be groups or individuals who protest that their rights have been
lost, or that they are suffering unduly through a maldistribution of resources, or that they can
do better with the available land than those who presently use it. Thus the need for good land
policy, policy which promotes equity and rights and also stimulates new investment and the
best use of available land for the greater social good.

3. A Bottom-up Approach: Opting for Family Farming

Why should we favor agriculture, and in
particular the family farm, as the focus Binswanger (1994) of the World Bank makes a clear
for development efforts? Binswanger case or focusing resources on small family rmns:
argues that small family farms have 'uring the early days of' "devetopret econoc"...
rural poverty was often explained by the
long been ignored for many reasons, but backwardne of traditional smallholder agriculture.
are in fact more productive than larger The sector was considered to ave almost on pottial
units and absorb far more labour. Not for dewlopment.I... In addition, international
only can they be the motor of commodity markets for agricultural gods wre
development in most countries where regarded as hostile, exposin countries, which reed
on tian for growth, to undue nrisk Agriculture could
they are the norm, but their development be taxed with little adverse consequence for economic
as economic units will also bring real growth or povrty reduction It is themoref not
human development benefits to the surprising that the solution to the reduction in rural
majority of people in the countries po t ws almost m unisall seen as being
d (e B associated with urban growth and rural-urban
migration. [...] These views have been thoroughly
discredited by research. Yet they also provided the
The greatest opportunity for the ideological justification for patterns of agricultural
promotion of agricultural progress policies and programs which hav been highly
therefore lies in the involvement and deimental to rual populations, especially the
leadership of producers themselves. poor.[... Both eomnust lcomies as wUl as many
market economies have paid an eormous price for
Given that family farmers are still the assuming without mch empirial evidee that
most widespread type of farmer in the large firms ae more efficient than mall on [...
world, and that they can be just as their economic costs of production usually exceed that
efficient as much larger units if of smaller en rises relying primarily on family
adequately supported and integrated or, in developing as wll as developed coxuries.
Their production is capital intensive and they generate
into markets, they form the "natural" y little employnt Because small farms haw less
focus of our analysis. In the following wealth and/or access to credit markets, they use an
pages we will examine the behavior of input mix which relies more on labor than capital, and
this complex group, and discuss ways thereby generates ma employment and self.
to analyze it. employment han their large counterparts"

4. A Systems Based Approach

All systems involve setting-up parts (components or sub-systems) that interact with each
other according to some process. Following Schilizzi (CIRAD) "A system, being organized,
is subject to an organizing factor. This is a set of rules governing the behavior of the system.
In human societies, internal rules are rules that a system applies to itself through some inner
process; external rules are those that the system appears to obey under theoretical analysis
[...] The set of internal rules is what governs the system. Governance (regulation) is the
activation of a set of rules and management of exogenous or endogenous perturbations
affecting the system. As these are typically irregular in pattern, the creation of a new internal
rule can be usefully viewed as dealing with uncertainty. The link between system governance
and uncertainty is memory: identification, storage and comparisons of patterns of
irregularities. Memory defines the scope or range of possible future events, which themselves
define the uncertainties affecting the system at present. The set of future possibilities defines
the system' s vision, which can deeply
affect its behavior. Vision is rooted in
past history". Stable development is the umangement and
conservation of the natural resource base, and the
The system approach aims to understand orientation of technological and institutional change
not only each component, but also in such a manner as to ensure the attainment and
interactions among components at continued satisfaction of human needs for present
_and future generations. Such sustainable
different levels. These interactions and future gen s. Such stainable
different levels. These teractions development (in the agriculture, forestry and
produce special and often identifiable fisheries sectrs) conserves land, water, plant and
characteristics through which each animal genetic resources, is environmentally non-
system can be classified and analyzed. degrading, technically appropriate, economically
Not all components are of the same viable and social acceptable" (FAO-SARD)
importance. When the degree of http/ht%,
complexity increases significantly, new
methods are needed to cope with
systems as a whole, and for taking into
account their internal and external dynamics.

5. Different Levels of Analysis and Inter-Relationships: Corresponding Concepts

Some basic definitions are needed before we proceed further. Classical concepts and notions
normally employed in agronomy, livestock production and agro-economics include:

Technical itinerary
Cropping or livestock pattern
Production system
Agrarian system

At a plot level the first concept we have to deal with is that of "technical itinerary". This
concept is found principally in French agronomic thinking and has almost no equivalent in
the Anglophone school of agronomy. It is normally defined as "the logical and ordinate
sequence of cropping operations, applied to a vegetable or animal species" (Prof. M.
Sebillotte). It encompasses the notion of "livestock management" which means the logical
and ordinate sequence of livestock operations applied to a domestic species.

If we move to a higher degree of complexity, we reach the stage of the cropping or livestock
pattern. One definition of cropping pattern proposed by Prof M. Sebillotte is "a surface of
land managed in a homogeneous way through different crops with their sequential order and
the technical itineraries which have been applied to them". In the same farm there are
normally many different cropping patterns, which together make up a cropping system (Prof
Mazoyer). The first level, or cropping pattern, in effect refers to a simplified system, although
this almost never exists in the small farmer units characterized by the production of various
crops, livestock and other forms of productive use of their surrounding resources.

The second level, or cropping system, does A Clusfifatdoa of Lvesteck Peductdio
allow for such diversity and therefore can be system
effectively used in the analysis of family farm
agriculture. We can make the same distinction Solely Llvetsek Systems: more than 90% of
between livestock pattern and system (see dry matter fed to animals comes from
Box). The rst term indicates animals rangelands, pastures, annual forages and
Box). The first term dictates animals of the feeds and less than 10% of the total
same species shared out among defined value of production comes from non-livestock
proportions by age and sex and managed with farming activities
the already identified technical itinerary. The Landless Livestock Produtin Systems: A
second term means "a subset of the [production] subset of the solely livestock systems in which
less than 10% of the dry matter fed to animals
systems, including cases in which livestock is farm produced and in which annual avenge
contribute more than 10 percent to total farm stocking rates are above ten livestock units
output in value terms or where intermediate (LU) per hectareof agricultital land.
contributions such as animal traction or manure Grslad Based Systems: A subset of solely
represent more than 10 percent of the total value livestock systems in which more than 10% of
the dry matter fed to animals is farm produced
of purchased inputs" (FAO, 1996). and in which annual average stocking rates are
less than ten LU per hectare of agricultural land
At this stage we are still working at the level of Mixed Farming Systems: more than 10% of
one or more plots and a possible combination the dry matter fed to animals comes from crop
with one or more animal species. The analysis by-products, stubble or more than 10% of the
total value ot production comes from non-
has so far taken taken little account of the total vah i of odctin vi mes f n
overall production unit. The combination of all Raed 5 d Fa bgi Systemem A subeet of
cropping and livestock systems, and the other mixed systems in which more than 90% of the
activities of a farm-household (such as value of non-livestock farm production comes
collecting, hunting, fishing, craft industry, and om rainfed land use.
Irrigated Mixed Famlnig Systems: A subset
off-farm incomes, etc.), set within its immediate of the mixed systems in ich mre than s10%
ecological, social and economic environment, of the value of non-livestock farm production
constitutes a higher order system, namely the comes fran irrigated land use. (AO- 1996)
production system.

Several production systems together and the interactions between them in turn make up an
agrarian system (see Figure 1). The agrarian system is the sum of relationships between the
production systems and the general social and economic organization of the whole society.

Figure 1

Prof. Mazoyer defines the agrarian system as: "a mode of exploiting the environment
historically created and sustainable, a system of production forces adapted to the
bioclimatic conditions of a given space and responsive to the social conditions and needs
of that moment".

The internal coherence of the mode of exploitation of the environment raises questions about
the overall technical, economic and social conditions of production. Mazoyer identifies the
following essential variables which combine in an agrarian system in one form or another:

- the cultivated ecosystem: original environment and its historical transformations
- the production elements: tools, machines and the biological material (cultivated plants,
domestic animals), and the social manpower (physical and intellectual) to manage them.
- the mode of transforming the environment resulting from i) and ii): reproduction and
exploitation of the cultivated ecosystem.
- the social division of labor between agriculture, craft industry and industry which allow
a) the reproduction of work tools, and b) the production of agricultural surplus and c)
the satisfaction of other social groups, beyond the needs of the farmers.
- the exchange relationships between these different but associated sectors of the economy,
the relations of ownership and strength which determine the share of the production
work, of the production and consumer goods.
- finally, the overall ideas and institutions, which allow the social reproduction: production
and exchange relationships and the sharing of production.

sub-system 2
production social
system a productive
system annnmic

The Agrarian System: a complex open system, made of two sub-systems
(based on Mazoyer, in Land Reform 1992-93. FAO)


Building upon the discussion above, we propose to use Agrarian Systems Diagnosis as our
basic methodology for analyzing family farm agriculture. This approach emphasizes
interactions among system components at different levels. It moves from the general to the
specific, using a holistic method, which respects a hierarchy of processes and determinants.

Agrarian Systems Diagnosis also gives a specific role to those systems which are "goal-
orientated", such as the production system managed by a farmer according to his or her own
purposes and taking into consideration his or her constraints and opportunities.


The sheer diversity and complexity of
land access and production systems Stnpifytng Diventry: The Search for Workable
often in the same national or regional SolhtieMm
context often demands that we
nt te and In Mozambique, FAO has been supporting the
identify patterns and simplify development of new land tenure policy and the drafting
analytical approaches. Finding a of new Land Legislation. With over 20 distinct ethnic
suitable unit of analysis to facilitate groups, each with its own land access and land
technical studies and give meaning to management systems, many people thought it would be
subsequent policy recommendations is impossible to integrate 'traditional' or customary land.
essential step along this path. practices into the new legislation. It would simply be too
an essential step along this pathdifficult and expensive to codify all the various systems
that exist, and the result would be too big and
Within such a unit of analysis, all those cumbersome to be useful.
present should share a common
understanding of social, technical and Instead, FAO has helped the Government devise a new
law which recognizes the relevance of customary land
economic norms. As we progress access as one of several channels through which Stale
upwards however, from cropping attributed land use rights are acquired. In other words,
system through production systems to the whole gamut of customary systems was reduced to
agrarian systems, it becomes more and one of several patterns of behaviour through which land
more difficult to adhere to this basic issues are addressed.
principle. We have already seen that This simple device, together with new concepts which
there are good reasons for focusing encapsulate local communities and allow customary
attention on family farms. Considering practices to continue within defined limits (both physical
these also as a kind of micro-system and constitutional), has facilitated the development of a
may also offer the best way forward, concise new Land Law which nonetheless gives real
.with te fm b g recognition to the validity of rights acquired through
with the family-farm-as-system being customary channels, and the customary land access and
the building block of the higher level use systems that manage these rights. The new law was
systems. approved in July 1997 and is now in force.

For better or worse, most technical literature is still strongly influenced by western modes of
thought which seek to identify discrete and easily studied units of production. 'The
household' is a classically western concept, approximating in the minds of many people to
the nuclear family of two adults and their children. Thus 'the household' has emerged across
the world as the universal unit of analysis. This unit lends itself to being analysed as a micro-
level system in its own right, summed up in the present context by the concept of the 'farm-
household'. This micro-system incorporates a range of production and social organisation

These variables include production information (land area, crop and livestock itineraries and
patterns, technology used), and indicators of labour availability, consumption, and overall
social wellbeing. In the human development context, it is often the last of these (for example,
young child nutrition within the unit) and not the more technical variables that are used as
outcome indicators of the overall success or otherwise of the micro-system in its wider
system context.

We shall see later that the farm-household concept is open to many interpretations reflecting
the many human situations and cultures around the world. This can often create confusions
between those who are 'thinking western' and want to see discrete units, and those who
appreciate the subtleties and nested structure of households and extended families in other
cultures (see the Box in the following section).

For the moment however, the concept of the farm-household as system is a useful starting
point. UNDP provides a helpful definition: "The farm-household system provides a good
framework for addressing human development concerns. It seeks to understand interactions
between the different components of the overall environment as they affect the decisions and
performance of each farm household. It also directly confronts the difficulty of reconciling
governmental objectives with individual priorities. Thus it acknowledges that the outcome of
governmental interventions in agriculture ultimately depends on the decisions of millions of
men and women farmers".

It is important to remember however that no family or farm-household exists in a vacuum, even
in western societies. Each unit is tied to others by a wide range of links. These begin with
straightforward kinship relations and marriages that tie family farms into wider extended family
groupings, and end in a web of commercial and trading relations that often mean that farmers in
widely separate regions are often equally affected by the same phenomena At local level this
means simply that farmers live and work in communities. At the wider level, it means that they
are embedded in social and political systems that are often beyond their control.

Moving from one level to the other from the individual 'family farmer' to the community
and on up to the society as a whole involves important shifts of analysis. Behavioral
patterns normally attributed to individual units or farmers as they deal with the larger forces
around them are sometimes attributed to groups of farmers or communities as if they were
individual units. When they do this, analysts often forget the heterogeneity of the
communities or even the farm-households they are dealing with. Perhaps the most well
known example are the still all-too-frequent assumptions that 'farmers' are always
'household heads' (ie the men of the household rather than women), and that income and
goods are equally shared by all household members.

Nevertheless in the quest for quantifiable data and discrete units of analysis, the FHS model
is useful. It must however be used correctly and due attention must be paid to ensuring that
the correct definition of 'farm-household' is used in any given cultural or socio-economic
setting. Bearing in mind these cautionary warnings, but aware of the systemic nature and
analytical utility of the farm-household unit, we propose to adopt the so-called Farm-
Household-System (FHS) as the main unit of analysis.

FAO has defined the FHS as "... three basic sub-systems (Habitat, Production, Consumption),
which are closely interlinked and interactive". The sub-systems are:

Habitat (those who live together): a decision-making unit; establishing general goals
for the system
Production (those who work together): other decision-making unit; establishing
technical goals for the systems
Consumption (those who eat together) other decision-making unit.

Thus the family farm becomes far more than just a series of cropping patterns or itineraries. It
incorporates units of consumption, and is the mechanism through which the family is linked
with others to form higher level production systems that deal with more complex forms of
production than the family alone can handle.

Unfortunately, entrenched thinking based of individ
on the western household model can stillis defined as a p of individuals
on the western household model can still o live on the same farm, work together on at least
result in inappropriate land and one parcel (for adults) and recognize the authority
development policies being developed. The of a single head of household in major decisions
wrong land holding unit can be identified relating to the farm enterprise. [...1 Fann production
for cadastral registration for example, or in the peanut basin of Senegal is organized at the
compound level. The compound consists of one or
the wrong people can be selected for credit more households. The nucleus of the compound is
and other technical assistance. In most typically one man who has the right of hatchet"and
African countries, nuclear families rarely his household (wives, children, older parents, aunts,
form the basis of the household unit and sisters, unmarried male relatives, etc.). Besides
FHS. Parents and children more often fit being the compound head, the male with right of
hatchet is also the head of his household within the
into wider extended family groups which compound. Other households in the compound are
sociological analysis can reveal as the most headed by married brothers, sons, or cousins of the
relevant FHS in a given cultural context compound head. These secondary households are
(see Box discussion on Senegal, for either dependent or independent, the difference
example). being that the latter prepare their own meals and are
responsible for meeting their own millet needs. A
holding or farm is the aggregate of all parcels held
1.1 Defining the FHS by all family members within the household. (Bruce
et Migot-Adholla)
How one defines and thus identifies a htp:l/www
household, and by extension the FHS, is
then the most obvious question.
Anthropologists have been debating this subject for years. From the layman's perspective it
often appears that they agree only that it is impossible to define precisely what a household is
- there are just too many cultural variations. The answer is that there is no simple definition.
What does exist however is a common approach which, if used correctly and completely, will
produce the most appropriate version of the FHS for a given cultural setting. This is the
systems approach advocated here, in which various aspects of the household are studied
together. Where there are points of contact, or where one result confirms another, we can
begin to define the most appropriate unit for our purposes.

This process is called triangulation, and is a widely used technique in many areas of social
and economic research. In this instance, we can start by considering each of the sub-systems
listed above: Habitat, Consumption, and Production.

The Household as a Habitat Unit

If people live in the same dwelling, it is natural enough to assume that they all form one
household. In most western economies this is a reasonable and correct assumption. Elsewhere
in the world however it is far less certain. In some African societies for example, where men
may have several wives, the husband lives in his own hut while each wife lives separately in
her own space with her own children. Afro-Caribbean societies are known for the high
incidence of female headed households, with several children from different fathers who may
or may not participate in the daily lives of their respective 'household'.

Using co-habitation as a criterion for defining a FHS in these contexts is decidedly risky.
Nevertheless, it is one of several tools that the investigator must have in his or her rucksack.
Views of what a house or dwelling is must also be opened out so that they may embrace
compounds with several distinct structures, or even several compounds which closer
investigation reveals are linked together by marriage and farm and eat together.

The Household as a Consumption Unit

The last observation leads on to another important criterion for identifying a household, and
by extension a FHS. The focus here is not the dwelling per se, but rather the cooking pot or
hearth (cooking place). If people eat from the same pot or share the same cooking place, then
they are a household.

A good example of this is found in Guinea Bissau, where the local term for a cooking place
(fogdo) defines a basic family unit which can approximate to a western nuclear family
(although this is not always the case and can vary between different ethnic groups). Moving
up from the fogdo we find the moranga. This expression refers to several cooking places
within one extended family unit. A moranga is often identified as a common unit of residence
however, sometimes clearly encircled with a high fence, and is used as 'the household' in
many studies. Even here however, where apparently clear local definitions exist, it is often
difficult to establish precisely where the FHS begins and ends. Even Guinean researchers
often disagree over the definitions, and argue at length over whether or not a fogdo or a
moranga is the best unit of analysis in 'household surveys' and similar field studies.

The Household as a Unit of Production

At this point it is useful to introduce two more variables into the discussion: labour and land.
In the Senegalese example above, Bruce and Migot-Adhollah emphasise the importance of
identifying units within which 'a group of individuals... work together on at least one
parcel... and recognize the authority of a single head of household in major decisions relating
to the farm enterprise'.

Working together on the same land holding is obviously an important indicator of whether or
not people belong to the same FHS. In many societies, units of production often include
people from neighboring dwellings, who share more difficult tasks such as land clearing and
ploughing. Yet working together on the same parcel is not always a good way of identifying
a FHS. This is particularly so in Africa, where youths who belong to traditional age-sets work
together on land especially set aside for them. In socialist settings, where cooperatives are

established, people also work together but cannot all be lumped together within the same
FHS. Some additional factor must be introduced to establish where the FHS begins and ends.
Thus, if neighboring dwellings can be shown to be related in some way using some of the
other criteria discussed here, then it is likely that we are getting near to a definition of the
FHS in a particular setting.

Land is the other factor of production that can bind people together in the same FHS.
Looking again at Guinea Bissau, thefogdo has been shown to be too low a level of analysis
when dealing with land issues. The moranga on the other hand does present a possible basic
unit of land occupation that is managed and farmed collectively between members of the
same extended family. Here, individual fogios and even individuals within them have their
own fields and plots, while their members also share labour with neighbouringfogdos. Add
the fact that they all accept the overall authority of a moranca head when it comes to overall
land management decisions, and it becomes clear that the moranca is the unit which most
closely resembles a FHS.

The management structure is in fact an essential first step in identifying the spatial extension of
the FHS. This structure in effect show us where the social borders of the FHS lie. Once these
are known, it is a relatively easy task to include all land areas used or claimed by all those who
accept the common authority of a particular individual or leader. It is also possible to look at this
from the other direction, and determine at what level certain groups of people can make basic
production decisions without reference to a higher management authority. At local level most
farmers even sharecroppers have at least a minimal degree of control over the decisions they
take. Thus the FHS can be equated to an area over which there is a certain autonomy over
production and consumption decisions (including how to dispose of surpluses). Establishing
how far this relative autonomy extends is another indicator of where the FHS lies.

To Sum Up

As far as family farm producers are concerned, the farm (or production unit) and the
household (family reproduction unit) cannot be considered apart. Focusing our attention on
the farm "per se" but not on the people managing it would prevent us from fully
understanding how the farm works, where its boundaries lie, and what rights it has over other
resources in both its immediate vicinity and further afield. Different versions of 'the family',
'households' and the FHS will have to be identified in different sociological and historical
contexts. And the resulting unit of analysis will not be the same in Africa, in Europe, in Asia.
What is important however is that those studying these diverse situations use the same
methods of analysis, and keep an open mind until such time as they are sure they have
identified the correct unit for the specific topic being addressed.

Where the FHS ends and higher-order systems begin is always going to be the focus of
debate amongst investigators and field workers (not to mention amongst local people
themselves). The answer is both vague yet indicative of the skills and sensitivity needed
when beginning to assess land issues and identify our basic unit of analysis. Each situation
requires careful examination, building up a complete picture of the social and other variables
that bind people together, underlie their decisions over production and consumption, and
determine the way in which they occupy and use the land and other natural resources around
them. Only when such a picture has been built up, and the objectives of a particular exercise

are clearly defined, can we concretely identify what the FHS is in a specific geographical and
cultural context.

By looking at each of the three sub-systems above in a given cultural context, we can identify
the most appropriate grouping of individuals who together make up the FHS. The foundation
stone of this process must be effective sociological analysis, producing a clear understanding
of the kinship and power relationships between individuals and the various forms of
residential and land occupation units encountered in a given setting. This approach will
establish where the social boundaries of the FHS are (Bruce and Adhollah's 'group of
individuals' defined in relation to habitat and a single authority figure). Once this is done,
further analysis of the production system from a more technical perspective can tell us which
land and other natural resources are used by the FHS. (This is done basically by asking all its
members what they use, when they use it, and where it is.)

In this we arrive at a location and culture specific definition of a particular FHS,
incorporating all the elements that make up the production system as defined above. These
elements include the land other resources over which the FHS has occupation, ownership,
extraction and other use rights, the individuals who make up the FHS, and the social and
economic relationships that bind them to each other and to the land they use.

1.2 Gender issues and social categories

As an FAO-ILO document has recently pointed out: "Households often change the
composition of their members through birth, death, marriage, migration, divorce or
abandonment. Instability in household structure, which tends to increase under pressures of
impoverishment and in periods of social and political turmoil, has created a relatively
frequent phenomenon of women-headed households. [...] Household members use natural
resources for production and consumption, which directly or indirectly has an impact on the
natural resource base. [...] Although roles between and within cultures can vary, women and
men are responsible for different although often complementary, productive activities,
including those that depend on natural resources. There are many types of households with
extensive variation from one place to another. Even one community, however, may include
extended family households, households that are polygamous, female-headed, male-headed
and that receive regular remissions from absent family members"(FAO-ILO).

Even when we consider the household as the most common interface between the individual
and the community, it is already clear from the discussion above that this is quite imperfect.
There is a need for the in-depth analysis of internal relationships between different members,
focusing on gender as well as differences between young and old. Yet our main concern is
not with the specific problems of each individual producer within the household or
community: our unit of analysis is the FHS. Individual rural producers do fit into social
categories within a community, the enhancement of which is the final purpose of our work.
These social categories are the main building blocks of the agrarian system. They are not just
mechanical components, as we would find in cybernetics. Some relationships between the
different parts are technically determined, but others are socially predetermined. Some of the
components are actors whose interests might be conflicting or contradictory.

Nevertheless an understanding of the behaviour of individual farmers, and the range of
possible gender and other characteristics of the FHSs within our agrarian system, will help
us to identify problems shared by groups of farmers within the FHS or production units.
Once these problems are identified, and are seen to be widespread and not the result of a very
particular situation in a give household, it is possible to devise improved agricultural and
economic policies for the group as a whole. Other mechanisms usually in the social domain
or built in as components of the new programme can then pick up individual problems and
deal with them on a case-by-case basis.

As indicated above, the 'farm household' and the FHS are best seen as frameworks within
which these social categories live and work together. Using the approach advocated here will
allow an appropriate unit of analysis, or FHS, to emerge in each specific cultural and
geographical situation. The systems approach is the glasses through which we commonly
view the landscape it is not necessary to change the glasses everytime a new landscape
comes into view.


Moving on from the household as the basic unit, we now consider some other characteristics
of familyfarms, or Farm Household Systems. These characteristics play an important role in
farmer decisions, as well as being of fundamental importance for developing tenure and land
use policies:

- Risk management
- The invisibility of women
- The family reproduction cycle
- The relationship between nuclear families and the rural community

2.1 Risks management at the farm level

Irregular yields are typical of agricultural production. Climate variations and recurring natural
disasters, plague and insect infestations, and unexpected price increases and decreases, are
amongst the many reasons why inter-annual yields fluctuate. The efficiency of family
farming systems very often comes from their ability to adopt non-capitalistic behavior in
order to cope with such difficult environmental conditions. Only one single year with bad
results would lead them to collapse. Thus family farmers often try to reduce risks as much
as possible. They prefer smaller yields and low-variability incomes over higher average
yields and higher incomes that also imply high risks.

Risk limitation strategies frequently involve product diversification contingent upon access to
a range of soil types in different agro-ecological areas. They also practice mixed cropping,
non-intensive methods that preserve soil fertility, and the limited use of commercial inputs.

2.2 The Invisibility of Women

As discussed above, the focus on the household as the unit of analysis often obscures the fact
that within this unit, social relations are anything but egalitarian or homogeneous. The role of
women in the FHS is often overlooked, both in terms of what they actually do to keep the
household going, and what they receive in return for their efforts.

There has long been an automatic assumption on the part of planners, technicians and even
field workers that 'the farmer' is the 'household head'. The household head is the person who
usually represents the household when dealing with the outside world. In other words it is
essentially a socio-political role rather than an economic one. As such, it is nearly always
occupied by a male household member, particularly in traditional societies where women are
not encouraged to talk to outsiders or engage in social relations beyond the immediate family

The result is that the male head of household is often assumed to be 'the farmer', and women
in effect become invisible to the outside world (see Box below). It has long been recognized
however that this perception is a fundamental error. In most developing countries, the bulk of
farm work is carried out by women, who also make a large number of production decisions.

While certain tasks are done by men -
for example clearing land or ploughing The Inisibility of Women
- many key agricultural tasks once the
land is prepared are mainly carried out "Invisibility occurs at .. levels, in the first instance
Sw ee. re inat the planning stage. Broad plans to improve
by women (see Figure 2 in the next agricultural performance do not usually include
page). consideration of the actors involved, but focus rather
on infrastructure, crop selection, technological
Apart from farm work, women are of innovations and administrative structures ... Women
course responsible for numerous other are also invisible at the later stage, when projects are
activities on the domestic front, ranging implemented. There is a tendency to implicitly
narrow the definition of target groups: the "rural
from the collection of firewood and poo" to on gets translated into "adult male
water, to preparing and cooking food household head", resulting in the failure to
and caring for children (see Figure 3 in recognize the economic role and contribution of
the next page). In short, they are the women. Conventional economics view the
backbone of the FHS household as one homogeneous unit, with a single
decision-maker per unit: therefore policies,
programs and projects address only the male head of
Moreover there is a growing body of household. But in fact, women and men producers
empirical evidence that shows that within the same household may have conflicting
women do not get a share of household interests or goals in regard to the cultivation of food
resources which is in line with the work crops. This conflict may be an obstacle to achieving
they carry out. Men usually eat first, and high levels of productivity in given food crops. Part
of the conflict is that in many cases women are
women will often give the best of what expected to work as unremunerated family labor on
remains to their children. Income from cash crops, while still being responsible for the
cash crops is often controlled by the provision of family meals" (Weeks-Vagliani)
man, even if it is the women of the
household who have done most of the
work to produce them.

UNICEF has long advocated greater attention being paid to this phenomenon, and the impact
it has on the key outcome indicators of child and maternal health. A crushing share of world
poverty is in fact borne by women, who suffer economic and social discrimination within the
FHS while being its most important productive asset.

0 20 40 60 80

Domsti work
Pirocessin ttlngrQps I


plantg, .W.^,ng^--- l

100 (percentages)

Women's work
SMen's work
SWomen in the
developing world are
almost entirely
responsible for
growing food for the
household. In many
countries they are
also responsible for
taking care of Larger
livestock, even
though the owners
are usually men.



One woman's day
in Sierra Leone

21:00 to 23:00
converse around fire while
shelling seeds and
making fishnets
20:00 to 21:00
clean dishes, clean children
18:00 to 20:00
process and prepare food,
cook dinner 22
17:00 to 18:00 s
fish in local pond
re 1600
15:00 to 17:00 4P
work in the gardens
14:00 to 15:00
wash clothes, carry water,
clean and smoke fish
12:00 to 14:00
process and prepare food,
cook lunch, wash dishes

4:00 to 5:30
fish in local pond

&Woo 6:00 to 8:00
light fire, heat washing
18:9 water, cook breakfast,
loi clean dishes,
11 sweep compound
8:00 to 11:00
work in rice fields
with four-year-old son
11:00 to 12:00 and baby on back
collect berries,
leaves and bark,
carry water

2.3 The Family Reproduction Cycle

Nuclear families do not always have the same configuration. The relationships between
family workers and family consumers change as children are born, grow older and enter the
FHS workforce, and eventually leave to make the own families.

Typically, household resources begin relatively small, increase over time, and eventually
decline again, as parcels of land and other resources are allocated to children and the next
generation of households is established (Figure 4). This process has a significant impact on
the occupation and use of land resources, both during the lifetime of a particular household,
and subsequently when the household head dies and land resources are redistributed (see

Distinctive inheritance patterns also influence the evolution of rural societies, and can impede
progress towards capital accumulation and investment.

It is therefore essential to fully consider the dynamic of household reproduction, both to
understand the behaviour of farmers in specific cultural contexts and the implications of the
FHS lifecycle for land occupation and use over time.

This dynamic is also important for sampling when doing field surveys. A majority of
households with limited resources may indicate an overall decline in resource availability.
This could however be merely the result of a larger number of households being at an earlier
stage in the developmental cycle. Where infant and maternal mortality rates are improving, it
is a hard fact that earlier checks on population growth no longer exist. Demographic growth
can thus threaten the balance between overall needs and the availability of land and other
natural resources that was maintained through the life-and-death, growth-and-disintegration
cycle of traditional households and land holdings.

2.4 Relationships between individuals, farm-households, and the rural community

Part of our work when we study individual production systems is to see how community level
and national institutions respond to incentives, strategies, and choices. Indigenous
communities usually have more experience of local conditions than non-indigenous people
when managing common resources such as forest and pasturelands, water, and wildlife.
National institutions or technical agencies often
have expertise and knowledge of the outside [--.. The main condition for the
world (for example, commodity prices) that is development of communities is not the
unavailable to those at local level. Traditionally, participation of the beneficiaries, although
the flow of information has been top-down, with this is important, but rather it is the
'experts' arriving to tell local people how to do creation of local ntiuti that can
ensure the continuation of that
things better. It is finally being accepted that in development. [...] Commuity institutions.
fact, local people often know more than outsiders are therefore the result of an awarness of
do about the management and use of local common problems, which are difficult to
resources. solve at the individual or family level, and
which require an agreement or consensus
among the different members of the
community. (Sanchez in Community
Development Journal)

(Factors limiting land occupation: labour and technology)


... .........

..... .. ...


". ... .. .. S.....

WOMEN (Smaller symbols are young children

o .. .............
Source: FAO 1999: Regional Training Courses in Community Land Delimitation. FAO/Land Commission, Maputo,
February-March 1999

What communities really need is well targeted support designed with their full participation,
and administered through community based institutions (see Box). Such support can make
existing activities more productive, and ensure that new ones are workable in the context of
the social, economic and ecological constraints identified by the local level end-users.

It is therefore essential to understand the different ways in which indigenous people or local
communities administer their common resources. The allocation and management of land and
other natural resources, as well as conflict-resolution mechanisms, are key elements of this
process. In this context it is necessary to look again at the basic unit of analysis the FHS -
and consider whether it is indeed the right unit for all purposes. Targetting resources at the
individual FHS level while the behind-the-scenes management role of a higher management
or judicial authority has been overlooked, may simply result in a failed project.

Guinea Bissau is again a good example to illustrate these points, and the importance of
understanding the relationships between the basic FHS unit (in this case the moranga) and
higher level orders of social and economic organisation (see Box below). FAO consultant
Paul De Wit has shown that by introducing land management systems into the analysis (as
opposed to land occupation and use), a higher level system is a more appropriate unit for
developing a workable approach to land policy. This unit is the village, or tabanca, and has
since become the main unit for delimiting community land rights within the new national
Land Law (approved in January 1998).

It is debatable however whether even the tabanca is the end of the story. Higher levels of
customary land management authority can also be identified in Guinea Bissau. These
correspond to old chieftancies (regulados) that have survived from pre-colonial times. De
Wit resolves this issue by analysing the relationships between the principal actors at local,
village, and regional level. In this way, it becomes clear that while the Regulo is recognized
as an overall authority on land matters and a repository of historical information, it is the
leader of the village chefe da tabanca who is the main player in land management

This analysis then provides us with the appropriate mechanism for identifying and
demarcating the land over which the tabanca as a whole has land rights. Within this area,
individual FHSs have clearly recognized and relatively strong (virtually private) rights over
specific areas of land, as well as shared rights over common resources such as forests for
hunting and gathering honey and other forest products.

Key local resources such as the fertile flood plains that are dotted over the landscape are
attribute to each FHS by the chefe da tabanca every year. Meanwhile, within each FHS, the
head of this unit exercises his management role over the land and other resources within land
held by the FHS for generations.

The distinction between land occupation and use, and land management, can also produce
very different pictures of the area over which a FHS or a community claims land rights. A
focus on occupation might produce a map with distinct, separate parcels of land used by a
number of FHS on a more or less regular basis. A focus on management on the other hand
could produce a much larger map which incorporates all the FHS parcels, as well as areas
in between them that are presently unused or unoccupied.

The land management system also
addresses the relations between F'aBechld and Conmanidae: Gui.s
individual land users. By analysing
these together with the spatial "'The tabanca farm system represents an integrated
occupation and use of natural strategy using land on three basic levels: bolanha
resources, we can arrive at an flood plains and river valleys; river banks and
assessment of community land rights drying out river bottom land, and upland raifd
So c areas. Access to forests, for hunting and other
as opposed to those exercised by products such as honey... is also an important part of
individual FHS units. the year-round subsistence strategy. For many
communities, fishing is a key activity, both for
It is also important to note that very consumption and for sale. Livestock...are important
few farm households rely solely upon for all ethnic groups requiring good dry-season
few farm house s ry grazing]. [...IThe production system is intimately
agriculture for their livelihoods. This is bound up with the social organisation of the tabanca.
particularly so for very small units, [...1 While individual fogdo [nuclear family]
where available resources simply members might make up separate production units in
cannot provide for family subsistence. their own right, labour and food exchanges are
Off-farm employment, remittances common at all levels. Most [ethnic] groups also
have some form of collective production, either on a
from relations abroad, trading and communal rice field or on fields cultivated by
other non-agricultural activities all members of age-sets or youth groups. [... The
contribute to household income. To labour force comprises the older children and adults
quote Conway and Barbier (cited in (especially women) of the immediate nuclear family
Carley): "With surprisingly few unit; plus workers it can rely on fom elsewhere in
the moranfa at critical periods. [... Each fogdo is
exceptions, developing country subject to the authority of the fogdo head, who
farmers, particularly in resource-poor allocates land of each type to his wives and
environments, do not rely exclusively unmarried sons, each of whom might have their own
on farming. Their aim is to secure a fields and a degree of control over at least part of
livelihood for themselves and their halis produced. Behind the fogdo head stands the
wider land allocation system and social organisation
families and to achieve this they of the tabanca and, m some ethnicj groups, the
usually pursue a range of productive noble clan and clan chief (Regulo). All land is
activities, only some of which involve allocated downwards by the village elder (Homen
crop or animal husbandry". Grande or chefe da tabanca), either through the
moranga or directly to eachfogdo head".
(Tanner 1991)
This observation means that a precise (
definition of the FHS is sometimes
very hard to pin down. With entrepreneurial producers, it is relatively easy to distinguish
between the economic production unit, the enterprise, and farm worker households. This is
not the case with the most common agricultural system, the family farm. In reality, the
production and consumption system of the farm-household or in other words its subsistence
strategy often extends far beyond the immediate geographical area of the farm or village.

In this way the 'system' can even reach into industrial suburbs in developed countries where
sons and brothers work to save money to send home, buy new land or invest in a truck or tractor,
or simply to get married. Does the FHS of these households extend from, say, a Manjaco village
in Guinea Bissau to incorporate a small flat in Marseilles? And what is the FHS of a household
where the women and older children work on a nearby cotton plantation to eam extra cash
income? Does it include the plantation itself, or is it confined to their farm plots and scrubland
where they gather berries and firewood?

These observations underlie the fact that our basic unit of analysis, the FHS, is set within a
complex structure of nested systems (see Figure 5), each one built upon or incorporating the
others within it. The tabanca system in Guinea Bissau is an excellent example of such a
'nested' system, starting at the level of individuals within a particular FHS (for example
women who get rights over certain plots attributed by their husband), moving up to the FHS
as a unit, and ending in the tabanca 'production system'. Beyond this, there are many
linkages to higher level systems the agrarian system, the society as a whole, and eventually
the international economy where terms of trade and other factors begin to exert their effect
right down to local level.

Figure 5: The Production System: An Open System

Few of the smaller systems close to the centre are sealed units without linkages of any kind
with the higher order systems around them. It is at this point that we introduce the distinction
between open and closed systems. Thus while we can identify a FHS using criteria such as
'working together on at least one parcel of land' or recognizingg the authority of a single head
of household', we must not forget that the FHS also has essential links to the outside world.
These links function through what is in effect an 'open border', and are often in integral part
of the subsistence strategy of the FHS. The FHS then becomes an 'open system', identified
as a distinct unit using the techniques above, but also with important two way exchanges
going on (including trade for example) between it and the higher level systems.

IL A' '1 i4 H L 0 .
Aii Airiti r i ] ii :-
A: l :ii .: .: ....i i .ii ...

iiij : i ji~i~;~: : ': ; : ': : ': : ^fB rU: ^ :^ ?::::^::::;iii~ ; i:: ::: i ^
I ....... ^ t^T: : .....p^ ^ 6:6^l

~ ~.~.ji~'j~~o~;~ ~PY~oH t'~'t ArA.'i IA Ai-
i/~~~~~~~~ I' :iiii : i ii


I .


The sustainability of production systems cannot be presented from a purely ecological or
technical point of view. Deforestation, soil erosion, decrease in topsoil organic matter content
in family farmers' production systems are often the result of much deeper causes which could
be described as a lack of economic sustainability. Deficient access to resources, to secured
land tenure, to markets, to production means, to labor, and to capital assets may force the
poorest of smallholders to adopt survival strategies and non-sustainable production systems.

Less-favored farmers, however, are not the only ones implementing non-sustainable systems.
Large profit making holdings may prefer to maximize their short-term benefits, and shift their
plantation sites (e.g. bananas companies in Central America) or their concessions (e.g logging
companies in many places) when natural resources are no longer available. Economic reasons
cause these different actors to adopt such negative types of behavior.

During the last decades, the world economy has undergone rapid change, generating
considerable transformations in agricultural production systems even in the remotest areas.
Nowadays, no pertinent analysis of agricultural systems can be done at local level without
taking into consideration socio-economic trends at local, national and even world level.

Depending upon their access to resources, and to the role they play in the production and
exchange relationships in which they are involved, farmers adopt very different attitudes in
order to maintain or to improve their livelihood as much as possible. The economic criteria
they try to optimize vary according to many different factors: land access and tenure security,
production means, family labor force, markets risks, off-farm income opportunities, etc.

However, farmers strive to use the scarcest resources as best they can. Farmer behavior may
generally be explained by materialistic concerns. The way they work and live is not a product
of low educational level, backward development, or only the expression of some specific
cultural feature: the production systems depend on agro-ecological and also socio-economic
potentialities and constraints, which have to be thoroughly identified. This is what we call
"family farming rationality". It is not always easy for external observers to understand those
different kinds of rationality, especially if they are technicians who are used to considering
themselves better educated than the farmers they are working with.

There are sometimes exceptions and individual farmers may not act in a "logical" way. But
we will never have historically significant social groups with irrational behavior.
Nevertheless, this behavior does not necessarily mean the best option per se: it only reflects
the best option available at a specific moment, and under specific ecological and
socioeconomic conditions. The so-called rationality has to be understood from the point
of view of the FHS, and does not always imply sustainability. In fact, as Parson mentioned
"It is not enough to postulate that people make rational choices among alternatives; for if
significant development is to be accomplished through choices, people must not only have
the appropriate abilities to perform but there must be something from which to choose.
Abilities are nourished by opportunities. The interrelations between abilities and
opportunities are fundamental for national development".

Access to land is of course one of the core issues
that will explain the behavior and the "strategy" of Pro ting Dvermt two different types
farmers. A detailed understanding of land constraints of diversity can be promoted [...]:
will be of paramount importance for rural diversity of production systems and
development. diversity of economic activities i...]
Diversified production systems in
agriculture are based on growing several
crops in association with several kinds of
livestocks; in forestry, they are based on
4. THE FRAMEWORK OF THE ECONOMIC multi-species forests; in fisheries, they
ANALYSIS are based on exploiting many different
species using different fishing methods
Before looking at the survey process in detail, it is mad gear. Diversification can also involve
important at this stage to have a clear idea of the integating elements of agriculture,
forestry and fisheries in a number of
fundamentals of the economic analysis that is the last different combinations. Diversification
step in the process. reduces the risks introduced by variations
in demand, changes in weather and
4.1 The theory of the ordinary farm climate, inter-seasonal variations in fish
stocks, and the occurrence of pests and
The concept of ordinary farm is the logical diseases. [... At the same time, yields can
consequence of applying classic economic theory to be inesed by grofrestr, multiple
cropping, the confinement of farm
the activity of valuation. The opinion that in animals and the production of some
valuation, priority attention should be given to inputs on the farm, recycling, multiple
ordinary incomes, was already being discussed in use of land and adding value to outputs
the XVIIIth century2, and was refined and by on-farm processing. [...1 The
consolidated by Ricardo and Marshall in the diversification o sorc of income is an
important part of any strategy for
following century. sustainable rural development.(FAO-
The basic idea underlying this theory is that if a
relatively homogeneous group of farmers, living and
producing in the same area, is distributed so that the great majority do not present qualities
very different from the average, these farmers will represent the ordinary producers who
manage ordinary farms. The "ordinary farm" is also understood to be an enterprise which
operates at zero level of profit, where the final value of production corresponds strictly to the
costs of production.

The first direct consequence of utilizing this concept is the way in which a survey
questionnaire is prepared. Instead of favoring hazardous data (mainly referring to techniques,
yields and prices) attention focuses on the analysis of the most common practices, or
"ordinary practices". This means, for example, that data presented for cropping yields do
not necessarily refer to the last agricultural cycle, but more probably to an ordinary cycle,
using ordinary prices expressed in constant terms.

4.2 Designing an economic model of a household

A comprehensive measure of farm-household income (Total Family Income) is a key
indicator which considers all household activities. As we are interested in agricultural
production systems, only that part concerning the farm is discussed below in details, while all
non-farm income is aggregated together under the expression 'off-farm net incomes'.

2See, for example: C. Trinci. "Tratato dlle time dei bani sabili" in Nuovo tratato di agricoltua, Venczia, 1778; Giri: "'agrimcre istrmito",
Venezia 1758; Pampani, "Breve metodo per timare i tenrri e fabbriche", Fenara, 1780 and M. Gioia, "Nuovo prospetto delle Scienze
Economiche", Milano, 1817


Total Family Income = Off-farm Net Incomes + Agricultural Net Income

Agricultural Net Income

The Agricultural Net Income (or Aggregated Value) is obtained by subtracting all costs from the
value of 'Farm Gross Production' (FGP), for the period under consideration (see Figure 6):

AV = Farm Gross Production (FGP) Costs of production


FGP = 2 [(production x surface)] x unit price


FGP represents all
production, for sale or
self-consumption. When
farm production or some
part of it is used as an
input for some other
production, it is
considered in that way.

Calculating the Gross
Farm Production
presents some
difficulties. For the sake
of simplicity, prices for
both marketed and self-
consumed production
are the same. Prices are
also referenced to the
same period of the year.
If there is a high
inflation rate, local
prices are given in a
stable currency (e.g. US





Cai- 't", l

Figure 6

The other problem is how to represent variations in livestock and plantation values3. The
variation in plantation values can be estimated by using the investment effectively made.

A practical way of calculating the livestock value is the equation (shown graphically below):

Livestock (year) production = {[(C-I) x unit price ]+ [M x unit price] + Fv} / n

n = "economic life" of a cow
n'= productive years (n'< n)
C = number of calves produced (whatever the sex) in n' years (total calf s less deaths)
M = quantity of milk produced and NOT used for calves (self-consumption + sold)
Fv = final (residual) value of the cow at the end of the "economic" life
I = one calf to be used for replacing the cow

None of these estimations will be perfect, but any rational estimation will be better than not
taking into account these hidden aspects of production.

Costs of Production

Among Costs of Production we can find proportional costs and fixed assets.

Proportional Costs (PC) are the sum of all costs incurred by producers for obtaining the final
product (gasoline, seeds, fertilizer, chemical inputs, etc.), directly proportional to the amount
of the production and which can be easily shared between different activities.

I3n some cases, it would be worthwhile also taking into consideration variations in stores and supplies.

Graphic Representation of Estimating Livestock Values


p *


Final value

One calf to
replace the cow

Milk sold

Fixed assets are the average annual consumption of capital, and are generally calculated on
the basis of a linear depreciation, in the following way:

(Fv Iv) / n


Fv = final or residual value
Iv = initial value
n = economic life expressed in years

The Aggregated Value can also be presented in the following way:

AV = [(FGP/Ha PC) x Ha] Fixed assets

This represents the equation of a line (y = mx + q) with x defined between 0 and the
maximum surface which can be managed by the unit of analysis in the present situation. In
order to study the relationships existing between capital, manpower, land and the productivity
of small farmer sector, we must start drawing a simple scheme of their relationships.

Such a scheme, following Mazoyer (1981) is shown in the diagram below (Figure 7), where:

Line 1 = Average gross product per equivalent worker
Line 2 = Average proportional costs
Line 3 = The proportional gross margin (1 2)
Line 4 = Average annual consumption of capital
Line 5 = Average productivity of agricultural labor
Line 6 = Maximum surface which can be managed within the existing technical system
Line 7 = Maximum labor productivity
Line 8 = Off-farm and non-agricultural income
Line 9 = Total Income
Line 10 = Present Reproduction Threshold

Figure 7

The Aggregated Value (line 7) corresponds to production factors that have been put into the
production cycle: labor (physical and intellectual), land and capital (operating capital).
Adding the different AV calculated for each activity will allow us to represent the ordinary
production system as seen in the figure 8 below (drawn from the Cooperation Project with
Incra, Brazil:

gado de corte
20.000 milho

15.000 feijso



20 40 60 80 100

ha /W

Figure 8

The next step is to show the aggregated values calculated using both the best as well as the
worst results (physical and economic) for each activity. This is illustrated in the following
figure (Figure 9) taken from a case study done in Itamaraju, Bahia State, Brazil.

12 000
Figure 9

Beans & Maize Fruit trees
8.000- J il

6 000 Maoc

Su4ar ,cn 0 0a" Reproduction Threshold No
4.000 ""'*Ba
Coconut "Go 'CO





1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0

rmal year
od year

4.3 Reproduction and accumulation thresholds

Under some economic thresholds, defined at household level, farmers will not have any
choice other than to adopt unsustainable systems.

The Reproduction Threshold (RT) of a Production System refers to the reproduction of
family manpower and the means of production. It covers:

a ) physiological needs (food) of the family members and all basic needs such as
clothing, housing, health services, etc. According to FAO estimates, the lowest level
of calories per capital needed to sustain life is about 2,000 kcal/person/day. This
approximates to the "cost of living" in a specific place at a given time. This clearly
dynamic concept can vary as a function
of the household age profile, especially
the number of growing children and An Example of Changes in Reproduction
pregnant women, and the type of labour d
(on farms, usually heavy manual) Th mor market-orentated the economic
mostly engaged in. The Total Family model, themore the Reproduction Threshold
Income (TFI) needed to meet needs is for that country will tend to be determined
equally dynamic, and depends upon by factors beyond the boundaries of its
the opportunities available to rural agricultural sector. A study carried out by a
FAO team in Chile looked at the evolution
people (both on-farm and off-farm or of the Simple Reproduction Threshold from
urban). Obviously farm-households the 1960s to the end of the 1980s. The study
which fail to cover basic needs are showed a major increase at the end of the
forced to join the rural exodus and/or 1970s and early 1980s. This increase was a
suffer infant deaths and other signs of direct result of a fully open market economic
model chosen by the Government of Chile,
systemic failure, in conjunction with major international
financial organizations.
b) the renewal of implements and
machines, farm buildings, animals, Explaining the reasons behind this policy
other capital, and also the fertility of and its subsuent impact is not ofinrest to
us here It is a good example however of the
the land. implications for rural producers. Effectively,
with an increase of almost 100% m the cost
In system analysis, the Reproduction Threshold is of living, improvements in productivity
the expression of external factors which the amongst small farmers during the same
hav y le h e of ig period (more or less at the same level) was
producers have very little hope of influencing ust sufficient to m ain the relative
The RT will be strongly affected by national position they previously had It was expected
economic policies, international terms of trade, that rural poverty would be reduced in the
and programmes like Structural Adjustment long term, yet nothing like that happened.
agreed with major lenders of capital. The impact Any positive effects of the increased
of these external factors can be quite rapid. Thus productivity were canceled out by te
parallel increase in the cost of
macro-economic changes that have no immediate living.(Grppo, 1991)
relationship with local farm production systems
can completely modify sustainability thresholds
and turn well-off producers into deprived ones. It is easier to understand such changes easier
if we use the Reproduction Threshold concept.

The RT is correlated with measuring poverty. The methods usually employed for measuring
poverty vary from one major group to another, but there are basically two approaches:

a) The Poverty Line (PL) approach.
b) The Unsatisfied Basic Needs approach

The Poverty Line Approach consists of the following steps:

define basic needs and its components; From a policy or program perspective, the
define a standard basket of what is alternative ways of documenting human welfare
essential for each household; c have a distinct contribution to make. The
measurements based on basic needs, such as the
calculate the cost of this basket, which incidence and severity of nutritional deficiency,
will represent the poverty, line; morbidity and mortality rates, water supply and
compare the poverty line with the sanitation facilities, housing conditions, and
household income education and health facilities, are most useful in
S classify each household falling below desigung programs or policies specifically
geared to those problems. They may also be
the poverty line as poor. useful in setting priorities, even though this may
entail subjective comparisons of the welfare
The Unsatisfied Basic Needs Approach is rather implications of, for example, a poor water supply
different. UNDP4 gives the basic steps as and a lack of education facilities. On the other
follows: hand, the minimum income (poverty line)
approach makes it possible to formulate policies
and programs that influence employment
define basic needs and its components generation, agricultural production, incomes, and
(as in the PL approach); prices. [...) As mentioned earlier, the income
select variables and indicators more approach attempts to determine people's ability to
appropriate for each component; meet some basic needs. Whether or not an
a individual actually does so can be influenced by
define a minimum level for each a host of factors [...] Some individuals who are
indicator, below which the household below the poverty line may have superior
will be considered as not meeting the managerial skills, make more efficient use of
basic needs, resources, and achieve the desired basic needs
classify households lacking one or more [.1 1 Nevertheless, the cutoff point is useful if it is
Sn based on "average" behavioral characteristics.
basic needs as poor. (Kumar in Mellor and Desai)

In both cases, all the members of a household
not reaching the required standards will be considered as poor.

In the Agrarian System Diagnosis whatever the method used, calculations should include not
only the income and other resources needed to meet basic needs, but also those needed to keep
thefarm-household above its Reproduction Threshold. The overall situation is also placed in a
development perspective to highlight recent trends within the ongoing evolution of the lives
and circumstances of the economic actors.

4"Desarrollo sin pobreza", Conferencia Regional sobre la pobreza en Amrrica Latina y el Caribe, UNDP, Quito, Ecuador, 1990

4.4 Modeling: a proposal

Modeling means the construction of physical, conceptual or mathematical simulations of the
real world. Models help to show relationships between processes (physical, economic or
social) and may be used to predict the effects of changes in land tenure systems.

As we said in the Introduction, the type oent would beneed
Another Development would be need-aiented, endogenous,
of development we are thinking of here self-reliant, ecologically sound and based on the
is quite different from that which has transformation of social structures. In delineating these
been advocated in the past. The need to characteristics, it [...1 emphasized that, though human needs
extrapolate from some models is used are both material and non-material, the basic needs of food,
only as a way to go a step further in the health, shelter and education should be satisfied on a
priority basis. It was further emphasized that development,
methodological discussion. In fact, being endogenous and self-reliant, should stem from the
almost all the available guidelines for heart of each society, and that it would acquire its full
carrying out such a diagnosis stop at meaning only if rooted at the local level and in the praxis of
this level. We know that within the each community. This, in turn, means that no development
households different strategies may model can be universal and that the richness of
ear or he different memes; at development consists in the plurality of its patterns
appear for the different members; at (Hamrell and Nordberg in Development Dialogue)
the same time we know that the poorer
the household is, the more important
will be other source of income, not
necessarily coming from the farm. Therefore it is quite difficult (maybe unrealistic) to model
such a situation. Our starting point must be that once we as technicians have decided that we
have the right to intervene in a specific
negative situation affecting a rural Recently I had the opportunity to contribute to a course [...]
population, our basic objective is to try tor a group of agriculture students in a West African
Satin, A university. Each student undertook a field project in a
and make things worse. As Richards village [.... The objectives of the project were to describe
has pointed out, in too many cases three "typical" farms, to provide, in scientific terms, an
farmers know their situation much assessment of the way each farm was run, and after detailed
better than the technicians who are discussion with the farmer, to propose technical solutions to
supposed to improve their welfare. his or her most pressing [... problems. The work was well
done, and the reports make fascinating reading. I think
many of the students were genuinely surprised to find out
Our starting point is that we want to how much farmers already knew about the ecological
work on the existing production system processes at work in their farms. [... Some lof the students]
(PS) and improve it where possible. were sufficiently impressed by this knowledge to ask
This approach is certainly cheaper than farmers' advice on problems they had come across in the
introducing a new (external) one. In course of experiments on the college farm. In the end,
tod a ne te lo. Ia however, it became clear that the students had learnt more
methodological terms, the logical from the farmers than the farmers had learnt from the
consequence of this is the need to check students. Few textbook solutions to agricultural
if the present PS, once improved within development problems seemed relevant or feasible given
the context of its own wider agrarian the realities of the farms described. The problems that
system will allow the FHS itself to enter farmers themselves listed as priorities were ones on which
the textbooks remained silent. The students came back from
what we call the Area of Possible the field not with a list of recommendations offered but an
Improvement (API). From this point on agenda of research issues upon which they would have to
we will work with an archetypal FHS start work from scratch. The project had demonstrated the
with average characteristics. width of the gap between what science has to offer and the
needs of typical West African small-scale farmers.

The methodological steps to follow are:

i) Identify the Basic Objective: Most FHSs are struggling in a market economy. We can assume
that the basic objective would be to reach a certain level of income and/or to reduce
uncertainty. Thus, if intervening in a given situation needs X years, and based upon an
historical analysis of the RT, we should make an assessment of what the RT will be in X years.
(see figure 10 below).
Reproduction Threshold

Historical ei solution


1980 1985 1990 1995 1999 ... 2000.

Figure 10

ii) Check the existence the Area of Possible Improvement (API): This means checking to see
if the present PS, allowed to reach its full limits, could ensure a level of income greater than the
predicted RT. If the answer is positive, we proceed with the analysis; if not we should think of
alternatives. We then ask FHS members what would be the maximum cultivable area given the
present technical system (the upper line presented in the previous scheme) without changing it
or introducing new tools or machinery. In the following figure (Figure 11) line PH represents
that physical limit. Using the Best Aggregated Value of values that have really occurred in the
past for crops in the same area, we then calculate the agro-economic limit (in the figure, line
SM represents the best results for a given cropping system). Obviously a certain level of
approximation is acceptable. The projection of line SM onto line PH will identify point A
which, together with the intersection of line PHwith the reproduction threshold (point P) and
the intersection of line AS with the reproduction threshold (point I), will show the API area.

$/W Area of Possible A

'Orecast reprouctlion ttueshold. P


Ha / Worker

S Surface available Maximum culthvatable surface
With present technical system

Figure 11

The identification of the API shows us how it might be possible to bring the existing PS to a
level above the foreseen threshold in X years. Each of the many possibilities within the area
represents a possible solution for obtaining that result. If that area had not been identified (point
A below the threshold) the overall PS should have been changed.

iii) Identify the Risk Management Component: The fact that the projection of present PS has
allowed us to identify the API does not mean that it should necessarily be the future PS of the
FHS. Another key aspect to consider is the risk management component, which is often part of
the overall strategy of a family farmer unit. This component is illustrated in the next diagram.
Suppose we are working with a simple PS with two crops A and B, whose lines a and b
represent their respective Aggregated Values. During the survey we have gathered data related
to good and bad past yields (and prices); we have also asked the farmers if bad or good yields
have occurred simultaneously for both crops.

In the Figure 12, a' represents the worst Aggregated Value for crop A and a" the best (a shows
the most probable (ordinary) from the FHS viewpoint). If the interview confirmed that
sometime in the past, poor production for B occurred at the same time as for A, then line b'
(showing the worst Aggregated Value for crop B) is drawn starting from point a'. Similarly in
the case of better yields, line b" will start at the end of a". The final points (called M and R)
then indicate the Surface ofMaximum Risk (SMR), starting from the origin S.

3 1

Figure 12

The two lines SR and the "ordinary" gross margin of the PS (line SX) are quite far from each
other. It is therefore quite probable that the strategy of the FHS will not necessarily aim at
maximizing Aggregated Value (reaching line SM), but will probably tend to reduce the SMR
by trying to improve the bad results which would be terribly difficult to overcome.



A further methodological step, certainly useful but not easy to do, would be to indicate the
frequency of the good, normal and bad yields. In the figure 13 below we have illustrated two
similar situations, equal in absolute terms (same good, normal and bad Aggregated Values), but
quite different in terms of frequency. The first PS has the following frequencies: 20% for good,
70% for normal and 10% for bad. The second has: 20% for good, 50% for normal and 30% for
bad. It is easy to see that the second FHS should have a more pronounced anti-risk strategy than
the first one.

The figures do not therefore give us any easy solution. They are better seen as tools for
visualizing other elements to be taken into account when trying to elaborate proposals. The
effective answer will only come from the interaction between the technician, who should have
a good knowledge of comparative agricultural systems, the FHS, the community to which it
belongs, and the Institution implementing the program. Consequently there is a need for a
validation and consensus meeting with all concerned parties before finalizing the proposals.

Figure 13


The method being presented here contains several distinct steps (Figure 14):

1. Zoning

2. Selecting the Sample

3. Carrying out an In-Depth Farm Household Survey

4. The Economic Analysis

The first two of these steps are largely preparatory exercises that are necessary before the
main exercise begins. Once the survey is completed, the final phase of data analysis can
begin, including feedback to the community to check the results and fully involve them in
drafting the main conclusions and recommendations.

5.1 Zoning (Step 1)

Zoning means the division of an area into smaller units, which have similar characteristics.
The objective of this activity is the identification and localization of agro-ecological and
socio-economic constraints and potentialities (ager, saltus and silva: individual productive
areas, the common, the reserve), which interfere with the dynamics of the different systems.

Zoning becomes necessary both for practical reasons (it does not make sense to study all the
FHS of the region) and for the identification of the Recommendation Domain. The RD is
generally conceived as a group of FHS sharing similar problems. We can interpret the RD
within each zone as the sub-unit for implementing possible solutions for overcoming any
problems identified.

The zone referred to here is often a One must constantly bear n mind that in predominantly
subdivision of agro-ecological zones agricultural communities, the form of tenure constitutes
produced by soil surveys and land use he social framework of production. The form of social
structure and productivity are, therefore, m direct and
planners. It is not a representation of a close relationship with the forms of land tenure and the
crop or livestock system seen as a purely concepts of land, particularly as it exists in customary
technical system. As indicated above, law. So, although it is not claimed that agricultural
the overlaying of social and economic productivity depends exclusively on tenure arrangements,
data on top of the technical aspects is the tenure system has a major mfluence on productivity. t
central to producing a complete picture would be unrealistic to deny any connection between the
central to producing a complete picture tenure system and agricultural productivity. In fact, it
of the FHS and higher level systems would be reasonable to postulate that there is a correlation
within which it functions. Sometimes the because agricultural production takes place within a
FHS or the RD are synonymous with framework of tenurial arrangements. The pattern of
agro-ecological zones but this is by no agricultural production also, in turn, could influence the
tenurial system thus establishing a cause-effect
means certain. Indeed while agro- relationship between the two (FAO Pacific)
ecological criteria are always important
for the determining RD or system-based
zones, the process of zoning itself must
focus on the problem facing us. As shown above, it is very likely that the identification of
social boundaries will be a more important first step than looking at agro-ecological aspects
(see Box). This is particularly the case where farm households use various types of land

Desk study (bibliography) Landscape analysis Oral history interviews

Report on the agrarian history : ecological
and socio-economic zomnng


tovisional Typology of present
production systems




-_- -t-- Macro conditions:
Regional, nanonal,
S Revision & International
of working

S 1 S Present

SuuLo-economic analysis of
production si stems
Quantitatle srud\

analysis & regional synthesis
balance of implemented politicies

-.4. I .

a Agrarian System Production System
Figure 14

within their overall production strategy, and depend upon long-distance grazing or access to
other seasonal resources to maintain their subsistence strategy through the year. The same
argument applies to policy and institutional criteria. In practice, depending upon the specific
situation, we will choose the more useful criteria, noting that for practical reasons, the total
number of zones cannot be so reduced as to avoid identification of differences and cannot be
so high as to be a non-sense for regional development program.

Zoning relies heavily upon secondary data (topographic maps, statistical rainfall data, earlier
surveys, etc), and the knowledge oflocal people, and is essentially an overlay of both kinds of
knowledge. In this kind of study, zoning is principally used to save time and to improve the
accuracy of further in-depth rural appraisal. Local empirical knowledge and secondary data
complement one another. This kind of zoning is a real participatory work because it is based
on a permanent dialogue between the local people and the technicians.

Secondary Data

Preparatory work by technicians is essential for them to be able to readily understand
proposals made by local people. This involves collecting together all available information on
the area in question. Types of information might include:

- topographic maps
- earlier surveys and data
- data from other sectoral work which may throw light on certain aspects (especially
key non-agricultural factors such as social organisation and consumption)
field workers with intimate knowledge of local conditions
key local informants

The most important point is to then select and prioritise only a few key variables among all
those available. Such choices will have to take into account historical data, and will depend
on the specific knowledge of the surveyor with regard to local production systems. Obviously
it would be absurd to ask local people to re-invent the wheel, for example by sketching a
topographic map if this information is already available. When a map is used in discussion
however, they may be able to interpret this existing information in a particular way, and
indicate on the map precisely where certain problems occur and areas where more detailed
investigation may be necessary.

Existing topographic maps are in fact one of the most important survey tools. These can be used
to sketch thematic simplified maps. In the case of a discussion of water-shed areas and water
management for example, they can be used to show just rivers, creeks, main irrigation canals,
and mountain ridges and valleys. Others might show just some contour lines so that our
understanding of erosion risk can be improved, or infrastructure to improve our understanding of
transport and markets. Another important sketch map may show just administrative boundaries,
a particularly important aspect in land access and management systems.

Topographic maps in fact nearly always exist. Even old ones can be very useful, particularly
if the discussion is about existing or acquired land rights. Being able to show on an old map
that a particular community has occupied a given area over a long period of time is
sometimes the only real proof needed to establish this point. New maps can of course be
made or other cartographic sources used, but this is a time-consuming and usually expensive
process. In this context it is essential to assess the probable utility of a given resources in

relation to its cost and other constraints. As Bruce (Bruce, 1989) says: "Satellite imagery is
available from several regional centers, but at current scales it is useful for orientation within
a large area such as a river valley or ecological zone is to be covered. The imagery and the
work with the imagery needed to make it useful in the field are expensive, especially if
alternatives exist. Aerial photography of the area, if it exists on an appropriate scale, is
usually preferable. This may be available through a government-mapping agency, a
geography department in a local university, or through a donor or contractor who has
previously planned or carried out project activities in the area Photography on a 1:20,000
scale will show roads and buildings and can be used to map the basic type of tenure niche.
[...] At 1:5,000 or lower one can map holdings in some farming systems quite comfortably.
At 1:1,000 one has excellent resolution; one millimeter on the map represents one meter on
the ground".

Other sources of secondary information may identify where local leaders live, where there are
storage units and markets, and where the boundaries between administrative units lie. It might
be helpful to come to a discussion with a community with important geographical reference
points already researched and marked down on a simplified diagram (mountains, rivers,
particularly large trees or forests, etc). It is likely that information will already exist on present
land use, crop and livestock s systems. If land tenure is the key focus, existing cadastral and
other records need to be checked before work begins. Data on infrastructure existing, planned,
and pre-existing can be significant (for example when local people refer in discussion to a
certain bridge which no longer exists or has been replaced, further upstream, by a new one, it is
useful to know that 'the bridge' in question may not be the one you can see today.)

Historical information about the area will obviously be gained from local level discussions, but
it is equally important to arrive there already with some idea of what has happened. In
Mozambique for example, some idea of the impact of the civil war on local population
movements and land occupation is important. When looking at spatial occupation, or discussing
new land borders, it is very useful to know beforehand whether or not the area in question was
the subject of villagisation programmes under the colonial and post-independence governments.
In Guinea Bissau, discussing the long term decline of flood plains in certain areas, knowing that
many fields were lost when dykes were bombed by the colonial power is crucial information to
arrive with. Having some knowledge of the local area already available to use in the discussion
also makes local people feel as if they are dealing with people who have taken some time to
understand their situation, and whom they can subsequently trust

Pre-pared maps and diagrams with all the information gathered can later be used to structure
and focus interviews with local people, or simply to facilitate discussion: a map can often be
a wonderful talking point which breaks the ice in many awkward situations. However, it is
essential not to arrive in an area with pre-ordained ideas about what it going on there. This
can be particularly important when discussing land borders delineating local communities for
example, where any kind of pre-judgement by the technician (based for example on existing
maps) can fundamentally distort the outcome of local discussions.

Local Level Discussion

Ensuring That All The Land Gets On the Map
Local level discussions are by
definition where the real work is In a recent field exercise in Southern Mozambique, the Swiss
done (De Wit 1998a, Tanner et NGO Heluas cared out interviews with local people to
l 1998). No matter how detennine where their community land boundaries are. The first
alo interview (with the women), rewvaled a complex pattern of plots
thorough the preparatory and fields for different crops and uses, residential areas, and other
process has been, no technician areas of social and common interest. This map appeared to reflect
should arrive in the field with all the land they felt was "theirs', and could be clearly delineated
pre-conceived ideas. This is and refernced on existing topographical maps. The wonmn also
different from arriving with indicated that fishing in the nearby riv was an important activety
hypotheses. Indeed, not having A later discussion with men from he same community revealed
some views on what is going on however that they used the surrounding forest extensiely, for
may in itself hinder discussion, hunting and exacting various forest product. Their perceptions
and make any structuring of the of this land area were completely different, not seeing it as 'land'
as such, but as a communal resource which they had always used
interview process very difficult and expected to go on using. The fact that it is an area of great
Yet the product of the overall mnt t to a neighboring conservation area and Reserve, now
process of ASD must reflect the undergoing a rehabilitation process under private management
local reality, and be seen by was source of anxiety. But in the inital discussion of what land is
local people as a product that being used, the focus as it always has been with technicians and
has direct relevance for their development workers was agricultural use and land that could be
has direct relevance for their clearly identfitled as beltging to one family or another.
lives. This will only happens if
they are completely involved, if While the forest was not subdivided in the same way, it was
their views are respected, and if clearly an essential resources for the overall subsistence strategy of
they are listened to. the commmity and a guarantee of food security in poor harvest
years. Including it in subsequent maps made a Vry large
difference to the area over which acquired rights are being
Of course it may well be the claimed, in the context of the new national Land Law.
case that local discussions will
confirm the ideas of the
technicians who arrive to conduct the local level investigation. The likelihood of this
happening is probably close related to the thoroughness of their secondary information
gathering, but it should not be overlooked that even good secondary data may reflect the
particular social or political parameters of the era in which they were collected. The interest
of the analyst in this context is to bring the data with him or her, and to have it confirmed or
to use it to provoke and open discussion ofpresent day issues.

In land matters, special attention will have to be paid to the way local people talk about
particular questions. Units of measurement are one key factor which is easily overlooked.
Again arriving well prepared is important: far better to have a table of hectare equivalents
already prepared, than to try to work out this out at the time (although again it is always wise
to check that they data you have is correct before relying on it completely is this really the
local term, is it really equal to 0.75 hectares, do local people have other terms, etc?).
Moreover, the exact quantifiable unit may not be important the concept of area and space is
what is probably more relevant, and good borders may be drawn up using social and
historical data that have no numerical value at all.

How local people classify things is often at odds with technical or official classifications.
Local land classifications are especially important. Land tenure discussions in fact must cover
a range of situations which do not necessarily always emerge when discussing 'land issues'.

These include the occupation and use of private or individually held land (with or without
title, but usually land that has been used for several generations by the same family), and
communal lands such as association and age-set plots grazing, forests, water resources, etc.
Local people often do not include certain areas in discussions because they simply do not
think that the interviewer is interested in them; or they themselves may not see these areas as
land over which they have rights that need to be protected (see Box).

Other issues to bear in mind are changes in land management; and the legal situation and legal
constraints. New titles may have been created that local people are unaware of, and old ones
may still be legal but not presently being exercised (for example, many pre-Independence
colonial titles in Mozambique are still in force, though the land is occupied by local people).
New leasing contracts may exist that are informal or even illegal in a situation where land laws
have long outlawed private transactions of any kind in land rights. Special reservations may
have been created by national government without adequate consultation over respecting
existing rights (to water, for example, or the continued and protected use of sacred sites within
Parks). Different state institutions may also have accorded use rights over the same area (a
common source of conflict in many countries), with these being interpreted as ownership rights
by their holders (hunting and mining are two examples while administrative boundaries are
always subject to change.

With such additional information, the basic map will be improved on if necessary, and a first
attempt at zoning can be carried out which sketches areas which share a set of development
problems and that can be treated as a group of FHS for policy or programming purposes. It
should be clearly emphasized throughout however that these maps are real communication
tools, to provide a floor for discussion of land tenure arrangements during the in-depth
survey (De Wit 1998a).

It is therefore also essential that the local community are fully involved in preparing more fully
worked out versions of these maps, and that they are subsequently in fill agreement with what is
contained in them. This is the core oftheparticpatory approach. (FAO 1999)

Once a more complete map is available, it can be used to identify areas that require more
detailed investigation. Obviously it is not possible or it may be too expensive to physically
inspect or survey the whole area designated on a map in collaboration with the community.
Local people can however indicate areas where they are uncertain of the exact boundary (for
example where there is no physical reference point or line such as a river, outcrops of stone
etc), or where they are having problems either with neighboring communities or with
'outsiders' who have come in to occupy land. Once these points are identified, two processes
are initiated: a) a review of the secondary data available with a focus on the precise area or
problem indicated; and physical inspection with the community to the site.

Fieldwork can include a range of techniques to check the map and resolve problems. These
may be transect techniques, a type of one-dimensional map of a line cut through a village
or area requiring more detailed investigation. It depicts a cross-section of an area along which
a number of issues are recorded. The purpose of a transect is to organize and refine spatial
information and to summarize local conditions in the area. The information is gathered from
direct observation while walking a straight line through the investigated area.

The other important technique, especially in cases of boundary conflicts or uncertainty, is
discussion with neighbours or land users who actually live in the area in question. Once
again, real information from real people is often of more use than more technical and
quantifiable data gathered by technicians who do not take the time to go out and talk to those
who occupy a given area. Confirmation of the land boundary, and of markers that show
where it is, is a key element of the land delimitation process, and can only done by getting
people together to argue out and resolve their differences. Only once this has been done can
the topographers be brought in to record the border markers etc using GPS or other
appropriate techniques.

5.2 Selecting the Sample (Step 2)

Two criteria determine the sample used for data collecting: the width of the survey and the
mechanisms for selecting individual case-fonts. Methods differ between two limits: i)
undertaking a large survey with more than one-hundred case-fonts for a micro-region,
selected according to statistical criteria; and ii) a limited survey, with case-fonts selected
according to a hypothesis of typologies
originating in the study of secondary
daKey founts when initiating data collection at field
level; the first step consists of interviewing key
The first method has the advantage of informants, selected among people having a deep
being more "systematic" and offering knowledge of the area where we are working. These
more quantifiable data. This approach is interviews are clearly less structured and more open
important if statistical relationships and than the others are, but they have to cover the same
fields because of their usefulness as "guidelines" for the
calculations are important for the preparation of the sample, which will be interviewed
investigation. Its disadvantage however later on.
is that it uses a great deal of human and
financial resources, and is repetitive. In Key informants can be identified both from the eldest
fact, practice has shown that such people living the area and within the institutions
Working locally. Where possible it is more suitable to
statistical surveys have rarely resulted in relate to ordinary people, preferably if they have been
viable development project or program living for a long time in the area. A typical interview
designs. This is because "the questions with a key informant will cover the following points: i)
asked reflect the expert's narrow historical development of the region; ii) basic food crops
perception of the problems facing and yields in the past, occurrence of famines in the past;
iii) type of techniques normally employed and
various localities. Moreover, the time introduction, when and why, of new procedures, crops,
needed for collection and analysis often etc.; iv) general problems, in the past and in the present
exceeds that available during this stage" as seen by them; v) identification of relevant "types" of
(FAO, 1984). households m the area.

Reliable information on the same subject can often be collected through alternative means at
a lower cost. These include reconnaissance surveys and rapid rural appraisals (RRAs) that are
much less structured and depend upon an open-ended process of questioning and observation.
Such surveys are often based on a preliminary diagnosis using secondary and other data, in
order to avoid possible underestimation of relevant land tenure and production systems. They
are carried out by experienced rural development specialists who concentrate on a) key
informants, and b) interviews with a range of different socio-economic categories within the
community or area in question.

Through a process of 'triangulation', when data from different interviews are cross-
referenced and checked, a complete picture of the real situation can be built up relatively
quickly. It is apparent that which of these two methods is chosen is very much a function of
the objectives of the survey. If statistical data are required, then clearly the first method
would be the logical choice. If the objective is to establish where the land borders of a
particular community might lie, and to understand the legal and other processes behind them,
then the second is not only cheaper but is likely to produce a more complete and correct

53 Carrying out an In-Depth Farm Household Survey (Step 3)

Agrarian System Diagnosis focuses on the farm-household system as a whole. It takes into
consideration simultaneously ecological, sociological and economic conditions and
constraints, which are beyond farm-household control. It is an approach which will enable us
to understand the different ways the farmers make use of resources and to find out about the
rational features of FHS' behavior. However, taking into account all production activities and
all family life-related aspects is not an easy task. Classical survey questionnaires usually are
not suitable for this type of undertaking. Too many questions would have to be asked; too
many different cases would have to be considered. Each class of farmer would require special

Methodological Key Points

Since the early 1990s, FAO has been supporting the development of new land policy and
legislation in Mozambique5. A new Land Law was approved in July 1997, and FAO is now
working with the Government to develop a new methodology for identifying and delimiting
land over which local communities claim use rights according to the provisions of the new
(1997) Land Law. This process has huge significance as it will determine the areas over
which communities have legally protected, customarily acquired land use rights. It will also
identify areas over which communities have a type of management jurisdiction, including the
right to participate in the official approval of requests for land by new, external investors.

This programme is an excellent example of the kind of participatory approach being
advocated in this document. It shows very well the main methodological points of the ASD,
which include:

a systemic approach to agrarian systems research in support of specific objectives
the use of participatory diagnosis as the main field methodology
focusing participatory techniques onto a given issue (in this case, establishing land
rights and land borders)
combining these techniques with other technical inputs when appropriate to produce
the final desired outcome (in this case, a legally approved and technically effective
method for identifying and demarcating land at reasonable cost)

5 FAO support to the Mozambique Government Land Commission began in 1993, with a TSS-1 project to establish guidelines for policy
development. In 1995/96, an FAO Technlical Cooperation Project implemented by the Land Tenure Service (SDAA) of the Sustainable
Development Department and the Development Law Service (LEGN) supported the development of the new Land Law approved in July
1997. FAO is now continuing its support to the consolidation and implementation of the new National Land Programme, with Netherlands
funded support to the Inter-ministerial Commission for the Revision of Land Legislation ('the Land Commission').

A System Approach

In system terms we are talking about community-based production systems enclosed by
'open borders' (i.e. open systems), through which linkages with the outside world and,
notably, external capital, are permitted and even encouraged under current Government
policy (see Box). These systems have already been shown to include many thousands of
hectares of land and other natural resources such as forests and grazing land, as well as sites
of cultural importance such as sacred forests and burial grounds. Much of this area is not
actually being used by the communities, and is best seen instead as being part of their
heritage or patrimony, handed down from previous generations and being safeguarded by
present generations for future use.

This approach is a radical departure from earlier land policy in Mozambique with focused on
land use as the main indicator of rights over a given area Thus any area a family was
actually cultivating or physically exploiting was protected by law, while the wider patrimony
was left unprotected. Such 'free land' has become increasingly sought after by private
investors, hence the need to reform the legislation and promote a more equitable form of land
access and land use.

The former land use-based model focused much more on the 'crop and livestock itinerary'
which is more closely related to the FHS level of analysis, while the newer approach
embraces the wider definition of a production or even agrarian system incorporating several
households within a single community. This model is the more appropriate in the African
context, where land rights are often attached not to individual FHSs, but to lineage or higher
level units of social organisation. It is obvious however that defining where the boundaries of
such a system are involves much more than just locating fields and pasture land and drawing
a line around them. As indicated above, the FHS and the wider production and agrarian
systems are a combination of the technical production-based picture of land occupation, and
the picture produced by grafting layers of social and economic relations over the top.

This understanding has enabled the FAO team in Mozambique to propose using systems
analysis to resolve the problem of how to identify and delimit community land. Together with
three key methodological tools participatory diagnosis, 'triangulation', and participatory
mapping this approach has produced an effective and viable method for approaching a very
complex problem, in culturally distinct and often very different contexts. (De Wit 1998a,
FAO 1999, Tanner et al 1998).

The FAO approach divides the production system referred to above into a series of sub-
systems, each one reflecting a specific aspect of local life. By comparing the results from an
analysis of each of these sub-systems, it has been possible to arrive at the most appropriate
definition of a 'local community' in any given setting. These sub-systems are:

the history of a given community, with emphasis on how it came to live where it is
and what historical evidence there is today of this longer-term occupation
social organisation (leading to the 'social borders' discussed above)
analysis of the production system as a dynamic and integrated concept
analysis of subsistence strategies and the resources needed to maintain them
analysis of land management structures within a given local context.

Each of these items is a kind of system in its own right. Together, data from each one can be
pieced together to make up a complete picture about land use, historical and present rights,
and how these are allocated and managed today. These data are secured using the
participatory diagnosis approach advocated by FAO, which is the subject of the next section.

Undert,.w zstLaqd hLaw. tis ie-p .didy. to.eeadry d.i bt a io.ul ca y i in sti tt ifa;n;tte
Law ,.is ,tty .-e ppUMlistEtb4xe. tllc ,.: -t -,. O 4,,,-m,
w^ al^:yWft bwalti^ Mt&M**u-an% 4W -

-if ait i Pam*a"ts 0, ao acW dd*a ( ay d MlAW&ta a1m. at e qji;. "*.. A .,O
U Is it a W co cpa s c <4w ( lSR A 'atwhi*w*n ,bwj ttw* i*Me Wa

paus3 lateaatit maucv po d u ttwri',,.pi-. .nam, .e

,- AU1 Ig-,V0*-MI f&W0. m pnw..oetlhfl*- iga x ts.

,. ,.:b a .e .*,.. .- .';;..' .-._
Threse quqtnf tor a4at *Vbujesm i ag fnkepi r i^ b a.#

for. Subsequent analysis of the production and management systems allows us to determine
witch i s t a e appropriate unit of organisation within th wid social boundaries t c e a
for. Sabe analysis ote p roductina mangeen sstesallw ssto deterin
(m tefl'4sK s e d

the 'local community' in a specific area. Production and other land use data can then be fitted
to this mainly socio-economic unit to show where its borders lie. These are then recorded on

a cadastral map, at a stroke conferring legal protection of all rights, whether being used or
not, within the delimited area.
a cadastral map, at a stroke conferring legal protection of all rights, whether being used or
not, within the delimited area.

Participatory Diagnosis

In Mozambique, the FAO team6 first helped to develop the systems-based theoretical
framework discussed above. Afield methodology was subsequently developed, within which
is embedded a central block of fieldwork, the participatory diagnosis. The outcome of the
process is a series of maps recording local community land borders and other rights (rights of
way, communal land or water resources shared by neighboring communities, etc), with the
final, technically cleared and fully geo-referenced map being recorded in official cadastral

Sequence ofActivities

The FAO team proposed a sequence of activities for carrying out a community mapping

Select the area to be investigated (using criteria such as incidence of land conflcits)
Visit the area before the survey to carry out preliminary sensibilization
- Participatory diagnosis to identify land borders and land rights
- Restitution (first): take results back to the community
- Geo-reference physical markers that do not appear on topograhical maps using GPS
- Restitution (second): confirm results with the community and neighboring
Record agreed boundaries on official cadastral maps

This process has three essential elements. Firstly, the task of sensibilization which might take
a considerable length of time but which is the foundation stone of a successful survey (see
below, Phase B). Secondly, the particpatory diagnosis itself (discussed in detail below). And
thirdly, the process of 'restitution', or going back to the community to discuss and confirm

Lastly, and most importantly, the community is given a copy of the results (in this case, the
map of their land which they produce, with support from but not directed by the technicians.
The ultimate outcome of the process is the initiation of new development activities, based
upon a stronger local awareness of their resources and the rights they have over them.

Participatory Diagnosis

The essence of this technique is that the community itself produces the final outcome picture.
The role of the external field team is to assist where needed, perhaps by focusing thoughts on
specific issues or by bringing in comparative information from other areas, and by helping the
community translate their results into a format appropriate for official or programme needs (in
the case of land rights in Mozambqiue, transferring community produced maps into officially
approved mapping formats which can then be registered formally in cadastral recordss.

6 FAO International consultants Paul De Wit, Sevy Madureira and Christopher Tanner, FAO National Project Coordinator Jafar Mussi and Land
Commission staff lead by ConceicAo Quadros.

The FAO and Land Commission team worked together to produce material on participatory
diagnosis which has since been disseminated through training programmes to field
technicians in Government departments and in the NGO sector. The particpatory approach
developed has the following main elements:

Use of particpatory techniques
Follow the system-based analytical model outlined above
Begin with an historical analysis leading into social organisation
Be on the alert for unexpected information and cross-check data from different
sources and from discussions on different topics (sub-systems)
Use semi-structured interview techniques
Use inter-disciplinary teams

Survey tools used include Venn Diagrams, Matrices of various sorts to cross-reference data, and
a check-list that essential topics are covered:

Meetings with local leaders and influential people
- The History of the Community
Local social organisation
Areas of jurisdiction/land management institutions
Spatial occupation (habitation)
Population dynamics
Land use processes
Perceptions of the future
Incidence of conflicts and conflict resolution mechanisms

Participatory Mapping

This is a central part of the diagnosis. Local community leaders and other groups are asked to
produce a map which corresponds to how each group sees the use and occupation of land and
other natural resources. A similar approach could be used for other issues too, with each group
giving views for example, on the causes of infant deaths in the community, or the main
problems besetting local farming.

The essential feature here is that these maps are produced by the communities, and not by the
technicians. The technical team subsequently help to transfer the data to official
topographical maps collected during the preparatory phase of the survey. This process of
cross-referencing or triangulation between different sources and sets of data is continuous
and two-directional, between the technical team and the local people being interviewed. Thus
in the case of the maps, the process identifies areas that are not already marked by some
physical feature (such as a river), and where more detailed investigation and use of GPS
equipment is needed. The approach therefore not only ensure local ownership and an accurate
representation of local reality; it also promotes the efficient use of scarce technical resources
by focusing them where needed as a consequence of the participatory process itself.

The Active Participation ofHousehold Members

It is clear by now that carrying out a farm-household survey is not simply a data-collecting
task. It requires the active participation of all the members of the household, and sometimes
that of some of the relatives. Basically, it is a participatory activity and requires complete
involvement by both the surveyor and the farmer. It is also meant to be a learning experience
for all concerned.

In order for the survey to be successful, it is of paramount importance for the outset:

To explain clearly to household members, both the aims and the different steps of the
To create the optimal conditions for communication and dialogue between the
farmers and the technician (adapted tools and skills, but also proper technician
To adopt a methodology which will enable the farmers and the technician to explain
to each other the rational behavior of the farm-household. Final submission of the
case studies at the end of the survey is not enough. The main findings must be
elaborated in full collaboration and by mutual consent with all concerned people.

Methodological considerations when looking at FHS and wider production systems

To be successful in carrying out these surveys, the technicians must take into consideration
the following points:

ecological, socioeconomic constraints and potentials of the environment of the
studied case (previous secondary data analysis is necessary so that the right
questions will be asked)
Sthe life history of the parents and recent trends of the farm-household
simultaneous system analysis of the farm and the household

The research will start with global issues. Details are only looked for as and when they are
needed ("holistic" method). The survey will consider the reconstruction of an "ordinary"
agricultural year and its economic assessment. This task will be carried out taking into
account the farm-household past history and future perspectives. At this stage, a vocabulary
in the local language covering the basic land tenure and production systems aspects should
have been developed. Some terms which may have a clear meaning in the language spoken
by the interviewers, may not have the same meaning for the farmers and vice-versa It is quite
difficult to clarify the use of technical terms in abstract discussion; if their meanings are tried
out in the field during the first surveys this will certainly simplify the overall task. The same
is true even for team members who speak the local language, but may not previously have
been involved with agro-economic terminology.

Direct observation of activities, types of behavior, relationships, farming methods in the field
are an important and complementary part of data gathering. Permanently making a parallel
with other farmers' cases, as well as permanently correlating secondary data will be needed so
that the carrying out of the survey will be constantly improved on.

Though flexibility may be necessary during the fieldwork, the survey requires the
implementation of a logical sequence, which will include several key events.

From a general point of view, several visits to the FHS would be required to realize a
comprehensive understanding of their viewpoints and strategies. In practice, in the concrete
situations where we are requested to intervene, the opportunity to return for subsequent
interviews rarely materializes. Therefore we should make the most of any visits made.

Effective and accurate fieldwork is also directly related to the knowledge, ability and
experience of the technicians and interviewers, who must be able to integrate and interpret
data and observations as and when they appear during the survey. The flow and volume of
data increases with the number of studies carried out, but after or two surveys it will be easier
to discern which information is essential and which is unnecessary. Carrying out two or three
case studies at the same time can in fact make this process easier and enable us to ask more
relevant questions.

The concern should be with quality and quantity. If a FHS seems uncommunicative, it is
preferable to move on to another FHS given the time constraints on the work schedule. The
main part of data gathering will be done in the field with the farmers. Another part of the
interview will be conducted with different FHS members separately where possible. Several
techniques may be used to gather data Information can be obtained through meetings with
individuals or groups of intended beneficiaries.

Group discussion can be used, for example, to identify constraints faced by producers or
variations in production techniques. Individual discussions can identify many details of
locality-specific production systems and concerns. Comparison of these beneficiary views
with those of government and other supporting agencies allows confirmation of important
design decisions such as the priorities assigned to various components.

Using a semi-structured interviewing method with a checklist and a list of pre-prepared
questions, partially closed and quantitative (for technical and economic aspects) and with a
strong emphasis on dynamic aspects, is the best and most rapid way obtain a good
understanding of the local agrarian situation. Obviously, it is possible that during the
implementation of field visits, new facts previously unknown will come to light through
analysis and interviewing, requiring additional collection of data This iterative process must
be considered as a central point and enough time and manpower should be reserved for this

The variables considered are referred to:

Composition and evolution of the household unit
Land tenure, land types and quality
Capital assets
Agricultural, forestry and livestock production
Technical-managerial profile
Off-farm agricultural activities and non-agricultural activities
Use of labour for different activities and in different seasons
Gender and youth relationships
Problems and perspectives.

Obviously, the environment where Agrarian System Diagnosis is carried out will help
determine the critical variables that must be taken into consideration. These variables include
the program components and institutional arrangements that are needed for an agrarian
reform/land settlement development program to be successful. These judgements must be
made in the context of the existing levels of human, economic and social infrastructure
development, and the social, cultural, economic, organizational, and political factors that will
affect program implementation.

In all cases the survey questionnaire must be tested on a limited number of FHS and
corrections made. A training workshop for the interviewers should also be held, including
practice sessions.

Different Phases of a Detailed Farm Household Survey

Phase A: Preparation

The Farm Household Survey consists of 8 distinct phases (see Table One below). Before the
fieldwork begins, the survey team will go through Steps One and Two above (zoning and
sample selection). These make up the Preparatory Phase of the survey process, designated as
Phase A in Table One. The team will review the secondary data and the results of the
previous steps (participatory appraisal conclusions, if any, maps, climate data, principal
constraints and potential,). They will then formulate hypotheses for each of the different
selected cases, so that s/he will be able to ask the farmer and others in the FHS the relevant
questions. Taking into account the main objectives of the study and the results of the
preparatory steps described above (zoning, local history, pre-classification of farmers), the
team will decide on the households to be selected and the number of cases to be surveyed in
each "homogeneous zone".

Table One: The Different Phases of A Detailed Farm Household Survey

Field Work

Office work

A Preparation: Review secondary data and
formulate hypothesis; select cases in light of
study objectives (whole team)
B Informing the Community ('sensibilization') Local meetings with whole community and its leaders, and with
individual case families, to explain the activity and secure
C First Farm Visit: Inventory of resources of farmer (both on farm
& off farm). Life history. Schedule of the next visits
D Initial consolidation and synthesis of data and
information gathered from Phase C
E Second Farm Visit: on the farm and in different fields with the
farmer to inquire about production, labour, and use of products
F Secondary data consolidation, production system
coherence assessment, and preliminary data
G Final meeting: Validation and consensus gathering with the
community. If necessary, recollection of complementary data,
modification of assessment tables. Reconstruction of economic
behavior. Discussion with the community on what can be done, on
what kind of projects or interventions could be suggested at the
farm level, at the municipality level, at global policy level.
H Fine-tuning of the proposals based upon the
assessment done by the Community as well as
the Institutional counterparts

Developing a 7)pology ofFarm Households

The economic analysis will be assisted if surveyed households can be ordered according to a
pre-survey typology. This can be done in the Preparation Phase, but only if sufficient
information is available in the review of secondary data to justify establishing a pre-survey
typology. If sufficient information is not available, putting households into presumed
typologies could distort or prejudge results and produce a picture of the real situation that
reflects the biases of the researchers and not the real situation on the ground. In this case,
putting households into a typology must wait until Phase D at the earliest (initial
consolidation of data).

Why is a typology necessary though, if the process of zoning (Step 1) has been effective? The
answer responds to a common concern often raised by observers of this kind of research, who
point out the big differences that exist between production/consumption units even within a
limited geographical area. There is therefore a need to highlight and classify this diversity in
order to produce specific programmes for each situation and type of problem

Devising an appropriate typology begins with a declared operational interest: trying to
simplify the heterogeneity of the surveyed FHSs through the identification of groups (types)
presenting similar potential and experiencing similar problems.

The objective is to identify the different capacities, rhythms and levels of possible
accumulation of the various production/consumption units. A number of methods exist for
devising the typology. The proposal presented here foresees three steps:

1. According to the FHS annual income (per capital, per worker of for the whole unit),
broadly separate the whole units into, say, three main groups: (a) units which are far
above the Reproduction Threshold; (b) the units averaging the RT; and (c) the units below
the RT.

2. Within each group, proceed to identify different production systems. Apart from main
livestock and cropping systems, Production Systems will be classified by fixed assets and
operating capital (for example, in a "dairy system", the farms featuring manual and
automated milking are placed in different categories).

3. Combine each group of units already identified with the historical origins of the present
types. This means for example taking into consideration the way they obtained their land,
capital and manpower in the last decades, or their sociological background.

Combining 1,2, and 3 will then allow us to devise a reliable typology of units which
represents the basic framework of the subsequent analysis.

Phase B: Informing the Farmers

The previous agreement of farmers is imperative before launching the case survey. The
survey team must gain their full confidence in the exercise before starting the fieldwork. This
is a process that can take many weeks in some cases, depending upon a range of factors.
These include the level of contact the community has with the outside world, the reasons why

the survey is being undertaken, and the previous experience of local people with technical
agents and other visitors to their area. If the survey is a response to local concerns for
example about land grabbing by outside investors then receptivity to the presence of the
team and a willingness to participate will be more or less assured. If the survey is in support
of some externally driven project, this acceptance is far less certain and time will have to
spent convincing the 'beneficiaries' of the good intentions of the survey team.

Before the fieldwork the survey team will therefore have to carry out several meetings at
local level, if possible accompanied by someone who already has the trust of the community
being surveyed (local priest, an NGO team already there, or maybe a team member who
comes from the area). These meetings should include general meetings with all the
community, to allay fears about why only certain people are being interviewed and to prevent
gossip undermining the survey work later in the day. In certain circumstances it may also be
necessary to make an appointment with each selected household to brief them on the purpose
of the activity. Only then will the team proceed to the next phases of the survey.

Phase C: Life history and inventory of household members and resources

Each of these lines of enquiry can throw light on the others. For example, discussion of the
life history of the farm household can help to identify its social boundaries and thus more
accurately determine who should be included or not in household composition data.
Similarly, inventory work can complement life history information history to fill in missing
data (for example, finding out the first purchase date of some production equipment, parcels
of land or new livestock).

Life history

Ask the farmer to relate the history of the family and the farm. Note all major events: starting
out as a farmer, marriages, deaths, illnesses, land inheritance, purchase or sale, purchases of
working animals, major changes to the production system, off-farm activities, and so on. If
necessary, ask for additional information based upon your own knowledge of local history.
The objective is to find out about major historical events (and possibly the reasons why they
took place). These can help to place events which older or illiterate people cannot put a clear
date to. In some field activities it is possible to use external references based on the life
period of "important" families in the community. Cemetery data can be useful. Thus the
technician can identify the period when someone of the X family died and subsequently ask
ask a farmer about change he made in the past, saying "at that time, was Mr. X still alive or
not? "

Household Members

List all residents of the FHS, with their ages and relationship to the FHS head(s). Ask about
people who are part of the household but not present during the survey (working abroad or
away for some other reason). Note what work each person does (on and off-farm tasks)
including children, during the different seasons. Draft a table summarizing this information,
highlighting the number of people who a) eat in the FHS and b) work, including field and
household activities. Distinguish between the tasks of men and women.


List all plots used by the family, their tenure status (ownership, tenant-contracts, share-
contracts, public or communal lands, etc), and all other land owned by the family (with or
without title). Asking the farmer to draw a blueprint of the different plots may be a good
starting point.

Improve on this with the help of the farmer.
Draft a table with each plot, its tenure status, The Commom: [...] how do we set about
present use, and agro-ecological constraints commons? Commons management has a
and capabilities. As Bruce says, "... particular community dimension, which cannot be
members of a household will often have captured through household interviewing
Alone. It must be approached initially through
specific tenure rights in particular parcels the small group and key informant interviews
and in fields within parcels. This is often [...] As a household may have a multi-tenure
particularly pronounced in situations where holding consisting of several parcels, so a
the production unit includes a number of community may have more than one
households or in the case of polygamous commons. It may have two pieces of commons
with the same tenure regime, or it may have
households where wives are assigned several commons under different uses and
separate individual fields... at any point in subject to different tenure rules It may, for
time certain parcels in the holding may be instance, have a communal forest; a common
held under tenure acquired by contract, such pasture on which trees grow, as well as
as leasehold, while other parcels which uncultivated interstices between parcels and
holdings. These commons areas must be
belong to the household may be encumbered identified and their various uses assessed. The
with such contractual obligations". managing group must be identified, its
membership clearly understood, its
Livestock Production institutional nature and potentials gauged, and
its various mechanisms for control of member
behavior evaluated [...1 Tenure in trees or the
List all animals owned by the family, common must however also be examined
including cattle, pigs, sheep, goats, poultry, from the viewpoint of the household.
fish production, etc., and include all types of Households' tenure extends to the commons:
draft animal. Try to get a preliminary idea of households which are members of the group
animal production. have rights to use the commons and may even
have specific rights in certain trees on the
commons under a system of tree nature. One
Infrastructure, Tools and Implements must evaluate the extent to winch those rights
provide effective incentives for households and
List all buildings, equipment, and main tools individuals to support and observe the rules,
and implements (hand tools, draft equipment, which control the use of the commons" (Bruce,
mechanical tools, etc). Note if possible the 1989)
purchase and present price of these items,
and purchase dates for the most important ones.

Phase D: Initial Consolidation of Data

After the first phase, an initial review of the data is carried out to check for missing
information and confirm (or not) the hypotheses driving the study. This work is of paramount
importance. It will avoid asking the same questions over and over again (which can annoy
informants and have a negative impact for the rest of the survey). If the data indicate that we
are on the wrong track, this process can indicate alternative hypotheses for testing in
subsequent phases.

Phase E: Second Farm Visit and Visits to Different Fields

Crop and livestock surveys must be done on site. Information from farmers obtained while
out in the fields is usually more accurate and can be immediately compared with visual
observation. If data are collected elsewhere (the village), it can be difficult to see the
difference between:

the specific situation of each plot during the year studied, giving us relevant
information to properly understand the problems of a particular farmer

the ordinary requirements for production, which would be almost the same for all the
farmers within a given locality.

Interviewers should ensure that farmers are not telling them what they want to hear (data
reflecting perceived ideals of production and practices). It will be necessary to explain to
farmers that only the specific data they themselves experimented with will help us in the
understanding of their production system bottlenecks. The same method will be applied to
livestock production.

In each plot or field location, the interview must:

check land-use during (at least) the last three years (crop rotation)
note all cultivation operations carried out in the plot as far as the studied period is
concerned, with their respective date, inputs (volume and cost), labor force
list all products and by-products obtained in the parcel of land
note the intended use of the various products (home consumption, sale, and barter, stock)
giving quantities, prices, and dates (animal feed by-products, non-timber products,
firewood, etc, should not be forgotten)

Sketching plant locations in the plot and measuring inter-plant distances would be useful to
assess yield component factors later on.

Care must be taken to distinguish family labor from hired labor, and any other forms of labor-
sharing systems between households. If this is not done, it will not be possible to find out what
the real return to family labour is within the farm production system. This is an essential
indicator for further economic assessment.

A similar process will be required for animal production. Inquiries will have to be made on
inputs and outputs, and the final use of each product. The interviewer will ask about all
aspects of animal husbandry, including feeding, grazing, use of purchased inputs, and
breeding. An assessment of variations in herd composition during the year may help to
understand the production system rationale and to evaluate gross animal production. Main
animal production indexes will have to be estimated, with appropriate questions, if livestock
production is an important component of the production system being studied.

The interviewers) must also investigate other activities in the FHS. Even if the bulk of field
production data is generally collected with the head (male or female) of the family, it is
equally important to check this and get additional information by talking to other members of
the household. Household consumption data and some production data will have to be
collected from women. These data will not only cover basic needs, family food consumption,

medicine requirements, clothing etc, but will also cover important areas of economic activity
such as vegetable gardening in backyard plots, trading activities, etc.

Phase F: Second Data Consolidation, Production System Coherence Assessment, and
Preliminary Data Analysis

The survey team will already have gathered a lot of data by now that needs to organised and
checked. Its overall coherence must also be checked before the final phase of the survey.
These two tasks are done concurrently in the second phase of data consolidation.

Four core time-schedule tables will be used as balance sheets to check data consistency and
to underline the principal constraints and opportunities:

- animal feeding balance sheet
- household staple food balance sheet
- labor requirements balance sheet (family, hired labor, ...)
- cash flow balance sheet

Each table will cover an "ordinary" agricultural cycle over a one-year period, and will give an
overview of seasonal problems. Illustrating each balance sheet with a diagram will facilitate
the dialogue with the farmer.

Animal feeding balance sheet

This balance sheet includes the year-round distribution of feeds for each type of animal,
rotative grazing, complementary feeds, etc. Other data reflecting seasonal livestock
production (calving period, milk production, etc) can be noted in the balance sheet to obtain a
better idea of problems and bottlenecks. This balance sheet can also be used to support a
dialogue with the farmer about animal husbandry problems.

Month May June July Aug Sept Oct Nov Dec Jan Feb Mar Apr
plot 1
plot 2
other feeds:
on-farm produced
critical periods

Labor requirements balance sheet

The balance sheet highlights critical periods for family labor. Some of these are only revealed
by the presence of hired labor or use of non-family labor even if family labor availability
seems to be higher than the normal labor requirement. Some agricultural tasks have to be
completed in a very short time (for instance, plowing, planting, harvesting). A monthly
partition is not an accurate indicator of all labor bottlenecks on the farm, but gives enough
information for a global view of the FHS situation. Balance sheet interpretation requires
thoughtfulness and a critical eye, and cannot be circumscribed as a mechanical exercise.

Month May June July Aug Sept Oct Nov Dec Jan Feb Mar Apr
plot 1
plot 2/
herd 1

other activities


other activities
Sub Total

Free labor
Hired labor


Household staple food balance sheet

This balance sheet assesses family food security. Harvesting dates, products kept for family
consumption, products sold and purchased are registered in the table. A line beginning at the
harvesting date shows the period during which each staple food is available for family
consumption. Women are usually key informants as far as family food is concerned.

The balance sheet will show food shortages before the harvesting season and the different
ways to resolve them. This can be established from the information on the production of each
plot or from the most important clusters of plots. Gathering and hunting activities have to be
included and seasonal variations in family composition must be allowed for when interpreting
the table.

May June July Aug Sept Oct Nov Dec Jan Feb Mar Apt
For each plot/ cluster of

- kept for family food
-* kept for sowing
- sold
- bartered
-> purchased

family requirements
hired labor requirements
animal feeding

Total requirements

Cash flow balance sheet

This reflects cash movements for both household and farm. The data in the table is therefore
very varied. For instance, cash inflows, aggregate production sales, labor earnings, and also
loan disbursements, and possibly an exceptional sale of some capital item.

May June July Aug Sept Oct Nov Dec Jan Feb Mar Apr
Cash inflows
- crops
- livestock
- non farm products
- capital items
- labor in agricultural tasks
- labor in non-agricultural
loan disbursements


|Cash outflows
- recurrent goods purchases
(crops, livestock, general farm
- hired labor
- loan repayments
-household consumption
- investments

Net Monthly

Confusion will be avoided by not going into too much detail. The purpose of the exercise is
to get an overall idea of the most important cash flows. This is not to be mistaken for an
accounting method. It is an analytical tool, and not an audit.

The cash flow balance sheet is worthwhile so as to reveal any recurrent cash problems, which
may have to be solved through adequate credit programs. Furthermore, it is a powerful tool
for checking and completing information that has already been collected. Although the table
seems complex at first sight, even illiterate farmers will find it easy to understand because it
corresponds with well-known and concrete problems, which they themselves have to deal

Other analyses, diagrams and tables

These balance sheets could be complemented by other tables to suit each case. It might be useful
for example to produce an abstract of the family life history highlighting the changes in the farm
and main trends shown by the data. Other sheets could trace the evolution of family land use
over the last 3-5 years, demonstrate a fertility-transfer scheme between the different plots, or
sketch rainfall distribution for each distinctive agricultural system.

Preliminary calculation of economic indicators

Once the balance sheets and tables have been prepared, the survey team can start the
provisional analysis of the data. Calculating the main economic indicators requires the use of
some specific concepts. Farming accounts are analyzed from two different points of view:

from the country point of view: indicators are used to compare the respective
efficiency of different types of farmers, and to classify them according to their
national interest. The most important indicator, the value-added, gives the amount of
new created wealth

from the FHS point of view: indicators and economic criteria will have to reflect family
farm behavior as much as possible. Distinctive analysis indicators will have to be used
for different kinds of farmers. For instance, it would be inaccurate to speak of any rate of
profit for a land and resource-less farmer who struggles for survival on a day to day

Annualfarm-household income is one of the most commonly used indicators for family farm
units. Calculating profit is done not by allocating an arbitrary value for each day worked by
each FHS member of their family (deducting both the real and the estimated costs from the
inclusive incomes), the efficiency of the FHS is calculated by estimating the real return to
family labour. This is done by deducting only real costs from annual gross-incomes (cash and 17
non-cash ones). During the farm survey, the results of this second approach will be the most
important ones to be shared with the farmer. (see section, The Framework of the Economic

Phases G and H: Validation and consensus with the community and fine-tuning of the

Towards the end of the survey process, it is essential to go back to the community and present
the preliminary results to the target households and proposed beneficiaries. This process of
'restitution' is the key moment of each diagnosis, for it underlines the participatory nature of
the whole exercise and allows all those involved to really ensure that the data and more
importantly, its interpretation is providing an accurate account of the situation.

This kind of meeting can take several hours. It has to be well planned in advance, to ensure
that all concerned are present. It is also advisable to break the session up into discrete
components, as it is unreasonable and also counterproductive to expect anyone to maintain
their interest in the subject matter over too long a period. Most importantly, the survey team
will have to be well prepared, and must go into the meeting with open minds ready to have
their conclusions challenged and possible altered by their local informants.

All the information collected during the survey and subsequently analysed will be shared
with the different households interviewed and other community members. They will be
requested to correct wrong interpretations, complete the data, and give their own point of

At this level, participation is no longer merely a formal exercise. Both the technician and the
farmers now have enough information about each other to be able to communicate. The
surveyor will have to change any points that may have been misunderstood, and may have to
improve on his calculations with the help of the farmers in order to obtain a consensus on all

This last meeting will also be used to explore alternatives and to devise feasible projects at
the farm level or at community level. New policy alternatives could also be discussed with
the farmers. Any proposal which emerges with the objective of moving from the present status

to a more desirable future status should be thoroughly thought through by all side. The
following questions need to be asked:

Is the proposal technically feasible?
Is the proposal economically profitable?
Will the proposal improve gender relations?
Is the proposal ecologically viable?
Is the proposal socially feasible?
Will the proposal respect local habits?
Will the proposal be politically feasible?


Given that we want to base our work on existing capacities and resources, we should not
propose technical changes that are far removed from local knowledge and expertise. This
means that while the technician requires experience of comparative agro-ecological systems,
the or she should also be able to match together the experiences of other FHS in the area.

If the technical proposal or something close to it has already been implemented by a farmer in
the same situation (but not a Model Farmer), there is a good chance that others will accept it.

Economic feasibility

As Morrisson said, "... bringing together the necessary resources is not sufficient to ensure
success, which depends not only on technical conditions but on economic ones as well. A
farmer will not adopt a system of cultivation merely because it is more advanced from a
technical point of view; it also has to be economically viable, in other words it has to provide
the farmer with a net income per hectare which is satisfactory in relation to the work required".

This question is also very much about risk. If the proposed activities expose the farm household
to high levels of risk, then they are unlikely to be adopted. When assessing this situation, the
security and regularity of other, off-farm income is of paramount importance; other external
inputs can also off-set risk during a start-up phase (for example, some kind of insurance scheme
or food-aid support while new fruit trees mature for example, or to cover farm labour costs until
returns to the investment come in).

Gender relations

The gender analysis carried out when doing the field surveys should not be considered as a
compulsory appendix to be included in each Guideline. It is an essential tool for understanding
the dynamics ofhousehold reproduction and the distribution oflabour and consumption linked
to certain types of activity. Male control of cash-crop incomes, even if the work is done by
women, is one example. Where this is the case, and men have no social obligations to support
the food budget of the household, there is no point in proposing new cash crop activities if the
objective is to raise the income of women and by extension promote better household food
security. The gender analysis must therefore enter right from the moment when proposals are
first elaborated, to see if they are going to produce a positive, neutral or negative impact in the

Ecological viability

Given that natural resources are the main production factor for the FHS, any external
intervention should not degrade these resources. Once again the technician should take account
of local environmental knowledge, as well us using his or her experience of comparative agro-
ecolgocial systems.

Social viability

The social viability of a proposal can be the most difficult aspect to consider. To let a PS reach
the API, or to establish a new FHS based on the technically viable PS proposed, two types of
action are normally needed: agro-economic improvement (moving toward line AE above); and
increasing the physical surface or area cultivated/exploited (moving toward line PH). The latter
action always means ensuring secure land access and longer-term rights over the resources
needed, not only over the land and other resources now being used, but also over the new
resources needed to facilitate the proposed changes in PS and the rising Reproduction
Threshold. ASD may only give us an estimate of what is needed, but the social feasibility of
such a program will essentially depend on the social relationships between FHSs, and between
the local farm community and wider agrarian system and national society in which it is

The survey should also produce a clear picture of local management and authority structures,
which may or not coincide with official or state institutions. Seeking the approval of these
structures is often an essential pre-condition for any proposal to be successful, and is one that is
surprisingly often over looked by technicians who fail to identify where the real authority lies
in a given community. At FHS level, this same question arises in relation to the roles of the
(male) head-of-household, who may not in fact be 'the farmer' and therefore should not be the
one to talk to when agreeing what may or may not the best farm development strategy for the

Local habits

New proposals should respect local habits. This may seem obvious but again is often
overlooked. For example, if for certain reasons a community does not work during a particular
period of the year or undertakes physically demanding obligations (such as fasting during
Ramadan), new proposals should not place demands on local people at precisely this time. This
might rule out an excellent technical proposal which requires large amounts of labour over the
period in question, but this simply has to be accepted.

A distinction must be made however between a practice which is generally shared by all
concerned, and one which is imposed by one group over another. Thus for example, a proposal
respecting a "traditional" habit of excluding women from the benefits of development
assistance should not be considered as a positive action, whether or not it respects local
"habits". Nonetheless, proposals that seek to by-pass such traditional practices and support the
development of women have to be sensitive to the overall social milieu and tread carefully.

Although the target beneficiaries may want to adopt a new technique or try a new crop,
technically or morally questionable constraints imposed by an elite or religious group may not
be desirable, but they may also underpin other important aspects of the local subsistence
strategy. A good example of this is kinship and godparent relationships between rich and poor
households, and strategic marriages which allow access to new resources or employment. In
Northeastern Brazil for example, this
has been shown to be an essential
factor which allows some poor Social Coping Mechanisms in the Face of Chronic Povertv
families to get through difficult d Ml ton
periods without suffering the loss of Detailed field research amongst a small fanner community
children through malnutrition (see in Northeastern Brazil in the 1980s revealed that all
Box). Such basic processes maybe households suffered the loss of at least one child through
exploitative or sustain an unequal the combined impact of malnutritin and disease.
system of chronic poverty, but they The key to understanding this information in the face or
should not be put at immediate risk other data which implied that such a disastrous FHS
by radical new proposals. outcome indicator was limited a samllpercentage of the
group of households being studied, was a detailed analysis
It is therefore of the utmost of their demlopmental cycles (the lifecycle discussed
earlier in this document). The same analysis did however
importance for the survey team to show that some households were able to avoid or minimize
hear from informants exactly what this disastrous and inevitable process. These households
their situation is, and whether going managed to maintain or develop social links with better off
against social norms and habits households and elite families as they went through the
established by or kept in place by 'developmental cycle'.
'higher' social group is really Thus, "Apart from affecting labour power and dependency
feasible or not. This observation ratios, developmental cycles also involve specific events
links in to some extent with the which profoundly affect the individual incomegathering
question of political feasibility strategies of households and their ability to cope with
discussed below. What is important malnutriin. Marriage and inheritance are particularly
e i t a S le ad important, offering new or expected combinations of
to recognize is that all FHS live and household labour, area fanned and tenure
work in a real world structured by relations...access to more lucrative or secure off-farm
real social and political relationships, often determined by patron-cliet or kinship
and that all proposals must be ties..."(Tanner 1987)
realistic in this context.


Finally, proposal that have the support of certain politicians are more willing to be implemented
than those that do not have this support. Unfortunately, the real situation we have to face is that
of plenty of technically interesting proposals produced by a good ASD have never been
implemented, while other politically interesting proposals that have no grounding whatever in
any kind of ASD have often been imposed on 'beneficiary' communities.

It is therefore very important that local leaders and politico-administrative authorities are fully
consulted during the Preparation Phases (A and B) of the survey, and at least feel that they have
been taken note of. While their interests are not necessarily the same as those of the people
they are supposed to represent, making them feel part of the participatory process is a good way
of ensuring the resulting proposals have their support.


Several knowledgeable sources have backed the use of Agrarian Systems Diagnosis in a wide
range of development programming and project planning situations. Nevertheless, although
the ASD approach can provide powerful insights into the process of agricultural development
at grassroots level, it does have a number of shortcomings. Opportunities and constraints tend
to differ between farm households, and a comprehensive picture requires familiarity with
several systems. Unless a suitably wide range of FHS is studied, programs may be developed
on the basis of an atypical stereotype. Thus the situation from the viewpoint of the farmer
needs to be supplemented by an analysis from the wider viewpoint of economic efficiency, in
order to develop a complete overview. These opinions voiced by UNDP are counterbalanced
by a strong statement in favour of ASD: "The basic information [produced by ASD] does,
however, offer a more immediately comprehensible and intuitive picture of sector problems
and opportunities. And it has the advantage that the conclusions and ideas developed have
been endorsed by what should be a representative segment of the farm population." (UNDP)

Finally, Agrarian System Diagnosis does not offer a development strategy, but only a set of
procedures aiming at the improvement of the standard of living of the target groups.
Understanding farm household systems helps to organize knowledge and direct data
collection, and can produce more effective interventions. On its own however, a holistic view
of the problems faced by a FHS is not enough to solve the problems face by a given FHS
situated within a given Agrarian System and wider politico-social context. Facing this
challenge requires a clear political commitment by Governments in favor of family farmers
and landless people. The link between technical anlaysis, strategic thinking, and policy
development is, as always, the least certain of all the variables discussed here.


Binswanger, H. (1994) Agricultural and Rural Development: Painful Lessons, revised version presented at the
32nd annual meeting of the Agricultural Economics Association of South Africa, Pretoria.
Binswanger, H., Deininger K. & Feder, G. (1993) Power, Distortions, Revolt an Reform in Agricultural Land
Relations, Working Paper, Washington: The World Bank
Bruce, J.W. & Migot-Adholla, S.E. (1994) Searching for Land Tenure Security in Africa, The World Bank
Carley, 1V (1994) Policy Management Systems an Methods ofAnalysis for Sustainable Agriculture and Rural
Development, IED-FAO
Centro de investigation y studios de la reform agraria (1981) Diagn6stico de base del Proyecto de
desarrollo rural integral PRONORTE, Managua: FIDA, BCIE, Gobiemo de Nicaragua
Chayanov (1985) La organization de la unidad campesina, Buenos Aires; Ed. Nueva vision
CIRAD (1994) Systems-Oriented Research in Agriculture and Rural Development, Montpellier: Cirad
Community Development Journal (1994) Vol. 29, N. 4 Latin America Special Issue
Development Dialogue (1977) Towards a Theory ofRural Development, Uppsala, Dag Hammarskjold
De Wit, P. (1998a) Community Land Delineation Case Studies in Mozambique: Lessons learned and
recommendations for future development. Maputo, Inter-ministerial Land Commission and FAO
De Wit, P. (1998b) Our Land Our Future: Decision Making on Future Land Resource Management The
FAO-UNEP Approach. Paper presented at the National Seminar on Land Use Planning for Sustainable Land
Resources Management, 17 September, Accra, Ghana. Rome, FAO, Technical Cooperation Programme and
De Wit, P. (1996a) Community Involvement for Promoting Rational Land Management. Bissau, Guinea Bissau,
Ministry of Public Works, Construction and Urban Planning/European Union. Project 'Valuing Land Resources
in Guinea Bissau'
De Wit, P., R Fonseca, B Biai, Q Nafatcham-na, B Naman, A Sami, M Reis & K Borges (1996) Natural
Resources,Land Tenure and Local Institutions. Bissau, Guinea Bissau, Ministry of Public Works, Construction
and Urban Planning/European Union. Project 'Valuing Land Resources in Guinea Bissau'
Dufumier, M. (1996) Les projects de developpement agricole Manuel d' expertise, Paris: CTA-Karthala
Dufumler, M. (1985) 'Systdmes de production et d6veloppement agricole dans le "Tiers Monde"', in Les
Cahiers de la Recherche Diveloppement N. 6, Paris
FAO (1999) Cursos e Experiencias Piloto na IdentificaVio e Delimitagao de Terras das Comunidades. Report
on the implementation of the training courses, and Course Content Manual. Maputo, Inter-ministerial Land
Commission and FAO Representation.
FAO, (1996) World livestock production systems. FAO Animal Production and Health paper n. 127.
FAO (1993) Guidelines for the conduct of a training course in farming systems development, Rome: FAO
FAO, (1984) Guidelines for designing development projects to benefit the rural poor, Rome: FAO
FAO, (1969) Farm Implements For Arid and Tropical Regions, Rome: FAO
FAO, El Ghonemy, Parson, Sinha, Uphoff, Wignaraja (1984) Studies on Agrarian Reform and Rural
Poverty, Rome: FAO
FAO, Institute of Pacific Studies of the University of the South Pacific (1984) Land Tenure and Rural
Productivity in the Pacific Islands, Rome: FAO
FAO-ILO (1996) Socioeconomic and Gender Analysis Programme, Draft
FAO-SARD, (1994) New Strategies for Agriculture, Forestry an Fisheries, Rome: FAO
FAO/UNDP (1994) Participatory Rapid Appraisal of Farmers'Agricultural Knowledge and Communication
Systems, mimeo
Groppo, P. (1992-93) 'El analisis comparative de los sistemas de producci6n' in Land Reform, Land Settlement
and Cooperatives, FAO
Groppo, P. Mekouar, M.A., Damals, G. Phouangphet, K (1995) Land Regularization Policy for Sustainable
Agriculture in Lao PDR, Rome: FAO (TSS-I)
IFAD (1995) Common Property Resources and the Rural Poor in Sub-Saharan Africa Rome: IFAD
Land Reform, Land Settlement and Cooperatives (1963), FAO
Madureira, S. (1999) Diagn6stico Participativo: Textos de Apoio Preparafao para osTrabalhos de Campo -
RecomendaVbes. Cursos e Experiencias Piloto na Identificacao e Delimitagao de Terras das Comunidades.
Maputo, Inter-ministerial Land Commission and FAO Representation.
Mazoyer, M. (1992-93) 'Pour des projects agricoles legitimes et efficace: thborie et methode d'analyse des
syst6mes agraires' in Land Reform, Land Settlement and Cooperatives, FAO

Mazoyer, M. (1981) The Agricultural Crisis in the Least Developed Countries, Paris: OECD Development
Centre, Draft
Mellor J. W. & Desai G.M. (1985) Agricultural Change and Rural Poverty, Variations on a Theme by Dharm
Narain, John Hopkins University Press
Merlet, M. (1995) Tipologia de productores agropecuarios. Estudios de casos defincas: guia metodologica,
Paris: IRAM
Merlet, M. (1996) Land Tenure and Production Systems in the Cordillera (The Philippines), Mission Report,
draft, Rome: FAO
Richards P. (1985) Indigenous Agricultural Revolution, London: Hutchinson
Rose T. (Edited by) (1985) Crisis and Recovery in Sub-Saharan Africa, Paris: OECD Development Centre
Tanner, Ch. R. (1991) Relations Between Ponteiros and Tabancas: Implications for a New Land Law in
Guinea Bissau. Report for USAID-Bissau. St Ives, England, CTC Consulting
Tanner, Ch. R. (1987) Malnutrition and the Development of Rural Households in the Agreste ofParaiba State,
Northeast Brazil. Journal of Development Studies, Vol 23, No 2, pp242-264, London, Frank Cass
Tanner, Ch. R., De Wit P. & Madureira, S. (1998) Proposals for a Programme of Local Community
Delineation. Document presented to the National Seminar on Community Land Delimitation and Management,
Beira, Mozambique 12-14 August 1998. Maputo, Inter-ministerial Commission for the Revision of Land
Legislation/FAO, Project to Support the Development and Implementation of a National Land Programme.
Trebuil G. & Dufumier, M.(1990) 'Diagnoses on Regional Agrarian Systems and Sustainability of Agricultural
Production Systems in Thailand' Asian Farming Systems Research and Extension Symposium. Bangkok
UNDP (1994) Sustainable Human Development and Agriculture
van Zyl, J., Barbosa T., Parker, A & Sonn, L. (1995) Decentralized Rural Development and Enhanced
Community Participation A Case Study from Northeast Brazil,The World Bank Policy Research Working
Paper n. 1498
The World Bank (1996) The World Bank Participation Sourcebook, Washington: The World Bank

Additional bibliography

AA.VV. (1994) Memorias del III Seminario Internacional: Desarrollo Sostenible de Sistemas Agrarios, Cali,
Colombia: IER-Universidad Javeriana, CIPAV, IMCA, Universidad de Ciencias Agricolas de Suecia
AA.VV. (1994) "Campesinado y recursos naturales", en Revista Agricultura v Sociedad, n. 10, Santiago, Chile:
AA.VV. (1993)Tierra, Economia y Sociedad, Colombia: FAO-PNUD-INCORA
AA.W. (1988) "Sistemas de producci6n campesinos: concepts y resultados", Revista Agricultura v Sociedad,
n. 6, Santiago, Chile: GIA
ABRA, (1992) n. 15-16, Heredia, Costa Rica: Revista Universidad Nacional
Adhikarya, R. (1994) Strategic Extension Campaign: A participatory-oriented method of agricultural
extension, Rome: FAO
Ancey, G. (1983) Monnaie et structures d'exploitations en pays Mossi Haute-Volta, Paris: Orstom
Association Europeenne des Economistes Agricoles (1989) (19eme seminaire europeen) Producteurs et
Consommateurs face auxpolitiques agricoles et alimentaires en Afrique (le defi de la connaissance),
Montpellier: AGRIPOLIS
Bebbington, A.; Ramon G. (Coordinadores) (1992) Actores de una decade ganada: tribus, Comunidades y
campesinos en la modernidad, Quito, Ecuador: Comunidec
Berdegue, J. & Nazif L Editores. (1988) Sistemas de producci6n campesinos, Santiago de Chile: GIA
Berdegue, J. & Ramirez, E. (1995) Investigacion con enfoque de sistemas en la agriculture y el desarrollo
rural, Santiago, Chile: RIMISP
BLACT-CFECTI-GRET-SEAR, (1984) Demarches en recherche-developpement appliquees au secteur de la
production rurale des pays en voie de developpement, Paris: GRET
Bojanic, A.; Canedo, M.E.; Gianotten, V.; Morales, M.A.; Ranaboldo, C. & Rijssenbeek, W. (1994)
Demandas campesinas. Manual para un analisisparticipativo, La Paz, Bolivia: Embajada Real de los Paises
Bonnemaire J.; Jouve Ph. et al. (1987) Appui pedagogique d l'analyse du milieu rural dans une perspective de
developpement, Collection syst6mes agraires, n. 8, Montpellier: CIRAD-ENSAA-CNEARC
Bruce, J. (1989) Rapid appraisal of tree and land tenure Rome: FAO CommunityForestry Note 5
Canall, M. (1994) La dinamica dei sistemi agrari, Bologna: CUSL
CEPAL/FAO (1986) Agricultura campesina en America latina y el Caribe. Santiago, Chile: CEPAL/FAO
Chonchol, J. (1994) Sistemas agrarios en America Latina, Mexico: Fondo de Cultura Econ6mica
Costa, F. de A. (1995) Agricultura Familiar em Transformagao na Amaz6nia: o caso de Capitao Pogo e suas
implicagoes para a political e planejamento agricolas regionais. Universidade Federal do Para, Belem, Brasil:
Clouet, Y.; Guilloneau, A. & Ruf, Th. (1986) Diagnostic du system agraire et des systems de production en
Segala Aveyronnais, CIRAD-DSA/CDA/INRA
Confino, M. (1969) Systemes agraires etprogres agricole: I'assolement triennal en Russie auxXVIII-XIX
siecles, Paris: Mouton
de Castro, J. (1946) Geografia dafome, Rio de Janiero: O Cruzeiro
de Grandi, J.C. (editeur) (1996) L evolutionn des systems de production agropastorale par rapport au
developpement rural durable dans les pays d'Afrique soudano-sahelienne, Rome: FAO
Dey, J. (1984) Women in Rice-Farming Systems, Rome: FAO
Dietze, R.; Kohler, A.; Ocampos, G.; Ramos, J. & de Zutter, P. (1993) Voces yPistas para un desarrollo
sostenible, Paraguay, Asunci6n: DPG/MAG-GTZ
Ducourtleux, 0. (1994) L'agriculture du plateau des Bolovens (Evolution du system agraire de la region de
Paksong -Sud Laos) Rome: FAO, Land Tenure Working Papers
Dudermel, Th. (1990) Les enjeux d'une agriculture en cruise: capitalistes et paysans du Nord-ouest du Rio
Grande do Sulface aur bouleversements economiques, PhD Thesis, Paris, France: INAP-G
Dufumier, M. (1990) Les politiques agraires, Paris:PUF, Col. Que sais-je?
Dufumier, M. (1985) Systemes de production et developpement agricole dans le Tiers Monde, in Les Cahiers
de la recherche d6veloppement, n.6 Montpellier
EMBRAPA-CPATSA-CIRAD-SAR. (1993-95) Sistemas de producao da agriculture familiar: i) A relagao
dos sistemas de produgao com a comercializagao e a trasformagao, num enfoque de pesquisa-desenvolvimento;
ii) enfoques de pesquisa sobre as agriculturasfamiliares do tr6pico semii-rido; iii) zoneamento por entrevista
de pessoas chaves, proposta metodoldgica para subsidiary o planejamento municipal; iv) a evolugao de un
program de pesquisa em sistema de produgao no Nordeste brasileiro; v) Estudo do process de
desenvolvimento e da construgao do espago rural para subsidiary o planejamento; vi) Inovagao institutional,

planejamento municipal e organizafao de produtores no Nordeste brasileiro: analise comparative de tris
experiencias. Petrolina, Brasil: CPATSA
Eresue M.; Brougere, A.M. (compiladores) (1988) Politicas agrariasy estrategias campesinas en la cuenca
del Cahete. Avances de una investigation Lima, Peri: UNALM/IFEA
Eresue, rM.; Gastellu, J.M.; Malpartida, E. & Poupon, H. (1990) Agricultura andina: unidady sistema de
production, Lima, Peri: ORSTOM/UNALM/Editorial Horizonte
Escobar, G.; Berdegue, J. (1990) Tipificacidn de sistemas de produccion agricola. Santiago. Chile: Rimisp
FAO (1993) Guidelines for Land Use Planning, Rome: FAO
FAO (1990) Guidelines for the conduct of a training course informing systems development Rome: FAO
FAO (1989) The Dynamics of Land Tenure and Agrarian Systems in Africa Rome: FAO
FAO. (1971) L'agriculture itinerante en Amerique latine Rome: FAO
FAO-UNA. (1995) Seminario Regional "Diagndstico de Sistemas Agrarios"-ompendio de sintesis teorico-
metodologicas de aplicacion sistimica a realidades agrarias, San Jose, Costa Rica
Franceschetti G. (a cura di) (1990) Governare il territorio. Milano: Franco Angeli
Ghai D., Khna A.R., Lee E., Radwan S. (1979) Agrarian Systems and Rural Development, London:
Macmillan Press Ltd
Gomez y Paloma, S. (1995) Developpement economique et systems agraires, Rome: FAO Land Tenure
Working Papers
Gomez y Paloma, S. (1993) Mediterranean and Baltic Essays on Contemporary Agricultural Systems,
Roskilde: Universitetscenter Federico Caff6 Center for Sydeuropaeiske Studier
Grandstaff, T.B. & Messerschmidt, D. A. (1995) A Manager's Guide to the Use of Rapid Rural Appraisal,
Thailand: FARM Programme, FAO/UNDP & Suranaree University of Technology
Groppo, P. (1991) Diagnostico de sistemas agrarios: una metodologia operative, Santiago de Chile: FAO-
Institute Panafricain pour le Developpement. (1981) Comprendre une economic rurale, guide pratique de
recherche. Paris: L' Harmattan
Junta del Acuerdo de Cartagena (1988) Metodologia para el studio de los sistemas de production agricolas
confines de desarrollo rural. Venezuela
Lecomte, B. (1986) L'aide par project, limits et alternatives, Paris: OCDE Development Centre
Lizet B. et de Ravignan F. (1987) Comprendre un paysage guide pratique de recherche, Paris, France: INRA
MAARA-INCRA (1995) Construindo o assentamento, Fortaleza, Brasil: Maara-Incra
Malpartida E.; Poupon, H. (Compiladores y Editores) (1987) Sistemas agrarios en el Peru., Lima, Peru:
Mazoyer, M. & Roudart L. (1998) Histoire des agriculture du monde, Paris, France, Seuil
Mazoyer, M. (1987) Dynamique des systems agraires. Rapport de synthese du colloque sur la dynamique des
systems agraires. Paris: Ministere de la Recherche et de la Technologie
Mazoyer, lM (1977) "Evolution et differenciation des syst6mes agricoles d' exploitation de la nature", in
Journal d'agriculture traditionnelle et de botanique applique, n. 2-3, vol. XXIV, Paris
Mercoiret, M.R.; Deveze, J.C. & Gentil, D. (1989) Les interventions en milieu rural. Principes et approche
methodologique. Paris, France: Ministere de la Cooperation et du Developpement
Mercoiret, M.R. (coordinator) (1994) L'appui aux producteurs ruraux: Guide a l'usage des agents e
developpement et des responsables de groupements, Paris, France: Ministere de la Cooperation/Karthala
Mondain Monval J.F., (1993) Diagnostic rapide pour le developpement agricole, Paris: Editions du GRET
Namdar-Irani, M. (1989) Campesinos y desarrollo tecnologico: una tipologia para la accion. Santiago, Chile:
Orstom. (1988) Cahiers des Sciences Humaines: Systemes de production agricole en Afrique tropical, vol. 24
Orstom. (1987) Le developpement rural: comprendre pour agir, Paris: Orstom
Passerini, L (a cura di) (1978) Storia orale vita quotidiana e cultural material delle classic subalterne.
Torino, Italy: Rosenberg & Sellier
Piccioni, N. (1995) Apoyo a las organizaciones de los pequehos productores para la comercializacion de
granos basicos -franja granifera San Ramdn-Esquipulas, Matagalpa, Nicaragua: MAG/PAN/FAO
Poats, S.V.; Schmink, M. & Spring, A. (1986) Gender Issues in Farming Systems Research and Extension,
London: Westview Press, London
Prestamburgo, M; Tempesta, T. (a cura di) (1994) Sistemi produttivi, redditi agricoli e political ambientale,
Milano, Italia: Franco Angeli
Romeiro, A.; Guanziroli, C.; Palmeira; M. & Leite, S. (organizadores) (1994) Reforma agraria: produFao,
emprego e renda o relatdrio da FAO em debate, Brasil: VOZES/IBASE/FAO
Ruthenberg H., (1980) Farming systems in the Tropics, Oxford: Clarendon Press
Saravia A. (1983) Un enfoque de sistemas para el desarrollo agricola, San Jose, Costa Rica : IICA

Schanner W.W. et al. (1982) Farming systems research and development. Guidelines for developing
countries, Colorado: Westview Press
Sebillotte M. (1976) Jachere, systeme de culture, system de production, Paris: INRA
Sereni, E. (1987) Storia delpaesaggio agrario italiano, Bari, Italy: BUL
Universidad Nacional (1981) Autobiografias campesinas, Heredia, Costa Rica
Villaret, A. (1994) El enfoque sistemico aplicado al andlisis del medio agricola. Sucre, Bolivia: Pradem/Cieda
Warner, K (1991) Shifting cultivators: local technical knowledge and natural resource management in the
humid tropics, FAO Community Forestry Note 8