SD Publication Series
Office of Sustainable Development
Bureau for Africa
Measuring the Impacts
of Natural Resource
Management Activities
in Mali's Upper
Niger Valley
Valerie A. Kelly
Department of Agricultural Economics
Michigan State University
Technical Paper No.
December 2003
117
i
Division of Economic Growth, Environment and Agriculture
Office of Sustainable Development
Bureau for Africa
U.S. Agency for International Development
Measuring the Impacts
of Natural Resource
Management Activities
in Mali’s Upper Niger Valley
Valerie A. Kelly
Department of Agricultural Economics
Michigan State University
December 2003
Technical Paper No.
Publication services provided by The Mitchell Group, Inc. (TMG)
pursuant to the following USAID contract: AFR/SD Support Services
Task Order # RLA-M-00-04-00009-00
117
ii
iii
Contents
Map, Figures and Tables v
Acknowledgments vii
Executive Summary ix
Forward xiii
Background 1
Objectives and Methods 3
Conceptual Framework 5
Rapid Appraisal Results 7
Expanding the Progress 11
What to do
Problems to resolve
Better Quantifying Progress Made to Date 17
Improvements in counting and reporting adoption
Improvements in reporting OHVN production and yield statistics
Case study approach to collecting and analyzing NRM farm-level income impacts
Rough extrapolations from case studies to the sector level
Survey of OHVN farmers using a representative sample
Development of an ongoing program of monitoring key income and food security indicators
Closing Remarks 25
References 27
Appendices
Appendix 1: Suggested Format for Periodic Reporting of NRM Village and Farm Adoption
 in the OHVN 29
Appendix 2: OHVN Case Study of Production and Income Changes for a Farmer Having
 Used NRM Practices During Nine Years 31
Appendix 3: Draft Questionnaires for Collecting Data from AVB/Animateur Notebooks 35
Appendix 4: Illustration of Budget Analysis Possible Using the Types of Data in Appendix 3 41
Appendix 5: Contents, Forward, and Introduction from A Methodology for Estimating Household
 Income in Rural Mozambique Using Easy-to-Collect Proxy Variables 47
iii
iv
v
Map: Map of the Upper Niger Valley Region of Mali, showing several villages in the OHVN study xv
Figure 1: Sustainable Economic Growth Strategic Objective Results Framework 5
Figure 2: Conceptual Framework of OHVN Impacts, Objectives and Activities 6
Table 1: Area, Production and Yield Data for the OHVN: 1991/1992–1998/1999 8
Table 2: Illustrative OHVN Adoption Report: Physical Indicators of NRM Adoption 18
Table 3: Illustrative OHVN Adoption Report: Villages, Farmers and Recovered Area 19
Map, Figures and Tables
vi
vii
Acknowledgments
So many people contributed to this report that it is not
possible to name them all individually. Mike McGahuey
(USAID/AFR/SD) must be singled out, however,
because without his keen interest in NRM and strong
desire to better understand how NRM activities fit
into the overall scheme of agricultural transformation
in Africa, I would not have had the opportunity to
visit the OHVN and prepare this report. Mamadou
Lamine Sylla also made a major contribution to this
report by organizing the field trips, providing me
access to a wide variety of OHVN documents, and
participating in many open, frank discussions about
the strengths and weaknesses of the NRM program
that he is managing.
Those who contributed the most to this effort—the
farmers who interrupted their daily routines to spend
the day with us during our field visits—are too
numerous to mention individually. These farmers
contributed to this report not only by chatting with us
and allowing us to visit their farms, but also by working
hard during the last 10–20 years to improve their
farming techniques, literacy skills and management
abilities.
Others who helped, either in the design of the rapid
appraisal or by commenting on the various presenta￾tions I have made concerning this work, are Gaoussou
Traoré and Mamadou A. Dembele of USAID/Bamako
and Bob Winterbottom of IRG.
viii
ix
Executive Summary
A visit to seven villages in the Office de la Haute Vallée
du Niger (Office of the Upper Niger Valley, or OHVN)
and discussions with about 100 farmers using natural
resource management (NRM) practices confirmed that
something good is happening in the zone (see Section 4):
+ Yields of all crops are increasing for farmers
adopting NRM intensification methods.
+ Farmers are unanimous that life is better now than
10 years ago.
+ Farmers are optimistic and enthusiastic about the
future.
These results come from a complex process that has
been going on for more than 15 years (Sections 2 and
3). Ingredients contributing to the current success
appear to be:
+ Identification of technologies capable of increas￾ing declining yields
+ Potential for increased cash income from improved
cotton production
+ Community approach to implementation
+ Focus on youth
+ Focus on villages/farmers most likely to benefit
from NRM actions
+ Use of demonstration effect through model farm￾ers and model villages
+ Incremental training (literacy, technical skills, com￾munity organization, management)
Support services offered have included:
+ Roads
+ Credit guarantees for limited period following man￾agement training
+ Input/output transport assistance
+ Regular supervision and support to trainees
+ Some free equipment for implementing NRM ac￾tivities
+ Market research by OHVN to help with crop di￾versification
Looking toward the future, two questions need to be
addressed:
1. Is it possible to extend these results by . . .
+ further increasing yields/incomes of current
NRM farmers?
+ reaching a broader group of OHVN farmers?
+ reaching farmers outside the OHVN area?
2. Is it possible to quantify the impacts of NRM in￾tensification activities in terms of . . .
+ benefits realized by farmers?
+ benefits realized by Malians in general?
+ benefits realized by the rest of the world?
The answer to both questions is yes. Suggestions for
accomplishing these tasks are contained in this report
(Sections 5 and 6).
x
xi
Glossary of Acronyms and
Abbreviations
AVB agents vulgarisateurs de base
APCAM Assemblée Permanente des Chambres d'Agriculture du Mali
BNDA Banque Nationale de Développement Rural
CLUSA Cooperative League of the USA
CMDT Compagnie Malienne pour le Développement des Textiles
FCFA CFA franc
IPM Integrated pest management
NGOs non-governmental organizations
NRM natural resource management
OHVN Office de la Haute Vallée du Niger
PASIDMA Projet d’Appui au Système d’Information Décentralisé du Marché Agricole
USAID/W U.S. Agency for International Development, Washington, D.C.
xii
xiii
The Office of Sustainable Development in USAID’s
Africa Bureau (USAID/AFR/SD) has long used formal
assessments of natural resource management (NRM)
programs it has assisted to evaluate impacts and take
stock of lessons from its investments. The experience
of USAID/Mali’s Upper Niger River Valley program
was—and continues to be—particularly fruitful. In
addition to providing lessons for other programs, it
offers a real foundation for hope that the degradation
of farmland can be halted and even reversed using a
multiyear, multifaceted approach.
Beginning in the late 1980s, this program has brought
together agriculture, NRM, microenterprise and
governance elements in promoting improved NRM
practices for farmers. Over the 12 years before the
study discussed in this publication, personnel from the
mission, AFR/SD, and various research institutes had
already conducted several informal assessments in the
program area, which is in the zone covered by the
Office de la Haute Vallée du Niger (OHVN). These
experts noted that, at least in some communities, there
appeared to be positive trends in terms of improved
livelihoods, decreased degradation, and strengthened
governance.
In order to improve its information on the lessons
produced in the OHVN, AFR/SD invited Dr. Valerie
Kelly of Michigan State University to assess people’s
perceptions in some of the communities involved. In
early 2000, Dr. Kelly visited seven communities in the
OHVN and interviewed about 100 farmers. While this
was a small sample of the total OHVN population, Dr.
Kelly conducted the assessment from a perspective
honed by over 20 years of experience of assessing rural
development in West Africa and elsewhere. In addition,
Dr. Kelly conducted extensive interviews with OHVN
personnel and reviewed critical reports.
Based on her interviews, Dr. Kelly noted the following
preliminary conclusions about the farmers she spoke
with: (a) yields for all crops increased for farmers
adopting intensified NRM practices; (b) farmers were
unanimous that life had improved over the last 10
years; and (c) farmers were optimistic and enthusiastic
Foreword
about the future. She went a step further and noted that
across the interviews, two things stood out: first, as
farmers moved from near-subsistence to commercial
agriculture, they started treating farming as a business.
Second, they used NRM practices as a way to increase
the efficiencies of fertilizers, improved seeds and other
investments necessitated by a transition to commercial
farming operations.
Dr. Kelly was particularly emphatic about the impacts
that the management/literacy training provided by the
Cooperative League of the USA (CLUSA) had on
farmers, both in developing entrepreneurial skills and in
helping people develop and manage cooperatives. As
members of competent coops, farmers could do a
number of things that they could not do as individuals.
She noted the importance of this new feeling of
empowerment on people’s attitudes. Overthrowing
feelings of helplessness has been paving the way for
people to take effective action to move out of poverty.
Dr. Kelly was clear that she thought that she visited
some of the better OHVN communities, and she had
hard questions about the extent of the impacts that she
saw. She noted, for example, that while the information
from participating farmers showed positive yield trends
over the last 10 years, yield changes for the zone as a
whole over the same period were stagnant. Moreover,
marketing missteps caused problems for a number of
farmers, leaving them with unsold stocks of some
grains and vegetables. More training in marketing for
farmers may be indicated, along with more and better
information on demand volume, transportation prob￾lems, and special requirements affecting the national,
regional and European/U.S. markets.
However, Dr. Kelly was also clear that there appeared to
be no particular constraint to scaling up the impacts
seen in the communities that she visited. Additional
work will be necessary to more clearly identify the
activities that would best contribute to such an
expansion, and she provided several options for taking
the next steps.
xiv
From AFR/SD’s perspective, Dr. Kelly’s work showed
that the lessons that were produced in the OHVN are
not crop- or site-specific. Most would apply to a broad
range of sustainable development problems. We
recommend that you read this thoughtful study, glean
the lessons, and add your own.
Carl M. Gallegos, Ph.D.
Acting Chief
Division of Economic Growth, Environment and
Agriculture
USAID/AFR/SD
Note: a previous version of this study, Measuring the
Impacts of Natural Resource Management Activities in
the OHVN, was made available on the web by IRG, Ltd.,
the firm which arranged for Dr. Kelly’s visit to Mali in
2000. The document may be found at http://
www.dec.org/pdf_docs/PNACP456.pdf .
xv
MAP OF THE UPPER NIGER VALLEY REGION OF MALI,
SHOWING SEVERAL VILLAGES IN THE OHVN STUDY
Map courtesy of Peter Freeman, Development Ecology Information Service
www.devecol.org

1
Background
Over time, the development community has talked
about the fact that different models of sustainable de￾velopment should include one in which entrepreneur￾ial farmers invest in systems that generate more secure
and prosperous livelihoods and decrease degradation
rates. By several measures, growing numbers of pro￾ducers in the OHVN (Office de la Haute Vallée du
Niger) zone of Mali appear to be on the road to this
type of sustainable development. Information available
from informal appraisals and the OHVN database sug￾gests that a significant number of producers are mov￾ing from subsistence systems to diversified, revenue￾generating systems where yields are increasing and
degradation rates are falling. The system is built on
production practices that integrate natural resource
management (NRM) with investments in inputs (fer￾tilizers, improved seeds). In principle, this integrated
system uses inputs more efficiently and allows pro￾ducers to practice intensified agriculture on less land.
Commercial credit is the source of capital for many of
these investments, and, judging by the repayment rates,
the producers have achieved a high level of compe￾to be progress toward community-financed extension
systems and community-financed support to improve
the delivery of health and education services. The
OHVN experience appears to merit closer study to
(1) better quantify the results and (2) draw lessons
that can be applied to other situations.1
1 This introductory paragraph is adapted from my scope of work,
which was drafted by Mike McGahuey.
tency in enterprise management. There also appears
2
3
Objectives and Methods
Given general perceptions of what has been happen￾ing in the OHVN during the recent past, it seems worth￾while for USAID and OHVN to measure and docu￾ment the OHVN program’s impacts better. This report
is a first step in that direction. Specific objectives are
to (1) confirm the general perceptions described above;
(2) recommend low-cost, easy-to-implement methods
for better quantifying the impacts of NRM/intensifi￾cation practices; and (3) recommend actions that can
be taken to increase adoption of promising NRM/in￾tensification practices.
To accomplish these objectives, I (1) reviewed a wide
range of documents describing activities in the OHVN
zone during the last 20 years (see References); (2) iden￾tified existing databases concerning the OHVN that
could contribute to current objectives; (3) developed a
format for conducting group discussions with farmers
and OHVN agents concerning their experiences with
NRM techniques (techniques adopted, factors influ￾encing adoption, impact on production, impact on in￾comes and standard of living, etc.—see Appendix 1);
(4) conducted the group discussions during four days
of field visits organized by OHVN; (5) discussed pre￾liminary findings and recommendations with USAID/
Bamako and OHVN staff; (6) made two presentations
of preliminary findings in Washington, D.C., to
USAID/Washington personnel and representatives of
organizations collaborating with USAID/W on NRM
activities; and (7) drafted the current report describing
key findings and recommendations.
In my work I have focused on describing—and, when￾ever possible, quantifying—changes in agricultural
productivity and incomes that have taken place among
farmers who adopted NRM practices during the last
decade. It is important to note from the start that these
changes cannot be attributed with certainty to any par￾ticular USAID investments or OHVN activities, since
the preconditions for doing an analysis of causality
over time are absent. The most important precondition
lacking is the ability to isolate USAID contributions
from other historical events. USAID is only one of
many actors in the OHVN, and during the last decade
many things have happened in Mali (e.g., structural
adjustment, market liberalization, restructuring of
OHVN, devaluation of the CFA franc, a military re￾gime replaced with a democratically elected govern￾ment, etc.) that have contributed to the higher level of
agricultural productivity and income that we find in
the OHVN today.
Another problem is the nature of the USAID contribu￾tion—it was a very diverse contribution covering a
wide range of interventions that varied across time and
space depending on initial conditions and the expressed
needs of different communities and farmers. Some ac￾tivities were specific to the OHVN project (e.g., sup￾port for extension services, road building, literacy
training, OHVN restructuring, credit guarantees),
while others were activities supported by the USAID
country program that had an impact in a number of
places besides the OHVN area (support for input/out￾put market liberalization, governance and democracy
activities, youth training/employment activities, etc.).
When appropriate, I call attention to some of the
USAID-funded activities that seem to have been par￾ticularly important components of the overall environ￾ment that stimulated productivity and income growth
in the OHVN, but it must be stressed that these obser￾vations are based on qualitative rather than quantita￾tive assessments.
4
5
Conceptual Framework
In any effort to evaluate the impacts of a program, it is
important to begin with a theoretical picture of how
the program activities are likely to affect selected indi￾cators and produce desired outcomes. Figure 1 is
adapted from the results framework used by USAID/
Mali to monitor activities contributing to their sustain￾able economic growth strategic objective. The major
change I have made is to add a row between the inter￾mediate result of “increasing sustainable dryland agri￾cultural and NRM practices” and the strategic objec￾tive of “increasing value added to national income.”
This intermediate row represents the positive impacts
on agricultural productivity and farm incomes that must
occur if the strategic objective is to be achieved. In
the longer-term process of quantifying contributions
of NRM activities to national income, I believe the
first step is collecting farm-level evidence that produc￾tivity and incomes are increasing in areas where NRM
practices are being adopted.
Figure 1:
Sustainable Economic Growth Strategic Objective Results Framework
USAID Strategic Objective
 Increased Value Added to National Income in Agricultural Sector
Intermediate Impacts
Increased Agricultural Productivity
Increased Farm Incomes
Intermediate Result
Increased Sustainable Dryland Agricultural and NRM Practices
Activity Results
Cropping Tenure Prolonged
Degraded Lands Rehabilitated
Afforested Area Increased
Integrated Pest Management (IPM) Technologies Increased
6
Figure 2 is adapted from the USAID results frame￾work designed specifically for USAID’s OHVN
activities. The strategic objective of the results frame￾work is “better production practices adopted by farm￾ers in the OHVN.” Figure 2 shows that adoption of
improved production practices is thought to be
fostered by working on the “facilitating variables”—
that is, by improving farmer access to commercial capi￾tal, decreasing transport costs, increasing community
control over local resources and improving farmer
knowledge of alternative production practices. For a
decade now, OHVN and USAID have been monitor￾ing changes in the facilitating variables as well as in￾creases in the adoption of improved production prac￾tices. These are all variables that can be monitored by
counting numbers of loans issued, kilometers of roads
built, number of villages managing their own forests,
etc. These indicators, however, do not provide us with
much information on how much (if at all) the adoption
of these improved technologies is improving agricul￾tural productivity and incomes. As these are the types
of impacts we now want to evaluate, I have added a
line of “impacts” above the strategic objective line.
In summary, what I am attempting in this report is to
go beyond the OHVN project’s strategic objective of
increasing adoption of NRM practices, to an evalua￾tion of the broader impacts that adoption of these
practices is having on agricultural productivity and in￾comes. I doubt, however, we are at a point where we
can begin quantifying how much OHVN’s NRM ac￾tivities contribute to the value added at the national
level. Although this remains the ultimate objective, I
do not believe it can be done in a credible way until
we are able to quantify key productivity and income
impacts at the farm level.
Figure 2: Conceptual Framework of OHVN Impacts, Objectives and Activities
 Capital Transport
Management Training
Impacts: Increased Agricultural Productivity and Farm Incomes
Banking Bank Negotiating Market
 Skills Policies Skills Liberalization
Strategic Objective: Better Production Practices Adopted by Farmers
 Community Knowledge
Literacy Training
 Village Forest Extension Farm Days
 Associations Services
7
My rapid appraisal was based on four days in the field,
seven village meetings that included group discussions
with an estimated 100 farmers, and a review of OHVN
documents. It leads me to confirm the impressions of
Mike McGahuey and others who have been working
in the OHVN zone for a number of years—something
good is happening in the zone. Evidence of progress
includes these trends:
+ Crop yields are increasing for farmers adopting
NRM intensification methods.
We don’t know how widespread this increase is,
as there is not yet strong evidence of it in the ag￾gregate data, but all villages visited provided nu￾merous illustrations based on individual farmer
records (see Appendix 2).
+ Village youth are staying at home to farm rather
than migrate.
This was very evident in all villages visited; youth
were present at all meetings, they played impor￾tant roles in managing farmer associations, and
they were very active participants in rapid appraisal
discussions.
+ Farmers are investing heavily in agricultural equip￾ment, traction animals and livestock.
When asked what they were doing with their in￾creased incomes, the most common response was
investment in equipment and/or livestock.
+ Farmers are diversifying, with many new forays
into dry-season crops and tree crops.
Increased production of horticultural products
during the dry season (green beans for Europe,
onions/tomatoes and bananas for Bamako, possible
increases in sorrel production for export to the
United States) is one of the reasons for the reduc￾tion in out-migration. Marketing remains a prob￾lem here, but the farmers’ associations appear to
Rapid Appraisal Results
have a level of management skills permitting them
to deal with the setbacks and move ahead (vs. the
old days when they expected the government to
bail them out).
Tree crop production (particularly teak for pro￾duction of construction poles) through develop￾ment of village and private woodlots is expanding
slowly, but most examples seen during the rapid
appraisal had not yet begun to generate income.
+ Farmers are unanimous that life is better now than
10 years ago.
‰ They eat better (more food and better variety)
‰ They dress better
‰ They travel more easily (motorbikes have
replaced bicycles in many cases)
‰ Schools and health services are more
accessible
‰ They are better educated (literacy programs
and management training by CLUSA)
+ Farmers are optimistic/enthusiastic about the
future.
There is always the possibility that the villages visited
were exceptional ones and not typical of the zone.
There is no way to know this for sure without doing a
survey with a large, randomly selected, representative
sample—a potentially costly endeavor. The only source
of representative information available for the zone is
longitudinal data on aggregate crop production statis￾tics. At present, these data suggest that the progress
noted in the rapid appraisal is not widespread enough
to have made a major impact on the aggregate picture.
Summary statistics (Table 1) on production, yields
and area cultivated in the zone present a picture of
impressive growth in production for most crops but
little growth in yields—i.e., most of the productivity
increases have been realized through area expansion
rather than through intensification and better resource
management.
8
Nevertheless, we have a growing base of rapid ap￾praisal results for approximately 20 villages (includ￾ing the results of two previous trips by Mike McGahuey
to different villages) that all point in the same direc￾tion. We also have the OHVN database showing con￾tinued expansion in cotton production (to which farm￾ers attribute their recent increases in income) and in￾creased adoption of intensification techniques (to which
participant farmers attribute their yield increases). Even
if the rapid appraisal results are not fully representa￾tive of the entire zone, it is clear that important progress
is being made in many villages and that important les￾sons can be learned about (1) what has been driving
the changes, (2) the magnitude of the increased house￾hold income being generated by program participants,
and (3) the expected impact that these changes in in￾come could have on national income if the types of
situations we saw in the rapid appraisals became wide￾spread.
Table 1: Area, Production and Yield Data for the OHVN: 1991/1992–1998/1999
 1991 1992 1993 1994 1995 1996 1997 1998 Trend 
Cotton area (ha) 10506 12201 8624 11692 14605 23158 30750 35816 + 
 prod (tons) 11842 12494 10684 13097 16167 21990 28927 33740 + 
 yield (kg/ha) 1127 1024 1239 1120 1107 950 941 942 – 
Tobacco area (ha) 209 285 331 237 100 83 77 87 – 
 prod (tons) 411 525 549 330 160 133 105 112 – 
 yield (kg/ha) 1971 1842 1661 1392 1600 1579 1853 1874 – 
Millet area (ha) 30906 31516 31892 34188 36660 35732 38149 37422 + 
 prod (tons) 30226 23900 26700 31800 32441 36095 38714 35595 + 
 yield (kg/ha) 978 758 837 930 885 1010 1015 951 stagnant 
Sorghum area (ha) 46603 48334 48140 51213 56009 59431 66390 72572 +
prod (tons) 50508 43911 44622 47904 50292 64638 73047 75901 + 
 yield (kg/ha) 1084 908 927 935 898 1088 1100 1046 stagnant 
Maize area (ha) 11099 11485 11648 12157 12834 13072 14411 15457 + 
 prod (tons) 13845 13110 13938 11214 12929 14594 16814 20033 + 
 yield (kg/ha) 1247 1141 1197 922 1007 1116 1167 1296 stagnant 
Rice area (ha) 4431 4656 4640 5243 5774 6333 7165 8596 + 
 prod (tons) 4679 4553 4420 5194 5033 7188 8184 9941 + 
 yield (kg/ha) 1056 978 953 990 872 1135 1142 1157 stagnant 
Groundnuts area (ha) 12297 12823 13331 13993 16210 16878 20286 23420 + 
 prod (tons) 10889 9415 11807 12473 13896 14488 17962 21773 + 
 yield (kg/ha) 886 734 886 891 857 858 885 930 stagnant 
Fonio area (ha) 749 1153 1084 1115 1344 1391 1271 + 
 prod (tons) 287 476 526 507 652 684 796 + 
 yield (kg/ha) 383 413 485 455 486 492 626 + 
Cowpeas area (ha) 255 312 521 +
prod (tons) 216 165 290 + 
 yield (kg/ha) 842 529 557 – 
Source: OHVN, Septième Session du Conseil d'Administration, Plan de Campagne 1999–2000, pg. 16. 
9
This brings us to the question of what is driving the
progress noted by the rapid appraisals. This progress
is the result of a complex 15- to 20-year process in￾volving multiple efforts by many actors. Nevertheless,
USAID has been a dominant actor, providing OHVN
with an important source of external financing since
the 1980s.2
Important contributions have also come from the Ger￾mans, who are supporting non-governmental organi￾zation (NGO) activities in the Ouélesseboughou sec￾tor. Their anti-erosion program (PAE) focuses on de￾veloping a gestion de terroir approach giving high
priority to improving village-level management of a
community’s natural resources. In addition, there are
an estimated 20–30 NGOs operating in various capaci￾ties in the OHVN (not all in the agriculture or NRM
sector). In other words, the progress is a result of ma￾jor investments in the zone over a long period of time.
Based on information gathered during the rapid ap￾praisals, discussions with USAID and OHVN person￾nel, and documents reviewed, the key ingredients con￾tributing to current progress appear to be:
+ Good identification of technologies capable of
reversing declines in yields
2 A review of the Procès-verbal (OHVN August 1998) showed
USAID annual contributions to the OHVN budget ranging
from $200,000 to $500,000 between 1995/6 and 1998/99,
with a planned increase to $1.3 million for support of the
agribusiness unit of OHVN in 1999/2000.
+ Potential for increased cash income from expan￾sion of cotton production
+ Community approach to implementation
+ Focus on youth
+ Focus on villages/farmers most likely to benefit
from NRM actions
+ Use of demonstration effect through model farm￾ers and model villages
+ Incremental training (literacy, technical skills,
community organization, management skills us￾ing the CLUSA model)
+ Support services offered, including:
‰ Roads
‰ Credit guarantees for limited periods following
management training
‰ Input/output transport assistance
‰ Regular supervision and support to trainees
‰ Some free equipment for implementing
NRM activities
‰ Market research by OHVN to help with
crop diversification
10
11
Expanding the Progress
WHAT TO DO
It is my opinion that the progress seen during the vari￾than from any single or limited number of activities or
investments. Nevertheless, certain components are
more essential than others if farmers are to make the
tices that characterized the zone in the 1970s and 1980s
to the level of commercial agriculture needed to stimu￾late agricultural transformation and generalized eco￾nomic growth:
• There must be a profitable cash crop with reliable
markets and stable prices.
• There must be improved, affordable technologies that
benefit both cash and food crops.
• There must be training programs to equip young farm￾ers with the literacy and management skills they need
to function as effective commercial farmers, both in￾dependently and in associations.
Without these basic ingredients, agricultural transfor￾mation will not take place. The OHVN program—at
least in the villages covered by rapid appraisals—ex￾hibits each one of these key ingredients.
Although the NRM program covers the entire OHVN
zone, it has recognized that farmers are unlikely to
adopt NRM practices if there is not a strong income
incentive. Hence, OHVN’s NRM program began by
targeting sectors where cotton production was already
underway, then expanding into zones where cotton pro￾duction was being introduced. This policy has worked
thus far, but both farmers and the OHVN administra￾tion recognize the need to identify alternative cash
crops for lower rainfall zones where cotton is not fea￾sible, and to reduce the risks of over-reliance on a
single cash crop in zones where cotton is currently
king.
The NRM program staff are to be commended for
their efforts to identify and promote (in collaboration
with Malian researchers) (1) truly effective anti-ero￾sion practices that have proven capable of recovering
highly degraded land and (2) improved methods of
collecting, composting and applying organic fertiliz￾ers. Although there is a long list of different tech￾niques promoted by the NRM department of OHVN,
the data show that it is the anti-erosion techniques
(rock lines and plugs, fascines (gulley plugs), and veg￾etative bands in particular) and the improved manage￾ment of organic matter (compost and manure pits and
use of crop residues) that are the most popular com￾ponents of the program. These techniques, combined
with the use of chemical fertilizers applied to cotton
that is rotated with (largely unfertilized) cereal crops
every 2-3 years, has resulted in substantial yield in￾creases over time for participant farmers (see illus￾trations in Appendix 1).
One of the most impressive components of the OHVN
program is the farmer training introduced by CLUSA
in the early 1990s. The CLUSA approach has a num￾ber of characteristics that make it stand out from other
farmer training programs—the most important being
that the ultimate goal is to empower farmers so they
can handle their own affairs as they make the transi￾tion from semi-subsistence to commercial agriculture.
Given this goal, CLUSA does not set up a training pro￾gram until farmers exhibit some initiative in (1) be￾coming literate in local languages and (2) creating an
association with a well-defined set of goals. At this
point, CLUSA offers training designed to help the
group meet its goals. In the villages visited, the most
common goal for newly formed associations was to
obtain bank credit for agricultural equipment and in￾puts. Our discussions with the many young farmers
the various programs that have been undertaken, rather
transition from the semi-subsistence production prac￾ous rapid appraisal trips results from the synergy of
12
who were managing the association finances and credit
left us with the impression that CLUSA has done an
outstanding job in this respect. Associations are help￾ing individual members prepare loan requests (includ￾ing proof of capacity to repay loans), making decisions
about the creditworthiness of association members,
submitting consolidated loan portfolios for all asso￾ciation members (written in Bambara) directly to local
bank representatives, dealing with several banks at once
(depending on the type of credit sought), negotiating
and contracting with input suppliers, and managing the
loan repayments which have been in the 95–98% range
during the last several years. Once the initial training
program is completed, CLUSA tends to move into the
background—remaining available for consultations
when needed (perhaps to undertake new activities), but
encouraging the associations they have trained to man￾age their own affairs.
If these three key ingredients are in place, I believe
the adoption of NRM practices, and the productivity
increases associated with them, can expand to vil￾lages and zones not yet reached. The presence of sup￾port services will, however, influence the speed of
the expansion. For example, assistance with equip￾ment to transport rocks for anti-erosion structures
appears to be needed by some farmers and associa￾tions, but not by others (depending on proximity of
rock supplies and number of carts already available in
the village). It will be important to carefully evaluate
each situation to avoid unnecessarily raising program
costs and stifling local initiative, while taking into ac￾count situations where a bit of help with rock trans￾port could stimulate an entire series of more produc￾tive activities. Another important support service is
improving rural infrastructure. Poor roads are a ma￾jor obstacle to farmers trying to diversify into pro￾duction and marketing of horticultural products and
to the acquisition of inputs (both problems mentioned
in several of our village discussions). In addition to
farmers, credit is important for farmers, input suppli￾ers and traders purchasing farm production as well.
USAID’s provision of funds to guarantee credit to
farmers’ associations during their first four agricul￾tural seasons may be one of the reasons that bank
representatives are now traveling from village to vil￾lage to deal directly with farmers (I have no evidence
to support this, but it is difficult to believe that the
guarantees did not provide some incentive). There are
still many villages (particularly in the newly established
cotton areas) where associations have not yet been
created and in these places input credit is being man￾aged by OHVN and distributed to individuals (rather
than to associations), with much less favorable re￾payment performance (see the OHVN report: Procès￾verbal de la 6ème session ordinaire du Conseil
d’Administration de l’OHVN, August 1998). Given
the poor performance to date for the individual loans,
providing guarantees for them does not appear to be
the best option. Rather, it appears more appropriate to
move as quickly as possible (without violating the ba￾sic CLUSA principles) through the stages of literacy
training, association creation, and management train￾ing so these villages can catch up with those in the
zones where cotton production is already better es￾tablished. This requires coordination of the training
efforts (now often carried out by Malian NGOs that
were trained by CLUSA) and OHVN/CMDT cotton
promotion/expansion efforts.
PROBLEMS TO RESOLVE
During the course of the rapid appraisal mission and
discussions with OHVN staff, a number of real or po￾tential problems surfaced that could hinder the desired
transition to commercial farming. They are described
briefly below.
Backsliding on development of private sector input
markets. Although progress was made in the mid-1990s
with the privatization of input markets, at present
farmer associations appear to be relying entirely on
OHVN for their cotton inputs. Both farmers and
OHVN reports (e.g., OHVN August 1998) explain that
the apparent backsliding came about because the prices
charged by the private sector distributors were sub￾stantially higher than those prevailing in the nearby
zones managed by the Compagnie Malienne pour le
Développement des Textiles (CMDT), where inputs
were still being provided through CMDT channels.
13
3
 For the 1996/97 campaign, association credit due to OHVN
was reimbursed at 94% while only 88% of individual credit
was reimbursed; taking all outstanding credit into account
associations are at 93% reimbursement while individual
borrowers are at 68%—a substantial difference.
Improvements were noted for the 1997/98 campaign when
individual borrowers repaid 97% of current debts and 92%
of total debts while the associations reimbursed 99% of
the current campaign’s and 98% overall (pg. 11, Proces
Verbal, OHVN August 1998).
This led OHVN farmers to protest the higher prices in
their zone vis-à-vis the CMDT zone. The OHVN re￾sponse was to rebuild their input supply network, re￾lying on CMDT connections to keep costs and prices
at the same level as those prevailing in the CMDT
zone. This is an issue that needs to be addressed at
the level of national policies: Mali needs to develop a
national fertilizer plan based on a thorough analysis of
the pros and cons of the continuing CMDT monopoly
on cotton inputs. This is not a simple issue, as there
tend to be important economies of size and scale as￾sociated with fertilizer imports. The small private sector
operators who attempted to market inputs in the OHVN
area were probably dealing in such small quantities
that they were unable to realize the economies accru￾ing to the CMDT—hence the inability to charge com￾petitive prices.
Continued OHVN financing of credit and high rates
of default for new cotton producers. For the 1998/99
campaign, OHVN financed 62.5% of input credit
(down slightly from 65% in 1997/98), with the Banque
Nationale de Developpement Agricole (BNDA) financ￾ing the rest. I found this information (from the OHVN
August 1998 Procès-verbal) surprising, as the villages
we visited were all getting their credit through the
BNDA. Understanding that the type of credit situa￾tions we saw represent only about one-third of the to￾tal input credit portfolio for the OHVN suggests, per￾haps, that the villages we visited are representative of
about one-third of the OHVN farm population—i.e.,
those that have succeeded in creating viable farmer
associations (not a very scientific way of getting at
representivity, but an interpretation that helps us get
closer to understanding how widespread the situations
we observed might be). This same OHVN report
(Partie Recommandations, pg.17) indicated that de￾faults are a problem in zones where farmer’s associa￾tions are not well-established and OHVN is obliged to
provide credit to individual farmers:
Le crédit individuel a représenté 33.15% du
crédit total accordé par l’ÓHVN. Ce type de
crédit est en nette progression depuis 3
campagnes. Cet état de faits est lié d’une part à
l’inexistence d’organisations paysannes
capables de gérer le crédit collectif dans le
secteur de Faladié et d’autre part à l’extension
de la culture du coton dans la zone de Kolokani
et de Kangaba.
These results are in sharp contrast to the high repay￾ment rates reported by the farmers’ associations we
visited (95–100% were the typical rates cited) and the
higher repayment rates reported by OHVN for asso￾ciation credit.3
 The lower rates for repayment of indi￾vidual credit and the need for OHVN, rather than pri￾vate banks, to provide the credit raises the question of
whether OHVN/CMDT is moving ahead too fast with
their plans to expand cotton areas. Is it a good deci￾sion for OHVN to be offering credit directly to indi￾vidual farmers who are just beginning to produce cot￾ton? How rapidly can these credit responsibilities be
transferred to the banking sector? Is there a role for
USAID credit guarantees in these zones where cotton
is now being introduced?
Decisions about financial or in-kind support for rock
hauling. As noted above, the issue of whether to pro￾vide equipment (carts, tools) for building anti-ero￾sion barriers appears to be one that needs to be evalu￾ated on a case-by-case basis. Too much assistance
(when it is not really needed) can pose problems for
sustainability if farmers become reliant on external
sources of help not only for building but also for
maintaining the anti-erosion structures and for ex￾tending their benefits to more farmers. On the other
hand, when carts are not available and rocks are far
away, building anti-erosion barriers can be an im￾possible task. Some intermediate options might be
providing credit or actually providing the equipment
as a gift to associations on the condition that they
develop a financial plan for replacing the equipment
once fully depreciated.
14
Decisions about which markets to develop for hor￾ticultural crops. USAID is putting a substantial
amount of new funding into the OHVN’s
agribusiness unit, which is charged with the task of
developing new markets for OHVN products. There
have been some signs of progress in developing
export markets. Following the CFA devaluation,
Mali was able to break into the European green bean
market, with most of the exported production com￾ing from the OHVN zone. And there are plans un￾derway to increase sorrel (hibiscus) production for
export to the United States.
Efforts to diversify into cash crops that are either
complements to, or substitutes for, cotton are to be
commended, but the extent to which Mali should
be targeting European and U.S. markets versus other
markets in the West African sub-region needs to be
better evaluated, in view of the serious problems
encountered in the green bean subsector this past
season. In the bean-producing village that we vis￾ited we were shown very large stocks of produce
that had not been picked up by the exporter as speci￾fied in the production contract. Apparently, the ex￾porter had not made adequate provision for the type
of packaging required by his buyers in Europe, so
he was unable to collect the produce from the vil￾lages and ship it on time to France. Although I would
not recommend that the OHVN agribusiness unit
ignore Europe and U.S. markets, I would suggest
that it divide its attention between those markets
(which are characterized by extremely high quality
standards and complicated transport arrangements)
and the markets that are opening up in Mali and
nearby countries such as Ghana, Ivory Coast and
Nigeria. (See INSAH, November 1998, which dis￾cusses the issue of European versus regional ex￾port markets for both horticultural and livestock
products).
Inadequate attention to cereals market development.
As farmers improve productivity, they are increas￾ingly capable of marketing cereals that were previ￾ously produced exclusively for home consump￾tion. Yet traditional views of cereals as a ‘social’
rather than a ‘market’ crop, and limited knowledge
about managing cereal stocks for profit, continue to
hamper cereal market development. OHVN may need
to improve farmers’ marketing skills as well as the
database on cereal production and stocks in the zone.
Given recent efforts of the USAID-funded PASIDMA
program to better estimate regional cereal availability
and encourage trade within Mali as well as the W.
African region, it is recommended that OHVN and
APCAM (which has local associations throughout the
OHVN zone) work together in an effort to improve
cereal marketing efficiency. In the villages visited, many
farmers and associations appear to be holding excess
cereal stocks because they (1) feel prices are too low
and (2) prefer building village cereal banks to hedge
against poor harvests. One association visited had re￾ceived a 9-month line of bank credit based on an 80
F/kg valuation of the associations’ cereal stocks. Us￾ing the line of credit, the association purchased cere￾als from members at 80 f/kg. To make good on the
loan, they need to sell their stocks at more than 80 f/
kg or members will need to buy back their own cere￾als. At the time of our visit they were quite concerned
about their ability to repay the loan, as current prices
were in the 60 F/kg range.
Rapid expansion of livestock herds. Most of the model
farmers visited were enthusiastic adopters of the NRM
themes involving increased use of manure (improved
stables, composting, etc.). With this enthusiasm comes
increased herd size—one farmer had increased his herd
from about 60 to approximately 120 head in about 5
years! As noted elsewhere, our impression is that we
were visiting the better-off farmers, and we do not have
to worry about most farmers owning 120 head of cattle
in the near future. Nevertheless, some thought needs
to be given to the long-term implications (e.g., over￾grazing) of the growth in herd size linked to the inten￾sive use of animal manures.
Need to improve integration and complementarity of
organic and inorganic fertilizers. At some point (sooner
rather than later) farmers will need to start increasing
the use of inorganic fertilizers. At present, inorganic
fertilizers are used almost exclusively on cotton. If
15
cereal and cotton yields are to increase beyond their
current—relatively mediocre—levels, use of inorganic
fertilizers on cereals will no doubt need to be part of
the picture (in addition to improved seed varieties and
continued improvements in management practices).
Finding the optimal combination of organic and inor￾ganic fertilizers for different crops and rotations may
require more research to identify the combinations that
are most efficient from both a private (profit) and a
social (environmental) perspective. During our rapid
appraisal visits, Mike McGahuey asked several differ￾ent farmers to describe how they saw organic and in￾organic fertilizers fitting into their production schemes.
The replies always indicated that farmers viewed the
two as complements rather than substitutes, suggest￾ing that farmers could be encouraged to use more in￾organic fertilizers if they could be convinced that it
would be a profitable investment.
Assuming there is good research evidence that inor￾ganic fertilizer use can be profitable for cereals grown
on fields where erosion has been controlled, OHVN
might want to consider an extension approach that re￾sembles that of Sasakawa Global (SG) 2000. SG 2000
trol plot (current practices) and a half-hectare test
plot (recommended doses of inorganic fertilizers). This
permits farmers to easily make comparisons of yields
for the two technologies and, given the literacy skills
of the OHVN village animateurs—local farmers trained
by extension agents to help other farmers in the area—
it should not be too difficult to make comparisons of
financial returns as well. For this approach to work
well, the farmers need to be closely supervised to make
sure the fertilizers are applied using optimal dates and
techniques. SG 2000 has been working in the millet/
sorghum areas of the Segou and Mopti regions for sev￾eral years, trying to introduce yield-enhancing tech￾nologies such as inorganic fertilizers (including rock
phosphates). Thus far, the evidence suggests that the
inorganic fertilizer is generally not profitable (see
Nubukpo et al. 1999 for a discussion of SG 2000 pro￾grams in the Segou Region). One hypothesis concern￾ing the lack of profitability is that SG 2000—in sharp
contrast to the OHVN program—did not begin with a
focus on improved NRM practices (anti-erosion invest￾ments and improved quality of organic amendments).
Thus, it seems important to invest some resources in
analyzing the potential to profitably use inorganic fer￾tilizers on cereals in the OHVN zone which are grown
on land that has been protected against erosion and
that benefits from increased levels of soil organic mat￾ter.
encourages farmers to cultivate a half-hectare con-
16
17
Although OHVN has made important progress in docu￾menting adoption trends for a wide range of recom￾mended NRM practices, and there is a mounting body
of anecdotal information concerning the positive farm￾level impacts of this adoption, we are still unable to
quantify the income impacts of NRM practices. Data
collection and analysis techniques need to be refined
and expanded if we want to better quantify both the
farm-level and the national-level income impacts of
NRM adoption.
Trying to quantify the impacts of NRM adoption over
a period of almost 20 years—years which were char￾acterized by major changes in the general economic
and political environment— raises numerous questions
concerning the real causes of any impacts measured:
NRM adoption? Economic reform? Political reform?
As noted above, it is impossible to scientifically deter￾mine the relative importance of the multiple factors
that have affected rural incomes in the OHVN zone
during the last 20 years. Nevertheless, a better analy￾sis of what has happened to farm incomes in the OHVN
during the last two decades will help us evaluate the
impact of NRM promotion in combination with all the
other political and economic reforms that have taken
place.
The proposal which follows is designed for incremen￾tal implementation. It starts with recommendations for
small improvements in data collection and analysis that
can be made using existing OHVN resources and
moves on to more costly but scientifically sound meth￾ods of gauging changes in rural incomes.
Six options are described for improving the measure￾ment of income impacts:
1. Improvements in counting and reporting adoption
of better NRM practices.
2. Improvements in reporting OHVN production and
yield statistics.
Better Quantifying Progress Made to Date
3. A case study approach to collecting and analyzing
NRM farm-level income impacts (e.g., impacts on
crop-based, livestock, and nonfarm incomes).
4. Rough extrapolations from case studies to the sec￾tor level.
5. Survey of OHVN farmers using a representative
sample.
6. Development of an ongoing program of monitor￾ing key income and food security indicators.
In addition to measuring changes in income, there are
a number of general environmental indicators that
should also be monitored in order to evaluate the over￾all impact of current crop and livestock production on
soil erosion and forest cover. We may be able to show
substantial increases in income at the farm level, but if
this is accompanied by increased clearing of wood￾lands and forests to accommodate larger numbers of
farmers and cotton fields (Table 1), the income gains
are unlikely to be sustainable over time. Hence, it will
be important to combine the income data with other
sources of information (e.g., aerial or satellite photos)
that show overall trends in land use and the extent to
which conservation efforts are outpacing or being out￾paced by growing enthusiasm for crop and livestock
production.
IMPROVEMENTS IN COUNTING AND
REPORTING ADOPTION
Over the years the OHVN NRM program has collected
statistics on the adoption of various practices or
‘themes’. Table 2 is a summary of what OHVN calls
the ‘physical’ results of their program, updated in De￾cember 1999. It shows the growth (1996–1999) in
physical measures (e.g., meters, hectares, number) of
22 practices promoted by the NRM program.
18
NRM Themes Level of Adoption (units) 
Prior to 1997 1997–1998 1998–1999 1999–2000 Sum
Rock lines (m) 79400 6485 10076 5329 101291 
Branch barriers (m) 18500 780 2011 1574 22865 
Small dikes (m) 38900 1492 775 457 41624 
Vegetative bands (m2
) 8998 1341 4000 3240 17579 
Living fences (m) 127022 12000 11831 9309 160162 
Permanent field markers (ha) 1098 599 846 544 3087 
Protected areas (ha) 450 450 615 750 2265 
Diversionary gullies (n) 1417 625 1171 50 3263 
Firebreaks (m) 5250 1406 615 500 7771 
Controlled land clearing (ha) 140 300 — — 440 
Village-managed forests (n) 1620 35 — — 1655 
Wells (n) 120 13 13 9 155 
Deeping of ponds (n) 68 2 1 2 73 
Improved bottom land (ha) 20 — — — 29 
Village tree nurseries (n) 57 15 5 28 105 
Plants from tree nurseries (n) 178800 13318 14640 45576 252334 
Village woodlots 447 23 19 18 507 
Improved cooking stoves (n) 2340 745 312 323 3720 
Manure pits (n) 2268 265 338 — 2871 
Stables for collecting manure (n) 13608 140 135 — 13883 
Improved animal pens (n) 146 8 — — 154 
Compost pit (n) 1303 399 490 — 2192 
Source: OHVN December 1999 and other OHVN data. Notes: ha= hectares, m = meters, n = number. 
Table 2: Illustrative OHVN Adoption Report: Physical Indicators of NRM Adoption
Although the table tells us nothing about how many
farmers are involved or the income impacts of adop￾tion, it does provide some insights about the relative
popularity of different themes and the extent to which
adoption is growing. Theoretically, this type of infor￾mation could be used to estimate income impacts for
the zone if we were able to estimate an average in￾come impact per unit of physical measure.
Table 3 sheds some light on what the physical adop￾tion statistics mean in terms of participating villages
19
Sector Villages Farms Recovered 
Area (ha) 
Kangaba 53 1529 3027
Bancoumana 57 2335 3221
Ouélessébougou 97 3628 7604
Dangassa 33 534 434
Fouani 110 3295 7264
Kati 70 1787 1303
Faladié 35 951 2274
Koulikoro 73 1358 2075
Sirakorola 79 2220 7656
Total OHVN 607 17637 34858
Source: OHVN 1999 data provided by M. Sylla. 
Table 3: Illustrative OHVN Adoption
Report: Villages, Farmers and
Recovered Area
and farms. It also attempts to evaluate impacts in terms
of hectares of previously degraded land recovered and
number of farms that have moved from shifting culti￾vation to working on fixed plots of land. Four improve￾ments that could be made to these statistics are de￾scribed below.
Report the Percentage of Villages and Farms
Adopting
At present OHVN is counting and reporting the num￾ber of villages and farmers adopting specific practices.
These absolute numbers would be much more useful
if presented along with numbers showing the relative
prevalence of the adoption that has occurred. If 50 of
5,000 farmers in a sector have adopted a theme (1%),
that is much less impressive than knowing that 50 of
100 (50%) have adopted it.
To accomplish this, OHVN needs to standardize how
they count villages and hamlets in their statistics—
some reports reviewed appear to be counting hamlets
as individual villages, while others count the mother
village and all hamlets as a single village. Without stan￾dardization and consistent reporting across time and
in different types of reports, it is difficult to know if
real progress is being made.
The issue of the changing boundaries of the OHVN
also poses problems for interpretation of the growth in
adoption counts and percentages. During the recent
past, two sectors (Banamba and Boro) have been
dropped from the OHVN and one (Faladié) has been
added. There is no ideal solution for dealing with such
changes when preparing statistical reports on changes
over time. Adding new zones where adoption is just
starting or dropping former zones where adoption was
high can give the impression that NRM is taking a big
step backwards if one looks only at the aggregate sta￾tistics for the entire OHVN area. When a time series
of statistics covers a period during which boundary
changes have occurred, there must be clear documen￾tation of when the changes took place and the number
of villages/households that were added or dropped from
the statistics for each year concerned. Without clear
documentation of these changes, comparing aggregate
OHVN data from year to year is clearly inappropriate.
Ideally, statistics should be disaggregated and reported
at the level of the units (sectors, circles, or
arrondissements, for example) that are likely to move
in or out of OHVN coverage.
Present More Detail to Show the Degree of
Adoption by Villages and Farms
At present, a village is counted as participating if only
one farmer adopts just one theme. This is a pretty weak
level of participation, and it is not very informative to
group this village with another village where 90% of
farmers are participating and most have adopted three
or more themes. Similarly, a farm is counted as par￾ticipating if it has adopted only one theme. For ex￾ample, a farm that is using a wood-conserving stove
but has adopted no other NRM theme is not differenti￾ated in these summary statistics from a farmer who
20
has made substantial investments in anti-erosion or
composting themes. Such a high degree of aggrega￾tion makes it harder to evaluate the potential impact of
the NRM program. Appendix 2 presents a more dis￾aggregated format for reporting village and farm level
adoption that would help OHVN better communicate
what is happening in the zone.
Clarify the Definition of ‘Recovered Land’ and
Disaggregate it into Different Categories
A total of almost 35,000 hectares ‘recovered’ (17% of
OHVN cultivated area in 1999) is impressive, but what
does it really mean? Is OHVN reporting the entire area
of a field if a rock line brought back into production a
small corner of the field that was unproductive due to
erosion? Or only the area of the small corner that was
affected? In my opinion, the latter is the preferred
method. What qualifies a field for being classified as
unproductive? No yield at all? The farmers’ qualita￾tive appraisal that the land was getting an unusually
low yield for the crop in question? The extension
agent’s appraisal that yield was below a specified level
for a given crop?
For these numbers to have real meaning, there needs
to be some standardization in classifying ‘recovered’
land. Perhaps OHVN is already using adequate crite￾ria. If so, the definitions need to be better explained in
reports so that the end users of the information grasp
the distinctions. Deciding on the criteria to be used is
more appropriately done by a soil scientist or agrono￾mist than by an economist. Nevertheless, there are cer￾tain elements of information that could facilitate eco￾nomic analysis if they could be taken into account. For
example, making the determination on the basis of
before/after yields for specified levels of technology
(e.g., seed variety, fertilizer and manure applications,
etc.) would contribute to improved economic analysis
of the impacts of NRM activities.
Make Some Effort to Measure ‘Disadoption’
Adoption data are collected annually by OHVN ex￾tension personnel and based on the activities of new
adopters that they supervise and/or observe during each
season. Each year the new adopters are added to the
previous ones to obtain a cumulative level of adoption
by theme, village, and farm. A major exception to this
was a survey conducted in 1999 that attempted to do
an exhaustive inventory of currently practiced NRM
themes (see OHVN December 1999).
Because the focus is on increasing adoption, there is
no year-to-year effort made to take into account cases
of disadoption. For example, if a household purchased
an improved stove but decided not to use it, the house￾hold would remain in the cumulative statistics as an
adopter. Similarly, if a farmer planted some living
fences but they all died and he made no effort to re￾place them, the farmer would still be counted in the
cumulative statistics for adoption. Given limited re￾sources for monitoring, the issue of ‘disadoption’
should not be turned into a major drain on OHVN re￾sources. Nevertheless, OHVN field personnel should
give the issue some consideration and try to develop
low-cost methods of monitoring ‘disadoption’ for those
themes where it is most likely to occur. This monitor￾ing should include some effort to identify the causes
of the ‘disadoption’ so that corrective actions can be
taken.
Improvements in Reporting OHVN Production
and Yield Statistics
The OHVN statistical service conducts surveys every
year to measure area cultivated, estimate the probable
harvest, and report final results for the entire agricul￾tural campaign. These surveys are designed to accu￾rately estimate aggregate production for the zone. More
effort is put into estimating cotton production (much
larger sample of fields per enumeration unit) than for
cereals and other crops because of the need to orga￾nize logistics for collecting and processing cotton.
It is recommended that the NRM service and the
OHVN statistical service examine the possibility of
adding a few additional variables to the annual pro￾duction survey in an effort to better grasp the extent to
which fields covered by the production survey have
benefited from NRM practices. Given the very limited
number of fields evaluated for non-cotton crops, it
would be best to limit this additional data collection to
21
the cotton fields. Since land is rotated from cotton to
cereals and back, collecting the following type of in￾formation on cotton fields only should provide infor￾mation on a representative sample of fields if the data
are collected consistently over at least 3–5 years. The
types of information that would be useful are:
1. Meters of anti-erosion structures (rock lines,
branch barriers, small dikes, vegetative bands) on
the field and dates established.
2. Use of parcellement or mise en défens on the field.
3. Carts of organic matter applied to the field in cur￾rent year.
4. Carts of organic matter applied to the field in pre￾vious year and crop cultivated that year.
5. Number of years since field was left in fallow.
6. Estimate of percent of each field currently suffer￾ing from erosion (particularly important for fields
where no NRM practices are being used).
Because the statistical service’s sample is randomly
selected and representative of the OHVN zone, add￾ing this type of information to the annual survey should
help the NRM service to get a better idea of how wide￾spread the use of these techniques is. It would also
permit them to do some analysis on whether yields for
fields having benefited from different NRM practices
are better, worse, or about the same as those of fields
not benefiting from NRM practices. Note that such
analyses will not permit OHVN to determine the yield
impact of the practices, because there is no way of
controlling for the initial condition of the field prior to
use of NRM practices. For example, it is reasonable to
assume that most of the fields benefiting from anti￾erosion themes were in a state of relatively low pro￾ductivity prior to adoption of the themes. If this is true,
we may find that yields on NRM fields are not any
better than on untreated fields, or even lower. Hence,
while this information cannot be used to evaluate the
contribution of NRM practices to yields, it can help us
better understand the general dynamics of NRM adop￾tion (location, percent of fields, length of use, most
common combinations of practices) and give us some
idea of current yields for a broad, randomly selected
sample.
The NRM program has made a point of focusing on
sectors and villages where certain preconditions favor￾ing NRM adoption exist. Among the criteria used are
the degree of socioeconomic disequilibrium, the re￾ceptivity of the milieu to NRM techniques, and the
demonstrated willingness of local populations to ac￾tively participate in identifying and implementing so￾lutions to their problems (OHVN December 1999). As
a result, NRM adoption is much higher in some sec￾tors of the OHVN (Ouélessébougou and Gouani, for
example) than others (Dangassa or Kati, for example).
This raises the question of the level of disaggregation
permitted by the OHVN sample design. For the pur￾poses of monitoring and evaluating the NRM program,
it would be helpful to be able to get statistically sig￾nificant results at the OHVN sector level. Even this
level of disaggregation remains problematic in some
cases, because, as noted earlier, the boundaries of
OHVN have changed, with some sectors (or parts of
sectors?) being added (e.g., Faladié) or removed (e.g.,
Banamba and Boron) from OHVN responsibility.
CASE STUDY APPROACH TO
COLLECTING AND ANALYZING NRM
FARM-LEVEL INCOME IMPACTS
During our field visits in the OHVN zone we were
presented with several case studies illustrating the
adoption of NRM practices and the corresponding
changes in land use, cropping patterns, yields, live￾stock holdings, and investments in animal traction
equipment (see Appendix 1 for one example). Although
the data presented differed from case to case, there were
some common aspects:
+ a time perspective starting with the first year of
adopting an NRM theme and continuing to present.
+ a list of NRM practices adopted (usually quanti￾fied in terms of meters or hectares per year).
22
+ annual yield and production figures for either (1)
selected NRM fields or (2) an aggregate picture
of all fields for the farm.
+ an inventory of animal traction equipment owned.
+ some information on inputs used each year (carts
of manure, sacks of fertilizer, pesticides and in￾secticides used).
These case studies were presented to us by village
animateurs in the presence of the case-study farmer.
The animateurs represent the final link in the exten￾sion chain. They are members of village associations
who have been selected by association members to
receive special training in NRM practices from OHVN
agents vulgarisateurs de base, or AVB (extension
agents). To become an animateur, one must have suc￾cessfully completed a literacy program. Most village
associations have several animateurs (2–5). Once
trained, the village animateurs help organize work/
training groups to assist individual farmers or groups
wanting to learn about or implement particular themes.
The animateurs’ own fields often serve as the initial
trial sites in the village. Animateurs are encouraged by
OHVN to keep records on participating farmers so that
they can track their progress. To date, there is no stan￾dardized format for this record keeping and no abso￾lute requirement that it be done for all participants.
Nevertheless, one gets the impression that the
animateurs are in possession of a substantial amount
of information that could be used as a starting point
for calculating the income impacts of NRM practices
if it could be transferred from personal notebooks to a
standardized reporting format. OHVN has already used
a couple of case studies in reports and a conference
paper to illustrate the impacts that NRM adoption has
had on selected farmers.
My recommendation is that the OHVN begin their ef￾forts to better quantify the income impacts of NRM
adoption by seeing how many case studies they can
put together from information currently recorded in
the notebooks of animateurs and/or AVBs. Although
we raised the issue of data availability with OHVN
personnel at all levels, no one seemed sure how much
information was currently recorded and how difficult
it would be to get it transferred to some type of stan￾dardized format. Appendix 3 contains some draft ‘ques￾tionnaires’ designed to collect information that is cur￾rently recorded in AVB’s and animateurs’ notebooks.
The questionnaires were drafted while I was in the field
and discussed with OHVN personnel (M. Sylla). My
recommendation is that OHVN do a trial run, filling
in about 10 copies before continuing with a larger num￾ber of cases, because the quality of the data in the first
10 copies needs to be evaluated to see if it is adequate
for calculating income impacts. Appendix 4 contains
an example of the type of calculations one could do if
the data were adequate. OHVN and USAID also need
to evaluate the cost (primarily AVB and animateur
time) of transferring the data to these questionnaires
and decide if doing another 50 to 100 questionnaires
would be desirable and feasible given their current re￾sources.
Rough Extrapolations From Case Studies to the
Sector Level
Getting another 50 to 100 examples of changes in crop￾ping patterns and yields over time would not permit us
to come up with statistically valid estimates of the con￾tribution of NRM to income because we would have
no way of knowing how representative these cases
were, but it would help us to get beyond the ‘anec￾dote’ stage (5–10 case studies) in which we currently
find ourselves. With 50 to 100 examples, we may be
able to say something about typical yield impacts over
time for the most popular themes and then develop
hypotheses about the aggregate impact that these yield
changes would have if more degraded land benefited
from the adoption of these techniques. This would prob￾ably require a small amount of additional consulting
time (5–10 days) from me or another agricultural
economist to develop a set of indicative yield change/
income scenarios based on the data collected and train
OHVN staff so they could do similar analyses in the
future.
Survey of OHVN Farmers Using a Representative
Sample
I am not presently recommending the development of
a stand-alone survey to evaluate income impacts of
23
NRM adoption. This decision is based on the follow￾ing factors:
1. My impression that neither OHVN nor USAID
want to commit the level of resources required.
2. OHVN dissatisfaction with the last major survey
effort in the zone (done in collaboration with the
Institut du Sahel [INSAH]).
3. My belief that it is more important to build OHVN
capacity for regular monitoring.
There is one area, however, that might warrant some
type of survey—the quantifying of income from NRM
themes that are not directly related to crop production.
My terms of reference included the task of identifying
and quantifying NRM-related income-generating ac￾tivities that are not normally captured in aggregate in￾come statistics. My impression is that the current NRM
program in OHVN is not promoting many themes that
would be generate these types of income, so I have not
made any concrete recommendations for trying to quan￾tify these impacts in the short run. Some background
on why I came to this conclusion follows.
Activities are underway in many villages to transfer
management of local forests from the forest service to
village associations. In all cases encountered during
our field visits, the objective was to manage the for￾ests for conservation purposes—permitting harvesting
only for personal use of village members. Although
one could place a value on the personal consumption,
this is not likely to represent a major contribution to
local or national income at the present time. In some
villages visited, karité (shea butter) harvesting and pro￾cessing was a major income-generating activity for
women. The NRM program does not have any themes
that relate directly to karité production per se, so valu￾ing this production to measure the contribution of the
NRM program to household and national income does
not appear justified at the present time. The value of
the karité harvested and processed should be consid￾ered in national accounts; I have not been able to con￾firm whether it is (I suspect that karité exports may
already be taken into account).
Perhaps the most likely NRM activity to be included
in this category is the establishment of woodlots by
both villages and individuals. During our fieldwork,
we saw a number of woodlots planted in teak for pro￾duction of construction poles. Most had been recently
planted and were not yet generating income. As the
currently planted woodlots mature and the total num￾ber of woodlots increases, OHVN should develop some
method for monitoring consumption and sales so that
the contribution of these woodlots to household, vil￾lage and national income can be taken into account.
At present—based on what was observed in villages
visited—it seems premature to put much effort into
quantifying woodlot incomes.
There are undoubtedly a number of other forest prod￾ucts that are gathered, processed, and sold (condiments,
herbal teas, medicines) by rural households in the
OHVN zone. It was difficult, however, to get a feel
for the importance of these incomes relative to income
from cropping and livestock activities. The focus of
the OHVN/NRM program has clearly been the pro￾motion of anti-erosion and soil fertility techniques. The
groups of farmers we met with spoke enthusiastically
about how NRM adoption had affected crop and live￾stock production practices and incomes but never men￾tioned any impact on other types of income. This could
be an omission on their part (and mine, for I did not
raise the issue). Had we been speaking with women,
we might have had more discussion of such incomes,
as they are more likely than men to gather and sell
forest products. Given the general lack of NRM themes
related to generating income from forestry products,
however, I suspect that the OHVN/NRM program has
not had much of an impact on the level of incomes
generated from these activities. If this is true, expend￾ing OHVN/NRM resources in an effort to quantify
these incomes is probably not warranted at present. As
more and more villages assume the responsibility for
managing their forests and OHVN assists with the de￾velopment of management plans, it may be important
to evaluate the extent to which villages or individuals
are able to increase the income generated from the
forest’s renewable resources.
24
DEVELOPMENT OF AN ONGOING
PROGRAM OF MONITORING KEY
INCOME AND FOOD SECURITY
INDICATORS
Household income growth is an important indicator of
program success for most of USAID/Bamako’s
projects. Unfortunately, the task of monitoring income
growth is so daunting that USAID staff and project
personnel usually opt for monitoring less informative
but easier-to-collect indicators. This has clearly been
the case with the OHVN project.
I am recommending that USAID/Bamako look into the
possibility of using some promising new methods for
income monitoring that were developed by MSU as
part of a USAID-funded project in Mozambique. At
present, the methods are also being tested in Kenya
(again, with USAID funding). There is an initial cost
in using these methods (see below) that is easier to
justify if it is applied to monitoring a large number of
diverse projects rather than to a single project such as
the NRM component of the OHVN program. Hence I
recommend that this type of monitoring be considered
by USAID rather than by the OHVN.
A comprehensive document describing the methods
used in Mozambique is available (Tschirley, Rose and
Marrule, 2000). An excerpt of a few pages from the
report is attached in Appendix 5 to give readers a bet￾ter idea of what this type of monitoring can do and the
level of survey work required. This income-proxy
method provides the possibility of obtaining regular
(for example, yearly) information on household income
without performing cumbersome quantitative surveys
each time.
In brief, the method requires conducting an initial sur￾vey to collect detailed information on a wide range of
both income and potential proxy variables (this is the
most costly part). These detailed data can then be used
to create econometric models that estimate total house￾hold income and permit analysts to identify appropri￾ate proxy variables. Data for the smaller set of proxy
variables are then collected in subsequent surveys and
used to monitor changes in income over time. Two
models were developed for Mozambique. The more
detailed model uses 40 variables to estimate both total
income and the amount of income earned in 10 sepa￾rate income categories; the less detailed model uses
16 variables to estimate total household income. In the
Mozambique case, this initial survey was funded by
USAID and conducted collaboratively by MSU and a
number of NGOs working on USAID projects, many
of which required some type of income monitoring to
satisfy USAID reporting requirements. In Mali, it might
be possible to use the upcoming budget/consumption
study or some other major survey now in the planning
stages as a base to which the proxy work can be added
rather than fund an entire survey.
If USAID decided to move in this direction, the issue
of monitoring income from forest products (see above)
could probably be incorporated into the initial surveys
and proxy variables—as could other project-specific
interests.
25
In summing up, I would like to reiterate that I was very
impressed with what I saw during the four days I vis￾ited OHVN farmers and farmers’ associations currently
working with NRM themes. The farmers were among
the most knowledgeable, motivated, and enthusiastic
farmers that I have met during the many years that I
have been working in the Sahel and elsewhere in Af￾rica. In a qualitative sense, I am very comfortable stat￾ing that the farmers visited have clearly improved their
food security and incomes because they adopted NRM
practices at a time when a wide range of policy changes
and sectoral investments made it particularly profit￾able to do so.
The limitation of this type of rapid appraisal is that I
cannot say anything concrete about how representa￾tive the farmers with whom we met are. Nor can I say
anything quantitative about the size of the income im￾Closing Remarks
pacts stimulated by the NRM program at either the
farm or the national level. These are two very impor￾tant types of information that both USAID and OHVN
need to gather in order to evaluate where they are and
what they need to be doing to further expand the
benefits of NRM practices. Implementation of the rec￾ommendations in “Better Quantifying Progress” sec￾tion of this paper should bring us all much closer to
understanding what is really happening with respect
to NRM in the zone.
Given that there appear to be a number of very useful
lessons to be learned from the OHVN experience, it
seems important to me that both USAID and OHVN
invest some resources in (1) improving their ability to
quantify the size and extent of the income impacts
stimulated by the NRM program and (2) documenting
and publicizing the OHVN story so that others in Mali
as well as elsewhere may benefit from the experi￾ence.
26
27
References Consulted
Bingen, R.J. and B. Simpson. 1995. “Technology
Transfer and Agricultural Development in
West Africa.” In Technology Transfer and
Public Policy, edited by Y.S. Lee. Westport,
CN: Quorum Books.
Crawford, E., and V. Kelly. November 2001. Evalu￾ating Measures to Improve Agricultural In￾put Use, Department of Agriculture Econom￾ics Staff Paper No. 01-55. E. Lansing, MI:
Michigan State University.
Crosson, P. and J.R. Anderson. March 1995. Achiev￾ing a Sustainable Agricultural System in Sub￾Saharan Africa. Building Blocks for Africa
2025, paper no. 2. Washington, D.C.: The
World Bank.
Hagen, R. 1993. NRM Timeline and Issues for West
Africa. Unpublished draft document.
Hijkoop, J., P. van der Poel, and B. Kaya. 1991. Une
lutte de longue haleine: Aménagements anti￾érosifs et gestion de terroir. Bamako and
Amsterdam: Institut d’Economie Rurale and
Institut Royal des Tropiques.
Institut du Sahel (INSAH). November 1998. Food
Security and Agricultural Subsectors in West
Africa: Future prospects and key issues four
years after the devaluation of the CFA franc.
A set of policy briefs on livestock, cotton,
horticulture, and food consumption prepared
for a CILSS Policy Conference, November
30–December 2, 1998. Bamako: Institut du
Sahel.
Maiga, A.S., B. Teme, B.S. Coulibaly, L. Diarra, A.O.
Kergna, and K. Tigana. September 1994.
Etude: Ajustement structurel et développement
durable—Cas du Mali. Bamako and London:
Institut d’Economie Rurale and Overseas De￾velopment Institute.
McGahuey, M. December 1998. Developing a Meth￾odology and Baseline Data for Monitoring
and Evaluation of Natural Resources Man￾agement and Environmental Impact of USAID
Interventions in Mali. Washington, D.C.:
USAID (AFR/SD/ANRE/NRM) trip report.
Nubukpo, K., V. Kelly, M. Yade, and M. Galiba. Au￾gust 2000. Accelerating Agricultural Intensi￾fication in the Riskier Environments of Sub￾Saharan Africa: A Case Study of the Sasakawa
Global 2000 Program in Mali. Select Paper
presented at the International Association of
Agricultural Economists meetings in Berlin.
OHVN. May 1992. Contribution de L’O.H.V.N. à la
Gestion des Ressources naturelles dans sa
Zone d’Intervention. Report of an annual semi￾nar held 12–14 May 1992.
OHVN. September 1995. Evaluation du Programme
de Gestion des Ressources Naturelles en Zone
OHVN. Bamako: OHVN.
OHVN. August 1998. Procès-verbal de la 6ème Ses￾sion Ordinaire du Conseil d’Administration
de l’Office de la Haute Vallée du Niger and
other documentation discussed at the Septième
Session du Conseil d’Administration de
L’OHVN. (This appears to be a set of docu￾ments given to participants at the 7th meeting
of the OHVN Administrative Council that met
in 1999. I have used August 1998 as the date
because that is what appears on the first docu￾ment in the set—the minutes of the 6th Ses￾sion.) Bamako: OHVN.
28
OHVN. December 1999. Communication de l’Office
de la Haute Vallée du Niger à l’Atelier
Régional sur les Expériences de la Gestion
des Ressources Naturelles: Evolution et Per￾spective. Koudougou, Burkina Faso.
Rose, D. and D. Tschirley. January 2000. A Simpli￾fied Method for Assessing Dietary Adequacy
in Mozambique. Research Report No. 36.
Maputo: Ministry of Agriculture and Fisher￾ies. Full document is available on the Michi￾gan State Web Site (www.aec.msu.edu/
agecon/fs2).
Ruben, R. and D.R. Lee. March 2000. Combining In￾ternal and External Inputs for Sustainable
Intensification. IFPRI 2020 Brief 65. Wash￾ington, D.C.: International Food Policy Re￾search Institute.
Sanders, J., D. Southgate, and J.G. Lee. December
1995.The Economics of Soil Degradation:
Technological Change and Policy Alterna￾tives. Soil Management Support Services
Technical Monograph No. 22. W. Lafayette,
IN: Purdue University.
Sanogo, A., B.S. Coulibaly, M.K. Ndiaye, M.Sidibe,
B. Teme, M. Togola. April 1993. Etude des
phénomènes de dégradation des terres au
Mali: Esquisse d’un programme national de
conservation et de restauration. Bamako:
Consulting report done for FAO.
Scherr, S. and S. Yadav. May 1996. Land Degrada￾tion in the Developing World: Implications for
Food, Agriculture, and the Environment to
2020. Food, Agriculture, and the Environment
Discussion Paper 14. Washington, D.C.: In￾ternational Food Policy Research Institute.
Shaikh, A., E. Arnould, K. Christophersen, R. Hagen,
J. Tabor, P. Warshall. 1988. Opportunities for
Sustained Development: Successful Natural
Resources Management in the Sahel, Volumes
1–3. Washington, D.C.: e/di (a member of
International Resources Group).
Speirs, M. and O. Olsen.1992. Indigenous Integrated
Farming Systems in the Sahel. World Bank
Technical Paper Number 179, Africa Techni￾cal Department Series. Washington, D.C.: The
World Bank.
Tefft, J. 2000. “Cotton in Mali: The ‘White
Revolution’and Development,” in Democracy
and Development in Mali. Edited by R.J.
Bingen, J. Staatz and D. Robinson. E. Lan￾sing, MI: Michigan State University Press.
Tschirley, D., D. Rose, and H. Marrule. February 2000.
A Methodology for Estimating Household In￾come in Rural Mozambique Using Easy-to￾Collect Proxy Variables. Research Report No.
38. Maputo: Ministry of Agriculture and Fish￾eries. Full document is available on the Michi￾gan State University Web site
(www.aec.msu.edu/agecon/fs2).
USAID/Bamako. various dates. Several documents,
reports, proposals, etc. concerning the OHVN
projects that were provided by Mike
McGahuey.
Weight, D. and V. Kelly. 1999. Fertilizer Impacts on
Soils and Crops of Sub-Saharan Africa. MSU
International Development Paper No. 21. E.
Lansing, MI: Michigan State University. Full
document is available on the Michigan State
University Web site (www.aec.msu.edu/
agecon/fs2).
Yanggen, D. December 1995. Understanding Small
Farmer Adoption of Natural Resource Con￾servation Techniques in the Sahel: A Concep￾tual Framework with an Application to the
Cotton Growing Zone of Koutiala, Mali. Un￾published mimeo.
Yates, R.A. and A. Kiss. January 1992. Using and Sus￾taining Africa’s Soil. Summary of the pro￾ceedings of a World Bank Seminar held in
Washington, D.C. in January 1992.Agricul￾ture and Rural Development Series N. 6, Tech￾nical Department, Africa Region, World Bank.
Washington, D.C.:World Bank.
29
OHVN Case Study of Production and Income Changes for
A Farmer Having Used NRM Practices During Nine Years
CASE STUDY TABLES FROM OHVN DECEMBER 1999
Introduction. The following pages contain an example of an OHVN case study taken from an OHVN confer￾ence paper (OHVN December 1999). Although it tells the story of only one farmer, it shows a good understand￾ing of the types of data that must be collected for a large number of farmers if OHVN is to do a more thorough
job of reporting on zone-level impacts of NRM adoption.
A number of improvements could be made in the economic analysis. Among the more important would be (1)
accounting for differences between the “with adoption” and “without adoption” scenarios; (2) accounting for
year-to-year changes in production and prices for the economic analysis (rather than a simple comparison of first
and most recent years); and (3) using real prices (nominal prices deflated by an index such as the consumer price
index) that reflect seasonal and interannual price risk (the current analysis uses a single price across all years to
value output). Note that Appendix 4 uses data for the same farmer, but with some changes in the method of
calculating benefits introduced.
Farmer: Masiamé COULIBALY, Bini village, Gouani Rural Development Sector (SDR)
 Themes used by farmer:
„ bandes enherbées (vegetative bands)
„ végétalisation
„ labour perpen-diculaire à la pente (contour plowing)
„ grattage à sec (light hoeing before rains)
„ utilisation fumure organique (manure use)
„ parcellement et piquets verts (marking long-term field borders)
„ labour de fin cycle (end-of-season plowing)
„ contour rock lines
Appendix 1
30
Table A.1.1 
Evolution of area, yields and production for Masimé Coulibaly 
1990/91–1998/99 
Crops Agricultural season 
1990/ 
1991 
1991/ 
1992 
1992/ 
1993 
1993/ 
1994 
1994/ 
1995 
1995/ 
1996 
1996/ 
1997 
1997/ 
1998 
1998/ 
1999 
Average 
Area 
 Area (ha) 2 3 2 2 3 — 4 2 4 2.7 
Millet Yield 
(tons/ha) 
0.8 0.95 1 1 1.2 — 1.28 1.3 1.3 
 Production 
(T) 
1.6 2.85 2 2 3.6 — 5.1 2.6 5.2 
 Area (ha) 4 3 5 6 3 5 3 4 4 4.1 
Sorghum Yield 
(tons/ha) 
0.95 1.2 1.43 1.63 1.7 1.8 1.8 1.8 1.85 
 Production 
(T) 
3.8 3.6 7.25 9.78 5.1 9 5.4 7.2 7.4 
 Area (ha) 3 4 3 4.5 4 4 4 3.5 4.5 3.8 
Maize Yield 
(tons/ha) 
1.25 1.3 1.8 1.95 2 2.15 2.4 3.05 3.2 
 Production 
(T) 
3.75 5.2 5.4 8.78 8 8.6 9.6 10.7 4.4 
 Area (ha) 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.75 0.75 
Rice Yield 
(tons/ha) 
0.85 0.9 0.9 0.95 0.95 1 1 1.02 1.2 
 Production 
(T) 
0.6 0.63 0.63 0.66 0.66 0.7 0.7 0.77 0.9 
 Area (ha) 0.5 0.7 0.7 0.7 0.5 0.5 0.5 0.5 0.75 0.6 
Peanut Yield 
(tons/ha) 
0.6 0.7 0.75 0.75 0.8 0.8 0.85 0.7 0.58 
 Production 
(T) 
0.3 0.49 0.53 0.53 0.4 0.4 0.42 0.35 1 
 Area (ha) 10 10 10 6 10 11 9 9 11 9.5 
Cotton Yield 
(tons/ha) 
1.41 1.55 0.93 2.59 1.52 1.52 2.11 2.11 1.41 
 Production 
(T) 
14.1 15.5 9.93 15.56 15.2 16.7 19 19 14.1 
 Area (ha) 0.5 0.75 1 0.5 0.5 0.5 0.5 1.25 1 0.7 
Cowpea Yield 
(tons/ha) 
0.3 0.4 0.5 0.5 0.5 0.55 0.6 0.6 0.5 
 Production 
(T) 
0.15 0.3 0.5 0.25 0.25 0.28 0.3 0.75 0.5 
Total area 20.7 22.2 22.4 23.4 24.7 24.7 24.7 23 28 
31
Table A.1.2: 
Analysis of Changes in Masimé Coulibaly’s Yields Between 1991 and 1999
 Years Yield differences Observations 
Crops 1990–91 
kg/ha 
1998–99 
kg/ha 
Yield 
kg/ha 
Percentage 
change 
Millet 
Sorghum 
Maize 
Rice 
Cowpeas 
Peanuts 
Cotton 
800 
950 
1,250 
850 
300 
600 
1,410 
1,300 
1,850 
3,200 
120 
500 
780 
1,414 
500 
900 
1,950 
350 
200 
180 
4 
62.5 
94.7 
156 
41 
66.6 
30 
0.003 
49.8 (1997–
1998) 
For cotton: low 
yield in 1999, but 
if 1990–91 is 
compared to 
1997–98, the 
yield difference 
was 703 kg. 
Table A.1.3: 
Economic Analysis of Masimé Coulibaly’s Farm 
Crops Mean 
area 
(ha) 
1990 
yield 
kg/ha 
1998 
yield 
kg/ha 
1990 
Prod 
(T) 
1998 
Prod 
(T) 
Value (FCFA) Yield 
increase 
(kg) 
Income 
increase 
(FCFA) 
Millet 2.7 800 1,300 1,600 5,200 112,000 364,000 325 252,000 
Sorghum 4.1 950 1,850 3,800 7,400 266,000 518,000 195 252,000 
Maize 3.8 1,250 3,200 3,750 4,400 262,500 308,000 117 45,500 
Cotton 9.5 1,410 1,410 14,101 14,101 1,198,585 1,198,585 0 0 
Peanut 0.7 850 1,200 600 900 72,000 108,000 150 36,000 
Rice 0.6 600 780 300 1,000 45,000 150,000 333 105,000 
Note: Constant prices used: Cereals 70 F/kg 
 Cotton 85 F/kg 
 Peanuts 120 F/kg 
 Rice 150 F/kg 
32
33
Appendix 2
Suggested Format for Periodic Reporting
 of NRM Village and Farm Adoption
in the OHVN
34
35
Appendix 3
Draft Questionnaires for Collecting Data From
AVB/ANIMATEUR Notebooks
36
SECTEUR: VILLAGE: AVB: AV: CHEF DE L'EXPLOITATION: NO. EXP.
 (HECTARES)
BOEUFS
ANNÉE HOMMES FEMMES HOMMES FEMMES HOMMES FEMMES HOMMES FEMMES
LABOUR
BOVINS
ANES
OVINS 
CAPRINS
CULTIVABLES
CULTIVÉES
COMMENTAIRES SUPPLÉMENTAIRES:
1991/1992
1992/1993
1993/1994
1994/1995
1995/1996
1996/1997
1997/1998
1998/1999
1999/2000
MAIN D'OEUVRE 
HIVERNAGE CONTRA-SAISON CONTRA-SAISON
AGRICOLE ACTIF AGRICOLE ACTIF
SUPERFICIES CHEPTEL MAIN D'OEUVRE MIGRATION
POPULATION, CHEPTEL, ET SUPERFICIES
EVOLUTION DES RESOURCES DE L'EXPLOITATION: 
DE L'EXPLOITATION
POPULATION TOTAL
37
SECTEUR: VILLAGE: AVB: AV: CHEF DE L'EXPLOITATION: NO. EXP.
COMMENTAIRES SUPPLÉMENTAIRES:
TRACTEUR
EVOLUTION DES RESOURCES DE L'EXPLOITATION: 
NOMBRE DE CHAQUE TYPE D'EQUIPEMENT FONCTIONNELAPPARTENANT A L'EXPLOITATION
ANNÉE
EQUIPMENTS
APPAREIL 
TRAITEMEN T
CHARRETTE
HANDY
AUTRES:
AUTRES:
TM 
CHARRUE
MULTI￾CULTURE
SEMOIR
1999/2000
1998/1999
1997/1998
1996/1997
1995/1996
1994/1995
1993/1994
1992/1993
1991/1992
1990/1991
38
SECTEUR: VILLAGE: AVB: AV: CHEF DE L'EXPLOITATION: NO. EXP.
DIGUETTES
PARES FEU
AUTRES:
PRATIQUES 
THÈMES PRATIQUÉES
CULTURALES
JACHERENOMBRE DE CHARRETTE UTILISÉ
FASCINES
AUTRES:
HAIES VIVESENGRAIS MINERALE FUMURE ORGANIQUE NOMBRE DE FOSSE (KILOGRAMMES)
LIGNES EN 
CAILLOUX
BARRIERE EN 
CAILLOUX
BANDES 
ENHERBÉS
PARCELLE￾MENT
COMPOSTE
FUMIER
COMPOSTE
AUTRE
 CUMULE DES HECTARES 
RECUPERÉS 
 FUMIER AVEC 
PAILLE
FUMIER SANS 
PAILLE
NPK ANNÉE
(METRES)
(METRES)
(METRES)
(HECTARES)
(HECTARES)
(HECTARES)
(METRES)
(METRES)
(METRES)
(METRES)
COMMENTAIRES SUPPL
1990/1991
1991/1992
1992/1993
1993/1994
1994/1995
1995/1996
1996/1997
1997/1998
1998/1999
1999/2000
ÉMENTAIRES:
DIGUETTES
PARES FEU
AUTRES:
PRATIQUES 
CULTURALES
JACHERE
FASCINES
AUTRES:
HAIES VIVES
LIGNES EN 
CAILLOUX
BARRIERE EN 
CAILLOUX
BANDES 
ENHERBÉS
PARCELLE￾MENT
COMPOSTE
FUMIER
COMPOSTE
AUTRE
 CUMULE DES HECTARES 
RECUPERÉS 
 FUMIER AVEC 
PAILLE
FUMIER SANS 
PAILLE
NPK
39
SECTEUR: VILLAGE: AVB: AV: CHEF DE L'EXPLOITATION: NO. EXP.
SUPER
PROD
REND
SUPER
PROD
REND
SUPER
PROD
REND
SUPER
PROD
REND
SUPER
PROD
REND
SUPER
PROD
REND
SUPER
PROD
REND
(HA) MT (KG) (HA) MT (KG) (HA) MT (KG) (HA) MT (KG) (HA) MT (KG) (HA) MT (KG) (HA) MT (KG)
99/00 OUI
NON
98/99 OUI
NON
97/98 OUI
NON
96/97 OUI
NON
95/96 OUI
NON
94/95 OUI
NON
93/94 OUI
NON
92/93 OUI
NON
91/92 OUI
NON
90/91 OUI
NON
COMMENTAIRES SUPPLÉMENTAIRES:
T TRAITEMEN
* LES CHAMPS SANS TRAITEMENT SONT CEUX QUI N'ONT PAS BENEFICIÉ DE PRATIQUES ANTI-EROSIVES ET AUXQUELS ON N'A APPLIQUÉ NI FUMURE NI ENGRAIS POUR AU MOINS 
TROIS CAMPAGNES.
M I L S O R G H O C O T O N A R A C H I D E N I E B E M A I S A U T R E
ANNÉE
EVOLUTION DE LA PRODUCTION
HIVERNAGE
40
SECTEUR: VILLAGE: AVB: AV: CHEF DE L'EXPLOITATION: NO. EXP.
SUPER
CULTURE 1: CULTURE 2: CULTURE 3: CULTURE 4: CULTURE 5: CULTURE 6: CULTURE 7:
PROD RENDSUPERPROD RENDSUPERPROD RENDSUPERPROD RENDSUPERPROD RENDSUPERPROD RENDSUPERPROD REND
(HA) (MT) (KG) (HA) (MT) (KG) (HA) (MT) (KG) (HA) (MT) (KG) (HA) (MT) (KG) (HA) (MT) (KG) (HA) (MT) (KG)
NON
OUI
NON
OUI
NON
OUI
96/97 OUI
NON
OUI
NON
OUI
NON
OUI
NON
OUI
NON
OUI
NON
OUI
NON
EVOLUTION DE LA PRODUCTION
CONTRE-SAISON
TRAITEMENT*
ANNÉE
REMPLIR LES NOMS DES CULTURES DANS LES CASES CI-DESSOUS
90/91
91/92
92/93
93/94
94/95
95/96
97/98
98/99
99/00
41
Appendix 4
The production data used in this appendix are taken
from the case-study farmer presented in Appendix 2.
The point of this appendix is not to do a full-blown
analysis of the net increases in income realized by farm￾ers adopting NRM practices, but to illustrate a number
of things that could be done to improve the analyses
currently done by OHVN. The tables below illustrate
three changes that OHVN could easily make in the
way they do their financial assessments of adoption.
(1) The first table quantifies the yields for a “without
project” scenario rather than simply comparing yields
in the initial starting year with current yields. A com￾parison of a with and without scenario does a better
job of showing the full extent of yield differences that
can be attributed to adoption of NRM practices. In the
example that follows I assume a rate of decline in yields
over time due to erosion and nutrient depletion that
approximates that shown in aggregate national yield
statistics for Mali.
(2) The second and third tables use both the with/with￾out scenario and two different price scenario to cap￾ture the potential impact of price instability on income.
The illustration values the nine-year cumulated differ￾ences in yields between the case-study farmer and the
without project scenario using both a favorable and
unfavorable producer price (prices for the illustration
were arbitrarily selected but reflect recent reality).
A more appropriate method would be to value the yield
difference for each year using the average price dur￾ing the harvest season (unfavorable scenario) at a ma￾jor OHVN market and the average price during the
hungry season (favorable price scenario), converted
to real terms using a price index. By using actual prices,
corrected for inflation, we get a better picture of how
Illustration of Budget Analysis Possible Using
the Types of Data in Appendix 3
price instability (which is generally high in Africa) af￾fects the value of agricultural production. I did not have
adequate time to get the price data needed for this type
of analysis during my visit to Mali, but the market data
available in Mali is adequate for this type of valuation.
(3) The fourth table adds an additional consideration—
the time value of income. The table uses the favorable
price scenario of the preceding table, but discounts the
stream of income using a 10% discount rate to obtain
a present value (PV) of the stream of annual incre￾ments to income obtained by the farmer adopting NRM
practices. This type of analysis takes into account the
likelihood that farmers place a greater value on present
than on future income. Doing this type of analysis tends
to reduce the benefits a farmer might realize from in￾vesting in NRM because the yield/income differences
tend to be larger toward the end of the nine years than
at the beginning of the period.
A major shortcoming of the analyses presented in these
four tables is that they do not account for differences
in farm-level costs between the with and the without
project scenario. If we are able to get more complete
information on levels of inputs used each year by par￾ticipant and nonparticipant farmers and the costs of
constructing some of the anti-erosion structures (see
Appendix 3 for details on types of data needed), a more
thorough analysis could be undertaken using a stan￾dard benefit/cost framework.
This type of framework has recently been applied to
an analysis of the use of Tilemsi rock phosphates in
Mali (IFDC 1999). If we are able to get at least 10
cases of the questionnaires recommended in Appen￾dix 3 filled in, some effort should be made to use them
in a benefit/cost framework similar to that used by
42
IFDC. A recent MSU staff paper (Crawford and Kelly,
2001) provides useful guidelines on how a simple
benefit/cost framework can be applied to analysis of
projects promoting input use and/or NRM practices
that have both private income and public environ￾mental impacts.
43
Table A.4.1
Illustration of NRM yield-increasing potential during 9-year period
Farmer: Masiamé Coulibaly
Summary of yield data (kg/ha)
1990/1 1991/2 1992/3 1993/4 1994/5 1995/6 l996/7 1997/8 1998/9
Millet
 Production/ha
 with NRM 800 950 1000 1000 1200 1275 1300 1300 8825 1103
 without NRM* 800 800 800 800 800 800 800 800 800 6400 800
 NRM millet increases/ha 150 200 200 400 475 500 500 2425 346
Cotton
 Production/ha
 with NRM 1410 1555 930 2590 1520 1520 2110 2113 1410 15158 1684
 without NRM** 1410 1368 1327 1287 1248 1211 1174 1139 1105 10058 1252
 NRM cotton increases/ha 187 -397 1303 272 309 936 974 305 3889 486
NRM yield increases for Rotation A : Millet/cotton
Millet 0 200 400 475 500 1575 315
Cotton 187 1303 309 974 2773 693
NRM yield increases for Rotation B : Cotton/millet
Cotton 0 -397 272 309 936 305 1425 237
Millet 150 200 500 850 283
Notes: Farmer didn't produce millet in 1995 so cotton yield difference is substituted in rotation B
* assumes yields are stagnant ** average annual yield decline of 3%
9-Year 
Totals
Annual 
Averages
44
Illustration of gross income-increasing potential of NRM adoption during a 9-year period:
Table A.4.2
Millet Cotton
Unfavorable price scenario: 
Summary of gross income data (FCFA/ha)
Farmer: Masiamé Coulibaly F/kg 70 F/kg 130
1990/1 1991/2 1992/3 1993/4 1994/5 1995/6 l996/7 1997/8 1998/9
 NRM millet increases/h
 without NRM* 56000 56000 56000 56000 56000 56000 56000 56000 56000 448000 56000
 with NRM 56000 66500 70000 70000 84000 0 89250 91000 91000 617750 77219
 Production/ha
Millet
a 10500 14000 14000 28000 0 33250 35000 35000 169750 24250
 NRM cotton increases/h
 without NRM** 183300 177801 172467 167293 162274 157406 152684 148103 143660 1307582 162776
 with NRM 183300 202150 120900 336700 197600 197600 274300 274690 183300 1970540 218949
 Production/ha
Cotton
a 24349 -51567 169407 35326 40194 121616 126587 39640 505552 63194
Rotation Total
Cotton 24349 169407 40194 126587 360537 90134
Millet 0 14000 28000 33250 35000 110250 22050
NRM income increases for Rotation A : Millet/cotton
0 24349 14000 169407 28000 40194 33250 126587 35000 470787 112184
Rotation Total
Millet 10500 14000 35000 59500 19833
Cotton 0 -51567 35326 40194 121616 39640 185209 30868
NRM income increases for Rotation B : Cotton/millet
0 10500 -51567 14000 35326 40194 121616 35000 39640 244709 50701
* assumes yields are stagnant ** average annual yield decline of 3%
Notes: Farmer didn't produce millet in 1995 so cotton income is substituted in rotation B
Totals
9-Year 
Averages
Annual 
45
Table A.4.3
Illustration of gross income-increasing potential of NRM adoption during a 9-year period: Favorable price scenario: 
Millet Cotton
Farmer: Masiamé Coulibaly F/kg 90 F/kg 150
Summary of gross income data (FCFA/ha)
1990/1 1991/2 1992/3 1993/4 1994/5 1995/6 l996/7 1997/8 1998/9
Millet
 Production/ha
 with NRM 72000 85500 90000 90000 108000 0 114750 117000 117000 794250 99281
 without NRM* 72000 72000 72000 72000 72000 72000 72000 72000 72000 576000 72000
 NRM millet increases/ha 13500 18000 18000 36000 0 42750 45000 45000 218250 31179
Cotton
 Production/ha
 with NRM 211500 233250 139500 388500 228000 228000 316500 316950 211500 2273700 252633
 without NRM** 211500 205155 199000 193030 187239 181622 176174 170888 165762 1508749 187819
 NRM cotton increases/ha 28095 -59500 195470 40761 46378 140326 146062 45738 583329 72916
NRM income increases for Rotation A : Millet/cotton
Millet 0 18000 36000 42750 45000 141750 28350
Cotton 28095 195470 46378 146062 416004 104001
Rotation Total 0 28095 18000 195470 36000 46378 42750 146062 45000 557754 132351
NRM income increases for Rotation B : Cotton/millet
Cotton 0 -59500 40761 46378 140326 45738 213703 35617
Millet 13500 18000 45000 76500 25500
Rotation Total 0 13500 -59500 18000 40761 46378 140326 45000 45738 290203 61117
Notes: Farmer didn't produce millet in 1995 so cotton income is substituted in rotation B
* assumes yields are stagnant ** average annual yield decline of 3%
9-Year 
Totals
Total 
Averages
46
Present value of gross income-increasing potential of NRM adoption during 9 years:
Table A.4.4
Millet Cotton
Favorable price scenario: 
Farmer: Masiamé Coulibaly F/kg 90 F/kg 150
Summary of gross income data (FCFA/ha)
1990/1 1991/2 1992/3 1993/4 1994/5 1995/6 l996/7 1997/8 1998/9
 NRM income increases/ha
 without NRM* 72000 72000 72000 72000 72000 72000 72000 72000 72000 576000 72000
 with NRM 72000 85500 90000 90000 108000 0 114750 117000 117000 794250 99281
 Production/ha
Millet
13500 18000 18000 36000 0 42750 45000 45000 218250 31179
 NRM income increases/ha
 without NRM** 211500 205155 199000 193030 187239 181622 176174 170888 165762 1508749 187819
 with NRM 211500 233250 139500 388500 228000 228000 316500 316950 211500 2273700 252633
 Production/ha
Cotton
28095 -59500 195470 40761 46378 140326 146062 45738 583329 72916
Rotation Total
Cotton 28095 195470 46378 146062 416004 104001
Millet 0 18000 36000 42750 45000 141750 28350
NRM income increases for Rotation A : Millet/cotton
0 28095 18000 195470 36000 46378 42750 146062 45000 557754 132351
Rotation Total
Millet 13500 18000 45000 76500 25500
Cotton 0 -59500 40761 46378 140326 45738 213703 35617
NRM income increases for Rotation B : Cotton/millet
0 13500 -59500 18000 40761 46378 140326 45000 45738 290203 61117
* assumes yields are stagnant ** average annual yield decline of 3% ***PV discount rate is 10%
Notes: Farmer didn't produce millet in 1995 so cotton income is substituted in rotation B
PV Rotation B 142636 fcfa/ha 190
PV Rotation A*** 327944 fcfa/ha man day equivalent 437
Totals
9-Year 
Averages
Annual 
47
Contents, Foreword and Introduction from
A Methodology for Estimating Household Income in Rural
Mozambique Using Easy-To-Collect Proxy Variables
Directorate of Economics
Ministry of Agriculture and Fisheries
Research Report No. 38, February 2000
Maputo, Mozambique
by
David Tschirley
Donald Rose
Htigino Marrule
Appendix 5
48
Table of Contents
Foreword Adapting INCPROX and INCPROX Lite to Other Data Sets
I. Introduction
II. Development of the Proxy Methodology
A. Data Collection and Processing
i. Sample Design
ii. Questionnaire Design
B. INCPROX: A Structural Approach to Estimating Income
C. INCPROX Lite: A Simpler Alternative
D. Statistical Results and Confidence Intervals
III. Performance of INCPROX and INCPROX Lite Across Zones
IV. Using INCPROX and INCPROX Lite
A. Conducting the Proxy Survey
B. Developing the Proxy Estimate of Household Income
Annex A Prices Used in Valuing Agricultural Production
Annex B Results of INCPROX Component Regressions
Annex C Goodness of Fit and Standard Errors of the Estimate for INCPROX and
INCPROX Lite
Annex D Complete INCPROX Ranking Performance Results
Annex E Sampling Guidelines for Income Proxy Surveys
Annex F INCPROX and INCPROX Lite Questionnaires
Annex G INCPROX and INCPROX Lite Manuals (Spreadsheet Version)
Annex H Procedures for Using SPSS/Windows to Generate INCPROX Estimates
of Income and Income Components
49
This report is a slightly modified version of a report
originally prepared for use by USAID-funded NGOs
in Mozambique in developing household income esti￾mates for evaluation of their programs and reporting
to USAID. Readers interested in the income proxy
methodologies but not specifically in Mozambique
might skip section II.A (Data Collection and Process￾ing), as it contains primarily information very specific
to Mozambique.
The methodologies reported on here represent a gen￾eral approach applied to specific circumstances. The
approach described in section II.B (INCPROX: A
Structural Approach to Estimating Income) and II.C.
(INCPROX Lite: A Simpler Alternative) could be ap￾plied in other countries or in other geographical areas
of Mozambique, but would need to be adapted to those
circumstances. Adapting INCPROX or INCPROX Lite
to other areas would involve:
1. Collecting or gaining access to an existing house￾hold level data set that contains all the data needed
to (a) directly calculate income for each house￾hold, and (b) develop income proxy variables for
each household similar to those utilized in this re￾port.
2. Utilizing regression techniques to develop
INCPROX or INCPROX Lite models based upon
this data set.
3. Developing standard procedures for (a) collecting
the proxy variables and (b) converting those proxy
variables into estimates of household income and
income components.
Income-expenditure surveys are done in many devel￾oping countries on a regular basis, for example every
three- to four years. Thus, one wishing to develop and
utilize these income proxy methodologies would typi￾Foreword
Adapting INCPROX and INCPROX Lite to Other Data Sets
cally not need to collect a data set specifically for that
purpose; work could focus on developing the models
and the standard procedures for utilizing the models
to obtain income estimates. Once these models and
procedures are developed, various organizations can
collect a much reduced set of simple proxy variables
on a regular basis (for example, yearly), and easily pro￾duce estimates of household income and income com￾ponents. These organizations do not need sophisticated
research capabilities, but do need access either in-house
or through consultants to data collection and manage￾ment skills typical of monitoring & evaluation opera￾tions.
Two key issues would benefit from further research.
First, how well do the models perform over time? The
value of these approaches as cost effective monitoring
tools is predicated on the income estimates they gen￾erate being acceptably accurate over the course of sev￾eral years (e.g., 2-4 years). If the models are robust
over such a time period, then a rich set of monitoring
information—household income and its structure—can
be tracked regularly without the burdensome, complex,
and costly work of collecting and processing income￾expenditure data sets.4
4
 These models are based on objective measures of the intensity
of a household’s involvement in each economic activity,
and on the productive resources the household had available
to dedicate to those activities. These simple proxy variables
are complemented by quantitative measures of the
production of two key crops–maize and cotton. Thus, this
approach should, in theory, be reasonably sensitive to
changes in weather (proxied by the production of maize
and cotton), in a household’s portfolio of economic
activities ( proxied by the intensity variables), and in the
quantity of productive resources available to the household
(proxied by production function variables). Factors not
accounted for in these models which could affect income
include changing relative prices, and pest or other
production problems which affect a crop other than maize
50
or cotton. Changes in the productivity of the household’s
productive assets will also affect income; these are partially
accounted for by the quantitative estimates of maize and
cotton production, holding constant the household’s
productive assets. The actual success of the approach in
controlling for all these factors is, of course, an empirical
issue requiring further analysis.
In Mozambique, the lack of comparable data sets sepa￾rated in time has not permitted testing the temporal
durability of these models. A country with comparable
income-expenditure data sets separated by 2-4 years
would be an ideal candidate for such research.
Second, how can the models better deal with chang￾ing relative prices? Agriculture is a key component of
income for most rural households in developing coun￾tries. Prices of agricultural commodities change every
year, often in unexpected ways, and these price changes
will affect income. Like the issue of temporal durabil￾ity, developing an approach to deal effectively with
changing relative prices requires comparable data sets
separated in time (since relative prices will in all like￾lihood be different for each data set).
Section I of the paper provides a brief introduction.
Section II reviews the work that was done to develop
the models in Mozambique, and presents basic statis￾tical results. Section III evaluates the performance of
the models over space within the research area, and
Section IV is a guide to NGOs on how to use the mod￾els–how to collect the proxy variables and develop the
income estimates. In all these sections, much of the
detail is in Annexes.
I. Introduction
This report outlines a method for estimating house￾hold income in rural areas of Mozambique using a
proxy approach. It is based on collaborative work be￾tween Michigan State University and USAID-funded
NGOs, and is meant for use by them in their areas of
operation.
The development of such a methodology prompts two
important questions. First, why focus on household
income? Second, why use a proxy approach?
An important overall development goal for
Mozambique is the reduction of poverty and improve￾ment in the incomes and well being of rural house￾holds. Thus, measurement of household income is a
logical choice for monitoring the effects of policies
and programs oriented towards accomplishing this
goal. To be sure, there are other measures of house￾hold well being. For example, some economists have
argued that welfare levels are more appropriately de￾termined by measuring household consumption expen￾ditures, in part because of the extensive data collec￾tion activities needed to accurately assess household
income. But, since so much of consumption in
Mozambique is from own production, accurately mea￾suring consumption in practice may be no easier than
measuring income.
Income is difficult to measure in rural settings of de￾veloping countries, in part because there are so many
different sources of income. Households in
Mozambique earn income from the production and sale
of seven different food staples, such as maize or man￾ioc, seven different cash crops, like cotton or tobacco,
and 20 different fruits and vegetables. In addition, in￾come is obtained from the production and sale of live￾stock, from fishing, from wage labor, and from any of
over three dozen different microenterprise activities,
such as the weaving of baskets or the production and
sale of alcoholic beverages. Thus, surveys attempting
to measure household income need to ask questions
on all of these activities and collect quantitative infor￾mation on each.
In addition to the sheer number of sources of income,
each of these sources presents different methodologi￾cal challenges. For example, to get information on in￾come from the production of maize, one needs to know
how much maize was produced. This involves getting
the farmer to remember how many bags or cans of
which size were obtained from the harvest as well as
the state of the maize, dried or fresh, on the cob or in
grain. Conversion factors are needed for the size of
the bag or can, and density factors are needed for the
state of the maize. While all this is doable for one or
two crops, it becomes very time-consuming and ex￾pensive when done for the vast array of crops that are
51
grown in Mozambique. The expense in human and
other resources is beyond the capacity of all but dedi￾cated research projects.
An income-proxy methodology provides the possibil￾ity of obtaining regular (for example, yearly) informa￾tion on household income without performing cum￾bersome quantitative surveys each time. This report
outlines the development and use of such a method￾ology.
52
53
SD Technical Papers
Office of Sustainable Development
Bureau for Africa
U.S. Agency for International Development
The series includes the following publications:
 1 Framework for Selection of Priority Research and Analysis Topics in Private Health Sector Devel￾opment in Africa
*2 Proceedings of the USAID Natural Resources Management and Environmental Policy
Conference: Banjul, The Gambia/January 18–22, 1994
*3 Agricultural Research in Africa: A Review of USAID Strategies and Experience
*4 Regionalization of Research in West and Central Africa: A Synthesis of Workshop Findings and
Recommendations (Banjul, The Gambia. March 14–16, 1994)
*5 Developments in Potato Research in Central Africa
*6 Maize Research Impact in Africa: The Obscured Revolution/Summary Report
*7 Maize Research Impact in Africa: The Obscured Revolution/Complete Report
*8 Urban Maize Meal Consumption Patterns: Strategies for Improving Food Access for Vulnerable House￾holds in Kenya
*9 Targeting Assistance to the Poor and Food Insecure: A Literature Review
 10 An Analysis of USAID Programs to Improve Equity in Malawi and Ghana’s Education Systems
*11 Understanding Linkages Among Food Availability, Access, Consumption and Nutrition in
Africa: Empirical Findings and Issues From the Literature
*12 Market-Oriented Strategies Improve Household Access to Food: Experiences From Sub￾Saharan Africa
 13 Overview of USAID Basic Education Programs in Sub-Saharan Africa II
 14 Basic Education in Africa: USAID’s Approach to Sustainable Reform in the 1990s
 15 Community-Based Primary Education: Lessons Learned from the Basic Education Expansion Project
(BEEP) in Mali
 16 Budgetary Impact of Non-Project Assistance in the Education Sector: A Review of Benin, Ghana,
Guinea and Malawi
*17 GIS Technology Transfer: An Ecological Approach—Final Report
*18 Environmental Guidelines for Small-Scale Activities in Africa: Environmentally Sound Design for
Planning and Implementing Humanitarian and Development Activities
*19 Comparative Analysis of Economic Reform and Structural Adjustment Programs in Eastern Africa
*20 Comparative Analysis of Economic Reform and Structural Adjustment Programs in Eastern Africa/
Annex
*21 Comparative Transportation Cost in East Africa: Executive Summary
*22 Comparative Transportation Cost in East Africa: Final Report
*23 Comparative Analysis of Structural Adjustment Programs in Southern Africa: With Emphasis on
Agriculture and Trade
*24 Endowments in Africa: A Discussion of Issues for Using Alternative Funding Mechanisms to Sup￾port Agricultural and Natural Resources Management Programs
*25 Effects of Market Reform on Access to Food by Low-Income Households: Evidence From Four
Countries in Eastern and Southern Africa
54
*26 Promoting Farm Investment for Sustainable Intensification of African Agriculture
*27 Improving the Measurement and Analysis of African Agricultural Productivity: Promoting
Complementarities Between Micro and Macro Data
*28 Promoting Food Security in Rwanda Through Sustainable Agricultural Productivity
*29 Methodologies for Estimating Informal Cross-Border Trade in Eastern and Southern Africa
*30 A Guide to the Gender Dimension of Environment and Natural Resources Management: Based on
Sample Review of USAID NRM Projects in Africa
*31 A Selected Bibliography on Gender in Environment and Natural Resources: With Emphasis on
Africa
*32 Comparative Cost of Production Analysis in East Africa: Implications for Competitiveness and Com￾parative Advantage
*33 Analysis of Policy Reform and Structural Adjustment Programs in Malawi: With Emphasis on Agri￾culture and Trade
*34 Structural Adjustment and Agricultural Reform in South Africa
*35 Policy Reforms and Structural Adjustment in Zambia: The Case of Agriculture and Trade
*36 Analysis of Policy Reform and Structural Adjustment Programs in Zimbabwe: With Emphasis on
Agriculture and Trade
 37 The Control of Dysentery in Africa: Overview, Recommendations and Checklists
 38 Collaborative Programs in Primary Education, Health and Nutrition: Report on the
Proceedings of a Collaborative Meeting, Washington, D.C., May 7–8, 1996
*39 Trends in Real Food Prices in Six Sub-Saharan African Countries
*40 Cash Crop and Foodgrain Productivity in Senegal: Historical View, New Survey Evidence and
Policy Implications
 41 Schools Are Places for Girls Too: Creating an Environment of Validation
*42 Bilateral Donor Agencies and the Environment: Pest and Pesticide Management
*43 Commercialization of Research and Technology
*44 Basic Guide to Using Debt Conversions
*45 Considerations of Wildlife Resources and Land Use in Chad
 46 Report on the Basic Education Workshop: Brits, South Africa, July 20–25, 1996
 47 Education Reform Support—Volume One: Overview and Bibliography
 48 Education Reform Support—Volume Two: Foundations of the Approach
 49 Education Reform Support—Volume Three: A Framework for Making It Happen
 50 Education Reform Support—Volume Four: Tools and Techniques
 51 Education Reform Support—Volume Five: Strategy Development and Project Design
 52 Education Reform Support—Volume Six: Evaluating Education Reform Support
*53 Checkoffs: New Approaches to Funding Research, Development and Conservation
 54 Educating Girls in Sub-Saharan Africa: USAID’s Approach and Lessons for Donors
 55 Early Intervention: HIV/AIDS Programs for School-Aged Youth
 56 Kids, Schools and Learning—African Success Stories: A Retrospective Study of USAID
Support to Basic Education in Africa
*57 Proceedings of the Workshop on Commercialization and Transfer of Agricultural Technology in
Africa
*58 Informal Cross-Border Trade Between Kenya and Uganda: Proceedings of a Workshop Held at the
Mayfair Hotel, Nairobi Kenya, December 6, 1996
*59 Unrecorded Cross-Border Trade Between Kenya and Uganda
*60 The Northern Tier Countries of the Greater Horn of Africa
55
*61 The Northern Tier Countries of the Greater Horn of Africa: Executive Summary
 62 Determinants of Educational Achievement and Attainment in Africa: Findings From Nine Case
Studies
 63 Enrollment in Primary Education and Cognitive Achievement in Egypt, Change and
Determinants
 64 School Quality and Educational Outcomes in South Africa
 65 Household Schooling Decisions in Tanzania
 66 Increasing School Quantity vs. Quality in Kenya: Impact on Children From Low- and High-In￾come Households
 67 Textbooks, Class Size and Test Scores: Evidence From a Prospective Evaluation in Kenya
 68 An Evaluation of Village-Based Schools in Mangochi, Malawi
 69 An Evaluation of Save the Children’s Community Schools Project in Kolondieba, Mali
 70 An Evaluation of the Aga Khan Foundation’s School Improvement Program in Kisumu, Kenya
 71 An Assessment of the Community Education Fund in Tanzania, Pretest Phase
*72 Stimulating Indigenous Agribusiness Development in the Northern Communal Areas of Namibia: A
Concept Paper
*73 Stimulating Indigenous Agribusiness Development in Zimbabwe: A Concept Paper
 74 How Do Teachers Use Textbooks? A Review of the Research Literature
*75 Determinants of Farm Productivity in Africa: A Synthesis of Four Case Studies
 76 Phoenix Rising: Success Stories About Basic Education Reform in Sub-Saharan Africa
*77 Agricultural Technology Development and Transfer in Africa: Impacts Achieved and Lessons Learned
*78 Innovative Approaches to Agribusiness Development in Sub-Saharan Africa—Volume 1:
Summary, Conclusions and Cross-Cutting Findings
*79 Innovative Approaches to Agribusiness Development in Sub-Saharan Africa—Volume 2:
Secondary Research Findings
*80 Innovative Approaches to Agribusiness Development in Sub-Saharan Africa—Volume 3:
East Africa
*81 Innovative Approaches to Agribusiness Development in Sub-Saharan Africa—Volume 4:
West Africa
*82 Innovative Approaches to Agribusiness Development in Sub-Saharan Africa—Volume 5:
Southern Africa
*83 Agribusiness Development in Sub-Saharan Africa: Optimal Strategies and Structures
 84 USAID’s Strategic Framework for Basic Education in Africa
*85 The Road to Financial Sustainability: How Managers, Government, and Donors in Africa Can
Create a Legacy of Viable Public and Non-Profit Organizations
 86 Girls’ Participatory Learning Activities in the Classroom Environment (GirlsPLACE)
 87 The Demand for Primary Schooling in Rural Ethiopia
*88 Unrecorded Cross-Border Trade Between Mozambique and Her Neighbors: Implications for Food
Security
*89 Unrecorded Cross-Border Trade Between Tanzania and Her Neighbors: Implications for Food Se￾curity
*90 Unrecorded Cross-Border Trade Between Malawi and Neighboring Countries
 91 Children’s Health and Nutrition as Educational Issues: A Case Study of the Ghana
Partnership for Child Development’s Intervention Research in the Volta Region of Ghana
*92 Malawi’s Environmental Monitoring Program: A Model That Merits Replication?
56
*93 Comparative Economic Advantage in Agricultural Trade and Production in Malawi
*94 Regional Agricultural Trade and Changing Comparative Advantage in South Africa
 95 Where Policy Hits the Ground: Policy Implementation Processes in Malawi and Namibia
*96 Thoughts About USAID’s Reforms: Perspectives From the Center for Naval Analysis
 97 A Comparative Cost Study of Community Schools in Mali
*99 Comparative Economic Advantage of Crop Production in Zimbabwe
*100 Analyzing Comparative Advantage of Agricultural Production and Trade Options in Southern Af￾rica: Guidelines
*102 Analysis of Comparative Economic Advantage of Alternative Agricultural Production in
Tanzania
*103 Comparative Economic Advantage of Alternative Agricultural Production Options in Swaziland
*104 Comparative Economic Advantage of Alternative Agricultural Production Activities in Zambia
*105 Transaction Costs Analysis of Maize and Cotton Marketing in Zambia and Tanzania
 106 Overview of USAID Basic Education Programs in Sub-Saharan Africa III
*107 Analysis of Comparative Advantage and Agricultural Trade in Mozambique
 108 Retrospective Study of Basic Education and Skills Training (BEST) USAID Assistance to
 Sector Reform in Zimbabwe
 109 Paradigm Lost? Implementation of Basic Education Reforms in Sub-Saharan Africa: Case
Studies of Benin, Ethiopia, Guinea, Malawi and Uganda
 110 From Information to Action: Tools for Improving Community Participation in Education
 111 Lessons From School-Based Environmental Education Programs in Three African Countries
*112 Agricultural Research in Africa and the Sustainable Financing Initiative: Review, Lessons
and Proposed Next Steps
* 113 Proceedings of the Investment Opportunities Workshop for U.S. and African Manufacturers
and Traders in Wood/Wood Products: The Case of Ghana
* 114 An Inventory of Agricultural Biotechnology for the Eastern and Central Africa Region
 115 Tips for Developing Life Skills Curricula for HIV Prevention Among African Youth: A
Synthesis of Emerging Lessons
 116 Helping Children Outgrow War
* Produced and disseminated under contract to USAID/AFR/SD by The Mitchell Group (TMG). For cop￾ies or information, contact Publications Specialist, at TMG, 1325 G Street, NW, Suite 400, Washington,
D.C. 20005. Phone: (202) 219-0456. Fax: (202) 219-0518. Titles indicated with an asterisk can be accessed
on the World Wide Web at http://www.afr-sd.org/pub.htm.
57
U.S. Agency for International Development
Bureau for Africa
Office of Sustainable Development
Division of Economic Growth, Environment
and Agriculture
1325 G Street, N.W., Suite 400
Washington, D.C. 20005-3104
Measuring the Impacts of Natural Resource Management Activities in Mali's Upper Niger Valley December 2003