THE FUTURE IMPACT OF RICE RESEARCH IN THE SENEGAL...
THE FUTURE IMPACT OF RICE RESEARCH
IN THE SENEGAL RIVER VALLEY
Monica G. Fisher
A. Abdoulaye Fall
Mamadou Sidibe
ISRA/BAME
September 1995
ACRONYMS
BAKE - Bureau dOAnalyses Macro-economiques
CIRAD - Centre de Cooperation Internationale en Recherche
Agronomique Pour le Developpement (France)
CNCAS - Caisse Nationale de Credit Agricole du Senegal
National agricultural credit institution (25% state
owned).
Created 1987-88 as part of the NAP to take
over the credit responsibilities of parastatal
agencies.
CSA - Comzissariat & la Securite Alimentaire
Xaintains a food security stock of 60,000 tons of rice,
purchases local cereals for distribution to deficit
areas and collects and publishes producer and consumer
prices for cereals.
CPSP - Caisse de Perequation et de Stabilisation des Prix
Parastatal which purchases domestic paddy (from SAED)
and imported rice and in turn sells it to licensed
private Lraders for distributicn.
DISEM - Division de Semence
General mission is to control q:zlity of seeds.
Kain
decision body for seed certification.
Provides
information about seeds to farmers, PWJA, SAED and
private multipliers.
Grants pezission for seed
multiplication and evaluates seeds.
GIE - Groupements d'Interet Economique
GIE's are producer unions established during the time
of the NAP (1984).
Producers organize into groups of
up to 50-100 individuals in order to make access to
credit easier (from CNCAS) as well as to reduce
production risks.
GOS - Government of Senegal
INGER - International Network for Genetic Evaluation of Rice
IRRI - International Rice Research Institute
ISRA - Institut Senegalais de Recherches hgricoles (1974)
NAP - New Agricultural Policy (1984)
ORSTOM - Institut Francais de Recherche Scientifique Pour de
Developpment en Cooperation
PAS - Projet Autonome Semencier
National seed multiplier.
Created g/15/90 to take over
seed multiplication duties of SAED during
implementation o-f rice sector privatization.
The
project is schedclled to end in 1995 at which time seed
multiplication activities will be completely
privatized.
PASA - Agricultural Sector Adjustment Program
Multi-donor program which has as its major objective
the elimination of distortions caused by the existing
rice sector policies.
PDRG - Plan Directeur de Developpement Integre Pour la Rive
Gauche de la Vallee du Fleuve Senegal (1992-2017)
PNVA - Programe Nationale de Vulgarisation Agricole
National extension program with World Bank funding.
Currently working with SAED on extension in the Senegal
River Valley.
SAED - Societe d'Amenagement et d'Exploitation des Terres du
Delta
Parastatal established in 1965 to construct irrigation
systems, provide extension services, distribute
production inputs, process grain and marke-, outputs in
the Senegal River Valley.
Although SAED's
responsibilities include all crops in practice it has
concentrated on rice.
URIC -
Unite autonome Rizeries Industrielles et Commerciales
Milling department of SAED in charge of purchasing
paddy from farmers to be turned over to CFSP.
WfiRDA - West African Rice Development Association
Intergovernmental research association formed in 1571
with a mandate to conduct rice research, training and
communications in West Africa.
I.
Introduction
Irrigated rice production systems were first introduced to the
Sahel in the 1920s.
Since that time, rice has become an important
staple crop throughout the zone and its importance continues to
urow.
From the mid-1960s to the mid-1980s, total rice production
in the Sahel grew -4% annually, while consumption increased 5%.'
The gan between domestic supply and demand has meant that Sahelian
countr-ies have become increasingly dependent on rice imports at a
time
when
these
nations
are
seeking to
attain food self-
sufficiency.
Indeed, between 1967 and 1988 consurnptionzof rice and
wheat imports increased an estimated 225% in Senegal.
The New Agricultural Policy of 1984 (NAP) and the Cereals Policy of
1986 articulated the Government of Senegal's (GOS) goal of 80% food
grain self-sufficiency by the year 2000.
Among the major emphases
of the policy statements is the expansion of irrigated production,
particularly irrigated rice production, in the Senegal River Valley
(SRV) .
The purpose of
the
investment is
to promote
the
substitution of locally produced cereals (rice, maize, sorghum and
millet) for imported rice, and decrease the production variability
associated with rainfed cropping systems.
Irrigated rice production is very costly.
Thus there is a clear
need for high-yielding rice
cultivars
not
only to meet the
increasing demand for rice,
but also to justify the high land
development and irrigation costs of the SRV scheme.
New cultivars
should also be early-maturing to allow for rice-rice double
cropping.
In addition, with the on-going liberalization of the
rice sector, there will be an increase in farmer demand for rice
cultivars,
adaptable to the specific physical,
biological and
socio-economic conditions of the Valley.
Today about 90% of all farmers in the Valley use only one or two
rice cultivars.
Medium-duration Jaya and short-duration I Kong Pao
(IKP), introduced from Asia in the 197Os, are still the most common
varieties despite requests from farmers for alternatives.' The IR-
1529 cultivar was released to farmers in the late 1980s.
Its
performance in terms of yield, growth cycle and grain quality is
almost as good as Jaya, but it has not been widely adopted mainly
because of inade uate
%
seed multiplication services and lack of
extension effort.
1 WARDA AMU! Report 1991, p. 70.
2 L.eaphvt Stqhmie. 1994, p. 3.
3 WARLI.4 Annual Report 1991. p. 74.
4 From cawusat~on at Disan md WARDA
The Sahel 108 and 201 were introducd from Asia and Sahel 202 from
Nigeria; and all have been screened, characterized, and introduced
in the SR'J. These three new rice cultivars iiere officially released
in November 1994.
This study estimates the potential impact cf research on the Sahel
cultivars.
Agricultural research has generally been found to have
high pay-offs, but there is a lack of infoc&ation on the returns to
research in Africa. Estimating, ex-ante, the impact of the Sahels
and identification of the constraints to their introduction and
adoption will be an extremely valuable exercise.
The results will
help determine whether rice research efforts are conducive to the
national goal of cereals self-sufficiency and aid in the allocation
of funds for future resesarch.
The report will complement other
BAME impact studies on millet, sorghum, maize, and cotton.
II.
Objectives
The general objective is to conduct an ex-ante evaluation of the
expected benefits and costs of rice research in the Valley.
The specific objectives include:
1 . Quantify the economic benefits to producers and consumers of
rice research.
2 . Kith the inclusion of data on research and extension costs,
calculate the internal rate of return tc the combined research-
extension efforts.
III.
Rice Consumption and Demand for Rice in Senegal
Although vast regional differences exist,
rice is an important
source of consumption in Senegal.
It is the basis of the major
Senegalese dish "Thiebou Dienne"
(literally,
fish and rice),
consumed in large quantities, particularly for the noOn time meal.
Nationally
rice
accounts
for 34% of
the
volume of
cereal
consumpticn,
making up 54% and 24% of the volume of cereals
consumed among urban and rural populations, respectively.
Food expenditures of urban and rural houVVI
c=holds are dominated by
rice across Senegal.
Recently it was est imated that the rural
population spends 25% of its food expenditures on rice, while the
urban pop*>lation spends 18% of its food budget on rice for home
consumpticn.'
Per capita consumption of rice in Senegal has fluctuated through
time, particularly during the drought years of the late 1970s to
2
early 1980s . During this period there was a massive substitution
O f
imported rice for local cereals.
As rice production and
consumption patterns did not return to pre-drought levels, imported
rice has served an important role in supplying ihe food needs
required by the 309 per annum population growth rate.
A 1992 TriARDA
study of rice consum?,?tion
in five Sahelian countries (Burkina Faso,
Mali, Mauritania, Niger and Senegal) found population growth and
rising consumption per capita to account for 40% and 60% of
the
ccnsumption increases, respectively.
"Regression analysis showed
that structural factors,
particularly urbanization, rather than
changing prices or income levels underlie the increased importance
of rice in the region's food consumption patterns."7
Ease of
preparation is the main reason rice is particularly popular in
urban areas.
Unless domestic rice production increases rapidly, rising demand
for rice will lead to increased dependency on imports, which is not
in line with GOS goals.
For this reason, the introduction of new
rice cultivars with high yields and short duration will become
increasingly important.
IV.
Rice Production in Senegal
The two major rice-growing areas in the country are 1) the Delta
and the Senegal River Valley (St. Louis to Bake11 and 2) the
Casamance Region, each with very different systems of production.
Each region has potential to increase productivity with new
technologies.
The Casamance Regicn in southern Senegal accounts for 61% of all
land planted to rice nationally, yet the region only contributes
29% of the national production.
Yields generally average only 1
ton/hectare due to the more traditional production system as well
as water constraints (mainly the upland system).
The principle
inputs are labor, land and seed, and farmers use little fertilizer,
pesticides or farm equipment.
Most of the rice in the area is
grown under rainfed and swamp production systems, mainly for home
consumption purposes.
The Senegal River Valley consists of four delegations:" Bakei in
the region of Tambacounda, Matam, Podor and Dagana in the region of
St. Louis.
Relative to the Casamance, rice grown in the Valle;l is
based on intensive and technologically advanced production methods,
6 World Bank Development
Report 1993.
7 WARDA Annual Report 19%. p. 40.
8 Kite, Rod, 1993. p. 2
9 Delegation is the SAED loid extension of&x.
with 67% of total production being marketed,
55% through the
official system and 12% through the parallel market.'"
The majority
of the active rural population is involved in irrigated rice
production.
The region accounts for a pproximately 89% of national
cash revenue from rice, and 41% of all agricultural cash revenue."
Most of the production in the Valley takes place in the Delta area
which accounts for about 62 % of all irrigated crop land and 71% of
paddy production.'2
Farms in the Dag ana delegation are larger in
size and use more mechanized inputs relative to the farms from
Podor to Bakel.
In general, production in Dagana is primarily
marketed whereas from Podor-Bake1 rice production is mainly for
home consumption.
Irrigated production quadrupled for the Senegal River Valley
between 1979-81 and 1989-91.
Most of the growth may be attributed
to expansion of area under cultivation, however productivity also
increased-l3
On average, the last decade (1982-1993) has seen an
increase of 1,976 ha/year of irrigated crop land in the Valley.'"
Yields have also increased in the Valley,
from an average of 3.3
t/ha (1979-81) to 5 t/ha (1989-91)."
Rice-rice double cropping is
practiced only on a limited basis.
Irrigation opened up the
possibility for two rice growing seascns in the Valley, but for a
variety of reasons (seasonal labor shortages, inadequate knowledge
of how to combine crop,
soil and water management,
access to
credit) cropping intensity continues to be low, rarely exceeding
1.=
The FA0 estimates that reaching the GOS's objective of cereals
self-sufficiency will require an annual increase of 3,800 hectares
of irrigated land, paddy rice yields of 6 tons per hectare, and a
coefficient of cropping intensity of 1.8."
Increased production in the Valley has involved and will continue
to involve high costs.
Irrigation system development in the Valley
has received the largest share of the GOS budget devoted to the
agricultural sector.
Likewise, the subsidies on production and
processing have been high.
In 1989 the total subsidy (irrigation,
1
0
S~dtbe. Mamadou. 1993. p. 7
11 Kite. Rod. 1993. p. 2
1’* S4ED. Feb. 1994, statls[1cs apply to 199Y93. p. 7.
1 3 U’ARLIA Annual Report 1992. Table 18. p. 40.
I4 SAED doctanents. 1992.1993.
1 5 WARDA A~ual Report 1992. p. 40.
1 6 Senegal Ag-imbal Sector Analysis. p. 72
1 7 Martin. 1988. p. 82.
producti .on and processing)
for local pa3 dy production was 88.4
CFA/kg,
which exceeded the CIF (cost
and freight) price for
imported rice of 85 CFA/kg." All costs and subsidies included it
L
costs approximately 300 CFA/kg to produce, process and market rice
in the Valley which in turn has been sold at a fixed retail p.rice
of 130 CFA from 1988-1993.
The domestic resource cost (DRC)-' of
rice produced in the Valley and marketed in Dakar has been
estimated as between 2-6 CFA (before devsluation).
This implies
that 2-6 CFA must be spent to replace 1 ZFA of imported rice.'"
Reduction of the DRC will require scme combination of increased
productivity
(through
increased
4.ields
and
rice-rice
doublecropping),
increased milling efficiency, higher retail rice
prices and adjustment for the overvaluaticn of the CFA.
V. GOS Rice Sector Policy
The agricultural sector in Senegal has a
history of heavy state
involvement.
The NAP and subsequently the Cereals Policy of 1986
set the stage for market liberafization and price deregulation in
the millet,
sorghum and maize sectors.
Although,
the domestic
rice sector has been fully decontrolled, =he imported rice sector
continues to be subject to state intervention, with fixed consumer
prices and heavy state involvement in marketing.
Until 1992 the GOS continued a policy cf import-substitution to
encourage
the substitution of local cereals for broken rice
imports.
The policy consisted mainly
of subsidies to paddy
producers in the SRV and taxes on urban rice consumers.
Although
the program in general has helped to achieve some level of import
substitution,
several of
its
components
have had conflicting
economic impacts.
For example, subsidized distribution of imported
rice has created price distortions for 1 ccal cereals and imported
rice regionally.
Likewise,
subsidies for :he irrigation system and
SAED processing and marketing functions has been at a high cost to
the economy.
In addition, the overvalue< exchange rate served as
an implicit subsidy to rice imports and a tax on paddy producers,
counter to the import-substitution policZ-.
For several years the GOS has worked v2:ith a multi-donor group and
as a result has developed the PAS> which will complete the
privatization of the entire cereals sector.
The three major
elements of the program which will be imp1 emented from 1994-1996
include: 1) elimination of all transporta -,ion subsidies along with
all fixed prices (paddy in 1994, imported broken rice in 1996) and
19 DRC swws as a mesure of comparative advantage. DRCs between zero and one imply cmpamve advantage: DRCs Iss tian zero or greater than one
imply no cmpamve rdvmtage.
Kite. Rod. 1993. p. 5.
5
margins, 2)
state disengagement from production, processing and
marketing locaily produced rice, and 3) an end to state involvement
in the distribution of imported rice.
The PASA also required
adjustment
for the overvalued exchange rate.
Prior to the 50%
devaluation of the CFA in early 1994, the Club du Sahel estimated
that the CFA was overvalued by 70%."
The devaluation will make
imports into Senegal more expensive and its exc,orts less e&xpensiTie.
It is expected that the program will result in lower prod*;cer
prices and higher consumer rice prices.
Paddy prices will likely
decrease with elimination of the guaranteed farm gate price, but
farm incomes are not expected to drop significantly for two
reasons.
First, the actual farm price has been about lo-20 CFA/kg
lower than the fixed price (offical and parallel market prices
weighted by shares marketed).
Secondly, as it will be legal for
private traders and processors to market rice, it is likely the low
local market prices will increase with some of the increase being
passed on to farmers.
Thus it appears there will be upward
pressure on retail rice prices.
Even if local
rice prices
eventually follow the lower import-delivered
(import price plus
transportation and handling costs) price path, the price of rice
received by farmers in Senegal will be higher, due to devaluation
of the CFA which has made imports more expensive.
Whether increased rice prices will lead to decreased demand for
rice depends on the substitutability between oLher local cereals
and rice,
a relationship which is not well understood.
The
empirical evidence would lead one to conclude either that it is not
possible
to determine (relevant cross-price elasticities not
obtainable) or that ric e and local cereals are in fact complements
(negative cross-price elasticities).
The studies which have found
rice and local cereals to be complements must however be qualified.
There is some question as to the reliability of the data used.
Likewise,
the
conciusions
have
typically
been
based on
UuncompensatedN demand elasticities whereas the economic concept of
substitution is usually based on compensated demand elasticities.
The difference between gross substitutes ("uncompensated") versus
substitutes ("compensated") is more than semantics in this case.
Indeed the two terms are approximately equal only in areas where
rice is produced, thus excluding much of the Senegalese population.
VI. Cultivar Development in the Senegal River Valley
The majority of the rice cultivars currently grown in the Sahel
were obtained through WARDA coordinated trials in the 1970s and
1980s.
Before the establishment of irrigation schemes in the 1920s
21 Sidibe. Mamadou. 1993, p. 6.
2 2 Kite. Rod. 1993. p. iv.
the only indigenous rice variety cultivated in the Sahel was Oryza
glaberrima.
Today must farmers in the SRV use only two rice
varieties:
medium-duration Jaya and short-duration IKP.
New
introduc: ions have not been able to match the existing varieties in
terms cf yield, stability or grain quality.
Table 1. Percent of Land Planted in Rice Cultivars,
Saint-Louis Region, r390
Zone
Jaya
IKP
Others
Cagana
44
43
13
Podor
72
28
Xstam
99
1
Region
67
27
6
Source : Eragantini, C., 1990.
Table 2. Survey Information, Zagana, STet Season, 93/94
Respondents
Frequency (%)
Land Ownership'
827
SAED
58
Private
42
Variety
818
Jaya
71
IKP
15
Otherss
14
Methc5 of Seeding
827
Direct Seeding
99
Transplantingc
1
Period of Seeding
807
<June 16
6
JunE 16-30
5
July 1-15
19
July 16-31
31
Aug. 1-15
30
Aug. 16-31
9
>Sept. 1
1
Source:
'Estimation de la Production
de PaiJy dans la Delegation de Cagana,
Hivernage 93/94' (SAED).
a Actual land-users.
" Others: Aiwu 3.67%; IR-1529 6.27%.
' Transplar.ting concerns only the z:ne of MEaxe.
Jaya, introduced from India in 1970, is very high yielding, with a
potential yield of about 9 t/ha.
It is a very high quality grain
and thus favorable for consumption.
Its two major weaknesses are
as follows: it is not tolerant to saline conditions which are ever
present in the Delta (north,s.-estern SRV) and the length of the
growth cycle is too long to allow for double cropping.
7
IKP was introduced from Thailand a few years before Jaya.
It has
the following major advantages: it may be grown in any season and
because
it has a short cycle it allc;%-s
for rice-rice double
cropping.
It is comparable to Jaya in terms of yield (potential
yield of 9 t/ha), hc>:eTJer it has poor grrin quality.
IR-1529 introduced in the late 198Os, also deserves mention. AS
Table 3 shcws, IR-1523 may be used by face,ys as much or more than
IKP, at least fcr Eke wet season.
Tables 1 and 2 lead one to
believe that IKP is more prominent hc~jever.
IR-1529 is high
yielding, with potential yield of 8 t/ha and almost as high quality
for consumption as Jaya.
It is medium-duration and may only be
grown in the wet season.
Table 3. Seeds Processed by Private Operators,
\\<et Seasons 91/92 and 92/93
91/92
92/93
Variety
Total (tons)
Percent
Total (tons)
Percent
Jaya
665.145
67
871.425
71
IKP
i21.672
12
log.160
9
IR-1529
156.8E5
16
224.750
18
Others
54.545
6
29.600
2
Jaya and
IK? in
general
have
Very
favcrable
agronomic
characteristics and It is not expected that potential yields can be
increased significantly above that of either variety.
However, the
identification and adaptation of new r ice cultivars is urgently
needed.
It is necessary to broaden the rl-Le germplasm to decrease
vulnerability to pests and disease.
Likewise, farmers have been
requesting alternatil.es to Jaya and IKP.
One short term research strategy at WARD.=: is to identify varieties
of at least the same yield potential and stability of Jaya and IKP,
but with improvements in other important criteria.
New short-
duration
varieties
are needed to all:-.g for rice-rice
double
croppinr.
Whereas in the past yields FZ~ e stressed to intensify
production, rice-rice doublecropping is nc;c recognized as presently
having the most potential for increasing productivity and farm
incomes
in the Valley.
There is also a need for varieties
specifically adaptable to the varying a gro-ecological conditions
across
the Valley.
For example,
soil
salinity
is a serious
constraint to irrigated rice production in the Delta area. In
addition,
with market liberalization, rice cultivars of high grain
quality will become increasingly important.
Previously, farmers
turned over their paddy production to URIC either as payment for
irrigation costs or for commercial sales at official producer
8
prices.
URIC has accepted paddy of any quality.
It is expected
that private mills will practice quality based pricing.
Each season WARDA screens about 300-400 varieties.
The goal of the
screening process is to identify and develop varieties which
combine the criteria mentioned above and enable high, stable yields
in varying seasons.
A succession of varietal evaluation and yield
trials forms the central element of WARDA's integrated research
approach.
WARDA currently has
implemented
experiments in
two
research
stations in the SR!' located at Fanaye and N'Diaye.
The locations
were chosen due to their representativeness of the major Sahelian
rice environments.
Fanaye is located inland, with typical semi-
arid, continental weather.
N'Diaye with its coastal location, has
a more moderate climate, higher air humidity and provides the
opportunity to expcse plants to saline soil conditions.
By the time the regional trials level is reached only about 23 of
the original 300 cuitivars remain.
In principal, responsibilities
are
then
transferred
from WARDA to
the
national
research
organisation,
ISRA,
and government parastatals which are then
responsible for multilocation trials and demonstration trials on
farmers' fields.
Those cultivars which advance frcm the regicnal
trials level become part of a varietal catalogue from which a
committee consisting of Division de Semence, PAS and ISRA decides
which seeds will be validated and released to farrr;ers.
Tables 4
and 5 summarize the results from experiment stations at NDiaye and
Fanaye (1988-93) and farm trials in the delegations of Podor and
Matam (19931,
Table 4. Yield, Yield Advantage (over IKP or Jaya), and Cycle,
Wet Seascn (Mean 1988-93)
Duration
Variety
Yield (t/ha)
Yield Advantage
Cycle
(days)
(%I
Short
Sahel lC8
6.118
14
113
IKP
5.213
116
Medium
Sahel 2Cl
6.088
16
124
Sahel 2.Z2
6.228
20
125
Jaya
4.913
123
Source: WARDA statistics
Table 5. Yield, Yield Advantage (over IKP), and Cycle,
Hot-Dry Season (Mean 1988-93)
Variety
Yield (t/ha)
Yield Advantage
Cycle (days)
(%I
Sahel 108
5.395
1
119
IKP
4.857
127
Source: W3.D.4 statistics
Currently all seed is distrib>Ated by PAS.
All extension services
in the Valley have been the responsibility of SAED, which has been
the only source cf information on seeds.
However, with market
liberalization,
some of the extension responsibilities are being
passed on to PW;A, the national extension service.
PNVA is
expected to improve the research-extension linkage in the Valley.
The provision of credit to the rural community has been the
responsibility of CNCAS, howe ':-er the greatest share of its funding
has gone to marketing campaigns for government parastatals.
CNCAS
has had a relatively high volume of loans which have not been
repaid, for example in Septem-
&=r 1989 the repayment rate for loans
was 7G%."
In response, C1JCP.S has denied credit to entire regions
and focused mainly on specific projects or GIEs.
VII. Model for Estimating Social Returns to Research
The model used to estimate the social benefits of research is the
partial-equilibri.Lm technique based on the concept of economic
surplus.
Eccnomic surplus methodology may be used for ex-post or
ex-ante evaluation of research. Ideally, ex-ante analyses should be
followed by ex-post analyses after the technology has been
introduced, for supplementati on and as a check for accuracy.
The economic surpl>is technique is commonly used for impact studies
because it is the simplest me-,hod for estirr;ating the joint impacts
of
research-extecsion
activities.
Because it is a partial-
equilibrium technique, all ot-.=
*--r economic sectors are assumed fixed
and thus an analysis
of s,pply
and demand changes in other
agricultural sectors
(i.e. rraize,
sorghum, millet) due to the
technological change in the rice sector is not possible. An
analysis of
several
ccmo5ity
sectors
would require a more
complicated general equi1ibri.n technique.
Another drawback of the econorrLc surplus method is that it does not
allow
for
the
separation of
the
individual
components of
technological change.
For example,
the individual contributions of
10
new
cultivars,
improved
agricultural practices
and
increased
fertilizer use are lumped together. The production function
approach would allow for individual parameter estimates of the
various factors related to increased productivity, however this
method is considerably more data demanding and unfortunately data
for the rice sector in the Valley is relatively limited. This may
not be an important issue in this case as the Sahel cultivars, for
a given season,
do not require addition21 input use over that of
exlstinz cultivars.
The eccnomic surplus technique estimates the gains and losses to
producers and consumers with the shift of the supply function.
Technological progress generated through agricultural research,
such as the Sahel cultivars,
leads to an upward shift in the
production function, implying that productivity has increased for
a given level of inputs.
The result is an increase in the supply
of rice as shown by the outward shift in the supply function of
Figure 2. Supply and demand curves may be linear or non-linear, and
the movement of the curves may be pivotal or parallel.
The choice
of linear versus nonlinear,
as well as that of pivotal versus
parallel
shifts can have a significant impact on the estimated
level of benefits and their distribution among consumers and
producers.
Whereas a parallel shift is easier to calculate, a
pivotal movement is often preferred for the more modest benefits it
generates and will be used here.
11
Figure 2: Technolcgical Change and Social Gains to Rice Reseaich
Quantity
Overall, the increase in economic benefits is represented by area
AOB plus the area .>3C.
Using the folio>.-ing formula, the economic
benefits can be calculated.
EB = kP(2 ,1/2P~(k(l+Es))'/(Ed+Es)
Which is equal to the triangles AOC plus ABC.
k = pro&--
--Ytional increase in yields with the new
cult i-.-ars times the proportion of the area in the nex
cclti-.-ars.
= L*(Yn-Yt)/Yn=L*(l-Yt/Yn
L = proiection for proportion of land expected to be
planred in Sahel cultivars
yt,
Yn = average yields of existing (traditional) and new
Sahel cultivars
12
E 5, E, = elasticities of s-apply and demand
P = expected price for different scenarios
VIII.
Methods of Calculation
Data
requirements for an
ex-ante
rate of
return
study
are
relatively low, consisting primarily of subjective estimates of
yield advantage, diffusion, price changes, quantity prodilced, and
research-extension expenditures.
Technological change was measured as: 1) the yield advantage of the
new cultivars, Sahel 108, Sahel 201 and Sahel 202,
in comparison
with the traditional varieties,
Jaya and IKP and 2) the double
cropping possibilities
offered by the new cultivars.
WARDA
experiment
station data from the N'Diaye and Fanaye sites is
available for the years 1988-1993.
XARDA trials compared Sahel 108
with I Kong Pao; and Sahel 201 and 2G2 were compared with Jaya.
In
the experiments the same level of inputs were used for each
cultivar.
Data from 1993 farm trials ccordinated by ISRA, WARDA,
Disem, SAED and PAS, of the Sahels in the delegations of Podor and
Matam is also available.
It is widely recognized that experiment station yields are higher
than those obtained by the average farmer.
Stations usually have
different physical and economic conditicns from the typical farm.
Farm level data is the relevant data for the study, but experiment
station and farm trial results were used in combination to
calcL;late the yield ad;rantage of eseh T:ariety. This was done to
ensure representativeness of the varying conditions in the Vallejr.
Dagaza farmers did not participate in the farm trials and since
soil-salinity problems are not present in the other delegations,
experiment
station
results from NDia?ye,
in particular,
were
necessary.
Also, farm trial results were only available for 1993
whereas
experiment
station data
covered lCZ8-93,
The yield
estimates (combination of experiment station and farm trial
results) have been reduced by 2:% to better reflect yields
obtainable under farm conditions.
WARCA provided estimates for the expected rate of diffusion for
each new cultivar over a IO-year period.
A more conservative
scenario has been added that considers a 50% reduction in the base
line diffusion rates.
Farm interviews were conducted with seven GIE chiefs to determine
the potential responsivenes of farmers to the introduction of the
Sahel cultivars. According to the GIE chiefs, the main constraint
to rice cultivation is that the necessary inputs are often not
available at the time of planting.
This obviously causes problems
in terms of planting on time to allow for rice-rice double
cropping,
a major objective
in the Valley.
The other main
13
difficulties mentioned were access to credit and the duration of
current seeds, each of which pose problems for double cropping.
Receiving payments for paddy from URIC six months to a year late
was mentioned by two farmers.
Six out of seven GIE cI.'-iefs stated that the main rice cultivar used
by faTLers is Jaya.
This was followed by IKP, Aiwu'4, and IR-1529.
No other cultivars were mentioned.
The most common response to why
farmers use a particular cultivar was because it is high yielding.
For charactristics of the "ideal" cultivar,
the most frequent
response was high yield,
s;ng followed by short cycle.
Resistance to
pests and consumption qualities were also frequent responses.
All village chiefs were familiar with the Sahels, as would be
expected since their farms were the sites of farm trials.
All were
familiar with Sahel 108 (trials began 1992), but some were not
familiar with 201 and 202 (trials in 1993). One village chief did
not recognize the names verbally, only visually.
He brought in the
trial plot signs to aid in discussion.
General impressions of the
Sahel cultivars is fay.rorable.
The main characteristics of Sahels
mentioned by farmers were high yielding, short-cycle and good for
consumption.
When asked if they wculd use the Sahel cultivars, 6 out of 7 GIE
chiefs responded yes.
Usually the response was qualified, yes, if
the restriction on multiplication is removed.
Of course, if the
cultivars are validated, multiplication will be allowed.
Responses
to whether the Sahels would replace existing cultivars or be used
in combination were divided evenly (one chief did not answer).
Farmers who said they would use the Sahels in combination with
currently
used
cultivars
said
they
have
the
objective of
diversifying for cons.Lmption and/or marketing purposes.
The representativeness of the surveys is uncertain, given that only
farms
participating in the trials were visited and only 7
interviews were cond*Lcted.
However,
because GIE leaders were
interviewed, it is expected that responses are
representative of
the particular villages.
Discussions with personnel at SAED, Disem and WARDA supplemented
information obtained from the intervie?;s.
A scientist at SAED
identified seed duration as the major constraint to increasing rice
production in the Valley.
Sahel 108 matures faster than IKP and
Jaya (see Tables 4-5); Sahel 201 and 202 are about equal to Jaya in
terms of duration.
A researcher at Disem stated that farmers from
Matam and Podor have come to his office requesting permission to
multiply the Sahel cultivars.
He also remarked on the past
ineffectiveness of extension.
He believes extension in the Valley
has improved with PNVA funding.
On August 5 and 17, 1994 the Sahel
cultivars
received
national
attention in
two
televised
24 Aiwu has not been cmtif-ied B was used in WARDA Earn trials from 1989 lo 19%.
14
documentaries.
A WARD.:. scientist who frequently interacts with
farmers in the 'Llalley said that many farmers have mentioned the
televised programs to him.
Although it is i-..u
'--ossible to know the extent of diffusion, ex-ante,
what is apparent is tL.s
"-t demand for the Sahel cultivars exists.
Farmers not only are requesting new seeds, but seeds with the
particular
characteristics of
the
Sahels:
short-cycle,
high
yielding,
resistance tc pests and saline-soil conditions, and high
quality grain.
Sahel varieties are popular among the farmers
familiar with them.
F"'*'"ver
.""I-
, extension efforts will be important
if the cultivars are to be widely adopted.
Research and extension
are
highly
complementary.
While
research
generates
new
technologies, extension enables the transfer of information among
farmers and researchers.
This facilitates the adoption of valuable
new
technologies
and
enables
researchers to
gain
important
information about the needs of farmers.
Labor requirements are also an issue in the SRV where seasonal
labor scarcity, due primarily to migration in search of employment,
is often a constraint to increasing cereals' productivity.
The
Sahels do not require additional labor per unit planted and
harvested, however if doublecropping increases there will be
increased labor requirements.
This is an important issue and
deserves further attention.
Increased availability of labor-saving
technologies may be
necessary to
allow
for
widespread
doublecropping.
Estimates of the quantity of paddy produced were calculated from
data on yields and the area in irrigated rice production for each
growing season.
These data are disaggregat ed by delegation as well
as by ownership (SAED or pril.-ate land).
Estimates for future area
in production are based on projected cropping intensities under two
scenarios: the first assumes that no land expansion occurs, the
second assumes that land ewsnsion does take place. Because prices
of rice were used it ?;as necessary to convert paddy produced to
rice produced.
This was done with a . 666 rate of transformation.'"
Research costs include costs of both WARDA and ISRA.
For the WARDA
related research, the following procedure was followed.
WARDA's
annual journals contain research program expenditures, 25% of which
goes to the Sahel Program.
Using WARDA figures for the area in
irrigated rice in each Sahelian country, an estimation of research
costs specific to the SRV *#;as calculated.
This figure was then
divided by 4 since WARDA has been involved in research on 8 other
Sahels.
Projections were based on past movements in research costs
(R2=.86).
Although ISRr? is involved in research of new cultivars,
WARDA provided funding for most of the ISRA research on the Sahels.
Because of difficulties obtaining research costs from ISFtA and
15
since it is expected that ISRA costs related to the Sahels are low,
these were not included in the analysis.
It is estimated that new variety testing takes 5 to 10 years. Since
the Sahel materials were selections rather than crosses, the lsi;er
limit of five years was used. So the research expenditures begin in
1990. Regional adaptation testing and agronomic work then continues
for five years after the introduction of the Sahels in 1995. Then
research on the Sahels is stopped.
Extension
costs
(for
additional distribution of
seeds
and
demonstrations)
were
estimated at
approximately
10 million
FCFA/year.
The costs of extension are included in the analysis for
the first five years following introduction of the seeds in 1995.
After the first five years it is expected that farmer-farmer
transfer will be the main form of diffusion of the cultivars.
The estimate for the price elasticity of demand for rice was taken
from Delgado (1988) and is E, = - .66.
The elasticity of supply for
rice is assumed to be E, = .7 since irrigated rice farmers can
rapidly respond to price changes.
16
IX. Results
The three critical determinants of the rate of return are the yield
advantage,
the rate of diffusion,
and the expected price. The yield
advantage estimates of the cultivars from experiment station and
farm trials were provided earlier.
Tithe yield differences were
discounted by 2c% to compensate for the mere timely operations
expected under experimental
conditicns. The expected diffusion
rates are provided in Table 6. Total dlffusi'on of the three new
cultivars reaches 50% after ten years.
Table 6.
Estimated
Diffusion
Rates
for
the
New
Cultivars
oercentaaei
YEAR
RAINY SEASON
HOT-DRY SEASON
NEW CULTIVARS
NEW CULTIVARS
I
I
I
I
Sahel
Sahel
Sahel
Sahel
Sahel
Sahel
108
201
202
108
201
202
1995
2
1
1
2
1
1
1996
4
2
2
5
2
2
1997
8
4
4
12
I
3
I3
19G8
16
6
8
22
I
4
I
4
1959
22
8
12
33
5
5
2000
25
10
15
40
5
5
2001
25
10
15
40
5
5
2032
25
10
15
40
5
5
20G3
25
10
15
40
5
5
2004
25
10
15
40
5
5
The price estimates were based on the economic outlook for rice,
and consultaticn with one of the authors of these reports. The
projected rice price over the next five years for Thai rice (35%
broken) would be approximately $200/ me:tric ton. We used world rice
prices of $200 and $250/metric tons and converted them to import
parity prices with an exchange rate of 487 FCFA/dollar, a
transformation rate of 66%, and estimation of the processing and
transportation costs.
These adjustments
resulted in an import
parity price range of 95 FCFA/kilo tc 119 CFA/kilo.
At the base line diffusion rates, the internal rates of return for
single cropping range from 66 to 72%.
With double cropping the
economic rates of return increase to 77 to 83% (see Table 7). One
of the main advantages of these shorter duration rice cultivars is
the ability to increase double cropping. There are additional labor
costs to get the one crop out and the other crop planted quickly
and we incorporated these costs into these estimates. Some farmers
may not be able to accomplish these additional activities without
mechanization. So additional study of the feasibility and the cost
of double cropping is appropriate.
17
Another advantage of the new rice cultivars mentioned by KX?DA
scientists is an improvement in quality over the traditional
cultivars. The addition of a price premium may be appropriate In
these calculations. However,
since there were no observations
available cn which to base this premium price estimate,
s,Jch
estimation was not done for this analysis.
Table 7.
Internal Rates of Return to Rice Research
(Ex-Ante)
for Different Expected Rice Prices With the
WARDA Diffusion Rates
No Area Expansion
Rice Price/Ton
Single Cropping
Double Cropping
(%I
(%I
95,000 FCFA
66
77
($200)
119,350 FCFA
72
a3
($250)
Note: 487 FCFA/US$ (1995)
Table 8.
Conservative Diffusion Rates for the New Cultivars
(50%
OF h3cnl;nn Fctim2tPc\\
fPFrrPntann\\
KI AJa3GAllLL Y-.lL&111L4
.-k-L,
,A
..A------^7-,
YEAR
RAINY SEASON
HOT-DRY SEASON
NEW CIJLTIVARS
NEW CULTIVARS
Sahel
Sahel
Sahel
Sahel
Sahel
Sahel
108
201
202
108
201
202
1995
l.G
0.5
0.5
1.0
0.5
0.5
1996
2.0
1.0
1.0
2.5
1.0
1.0
1997
4.c
2.0
2.0
6.0
1.5
1.5
1998
a.0
3.0
4.0
11.0
2.0
2.0
1999
11.0
4.0
6.0
16.5
2.5
2.5
2000
12.5
5.0
7.5
20.0
2.5
2.5
2001
12.5
5.0
7.5
20.0
2.5
2.5
2002
12.5
5.0
7.5
20.0
2.5
2.5
2003
12.5
5.0
7.5
20.0
2.5
2.5
2004
12.5
5.0
7.5
20.0
2.5
2.5
With the more conservative estimates of the diffusion rates, the
internal rates of return fall to 50% to 55% in the single cropping
case and 60% to 65% in the double cropping case (see Table 9).
These
are
still
very
credible
rates of
return
and
these
conservative estimates of diffusion are considered lower boundary
estimates. Hence, rates of return to the new rice cultivars appear
to justify the investments of WARDA, ISRA, and the farmers,
Table 9.
Internal Rates of Return to Rice Research (Ex-
Ante) With More Conservative Diffusion Rates
No Area Expansion
Rice Frice/Ton
Single Cropping
Double Cropping
(%)
(%I
95,000 FCFA ($200)
50
60
119,350 FCFA ($250)
55
65
487 FCFA/US$ (1995)
Only investments in Senegal on these rice cultivars were included
in the analysis as previous rice research done outside the country
was considered to be available to Senegal at minimal cost. The
rates of return then are from the perspective of investments made
in Senegal given the present state of rice technology in the rest
of the world. These esimates of the rate of return respond to the
following question: What are the rates of return to society in real
terms that Senegal will receive
from rice research based upon the
expected returns from the new cultivars developed and just starting
to be extended?
In the Senegal Valley there is substantial interest in further
extensions of the irrigated area. The analysis was also done with
land area expansion through increased irrigated areas by projecting
the past historic area expansion into the future. However, the
costs of this operation are so high that the combined investments
in new cultivars and expanded areas become marginal investments
except at the higher price and with double cropping. The costs of
this area expansion is an estimated 400,000 FCFA/ha.
This estimate
for drainage and land-leveling was obtained from farm interviews.
To this, must be added the additional variable costs of expanding
the land area.
Both irrigation and input costs were included as
additional costs of production in the analysis to account for the
expected increase in such costs as farmers begin to practice rice-
rice double cropping.
Irrigation costs are higher than input
costs, and account for 1.6% to 5.5% of the total amount earned from
rice production between 1995 and 2004.
19
Table 10.
internal Rates of Return to the Combined
Investments in the New Rice Technologies and Further
Extension of the Irrigated Area With the WARDA Diffusion
Rates
Area Expansion
Rice Price/Ton
Single Cropping
Double Cropping
(%I
(%I
95,000 FCFA ($200)
-
18
119,350 FCFA ($250)
10
28
487 FCFA/US$ (1335)
For the Sahels to be widely adopted, extension will be necessary to
inform farmers who have not been involved in WARDA farm trials
about
the existence
of the Sahels.
Extension will also be
important for providing
information on appropriate use of the
cultivars.
For example if farmers use Sahel 108 for rice-rice
doublecropping,
they will need information on soil and water
management and perhaps most importantly,
information about crep
calendars since doublecropping requires a precise crop calendar.
Improvements in access to credit and prompt payment for paddy will
be of critical importance in terms of allowing for rice-rice double
cropping and high yields.
Farmers must be able to purchase the
seeds and complementary inputs at the start of the hot-dry season
to allosJ for doublecropping.
Likewise, yields are quite sensitive
to the time of planting, as shown in Table 11.
Table 11: A-ierage Yields with Respect to Period of Seeding,
Wet Season 1393, Delegation of Dagana.
Period of Seeding
Average Yield (t/ha)
Before June 30
5.322
July
4.319
After Aug. 1
4.459
Access to credit will also be necessary to lessen the other major
constraint to doublecropping, seasonal labor shortages, as farmers
may need to borrow money to hire non-household laborers.
Aside
from its necessity to producers, a well functioning credit system
will also facilitate privatization of the rice sector.
Private
agents will need to build processing plants and storage facilities
or buy such sites from SAED.
These investments are costly.
.If -
20
credit is not available for such purposes or the interest rate is
too high, privatization will not tc a smooth process.
It is
expected that with market liberalization, the problem of delayed
payment for paddy will be eliminated, but the need for a well
functioning credit system will remain an issue.
Access to credit
is difficult in Senegal,
and this
is particularly true for
agricultural production.
Recently it was estimated that only 3% of
all
commercial bank credit was allotted to the agricultural
sector."
X: Conclusions:
The future rates of return to the development of these three new
cultivars are high even with the present land area and without
double cropping.
These rates are further increased with double
cropping.
Surprisingly, extending the irrigated land area is only
a marginal
investment combined with these new cultivars. This
results from the high costs of extending the irrigated area.It
appears to be worthwhile to examine these costs in more detail and
to identify the conditions under which it would be profitable to
extend the irrigated area.
The Sahel cultivars have higher yields than existing cultivars.
Sahel 108 is shorter duration than IKP and thus can increase the
likelihood of rice-rice double cropping.
The three cultivars are
of high grain quality, the position of the panicle leaf for Sahel
108 makes access to the seed by birds difficult, and Sahel 201 and
Sahel 108 have shown very modest decreases in yield with exposure
to salinity in ongoing experiments.
However, whether technological
change leads to increased productivity in the Valley will depend on
many interdependent factors: extension, credit services, changes in
marketing and processing, transportation infrastructure, azcess to
information.
It is hoped that researchers at ISRA and/or WARDA will carry out an
ex-post study of the Sahel cultivars.
Such an analysis would
provide more accurate figures for the rates of return to investment
on the new cultivars.
Likewise, such a study would allow for the
comparison between ex-ante and ex-post results.
In addition to
calculating the rate of return, future studies should also consider
distributional issues.
Substantial variation in adoption is
expected between regions of the SRV, among small and large farmers,
and by gender of the farmer.
It will be important to look at
distributional issues so that it may be determined if changes are
necessary
so that adoption of the cultivars does not lead to
regional, between household or within household inequities.
26 Robert R Nathan Associates. Ix., 1987. Annex 3.9.
21
XI.Bibliography:
Arcia, Gustavo, Mary Young Rice, Mamadou Sidibe, and Khardlata Lo
N'Diaye, "Senegal Grand Model Technical Description:
Improving Family Welfare Through Women's Work in Farming
Systems in Senegal", Research Triangle Institute, North
Carolina, April 1990.
Bragantini, C., "Etude Sur la Qualite des Semences de Riz a-~ I;ive~.~
des Paq'sans de la Eegicn de Saint-Louis", M.D.R.H., Pr,cjet
A.P.S., Dakar, 1990.
Comite perrr,anent Interetats de Lutte centre la Secheresse dans la
Sahel (CILSS),
"Atelier de Formation sur L'Impact Econcmique de
la Recherche Agricole", Mali, June 6-10, 1994.
Daniels, Lisa, Julie Howard, Mywish Maredia, James Oehmke, and
Richard Bernsten, "Assessment of Agricultural Research: Ex-
Post, Ex-Ante, and Needed Methodologies", Department of
Agricultural Economics, Michigan State University, no date.
Either, Carl K. and John M. Staatz, "Food Security Policy In Sub-
Sahara2 Africa", Procey-nas
-4i
of the Intermnal Conf.s r c .P
e&-.-e c'
a Turbulent WC.r--d
F;zonor!b' , Malaga, Spain, August 26 - September 4, 1985.
Gittinger, J. Price, EconQCic &l.&ysis of Acrr~cl1ltu.ra1 Pro-ects
I .
cecoqd Edltlcn I The Johns Hopkins University Press, Baltimoie
and London, 1982.
International Food Policy Research Institute (IFPRI), Accelerat~irs
an Bfrica edited by John ?:.
Mellor,
Christopher L. Delgado, and Malcolm J. Blackie, The
Jo'nns Hopkins Universit:? Press, 1987.
International Monetary Fund, uternational Fiual Statl.
. .l,s
=+ ' ,m I
Jan . - Feb. 1994.
Internaticnal Rice Research Institute (IRRI), Economic
c&J -
c
af,
a:.
'
Los Banos, Philippines, 1:?8.
Kite, Rod, "Senegal: Ilr.plications for the Rice Policy Dialcgue",
USAID/Dakar, Revised S&ptember 1993.
Martin, Frederic, "Food Security and Comparative Advantage in
Senegal: A Micro-Macro Approach", unpublished dissertation
submitted to Michigan State University, 1988.
Norton, Gecrge W. and Jeffrey Alwang, ljltroductlon to Econcfics c,F
ultural Develooment, McGraw-Hill, Inc., 1993.
Republique Francaise et Ministere de la Cooperation, aux de
. - llep du Sene g,gd, 1986.
22
Republique d~1 Senegal, "Conseil Interministeriel sur Le Plan
Directeur Rive Gauche", March 1994.
RepI;? ique du Senegal, "Plan Directeur de Development Integre Pour
la Rive Gauche de la Vallee du Fleuve Senegal: Rapport Final",
;pril I-31.
Robert R. Nathan Associates, Inc.: Clark Ross, Philip Boyle, Don
Harrison,
;
and Charles Nicholson, me Eifect s of Structural
2&1JqtvC-t irt Seye. &, Columbia, Maryland: USAID Africa Bureau,
1387.
. .
Rosen, Harvey S., &&lic: Finance, Second Edltlo,."1 Richard D. Irwin,
Inc., Chicago, 1988.
Sanders, John H. and Mohamed M. Ahmed, "The Economic Impacts of
Hageen Dura-I in the Gezira Scheme, Sudan", Proceed;Lnas of the
*-
mat 4 ona
O+-
let CRSP Conference, Corpus
Christi, Texas, July 8-12, 1991.
Sere, Carlos and Love11 S. Jarvis, "The Betting Line on Beef: Ex
>nte Estimates of the Benefits of Research on Improved Pasture
for the Latin American Tropics", Procwtias of the
University, New Jersey, July 6-8, 1988.
Sidibe,
Ma,m.adou, ISRA/BAME, "Effets des Politiques Macro-
Economiques sur le Niveau de Protection de la Filliere du Riz
Local Irrigue", First Draft, October 1?93.
Societe Nati onale d'Amenagement et D'Exploitation des Terres du
"ol+a
--i- (SAED) , "Estimation des Productions en Riz-Mais-Sorgho:
Hivernage 1993/94",
Saint-Louis, April 1994.
Societ e Nationale d'Amenagement et D'Exploitation des Terres du
Telta (SAEDj , "Estimation de la Production de Paddy dans la
r)elegation de Dagana: Hivernage 93/94", Saint-Louis, April
1994.
Societe Nationale d'Amenagement et D'Exploitation des Terres du
Delta (SAED) , "Politiques et strategies de developpement rural
dans la vallee et le delta du fleuve Senegal, Annexe 1:
=-rolution
-.
des superficies amenagees", Contribution de la DPDR,
February 26, 1994.
Societe Nationale d'Amenagement et D'Exploitation des Terres du
Delta (SAED), "Statistiques Agricoles: Evolutions de 1981 a
1991", Saint-Louis, May 1992.
USAID/Senegal/ADO,
"Senegal Agricultural Sector Analysis", January
1991.
__-.
23
WAFSA , "Annual Report", Cote d'Ivoire, 1990, 1321, 1992.
World Eank, Sub-SaWar. Africa: Fram Growth to q.3
WgrLd Bank, Washington D.C., 1987.
b;;crld E;nk, >;0-‘2 ppv~i -,
; 902.
TP.xlcz * n ip Cc:1 h
- I
Cxf zrd C'xiversity Fress, Inc., Ne,cJ York, NY.
24
IN THE SENEGAL RIVER VALLEY
Monica G. Fisher
A. Abdoulaye Fall
Mamadou Sidibe
ISRA/BAME
September 1995
ACRONYMS
BAKE - Bureau dOAnalyses Macro-economiques
CIRAD - Centre de Cooperation Internationale en Recherche
Agronomique Pour le Developpement (France)
CNCAS - Caisse Nationale de Credit Agricole du Senegal
National agricultural credit institution (25% state
owned).
Created 1987-88 as part of the NAP to take
over the credit responsibilities of parastatal
agencies.
CSA - Comzissariat & la Securite Alimentaire
Xaintains a food security stock of 60,000 tons of rice,
purchases local cereals for distribution to deficit
areas and collects and publishes producer and consumer
prices for cereals.
CPSP - Caisse de Perequation et de Stabilisation des Prix
Parastatal which purchases domestic paddy (from SAED)
and imported rice and in turn sells it to licensed
private Lraders for distributicn.
DISEM - Division de Semence
General mission is to control q:zlity of seeds.
Kain
decision body for seed certification.
Provides
information about seeds to farmers, PWJA, SAED and
private multipliers.
Grants pezission for seed
multiplication and evaluates seeds.
GIE - Groupements d'Interet Economique
GIE's are producer unions established during the time
of the NAP (1984).
Producers organize into groups of
up to 50-100 individuals in order to make access to
credit easier (from CNCAS) as well as to reduce
production risks.
GOS - Government of Senegal
INGER - International Network for Genetic Evaluation of Rice
IRRI - International Rice Research Institute
ISRA - Institut Senegalais de Recherches hgricoles (1974)
NAP - New Agricultural Policy (1984)
ORSTOM - Institut Francais de Recherche Scientifique Pour de
Developpment en Cooperation
PAS - Projet Autonome Semencier
National seed multiplier.
Created g/15/90 to take over
seed multiplication duties of SAED during
implementation o-f rice sector privatization.
The
project is schedclled to end in 1995 at which time seed
multiplication activities will be completely
privatized.
PASA - Agricultural Sector Adjustment Program
Multi-donor program which has as its major objective
the elimination of distortions caused by the existing
rice sector policies.
PDRG - Plan Directeur de Developpement Integre Pour la Rive
Gauche de la Vallee du Fleuve Senegal (1992-2017)
PNVA - Programe Nationale de Vulgarisation Agricole
National extension program with World Bank funding.
Currently working with SAED on extension in the Senegal
River Valley.
SAED - Societe d'Amenagement et d'Exploitation des Terres du
Delta
Parastatal established in 1965 to construct irrigation
systems, provide extension services, distribute
production inputs, process grain and marke-, outputs in
the Senegal River Valley.
Although SAED's
responsibilities include all crops in practice it has
concentrated on rice.
URIC -
Unite autonome Rizeries Industrielles et Commerciales
Milling department of SAED in charge of purchasing
paddy from farmers to be turned over to CFSP.
WfiRDA - West African Rice Development Association
Intergovernmental research association formed in 1571
with a mandate to conduct rice research, training and
communications in West Africa.
I.
Introduction
Irrigated rice production systems were first introduced to the
Sahel in the 1920s.
Since that time, rice has become an important
staple crop throughout the zone and its importance continues to
urow.
From the mid-1960s to the mid-1980s, total rice production
in the Sahel grew -4% annually, while consumption increased 5%.'
The gan between domestic supply and demand has meant that Sahelian
countr-ies have become increasingly dependent on rice imports at a
time
when
these
nations
are
seeking to
attain food self-
sufficiency.
Indeed, between 1967 and 1988 consurnptionzof rice and
wheat imports increased an estimated 225% in Senegal.
The New Agricultural Policy of 1984 (NAP) and the Cereals Policy of
1986 articulated the Government of Senegal's (GOS) goal of 80% food
grain self-sufficiency by the year 2000.
Among the major emphases
of the policy statements is the expansion of irrigated production,
particularly irrigated rice production, in the Senegal River Valley
(SRV) .
The purpose of
the
investment is
to promote
the
substitution of locally produced cereals (rice, maize, sorghum and
millet) for imported rice, and decrease the production variability
associated with rainfed cropping systems.
Irrigated rice production is very costly.
Thus there is a clear
need for high-yielding rice
cultivars
not
only to meet the
increasing demand for rice,
but also to justify the high land
development and irrigation costs of the SRV scheme.
New cultivars
should also be early-maturing to allow for rice-rice double
cropping.
In addition, with the on-going liberalization of the
rice sector, there will be an increase in farmer demand for rice
cultivars,
adaptable to the specific physical,
biological and
socio-economic conditions of the Valley.
Today about 90% of all farmers in the Valley use only one or two
rice cultivars.
Medium-duration Jaya and short-duration I Kong Pao
(IKP), introduced from Asia in the 197Os, are still the most common
varieties despite requests from farmers for alternatives.' The IR-
1529 cultivar was released to farmers in the late 1980s.
Its
performance in terms of yield, growth cycle and grain quality is
almost as good as Jaya, but it has not been widely adopted mainly
because of inade uate
%
seed multiplication services and lack of
extension effort.
1 WARDA AMU! Report 1991, p. 70.
2 L.eaphvt Stqhmie. 1994, p. 3.
3 WARLI.4 Annual Report 1991. p. 74.
4 From cawusat~on at Disan md WARDA
The Sahel 108 and 201 were introducd from Asia and Sahel 202 from
Nigeria; and all have been screened, characterized, and introduced
in the SR'J. These three new rice cultivars iiere officially released
in November 1994.
This study estimates the potential impact cf research on the Sahel
cultivars.
Agricultural research has generally been found to have
high pay-offs, but there is a lack of infoc&ation on the returns to
research in Africa. Estimating, ex-ante, the impact of the Sahels
and identification of the constraints to their introduction and
adoption will be an extremely valuable exercise.
The results will
help determine whether rice research efforts are conducive to the
national goal of cereals self-sufficiency and aid in the allocation
of funds for future resesarch.
The report will complement other
BAME impact studies on millet, sorghum, maize, and cotton.
II.
Objectives
The general objective is to conduct an ex-ante evaluation of the
expected benefits and costs of rice research in the Valley.
The specific objectives include:
1 . Quantify the economic benefits to producers and consumers of
rice research.
2 . Kith the inclusion of data on research and extension costs,
calculate the internal rate of return tc the combined research-
extension efforts.
III.
Rice Consumption and Demand for Rice in Senegal
Although vast regional differences exist,
rice is an important
source of consumption in Senegal.
It is the basis of the major
Senegalese dish "Thiebou Dienne"
(literally,
fish and rice),
consumed in large quantities, particularly for the noOn time meal.
Nationally
rice
accounts
for 34% of
the
volume of
cereal
consumpticn,
making up 54% and 24% of the volume of cereals
consumed among urban and rural populations, respectively.
Food expenditures of urban and rural houVVI
c=holds are dominated by
rice across Senegal.
Recently it was est imated that the rural
population spends 25% of its food expenditures on rice, while the
urban pop*>lation spends 18% of its food budget on rice for home
consumpticn.'
Per capita consumption of rice in Senegal has fluctuated through
time, particularly during the drought years of the late 1970s to
2
early 1980s . During this period there was a massive substitution
O f
imported rice for local cereals.
As rice production and
consumption patterns did not return to pre-drought levels, imported
rice has served an important role in supplying ihe food needs
required by the 309 per annum population growth rate.
A 1992 TriARDA
study of rice consum?,?tion
in five Sahelian countries (Burkina Faso,
Mali, Mauritania, Niger and Senegal) found population growth and
rising consumption per capita to account for 40% and 60% of
the
ccnsumption increases, respectively.
"Regression analysis showed
that structural factors,
particularly urbanization, rather than
changing prices or income levels underlie the increased importance
of rice in the region's food consumption patterns."7
Ease of
preparation is the main reason rice is particularly popular in
urban areas.
Unless domestic rice production increases rapidly, rising demand
for rice will lead to increased dependency on imports, which is not
in line with GOS goals.
For this reason, the introduction of new
rice cultivars with high yields and short duration will become
increasingly important.
IV.
Rice Production in Senegal
The two major rice-growing areas in the country are 1) the Delta
and the Senegal River Valley (St. Louis to Bake11 and 2) the
Casamance Region, each with very different systems of production.
Each region has potential to increase productivity with new
technologies.
The Casamance Regicn in southern Senegal accounts for 61% of all
land planted to rice nationally, yet the region only contributes
29% of the national production.
Yields generally average only 1
ton/hectare due to the more traditional production system as well
as water constraints (mainly the upland system).
The principle
inputs are labor, land and seed, and farmers use little fertilizer,
pesticides or farm equipment.
Most of the rice in the area is
grown under rainfed and swamp production systems, mainly for home
consumption purposes.
The Senegal River Valley consists of four delegations:" Bakei in
the region of Tambacounda, Matam, Podor and Dagana in the region of
St. Louis.
Relative to the Casamance, rice grown in the Valle;l is
based on intensive and technologically advanced production methods,
6 World Bank Development
Report 1993.
7 WARDA Annual Report 19%. p. 40.
8 Kite, Rod, 1993. p. 2
9 Delegation is the SAED loid extension of&x.
with 67% of total production being marketed,
55% through the
official system and 12% through the parallel market.'"
The majority
of the active rural population is involved in irrigated rice
production.
The region accounts for a pproximately 89% of national
cash revenue from rice, and 41% of all agricultural cash revenue."
Most of the production in the Valley takes place in the Delta area
which accounts for about 62 % of all irrigated crop land and 71% of
paddy production.'2
Farms in the Dag ana delegation are larger in
size and use more mechanized inputs relative to the farms from
Podor to Bakel.
In general, production in Dagana is primarily
marketed whereas from Podor-Bake1 rice production is mainly for
home consumption.
Irrigated production quadrupled for the Senegal River Valley
between 1979-81 and 1989-91.
Most of the growth may be attributed
to expansion of area under cultivation, however productivity also
increased-l3
On average, the last decade (1982-1993) has seen an
increase of 1,976 ha/year of irrigated crop land in the Valley.'"
Yields have also increased in the Valley,
from an average of 3.3
t/ha (1979-81) to 5 t/ha (1989-91)."
Rice-rice double cropping is
practiced only on a limited basis.
Irrigation opened up the
possibility for two rice growing seascns in the Valley, but for a
variety of reasons (seasonal labor shortages, inadequate knowledge
of how to combine crop,
soil and water management,
access to
credit) cropping intensity continues to be low, rarely exceeding
1.=
The FA0 estimates that reaching the GOS's objective of cereals
self-sufficiency will require an annual increase of 3,800 hectares
of irrigated land, paddy rice yields of 6 tons per hectare, and a
coefficient of cropping intensity of 1.8."
Increased production in the Valley has involved and will continue
to involve high costs.
Irrigation system development in the Valley
has received the largest share of the GOS budget devoted to the
agricultural sector.
Likewise, the subsidies on production and
processing have been high.
In 1989 the total subsidy (irrigation,
1
0
S~dtbe. Mamadou. 1993. p. 7
11 Kite. Rod. 1993. p. 2
1’* S4ED. Feb. 1994, statls[1cs apply to 199Y93. p. 7.
1 3 U’ARLIA Annual Report 1992. Table 18. p. 40.
I4 SAED doctanents. 1992.1993.
1 5 WARDA A~ual Report 1992. p. 40.
1 6 Senegal Ag-imbal Sector Analysis. p. 72
1 7 Martin. 1988. p. 82.
producti .on and processing)
for local pa3 dy production was 88.4
CFA/kg,
which exceeded the CIF (cost
and freight) price for
imported rice of 85 CFA/kg." All costs and subsidies included it
L
costs approximately 300 CFA/kg to produce, process and market rice
in the Valley which in turn has been sold at a fixed retail p.rice
of 130 CFA from 1988-1993.
The domestic resource cost (DRC)-' of
rice produced in the Valley and marketed in Dakar has been
estimated as between 2-6 CFA (before devsluation).
This implies
that 2-6 CFA must be spent to replace 1 ZFA of imported rice.'"
Reduction of the DRC will require scme combination of increased
productivity
(through
increased
4.ields
and
rice-rice
doublecropping),
increased milling efficiency, higher retail rice
prices and adjustment for the overvaluaticn of the CFA.
V. GOS Rice Sector Policy
The agricultural sector in Senegal has a
history of heavy state
involvement.
The NAP and subsequently the Cereals Policy of 1986
set the stage for market liberafization and price deregulation in
the millet,
sorghum and maize sectors.
Although,
the domestic
rice sector has been fully decontrolled, =he imported rice sector
continues to be subject to state intervention, with fixed consumer
prices and heavy state involvement in marketing.
Until 1992 the GOS continued a policy cf import-substitution to
encourage
the substitution of local cereals for broken rice
imports.
The policy consisted mainly
of subsidies to paddy
producers in the SRV and taxes on urban rice consumers.
Although
the program in general has helped to achieve some level of import
substitution,
several of
its
components
have had conflicting
economic impacts.
For example, subsidized distribution of imported
rice has created price distortions for 1 ccal cereals and imported
rice regionally.
Likewise,
subsidies for :he irrigation system and
SAED processing and marketing functions has been at a high cost to
the economy.
In addition, the overvalue< exchange rate served as
an implicit subsidy to rice imports and a tax on paddy producers,
counter to the import-substitution policZ-.
For several years the GOS has worked v2:ith a multi-donor group and
as a result has developed the PAS> which will complete the
privatization of the entire cereals sector.
The three major
elements of the program which will be imp1 emented from 1994-1996
include: 1) elimination of all transporta -,ion subsidies along with
all fixed prices (paddy in 1994, imported broken rice in 1996) and
19 DRC swws as a mesure of comparative advantage. DRCs between zero and one imply cmpamve advantage: DRCs Iss tian zero or greater than one
imply no cmpamve rdvmtage.
Kite. Rod. 1993. p. 5.
5
margins, 2)
state disengagement from production, processing and
marketing locaily produced rice, and 3) an end to state involvement
in the distribution of imported rice.
The PASA also required
adjustment
for the overvalued exchange rate.
Prior to the 50%
devaluation of the CFA in early 1994, the Club du Sahel estimated
that the CFA was overvalued by 70%."
The devaluation will make
imports into Senegal more expensive and its exc,orts less e&xpensiTie.
It is expected that the program will result in lower prod*;cer
prices and higher consumer rice prices.
Paddy prices will likely
decrease with elimination of the guaranteed farm gate price, but
farm incomes are not expected to drop significantly for two
reasons.
First, the actual farm price has been about lo-20 CFA/kg
lower than the fixed price (offical and parallel market prices
weighted by shares marketed).
Secondly, as it will be legal for
private traders and processors to market rice, it is likely the low
local market prices will increase with some of the increase being
passed on to farmers.
Thus it appears there will be upward
pressure on retail rice prices.
Even if local
rice prices
eventually follow the lower import-delivered
(import price plus
transportation and handling costs) price path, the price of rice
received by farmers in Senegal will be higher, due to devaluation
of the CFA which has made imports more expensive.
Whether increased rice prices will lead to decreased demand for
rice depends on the substitutability between oLher local cereals
and rice,
a relationship which is not well understood.
The
empirical evidence would lead one to conclude either that it is not
possible
to determine (relevant cross-price elasticities not
obtainable) or that ric e and local cereals are in fact complements
(negative cross-price elasticities).
The studies which have found
rice and local cereals to be complements must however be qualified.
There is some question as to the reliability of the data used.
Likewise,
the
conciusions
have
typically
been
based on
UuncompensatedN demand elasticities whereas the economic concept of
substitution is usually based on compensated demand elasticities.
The difference between gross substitutes ("uncompensated") versus
substitutes ("compensated") is more than semantics in this case.
Indeed the two terms are approximately equal only in areas where
rice is produced, thus excluding much of the Senegalese population.
VI. Cultivar Development in the Senegal River Valley
The majority of the rice cultivars currently grown in the Sahel
were obtained through WARDA coordinated trials in the 1970s and
1980s.
Before the establishment of irrigation schemes in the 1920s
21 Sidibe. Mamadou. 1993, p. 6.
2 2 Kite. Rod. 1993. p. iv.
the only indigenous rice variety cultivated in the Sahel was Oryza
glaberrima.
Today must farmers in the SRV use only two rice
varieties:
medium-duration Jaya and short-duration IKP.
New
introduc: ions have not been able to match the existing varieties in
terms cf yield, stability or grain quality.
Table 1. Percent of Land Planted in Rice Cultivars,
Saint-Louis Region, r390
Zone
Jaya
IKP
Others
Cagana
44
43
13
Podor
72
28
Xstam
99
1
Region
67
27
6
Source : Eragantini, C., 1990.
Table 2. Survey Information, Zagana, STet Season, 93/94
Respondents
Frequency (%)
Land Ownership'
827
SAED
58
Private
42
Variety
818
Jaya
71
IKP
15
Otherss
14
Methc5 of Seeding
827
Direct Seeding
99
Transplantingc
1
Period of Seeding
807
<June 16
6
JunE 16-30
5
July 1-15
19
July 16-31
31
Aug. 1-15
30
Aug. 16-31
9
>Sept. 1
1
Source:
'Estimation de la Production
de PaiJy dans la Delegation de Cagana,
Hivernage 93/94' (SAED).
a Actual land-users.
" Others: Aiwu 3.67%; IR-1529 6.27%.
' Transplar.ting concerns only the z:ne of MEaxe.
Jaya, introduced from India in 1970, is very high yielding, with a
potential yield of about 9 t/ha.
It is a very high quality grain
and thus favorable for consumption.
Its two major weaknesses are
as follows: it is not tolerant to saline conditions which are ever
present in the Delta (north,s.-estern SRV) and the length of the
growth cycle is too long to allow for double cropping.
7
IKP was introduced from Thailand a few years before Jaya.
It has
the following major advantages: it may be grown in any season and
because
it has a short cycle it allc;%-s
for rice-rice double
cropping.
It is comparable to Jaya in terms of yield (potential
yield of 9 t/ha), hc>:eTJer it has poor grrin quality.
IR-1529 introduced in the late 198Os, also deserves mention. AS
Table 3 shcws, IR-1523 may be used by face,ys as much or more than
IKP, at least fcr Eke wet season.
Tables 1 and 2 lead one to
believe that IKP is more prominent hc~jever.
IR-1529 is high
yielding, with potential yield of 8 t/ha and almost as high quality
for consumption as Jaya.
It is medium-duration and may only be
grown in the wet season.
Table 3. Seeds Processed by Private Operators,
\\<et Seasons 91/92 and 92/93
91/92
92/93
Variety
Total (tons)
Percent
Total (tons)
Percent
Jaya
665.145
67
871.425
71
IKP
i21.672
12
log.160
9
IR-1529
156.8E5
16
224.750
18
Others
54.545
6
29.600
2
Jaya and
IK? in
general
have
Very
favcrable
agronomic
characteristics and It is not expected that potential yields can be
increased significantly above that of either variety.
However, the
identification and adaptation of new r ice cultivars is urgently
needed.
It is necessary to broaden the rl-Le germplasm to decrease
vulnerability to pests and disease.
Likewise, farmers have been
requesting alternatil.es to Jaya and IKP.
One short term research strategy at WARD.=: is to identify varieties
of at least the same yield potential and stability of Jaya and IKP,
but with improvements in other important criteria.
New short-
duration
varieties
are needed to all:-.g for rice-rice
double
croppinr.
Whereas in the past yields FZ~ e stressed to intensify
production, rice-rice doublecropping is nc;c recognized as presently
having the most potential for increasing productivity and farm
incomes
in the Valley.
There is also a need for varieties
specifically adaptable to the varying a gro-ecological conditions
across
the Valley.
For example,
soil
salinity
is a serious
constraint to irrigated rice production in the Delta area. In
addition,
with market liberalization, rice cultivars of high grain
quality will become increasingly important.
Previously, farmers
turned over their paddy production to URIC either as payment for
irrigation costs or for commercial sales at official producer
8
prices.
URIC has accepted paddy of any quality.
It is expected
that private mills will practice quality based pricing.
Each season WARDA screens about 300-400 varieties.
The goal of the
screening process is to identify and develop varieties which
combine the criteria mentioned above and enable high, stable yields
in varying seasons.
A succession of varietal evaluation and yield
trials forms the central element of WARDA's integrated research
approach.
WARDA currently has
implemented
experiments in
two
research
stations in the SR!' located at Fanaye and N'Diaye.
The locations
were chosen due to their representativeness of the major Sahelian
rice environments.
Fanaye is located inland, with typical semi-
arid, continental weather.
N'Diaye with its coastal location, has
a more moderate climate, higher air humidity and provides the
opportunity to expcse plants to saline soil conditions.
By the time the regional trials level is reached only about 23 of
the original 300 cuitivars remain.
In principal, responsibilities
are
then
transferred
from WARDA to
the
national
research
organisation,
ISRA,
and government parastatals which are then
responsible for multilocation trials and demonstration trials on
farmers' fields.
Those cultivars which advance frcm the regicnal
trials level become part of a varietal catalogue from which a
committee consisting of Division de Semence, PAS and ISRA decides
which seeds will be validated and released to farrr;ers.
Tables 4
and 5 summarize the results from experiment stations at NDiaye and
Fanaye (1988-93) and farm trials in the delegations of Podor and
Matam (19931,
Table 4. Yield, Yield Advantage (over IKP or Jaya), and Cycle,
Wet Seascn (Mean 1988-93)
Duration
Variety
Yield (t/ha)
Yield Advantage
Cycle
(days)
(%I
Short
Sahel lC8
6.118
14
113
IKP
5.213
116
Medium
Sahel 2Cl
6.088
16
124
Sahel 2.Z2
6.228
20
125
Jaya
4.913
123
Source: WARDA statistics
Table 5. Yield, Yield Advantage (over IKP), and Cycle,
Hot-Dry Season (Mean 1988-93)
Variety
Yield (t/ha)
Yield Advantage
Cycle (days)
(%I
Sahel 108
5.395
1
119
IKP
4.857
127
Source: W3.D.4 statistics
Currently all seed is distrib>Ated by PAS.
All extension services
in the Valley have been the responsibility of SAED, which has been
the only source cf information on seeds.
However, with market
liberalization,
some of the extension responsibilities are being
passed on to PW;A, the national extension service.
PNVA is
expected to improve the research-extension linkage in the Valley.
The provision of credit to the rural community has been the
responsibility of CNCAS, howe ':-er the greatest share of its funding
has gone to marketing campaigns for government parastatals.
CNCAS
has had a relatively high volume of loans which have not been
repaid, for example in Septem-
&=r 1989 the repayment rate for loans
was 7G%."
In response, C1JCP.S has denied credit to entire regions
and focused mainly on specific projects or GIEs.
VII. Model for Estimating Social Returns to Research
The model used to estimate the social benefits of research is the
partial-equilibri.Lm technique based on the concept of economic
surplus.
Eccnomic surplus methodology may be used for ex-post or
ex-ante evaluation of research. Ideally, ex-ante analyses should be
followed by ex-post analyses after the technology has been
introduced, for supplementati on and as a check for accuracy.
The economic surpl>is technique is commonly used for impact studies
because it is the simplest me-,hod for estirr;ating the joint impacts
of
research-extecsion
activities.
Because it is a partial-
equilibrium technique, all ot-.=
*--r economic sectors are assumed fixed
and thus an analysis
of s,pply
and demand changes in other
agricultural sectors
(i.e. rraize,
sorghum, millet) due to the
technological change in the rice sector is not possible. An
analysis of
several
ccmo5ity
sectors
would require a more
complicated general equi1ibri.n technique.
Another drawback of the econorrLc surplus method is that it does not
allow
for
the
separation of
the
individual
components of
technological change.
For example,
the individual contributions of
10
new
cultivars,
improved
agricultural practices
and
increased
fertilizer use are lumped together. The production function
approach would allow for individual parameter estimates of the
various factors related to increased productivity, however this
method is considerably more data demanding and unfortunately data
for the rice sector in the Valley is relatively limited. This may
not be an important issue in this case as the Sahel cultivars, for
a given season,
do not require addition21 input use over that of
exlstinz cultivars.
The eccnomic surplus technique estimates the gains and losses to
producers and consumers with the shift of the supply function.
Technological progress generated through agricultural research,
such as the Sahel cultivars,
leads to an upward shift in the
production function, implying that productivity has increased for
a given level of inputs.
The result is an increase in the supply
of rice as shown by the outward shift in the supply function of
Figure 2. Supply and demand curves may be linear or non-linear, and
the movement of the curves may be pivotal or parallel.
The choice
of linear versus nonlinear,
as well as that of pivotal versus
parallel
shifts can have a significant impact on the estimated
level of benefits and their distribution among consumers and
producers.
Whereas a parallel shift is easier to calculate, a
pivotal movement is often preferred for the more modest benefits it
generates and will be used here.
11
Figure 2: Technolcgical Change and Social Gains to Rice Reseaich
Quantity
Overall, the increase in economic benefits is represented by area
AOB plus the area .>3C.
Using the folio>.-ing formula, the economic
benefits can be calculated.
EB = kP(2 ,1/2P~(k(l+Es))'/(Ed+Es)
Which is equal to the triangles AOC plus ABC.
k = pro&--
--Ytional increase in yields with the new
cult i-.-ars times the proportion of the area in the nex
cclti-.-ars.
= L*(Yn-Yt)/Yn=L*(l-Yt/Yn
L = proiection for proportion of land expected to be
planred in Sahel cultivars
yt,
Yn = average yields of existing (traditional) and new
Sahel cultivars
12
E 5, E, = elasticities of s-apply and demand
P = expected price for different scenarios
VIII.
Methods of Calculation
Data
requirements for an
ex-ante
rate of
return
study
are
relatively low, consisting primarily of subjective estimates of
yield advantage, diffusion, price changes, quantity prodilced, and
research-extension expenditures.
Technological change was measured as: 1) the yield advantage of the
new cultivars, Sahel 108, Sahel 201 and Sahel 202,
in comparison
with the traditional varieties,
Jaya and IKP and 2) the double
cropping possibilities
offered by the new cultivars.
WARDA
experiment
station data from the N'Diaye and Fanaye sites is
available for the years 1988-1993.
XARDA trials compared Sahel 108
with I Kong Pao; and Sahel 201 and 2G2 were compared with Jaya.
In
the experiments the same level of inputs were used for each
cultivar.
Data from 1993 farm trials ccordinated by ISRA, WARDA,
Disem, SAED and PAS, of the Sahels in the delegations of Podor and
Matam is also available.
It is widely recognized that experiment station yields are higher
than those obtained by the average farmer.
Stations usually have
different physical and economic conditicns from the typical farm.
Farm level data is the relevant data for the study, but experiment
station and farm trial results were used in combination to
calcL;late the yield ad;rantage of eseh T:ariety. This was done to
ensure representativeness of the varying conditions in the Vallejr.
Dagaza farmers did not participate in the farm trials and since
soil-salinity problems are not present in the other delegations,
experiment
station
results from NDia?ye,
in particular,
were
necessary.
Also, farm trial results were only available for 1993
whereas
experiment
station data
covered lCZ8-93,
The yield
estimates (combination of experiment station and farm trial
results) have been reduced by 2:% to better reflect yields
obtainable under farm conditions.
WARCA provided estimates for the expected rate of diffusion for
each new cultivar over a IO-year period.
A more conservative
scenario has been added that considers a 50% reduction in the base
line diffusion rates.
Farm interviews were conducted with seven GIE chiefs to determine
the potential responsivenes of farmers to the introduction of the
Sahel cultivars. According to the GIE chiefs, the main constraint
to rice cultivation is that the necessary inputs are often not
available at the time of planting.
This obviously causes problems
in terms of planting on time to allow for rice-rice double
cropping,
a major objective
in the Valley.
The other main
13
difficulties mentioned were access to credit and the duration of
current seeds, each of which pose problems for double cropping.
Receiving payments for paddy from URIC six months to a year late
was mentioned by two farmers.
Six out of seven GIE cI.'-iefs stated that the main rice cultivar used
by faTLers is Jaya.
This was followed by IKP, Aiwu'4, and IR-1529.
No other cultivars were mentioned.
The most common response to why
farmers use a particular cultivar was because it is high yielding.
For charactristics of the "ideal" cultivar,
the most frequent
response was high yield,
s;ng followed by short cycle.
Resistance to
pests and consumption qualities were also frequent responses.
All village chiefs were familiar with the Sahels, as would be
expected since their farms were the sites of farm trials.
All were
familiar with Sahel 108 (trials began 1992), but some were not
familiar with 201 and 202 (trials in 1993). One village chief did
not recognize the names verbally, only visually.
He brought in the
trial plot signs to aid in discussion.
General impressions of the
Sahel cultivars is fay.rorable.
The main characteristics of Sahels
mentioned by farmers were high yielding, short-cycle and good for
consumption.
When asked if they wculd use the Sahel cultivars, 6 out of 7 GIE
chiefs responded yes.
Usually the response was qualified, yes, if
the restriction on multiplication is removed.
Of course, if the
cultivars are validated, multiplication will be allowed.
Responses
to whether the Sahels would replace existing cultivars or be used
in combination were divided evenly (one chief did not answer).
Farmers who said they would use the Sahels in combination with
currently
used
cultivars
said
they
have
the
objective of
diversifying for cons.Lmption and/or marketing purposes.
The representativeness of the surveys is uncertain, given that only
farms
participating in the trials were visited and only 7
interviews were cond*Lcted.
However,
because GIE leaders were
interviewed, it is expected that responses are
representative of
the particular villages.
Discussions with personnel at SAED, Disem and WARDA supplemented
information obtained from the intervie?;s.
A scientist at SAED
identified seed duration as the major constraint to increasing rice
production in the Valley.
Sahel 108 matures faster than IKP and
Jaya (see Tables 4-5); Sahel 201 and 202 are about equal to Jaya in
terms of duration.
A researcher at Disem stated that farmers from
Matam and Podor have come to his office requesting permission to
multiply the Sahel cultivars.
He also remarked on the past
ineffectiveness of extension.
He believes extension in the Valley
has improved with PNVA funding.
On August 5 and 17, 1994 the Sahel
cultivars
received
national
attention in
two
televised
24 Aiwu has not been cmtif-ied B was used in WARDA Earn trials from 1989 lo 19%.
14
documentaries.
A WARD.:. scientist who frequently interacts with
farmers in the 'Llalley said that many farmers have mentioned the
televised programs to him.
Although it is i-..u
'--ossible to know the extent of diffusion, ex-ante,
what is apparent is tL.s
"-t demand for the Sahel cultivars exists.
Farmers not only are requesting new seeds, but seeds with the
particular
characteristics of
the
Sahels:
short-cycle,
high
yielding,
resistance tc pests and saline-soil conditions, and high
quality grain.
Sahel varieties are popular among the farmers
familiar with them.
F"'*'"ver
.""I-
, extension efforts will be important
if the cultivars are to be widely adopted.
Research and extension
are
highly
complementary.
While
research
generates
new
technologies, extension enables the transfer of information among
farmers and researchers.
This facilitates the adoption of valuable
new
technologies
and
enables
researchers to
gain
important
information about the needs of farmers.
Labor requirements are also an issue in the SRV where seasonal
labor scarcity, due primarily to migration in search of employment,
is often a constraint to increasing cereals' productivity.
The
Sahels do not require additional labor per unit planted and
harvested, however if doublecropping increases there will be
increased labor requirements.
This is an important issue and
deserves further attention.
Increased availability of labor-saving
technologies may be
necessary to
allow
for
widespread
doublecropping.
Estimates of the quantity of paddy produced were calculated from
data on yields and the area in irrigated rice production for each
growing season.
These data are disaggregat ed by delegation as well
as by ownership (SAED or pril.-ate land).
Estimates for future area
in production are based on projected cropping intensities under two
scenarios: the first assumes that no land expansion occurs, the
second assumes that land ewsnsion does take place. Because prices
of rice were used it ?;as necessary to convert paddy produced to
rice produced.
This was done with a . 666 rate of transformation.'"
Research costs include costs of both WARDA and ISRA.
For the WARDA
related research, the following procedure was followed.
WARDA's
annual journals contain research program expenditures, 25% of which
goes to the Sahel Program.
Using WARDA figures for the area in
irrigated rice in each Sahelian country, an estimation of research
costs specific to the SRV *#;as calculated.
This figure was then
divided by 4 since WARDA has been involved in research on 8 other
Sahels.
Projections were based on past movements in research costs
(R2=.86).
Although ISRr? is involved in research of new cultivars,
WARDA provided funding for most of the ISRA research on the Sahels.
Because of difficulties obtaining research costs from ISFtA and
15
since it is expected that ISRA costs related to the Sahels are low,
these were not included in the analysis.
It is estimated that new variety testing takes 5 to 10 years. Since
the Sahel materials were selections rather than crosses, the lsi;er
limit of five years was used. So the research expenditures begin in
1990. Regional adaptation testing and agronomic work then continues
for five years after the introduction of the Sahels in 1995. Then
research on the Sahels is stopped.
Extension
costs
(for
additional distribution of
seeds
and
demonstrations)
were
estimated at
approximately
10 million
FCFA/year.
The costs of extension are included in the analysis for
the first five years following introduction of the seeds in 1995.
After the first five years it is expected that farmer-farmer
transfer will be the main form of diffusion of the cultivars.
The estimate for the price elasticity of demand for rice was taken
from Delgado (1988) and is E, = - .66.
The elasticity of supply for
rice is assumed to be E, = .7 since irrigated rice farmers can
rapidly respond to price changes.
16
IX. Results
The three critical determinants of the rate of return are the yield
advantage,
the rate of diffusion,
and the expected price. The yield
advantage estimates of the cultivars from experiment station and
farm trials were provided earlier.
Tithe yield differences were
discounted by 2c% to compensate for the mere timely operations
expected under experimental
conditicns. The expected diffusion
rates are provided in Table 6. Total dlffusi'on of the three new
cultivars reaches 50% after ten years.
Table 6.
Estimated
Diffusion
Rates
for
the
New
Cultivars
oercentaaei
YEAR
RAINY SEASON
HOT-DRY SEASON
NEW CULTIVARS
NEW CULTIVARS
I
I
I
I
Sahel
Sahel
Sahel
Sahel
Sahel
Sahel
108
201
202
108
201
202
1995
2
1
1
2
1
1
1996
4
2
2
5
2
2
1997
8
4
4
12
I
3
I3
19G8
16
6
8
22
I
4
I
4
1959
22
8
12
33
5
5
2000
25
10
15
40
5
5
2001
25
10
15
40
5
5
2032
25
10
15
40
5
5
20G3
25
10
15
40
5
5
2004
25
10
15
40
5
5
The price estimates were based on the economic outlook for rice,
and consultaticn with one of the authors of these reports. The
projected rice price over the next five years for Thai rice (35%
broken) would be approximately $200/ me:tric ton. We used world rice
prices of $200 and $250/metric tons and converted them to import
parity prices with an exchange rate of 487 FCFA/dollar, a
transformation rate of 66%, and estimation of the processing and
transportation costs.
These adjustments
resulted in an import
parity price range of 95 FCFA/kilo tc 119 CFA/kilo.
At the base line diffusion rates, the internal rates of return for
single cropping range from 66 to 72%.
With double cropping the
economic rates of return increase to 77 to 83% (see Table 7). One
of the main advantages of these shorter duration rice cultivars is
the ability to increase double cropping. There are additional labor
costs to get the one crop out and the other crop planted quickly
and we incorporated these costs into these estimates. Some farmers
may not be able to accomplish these additional activities without
mechanization. So additional study of the feasibility and the cost
of double cropping is appropriate.
17
Another advantage of the new rice cultivars mentioned by KX?DA
scientists is an improvement in quality over the traditional
cultivars. The addition of a price premium may be appropriate In
these calculations. However,
since there were no observations
available cn which to base this premium price estimate,
s,Jch
estimation was not done for this analysis.
Table 7.
Internal Rates of Return to Rice Research
(Ex-Ante)
for Different Expected Rice Prices With the
WARDA Diffusion Rates
No Area Expansion
Rice Price/Ton
Single Cropping
Double Cropping
(%I
(%I
95,000 FCFA
66
77
($200)
119,350 FCFA
72
a3
($250)
Note: 487 FCFA/US$ (1995)
Table 8.
Conservative Diffusion Rates for the New Cultivars
(50%
OF h3cnl;nn Fctim2tPc\\
fPFrrPntann\\
KI AJa3GAllLL Y-.lL&111L4
.-k-L,
,A
..A------^7-,
YEAR
RAINY SEASON
HOT-DRY SEASON
NEW CIJLTIVARS
NEW CULTIVARS
Sahel
Sahel
Sahel
Sahel
Sahel
Sahel
108
201
202
108
201
202
1995
l.G
0.5
0.5
1.0
0.5
0.5
1996
2.0
1.0
1.0
2.5
1.0
1.0
1997
4.c
2.0
2.0
6.0
1.5
1.5
1998
a.0
3.0
4.0
11.0
2.0
2.0
1999
11.0
4.0
6.0
16.5
2.5
2.5
2000
12.5
5.0
7.5
20.0
2.5
2.5
2001
12.5
5.0
7.5
20.0
2.5
2.5
2002
12.5
5.0
7.5
20.0
2.5
2.5
2003
12.5
5.0
7.5
20.0
2.5
2.5
2004
12.5
5.0
7.5
20.0
2.5
2.5
With the more conservative estimates of the diffusion rates, the
internal rates of return fall to 50% to 55% in the single cropping
case and 60% to 65% in the double cropping case (see Table 9).
These
are
still
very
credible
rates of
return
and
these
conservative estimates of diffusion are considered lower boundary
estimates. Hence, rates of return to the new rice cultivars appear
to justify the investments of WARDA, ISRA, and the farmers,
Table 9.
Internal Rates of Return to Rice Research (Ex-
Ante) With More Conservative Diffusion Rates
No Area Expansion
Rice Frice/Ton
Single Cropping
Double Cropping
(%)
(%I
95,000 FCFA ($200)
50
60
119,350 FCFA ($250)
55
65
487 FCFA/US$ (1995)
Only investments in Senegal on these rice cultivars were included
in the analysis as previous rice research done outside the country
was considered to be available to Senegal at minimal cost. The
rates of return then are from the perspective of investments made
in Senegal given the present state of rice technology in the rest
of the world. These esimates of the rate of return respond to the
following question: What are the rates of return to society in real
terms that Senegal will receive
from rice research based upon the
expected returns from the new cultivars developed and just starting
to be extended?
In the Senegal Valley there is substantial interest in further
extensions of the irrigated area. The analysis was also done with
land area expansion through increased irrigated areas by projecting
the past historic area expansion into the future. However, the
costs of this operation are so high that the combined investments
in new cultivars and expanded areas become marginal investments
except at the higher price and with double cropping. The costs of
this area expansion is an estimated 400,000 FCFA/ha.
This estimate
for drainage and land-leveling was obtained from farm interviews.
To this, must be added the additional variable costs of expanding
the land area.
Both irrigation and input costs were included as
additional costs of production in the analysis to account for the
expected increase in such costs as farmers begin to practice rice-
rice double cropping.
Irrigation costs are higher than input
costs, and account for 1.6% to 5.5% of the total amount earned from
rice production between 1995 and 2004.
19
Table 10.
internal Rates of Return to the Combined
Investments in the New Rice Technologies and Further
Extension of the Irrigated Area With the WARDA Diffusion
Rates
Area Expansion
Rice Price/Ton
Single Cropping
Double Cropping
(%I
(%I
95,000 FCFA ($200)
-
18
119,350 FCFA ($250)
10
28
487 FCFA/US$ (1335)
For the Sahels to be widely adopted, extension will be necessary to
inform farmers who have not been involved in WARDA farm trials
about
the existence
of the Sahels.
Extension will also be
important for providing
information on appropriate use of the
cultivars.
For example if farmers use Sahel 108 for rice-rice
doublecropping,
they will need information on soil and water
management and perhaps most importantly,
information about crep
calendars since doublecropping requires a precise crop calendar.
Improvements in access to credit and prompt payment for paddy will
be of critical importance in terms of allowing for rice-rice double
cropping and high yields.
Farmers must be able to purchase the
seeds and complementary inputs at the start of the hot-dry season
to allosJ for doublecropping.
Likewise, yields are quite sensitive
to the time of planting, as shown in Table 11.
Table 11: A-ierage Yields with Respect to Period of Seeding,
Wet Season 1393, Delegation of Dagana.
Period of Seeding
Average Yield (t/ha)
Before June 30
5.322
July
4.319
After Aug. 1
4.459
Access to credit will also be necessary to lessen the other major
constraint to doublecropping, seasonal labor shortages, as farmers
may need to borrow money to hire non-household laborers.
Aside
from its necessity to producers, a well functioning credit system
will also facilitate privatization of the rice sector.
Private
agents will need to build processing plants and storage facilities
or buy such sites from SAED.
These investments are costly.
.If -
20
credit is not available for such purposes or the interest rate is
too high, privatization will not tc a smooth process.
It is
expected that with market liberalization, the problem of delayed
payment for paddy will be eliminated, but the need for a well
functioning credit system will remain an issue.
Access to credit
is difficult in Senegal,
and this
is particularly true for
agricultural production.
Recently it was estimated that only 3% of
all
commercial bank credit was allotted to the agricultural
sector."
X: Conclusions:
The future rates of return to the development of these three new
cultivars are high even with the present land area and without
double cropping.
These rates are further increased with double
cropping.
Surprisingly, extending the irrigated land area is only
a marginal
investment combined with these new cultivars. This
results from the high costs of extending the irrigated area.It
appears to be worthwhile to examine these costs in more detail and
to identify the conditions under which it would be profitable to
extend the irrigated area.
The Sahel cultivars have higher yields than existing cultivars.
Sahel 108 is shorter duration than IKP and thus can increase the
likelihood of rice-rice double cropping.
The three cultivars are
of high grain quality, the position of the panicle leaf for Sahel
108 makes access to the seed by birds difficult, and Sahel 201 and
Sahel 108 have shown very modest decreases in yield with exposure
to salinity in ongoing experiments.
However, whether technological
change leads to increased productivity in the Valley will depend on
many interdependent factors: extension, credit services, changes in
marketing and processing, transportation infrastructure, azcess to
information.
It is hoped that researchers at ISRA and/or WARDA will carry out an
ex-post study of the Sahel cultivars.
Such an analysis would
provide more accurate figures for the rates of return to investment
on the new cultivars.
Likewise, such a study would allow for the
comparison between ex-ante and ex-post results.
In addition to
calculating the rate of return, future studies should also consider
distributional issues.
Substantial variation in adoption is
expected between regions of the SRV, among small and large farmers,
and by gender of the farmer.
It will be important to look at
distributional issues so that it may be determined if changes are
necessary
so that adoption of the cultivars does not lead to
regional, between household or within household inequities.
26 Robert R Nathan Associates. Ix., 1987. Annex 3.9.
21
XI.Bibliography:
Arcia, Gustavo, Mary Young Rice, Mamadou Sidibe, and Khardlata Lo
N'Diaye, "Senegal Grand Model Technical Description:
Improving Family Welfare Through Women's Work in Farming
Systems in Senegal", Research Triangle Institute, North
Carolina, April 1990.
Bragantini, C., "Etude Sur la Qualite des Semences de Riz a-~ I;ive~.~
des Paq'sans de la Eegicn de Saint-Louis", M.D.R.H., Pr,cjet
A.P.S., Dakar, 1990.
Comite perrr,anent Interetats de Lutte centre la Secheresse dans la
Sahel (CILSS),
"Atelier de Formation sur L'Impact Econcmique de
la Recherche Agricole", Mali, June 6-10, 1994.
Daniels, Lisa, Julie Howard, Mywish Maredia, James Oehmke, and
Richard Bernsten, "Assessment of Agricultural Research: Ex-
Post, Ex-Ante, and Needed Methodologies", Department of
Agricultural Economics, Michigan State University, no date.
Either, Carl K. and John M. Staatz, "Food Security Policy In Sub-
Sahara2 Africa", Procey-nas
-4i
of the Intermnal Conf.s r c .P
e&-.-e c'
a Turbulent WC.r--d
F;zonor!b' , Malaga, Spain, August 26 - September 4, 1985.
Gittinger, J. Price, EconQCic &l.&ysis of Acrr~cl1ltu.ra1 Pro-ects
I .
cecoqd Edltlcn I The Johns Hopkins University Press, Baltimoie
and London, 1982.
International Food Policy Research Institute (IFPRI), Accelerat~irs
an Bfrica edited by John ?:.
Mellor,
Christopher L. Delgado, and Malcolm J. Blackie, The
Jo'nns Hopkins Universit:? Press, 1987.
International Monetary Fund, uternational Fiual Statl.
. .l,s
=+ ' ,m I
Jan . - Feb. 1994.
Internaticnal Rice Research Institute (IRRI), Economic
c&J -
c
af,
a:.
'
Los Banos, Philippines, 1:?8.
Kite, Rod, "Senegal: Ilr.plications for the Rice Policy Dialcgue",
USAID/Dakar, Revised S&ptember 1993.
Martin, Frederic, "Food Security and Comparative Advantage in
Senegal: A Micro-Macro Approach", unpublished dissertation
submitted to Michigan State University, 1988.
Norton, Gecrge W. and Jeffrey Alwang, ljltroductlon to Econcfics c,F
ultural Develooment, McGraw-Hill, Inc., 1993.
Republique Francaise et Ministere de la Cooperation, aux de
. - llep du Sene g,gd, 1986.
22
Republique d~1 Senegal, "Conseil Interministeriel sur Le Plan
Directeur Rive Gauche", March 1994.
RepI;? ique du Senegal, "Plan Directeur de Development Integre Pour
la Rive Gauche de la Vallee du Fleuve Senegal: Rapport Final",
;pril I-31.
Robert R. Nathan Associates, Inc.: Clark Ross, Philip Boyle, Don
Harrison,
;
and Charles Nicholson, me Eifect s of Structural
2&1JqtvC-t irt Seye. &, Columbia, Maryland: USAID Africa Bureau,
1387.
. .
Rosen, Harvey S., &&lic: Finance, Second Edltlo,."1 Richard D. Irwin,
Inc., Chicago, 1988.
Sanders, John H. and Mohamed M. Ahmed, "The Economic Impacts of
Hageen Dura-I in the Gezira Scheme, Sudan", Proceed;Lnas of the
*-
mat 4 ona
O+-
let CRSP Conference, Corpus
Christi, Texas, July 8-12, 1991.
Sere, Carlos and Love11 S. Jarvis, "The Betting Line on Beef: Ex
>nte Estimates of the Benefits of Research on Improved Pasture
for the Latin American Tropics", Procwtias of the
University, New Jersey, July 6-8, 1988.
Sidibe,
Ma,m.adou, ISRA/BAME, "Effets des Politiques Macro-
Economiques sur le Niveau de Protection de la Filliere du Riz
Local Irrigue", First Draft, October 1?93.
Societe Nati onale d'Amenagement et D'Exploitation des Terres du
"ol+a
--i- (SAED) , "Estimation des Productions en Riz-Mais-Sorgho:
Hivernage 1993/94",
Saint-Louis, April 1994.
Societ e Nationale d'Amenagement et D'Exploitation des Terres du
Telta (SAEDj , "Estimation de la Production de Paddy dans la
r)elegation de Dagana: Hivernage 93/94", Saint-Louis, April
1994.
Societe Nationale d'Amenagement et D'Exploitation des Terres du
Delta (SAED) , "Politiques et strategies de developpement rural
dans la vallee et le delta du fleuve Senegal, Annexe 1:
=-rolution
-.
des superficies amenagees", Contribution de la DPDR,
February 26, 1994.
Societe Nationale d'Amenagement et D'Exploitation des Terres du
Delta (SAED), "Statistiques Agricoles: Evolutions de 1981 a
1991", Saint-Louis, May 1992.
USAID/Senegal/ADO,
"Senegal Agricultural Sector Analysis", January
1991.
__-.
23
WAFSA , "Annual Report", Cote d'Ivoire, 1990, 1321, 1992.
World Eank, Sub-SaWar. Africa: Fram Growth to q.3
WgrLd Bank, Washington D.C., 1987.
b;;crld E;nk, >;0-‘2 ppv~i -,
; 902.
TP.xlcz * n ip Cc:1 h
- I
Cxf zrd C'xiversity Fress, Inc., Ne,cJ York, NY.
24