Discussion Paper No. 7 17- cO
KENYA: MAIZE* DISCU S~10t ONLY-
Over a 15-year period Kenya carried out a maize production program
which resulted in the situation that today a very large part of maize area is
sown to high yielding hybrids**and Kenyn has become a not exporter of mize.
Several key factors contributed significantly to the success of the program.
First, research was aimed at developing a complete package of recommend1-
tions to maximize farmers yields and profits.
The first step was to produce a'higher yielding variety responsive to
better conditions but with a stable yield superior or at least equal to local
maize under poor condition. A comprehensive breeding system was developed which
was flexible, leaving the options open as to choice of hybrid or improved open-
pollinated variety. The system provided for continuous long-term yield improve-
ment: yield superiority of hybrids over local varieties progressed from an
average of thirty percent in 1961 to 60 percent currently.
The next step was to identify which factors affected yield most and to
study the interactions of those factors to arrive at the complete recommended
package. Table 8.3, considering only the main treatment effects, shows that
*This report is based on (28, 29, 30, 31, and 32). r-i'ic [arri'
helped from his first-nand experience in drafting this account.
**Gerhart, whose study was made West of the Rift, estimated that by 19'(
the use of hybrid maize had been extended to 325,000 hectares, accounting, for
Spprox imu.t, ly harolF of t.hli (roni]try's total pr'oduil, ion. In the higi h a.nd metdil
pi tt'iti l i.roai:; o i' i. HiI'l. V;ll l y, ;io tIwo-l.h i rd:; V ti.h :in, I I-: 1
1'armers had adopted the hybrid:s iinculudinog up tL( 90 wpercciiil o1' [,'i t';am rir :; in
1 h er altitude 'n d ri.i n al 1 on (P ) Hlii snn tir o t i n.I.d int
y tools ) seed sold in i19''() a.t. A (C) kilos per hofl.!,r,' wo ,n !;u f'.i ci en I, l
plant e~i bhlcct.n.res (since hiybri-d -,eed c o ts os- ('in- timoni the grani.n
price, it can be safely assumed that all seed bought is planted). Kenya
grows a little more than one million hectares of maize, but a substantial
percentage of seed sales is to other countries, especially Tanzania, Ethiopia,
RESULTS FROM ELEVEN DISTRICT HUSBANDRY TRIALS
IN WESTERN KENYA 1966 AND 1967
Differences Between Treatments
in in in in net
Yields yield returns cost profit
Factor Treatment tons/ha tons/ha $/ha $/ha $/ha
Time of planting Start of rains 5.8 2.4 96 3 3
4 weeks later 3.4
Plants per 39,000 5.1 1.0 4o 3 3
hectare 20,000 4.1
Type of maize Hybrid 5.4 6 6 60
Weeding Thrice, early 5.2 1.2 8 7 1
Once, late 4.0
P, kg/ha 56 1.7 o0. 4 1 -
at planting 0 4.6
N, kg/ha 78 4.9 25 -
top dressing 0 4.3
Note: Results from a. trials of 6 factors each at two levels, a 2 factorial
trial with a total of 64 treatment combinations.
(Maize selling price taken at then-current Kenya rate, equivalent
U.S. $110 per ton)
the factors which cost least (time of planting, seed weeding) tended to give the
greatest returns. The use of fertilizer by itself, on average of all treatment
combinat-'ins.;, resulted in financial loss. Figure 8.5, dj.: play-ing four treRt-
ment combinations of greatest interest, brings out clearly Lhe considerable
positive interaction among factors. Hybrid plus fertilizer increased yields by
1.3 tons/ha with poor husbandry (late planting, poor stand, and poor weeding)
but by 3.14 tons/ha with good husbandry. Similarly, good husbandry increased
yields by 2.92 tons/ha without hybrid and fertilizer but by 4.76 tons/ha with
them. When these four treatment- combinations were later widely used as demon-
strations on farmers fields, the'lesson to be learned was clear: hybrids and
fertilizers by themselves are no substitute for good farming and only give
worthwhile profit when combined together. If the full recommendations; were
followed, farmers could quadruple yield and get a good profit.
Second, there was no conventional and formal planning. There were,
however, frequent, though irregular, exchanges of ideas, mostly between indivi-
duals of the Maize Improvement Section and the Kenya Seed Company, and bets
were laid on attaining targets by set dates (with settlement in bottles of
beer). This informality provided a desirable flexibility to take advantage
of unforseeable opportunities as.they arose. Three types of the general
SFive aspects were considered most important--the breeding program,
agronomy research, extension, seed production, and market ngl ul,i l.i zat;ioin.
'The' vi. w w:i:: hl l iht:. lth. :; l t tl orf a n .l. i on l c.i'. prt,',;r:.! dc'ciih'd on l.1'
weakest link. 1 t was; consider d ijmporta nt to identify l.hi., wnkrest link
and concentrate effort accordingly.
Research was evaluated by its effect on farmers yields and national yield
levels, not by published papers and annual reports.
THE KENYA MAIZE DEMONSTRATION DIAMOND;
DATA OF FOUR TREATMENT COMBINATIONS
SHOWING POSITIVE INTERACTIONS
A. Previous Local Practices
Yield 1.97 tons/ha
B. Hybrid and Fertil-
izer only. Yield
Extra net over A.,
C. Better Husbandry only.
Yield 4.89 tons/ha
Extra net over A, 442
D. Better Husbandry Plus
Hybrid and Fertilizer.
Yield .8.03 tons/ha
Extra net over A, -$7f6
"(mrco: ,I:; T,,fldo 8.
SStrong and effective cooperation was formed with exLensionr staff by
seminars, demonstration plots, and district tours. Annual seminars were
started under the chairmanship of provincial extension officers. Research
staff presented their latest trials results and extension staff decided the
package of recommendations (from one such seminar the demonstration diamond
was born when trying to digest the implications of the district husbandry
trials). The .use of demonstrationplots was advocated and dummy field runs were
practiced during the seminars. The district tours enabled at least one
research officer each year to visit each district to maintain contact with
extension staff, to monitor progress and identify weak links; and weaknesses
found were then emphasized at the next seminar. Kenya Seed Company and staff
of other commercial companies played a full part in these seminars and policy
coordination activities. This avoided conflicting advice reaching the farmer
and the considerable strengths of commercial enterprise were combined and
harmonized with government effort.
Third, European settlers, with their own money and commercial initiative,
had provided the necessary infrastructure. Fertilizer, seed and other inputs
were available at the right time, in sufficient quantity, and at reasonable
cost. There was a marketing organization with a network of buying points
and a price fixed annually before planting that remained constant throughout
the year (with small fixed and known storage allowances). Local African
shopkeepers were encouraged to extend all. this network to the small eale
farming area at theI time when the mai o reearch .ckaf bIecaunt, nI.v.i I .hl T''
Kenya Ieed Company, w.i ll I..,t ontlanding o 'fii 'en'y, pl.niyel a r m.rkab I
role in this development. All hybrid seed is certified to the same or higher
standards as the International Crop Improvement Association, seed price is
cheaper than hybrids anywhere else in the world, and there are now over
5,000 registered se]liers. The Seed Company soon realized its main market
was with African small scale farmers and it become very active in good extension
and now employs its own maize breeder. Originally, the Company was entirely
private, being formed by a group of settler farmers, when research first
produced improved strains of indigenous pasture grasses; and legumes. Now
there are very 'few settlers left, most of the seed producers are African,
Government has taken a majority controlling interest, and continuity of manage-
ment has ensured improvement of standards. Bulk handling facilities have
been installed on the railway and at the port to facilitate exports. Large
amounts of maize are being used for livestock feed, especially for pigs and
poultry. Corn Products Inc. has built a maize processing factory. As
happened in the U.S. Corn Belt, efficient maize production is generating a
spread of other economic activity. The mutually reinforcing interaction of
the public and private sector played the same beneficial role in Kenya as
already noted in the U.S. and Japan.
Fourth. it was not necessary to overcome resistance to change by any class
of farmers. The large-scale expatriate commercial farmers initially
insisted on hybrids. When these farmers got the hybrids, a few small-scale
peasant subsistence farmers, seeing and appreciating the advantages, demanded
the hybrids for themselves. Against the advice of their extension officers.
they traveled long distances to buy the seed because distribution points were
i(1 p1IV i dh.'d in tlhi i 1 I 1i n i.i ': '1'h I'ic l l: I' hl.l 'i..- whli- :1 '"'p l.'d >.li,'
charl lenge, to prove tlhey were ju:-st a.-, ood farmers ar, the sef;tlc lr, p-rovii.('d
the first and best demonstration plots. Thereafter, the problem was to expand
seed supplies with sufficient rapidity. For the four years 1964-67 the
Seed Company almost doubled production each year and still were embarrassed
by seed shortage, were unable to meet the growing demanund, and did not dare
advertise. All major large-scale farming areas reached over 80% of their total
maize acreage planted to hybrids in six years from first introduction. West
of the Rift Valley, because of delays already mentioned due to seed shortage
and initial official opposition, the time required for 80 percent, of peasatn
subsistence farmers to adopt varied from eight years in districts ecologically
suitable for the hybrids, to eleven years in other districts. When medium
term hybrids were bred for East of Rift then participation of peasant subsis-
tence farmers of Center Province reached 80 percent in five years. This
compares to ten years for U.S. Corn Belt farmers. The rate of adoption
of fertilizer, from none being used on maize by subsistence farmers prior to
hybrid introduction, reached 50 percent recommended rate in ten yeaOrs west of Pift
in suitable ecologies. Fertilizer adoption rates in other areas were expected
to vary proportionately but it remains to be seen how the rates will be
affected by recent world price increases.
Farm size, educational level, income from cash crops, distance
from source of inputs, and work experience on large-scale commercial farms,
were all found to have had some effect on rate of adoption of some aspects
of the recommended hybrid maize package. In any free society, there always
will be some individuals who are good farmers and some who are bad farmers.
However, in Kenya it was found that none of the factors studied delayed
adoption by an average of more than two years.
Thle l';1rorn:; w ,ho gi'w hybri d i i o' now : ir, tl ,e:; wiro ,r'w Oli:,.
1 fwr'it--iin Ih\ e r'nle' :]'(nrl:. Thlre'r" wore ll lilt it. in : l l icd t. C l re pr'bl.1 :;
or any large class of landless labourers, ,so tl1.L hybri"dls have i ot broulghlt
about any major changes except increased yields. In some low altitude areas,
high maize yields have resulted in e aize monocult.ure repL.cing mixed
cropping with beans but this has not taken place on a large scale.
The two extremes of farm scale, expatriate commercial farmers whose
farms averaged about 500 hectares and peasant subsistence farmers averaging
P to 20 hectares, both adopted hybrids in no more year, than in Ith I1.. Corn.
Belt; A five poercont increased: .in Lota.l crop ca.unnied(l :la eep(: drop in price or
maize from 37/50 shillings per 200 lb bag in' 1967 to 28/- in 1968. This caur
a considerable reduction in acreage by large scale farmers but scarcely
affected the rate of increased uptake of hybrid by small scale farmers. The
small scale farmers have much lower production costs and, in predominantly
subsistence farming, are less affected by maize prices. If Kenya hybrid maize
technology is affected by size of farming operations, a case could be ac'rgue
that it has benefited the small scale farmer more than the .large scale
Fifth, some credit and some tractor nire uoLn governmnl'lc, Iand pr.iv;Lt, wve,'
available when hybrids were introduced. Neither were ever made available
specifically for maize and cases were all too common of the official schemes
having the opposite effect of that intended. Poor credit return rates were
followed by investigations which delayed availability of credit. Tractor hire units
would not begin work until credit had come through. Time of planting affected
yield more than any other factor. West of the Rift most farmers use ox-drawn
ploughs thi.t were i.nlt.roduced during food can.ipaigns or' I ho ;reconed world war.
lome oC ith hilghll(' y. el dN pet r unil t area we() ? I.ai ni d by It.,hins;e ( x;.p'.:l :;lt
farmers who told the wife to go take the hoe and get p:la:.nted on time.
Sixth, in both research and seed production, continuity of leadership was
remarkable. Michael Harrison as maize breeder and Alistair Allan as maize
agronomist each worked about fifteen years in the program. Festuis Ogaj.da,
after three years as maize breeder,took over from Harrison as Senior Maize
Research Officer, while Harrison stayed on as CIMMYT regional coordinator for
two years, a total transition period of five years; Ogada is still Director
of the Kitale station. Wim VerblnrghT1 and Ted Hazeldcn, managing director
and chief of the maize work, have been with the Kenya Seed Company throughout
the development of hybrid production. Hazelden was once assistant maize
breeder and his transfer typifies the close cooperation between the Maize
Improvement Section and the Kenya Seed Company.
Seventh, effective use was made of strategically situated foreign tcjnical
assistance to supplement local resources. The Kenya Government recruited
their first full-time.maize research worker, a maize breeder in 1955, and
British government assistance soon afterwards extended this effort to other
ecological areas. Promising results attracted a Rockefeller Foundation grant
to add an agronomist and a U.S. AID/USDA supporting project of a comparative
study of maize breeding methods. Both the latter additions were also designed
to enable the Kenya Government to work for Eastern Africa as a whole, from
Ethiopia to Malawi, and start a cooperative regional program. CIMMYT assisted
the regional program for a short period. The British government added an
agronomy team of four expatriates after the Rockefeller grant ended. Mean-
while the Kenya Government established and built up the Maize Improvement
Sectti.on. Tt operal.s from threee main research ntationii: in li.f ferent cc ol gi cn.al
".('. n b lu l. i ::1.i ll i I ,m ,p:or:i.l.i w 'ly :'iy :ill rv ,r;r'iu, w i l.l i ;i,', i i :;u .'.i l.l2 i '
th:u0 l) e (- c.in ,ui i.$ri( luin s\ 1 1 iy i. uctl. i i Vi1 1'1 :mlh] I I I '
single source of funds was a research grant from the Mliize & PTroduce IMarket,.inr
Board. This grant involved farmer membership on the Cereals Research
Advisory Committee and so ensured close farmer participation right down from
the highest budget and policy level.
Large capital investment was not needed in any aspect of developing
hybrid maize since full advantage was taken of existing resources and organ-
izations. Outside money for research was considerable but supplemented
existing Kenyan research and was used partly to develop regional, research
for other countries. Ways were found of cooperating effectively with the
extension service. Commercial companies supplying the infrastructure financed
expansion mainly out of profits. Large scale farmers had mechanization
and small scale farmers did not need it to adopt hybrid maize. Credit to
farmers played a negligible role. The quality of the technology was such,
and the infrastructure already operating was such, that it would have been
difficult, once started, to stop hybrid maize in Kenya.
Some adoption for other crops has taken place based on methods that
were developed for maize. Demonstration plots for cotton were improved and
extension linked more closely to research. Other countries have also picked
up ideas. The comprehensive breeding system, or modifications of it, were
adopted.in Tanzania, Malawi and Zambia. The demonstration diamond was taken
up by CIMMYT and, through their advocacy, by other countries.
The development of hybrid maize in Kenya is in no way a "forced-pace
campaign" but is included, here as an example of development without planning.
It in a. pri me example off what a national program can a ch-ivev wit ih a miniilmulm f"
()Il. :i dc al-,;; i !;-I-tanct .
(28) John Gerhart,, The Di ffnsion of Hybrid Maize in Western Kenya (Londr.es,
Mexico: CIMMYT, 1975).
(29) M. N. Harri son, et. al ., "How Hybrid Seed i. ; Revolnt ion izingl Maize
Growing in Kenya," International Developmcnt Review, vol.. 10, no.
( ISqB) 9-13.
(30) G. F. Sprague, "Factors Affecting the Adoption of Hybrid Maize in the
United States and Kenya," in Change in Agriculture, ed. by A. H. Bunting
(31) M. N. Harrison, Chapter 1 in Crop Improvement in East Africa, ed.
by C.L.A. Leakey. Commonwealth Agricultural Bureaux, U.K.
(32) S. A. Eberhart and G. F. Sprague, "A Major Cereals Project to Improve
Maize, Sorghum and Millet Production in Africa," Agronomy Journal, vol. 65