ti.
5:4
. {.35
5.3..
:15
’13:}...
11:11..)
‘355i3t‘3rbi
.a..::..xx:::)..,x
ii): .5:
.Iv
,-

35.1.2;
12)....
xINNS»;
.3
:5....!.3I..I§93:.
.. v. :3:an .1... A:
551...: ! n.roa..!i
{53:32: ..:1\..:)..)
31:1! 2. x i.
3.53.. .3... a: a!
a...
$31.41 )3
{in}
3...
1391:3231
9.:
at Hannah?!»
.5 .12
{323‘

V at
. 31“}, T
3! n): {4 av
22:45... {5.33. .
, (11.513...)

I...

A .i i.

,n)...21

 

ERSITY LIBRARIES

lllllllllllllllllllllll HHH II
n 3 1293 01

020 6203 r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

/ thesis entitled
(HE PAVOPF To (“{ng vw‘HzE ZGQWCH
AND foeuguow no 2mgfiswg : Aw egowvmc.

FWD H’LSTIT‘IKTIQN m, MAL‘I—gl 3
presented by

KWDWMA , gaze/11w

l
has been accepted towards fulfillment ‘
of the requirements for

W4. 0 S 1 (1953er in W49 Eamon/tics

fwfifiw

[ Major professor

Date El 30 W

 

0—7639 MS U is an Affirmative Action/Equal Opportunity Institution '-

 

 

LIBRARY I
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove We checkout Irom your record.
TO AVOID FINES return on or before date due.

  

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU le An Affirmative Action/Equal Opportunity Inetltution
Warns-9.1

 

THE PAYOFFS TO HYBRID MAIZE RESEARCH AND EXTENSION IN ZIMBABWE:
AN ECONOMIC AND INSTITUTIONAL ANALYSIS.

BX

Bernard Kupfuma

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of
MASTER OF SCIENCE
Department of Agricultural Economics

1994

ABSTRACT
THE PAYOFF TO HYBRID MAIZE RESEARCH AND EXTENSION IN
ZIMBABWE: AN ECONOMIC AND INSTITUTIONAL ANALYSIS
By
Bernard Kupfuma

This thesis evaluates the rate of return to public investment in hybrid maize
research and extension from 1932 to 1990. The economic surplus approach (an ex post
technique) is used to assess the returns using data collected in Zimbabwe in 1992. The
average annual internal rate of return was found to be 43.5 percent which means that
every Zimbabwe dollar of public expenditures on maize research and extension over the
1932-90 period earned a return of 43.5 cents. This is an impressive rate of return which
is higher than the return from other capital investments. The benefits were captured by
consumers, the bulk of them being communal farmers producing primarily for home
consumption.

This study also examined a number of institutional factors. Findings from the
institutional analysis recommends leaner public maize research teams compatible with
budget bottlenecks; retention of gifted scientists for long periods; encouraging communal
farmers to lobby for agricultural research and extension in the political arena; and that
public maize research and extension programs should be retained but with more
emphasis on basic public responsibilities such as maintenance research and production

problems in marginal communal areas.

ACKNOWLEDGEMENTS

I am heavily indebted to my major advisor, Professor Carl, K. Eicher who
diligently and enthusiastically guided me through my study program. My thesis advisor
Professor Eric, W. Crawford, meticulously and patiently helped me with the technicalities
of calculating the research and extension payoffs. Professor Richard Ward helped ensure
that the agronomic aspects of maize research are considered in the study.

I am also indebted to Drs Derek Byerlee, Ruben Echeverria, Johannes
Roseboom and Philip, G. Pardey and Professor Mandivamba Rukuni for taking time to
attend to me and enlighten me on the various issues on rate of return studies. I drew a
lot of insights from their vast experience and knowledge. Mr Rex Tattersfields, Head of
Rattray-Arnold Research Station assisted me understand the past, present and future of
maize research in Zimbabwe.

The Kellogg Foundation and the Government of Zimbabwe funded my study
leave and made all this possible. Dr Enos Shumba, Dr Ephrem Whingwiri and Mr Ron
Fenner (Department of Research and Specialist Services) made it possible for me to
take a study leave and pursue an MSc degree. Dr Shumba, Mr Fenner, and Mrs Danisile
Hikwa (Agronomy Institute) became my field advisors when I was collecting data for this
study. I also express my gratitude to fellow officers in the Agronomy and Crop Breeding
Institutes for their encouragement and interest in the study. I single out maize breeders,
Mr Takesure Rukweza and Mr Lewis Machida for providing me with detailed
information on maize breeding. I also acknowledge the help I got from the Central
Statistical Office Library, National Archives and the Ministry of Lands, Agriculture and
Water Development’s Central Library.

I am grateful to students, faculty and staff in the Department of Agricultural
Economics at Michigan State University who in so many ways made my studies a
pleasant experience. My last, but by no means least, thanks go to my mother and father
whose love I thrived on even when so far away. To my brothers John, Benedict, and
Farai ”Rhodzi" and sisters Pepetua, Margaret, and Miriam "Makebo" , this would have

not been possible without your support and encouragement since 1971!.

 

DEDICATION
Mukoma Wilfred Nyasha and Mukoma Linos "Mwana waMwari", my dearest brothers

this is for you.

TABLE OF CONTENT S

LIST OF FIGURES .................................................. iv
LIST OF TABLES ................................................... vi
CHAPTER I: INTRODUCTION ........................................ 1
Background ................................................... 1

An Overview of Zimbabwe and its economy .......................... 2
Location and the natural resource base ........................ 2

Agriculture in the National Economy .......................... 3

The Evolution of Agricultural Research ........................ 8

Problem Statement ............................................ 12
Objectives ................................................... 14
Organization of the Study ....................................... 14

CHAPTER II: OVERVIEW OF MAIZE RESEARCH AND SUPPORT

INSTITUTIONS .............................................. 15
Introduction .................................................. 15
Maize Research in Zimbabwe .................................... 17
Breeding .............................................. 17
Agronomy and Crop Management ........................... 19
Support Institutions ............................................ 22
Extension .............................................. 22
Seed Production and Marketing ............................. 23
Maize Marketing ........................................ 25
Credit ................................................. 27
Other Institutions ........................................ 28

 

Summary .................................................... 29

CHAPTER III: MEASURING THE PRODUCTIVITY OF AGRICULTURAL

RESEARCH AND EXTENSION ................................. 31
Introduction .................................................. 31
Ex Post Evaluation Approaches ................................... 32
The Economic Surplus Approach ............................ 32
The Production Function Approach .......................... 36
Impact on National Income Approach ........................ 38
The Nutritional Impact Approach ........................... 39
Ex Ante Evaluation Approaches ................................... 39
Rate of Return Studies in Africa .................................. 40
Evaluations of Payoffs to Maize Research in Zimbabwe ................ 43
Model Specification and Assumptions ........................ 44
Graphical Representation .................................. 46
Algebraic Representation .................................. 48
Summary .................................................... 50

CHAPTER IV: RETURNS TO HYBRID MAIZE RESEARCH AND

EXTENSION ................................................ 51
Introduction .................................................. 51
Calculating the Rate of Return ................................... 52
Data Collection ......................................... 52
The Shift Parameter ...................................... 54
The Demand and Supply Elasticities ......................... 58
Technology Generation and Transfer Costs .................... 60
Market Distortion Induced Costs ............................ 62
Results and Discussion ......................................... 65

Returns to Research Investments ............................ 65

Sensitivity Analysis ....................................... 68
Distribution of Benefits Among Producers ........................... 70
Summary .................................................... 75

CHAPTER V: INSTITUTIONAL FACTORS INFLUENCING THE

PERFORMANCE OF THE MAIZE RESEARCH SYSTEM ............ 77
Introduction .................................................. 77

Size of the Maize Research Team ................................. 78
Financing Agricultural Research: The Role of Client Organizations ....... 81

The Role of Commercial Farmers’ Organizations ............... 83

The Role of Communal Farmers’ Organizations ................ 88

Management of Human Capital in Agricultural Research Institutions ...... 92
Recruitment, Rewarding and Retention of Researchers ........... 92

Managing the Scientific Discovery Process ..................... 97

The Public-Private Sector Division of Labour ....................... _. 98
Summary .................................................... 103
CHAPTER VI: CONCLUDING REMARKS .............................. 107
REFERENCES ..................................................... 112
ANNEX A: THE DATA .............................................. 124
ANNEX B: THE RESULTS ........................................... 129

 

LIST OF FIGURES

Figure 1.1: Zimbabwe in the Southern African Sub-Continent ................... 4
Figure 1.2: Zimbabwe: The Agro-Ecological Regions. ........................ 5
Figure 1.3: The structure of the Department of Research and Specialist Services . . . 11

Figure 3.1: Social benefits from a downward shift in the supply curve due to
agricultural research. ........................................... 33
Figure 3.2: Zvi Griliches’ models for measuring the returns to hybrid maize
research in the United States. .................................... 34
Figure 3.3: The impact of technical change. ............................... I 47
_ Figure 4.1: Zimbabwe: Shifts in maize supply functions for communal and
commercial areas, 1950-1990. ..................................... 57

Figure 4.2; The impact of technical change on the cost of target pricing and

consumer subsidies in Zimbabwe: A closed economy model. ............. 63
Figure 4.3: Trend in real maize prices, 1945-1990. ........................... 64
Figure 4.4: The impact of technical change on farmers’ cash income. ............ 72

Figure 4.5: Zimbabwe: The effect of maize hybrids on annual farm cash income
by sector, 1949-1990 ............................................ 74

Figure 5.1: Zimbabwe: Real research and extension expenditures from 1954 to

1990 (1980= 100) ........................................ 85
Figure 5.2: Zimbabwe: Expenditures in extension by sector from 1950 to 1982 ...... 90
Figure 5.3: Zimbabwe: Salary ratios for researchers from 1930 to 1964. .......... 95

iv

LIST OF TABLES

Table 1.1: Zimbabwe: Percentage Distribution of Land Area by Natural
Region and land use. ...................................... 6
Table 1.2: Zimbabwe: Sectoral contribution to GDP, 1970 to 1988. ............... 7
Table 2.1: Zimbabwe: Percentage Adoption of Hybrid Maize Seed by
Commercial and Communal Farmers 1945-1985. ...................... 19
Table 2.2: Zimbabwe: Sources of increases in maize yield, (1950 to 1980) ......... 21
Table 2.3: Zimbabwe: GMB Maize Intake From Smallholders, Percentage of
Total Grain Intake and Rainfall, 1981-89. ........................... 27
Table 3.1: Summary results of rate of return studies in Africa. ................. 42
Table 4.1: Zimbabwe: The shifts (k-factor) in the maize supply curve for
communal and commercial farmers, 1950-1990. ...................... 56
Table 4.2: Zimbabwe: Gross benefits and research and extension expenditures,
1932-1990 (’000 ZS). ........................................... 66
Table 4.3: Zimbabwe: Returns to investments in maize research, 1932-1990. ....... 67
Table 4.4: The effect of changing supply and demand elasticities on the internal
rate of return. ................................................ 68
Table 4.5: The effect of reducing the supply shift parameter by half on measures
of payoffs to research and extension. ............................... 70
Table 4.6: Zimbabwe: Average annual change in farmers’ income by farm type at
different total supply elasticities levels. ............................. 75
Table 5.1: Annual Research Budgets per Researcher in Four African Countries,
1961-1985. ................................................... 80
Table 5.2: Conditions of service for researchers in Zimbabwe (1984), Dominican

Republic and Thailand (1983). .................................... 93

W
INTRODUCTION

mung

The performance of Sub-Saharan African (SSA) economies has been
deteriorating since the 1970’s (World Bank, 1990). In fact, in many countries living
standards have fallen below the 1960s’ levels. Because of the dominant role of
agriculture in most SSA countries, the road to economic recovery must begin in the
agricultural sector. Agricultural research to develop better production techniques is one
of the prime movers in "Getting African Agriculture Moving" (Eicher and Rukuni, 1990).

The role of agricultural research in agricultural grth and development has been
documented in numerous studies in industrial countries and in Asia and Latin America‘.
These studies have shown that investments in agricultural research generally have high
rates of return. However, there are only a handful of published and unpublished studies
of returns to research in Africa (Oehmke, et. al., 1992). Moreover numerous studies
point out that most African national research systems are under-funded, poorly
organized and managed, and inadequately focused (Mudirnu, 1986).

Zimbabwe’s impressive agricultural research experience is well known throughout
the World. Between 1900 and 1930, the country laid the foundation of a productive
agricultural research system, developed intensive research programs from 1930 to 1950
and began to reap the flows of benefits in the 1950’s (Rukuni, 1994). Maize research has
achieved an outstanding record of success. Zimbabwe was the second country after the
United States to develop hybrid seed in 1949 and in 1960 it became the first country in
the world to release a single cross hybrid maize variety (Tattersfields and Havazvidi,
1994). The research system also developed improved varieties and production practices
for other crops, notably cotton, wheat, soybeans, tobacco, groundnut and sunflowers

(Weinmann, 1975). Unlike many other African countries, staff turnover in the national

 

1 See Eicher, 1990a, 1990b and Tattersfields, 1982 among many other studies.

1

 

 

agricultural research system (NARS) was low before independence in 1980. The maize
breeding program had only three changes in scientific leadership between 1930 and 1988
(Rattray, 1988). Until the 1980’s government support to agricultural research was "strong
and consistent" (Tawonezvi, 1994; Avila, et al, 1989). A relatively high percentage of the
funds allocated for agricultural research was used to finance operating costs (Mudimu,
1986). By contrast, in many other countries 80-95 percent of the total research budgets
is used to pay salaries and wages.

The payoff to investments in agricultural research and support services in
Zimbabwe has been spectacular in terms of output and exports. Maize yields increased
by 312 percent between 1946 and 1980 (Tattersfields and Havazvidi, 1994). Zimbabwe
was a net exporter of maize for 20 out of the 22 years from 1970 to 1992. Although the
distribution of the gains from research has not been equitable across natural resource
regions and farm types, the contribution of communal farmers to marketed output
increased sharply after 1980 (Kupfuma, et al 1992). Divisible on-shelf maize
technologies is one of the major factors responsible for increasing communal maize
production. Virtually all communal farmers have adopted maize hybrids (Tattersfields
and Havazvidi, 1994). Without question, these achievements would not have been
possible without a well-managed and productive agricultural research system and efficient
supporting services. But to date nobody has calculated the returns to past investments in

agricultural research and extension for any commodity in Zimbabwe.

An Overview of_ Zimbabwe and its economy
Location and the natural resource base
Zimbabwe is located between 15 and 22 degrees south, latitude and 26 and 34
degrees east, longitude. Figure 1.1 shows the location of Zimbabwe in the southern
African sub-continent. It has a total land area of approximately 390,700 square

kilometers (150,800 square miles). In 1992 the population was estimated to be around
10.5 million.

 

Three altitudinal planes dominate the natural landscape. At 1,200 to 1,500
meters above sea level (m.a.s.l.), the high-veld plateau runs across the country from the
south-west to the higher and mountainous eastern edge. This plateau forms the
watershed region and experiences higher rainfall and cooler temperatures. Surrounding
this watershed region is an area of between 900 and 1,220 m.a.s.l. called the middle-veld.
The basins of the rivers that borders the country to the south and north (Limpopo and
Zambezi, respectively) form the low-veld region lying between 300 and 900 m.a.s.l.. This
region experiences lower rainfall and hotter temperatures.

Climatic factors are important as natural resource base quality determinants.
The detailed classification of the country into five agro-ecological regions primarily based
on rainfall is shown in figure 1.2. Table 1.1 shows the skewed distribution of land in
favor of large scale commercial farmers. Over 74 percent of the land in the drier regions
(IV and V) is under communal farming.

The hot and wet growing season runs from late-October to mid- or late-March.
The rainfall season is unirnodal. The other major season is the cool and dry winter
running roughly from April to early September. The length of the rain season and its
quality of the season vary considerably from year to year. The quality in terms of
rainfall, declines from region I to region V as shown in figure 1.2. As a result, about 40

percent of the country’s land is classified as arable, and 55 percent is suitable for grazing.

AM“ in the National Economy
Records by early missionaries indicate that "Crop production was the most
important single economic activity of the peoples of Matebeleland and Mashonaland...”
(Reid, 1977). When the European settlers arrived in what is now Zimbabwe in 1890
they were forced to turn to agriculture after their failure to discover large mineral
deposits. While the contribution of agriculture to GDP has declined over the years
(table 1.2) the economic performance of the economy is highly dependent on the

performance of the agricultural sector (Chidzero, 1990).

 

 

   

m/ [fl Ell IIUPIU

entral AfrDan
Republic

/

Somalifl/
«5 ”0
;ason 0° Zaire Kenya
$“' Rwanda
'

Cameroon

V Burundi
Tanzania

Angola / ’44

    
     
      

Malawi
Zambia

\\

‘ Mozambique

i‘ZIMBABWE

Namibia

/

  

Swaziland
/\_/

 

 

 

Figure 1.1: Zimbabwe in the Southern African Sub-Continent.

 

 

       

/fl/I/H/fi//”IM //W////¢//////
7/// 2%fl I/lZ/l/

.fl

          

’4!/!.

      
  

  

////

      

//

/
//

                  

//

/
4‘

W
W/

I

_%Z/
W. /

    

 

m.
mmm
Am...

°“" 950m
600

mm
W.W.M

numm:

 

 

 

Table 1.1: Zimbabwe: Percentage Distribution of Land Area by Natural

Region and land use.

 

 

 

 

 

 

 

 

 

 

 

 

Item ll 1.8

 

 

 

 

 

Land Use N u a R e g 1 o n
I II III IV V X II Total
I:ommunal Areas If 0.2 3.1 7.0 18.1 12.8 1.5 i 42.7
Large-scale Areas 1.2 9.8 5.6 9.7 5.7 0.1 32.1
Small-scale Areas - 0.6 0.7 1.5 0.5 0.2 3.5
Resettlement Areas 0.1 0.4 2.4 0.5 1.1 - 4.5
National Land 0.3 0.9 2.1 6.5 6.0 1.4 17.2
14.8 17.8 36.3 26.1 3.2

 

 

Source: Avila et al (1989).

Thus agriculture’s contribution of 15 percent to the GDP masks its strategic importance

in the national economy.2

The agrarian structure of Zimbabwe consists of four major types of farms.3

These farm types illustrate the unequal nature of the agrarian structure.

 

2 Agriculture contributes over forty percent of Zimbabwe’s

foreign currency

earnings, employs over thirty percent of the labour force, directly supports over seventy
percent of the population, and supplies raw materials to several manufacturing industries.

3 The terms commercial, and communal areas will be used to refer to large-scale
commercial areas and smallholder areas respectively. In terms of total area these are the

major farm types.

 

Table 1.2: Zimbabwe: Sectoral contribution to GDP, 1970 to 1988.

 

 

 

 

 

 

 

 

 

 

I Year II Agriculture Manufacturing Services Others

1970 I 14.7 34.0 45.5 6.3

1973 13.8 36.6 42.9 6.6

1976 16.2 35.9 43.3 4.7

1979 11.4 35.9 46.6 6.1 "
1982 12.9 31.0 45.7 10.4

1985 13.7 36.3 39.5 10.6

1988 . 11.0 37.0 40.0 12.0

 

 

 

 

 

 

Source : Shapouri and Missiaen (1991)

a) Large-scale commercial farming areas At independence in 1980, there were about 5,000
commercial farmers under freehold title but this number has declined to around 4,000 by
1990. The average farm size is around 2,200 hectares. This farm type covers 40 percent
of the total area of the country and, in the early eighties, it contributed about 85 percent
of the total marketed agricultural output in the country.

b) Small-scale commercial farming areas This farm type consists of around 8,700 farms
with an average farm size of 124 hectares. These farms are held under freehold or
leasehold title. The sector covers only 4 percent of the total area of the country
(Mudimu, 1986).

c) Communal farming areas This farm type occupies about 42 percent of the land. The
population density is highest in these areas averaging around 25 persons per square
kilometers. About one million households make up this sector. The average land area
available to each household is about three hectares. There is no legal title to land but
each household has use rights to cultivate the land and access to communal grazing land.
Most of the communal areas are in unfavorable natural resource regions where

agricultural management practices are inferior to those in other farming systems.

 

d) Resettlement farming areas These types of farms emerged after 1980 on land purchased
by the government from large-scale farmers. About 41,000 of a planned 162,000
households chosen from densely populated communal areas have been resettled on an
individual or cooperative basis. The land is owned by the government but settlers enjoy
life long use rights. In 1990 these farm types occupied about 5 percent of the total land
area in the country.

Since about 4,000 commercial farmers control about half of the arable land in the
country, studies of technology adoption must address the issue of the distribution of
benefits from agricultural technology development programs. The task of developing
appropriate technologies for the different farm types is the major responsibility of the

Department of Research and Specialist Services (DR&SS).

The Evolution of Agncu_lt_u__ral° Research
The Department of Agriculture, formed in 1897, deve10ped the basic agricultural

research infrastructure over the 1900-1940 period. After the Second World War,
lucrative markets for agricultural exports opened up in Europe, giving rise to an increase
in demand for new farm technology. The formation of DR&SS also came at a time
when there was an increase in agricultural activities caused by an increasing urban
demand for agricultural products (Weinmann, 1975). Thus the establishment of DR&SS
can be partially viewed as a demand-driven institutional innovation.

The DR&SS is the primary public agricultural research organization in
Zimbabwe. It is entrusted with the welfare and technical progress in agriculture (Avila,
et al, 1989). The department is under the Ministry of Lands, Agriculture and Water
Development (MLAWD). Its structure is outlined in figure 1.3. When DR&SS was
formed in 1948 , it was organized along disciplinary and commodity lines ranging from
tobacco to horticulture. Disciplinary programs ranged from soil chemistry to animal
husbandry. The thirteen branches included, Animal Husbandry, Botany, Conservation,

Crop Production, Dairy, Entomology, Extension, Horticulture, Pasture Research, Plant

 

Pathology, Poultry, Tobacco, and Soil Chemistry. The interests of the large scale farmers
were articulated to the researchers through commodity committees of farmers and
members of the industry at large. The evolutionary pathway of the department was
strongly influenced by the emergence of a powerful commercial farming sector.

From 1948 to 1970 DR&SS released improved varieties and crop and livestock
production practices (Tawonezvi, 1994; Mudimu, 1986). During this period, the
department established a tradition of excellence in research (Avila, et al, 1989).
Agricultural research was managed at a federal level from 1953-63 during the Federation
of the Rhodesias (now Zambia and Zimbabwe) and Nyasaland (now Malawi).

In 1970 the Agricultural Research Council (ARC)‘ was formed to give farmers
and other interested parties a greater role in the planning and setting research priorities.
The mandate of the ARC was to: a)- strengthen the national agricultural research system
for commercial farmers; b)- improve the nutritional quality and yield of the basic food
crops through breeding and plant protection; and c)- improve animal health and
production primarily for the commercial farmers (Tawonezvi, 1994, p. 7)

The ARC freed DR&SS from the bureaucracy of the civil service by placing the
management of research expenditures in the hands of commodity sub-committees in
which farmers played an active role (Tawonezvi, 1994). In 1971, a special ARC sub-
committee recommended that DR&SS be turned into a semi-autonomous organization
but the proposal was not approved. However, the special sub-committee was influential
in pushing through a reorganization of DR&SS that improved the administrative
efficiency and the coordination of research activities. The thirteen branches were
consolidated into three divisions: Crops Research, Livestock and Pastures Research and
Research Services. The research stations and institutes were given specific mandates

based on a particular commodity or discipline.

 

‘ This was part of the Agricultural Diversification Scheme which was launched in
1967 in response to the imposition of economic sanctions following the Unilateral
Declaration of Independence (UDI) in 1965.

9

 

 

 

 

may and soil

leeeuch Institute _ cotton leseuch xutituu- Kuh- ls “to”. Is _
__ Coffee as
Plant Murine
__ lexticultuxe and out: bureau“. 1c Deity Service- —

leseesch institute

lstianl lezhuiu Hanna: .3 me 6,3.qu __
end laconic Guden 4211me Is Services

”on: Protection Weld useuch Stations mismje 63...],m u —
Ieseuch Institute Snbi Vsiiey

—— Seed Services 1d P
m". ” Ve a nature —

l— uop Breeding Institute—C Geebi “a“, we Research
00's It Rune. Henderson. mun —

hoe-s, mono]! end utopos

_ Iio-etr ics Bureau

__ Lem-e Inneculetion factory

 

 

_ mm Institute —[ -
Pen-use Henderson Its —

 

 

__ xntomtion Services

 

 

Figure 1.3: The structure of the Department of Research and Specialist Services

Independence in 1980 was followed by a sudden departure of experienced staff, a
slow recruitment of recent university and agricultural college graduates and a shift in
government policy to place more emphasis on the research needs of communal farmers
(Republic of Zimbabwe, 1983). Because the communal farmers face production and
institutional constraints that are more technically challenging to researchers than those of
commercial farmers, DR&SS’s institutes and stations, with the assistance from
international agricultural research centers, initiated on-farm research programs in 1981.
These on-farm research programs have successfully brought researchers to the farmers’
fields, improved research-extension-farmer linkages and channeled problems of
communal farmers to on-station researchers. But a 25 percent reduction in the
DR&SS’s real (inflation-adjusted) budget from 1980 to 1989 forced DR&SS
administrators to scale back on-farm research activities (Shumba, 1990; MacMillan et al
1991).

Moreover a 1981 Act of Parliament stripped the ARC of its statutory powers to
allocate research funds and created an Agricultural Research Fund (Tawonezvi, 1994).
The Fund is under the control of the Director of DR&SS on behalf of the Secretary of
Agriculture. Thus the role of commercial farmers, who can only express their views

through the ARC, is now a purely advisory one.

Problem Statement

Because of the budget crisis facing DR&SS and the rising skepticism about the
need for research, there is an urgent need for information on the payoff to agricultural
research. At the root of this crisis are several factors, including a decline in farmer
support for research. Several examples illustrate the commercial farmers’ skepticism
about the ability of DR&SS to maintain high quality research programs after
independence. With this fear in mind, commercial farmers made two key decisions to
ensure that their research needs would continue to be met. First, in 1974 commercial

farmers pooled their resources and purchased a farm 30 kilometers north of Harare and

11

converted it into a research station (Tattersfields, 1992). Second in 1981 the Agricultural
Research Trust (ART) farm, an independent research body, was established by
commercial farmers in order to carry out adaptive research and fine tune
recommendations for commercial farmers. Several commentators view the establishment
of the ART farm as a symbol of a loss in confidence by commercial farmers in the
productivity of DR&SS (Mudimu, 1991). Moreover, the ART farm recently started
conducting research trials in the communal areas in direct competition with DR&SS.

Zimbabwe’s national agricultural research system faces some enormous
challenges in maintaining its momentum and meeting the new challenges of the 1990’s.
The population growth rate is estimated to be around 3.0 percent (Central Statistical
Office, 1989). Blackie (1984) contends that because the country has closed its land
frontier, yield enhancement research should be intensified to meet the increased food
demand in the next twenty years. Maize yields have been growing at a modest rate of
only 0.3 percent over the years 1970 to 1980 (Roth, 1990). Thus new technologies are
necessary in the 1990’s and beyond, not only to increase food production, but also to
ensure stability in supply across highly variable seasons and agro-ecological regions.

To meet these challenges, the agricultural research system in Zimbabwe has to
deal with the following issues: a)- retain experienced staff and recruit new staff members;
b)- maintain existing programs in the face of a decline in the real budget for research;
c)- maintain the productivity, innovativeness and flexibility of the research system; c)-
effectively implement Growth with Equity policies.(ISNAR, 1988).

To address these concerns, research is urgently needed on the payoffs to past research
efforts, including an analysis of the institutional factors that shape the scientific discovery
process and the financing, productivity and sustainability of the research programs. This
information will help inform decision-making on priority setting, resource allocation and
research organization as well as the debates on the potential role of research in

increasing agricultural productivity.

12

 

Obiectives

The general objective of this study is to evaluate the returns to investment in
hybrid maize research and extension in Zimbabwe from 1932-1990. The specific
objectives are to;

(1) - present an overview of the maize sub-sector in Zimbabwe,

(2) - critically review the literature on methodologies to assess returns to

investment in research.

(3) - calculate the rate of return to investment in hybrid maize research in

Zimbabwe between 1930 and 1990.

(4) - identify the institutional factors that shape the process of technology

generation and diffusion.

(5) - draw some insights for the effective design and implementation of maize

research in the 1990’s.

This analysis will be based on data and information that was collected in
Zimbabwe during June to September 1992. Most of the data were obtained from
archival sources, interviews with key informants and secondary sources. The analysis
involves the use of an economic surplus approach in estimating returns to investment in

maize research.

W

The study has five chapters. The second chapter presents an overview of the
maize sub-sector and supporting institutions. The third chapter evaluates various
methodologies to calculate the rate of return to investments in maize research. The
fourth chapter presents the results of the study on returns to maize research. The
objective of the fifth chapter is to examine the institutions that shaped the technology
generation and transfer during the period 1932 to 1990. Chapter six summarizes the

study and draws implications for the future of maize research.

13

 

 

MILK.
OVERVIEW OF MAIZE RESEARCH AND SUPPORT INSTITUTIONS

Introduction

White maize from South America was introduced into southern Africa by
Portuguese and Arab traders around the fifteenth century (Reid, 1977). Today white
maize is the major food staple and it supplies around half the calories in the average diet
in Zimbabwe. Moreover, maize ranks fourth in value as Zimbabwe’s principal
agricultural commodity after tobacco, sugar and cotton, respectively (Avila, et al, 1989).
In the communal sub-sector, maize ranks first in terms of the value of output (Mudimu,
1986).

Maize production trends in Zimbabwe are sharply influenced by government
policies and marketing arrangements. Before 1931 maize production was concentrated in
areas surrounding the urban and mining centers. Because maize marketing was not
controlled by the government, communal farmers contributed a high proportion of the
total marketed output of maize in local areas (Muir and Takavarasha, 1989). The
government started controlling maize marketing in 1931 with the passing of the Maize
Control Act and the subsequent creation of the Maize Control Board in 1934. The
objectives of the 1931 act were to stabilize farm incomes and maintain the viability of
EurOpean commercial farms (Muir and Takavarasha, 1989; Muchero, 1987). The
introduction of government controlled marketing led to a reduction in the contribution of
the communal farmers to marketed maize output. The commercial farmers produced
the lion’s share of the marketable surplus of agricultural products thereafter.

At independence in 1980, communal farmers producing maize faced the following
problems:

-inappropriateness of the technology generated by the research institutes,

-limited access to support institutions especially credit, extension and

marketing facilities, and

14

 

-weak rural infrastructure because of the 1965-79 liberation war.

In 1980 the new government headed by Robert Mugabe abolished discriminatory
policies, directed public and private support institutions (credit, extension and research)
to put more emphasis on the problems of communal farmers, increased the number of
rural buying points for communal farmers, and increased incentives by doubling official
maize producer prices over the 1980 to 1989 period. The results of these changes led to
a dramatic increase in maize production.5 But these changes mask several important
problems. Over 80 percent of the maize from communal farmers is produced from
about 20 percent of farmers in the favorable agro-ecological regions (Rohrbach, 1988;
Shumba, 1990). For instance, while maize yields in Mashonaland (a predominantly high
potential province) increased by a modest 1.8 percent they declined by 6.1 percent in the
drought-prone Masvingo province in the eighties (Roth, 1990). Communal farmers in
the dry regions face food security problems because most of them are net-buyers of grain
(Stack, 1994). Consumer tastes are shifting from small grains6 to maize (Lowveld
Research Stations, 1989).

These basic trends require multiple responses from the research community.
Researchers must find alternative and/or new cr0p varieties for the dry regions through
an aggressive agricultural diversification program. The growing land shortage should be
countered with yield-increasing technologies. But the technology development process
has a relatively long gestation period. In fact, it took 17 years (1932-49) for maize
researchers to develop a hybrid variety for Zimbabwean conditions. Cotton research
began in 1920’s in Zimbabwe but these efforts only started paying off in the 1960’s
(Blackie, 1987).

 

5 This is documented in studies by Rohrbach (1988) and (1989); Blackie, (1984);
Stanning, (1988) and (1991); Kupfuma et al (1989); Mombeshora et al (1989) among
many others.

‘ Consumers prefer maize flour to that from small grains such as sorghum, pearl-
millet and finger-millet.

15

 

Maize Research in Zimbabwe

ML":

Maize has received substantial research attention during the past six decades
(Weinmann, 1975). Early work on maize improvement concentrated on the evaluation
and improvement of local, South African and American varieties. By 1924 four open-
pollinated varieties namely Salisbury White, Hickory King, Louisiana Hickory and
Potchefstroom Pearl were found to be superior to local varieties in Mashonaland
(Weinmann, 1975). Salisbury White is an improvement of a local variety through
continuous selection. It was more suitable for the heavy soils than Hickory King, a
variety imported from the United States of America via South Africa. The latter gave
higher yields than the former on lighter and poorer soils (Weinmann, 1975). These four
varieties were later used to initiate a hybrid development program in 1932.

The breeding of hybrid maize started with the selfing" of Salisbury White, and
later Hickory King and Southern Cross to produce inbred lines. By 1940 over 300 inbred
lines had been developed and a program of crossing the inbred lines began (Weinmann,
1975). The first experimental hybrids were tested in 1938 (Mashingaidze, 1994). Single
hybrids (a cross of two inbreds), double hybrids (a cross of four inbreds or two single-
cross hybrids) and top-crosses (a cross where the female parent is an open-pollinated
plant and the male an inbred) were produced and tested extensively under field
conditions at six locations differing in rainfall, soil, altitude and other physical attributes.
In 1948/49 season the country released the first top-cross hybrid seed commercially and
the first conventional double-cross hybrid (SR1) became available the following season.
Zimbabwe thus became the second country after the USA to release locally developed
maize hybrids for commercial production. During the 1948/49 season about 22 percent

of the total commercial maize area was planted to new hybrid seed and produced about

 

7 Selfing refers to the crossing of plants from the same parent over and over again to
produce inbred (pure) lines.

16

 

33 percent of the total commercial crop harvested (Weinmann, 1975). The superiority of
hybrid maize over open-pollinated maize, especially during seasons when rainfall is below
average, led to the evolution of an exclusively hybrid oriented maize breeding program
(Mashingaidze, 1994).8

Maize production further accelerated with the release of a single-cross hybrid
SR52 in 1960 when Zimbabwe became the first country in the world to grow a single
hybrid commercially (Tattersfields and Havazvidi, 1994). SR52 had greater uniformity
and better standability and resistance to leaf and cob diseases than other varieties
available at that time. With improvements in inbred lines and seed production
management practices, the production of single-crosses commercially became feasible
economically for seed producers and farmers (Mashingaidze, 1994). Single-cross hybrids
have higher potential than double-cross and top-cross hybrids.

The double-cross and single cross hybrids released in the 1960’s were long-season
or late maturing best suited for commercial farmers in the high rainfall areas of the
country. The development of varieties suited for the marginal areas, where the bulk of
the communal farmers are found, was accomplished by focusing research on short-season
heterogeneous hybrids. In the early 1970’s three early maturing and stress tolerant
varieties were released (R200 R201 and R215), leading to wider adoption of hybrids over
the country including many communal farmers. Table 2.1 shows that the adoption of
hybrid seed by communal farmers rapidly increased in the 1970’s. The hybrids released
in the 1970’s were three-way crosses.9

The development of maize hybrids was done by scientists in what is now the
Crop Breeding Institute (CBI) of DR&SS. CBI shoulders the responsibility of managing

the maize and other crop breeding programs. Other CBI-bred varieties include 28225

 

8 The other reason for the push for hybrid development was the instability of the
early hybrids. It took researchers another ten years to nail down this factor.

9 Three-way crosses have properties of drought escape and avoidance due to early
maturity, a good pollen-to-silk synchronization, a relatively long pollen shedding period
(about 3 weeks) and tolerance to heat stress (Mashingaidze, 1994).

17

Table 2.1: Zimbabwe: Percentage Ad0ption of Hybrid Maize Seed by Commercial and
Communal Farmers 1945-1985.

 

 

Period Commercial farmers Communal farmers
1945-55 22 3
1955-65 88 20
1965-75 93 47
1975-85 100 94

 

Source: Mashingaidze, 1991.

and 28107 for high potential areas and 28232, Z8240 and 28206 that are yellow dents
mainly for animal feed and other industrial use. Maize hybrids are available for
commercial production from the Seed Co-op Company of Zimbabwe, Pioneer Hi-Bred
International, CIMMYT, and Pannar Seeds Company. DR&SS’s CBI, however, remains
the main maize breeding organization in the country.

After independence in 1980 the CBI’s maize breeding program remained hybrid
oriented but it expanded its testing sites to include the communal areas. By expanding
into communal areas, CBI, aims to; identify hybrid varieties that perform best under
communal area conditions, demonstrate good production techniques to the farmers, and
identify the main factors limiting maize production in these areas (Mashingaidze, 1994).
The main thrust of the breeding program in the 1980’s and 1990’s was to develop more
drought-tolerant short-season heterogeneous hybrids for farmers in the marginal rainfall

areas.

18

   

Agr_onom and Crop Management

In many research circles, plant breeding is regarded as the dominant source of
increased crop productivity. However, a recent international survey on agricultural
research budget allocation by discipline revealed that agronomic and other crop
management research accounted for about 50 percent of total crop research budgets
(Traxler. and Byerlee, 1991). It is also well known that crop breeding and agronomic
research are complementary activities. For example, agronomic and other crop
management research work is enhanced by improving the genetic potential of crops
through breeding programs (Kupfuma et al, 1990). This is confirmed by tracing the
development of agronomic research in Zimbabwe. Agronomic research on maize
nutrition (fertilizer management and crop rotation), crop geometry (plant populations
and spacing), timing of operations, weed control and pest and disease control did not
yield significant results until hybrid seed maize was used in the trials during the 1947/48
season (Mashingaidze, 1994). Thus optimal agronomic practices can establish the genetic
potential of a given crop variety but cannot enhance it (Mashiringwani, 1990).

The evaluation of the contribution of improved agronomic practices to yield
increases is an area of contention. However, authorities are in agreement that the use of
nitrogenous fertilizer gave the highest yield response”. Zimbabwean soils are known to
be deficient in nitrogen, phosphorus and sulphur (Grant, 1981). Potassium is abundant
in Zimbabwean soils but it must be replenished if the land is continuously cropped
(Mugwira, 1984). Since many trials were conducted to establish calibration curves for
major nutrients, especially phosphorus and nitrogen, it is possible to estimate the fertility
of a particular soil type for maize production. Other work on crop nutrition
encompassed other macro-nutrients and trace elements. Most of this work was done in

high rainfall areas and the results are therefore not applicable to the bulk of communal

 

1° Tattersfields (1982) estimated that 62 percent of the 325 percent (200%) increase
in maize yields of commercial farmers between 1950 and 1980 can be attributed to
nitrogenous fertilizer. Table 4 shows a complete breakdown of the contributions of
various other factors.

19

 

farmers in the low rainfall areas. Fertility work in the post-independence period is thus
focused on the establishment of calibration curves for the marginal rainfall areas.

Weed control and crop geometry contributed modestly to yield increases. Timing
of operations and pest and disease control experiments were ranked as marginally
favorable. The challenge presented by the communal farm setting is to develop less
capital intensive crop management practices compatible with the socio-economic
environment of the small farmers. It is scientifically more challenging to generate a
significant impact on productivity from agronomic research work in the communal areas
than in the well endowed commercial farming areas.

Table 2.2: Zimbabwe: Sources of increases in maize yield, (1950 to 1980)

 

 

 

Source Yield Increase (%)
Fertilizer 200
Hybrid varieties 45
Weed Control Practices 30
Recommended Plant Geometry 20
Early Planting Practices 15
Disease and Pest Control 10
Harvest Loss Reduction Practices 5
TOTAL 325

 

Source: Annual Report and Accounts, ARC 1981.

 

The largest impact of maize research has been on commercial farmers where

yields increased by 325 percent between 1946 and 1980 as shown in table 2.2 above. The

20

higher yielding varieties were fertilizer-responsive making it possible for farmers to use
fertilizers and other improved cultural practices. Table 2.2 shows the breakdown of the
contributions of the different types of technology. The use of fertilizer had the largest
contribution to increased yields followed by plant breeding. This table should be
interpreted with caution given the complimentary relationships between the different
technologies.

Today the almost universal use of hybrids by communal farmers demonstrates
that these farmers benefited from maize research (Mudimu, 1986). However, the
adoption of complementary and supplementary technologies has influenced the
magnitude of the gains to these farmers (Mombeshora, et al, 1989; Tattersfields, 1982;
Rohrbach, 1988). Thus it is important to take into account the different contributions

from other factors in the assessment of the returns from research investments.

Sum" Institutions

Enension

The Department of Agriculture provided both research and extension services to
the commercial farmers from the time of its establishment in 1897 until 1950 when the
Conservation and Extension (CONEX) branch within the newly constituted DR&SS was
spun off into a separate department to assist the commercial farmers (Tawonezvi,]994).
CONEX had specialists in crop and livestock extension who effectively linked extension
field staff with research stations. They had good access to farmers (Mudimu, 1986).

The extension needs of communal farmers remained unmet until the Department
of Native Development was established in 1920. Instructors were hired to work with
communal farmers. In 1944 the Department of Native Agriculture was formed after the
re-organization and restructuring of the Department of Native Development
(Weinmann, 1975). The first Director of the department, E.D. Alvord, an American

missionary agriculturist, developed a program of farmer training, crop rotation, and soil

21

conservation. In 1970 the Department was re-organized and renamed the Department of
Agricultural Development (DEVAG). But DEVAG did not have close links with
DR&SS partially because it was administered by the Ministry of Lands and Internal
Affairs (Mudimu, 1986). By contrast CONEX and DR&SS were administered by the
Ministry of Agriculture.

At independence in 1980, the two extension departments were merged into a
Department of Agricultural, Technical and Extension Services (AGRITEX) under the
Ministry of Agriculture. AGRITEX has three divisions: Field, Technical, and Irrigation.
The Field division manages a decentralized structure of provincial information
dissemination teams. The Technical and Irrigation divisions back-up the Field division
through a subject-matter specialist who is assigned to a provincial team. The provincial
teams in turn manage district teams, each of which is responsible for planning and
implementing extension programs. Field workers, each managing about 8 villages, are
responsible for program implementation at the field level. Each village has about 100
farming households giving an extension worker-to-farmer ratio of 1:800 (Avila, et al,
1989). A group approach is used to reach a larger number of farmers than the
individual approach whereby extension agents work with farmers on a one-to-one basis.

The failure of AGRITEX to reach most of the farmers in the communal farmers
limits the diffusion of new technology in these areas. Although AGRITEX has improved
its interaction with the DR&SS, the research-extension-farmer linkages need to be
further strengthened. The fact that most on-farm trials are researcher-managed is
evidence of these weak linkages. AGRITEX needs to improve its coverage of communal
farmers to identify farm level problems and constraints (Shumba, 1990). In addition
AGRITEX field workers need to be upgraded in terms of their ability to diagnose and
identify problems.

22

Seed Produaion and Marketing

About forty years after agricultural research was initiated in 1897 a regulated
seed industry was established when it became apparent that home grown seed had
problems such as reduced variety purity and quality, poor germination, and proneness to
insect attack and seed borne diseases (Tattersfields and Havazvidi, 1994). These
problem reduced crop yields and quality. A small group of commercial farmers formed
the Southern Rhodesia (now Zimbabwe) Seed Maize Association (SMA) in 1940 to
produce certified maize seed (Weinmann, 1975). The SMA established close liaison with
the Department of Agriculture and produced certified seed for the open-pollinated
maize varieties. When commercial hybrid seed production started in 1949 the SMA
assumed the lead role and it marketed about 100 metric tons that year”. The
establishment of other seed associations and schemes followed the success of the maize
seed program. In 1957 three associations were established; the Crop Seeds Association
(other field crops), Tobacco Seed Association, and Potato Seed Producers Association.
A pasture seed association started operations in 1964. Seed production for other
commodities such as cotton, horticultural crops, and coffee is carried out through a
variety of arrangements. The common features of these organizations are the high
priority given to the regulation of seed production and a partnership between the
farmers and the government.

The SMA, like other seed associations, established close working relationships
with the government. This informally assured the SMAs access to the varieties produced
by government’s DR&SS. In 1970 a Tripartite Agreement was signed between the
Ministry of Agriculture, the SMA and the Commercial Farmers Union (Tattersfields and
Havazvidi, 1994). The agreement granted SMA exclusive rights to produce and market
seed for all government bred varieties. The aim of the agreement was to ensure

government control over seed marketing and the maintenance of quality standards in

 

11 This compares with 35 000 metric tons in the late 1980’s (Tattersfields and
Havazvidi, forthcoming).

23

 

, \

seed production. The Seed Act of 1965 added the requirement that all varieties must be
approved by a broadly-constituted Crop Variety Release Committee to provide the
necessary checks and balances. Finally the passage of the Plant Breeders Rights Act in
1973 offered protection to varieties bred locally and outside Zimbabwe.

The Seed Maize and Crop Seeds Associations were merged in 1982 to form the
Seed Co-op Company of Zimbabwe which arranges local and export sales of seed. The
Seed Co-op has expanded its research activities and its maize research program is
currently as large as the government’s research program”. The Seed Co-op’s maize
research program, however, concentrates on breeding for heat and stress tolerance in
early maturing varieties, streak virus resistance and the agronomy of seed production
(Tattersfields and Havazvidi, 1994). There is close collaboration among the government,
international research centers, and the Co-op’s programs in terms of exchange of
materials and information, and use of sites and other resources.

The Seed Co-op has done a good job in making seed available to most farmers
on time. It makes good use of the private marketing system to achieve comprehensive
nation-wide coverage. The main participants are distributors, wholesalers and retailers
who have made it possible for over 90 percent of the communal farmers to adopt maize
hybrid seed (Tattersfields and Havazvidi, 1994; Mudimu, 1986). The introduction of
small packages varying in sizes from 500g, 2kg, 10kg, 25kg and the standard 50kg was
instrumental in the rapid and widespread adoption of hybrid seed. The seed industry
has also generated foreign currency through its exports to neighboring countries, partially
fulfilling its food security role under the SADC regional food security plan (Tattersfields
and Havazvidi, 1994).

The Ministry of Industry and Trade controls hybrid seed maize pricing. The cost

of production is used as the basis of pricing decisions but the establishment of new local

 

‘2 These activities are carried out at Rattray-Arnold Research Station located 30 km
north of Harare. This area is typically classified as high potential and is in many ways
similar to the government’s Harare Research Station except for altitude.

24

and international seed companies may make this system unworkable as different

companies face different production costs.

Maize Marketing

The creation of the Maize Control Board in 1934 marked the beginning of
controlled marketing of maize in Zimbabwe (Muir and Takavarasha, 1989; Muchero,
1987). In 1957 the replacement of the Maize Control Act with a Grain Marketing Act,
brought other crops under controlled marketing. The act promulgated the establishment
of the Grain Marketing Board (GMB) and includes a provision which allows the GMB to
prohibit grain marketing between the large scale and communal farms and between the
large scale farmers. It, however, authorizes trade between the smallholders (Muir and
Takavarasha, 1989). These regulations were based on the assumption that communal
farmers should be self-sufficient in food allowing the government to deal with supplying
food to the urban market.

The GMB is a single channel marketing system designed to service the needs of
commercial farmers. Its grain silos and depots are concentrated in commercial farming
areas. The GMB has effectively fulfilled its objective of national food self-sufficiency
and exports in normal rainfall years with rare exceptions during epic drought years. At
independence in 1980, the GMB was given a broader mandate to serve communal
farmers. As a result grain intake from communal farmers increased sharply as shown on
Table 2.3. Rohrbach (1987) cited increased access to markets as one of the sources of
the large post-independence increase in communal maize production and sales shown in
table 2.3. But this expansion increased the GMB operating costs, raising questions about
the efficacy of a single channel marketing system13 in communal farming areas

(Muchero, 1987). It is costly for the GMB marketing system to collect small sparsely

 

‘3 A single channel marketing system refers to a centrally controlled grain collection
scheme that takes grain away from the farms and has no mechanism to bring it back
when need arises. This is particularly necessary during drought years.

25

distributed quantities of grain from these areas. The challenge now is to develop new

marketing systems compatible with austerity measures and the need to serve the interests

of small, medium and large scale farmers.

Table 2.3: Zimbabwe: GMB Maize Intake From Smallholders, Percentage of Total Grain

Intake and Rainfall, 1981-89.

 

Factors 1981

1983 1984 1985 1986 1987 1988

1989

 

GMB Intake 86

(1000 MT)
Percent 10.6
Rainfall" 609
(mm /year)

369 152 389 819 682 156

26.6 24.7 41.4 44.8 42.8 38.7

426 352 468 712 662 «-

756

63.2

 

* Rainfall is a "national simple average" in millimeters per year

Source : Central Statistical Office; Statistical Yearbook, (several).

Another marketing problem stems from the fact that over 80 percent of marketed
output from the communal farmers comes from about 20 percent of farmers who were
located in favorable (high rainfall) natural resource regions (Rohrbach, 1988; Shumba,
1990). Stack (1994) found that relatively higher percentage of farmers running out of
food during the year as one moves from higher to lower potential areas. The single
channel marketing system seems ill-equipped to serve food deficit communal households
(Jayne, et a1 1994) because most of these households rely on purchases of maize meal
rather than grain (Stack, 1994). But the local markets do not sell grain because it is
channelled by the GMB to the urban centers for processing, export, drought relief resold
as maize meal in communal areas (Jayne and Rubey, 1993: Jayne et al, 1994). At the

policy level, the question becomes one of how much emphasis should be placed on the

26

development of technologies suitable for these unfavorable areas versus the development

of efficient and effective marketing systems.

Credit

Because most communal farmers did not have access to agricultural credit, the
Agricultural Finance Corporation (AFC), a parastatal organization, expanded its credit
facilities to communal farmers at independence in 1980. The problem of lack of
collateral has hindered further expansion of the program. The AFC reported a sharp
downward trend in both number of loans and amounts extended to communal farmers as
from the mid-1980’s (AFC Annual Report, 1991). The bulk (57%) of these loans went
to farmers in the high potential Mashonaland Central and West provinces (Chimedza,
1994).

Communal farmers also get credit from private organizations such as the Silveira
House, Savings Clubs, input supply companies and informal community associations
because the AFC issued credit to less than 10 percent of the communal farmers in the
eighties (Chimedza, 1994).14 These institutions need to be studied to identify ways of
improving or reorganizing existing institutions such as the commercial banks and the
AFC. If effective rural credit institutions are developed, then there are greater chances
that technologies developed by research organizations will be adopted. The benefits of
technology development will therefore be enjoyed by a wider clientele.

Agricultural credit for commercial farmers can be obtained from the AFC,
commercial banks, financial institutions, farmers’ cooperatives and private companies
(Avila, et al, 1989). While other institutions supply only short-term and medium-term
lending, the AFC finances long-term loans. Therefore, for the commercial farmers,

credit services are adequate and available. The farmers are able to make investments in

 

1‘ A May 1985 IFAD study showed that Savings Clubs had 18.8 % communal
household membership, the Silveira House had 7.4% and the input supply companies
extended credit to 7.5% of the rural households (Chimedza, 1994).

27

farm inputs and infrastructure that enables them to maintain productivity levels higher
than that of the credit-deficient communal farmers. The importance of credit

institutions, therefore, need not be over-emphasized.

Odierlnstitutions

Besides efficient input and product market delivery systems for communal
farmers it is important to ensure that other infrastructure is in place for communal
farmers. This includes the provision of all-weather roads and where rainfall is a limiting
factor, suitable irrigation facilities. While these facilities are well-developed and
efficiently utilized in the commercial farming areas, the communal farmers have very
limited access to these facilities. This study will examine the differences in technology
adoption among sectors in order to contribute to the debate and provide information on
setting priorities on whether to emphasize technology development specifically for the
communal sector (in the unfavorable environments) or to continue to invest in regions
giving the highest returns.

Another factor for analysis is the role of farm organizations in setting research
priorities, mobilizing financial support for agricultural research and extension and
generally lobbying for the welfare of agriculture in the policy arena. Commercial farmers
have a strong and well-organized lobby group dating back to 1903 when the Rhodesian
Farmers Association was established. This lobby group masterminded most of the
changes in agricultural rules, regulations and policies especially before 1980. On the
other hand about 10 percent of the communal farmers constituted the membership of
the National Farmers Association of Zimbabwe (NFAZ)‘5(Ndoro, 1990). The
commercial farmers’ organization has almost 100 percent membership and is able to

collect levies from its members.

 

‘5 NFAZ was an organization which represented communal farmer interests at
national level before 1992 when it merged with the Zimbabwe National Farmers Union
(ZNFU) representing small-scale commercial farmers.

28

 

SM!

Maize is a crop of strategic importance in Zimbabwe, especially for the
communal farmers who depend on it as a source of both food and income. Increasing
maize production in the communal areas will help ensure an adequate food supply and
increase communal farm incomes. Increased production can be realized from yield
improvement through research as the land frontier nears exhaustion. Most of the
research work done in the past was not focused to meet the needs of communal farmers
found mainly in the marginal areas.

The development of maize hybrids and calibration curves for major
macronutrients and optimal agronomic practices was accomplished by careful sequencing
of breeding and crop management research work. It also required that a number of
institutions be in place as most of the past improvements in maize yields have come
from the use of fertilizer and other expensive industrial inputs. Access to inputs was
found to be a critical determinant of the adoption of hybrid maize in the 1960’s and
1970’s. The role of extension, credit institutions, input and output markets in technology

adoption cannot be over-emphasized.

29

HAPTER III

ME_A§ULIN§ THE PRQDUQTIVITY OF AGRICULTURAL RESEARQH AND
w

1%

Nations invest in research and extension because they are a primary source of ‘
increase in productivity in both the agricultural and non-agricultural sectors (Evenson,
Waggoner and Ruttan, 1979). Both the private and public sectors invest in agricultural
research and extension in order to increase factor productivity. Because research and
extension activities require an investment of scarce financial and manpower resources,
the returns thereof must be calculated and compared with those obtained in alternative
investment opportunities to guide decision makers on how much to invest. The
measurement of research productivity is critical for such an analysis.

The accuracy of measurement depends on the nature of the variables in question.
Agricultural research and extension, in most cases, produces products that are endowed
with attributes of public goods. In such cases the inability of the supplier of agricultural
research products to internalize the benefits thereof renders the production prOCess non-
competitive (Akino and Hayami, 1975).

The agricultural research process, by producing mainly public goods, contributes
significantly to societal welfare if the benefits exceed the costs. Private firms are rarely
expected to undertake activities which do not have direct payback for they are interested
in products from which they can ”expropriate the benefits". Thus, public institutions are
left with the major task of ensuring that society optimizes the gains from agricultural
research in cases where benefits cannot be internalized. To decide on the socially
optimal level of investment in agricultural research, decision makers need information on
research payoffs. This information is essential given the credibility problems that are

faced by many agricultural research organizations (Norton and Davis, 1981).

30

A number of techniques have been developed to evaluate the payoffs to
agricultural research. There are several reviews of these techniques.16 Schuh and
Tollini (1979) categorized the techniques in two groups: at post and ex ante evaluations.
Ex post evaluations assess past research efforts while ex ante evaluations assess the likely
future impact of research activities before they are implemented or before the impact of
these activities is felt. This chapter reviews agricultural research evaluation methods
with a view of identifying one which has the properties most appropriate to assess

investments in maize research in Zimbabwe.

Ex Post Evaluation Aggy—chew

Norton and Davis (1981) identified four classes of ex post evaluations:
(a) the economic surplus approach;
(b) the production function approach;
(c) the impact on national income approach and
(d) the impact on nutritional status approach.
Schuh and Tollini (1979) added an inputs-saved approach as a fifth category. This

method was classified under the economic surplus approach by Norton and Davis (1981).

The Economic Sgplus Approach

As shown in figure 3.1 above, the economic surplus approach assumes a
downward shift in the supply function from So to 81 that improved technology brings.
The benefits are estimated by netting out the gains and losses in producer and/or
consumer surpluses over time. This yields a net gain represented by the area abcd in
figure 3.1. The shape and hence the size of this area depends on the functional form of

the supply and demand curves as well as the magnitude of the supply shift. The

 

’6 Reviews of approaches to evaluate research payoffs include those by Norton and
Davis (1981); Oehmke et al. (1991); Schuh and Tollini (1979); and Scobie (1979) among
many others.

 

31

 

Price SO

\ // * ’/,51

 

. Demand
Curve

 

 

Q0 O1

Amount Produced

Figure 3.1: Social benefits from a downward shift in the supply curve due to agricultural
research.

cumulative benefits are then compared with cumulative research costs by calculating an
average rate of return to research expenditures”.

T. W. Schultz (1953) carried out the first study to quantify returns to agricultural
research by using an inputs-saved approach (Schultz, 1953). Schultz compared the value
of inputs saved in the US. by using 1950 technology with the costs of research and
development at 1910 technology levels. He found that the 32 percent increase in the
productivity of inputs translated into an increase in consumer surplus as the implicitly
assumed perfectly elastic supply function shifted down on a perfectly inelastic demand

function.

 

‘7 The net benefits stream is calculated as benefits minus costs. The economic surplus
is the sum of the discounted net benefits over the project period. The average rate of
return is the internal rate of return or the discount rate that reduces the present value of
the net benefit stream to zero.

32

Griliches (1958) extended Schultz’s pioneering study by quantifying the returns to
hybrid maize research in the United States under two scenarios (Fig. 3.2). Griliches
assumed unitary demand elasticity. In Case I, the supply function was assumed to be
perfectly elastic whilst it was perfectly inelastic in Case II. The net change in economic
surplus in the perfectly elastic supply function case (Case I) affects only consumers.
There is no producer surplus in a situation depicted by Case I. Producers gain in terms

of reduced production costs if the elasticity of demand is less than unit in absolute terms.

Case I: Perfectly elastic supply curve.

Price

SO

31

 

.\\
\.Demand
Curve

 

 

Q0 Q1 Quantity
Case II: Perfectly inelastic supply curve.
SO

Price

 

 

Quantity
Figure 3.2: Zvi Griliches’ models for measuring the returns to hybrid maize research in

the United States.

In this case savings made (area abde) are greater than the extra costs from increased
output (area Qochl). In the perfectly inelastic supply function case (Case H), consumer

surplus includes the portion lost by producers represented by area P0ch1 in figure 3.2.

33

 

However, the producers gain by being able to produce more, contributing area abde to
the net economic surplus area (acde). Although elasticities have no effect on the level of
research payoffs, a more elastic demand curve will yield a more equitable distribution of
benefits between consumers and producers after a supply-shifting technical change.

The net change in economic surplus has been estimated with different
assumptions about the elasticity of the demand and supply functions and the nature of
the supply shift. Peterson (1967) estimated the change in the net social surplus for
poultry research with non-linear functions, thus eliminating the restrictions imposed on
the Griliches demand and supply functions. Peterson’s supply function shift was a
proportional one. Linear functions and a parallel shift in the supply function
assumptions were used to estimate the returns to research on tomato harvesting
techniques by Schmitz and Seckler in 1970. They also included the value of inputs saved
as part of the benefit stream in their calculations.

Internationally the economic surplus approach has been used extensively to
evaluate returns to research under different assumptions about the shift of the supply
function and the elasticities of both the supply and demand curves. Ayer and Schuh
(1972) and Akino and Hayami (1975) assumed a pivotal shift in their studies on returns
to cotton research in Brazil and rice research in Japan, respectively. In Australia,
Duncan (1972) calculated the net benefits from research that increased the productivity
of pasture inputs expressed through shifts to the right of perfectly inelastic derived
demand curves of the inputs.

These studies18 illustrate the diversity of methods of specifying the supply and

demand functions, in terms of elasticity, and the shift parameter. This approach requires

 

‘3 Other studies using the economic surplus approach include Scobie and Posada for
rice in Colombia and the distributional impacts; Barletta (1970) for maize and wheat in
Mexico; Flores-Moya, Evenson and Hayami (1978) for rice in the Philippines. Lindner
and Jarrett (1978) took the differential effects of the different technologies into account
by looking at the aggregate effects at the industry level. Rose (1980) modified this
approach by using a supply curve that is kinked at the point of intersection with the
demand curve before the innovation.

34

these parameters to be accurately estimated and for the functions to be appropriately
specified. The measurement of the shift factor is particularly critical because the results
can vary according to the specification (Norton and Davis, 1981). The elasticity
assumptions are important in that the relative elasticity of supply and demand functions
determines the distribution of benefits and/ or costs between consumers and producers.
These considerations and the problem of defining research costs reduce the confidence
of using the results of economic surplus approach to evaluate returns to agricultural

research.

This approach estimates the contribution of each input to increased output by
postulating a causal relationship between inputs (such as research, extension, prices, etc)
as independent variables and output as the dependent variable. The coefficients of the
independent variables, if found statistically significant, are used to estimate the marginal
productivity of these variables. The production function approach makes it possible to
apportion the benefits of research to different variables. The crucial element of the
production function approach is the way the lagged response of output to research inputs
is specified in terms of length and the pattern of persistence of benefits from the time
they are first felt.

Griliches (1964) pioneered the use of the production function approach when be
evaluated returns to agricultural research in the United States from 1949 to 1959. Using
single-year lagged aggregate cross-section data Griliches estimated the marginal rate of
return of agricultural research and education of farm workers in the United States to be
35 to 40 percent. Later Evenson (1967) used the same time period, a mean lag length of
six to seven years and a concave shaped distribution of benefits which generated a 47

percent rate of return to agricultural research. Fishelson (1971) and Cline and Lu

35

(1976) used the same data set to refine Evenson’s work by calculating the marginal
returns to research for ten production regions in the United States.19

Attempts to use time-series data to calculate the payoffs to research have not
been successful due to problems of data shortage and high correlation of independent
variables (Norton and Davis, 1981). To avoid these problems, duality approaches using
profit and/or cost relationships have been used (Binswanger, 1974). This approach
involves the aggregation of those variables that are likely to be interrelated.

The production function approach has been found useful when it is possible to
separate the effects not only by input type, but also by geographical location and type of
research (commodity-specific versus non-commodity specific; applied versus basic
research programs)”. This made it possible to assess the spill-over effects between
regions and the benefits of decentralization of the research system. Most of these
findings pertain to the United States. Kahlon et a1. (1977) used a production function
approach with aggregate data to evaluate the productivity of agricultural research in
India. Flores-Moya, Evenson and Hayami (1978) evaluated the payoffs to research in
the Philippines paying particular attention to international spill-over effects. Karanja
(1990) used a production function approach to calculate the average rate of return to
maize research in Kenya (68 percent) for the period 1955 to 1988.

More data is required for the production function analysis than that required for
the economic surplus approach. This imposes a severe limitation on the use of the
production function approach in many developing countries. Another shortcoming is the
treatment of research, extension and other variables as exogenous variables in

formulating the causal relationships. Evenson (1991) illustrated this endogeneity

 

’9 These extensions involved the use of other variables to explain the sources of
productivity increase, such as the number of scientific papers published as a proxy for
research output (Evenson and Kinsley, 1975).

2° See Evenson (1958) and other studies by Flores-Moya (1978); White and Havlicek
(1979); Evenson and Kinsley (1973, 1975); Evenson, Waggoner and Ruttan (1979)
Evenson and Binswanger ( 1978); and Bredahl and Peterson (1976).

36

problem by using extension expenditures as the explanatory variable for technology
adoption. The induced institutional innovation model predicts that the development of
technologies is, in most cases, driven by the farmers’ desire to exploit profitable
opportunities that arise after relative price changes over time (Hayami and Ruttan,
1990). The research input therefore becomes endogenous to the system rather than

exogenous.

[mm on National Income Approach

This approach to estimating the returns to research is similar to the economic
surplus approach in that it involves the calculation of additional national income
generated by agricultural research and compares it with the costs of research and other
related agricultural activities. This approach is essentially a benefit-cost analysis
approach. Benefit-cost ratios as well as rates of return can be calculated by this
approach. Depending on the availability of data, this approach can be used to assess the
impact of agricultural research on employment, foreign exchange generation, and the
distribution of income between income groups and also between producers and
consumers.

The most widely cited study using this approach was carried out for the United
States by T‘weeten and Hines (1965). The study calculated the most likely level of
national income if: (a) the ratio of farmers to non-farmers remained constant at 1910
levels; and (b) the income levels of the assumed additional farmers were the same as the
level of the actual farmers. Part (b) equates the income of additional farmers to that
earned by actual farmers and not what they are currently earning in the non-farm sector.
The data required for this approach include time-series data on national income,
research and other related agricultural program expenditures, farm and non-farm
populations and the movement of people between farm and non-farm occupations. The
non-availability of such data in developing countries limits the use of this approach in

evaluation of research payoffs.

37

The Nutritional 11¢ng Approach

This approach estimates the contribution of agricultural research to food
availability both quantitatively and qualitatively. It does not by itself provide an estimate
of the rate of return, but gives a measure of the how well agricultural research is
achieving the goal of improving the nutritional status of the population (Oehmke _e_t_al,
1991). This analysis can be done by consumer income groups.

Pinstrup-Andersen, Londono and Hoover (1976) developed a two-phased model
to measure the nutritional impact of technology. The first task is to examine the
demand structure in terms of elasticity and the differences across income strata and
market segments. The second phase involves the assessment of the availability of the
commodity, distribution among income strata and the aggregate impact on protein and
calorie intake again by income strata. The change in consumption baskets must be

included in the assessment of the impact of research on nutrition.

Ex Ante Evaluation Approaches

Er ante approaches calculate the likely future rate of return before research has
been done. Ex ante evaluations can also be classified into four categories:
(i) scoring models;
(ii) benefit-cost analysis;
(iii) simulation models, and
(iv) mathematical programming models. (Norton and Davis, 1981).
The scoring models and benefit-cost analysis are simple to use. A panel of
experts use a predetermined criterion when ranking the alternative planned research

programs.21 Benefit-cost approaches are essentially ex ante economic surplus

 

2‘ Priorities were set for future research and other agricultural activities after studies
utilizing simple scoring models such as the 1966 Department of Agriculture and the
National Association of State Universities and Land Grant Colleges study on agricultural
and forestry research activities; the Minnesota model by Pinstrup-Andersen and Franklin
(1977); the Iowa model by Paulson and Kaldor (1968) and Mahlstede (1971); the North
Carolina model by Shumway and McCraken (1975); among many others.

38

evaluations of planned research programs. Research priorities are formulated on the
basis of benefit-cost ratios or rates of return obtained.22 These approaches incorporate
expert opinion on probabilities of research success and allow incorporation of benefits
and goals that are difficult to quantify.

Simulation models are flexible, step-wise procedures used to predict the relative
contributions and costs of alternative research strategies or plans. The best known
example of such a model is the one constructed by Pinstrup-Andersen and Franklin
(1977) for research resource allocation for developing countries. Later Scobie (1979)
constructed a simulation model that estimated annual research investments required to
generate given rates of return. But the data required to build these models, as in the
case of mathematical programming models, are extensive and difficult to find.23
Russell’s (1969) linear programming study in the United Kingdom maximized three goals
(consumption, security and equity) subject to resource and statutory constraints.

Specifying the objective function is one of the major difficulty with this approach.

Rate at Return Studies in Am

The role of agricultural research has long been recognized by both African

governments and the international community. Nevertheless, only eleven studies on

payoffs to research have been conducted in Africa. This compares with 25 and 60 studies

of the payoffs to research in Asia and Latin America, respectively (Oehmke and
Crawford, 1993).

Because land is becoming scarce in many countries in Africa, future growth in

agricultural output will have to come from a science-based agricultural revolution which

increases yields per unit of land area. The expected intensification of agricultural

 

22 Norton and Davis (1981) reviewed studies by Fishel (1971), Ramalho de Castro
and Schuh (1977), Easter and Norton (1977), Araji et al (1978) and Eddleman (1977).

23 The data problem is more pronounced when using mathematical programming
models because the objective is to optimize a multiple-goal objective function for a given
bundle of research resources.

39

research activities in Sub-Saharan Africa will increase the utility of measuring the
productivity of research to guide investment and resource allocation decisions. Thus the
number of rate of return studies is expected to increase. The results of the eight rate of
return studies are shown in table 3.1.

Abidogun (1982) evaluated the payoff to cocoa research and extension and found
a 37 to 42 percent rate of return. The average rates of return to World Bank cotton
development projects in Burkina Faso, Cote d’Ivoire and Togo demonstrated the role of
a ”package approach" in increasing cotton production“. Makau (1984) found a 30 to 40
percent rate of return on wheat research in Kenya which he attributed mainly to
improved varieties and modern inputs. Using data from twenty-four countries, Evenson
(1987) found that the rate of return on research on staple crops in Africa was about half
that of Latin America and Asian research on staples and cereals.

Noorgard reported a 149:1 benefit-cost ratio from research on the control of the
cassava mealybug in West Africa in 1988. Schwartz e_t_al_ (1990) found that the rate of
return to cowpea research, extension and distribution in Senegal ranged between 38 and
96 percent. The average rate of return to hybrid maize research in Kenya over the 1955-
88 period was found to be 68 percent (Karanja, 1990). Mazzucato (1991) drew on
Karanja’s data and fertilizer consumption data and found a rate of return of 58 to 60
percent.

Table 3.1 presents the results of fourteen studies and reveals that the payoff to
research in Africa is generally high, averaging well above 15 percent”. The variability
in the estimated rate of return especially across countries is an indicator of the effect of
country-specific factors. The higher rate of return to research on maize and cereals in

Eastern and Southern Africa is probably due to the effects of strong maize and cereals

 

3‘ The package includes not only improved technology but also the provision of
services such as input supply, extension, processing and marketing.

25 The assumption is that the opportunity cost of investment funds is between 10 and
15 percent.

40

Table 3.1: Summary results of rate of return studies in Africa.

 

 

 

Author(s) Year Country or Crop(s) Period Rate of
Region Return
World Bank 1982 W. Africa. Cotton 11 / a 11-41
Abidogun 1982 Nigeria Cocoa 1945-1985 37-42
Makau 1984 Kenya Wheat 1924-1974 33
Evenson 1987 Africa Maize and 1962-1980 30-40
staples
Norgaard 1988 Africa Cassava 1977-2003 149:1"
Schwartz et al. 1989 Senegal Cowpeas 1981-2015 63
Karanja 1990 Kenya Maize 1955-1988 40-60
Mazzucato 1991 Kenya Maize 1955-1988 58-60
Ahmed & Sanders 1991 Sudan Sorghum 1977-2013 22-39
Boughton & 1992 Mali Maize 1969-1991 135
Teme
Laker-Ojok 1992 Uganda Sunflower 1986-1991 < 0
COWpea &
Soybeans
Howard 1993 Zambia Maize 1979-1991 20
Mazzucato 1992 Niger COWpea, 1975-2010 2-21
Sorghum
& Millet
Sterns and 1992 Cameroon COWpea 1979-1998 3- 15
Bemsten

 

’ The west African countries are Burkina Faso, Cote d’Ivoire and Togo.
" Benefit:cost ratio.

n/a: Data not available.

Source: Oehmke and Stems (1992)

research programs in Kenya and Zimbabwe and the spill-over effects into neighboring
countries. Thus, although it appears appropriate to estimate research payoffs at the
country level, one should not lose sight of regional spill-over and spill-in effects. The
current sale of Zimbabwe’s SR-52 hybrid maize seed in a dozen or more countries in

Africa adds further weight to the role of research spillovers.

41

Evaluations o P to Maize Research in Zimbabwe

Weinmann laid the groundwork for the estimation of payoffs to research in the
country by documenting the detailed historical evolution of Zimbabwe’s research system
from 1900 to 1975. This information is valuable in tracing the impact of policy and
institutional changes on the productivity of the research system. Another attempt to
quantify the returns to maize research was carried out by Tattersfields in 1982.
Tattersfields divided the 1950 to 1980 period into five-year periods and calculated the
average yields per hectare for each period. An analysis of yield trends over time
revealed that the maize yields had increased by 325 percent from 1950 to 1980. Just
under 14 percent of the increase in yields (45% divided by 325%) was attributed to the
introduction of maize hybrids (Tattersfields, 1982). Even though these estimates were
widely used as reasonable estimates of partial effects of individual factors on maize
yields, Tattersfields (1988) agreed that this exercise was an over simplification because
interaction effects were ignored.

Rorhbach (1988) found that, in communal areas, the area under hybrids
increased from over 30 percent in the late seventies to over 80 percent in the mid-
eightiesz‘. Maize yields in the communal areas have consistently been about one-third
of the yields on commercial farms (Tattersfields,1982). A review of crop production
from 1890 to 1980 by Muir-Leresche (1984) is an important source of historical
information for this analysis. In the first published study of the returns to agricultural
research in Zimbabwe, Thirtle, et al. (1993) regressed the logarithms of the total factor
productivity (TFP) index on the polynomial distributed lag of ”R&D expenditures",
lagged extension expenditures and a weather index and obtained a rate of return
between 40 and 60 percent on investments in agricultural research and extension. This

study covered the period 1970 to 1990. But since Zimbabwe started to invest heavily in

 

2‘ Mashingaidze (1994) reported that maize hybrid adoption was 46 percent in the
communal areas in the 1970’s and increased to over 90 percent in the 1980’s. The
adoption rates are shown in Table 3 in chapter 2.

42

research in the 19203 (cotton) and the 19308 (hybrid maize) it is important to include

these early research costs in the study of the returns to research.

Model SpgciM’ n and A_s§u_mptions

The purpose of this study is to carry out an ex post evaluation of the payoffs to
hybrid maize research from 1932 to 1990. An economic surplus approach will be used
because the incidence of net benefits can be quantified. Even though the concept of
economic surplus is a subject of considerable controversy, this approach will be used to
analyze the differential impact of the new technology on large-scale farmers, communal
farmers and consumers. The underlying model is a market model for maize where the
change in economic surplus is an indicator of change in welfare. Economic surplus is
made up of consumer and producer surplus. The consumer surplus is a measure of
welfare changes associated with the consumption of both maize and other
commodities”. Producer surplus is a measure of gains to producers or a proxy for
returns to land, labour and capital used in maize production.

Once the benefits, in the form of consumer and producer surplus, have been
calculated, the research and transfer costs will be subtracted to obtain the net benefit of
investing in maize research. The research costs consist of all the costs on work on
breeding, agronomy, plant protection and maize seed agronomy from 1932 to 1990.
Transfer costs are made up of the cost of maize technology extension programs. The
seed production and marketing cost are fully recovered since these activities are carried
out by a private organization. The net benefits are discounted to take into account the

time value of money. The general form of the model is presented below:

 

27 The fall in the price of maize has both income and substitution effects on maize
consumers. With inelastic demand for maize the income effect dominate the substitution
effect. Thus less income is used to purchase maize and the excess income is available to
consumers who spend it on other commodities. The benefits from the consumption of
other goods is part of the change in the consumer surplus area.

43

 

with

T
DB = 2 rt(BNt — Cr) 3'1
t= l

1
(1 + r)’

r,=

 

Where DB is the summation of net benefits; that is BN, which is the summation

of consumer and producer surpluses, less C,, the research and transfer costs over the

years t=1932,...,1990. The discount rate r, used to calculate r, (the discount factor), is a

real discount rate. Calculating the rate of return involves solving for the internal rate of

return, which is equal to the discount rate that reduces DB to zero.

The model used in this study is adapted from Akino and Hayami (1975). The

major assumptions in this study and their justification are as follows:

1.

The maize industry in Zimbabwe is effectively closed from world markets
(outside Southern Africa) because of high transport costs from the main ports of
Durban, Maputo and Beira.

Before 1965, the maize industry operated more or less as if it was a free market
(Blackie, 1994). After 1965 government interventions which were designed to
benefit the commercial sector had little impact on communal production until
1980.

Over 90 percent of traded maize volume is through the Grain Marketing Board
(GMB) the statutory marketing authority. Farmers are paid a "cost-of-production
base" price. The consumption of processed maize meal is subsidized. But this
applies to less than one third of total maize consumed. The increase in social
costs due to this policy is ignored because it is relatively insignificant.

Prior to 1980 the influence of government pricing policy was minimal in the
communal areas. Because intra-communal area trade was limited, the

government prices will be taken as the opportunity costs of maize in this sector.

44

 

5. Constant elasticity supply and demand functions are assumed. The new maize
hybrids shift the supply curve downwards. Alternatively this can be interpreted as
a reduction in per unit costs of production. The downward shift is assumed to be
pivotal. A pivotal shift results in the most conservative measure of benefits
(Lindner and Jarrett 1978, 1980).

6. Most of the maize produced in Zimbabwe is for human consumption. Price
differentials are small with maize for stock feed being valued equally with that for
human consumption. It is assumed that the maize product is homogeneous in the
national market.

7. The analysis will be conducted at the farm level where most of the consumption
takes place. Urban consumption, it is assumed, does not involve a significant

departure from rural consumption in terms of processing requirements.

Graphical RQresentation

Figure 3.3 represents a general case from which to construct measures of total
research benefits and their distribution. The supply curves (80 and 81) represent a
horizontal summation of communal and commercial supply functions. Similarly the
demand curve (Dm) is a horizontal summation of the demand functions of rural and
urban households, stockfeed companies and occasionally exports. The benefit of

research is represented by the area between the supply curves and below the demand

function:
TB =ACS +APS =0bc (32)
where;
A CS = PObaPl + abc = PobcP1 (33)
and;

45

 

 

 

 

 

oo 01
Out
Figure 3.3: The impact of technical change.
A PS = PICO — PobO = Oac - PobaP1 (3.4)

Thus the area Obc is made up of abc, which is the consumer surplus and Oac which is
the producer surplus. The net position of producers depends on whether the area Oac is
greater than the loss in producer surplus represented by area PobaPl. The sizes of these
areas depend on the relative elasticities of demand and supply functions. The only time
consumers do not benefit from a research-induced shift in the supply curve is when the
demand curve is perfectly elastic. This means that the more elastic the demand curve is
the smaller the benefits to consumers. Similarly producer benefits become smaller as the

supply curve becomes more elastic.

46

 

Algebraic Representation

The formal mathematical representation of the measures of research benefits in

terms of supply and demand functions is as follows:

Qd' = (1! Pi" Demand Equation....(3.5)
Q50 = 6 P050 1) Supply Equation without hybrids ..... (3.6)
J r . -

e . . _
Q5” = 31(1 + kr)Pl.(r- 1) Supply Equation wrth hybnds .......... (3.7)

In equilibrium the following identities hold;

Qd, = Q30; Without hybrids....(3.8)
and
Qd, = qu With hybrids ....... (3.9)

Equations 3.5 through 3.9 can be rearranged to give the following form;

k

 

 

 

P1=Pu=Por(1"€+n) (3'10)
and
11k
= = 1 + l (3°11)
Q. 9,, Q..( 8 + n)
and where;
k, = [Yu “Yul [Au] (3.12)
Y1; A:

Where;
Pt = Farm level price of maize in year t;
Qd, = Amount of maize consumed in year t;
Qso, = Amount of maize that would have been produced in year t;

Qsm = Amount of maize produced in year t with hybrids;

47

 

k, = The shift in the supply curve due to maize hybrids in year t;

Y0, = The average non hybrid maize yield in year t;

Y1,t = The average hybrid maize yield in year t;

A1,t = The area planted to maize hybrids in year t;

A, = The total area planted to maize in year t;

e = price elasticity of maize supply;

11 = absolute price elasticity of maize demand; and

a and 6 represent all other omitted variables that affect the demand and

supply functions, respectively;

These equations can be used to calculate the total research benefits. Area Obc

which is equal to TB in equation 3.2 becomes;

(a + 1)2k.
2(8 + 11)

Similarly, the research benefits captured by consumers can be written as follows;

EN, = errtPuIl + ] (3J3)

 

 

1 1 + e
CBN, = Pug/(ll - E1c,(1 + c)[8 ’1" -1 + a + "1] (3.14)

The benefits that accrue to producers in each year are obtained by subtracting the
benefits to consumers (CBN,) from total research benefits (BN,).

To arrive at these measures of research benefits elasticities of demand and supply
have to be estimated. There is no comprehensive study on estimating demand and
supply functions in Zimbabwe. The elasticities used in this study, as outlined in the next
chapter, are based on estimates made by various researchers and studies from similar
market situations in other developing countries. A sensitivity analysis approach will be

used to consider different specifications of the supply and demand elasticities.

48

 

SW

The review of literature reveals that there are many methods which have been
designed to measure the returns to investments in agricultural research. The methods
are classified either ex post or ex ante. Ex post approaches are used in evaluating
research payoffs after the research has occurred while ex ante approaches are used to
evaluate likely returns to research investments before the research program has been
implemented. Because most evaluations are done on past research ex post techniques
have been used more often. A more detailed review of ex post techniques was done to
determine the appropriate method to use in assessing the returns to maize research and
extension in Zimbabwe.

Just as there are few studies on payoffs to agricultural research in Sub-Saharan
Africa, there are no studies of maize in Zimbabwe that we can draw upon as sources of
data for this study. Other than estimating the rate of return to maize research and
extension, it is also important to determine the distribution of these benefits between
producers and consumers on one hand and commercial and communal farmers on the
other. Thus data limitations and the advantage of being able to disaggregate benefits
between producers and consumers led to the choice of the economic surplus technique.
The data required to quantitatively specify the form of demand and supply functions, and
estimates of elasticities are derived from other studies in Zimbabwe and countries with

similar conditions.

49

 

W
RETUMS TO HYBRID MAIZE RESEARCH AND EXTENSIQN

M

Several studies have highlighted the fact that the growth of the hybrid maize
industry in Zimbabwe is a success story par excellence. The role of research in this
success story is an important one. The release of long season maize hybrids in 1949, and
1960 and short season varieties in the 1970’s is testimony to the creativity of Zimbabwe’s
indigenous researchers”. However, no one has quantitatively documented this hybrid
maize success story. The computation of the payoff to investment in maize research and
extension can provide objective information to DR&SS and the Ministry of Finance on
the strategic role of research and extension in national development. This chapter
documents an attempt to measure quantitatively the payoffs to investments in maize
research and extension in Zimbabwe from 1932 to 1990. This quantitative assessment is
long overdue.

The calculation of the rate of return to maize research and extension is based on
the model outlined in the last chapter. This chapter begins with a description of the data
collection process. The shift parameter is calculated and elasticities of demand and
supply are derived from estimates from other studies of the Zimbabwe maize market
and similar commodity markets in other countries. The total public maize research and
extension expenditures over the 1932-1990 period are then computed from historical
records. The computation of the total benefits from maize research and extension is a
first step in computing the rate of return to investments in maize research and extension.
The last section of this chapter present the results of an analysis of the distribution of

benefits between commercial and communal farmers.

 

2" Long season varieties take about 160 days to mature and short season varieties
take about 140 days to reach physiological maturity.

50

 

Calculatin theRateo Return

Precision is lost when calculating the benefits and costs of agricultural research
and extension if the appropriate data are unavailable or are in forms that are difficult to
disaggregate into the required components. Some assumptions have to be made in filling
these data gaps. Whenever an assumption had to be made, extra care was taken to
avoid over estimation of benefits or under estimation of the costs. This is in line with
the conservative approach that was adopted by Griliches (1964) in his pioneering study

of the payoff to hybrid maize research and extension in the United States.

Dam

Calculating a rate of return requires data on costs and benefits over time. The
maize production data collected by the Central Statistical Office (CS0) and estimates of
demand and supply elasticities are used in this study to estimate research benefits.
These data were collected during three months of field research from June to September
of 1992. Some yield data were obtained from annual reports of DR&SS’s Crop Breeding
Institute.” During field research, intensive interviews were conducted with key
participants in the maize industry to obtain a comprehensive understanding of the role of
research and extension in the development of the industry.

The production data from the commercial areas is based on annual farm census
returns”. Maize production. data for these farm types are detailed enough to allow
categorization by type of hybrid, ICA (Intensive Conservation Areas), districts, and
province. The data on the communal areas are based on crop forecasts by extension

workers in collaboration with the C80 Crop Forecasting Committee. The communal

 

2’ These were mainly the results of advanced varietal trials - the final stage in varietal
assessment before release. This also included the dates of release and the parental
combinations of varieties released to date.

3° The commercial areas here refers to large-scale and small-scale commercial areas.

The censuses are administered by extension staff and these areas are well served by the
extension service.

51

area data are not as accurate as those obtained from farm-by-farm census returns in
commercial areas (Muir-Leresche, 1992). However, the Crop Forecasting Committee
production statistics remain the best available ‘measure of production patterns in
communal areas.

The most accurate data on production in communal areas are the official GMB
intakes. Household surveys in different areas and seasons have shown that, on average,
communal farmers market about a third of their harvested maize crop in a good season
and 10 percent in a poor season”. In this study the Crop Forecasting Committee
figures are adjusted by restricting the marketed output to less than 40 percent of total
production.32

The data on costs of maize technology generation and transfer, in the form of
expenditure by government departments, were obtained from the C80. Particular
attention was paid to the changes in government accounts presentation, the changes in
structure and composition of government departments and ministries. Detailed data on
DR&SS and Agritex (and its predecessors) was also gathered to get an accurate measure
of the actual cost of generating the technologies and getting them adopted.

Normally research expenditures include both public and private sources.
However the contribution of the private sector to maize research was minimal for the
1932-1970 period. The Seed Co-op’s Rattray-Arnold Research Station became an active
participant in agricultural research in the late seventies and early eighties. However, the
impact of the Seed Co-op’s activities can be assumed to be minimal during the sample

period, because the Co-op released its first maize hybrid (8C501) in 1987. In 1989 the

 

3’ Kupfuma (1990) in Gutu; Chipika (1987-88) in Gutu; Stanning-Stark (1988) in
Bushu, Binga and Hurungwe; Shumba (1985) in Mangwende and Chivi; and MIAWD
(several years and seasons).

32 There are cases were, due to data inaccuracy, GMB maize intakes from the
communal areas were higher than 40 percent of the Crop Forecasting Committee
production figures.

52

Seed Co-Op hybrids made up two percent of the total domestic sales of maize hybrid
seed.

The maize research program did not receive international financial and/or
technical assistance from 1965 to 1980 because of the economic and socio-cultural
sanctions imposed on the country. Because the CIMMYT maize research program was
established shortly after independence, it can be assumed that it did not contribute to
maize production during the study period ending in 1990. Thus during the study period
the DR&SS and Agritex (and its predecessors) were the dominant actors in maize
research and extension in Zimbabwe. Therefore, this study shall include only public

expenditures incurred by the government of Zimbabwe on maize research and extension.

The Ship Parameter

The formula for the shift factor is shown in equation 3.12 in the last chapter.
The shift factor attempts to quantify the extent to which the supply curve shifted
downwards due to hybrids. The shift factor is a product of two parameters. The first
measures the yield change as the ratio of the yield differences (hybrid less non-hybrid)
divided by the improved (hybrid) yields. The second term is a measure of adoption with
area under maize hybrids being divided by total maize area.

Without question, it would be erroneous to attribute all the increase in maize
yield to research and extension. Institutional changes in input and output markets and
other spheres are important in that they encourage farmers to adopt improved
production practices. But it is difficult to decompose the contribution of each into its
constituent components (technological and non-technological components). However,
the yield differences used in this analysis can be assumed to be due to technology
generation and transfer as institutional changes affect both improved and unimproved
maize varieties equally.

The other important data series required in estimating the shift factor are the

yield levels without hybrids (Yo, from equation 3.12). The information on maize

53

produced by commercial farmers is sufficient as production data is broken down by
variety type. However, this information is not available for communal maize production.
Nevertheless, with the fact that hybrids in the high potential areas have a 45 percent
advantage over open pollinated varieties (Weinmann, 1975) or 68 percent during low
rainfall years (Mashingaidze, 1994) and the adoption rates available, one can proceed to
approximate the without hybrid yield levels. The without hybrid (unimproved) maize
yield levels for each year is derived from the following relationship:

Y = (YH *AH) + (YN *AN)

(4.1)
T A T

 

when YN = XYH, and 5 = AN/AT are plugged into equation 4.1 it reduces to;

 

 

Y 61’
Y” = T with YN = T (41)
6+)\(1-6) 5+).(1-6)

where;

Y.r = Total average yield for communal areas;

Ya

Average yield for areas under hybrids;

YN = Average yield for areas under non-hybrid and other varieties;

AT = Total area under maize in communal areas;

AH = Area under maize hybrids33

AN = Area under open pollinated and other varieties

A = The yield advantage of hybrids over non-hybrids
The with and without hybrid yields in the two farming sector are shown in the annex for
the period 1945 to 1990.

We now have all the parameters we need to estimate the extent of the shift in
the supply curve, k,. The shift parameter is calculated for communal and commercial

farmers using equation 3.12. The overall k, is obtained by getting a weighted average of

 

3‘3 This area is calculated from the given adoption rates of hybrid technology by
communal farmers. The adoption rates are shown in Table 3 in Chapter II.

54

the k,’s from the two farm types. The weights used are proportions of each farm type’s
contribution to total production. Table 4.1 illustrates the difference between the sub-
sector shift parameters.

Table 4.1: Zimbabwe: The shifts (k-factor) in the maize supply curve for communal and

commercial farmers, 1950-1990.

 

 

 

 

 

 

 

 

 

 

 

Year Commercial farms“ Communal farms" National"
1950 0.12 0.00 0.05
1955 0.21 0.03 0.12
1960 0.28 0.08 0.19
1965 0.28 0.13 0.20
1970 0.26 0.17 0.24
1975 0.31 0.24 0.29
1980 0.31 0.34 0.32
1985 0.31 0.38 0.35
1990 0.31 0.39 0.34

 

 

 

 

 

* The figures are rounded to two decimal places so that anything less than 0.5
percent is shown as 0.00.
Source; Table A1

It took 17 years for the benefits to maize research and extension to be realized.
The first hybrid SR-l was released in 1949 and it was adopted mainly by commercial
farmers. However, thereafter the release of new hybrids became more frequent. Table
4.1 shows that the shift parameter for commercial farmers was more than that of
communal farmers before 1980. The trend in the shift parameters for the two sectors is
also shown graphically in figure 4.1 which highlights the differences in increases of maize
supply shifts between the commercial and communal farmers. This can be explained by

the fact that all the hybrids released in the 1950’s and 1960’s were suitable for the high

55

 

potential areas where most of the commercial farmers are located.34 It was not until
the mid-seventies that the supply curve for communal farmers shifted significantly
downwards with the introduction of the short season three-way hybrids in the early

seventies.

 

040

0.354

 

0J0—

015‘

020-

00‘5 -

0.10-

035-

 

 

 

 

I1 IIIIHIIIIIIIIIIIIIIIIIIIIIIII
I965 I970 1975 1980 1985 1990
Yea

urn III
1955 1950

 

[— Cmcld Im "" OW tame “" Weld‘lted overdue J

 

Figure 4.1: Zimbabwe: Shifts in maize supply functions for communal and commercial

areas, 1950-1990.

 

3‘ These hybrids are late maturing (averaging 165 days) and have a short silk to
pollen synchronization period. In dry areas, where about 75 percent of the communal
farmers are located, these hybrids do not do well because the silk to pollen
syrlchronization period frequently coincides with a dry spell (Hussein, 1987).

56

The Demand and Supply Elasticities

The previous chapter pointed out that demand and supply elasticities are
required to estimate the change in consumer and producer welfare. Because of the lack
of information on elasticities in Zimbabwe, it is necessary to conduct a sensitivity
analysis. It is also important to highlight the effect of the dualistic nature of the
agricultural economy on both the demand and supply of maize in the country. In all
cases the elasticities used must be weighted to reflect the differential contributions of
communal and commercial farmers on the supply side and rural and urban consumers on
the demand side. The study period is long enough for some structural change to have
occurred in both demand and supply functions. For example, tastes have gradually
shifted from sorghum and millets to maize over the 58 year study period. However, we

do not have estimates of changes in elasticities of demand due to structural change.

Own-Price Elasticity of Demand

The demand for maize can be broken down into urban and rural components.
Because maize is a basic staple for both consumer groups and available in adequate
amounts in normal rainfall years, it is assumed that there is no difference in the amount
of maize consumed per capita between the groups. However, the differences in
availability of substitutes, tastes and incomes between rural and urban consumers could
result in differences in the price elasticities of demand. Studies in South Africa by Cadiz
(1984), Van Zyl (1986) and Elliott (1991) report higher price elasticities of demand for
urban consumers than for all consumers, seemingly implying lower price elasticity of
demand for rural consumers. This is in line with conventional wisdom and findings from

other countries.35

 

’5 Such studies include, Consejo de Bienestar (1965) in Venezuela and Pinstrup-
Anderson (undated) for Colombia all of which are on rice; a major food staple in both
countries.

57

Shapouri, Missiaen and Rosen (1992) is the only study that breaks down
elasticities by income groups in Zimbabwe. But they used a price elasticity of demand
for high income market that is lower than the low income market in absolute terms (-0.2
and -0.5 respectively). We abstract from the issue of correctly specifying demand
elasticities on the basis of income level. Assuming an average urbanization percentage
of 30 percent over the 58 years of the study period, 1932-90, and equivalence between
high income and urban and low income and rural, yields a weighted price elasticity of
demand of -0.41. Buccola and Sukume (1991) used a value of -0.6 as the price elasticity
of demand. All these values are within the range of the price elasticities of demand for
staples in other countries“. The value of ~04] will be used in this study as the base

level. The values of -0.2 and -1.0 will be used in the sensitivity analysis.

Price Elasticity of Supply

The price elasticities for communal and commercial farmers supplying maize are
required to obtain the weighted price elasticity of aggregate supply. The supply of
agricultural commodities is generally relatively inelastic in the short and medium term
due to the biological nature of the production process and specifically the time lag
between production decisions and harvesting. The extent to which supply is inelastic also
depends on the commercial orientation of the farmers. The commercial farmers’ supply
function is expected to be more elastic than that of communal farmers”.

The study by Shapouri, Missiaen and Rosen (1992) is the only one that has
separate estimates of price elasticity of supply curves for communal and commercial
farmers. Their estimates are 0.2 and 0.7 for communal and commercial farmers,

respectively. Although farmers respond to prices by changing area planted, actual

 

3‘ Ahmed and Sanders (1991) used values ranging from -0.25 to -0.79 for sorghum in
Sudan;

37 This is consistent with findings from work in The Philippines (FAO, 1971);
Thailand (Behrman, 1968); Bangladesh (Cummings, 1974); and Brazil (Ayer and Schuh,
1971).

58

production will be used to construct the weights. The land under maize in communal
areas constitutes over 65 percent of the total maize area in the country. However, they
produced about 40 percent of the total maize from 1945 to 1990 (see Table a.1 in the
annex). Therefore the weighted price elasticity of supply, to be used as the base level, is
0.5. Again this is within the range of price elasticities of supply used in other studies.

Buccola and Sukume (1991) and Jayne and Chisvo (1991) used levels higher than
the weighted average, 0.97 and 0.59 respectively. But these values are price elasticities of
supply of maizeintake by the Grain Marketing Board. These elasticities are probably
close to the actual ones in the case of commercial production where most of the output
is marketed. For the communal farmers who sell a small percentage of their output
these values are questionable. The values used for sensitivity analysis are the 0.2 and
1.2. The latter value was obtained from the study by Buccola and Sukume (1992).

The parameters required to estimate gross benefits using equations 3.13 and 3.14
in chapter 111 have now been calculated and the quantity OM and price P1,, after the
adoption of hybrids is taken as the actual annual maize production levels and prices over
the 58-year period, respectively. The calculation of the rate of return to hybrid

technology generation and transfer requires an estimate of the expenditures incurred.

Technology Generation and Transfer Costs
The data on the costs involved in the development and transfer of hybrids to
farmers are difficult to assemble because government expenditures are available for the
various research institutes within DR&SS rather than on a crop by crop basis. For
example, because maize research is carried out by scientists in several research
institutes38 it difficult to estimate the actual cost of maize research. One way to get

around the problem of institute-based expenditure records is to use staff establishment

 

3" Maize breeding is done in the Crop Breeding Institute, maize agronomy in the
Agronomy Institute and so on. In addition, an agronomist working on legumes or in the
Farming Systems Research Unit may spend over 40 percent of his time on maize in an
attempt to develop an appropriate system compatible with farm level practices.

59

records to estimate the person years of scientists and technicians working on maize
research and extension. However, the staff establishment records are not only
incomplete, but are only available for senior staff. Therefore, this study relies heavily on
interviews with maize researchers and research administrators on financial and human
resources outlays on maize research and extension from 1932 to 1990.

In Zimbabwe, it is estimated that 12 percent of the resources allocated to crop
research (agronomy and breeding) and research services (crop protection, seed
production and crop/soil nutrition) is spent on maize research (ISNAR, 1988: Fenner,
1992). The aggregate cost records show that the expenditures of crop research and
research services divisions make up about 60 percent of the total research expenditure of
DR&SS. Because this does not take into account the contribution of other scientists not
officially designated to work on maize, the 60 percent was adjusted upwards, to 70
percent, to take their contribution into account. Following the tradition of making
conservative assumptions in rate of return studies, the base level cost of research will be
calculated using double the percentage of the resources used on maize that was obtained
from the DR&SS directorate”. This translates to a maize research cost equal to 8
percent of the total DR&SS annual budget“. The costs of extension were calculated
using the same percentages in the absence of estimates like those obtained from
DR&SS. The sensitivity analysis will include the calculation of benefits when costs are at

6 and 12 percent of the total research and extension departments’ annual budgets.

 

3’ The tradition of conservatism refers to the common practice of understating

research benefits and taking research and extension costs at their highest level possible
as Griliches did in his pioneering 1964 study.

‘0 The 8 percent is obtained as follows:
0.12 (Crop Research Budget) = 0.12(.7(DR&SS Budget)
= 0.084(DR&SS Budget) 2: 0.08(DR&SS Budget)

60

Market Distortion Induced Costs

Governments intervene in agricultural markets for various reasons and in various
ways, incurring some costs in the process. When technology shifts the supply curve down
and to the right, the increase in marketed output also increases the amount the fiscal
authority has to pay if they do not adjust their prices and policies accordingly. In the
case of subsidies the costs to society increase with increases in output. The subsidies are
to some extent a result of using target prices to protect farmers’ incomes“. In
Zimbabwe target prices have been used to protect commercial farmers while subsidies
have benefitted urban consumers and other net food buyers through lower food prices.
This means that the maize market was not in equilibrium as shown in figure 4.2. Also
evident in figure 4.2 is the fact that the shift in the supply curve due to technology
generation and transfer resulted in an increase in the amount the government has to pay
to support this pricing policy (area PT lchCl compared to PToabPCo). The social cost of
such a policy, the dead weight loss, increases from fab to gcd after adoption of improved
technology.

The subject of estimating research benefits in distorted closed markets has been
dealt with by several researchers. Alston, Edwards and Freebairn (1988) used a model
similar to that shown in figure 4.2 with the exception that they assumed the target price
to remain fixed after the adoption of the new technology. Oehmke (1988) assumed that
the level of the subsidy does not change from the without technology case to the case
when technology is adopted.

In Zimbabwe the target price is a function of level of GMB ending stocks, which
is itself a function of season quality, and the general price level (Muir and Takavarasha,
1988; Jayne and Chisvo, 1991). The main target price determinant during the UDI
period (1965-1980) included the adoption of the Crop Diversification Program and an

 

‘1 This is only an intention. Price stabilization does not stabilize incomes if the
production is not stable. Farmers cannot stabilize production because the variability in
weather, mainly rainfall.

61

Supply Curve

 

 

 

 

 

 

Price
Without Hybrids
Supply Curve
With hybrids
Pra ................................................................... e
'c
\\\\T‘\\~‘__Demand Curve

 

 

 

Q0 Qd Ql Marketed Output

Figure 4.2; The impact of technical change on the cost of target pricing and consumer

subsidies in Zimbabwe: A closed economy model.

import substitution strategy to circumvent the economic sanctions imposed on the
Rhodesian government. With the government losing control of most of the country-side
due to rising political agitation, containing urban-consumer discontent with lower food
prices became a priority. Lower food prices were also required to support the import
substituting industries which relied on cheap labor. These factors and population growth
influenced the target price level.

Even if the above factors were not important, the seasonality of marketed maize
supply makes it difficult to determine the optimal level of maize supply the government
required and the target price to generate it. If it was possible to determine annual maize
supply targets, then the target prices that would have induced farmers to produce these
targets could be determined by plugging in the known quantity in the without research

supply equation and then solving for the price. In figure 4.2 this would equal PTa, which

62

 

150.00
140.00 - .............. ' . .“. ..............................................................................

\
120.00 - ----------

100.00 -l

00.00 J

Zlm Dollars

60.00 - '

40.00 *

 

20.00 -

 

 

 

I T

IlrrrTlrrrrrrrrr Irrlinr IlirrrrrinT
1980 1985 1990

I I I I I I I
1950 1955 1960 1965 1970 l975
M

0.00 1 r
1045

L— At 1070 Prlcee ---- At 1900 Prion}

Figure 4.3: Trend in real maize prices, 1945-1990.

in a country like Zimbabwe would equal the import parity price. The target output (01)
would be met by domestic production (00) and imports (Ql-Qo). Under this scenario the
government saves PTaecPTl.

The Alston et al. (1988) and Oehmke (1988) study do not take this approach. As
shown in the graph in figure 4.3, there is no conclusive evidence of a constant real target
price in Zimbabwe. The annual adjustments and a downward trend are clearly apparent.
The sharp increase in real prices in the early eighties and the decline from the mid-
eighties thereafter is also apparent in figure 4.3. This shows the new 1980 government’s
attempt to reduce subsidies as the expensive production system could no longer be
justified and maintained (Blackie, 1994). The Economic Structural Adjustment Program

(ESAP) which was started in 1989 includes a reduction in subsidies in agriculture.

63

The social cost (dead weight loss) of price support will depend on the assumption
on the level of the target price after adoption of hybrid maize varieties. Figure 4.3
shows evidence of a downward adjustment of the real target price from the 19503 to the
1980’s; a time when the adoption of hybrid maize technology was it its peak. The change
in the dead weight loss depends on the actual level of PT}. In this study our base case is
one where the target price is adjusted so that the dead weight loss after adoption of

hybrid maize technology is just equal to that without.

Results and Discussion

Total maize research benefits are calculated by solving equation 3.13. Research
and extension expenditures over the period 1932-1990 are also calculated as outlined
above. Gross benefits and research and extension expenditures for the study period are
shown in table 4.2.

Table 4.2 reveals that the benefits from hybrid maize research were not realized
until 17 years of research had elapsed (1932-1949). Assuming that it took an average of
ten years of research and extension to generate benefits thereafter, equation 3.1 in
chapter III is now applied. This gives us a stream of annual discounted net benefits
which can be used to calculate the net present value (NPV) and the internal rate of
return (IRR). The NPV is calculated at two levels of the real discount rate r (0.1 and
0.15). Equating the right hand side of equation 3.1 to zero and solving for r, gives the
IRR.

Returns to Research Investments

The measures of the return to and efficiency of investments in maize research
and extension used in this study include the net present value (NPV), and the internal
rate of return (IRR). These two measures, calculated with demand and supply
elasticities at base level (1;=-0.41 and e=0.5), are presented for the three cost levels in
table 4.3. This table reveals that investments in the development and transfer of the

maize hybrids have an average annual internal rate of return of 43.5 percent when the

64

Table 4.2: Zimbabwe: Gross benefits and research and extension expenditures, 1932-1990

('000 23).

[”9

[”0

 

Notes:
1. no benefits are denoted by a culmuoo.

2. For the breakdown of research and extension costs see the

Wadi!-

supply shift, demand and supply elasticities, and research and extension expenditures are
65

set at the base levels. At the cost level reported in the 1988 ISNAR report the rate of
return is 46.5 percent and is 41.2 percent when the base level costs are doubled. These
rates of return are higher than the rates of return obtained from investments in non-
agricultural industries which in Zimbabwe is in the 10 to 15 percent range (average of
12%, according to the World Bank, 1983). The investments in maize research represent
a highly efficient use of public funds during the period 1932 to 1990. Such high rates of
return on maize research reveal an under-investment of public expenditure on maize
research and extension.

Table 4.3: Zimbabwe: Returns to investments in maize research, 1932-1990.

 

 

Return Measure Base Level
Net Present Value at 15 % (Z3) 2,413
Net Present Value at 10 % (ZS) 12,740
Internal Rate of Return (%) 43.5

 

Note: The NPVs are discounted back to 1930 and are in constant 1970 thousands

Zimbabwe dollars.

Source Table A2

In terms of constant 1970 dollars, investments in maize research generated 2.413
million dollars of 1930 NPV“. This translates to 10.579 billion 1990 NPV dollars
generated as returns over the 58 year period at the base level of supply and demand
elasticities and research and extension costs. This translates into an average annual

dollar-equivalent return of 182.4 million constant 1970 dollars which is over eighteen

 

‘2 The costs and benefits are deflated using the 1970 CPI as the base level. Then they
are discounted using 1930 as the base year.

66

times higher than the total real terms DR&SS and Agritex budgets in 199043. If these
benefits could be translated into monetary values, then research and extension on maize

hybrids more than paid for the original expenditures.

Sensitivity Analysis

In this section we examine the impact of price support policies and changing
supply and demand elasticities on returns to agricultural research and extension in
Zimbabwe. We first consider changing the levels of supply and demand elasticities. We
then proceed to examine what happens to the measures of return when the supply shift
parameters are reduced by half from the levels assumed in this study.

As expected all of the nine supply and demand elasticity combinations result in
roughly the same internal rates of return. This is shown in table 4.4 and demonstrates
that the absolute size of benefits is not sensitive to changes in supply and demand
elasticities. Elasticities, however, have a significant effect on the distribution of benefits
between consumers and producers. Consumers gain more than producers when the
supply function shifts due to changes in technology. When the supply and demand
functions are relatively inelastic producers lose part of their producer surplus and
consumer gains can exceed the total research and extension benefits. While we claim
that consumers capture the bulk of the gains from research and extension, we should
note that since most producers are also consumers, the research and extension benefits
are captured by both rural and urban populations. In Zimbabwe, a significant portion of
the communal farmers, especially in the semi-arid regions, are net food-buyers.

When measuring returns to research, Echeverria (1988) included a parameter
that excludes the impact of other non-technological factors on the shift in the supply

curve. We have already argued that if non-technological factors do not affect hybrids

 

‘3 The DR&SS and Agritex budgets in 1990 were 19.55 and 42.42 million dollars,
respectively. The 1970 CPI for 1990 is 6.33 so that the combined DR&SS and Agritex
1990 budget becomes 9.79 million 1970 Zimbabwe dollars. Dividing 182.4 by 9.79 yields
18.6.

67

.r

Table 4.4: The effect of changing supply and demand elasticities on the internal rate of

 

 

return'.
Price elasticity Price elasticity of demand
of supply
-0.2 -0.4l’"I -1.0

 

I=

0.2

 

0.5..

 

1.2

 

 

 

 

 

 

Notes:
‘ These Internal Rates of Return are calculated at the base levels of maize research
and transfer costs.

” These are the base level elasticities.

and non-hybrids differently then there is no need to include such a parameter. This is
the approach taken by Akino and Hayami (1975) and the one adopted in this study. We
explore the impact of including such parameter on the measures of payoff to maize and
research and extension in Zimbabwe. We look at a case when the half of the supply
shift is attributable to other factors. The results are shown in table 4.5.

Reducing the shift parameter by half reduces the internal rate of return by just
under 10 percentage points as evident from table 4.5. The resulting rates of return (32-
37%) are still very high when compared to the opportunity cost of investment funds in
Zimbabwe during the period 1932 to 1990. The largest impact of reducing the supply
shift parameter is on the net present value, which falls by more than 50 percent (59 % at
the base cost level). However, the annual average 1970 dollars, compounded to 1990

values, is about eight times higher than the DR&SS and Agritex budgets combined and

68

 

Table 4.5: The effect of reducing the supply shift parameter by half on measures of

payoffs to research and extension.

 

 

 

 

 

Cost Level Base level supply shift Half supply shift
IRR NPV at 15% IRR NPV at
15%
Cost level 1 (High) 41.2 2,378 32.2 965
Cost level 2 (base) 43.5 2,413 34.3 1000
Cost level 3 (actual) 46.5 2,447 36.9 1,034

 

 

 

 

 

 

 

Notes:
The NPVs are expressed in constant 1970 thousand Zimbabwe dollars discounted

back to 1930.

deflated to 1970 terms. It is clear that the payoff to maize research are relatively
insensitive to changes in some crucial parameters such as the shift factor. The results

still reveal high payoff to maize hybrid research and extension.

Distribution of Hem Among Producers

The distribution of benefits between different types of producers is also of
interest because the critics of the Green Revolution argue that most of the benefits were
captured by large scale producers. Zimbabwe’s dualistic agrarian structure presents an
opportunity to examine how the benefits of the improved maize technology were
distributed between communal and commercial producers. The distribution of benefits

among communal and commercial farmers is also a subject of considerable interest to

69

policy makers in Zimbabwe.“ The 1980 government’s directive to focus DR&SS
towards the needs of communal farmers is an attempt to address the belief that the
commercial farmers were getting the lion’s share of the gains from hybrid maize research
and extension.

The dominance of commercial farmers in official maize marketing, especially
before 1980, undoubtably gave rise to this view. This view was reinforced by the
availability of single cross maize hybrids which are suitable in high potential areas. Less
than a quarter of the total communal land area is in the high potential areas. The
perception that commercial farmers were the primary beneficiaries of hybrid maize
research would be accurate if they were the only farmers using the single cross hybrids.
But some communal farmers did and still use these single hybrids. Moreover, the
release of the three-way cross hybrids in the early seventies increased the communal
farmer utilization of hybrid maize seed.

The communal farmers can be characterized as semi-subsistence farmers who
produce primarily to meet their family food requirements. Communal farmers in semi-
arid areas market less than 30 percent of their produce (Kupfuma, 1992). When
communal farmers adopt maize hybrids they gain both as producers and consumers.
Many commercial farmers retain about 15 percent of their maize crop for themselves,
their workers, animals and other uses”. Thus commercial farmers also gain as maize
consumers. Thus all producers who use hybrids gain as consumers. The focus of this
analysis is, however, not on the size of consumer surplus captured by these two groups of
producers which is a function of the amount of maize consumed on-farm. The analysis

focuses on changes in farmers’ total cash income that are brought about by changes in

 

“ Several early assessments of the Green Revolution in Asia concluded that the rural
poor (smallholder peasants and/or tenants) were made worse off. (Frankel, 1971;
Griffin, 1974). But later assessments by Hayami (1990); Sing (1990) and others have
shown that the Green Revolution has pushed wage rates, kept down food prices and
favored the poor as workers and consumers.

‘5 This is obtained from subtracting the amount marketed from the total amount
produced. The difference is assumed to represent the amount retained.

70

technology. This is inversely related to the percentage of produce that is marketed

commercially (Hayami and Herdt, 1977).

Amount consumed at home - 0c
Price
of supply without hybrids . so

maize

Supply with hybrids - $1
/

 

 

 

 

 

 

 

 

 

 

 

O QC 00 91 Quantity of maize

Figure 4.4: The impact of technical change on farmers’ cash income.

The model used in this study to relate adoption of maize hybrids to the
distribution of the benefits follows from that developed by Hayami and Herdt (1977).
Figure 4.4 depicts the change in cash income for any group of farmers as a summation of
the changes in cash sales and costs of production. The model assumes that all additional
production after technology adoption is sold (a perfectly inelastic home consumption
demand function). If this assumption is changed to allow producers to retain more
maize for home consumption, then the loss due to a fall in price would be smaller than
area abde shown in figure 4.4. Another possible formulation is to allow for different
elasticities for communal and commercial farmers with the latter having a more elastic
home consumption demand function. This study uses an inelastic home consumption

demand function for both communal and commercial farmers as a base case. From

71

figure 4.4, which assumes all additional production is sold, this can be represented as

follows:

ACAS‘ = QoflcQ” - abde Change in cash sales ..... (4.3)

ACCP: = ObQOJ - OCQlJ Change in production costs ............ (4.4)

The addition of these expressions gives a measure of the change in farmers’ cash income
due to technological change. The algebraic form of the change in farmers income is

shown in equation 4.5.

= 9,19. (e +11) [, _ (k. a.) _ (k Q...) (.5,
(8 + n - k)(8 + n(1 + k))['” 1 + 8. Qp. (8 + 1?)

 

Where P,, e, n, and k, are as defined before and;

QPt = Annual maize production by farm type

on: = Annual maize marketed by farm type

kfl = Supply shift in year t by farm type.

c, = Supply elasticity by farm type.

The results of the computation of equation 4.5 are shown in table 4.6. Three
cases were considered all involving the different levels of the supply elasticity of
commercial production. This was done because maize supply from the commercial
sector is more elastic than that from the communal areas. The three levels considered
are 0.5, 0.7 and 1.2. The latter sector’s elasticity of supply was assumed to change very
little from the 0.2 level. This resulted in weighted elasticities of supply of 0.38, 0.5 and
0.84, respectively.

Figure 4.5 looks at the trend in cash income gains by the two farm types since the
release of the first hybrid variety SR1 (See table b.3 in the appendix). Figure 4.5 shows
that communal farmers’ cash income did not change as much as did that of commercial
farmers especially during the first ten years after the release of the first hybrid. During

the sixties communal farmers cash income began to increase as their sector caught on

72

 

25

 

20.. ................................................................................... I} ...........

n

it

l\

l\

at

l \
fi ‘5 .' l .x‘
v .1 i .‘x'
O l I a t
E : l i \
o .' 1: \
On I I I \
c C : .I: ‘
£510- ......................................................................... .I\.....: ..... i'i ......
”V r/ I 1’ ii
8 x i : H
'6 i ii if
:3 . I H l:
c 5.. ..................................................................... : ...... Y." ....... 'g' ........
E {I r u I

\'r e I

 

 

 

 

 

 

 

-5IlllITllIlllrllllllllfllllllllllllllilirrj
1949 1954 1959 1964 1969 1974 1979 1934 1999
Year
[— Comerclai tamer: """ Comm! («more I

 

Figure 4.5: Zimbabwe: The effect of maize hybrids on annual farm cash income by

sector, 1949-1990.

with modern production practices. The availability of suitable varieties in the seventies
is reflected by the upswing in income gains to communal sector. The effect of favorable
complementary policies in the eighties is also particularly evident.

As pointed out before the size of income gain to farmers is an inverse function of
the percentage they sell in the market and the relative elasticity of sub-sector supply.
Commercial farmers, contrary to popular belief, did not gain as much as the communal
farmers did. In fact they could have gained more than communal farmers only if the
elasticity of supply for the commercial sector was 0.5. In the case when elasticity of
supply was 1.2 commercial farmers lose cash income. These findings confirm those

found by Hayami and Herdt (1977) in their study of rice producers in The Philippines.

73

Table 4.6: Zimbabwe: Average annual change in farmers’ income by farm type at

different total supply elasticities levels'.

Total supply elasticities . Commercial Communal
V

1,38,792 826,398

 

583,286 826,398

 

-653,371 826,398

 

 

’ The benefits are in 1990 Zimbabwe thousand dollars.

Thus it does not follow that most of the benefits were captured by commercial
farmers. It is also untrue that producers lost from the adoption of hybrids for they too
are also consumers. It is only those who sell the bulk of their produce to the market
who ended up worse off, especially if their supply elasticity is large. We have established
the fact that, after a change in technology, farmers lose as producers but can end up with
more cash income depending on the elasticity of their supply function and the
percentage marketed. In addition farmers also gain if they consume part of their

production.

m
Like in cases similar to the Zimbabwe maize research and extension situation,
the accuracy of the data used in this study has several limitations. The method used,
however, required the least amount of data as compared to other methods. The data on
commercial maize production is adequate while that on communal maize production and
that of research and extension costs are open to question. This analysis is conservative

in that the form of the shift of the supply function that is used to estimate returns to

74

hybrids is one which gives the lowest estimate. Thus if the assumptions are changed
then the estimates of benefits to research will be higher than given in this study.

The internal rate of return of 43.5 percent is greater the opportunity cost of
public funds. The net present value shows that the payoffs to research can more than
pay for the investment costs incurred. The distribution of the benefits is such that they
are captured mostly by consumers. However, because most producers are also maize
consumers, the producers also gained the largest portion of this consumer surplus. The
small farmers who do not sell most of their produce on the market increased their

income as a result of maize hybrid research and transfer.

75

CHAPTER V

INSIIIUIIQNAL FAQLORS INFLQENCINQ THE PERFORMANCE OF THE MAIZE
W

Introduction

Public investments in hybrid maize research and diffusion in Zimbabwe
dramatically increased productivity in the maize industry and the agricultural sector. But
it would be erroneous to attribute the entire increase in productivity to hybrid technology
alone without examining the institutional environment that made it possible. Goldsmith
(1991), Hayami and Ruttan (1990) and others have shown that economic progress does
not occur in an institutional vacuum. Bonnen (1990, p. 121) reached a similar conclusion
about the relationship between technical change and institutions:

It takes more than research to increase productivity. It takes an articulated-
systems approach that coordinates a broad research investment with the creation
of biological and physical capital, new human capital, new technologies and an
adaptive response to their use...(and) sustained national policy and institution
building focused on clear common goals over long periods of time. (emphasis

£144)

Thus it is important to examine the interactive relationship between technology
generation and transfer, on one hand, and the institutions and capital stocks (human and
physical), on the other.

Institutions shape the evolution of societies over time through their effect on the
processes that societies use in production and exchange (North, 1990). The process is
shaped by the rules and conventions that members of society impose on themselves
whilst pursuing some common goal to increase their well-being. The legitimacy of these
rules and conventions and the mutual trust between individuals are key determinants of
institutional stability. The rules and conventions that determine profitable opportunities
individuals can exploit, are a function of the societal power structure and the ability of

individuals in a particular society to accurately interpret their environment.

76

The development, transfer and utilization of maize hybrid technology in
Zimbabwe, represents an interactive system. The rules and conventions under which
maize hybrids were developed and transferred to farmers will be examined with a view to
inform the debate on the institutional factors affecting the productivity of research
systems. It is not possible to quantitatively assess the role of institutions in bringing
about technical change in maize production because it is difficult to estimate what would
have occurred without the institutions in place. In addition, the rules and conventions
that shaped the maize hybrid development process are not strictly maize industry-
specific. Thus it will be difficult to confine the analysis of the impact of the evolution of
institutions and organizations to maize research. However, the general institutional
issues were instrumental in the modernization of the maize industry.

The chapter begins with a discussion of the issue of the size of the maize
research team because it is one of the strategic factors affecting the ultimate
performance of a national research system. The question of size of the research team is
closely related to the issues of financing agricultural research which is addressed next.
The role of farmer organizations in mobilizing political and financial support for research
is particularly important and is given due attention. The next section synthesizes the
issues of size and financing research by addressing the issue of how to recruit, retain and
reward research workers to ensure productivity. A related issue is the management of
the scientific discovery process. The chapter closes with one of the most topical issues in
the debates on the organization of agricultural research - the division of labour between
the private and public sectors in shouldering the national responsibility of technology
generation and transfer. The implications of different mixes of public/private efforts on

the issues of size of the research team, its financing and support base are examined.

Size 0 the Maize Research Team
The size of national agricultural research systems (NARS) is a key issue in

current debates about the revitalization of these systems. Many of these NARS, due to

77

some political considerations, were expanded beyond available resources. But in a
number of countries, the expansion has proved unsustainable. For example, the
government of Mali increased the number of scientists in its NARS from 9 in 1965-69 to
275 in the mid 1980’s. Likewise, Nigeria increased the number of scientists from around
100 at independence in 1960 to around 1000 by 1985. It has subsequently reduced the
number to around 800 (Eicher, 1991).

The case of DR&SS provides some insights into this debate over the size of a
NARS. The size of DR&SS is based on the government’s desire to ensure adequate and
efficient scientific services to farmers within the capability of the national budget
(ISNAR, 1988). Because of the government emphasis on ensuring that researchers have
adequate operating budgets, DR&SS only increased the number of researchers from an
average of 107 during the years 1961-65 to a 1981-85 average of 166. In many African
countries the rapid increase in number of researchers has resulted in a high percentage

of the budget being used to pay salaries, leaving little for running expenses.

In Zimbabwe, unlike in most other African countries, the rationale has always
been that "employing fewer staff...with better salaries and resources (is the only option)
in a tightly constrained economy." (Blackie, 1984 p.11). The amount spent on a
researcher has increased consistently from the early sixties to the late seventies. The
increase in amount spent per researcher in the late seventies could be attributed to the
mass resignations triggered by the imminent changes in the political arena. The falling
research budget per researcher in Zimbabwe in the eighties, is due to the reduction in
real (inflation-adjusted) government support for DR&SS and the modest increase in the
number of researchers.

The question of size of research teams needs to be viewed in the light of the
ability of the organization to retain the experienced scientists for periods long enough for
them to be productive. Although the size of Zimbabwe’s maize breeding team has never

been larger than three person years, the achievements of these teams have been

78

Table 5.1: Annual Research Budgets per Researcher in Four African Countries, 1961-

 

 

19853
Countries 1961-65 1966-70 1971-75 1976-80 1981-85
Kenya 10.00 9.19 7.92 10.03 5.89
(129) (209) (332) (320) (462)
Zimbabwe 7.76 9.14 10.98 14.87 10.00
(107) (116) (133) (119) (166)
Nigeria 11.98 11.99 18.06 11.57 7.99
(172) (306) (348) (903) (1003)
Ghana 9.14 6.94 4.05 3.89 1.95
(70) (98) (111) (126)

 

‘ The figures are in thousands of 1980 PPP dollars while the figures in parentheses
represent the average total number of researchers.

Source: Pardey et.al (1991)

outstanding. Rattray observed that:

One remarkable feature of maize breeding in Zimbabwe is that since 1932 only
four people have been in charge of the program. [....] Thus, although the team of
maize breeders was always very small, in most cases consisting of two breeders
only, there was a considerable degree of continuity [which] has been highly
beneficial for maize breeding in Zimbabwe. (1988, p.5).

There is no doubt that this continuity was made possible by the availability of operating
funds and adequate financial rewards for the researchers. Thus it is difficult to divorce
the issues of financial rewards and resource availability from that of size. The
ISNAR/DR&SS (1988) report to the Government of Zimbabwe, recommended that the
gradual increases of professional and technical establishments should be accompanied by
an increase of financial support, in real (inflation-adjusted) terms, of 5 percent per year.
This increase was not only supposed to cover expenses related to new staff but also the
cost of keeping existing staff as dictated by normal career progression.

The paucity of government support for research is highlighted by a simple

comparison of maize research in Kenya and Zimbabwe. Kenya hired its first full-time

79

maize breeder, an expatriate, in 1955 - a time when Zimbabwe also had one maize
breeder. In 1990 Kenya had over fifteen person years engaged in maize breeding whilst
Zimbabwe had only five. In 1990, the total person years in each country’s maize
research system were about 60 and 16, for Kenya and Zimbabwe, respectively
(Echeverria, 1990). Table 5.1 shows that on average Kenyan researchers had nearly half
the research budget than Zimbabwean researchers in 1981-85. It is an empirical
question as to which maize research system has performed better than the other. What
remains clear, however, is that adequately financed research systems are more efficient
and effective in the retention of the experienced and talented scientists and in the
execution of research projects.

The size of the research team should be closely related to the level of financial
support. The major issue is how to finance national agricultural research given resource
scarcity. The high returns to agricultural research in Zimbabwe, and indeed in many
other developing and developed countries, indicates chronic under-investments in
agricultural research and is clear testimony of the need to increase financial support for

agricultural research.

Financing Agricultural Research: The Role of Client Organizations

The World Bank (1981) reported that industrial countries were allocating about
two percent of their agricultural gross domestic product to agricultural research. A
number of donors converted this into a target and urged African countries to increase
their outlay in research and achieve the 2 percent target. This level may be
unsustainable in many African countries which lack the fiscal capacity to generate such
support. About 75 percent of the research budget for Burundi’s agricultural research
system comes from donors (Eicher, 1991). Bitter experience has shown that donor
supported agricultural research projects have contributed to an expansion in the size of

NARS beyond their capacity to finance the system after donor funding is phased out.

80

 

The ability of Zimbabwe to finance its research program during the entire length
of the study period from domestic sources warrants an inquiry. We shall now examine
how the government and the various clientele groups interacted to ensure that
researchers had reasonably adequate funds to carry out research projects. We shall then
address the reasons for the current financial stress in DR&SS.

Client organizations played an instrumental role in the financing of research
through advocacy. Client organizations refer to formally recognized representative
bodies formed by individuals for the sole purpose of advancing these individuals’
interests. In Zimbabwe, groups that have interests in the operation of the research
system include farmers (commercial and communal), the seed industry and the input
supply industry. This analysis will concentrate on role of farmers’ organizations in the
evolution of rules and conventions under which maize research was conducted. The
farmers, especially commercial (both small-scale and large-scale), have been the
promoters of agricultural research because they believed in it to the extend that they
even financed it whenever possible (Tattersfields, 1992).

Zimbabwe’s two current farmer organizations reflect the country’s dualist
agrarian structure. These organizations, however, are both pursuing the common
objective of protecting and advancing the interests of farmers (Bratton, 1991). Their
formation and history, just as the farmers they represent, are diametrically different.
The Commercial Farmers’ Union (CFU) was established in 1942 to look after the
interests of large scale commercial farmers. The Zimbabwe National Farmers Union
(ZNFU) was formed in 1945 to serve the interests of small-scale commercial farmers.
The National Farmers’ Association of Zimbabwe (NFAZ) was established only in 1980
to represent communal farmers.46 The pattern of development ‘of farmer

representation in Zimbabwe has important implications for the role that each of the

 

‘6 The ZNFU and NFAZ merged in 1991 to form the Zimbabwe Farmers’ Union
(ZFU), which now represents the interests of small scale commercial and communal
farmers. This merger was orchestrated by the government which is promoting the single
union idea so that farmers could "speak with one voice".

81

unions played over time in communicating the interests of its constituent members to

researchers and policy makers.

The Role of_ Commercial Farmers’ manizations

The CPU became a national farmers’ organizations in 1942 after the
amalgamation of several provincial white commercial farmers’ associations.47 The
Mashonaland provincial associations, for example, played a key role in getting the
research department to develop maize varieties suitable for Mashonaland conditions
(Tattersfields, 1992). It is not surprising that the Mashonaland farmers formed a seed
co-operative in 1940 to distribute the improved maize varieties. When the first maize
hybrid was released in 1949 or nine years after the establishment of the Mashonaland
farmers maize seed co-operative, the Government asked the Seed Co-op to take
immediate responsibility in distributing the new hybrid seed. The farmers agreed to
shoulder'the responsibility provided that the government gave them sole rights to market
the government bred crop varieties (Rattray, 1988).48

The commercial farmers utilized their political strength even before DR&SS was
established in 1948. Their strength came from their majority status among the ruling
settler population. The driving force was to develop production systems that could
support European standards of living (Riddell, 1978). A successful client-contractor
relationship was developed and the clients were informed and participating citizens. This
illustrates the importance of having the users’ say in setting research priorities. Lord

Rothschild observed that;

 

‘7 Some of the member organizations involved in this merger include the Rhodesian
Farmers Union formed in 1909; the Grain Producers Association formed in 1919;
Tobacco Growers Association formed in 1935; and the Rhodesia Seed Maize Association
established in 1940 and which evolved into what is now The Seed Co-op of Zimbabwe.

‘8 In the absence of an institution that produces and markets seeds, the benefits from
improved varieties are often delayed. Ahmed and Sanders (1991) report that while a
hybrid sorghum variety, HD-l, was released in the Sudan in 1983, its distribution has
been constrained by the lack of a reliable seed distribution system.

82

"...however distinguished, intelligent and practical scientists may be, they cannot
be so well qualified to decide what the needs of the nation are, and their
priorities, ..... ". (Saunders, 1978; p. 7)

The responsiveness of researchers to the needs of commercial farmers - a
product of the close coordination of farmers need and research objectives - was
illustrated in the aftermath of economic sanctions that were imposed by the international
community in 1965. Because sanctions imposed a severe constraint on tobacco exports,
tobacco farmers were forced to diversify and find new crops. DR&SS responded
magnificently and developed short season three-way maize hybrids in just five years. The
new hybrids were suitable for sandy soils common on most of the tobacco farms (Billings,
1984: Mashingaidze, 1994). However, the genetic material (inbred lines, etc) for these
varieties had been developed in the 1950’s (Tattersfields, 1992). This attests to a very
cordial relationship between DR&SS and the commercial farmers - a necessary
ingredient in the development of a productive research system.

The strong DR&SS-commercial farmer relationship during most of the pre-
independence period ensured that issues related to the well-being of DR&SS were
brought to the fore in the political arena. Other than the benefit of the immediate
attention these topical issues received, this relationship had positive financial
implications. Saunders (1978, p. 5) acknowledges the provision of funds for agricultural
research "on a more generous scale" with the DR&SS budget increasing, in real terms, by
an annual average of almost 7.8 percent from 1964 to 1974 (figure 5.1). This coincided
with the time when the majority of the buildings (offices, laboratories and greenhouses)

at the DR&SS headquarters were built.

83

 

20

18‘

0’
l

14-

Real Expenditures (Z$)
(Mtfllons)

 

 

‘q
"0

2 I I I I I I I I I I I I I I I I I I I I
1954 1959 1964 1969 1974
Year

 

I I I

ITFIIIIIIIII
1979 1984 1989

 

+ Research costs —+— Extension costs —>K- Total costs

 

 

 

Figure 5.1: Zimbabwe: Real research and extension expenditures from 1954 to

1990 (1980= 100)

The CPU and other farmer organizations have provided about 10 percent of
DR&SS’s operational expenses since it was formed in 1948.49 The research system has
benefited from the close relationship with the commercial farmers. The farmers used
their political clout to ensure that DR&SS obtained appropriate financial resources from
the Treasury, which makes decisions on the overall allocation of public funds. The

commercial farmers were particularly effective prior to independence in 1980.

 

‘9 For example in 1990/91 financial year the CPU contributed Z$323,000 to
DR&SS’s recurrent expenditure. The government contributed Z$24,888,000 of which
235,950,000 was allocated for recurrent expenditure.

84

Commercial farmers have also been actively involved in the debate on the scope
and priorities of DR&SS since from its formation in 1948. Commercial farmers
supported turning some DR&SS units into independent entities. For example, the
conversion of the Tobacco Branch of DR&SS into an independent research organization
by the Tobacco Research Board, whilst considered divisive by scientists50 was
considered by commercial farmers to be the best way to serve the interests of the
commercial farming community.

Whilst most of the advocacy of commercial farmers was done informally by
commodity groups, the involvement of farmers in DR&SS was formalized by the
formation of the Agricultural Research Council in 1961. In 1970 the ARC’s structure
was altered to give farmers some advisory powers on the planning and priority setting in
research institutes. Three years later, a sub-committee was set—up on communal
agriculture to redress the institutional illegitimacy at a time when it became apparent
that the ultimate success of a national research system is related to its ability to serve
both communal and commercial farmers?l

In the 1970’s the political future of the minority white government, which the
farmers controlled, was becoming uncertain. This uncertainty fueled a drive by
commercial farmers to turn DR&SS into a independent research unit. The

reorganization of the ARC in 1975 gave farmers a greater role in the determination of

 

5° Many Zimbabwean scientists favored the Land Grant system in the USA. in
which research, extension and education (universities) were coordinated by a single
administrator - the Dean of the College of Agriculture. Zimbabwean farmers, on the
other hand, advocated the reorganization of research along commodity lines which
DR&SS researchers felt would result in unnecessary duplication of effort.

5‘ North (1990) defines institutional illegitimacy in terms of how popular rules and
conventions that govern societal production and exchange activities are among the broad
spectrum of society.

85

research priorities and expenditures (DR&SS, 1975). The climax of attempts to turn
DR&SS into an independent unit came with the appointment of a Commission of
Inquiry under Dr. Graham Child in 1978. The Child Commission conducted an in-depth
study of DR&SS and recommended turning it into an independent unit (ARC, 1979).

When the Child Report was completed in 1979, the government delayed a
decision because it was pre-occupied with the escalating political unrest. With the
advent of independence in 1980, the ARC made a spirited attempt to get the new
government to accept the recommendations of the Child Report. The 1980 ARC annual
report documented the strategic contribution of agricultural research in increasing crop
yields. In 1981 the ARC reported that "...it was with some relief that the Council was
informed [that] the Prime Minister [Robert Mugabe]....had decided against" the concept
of a statutory DR&SS. This announcement cleared most of the uncertainty prevailing
among professional staff members allowing them to "reassess the position and plan for
the future" (ARC, 1981). In the next three years (1981 to 1983) there was a mass
resignation of experienced white researcher officers was followed by an inflow by young
black officers. During the 1981-83 period, white professional officers made up nearly 84
percent of total staff resignations and only 21 percent of the newly appointed officers
(ISNAR, 1988).

The loss of influence by commercial farmers over research policy formulation, in
the 1980’s, is not surprising because DR&SS was not serving the needs of the majority of
farmers. This is the fate of institutions that do not have the support of the broad
spectrum of society. The challenge is to establish a set of popularly acceptable rules and
conventions which will serve the agricultural industry in its entirety. Farmers are
interested in a system that offers them opportunity sets they can all exploit to their

advantage. For most of the time, communal farmers’ interests were relegated to the

86

sidelines. Their new representative body , the Zimbabwe Farmers Union (ZFU)52 will

have to take a stance on the products they require from the DR&SS.

13 Role at Communal Farmers’ Qrganizations

To understand why communal farmers did not have a national organization to
advance their interests before 1980, one needs to examine the political evolution of the
country from the time of European occupation. The settlers drove most of the
indigenous population out of the fertile and high rainfall areas starting with the Land
Apportionment Act of 1931 (Riddel, 1978). Since the land issue was the main grievance
of communal farmers, any organization that represented the interests of communal
farmers during the UDI period, ran the risk of being identified with the liberation
movement which the minority government had outlawed. The liberation movement
considered land as the root cause of economic inequality between black and white
Zimbabweans.

NFAZ was established in 1980 as an outgrth of the non-political Master
Farmer movement that was active in the Masvingo Province (Herbst, 1990). Although it
is extremely difficult to transform a technically-oriented provincial organization into a
national organization which articulated the political interests of communal farmers,
NFAZ expanded rapidly in early 1980’s with a membership base that was national in
character. However, although NFAZ was considered as the legitimate representative of
communal farmers, it was dogged by financial difficulties throughout the 1980’s because

it only attracted about 6 percent or 65 000 paid-up members out of a possible million

 

52 ZFU is a product of the amalgamation of NFAZ (National Farmers Association of
Zimbabwe for communal farmers) and ZNFU (Zimbabwe National Farmers Union for
small-scale commercial farmers) in August of 1991.

87

communal households (Bratton, 1991). Now that NFAZ has been incorporated into
ZFNU, it is an open question on how well the interests of one million communal farm
families will be represented in this new organization.

It appears that, given these institutional weaknesses, the interests of communal
farmers will be difficult to communicate to institutions like the DR&SS. Some small-
scale farmers, after the merger, will become officials who have to articulate the interests
of communal farmers in addition to those of small-scale commercial agriculture. This
poses potential difficulties if, as expected, these small-scale farmers do not understand
the problems of communal agriculture. The question of whose interests will count
remains an open one given the experience of how effective the former NFAZ has been
in drawing the attention of policy makers. Some people speculate that the small-scale
commercial farmers, despite the pressure from the government, decided to merge with
the NFAZ to take advantage of NFAZ’s national transparency (Bratton, 1991).

ZFU could help in dispelling the pre-independence (1980) notion that the
provision of extension services is the most effective way to develop communal areas.
The importance attached to extension, and not to research, as an answer to communal
agricultural production problems is evident in the nature of development work carried
out in these areas from the time Alvold was appointed to establish the Department of
Native Agriculture to the formation of an ARC sub-committee on communal agriculture
in 1973. The 1973 ARC sub-committee was established after the limited adoption of
Alvold system which emphasized improved production techniques. Sociological and
economic dimensions of communal farmer behavior were considered to be the major
constraints on the development of communal agriculture. The development of

appropriate technologies was not seen as important as research programs to understand

88

peasant behavior patterns. Extension was assumed to be the best vehicle to change
behavior patterns of communal farmers.

Figure 5.2 shows the increase in extension expenditures for the department
(DEVAG) working in communal areas as compared to the department working in
commercial areas (CONEX). Thus if the problems of communal agriculture are
correctly identified as research issues then it is possible that DR&SS and other research
organizations will see their research budgets increase. The ZFU can play an important

role in getting policy makers to correctly identify the complex production problems in

12000

 

10000-

8000-

6000‘

4000‘

Expenditures ('0 O O Z$)

2000 - A 4‘

 

 

 

I

o I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 7
1950 1955 1960 1965 1970 1975 1980
Year

 

"I- CONEX + DEVAG

 

 

 

Figure 5.2: Zimbabwe: Expenditures in extension by sector from 1950 to 1982.

communal areas that require sustained research attention. Extension, after all, is

89

effective only if on-shelf technology is available.53 Future maize production grth will
have to come from new technologies tailored to the communal conditions - technologies
that are not currently available on-shelf. Unless the ZFU representing communal
farmers is able to bring pressure to bear on policy makers to allocate more funds to
DR&SS for technology development for communal conditions, then the benefits
communal farmers will get from DR&SS will continue to be limited. The other option is
for ZFU to provide research funds to DR&SS or other research organizations to conduct
research that directly benefit its members. Given limited financial resources, this option
is remote. The bottom line is that ZFU needs to create a client-contractor relationship
with DR&SS that allows its voice to be heard in setting agricultural research priorities.

The active voice of clients in agricultural research programming is essential in
ensuring that research programs are relevant, timely and productive. We have
demonstrated how the influence of commercial farmers on DR&SS impacted positively
on research programs especially major ones such as the maize research program. The
main benefits of getting clients involved in the programming of research include:

* the assurance that high priority problems of farmers are identified and are

tackled;

* the availability of financial resources from farmer organizations and the

treasury;

"‘ more rapid dissemination of technologies;

* the improvements in research organization through exertion of pressure on

policy makers.

 

’3 The coincidental availability of three-way hybrids that spurred productivity
increases in communal maize production in the 1970’s may have masked the absence of
research work for conditions in the majority of communal areas.

90

Management of Human gaital in Am’ ultuml Research Institutions

The rules and conventions which govern human capital stocks and its
management are important determinants of the productivity of research programs. This
mainly relates to the incentive structure which is used to recruit, reward and retain the
scientific manpower. It also is concerned with the management of the work-place in
terms of flexibility and accountability. Lastly, the training and experience level of
researchers are important determinants of the innovativeness of the research system.
These issues will now be discussed in as far as they apply to the Zimbabwe research

system and especially the maize research program.

Recruitment, Rewarding and Retention of Researchers
Lord Zuckerman, as quoted in Saunders (1978), highlighted the national

importance of retaining productive research scientists:

The capital of any research organization lies in its brains and it is vital to provide
the right climate for the recruitment and retention of good scientist. (p.8)

Thus to maintain a productive research program, experienced and talented scientists
must be developed, motivated and retained (Elliott and Kang, 1985). It takes a long
time for newly recruited scientist to be productive. The research institution should get
returns from its in-service training by retaining scientist for a time long enough for them
to be productive in the same manner that firms capture benefits from investments in
long-term physical capital. Otherwise it becomes a training ground for other institutions,
especially in the private sector. This can be considered a loss to society given higher
rates of return on investments in public utilities such as agricultural research than those

in the private research organizations.

91

DR&SS has been successful in maintaining research continuity by retaining its
experienced scientists in its maize research program. Only four researchers have led the
maize breeding program from 1932 when it was initiated until 1988“. This continuity is
quite remarkable and is an envy of many research directors in other countries.

Table 5.2: Conditions of service for researchers in Zimbabwe (1984), Dominican

Republic and Thailand (1983).

 

 

Item Zimbabwe Dominican Rep. Thailand
Salary ratio: Entry/ Max 0.18 0.33 0.17
Salary ratio: Entry/ Mean 0.42 0.62 0.39
Average Years of Service 8.68 6.44 11.48
Years Served by Oldest Researcher 41 21 37
Average Age 34.9 32.9 36.2

 

Source, Elliott and Kang (1985)

Table 5.2 shows the findings of an ISNAR study on conditions of service in
Zimbabwe, the Dominican Republic and Thailand.55 The study is based on the premise
that the ability to recruit, retain and motivate researchers is mainly a function of the
salary structure. The length of service and education levels6 were found to significantly
affect earnings .in Zimbabwe were. Table 5.2 reveals that Thailand rewards its

experienced staff more than Zimbabwe does with a lower ratio of entry level salary to

 

5‘ The maize breeding program was led by Arnold (1932-1950), Rattray (1950-1968),
Nelson (1969-1975) and Olver (1975-1988).

55 The Thailand system is designed to reduce turnovers and the Dominican Republic
is a typical case of a developing country.

5‘ In Zimbabwe a PhD earns significantly more than a MSc, who in turn earns
significantly more than a BSc. The coefficient on a the years of schooling variable was
found to be highly significant (t(,_025 = 13.63).

92

the level earned by the most senior researchers. In the Dominican Republic the
retention of researchers is problematic as the average years of service (6.44) is lower
than in both Zimbabwe and Thailand. Thus staff retention is clearly a function of how
experience and higher academic attainment are rewarded in the salary structure.

Whilst the reward structure is important, it is well-known that government
salaries for scientists tend to level off faster than they do in the private sector (Saunders,
1978). Thus the salary structure must not only reward experience and academic
attainments, but also consider the relative structures of salaries in various sectors that
can employ researchers. Figure 5.3 shows the gap between the annual salaries .of a
research technician (diploma holders), chief research officer (more than ten years as a
researcher with at least a BSc degree) and the director of DR&SS and the secretary of
agriculture.”

Figure 5.3 covers the period 1930 to 1964 because the salary data was not
recorded in government expenditure reports after 1964. The figure shows that the chief
research officer and the DR&SS director received about the same salary as the Secretary
of Agriculture58 during this 34 year period. Thus the research director and his chief
research officers were given salaries that are at parity with those of the most senior '
public servants. The relatively high salaries of chief research officers meant that the
experienced scientists could spend their entire career in research. This helps explain why

some researchers stayed with DR&SS for more than 40 years as shown in table 5.2.

 

’7 Because of the difficulties of securing data on the salaries for researchers in the
private sector, the salary of the Secretary of Agriculture is used as the ratio’s numerator.

5’ From 1954 to 1969 the salary of the Secretary of Agriculture, the highest civil
servant in the ministry, was higher than that of the Minister of Agriculture.

93

\l

 

O
l

on
1

b
I

0‘
L

N
1

it ¥*¥;E-; A‘P+—+—P++¥‘+

a.
1

Secretary of Agric / Researcher salary

 

 

O

 

I I I I I I I I I I I I I I I I I I I I I I I I I T I I I I I I

I I I
1930 1935 1940 1945 1950 1955 1960
Year

 

+ Junior officer "-1-" Senior officer “*6" Director of DIES

 

 

 

Figure 5.3: Zimbabwe: Salary ratios for researchers from 1930 to 1964.

The salary of junior research officers was about one fifth that of their senior
counterparts during the period 1930 to 1964. The increase in salaries with experience
was quite rapid. During this period it took eighteen years for a junior officer to earn the
highest salary of a research officer. This on average implies a doubling in salary in
every three years. This is equal to an average annual increase in salary of about 20
percent. The percentage increase is normally higher in the earlier years (Elliott and
Kang, 1986). This and the relatively high rewards for higher academic attainment kept
young researchers at DR&SS.

Most ARC annual reports have decried the falling levels of real (inflation-
adjusted) salaries from the 19708. The ARC, for example, reported that "salaries were

lagging far behind commerce and industry and the wage freeze resulted in a

94

retrogression of living standards as the costs of...[living] escalated monthly" (ARC, 1982,
p. 6). This was one of the three principal reason for the heavy loss of staff DR&SS in
the early eighties.

Many of the white research officers left immediately following independence in
anticipation of the promotion of less experienced black officers to positions of authority
in order to help redress racial imbalances in the civil service. Indeed the number of
research officers leaving DR&SS fell from an annual average of about 26 in the early
eighties (1981-1983) to 15 in the mid eighties (1984-1986) (ISNAR, 1988). This figure
increased in the late eighties to 19 (1987-89). The turnover levels are likely to increase in
the 1990’s as the research budget squeeze becomes tighter with the structural adjustment
program in place.

The salary structure has also been changed so that a chief research officer earns
2.32 times the salary of a junior officer compared to 5 times in the period 1930 to 1964.
In the late eighties, a chief research officer could double his salary by joining the private
sector (Fenner, 1990). This means that experience and academic achievement are not
being rewarded as they used to be in the period 1930 to 1964. There is currently little
incentive for researchers to stay with DR&SS. The research base has been depleted by
the mass exodus of experienced research officers in the early eighties. It appears that
the staff recruitment exercise instituted after independence in 1980 will not be a fruitful
exercise unless something is done to reduce staff turnover.

The rewarding of scientists is a difficult exercise given the problems of relating
remuneration level to individual performance. Although measuring scientific
performance is not easy, the failure to reward scientists is costly to society given the high
social returns from research and extension. The issue of measuring individual

performance is related to that of managing the scientific discovery process. Whilst

95

adequate physical resources for scientists are important, the stability of funding for the
organization and the continuity of scientific leadership is important given the long

gestation periods for agricultural research”.

Managm g the SM ' Discovgz Process

The motivation of agricultural scientists to be productive is a subject that is not
explored in the development literature in Africa. While not denying the possible
existence of individuals who derive utility from the opportunity to venture into the
mysteries of science, the leadership style employed in the management of researchers
play a crucial role in what is produced by the research system. Most of the issues in this
area falls into the realms of management, psychology and sociology. The discussion will
abstract from the sociological and psychological dimension of this subject and tackle the
issues of leadership and staff advancement in a simplistic but realistic approach.

The management of researchers is admittedly complex. In most instances
informal patterns of leadership and communications are recommended in place of very
formal ones. The important point, however, is that the management system should meet
the needs of the researchers involved in terms of motivating them and providing job
satisfaction and opportunities to advance their professional careers (Saunders, 1978).

Flexibility in management recognizes that the scientific discovery process is, in
many ways, a random walk process with breakthroughs occurring, sometimes, from pure
serendipity. Yet there is danger in having ”total freedom". The abundance of areas and
problems a scientist can pursue is limited by the scarcity of resources required to decide

what to do, doing it and getting the results to the clientele (Dagg, 1992). While

 

5’ The gestation period for crop research on average is 10 years and 15 years for
livestock research.

96

researchers should not operate under a rigid framework or in completely unfocused
conditions, they need to be guided "like a snake in a tunnel". The maize research teams
from the 1940’s pursued the objective of increasing maize production by producing
hybrids. The clear focus on productivity and hybrids produced results which may have
taken more resources if the policy was broad enough to allow breeders to work with both
hybrid and non-hybrid varieties.

Researchers also need to advance their careers. They may need to publish their
work in publications other than the official ones for professional reasons. Allowing
researchers to pursue higher studies, attend meetings with colleagues inside and outside
the country helps in the interchange of ideas and experiences enriching the knowledge
base of the research staff‘”.

The financial rewards as already pointed out, are essential for keeping talented
scientists engaged in research. If the financial rewards are not adequate for researchers,
there will likely be a loss of gifted scientists to administration duties. Much of the
success of the research system in Zimbabwe prior to independence is attributed to a
reward structure that kept researchers at work for decades, allowing full extraction of the

benefits of cumulative knowledge.

The Public-Private Sector Division at Labour

The emergence of the private sector is best understood if one takes a

evolutionary perspective of the process of agricultural modernization (Pineiro, 1986).

 

6° The experience of scientists in Zimbabwe before 1980 may appear to contradict
this view. Researchers had fewer overseas visits and training yet they managed to
produce world class scientific results. Their research agenda was not as demanding as the
one shouldered by DR&SS since 1980. In addition many feel that there were key people
in each area who trained most of the new, young researchers. Most of these key people
left DR&SS around 1980.

97

The results of work carried out by research institutions are instrumental in bringing
about the modernization of agricultural production. The stage at which a particular
country’s modernization process is determines the nature of agricultural research that is
carried out and this in turn determines the nature of the research organization to carry it
out. Pineiro (1986) defines these stages as technological cycles identified by the type of
the most dominant technology that the research system focuses on during a particular
time period. In the earlier stages of the modernization process, adaptive and applied
research are the most dominant type of agricultural research undertaken. When the
research system is focused on defining agronomic practices the profit-seeking private
firms are not interested in research results because the creation and diffusion of these
production practices have all the properties of a public good“. Gradually the
modernization process induces research institutes to focus on mechanical, chemical and
biological technologies. The evolution opens opportunities for the private sector to
exploit as some of the technologies require industrial production and some distribution
infrastructure. Economic progress gradually forces the public research system to focus
more on upstream type of research while the private firms do most of the applied and
adaptive research. The private sector becomes an active participant in national
agricultural research.

The sophistication level of the research agenda increases with modernization.
Most important is the emergence of the private sector from an complimentary role to a
more or less competitive one. This gives rise to two crucial issues. First, the

participation of the private sector will force the public research institutions to evolve in

 

‘1 Public goods are characterized by properties of non-exclusivity and non-rivalry.
There are no incentives for private firms to invest in the production of public goods
because they cannot internalize the benefits (Griliches, 1964).

98

ways which permit them to interact with the private sector more effectively. The second
but related issue is the need to compete with the new private research firms for both
financial and scientific resources. These are some of the issues that the maize research
program in Zimbabwe has to address, given the level of sophistication it has attained.

The early stages of maize research in Zimbabwe covers the time from 1900 to
1930 when farmers were trying to find ways of growing the crop. They tested production
practices and materials used in their home countries (Europe). This was conveniently
done by the government’s Department of Agriculture. When the hybrid breeding
program was started in 1932, it only took eight years for the formation of a non-
government organization, the Maize Seed Cooperative, to help with the distribution of
the new maize hybrids to farmers when they became available. The development of the
fertilizer and other crop chemical industries greatly facilitated the diffusion of maize
hybrids in the 1950’s and 1960’s (Rattray, 1988). This demonstrates that the public
research institution like DR&SS must work closely with the private input supply firms if
technologies are to reach farmers quickly.

The fertilizer and other input supply companies did not present a competitive
threat to DR&SS. However, the establishment of a maize breeding program at Rattray-
Amold Research Station by the Seed Co-op in 1974, which has grown in size to the level
of the DR&SS maize breeding program by the late eighties, is an example of a
competitive private initiative. Even though ART farm is seen as a compliment to
DR&SS, the establishment of the maize agronomy at ART Farm in 1981 is clear
evidence of how the evolution process give rise to the emergence of the private sector as
a competitor with the public research institution (Pineiro, 1986). The emergence of
private sector research is a diminishing image of DR&SS as an institute that generates

high utility technologies. To the public, especially politicians, DR&SS has a diffuse role

99

as the private research firms are seen as the source of most of the technological inputs
and advice farmers use in production. Most damaging is the rift that private firms have
created between DR&SS and the farmers who have been the key voice for adequate
financial support for the department.62

The private sector has also been a strong competitor with DR&SS for scarce and
expensive human resources. The managers and scientists in most of Zimbabwe’s private
research organizations are former DR&SS personnel who were attracted by the higher
salaries and a chance to escape the institutional rigidities of public service. If DR&SS is
interested in avoiding the progressive depletion of its research capacity, then it has to
redress the institutional rigidities of the civil service, especially the salary structure.
Otherwise it will continue to serve as a training ground for the private sector firms. One
option DR&SS can consider is to turn itself into an independent body that is not subject
to civil service rigidities.

The private sector has not only diminished the perceived importance of DR&SS,
but present a big dilemma to policy makers. The question is whether public research
institutions should continue to work on applied and adaptive or basic research. Applied
and adaptive research results are tangible practical products which can restore the image
of DR&SS as a productive institution. But these are the activities private firms

concentrate on because the products have a market. The public sector can avoid

 

‘2 The presence of private research firms diverts attention from institutions like
DR&SS. During a 1988 cereal conference by the CPU dominated by the issue of
funding agricultural research, Permanent Secretary Ndimande had to remind the
commercial farmers that ”DR&SS is there to serve the whole state" (The Farmer, July
1988, p.10).

100

 

duplication of effort by concentrating on other lines of research.63 Pineiro (1986) and
Jaffee, et a1 (1991) identified the areas that, depending on the stage of economic
development, the public research institutions should concentrate on, to include basic
research and research for marginal farmers/conditions. The case of DR&SS warrants
similar considerations especially with respect to maize research and the semi-arid regions
where the bulk of communal farmers are found.

DR&SS faces a changing agricultural industry. The department should change its
mode of operation to ensure that it continues to play a meaningful role in the continuing
evolution of the country’s agricultural sector. Changing the rules and conventions that
DR&SS operates under, as in any case involving institutional change, will be a protracted
process. Institutions do not change overnight (North, 1990). Yet it is vital that DR&SS
change its orientation to ensure efficiency in resource use, change its organizational
structure to position itself adequately to compete for financial and human resources with
the private sector, and reassess the role it can play in further promotion of the
modernization of agriculture, especially in communal areas.

The question many people are asking is how much attention should DR&SS give,
in terms of financial and manpower resources, to research programs where the private
sector is no longer a complement to public efforts but an outright competitor. Several
issues have to be considered in the search for answers to this question. For research

programs on commodities that are as entrenched as traditional foods, it is politically

 

‘3 Some do not view duplication as a problem. Professor Malcolm Blackie as
chairman of the ARC argued "I suggest duplication is not a problem. As an
unreconstructed neo-classical economist, I believe in the avoidance of monopoly even
amongst researchers.” (1984, p.10). Richard R. Nelson concurs; "...I maintain that by and
large the experience is that technical advance proceeds much more rapidly when a
considerable number of parties are engaged in competitive efforts, than in a context
where one or a few parties are in a position to control developments". (1990, p.61)

101

suicidal for a nation at Zimbabwe’s stage of development to entrust research
responsibility for these commodities to private firms. We also know that it is important
to preserve both the public and private good aspects of technology (Nelson, 1990).
These two aspects are complements in that private firms can thrive from the basic
information generated by the public sector. This public information enables private firms
to provide consumers with higher quality and relatively inexpensive product. Therefore
we can conclude that it is unwise for DR&SS to discontinue financing maize research.
The other question concerning the participation of private firms in agricultural
research relates to what DR&SS’s priorities should be. We know that DR&SS should
not discontinue its maize breeding program. Maize research efforts should focus on
maize in the areas that are considered unsuitable for growing maize as we have seen an
almost irreversible change in tastes from sorghum and millet to maize. The DR&SS
should not also drop maize research in the high potential regions because maize is a
political crop. We also know that private firms are less likely to invest in research that is
targeted for the marginal areas given the high risk of poor financial returns. Therefore,

DR&SS should focus on the problems of production in the communal areas.

m
The process of technological change is shaped by the rules and conventions that
guide the production and exchange activities of its members. Public institutions such as
DR&SS are particularly important in cases involving the provision of public goods. The
development and transfer of maize hybrids in Zimbabwe came about through an
interactive process involving various public and private institutions. Particularly
important was the role of commercial farmers in the political process. The development

of DR&SS to an unquestioned position of leadership in Africa was greatly facilitated by

102

 

political support from commercial farmers. Using their powerful position, commercial
farmers were able to ensure that DR&SS received enough financial support from
domestic sources. In other African countries donor supported research programs proved
unsustainable. The role played by farmers was important in Zimbabwe.

Before 1980 communal farmers could not influence the policies and operations of
DR&SS because of the political rules then did not allow the existence of a legitimate
representative organization for communal farmers. The change in government in 1980
brought the issues affecting communal agriculture to the fore in DR&SS activities. This
has left DR&SS is a state in which it no longer enjoys a high level of support from
commercial farmers. Unless communal farmers become active advocates for research for
their environments, then they will not systematically capture the full potential of new
technology.

Four critical institutional issues affected maize research program: size and
continuity of maize research; the financing of research and technology transfer; the
management of the scientific discovery process in terms of recruiting, rewarding and
retaining researchers and the role of the private sector.

The key issue is that Zimbabwe opted for a small number of maize researchers in
the 19308 and kept them adequately financed and supported for decades. It is important
to ensure that all researchers are adequately rewarded in terms of salary, career
prospects and a conducive working environment. The objective is to create an enabling
environment so that scientists stay long enough to make a contribution given the average
gestation of a decade or longer to produce new crop varieties. A comparison of Kenya
and Zimbabwe’s maize research programs showed that the former increased the number

of researchers without a consumerate increases in research budget. While it remains an

103

empirical question as to which system is more productive, it is clear that researchers
require stability of funding to carry out their research projects.

The high productivity of DR&SS, especially its maize research program, has been
achieved through administrative, financial and scientific continuity. Before the eighties,
the salary structure rewarded experience and academic achievements. Scientists stayed
longer with DR&SS as demonstrated by only four leaders of its maize breeding research
from 1932 to 1988. The clear focus on hybrids played a pivotal role. The importance of
financial rewards and prospects of professional advancement of scientists cannot be
overemphasized.

Today there are signs of stress in DR&SS in the nineties. Financial support has
fallen in real terms and the salary structure is no longer attractive for experienced
scientists to stay with DR&SS. The renewal of DR&SS will have to give high priority to
rebuilding DR&SS’s human capital base and retaining it through a greatly overhauled
incentive structure including support for research, travel to scientific conferences, library
and computer support and overseas training.

The role of private research firms in agricultural modernization is important and
in the process of being reexamined. In Zimbabwe the private sector started conducting
maize research in the seventies after DR&SS had successfully developed embodied
technologies which could be produced and/or distributed in ways that yield profits to
private firms. Today DR&SS faces evolving research needs and should face up to the
competition with the private sector for financial and human resources and to redefine its
focus of research work. Specifically DR&SS, while not abandoning maize research
outrightly should concentrate its work in the marginal areas where private firms are

reluctant to venture into. DR&SS can compete more effectively for human and financial

104

resources with the private firms if it sheds its civil service status. This will enable it to

design competitive salary structures as well as other ways of securing funds for research.

105

W
CONCLUDING REMARKS

Maize is a crap of strategic importance in Zimbabwe. For the communal
farmers, especially, it is a source of both food and income. Thus increasing maize
production will not only ensure adequate food supply but also increase incomes. This
increase in production can be realized from yield improvement research as area
expansion is no longer possible. The communal farmers in the high potential areas have
benefited most from past yield enhancing research. But improved maize varieties are
not available for marginal area in which most communal farmers are found.

The development of maize hybrids and agronomic practices was accomplished by
careful sequencing of breeding and crop management research. A number of institutions
were already in place or subsequently established to enable maximum utilization of the
new hybrids starting in the 1950s. These include extension, credit institutions, input and
output marketing organizations. These organizations along with research and extension
make up successful technology generation and transfer systems.

The costs and benefits of technology generation and transfer need to be
quantified to guide policy on future investments. But there are few studies on payoffs to
research that have been conducted and published in Africa. These studies are important
in guiding research allocation and planning decisions at the research department and
national levels.

There are several approaches available to evaluate quantitatively the payoff of
maize research. However, by definition, ex post approaches are better at addressing the
question of past research productivity than ex ante approaches. The latter are good as
tools of research planning. The ex post techniques that employ the economic surplus

approach are more appropriate in assessing the returns to maize research in Zimbabwe.

106

But estimates of supply and demand elasticities are required. Since there are no studies
on maize in Zimbabwe that specify quantitatively the form of demand and supply
functions, estimates of elasticities are derived from studies in Zimbabwe and other
countries. A sensitivity analysis was used to test the effect of using other levels of supply
and demand elasticities.

The accuracy of the communal area and cost data is open to question. Also
because conservative assumptions were used the results of this analysis represent a
lower-bound estimate of actual returns. The downward bias was incorporated in the
specification of the shift of the supply function and the research and extension
expenditures.

The estimates of payoffs to maize research and extension under these
conservative assumptions are very high. The annual internal rate of return was
estimated to be 43.5 percent which means that for every Zimbabwe dollar invested in
hybrid maize research yielded an annual return of 43.5 cents. This is an impressive
achievement and it is greater than the opportunity cost of public funds. Also, the Net
Present Value shows that the benefits from maize research and extension can more than
pay for the initial investment costs. Most of these benefits were captured by consumers
through lower maize prices. Because the majority of producers are also consumers
and/or net food buyers, the producers gained most of the consumer surplus. In addition
communal farmers, who do not sell most of their produce and have relatively inelastic
maize supply functions, increased their cash income as a result of maize hybrid research
and transfer.

The process of technological change is shaped by the rules and conventions that
society designs to guide the production and exchange activities of its members.

Institutions are particularly important in cases involving the provision of public goods.

107

 

The development and transfer of maize hybrids in Zimbabwe came about through an
interactive process involving various public and private institutions. The institutional
issues affecting technology generation and transfer include the size of the maize
research, financing of research and technology transfer, management of the scientific
discovery process in terms of recruiting, rewarding and retaining researchers and the
public and private sector division of labor. Decisions on these issues were shaped by the
nature of rules and conventions prevailing then and their subsequent evolution over time.

The size of the maize research team should be decided based on the expected
payoff of the research project being undertaken. It is also important that researchers are
adequately rewarded so that they stay long enough if they are to make a contribution. A
comparison of Kenya and Zimbabwe maize research programs showed that the former
increased the number researchers without a commensurate increase in research budget.
While it remains an empirical question as to which system is productive it is clear that
researchers are only productive if they have enough funds to carry out research projects
and are adequately rewarded.

The financing of development and transfer of maize hybrids starting in the 19508
was greatly facilitated by commercial farmers who exploited their powerful position
within the polity effectively and ensured that DR&SS received enough financial support
from the national treasury. They did not rely on donor support. Experience from other
countries has shown that donor supported research programs have a tendency to increase
and sustain dependency.

However, before 1980, communal farmers could not influence the policies and
operations of DR&SS because the political rules then did not permit communal farmers
to form a legitimate representative organization. The change in government in 1980

brought the issues affecting communal agriculture to the fore within DR&SS without an

108

 

input from the communal farmer organizations which became legal. This has left
DR&SS is a state in which it no longer enjoys the support it usually got from commercial
farmers. Unless communal farmers become active advocates for research for their
environments, then they will not systematically capture the full potential of new
technology as commercial farmers did over the years.

The productivity of DR&SS, especially its maize research program, has come
about through administrative, financial and scientific continuity. Before the eighties, the
salary structure rewarded experience and academic achievements. Scientists stayed
longer with DR&SS as demonstrated by only four leadership changes in the maize
breeding program from 1932 to 1988. The importance of financial rewards and prospects
of professional advancement of scientists is apparent if one considers the effects of the
emerging private sector in research and technology transfer.

The DR&SS is under severe stress in the 19908. Financial support has fallen in
real terms and the salary structure is no longer attractive for experienced scientists to
plan on devoting their career to research in the DR&SS. The renewal of DR&SS will
have to include the arresting of the erosion of its human capital base, provision of
adequate funding, and the development of strong grass roots support for agricultural
research and extension.

Private research firms have an important role to play in agricultural
modernization in Zimbabwe. The private sector started conducting maize research in
the seventies after DR&SS had successfully developed embodied technologies“ which

could be produced and/or distributed in ways that yield profits to private firms. The

 

‘4 Embodied technology refers to that technology which the adopter can prevent a
non-adopter from using without paying for it. Examples include chemical and mechanical
inputs that research develops.

109

 

DR&SS should deal with the competition for financial and human resources and
redefine its focus of research work. Specifically DR&SS while not abandoning maize
research outrightly should concentrate its work in the marginal areas were private firms
are reluctant to venture. International centers such as CIMMYT provide not only
manpower support but also increase the gerrnplasm pool with which DR&SS work with.
The country can benefit from the competition between DR&SS and the private firms
and international research centers rather than lose from the duplication of effort that
results. However, DR&SS have to compete for human and financial resources with
these organizations.

DR&SS can compete more effectively for human and financial resources with the
private firms if it sheds its civil service status and become an independent body. The
advantages of reorganization have been identified as sustainable financing through cost
recovery and the involvement of the private sector, flexibility in the recruiting, rewarding
and retaining of gifted scientists, and the ease of getting clientele farmers’ inputs. Thus
DR&SS will be able to design competitive salary structures as well as find other ways of
securing funds for research. The danger of powerful interest groups dominating the
research priority setting and planning activities is a real one and stands as the major

disadvantage of an independent DR&SS.

110

REFERENCES

Abidogun, A. (1982) Cocoa Research in Nigeria: An Ex-Post Investment Analysis. The
Nigerian J. of Econ. and Soc. Studies. 21-35.

Agricultural Finance Corporation (1990) Bi-annual Statistical Digest. Harare.

Agricultural Research Council, (1979) Annual Report for the Period July 1978 to June
1979. Government Printers, Harare.

Agricultural Research Council, (1980) Annual Report for the Period July 1979 to June
1980. Government Printers, Harare.

Agricultural Research Council, (1981) Annual Report for the Period July 1980 to June
1981. Government Printers, Harare.

Agricultural Research Council, (1982) Annual Report for the Period July 1981 to June
1982. Government Printers, Harare.

Ahmed, M.M. and J. H. Sanders (1991) the impact of Hageen Dura I in the Gezira
Scheme, Sudan. Paper presented at the Technology Assessment Workshop,

Department of Agricultural Economics, Michigan State University, East Lansing,
Michigan.

Akino, M. and Y. Hayami (1975) Efficiency and Equity in Public Research: Rice
Breeding in Japan’s Economic Development. Amer. J. Agr. Econ, 57(1), 1-10.

Alston, J.M.; G.W. Edwards, and J.W. Freebairn, (1988) Market Distortions and
Benefits from Research. Amer. J. Agr. Econ, 70:2, 281-288.

Araji, A.A.; R.J. Sim and R.L. Gardner (1978) Returns to agricultural research and
extension programs: An ex ante approach. Amer. J. Agr. Econ. 60; 964-8.

Avila, M; Whingwiri, BE. and B.G. Mombeshora (1989) Organization and management
of on-farm research in the Department of Research and Specialist Services,
Ministry of Lands, Agriculture and Rural Resettlement: OFCOR-Case Study,
Number 5, ISNAR, The Hague.

Ayer, H.W. and GE. Schuh (1972) Social Rates of Return and Other Aspects of
Agricultural Research: The Case of Cotton Research in Sao Paulo, Brazil. Amer.
J. Agr. Econ, 54:557-569.

Barlett, A.N. (1970) Costs and social benefits of agricultural research in Mexico.
Unpublished Ph.D. thesis, University of Chicago.

Billings, K]. (1985) Zimbabwe and the CGIAR Centers: A Study of Their Collaboration

in Agricultural Research. CGIAR Study Paper, No. 6 World Bank, Washington,
DC.

111

 

Binswanger, HP. (1974) The measurement of technical change biases with many factors
of production. Amer. J. Agr. Econ. 642964-76.

Blackie, MJ. (1994) Realizing the potential of smallholder agriculture. (In "Beneath the

Surface of Zimbabwe’s Agricultural Revolution: Implications for Southern Africa"
Eds Eicher, C.K. and M. Rukuni), University of Zimbabwe Press, Harare.

Blackie, M.J. (1987) The elusive peasant: Zimbabwe agricultural policy, 1965-86. (In
Food securig for Southern Africa. Edited by M. Rukuni and C. K. Eicher),
UZ/MSU Food Security Research Project, Department of Agricultural
Economics and Extension, University of Zimbabwe, Harare.

Blackie, M.J. (1984) The Agricultural Research Council of Zimbabwe: Proposals for
Chance. Report to the ARC, Causeway, Harare.

Bonnen, J.T. (1990) Agricultural Development: Transforming Human Capital,

Technology and Institutions. (In Agpicultural Development in the Third World.
Edited by Carl K. Eicher and John M. Staatz), Second Edition, pp 262-279, John
Hopkins University Press, Baltimore, Maryland.

Bratton, M. (1991) Agricultural Interest Groups in Zimbabwe: Towards Institutional
Merger? A consultancy report to the Ministry of Lands, Agriculture and Rural
Resettlement, Causeway, Harare.

Bredahl, M. and W. Peterson (1976) The productivity and allocation of research: U.S.
Agricultural Experiment Stations. Amer. J. Agr. Econ 58:684-92.

Buccola, S. and C. Sukume (1988) Optimal Grain Pricing and Storage Policy in
Controlled Agricultural Economies: Application to Zimbabwe. World
Development. Vol. 16(3), pp. 361-71.

Buccola, S. and C. Sukume (1991) Regulated-price and stock policies: Interaction effects
and welfare preferences. J. Development Economics, 35:281-305.

Byerlee, D. (1993) Private sector research, hybrid seed and the small farmer. Seminar
paper presented at Michigan State University, June 25, Department of
Agricultural Economics, East Lansing.

Central Statistical Office (1989) Statistical Yearbook. Government Printers, Harare.

Chidzero, B.T.G. (1990) Economic Policy Statement. Government Printers, Harare.

Chigaru, P.R.N., (1984) Future Directions of the Department of Research and Specialist
Services. Internal Paper, DR&SS, Ministry of Agriculture, Harare.

Chimedza, R. (1994) Rural financial markets. (In "Beneath the Surface of Zimbabwe’s

Agpicultural Revolution: Implications for Southern Africa" Eds Eicher, GK. and
M. Rukuni), University of Zimbabwe Press, Harare.

112

Cleaver, K. (1991) Agricultural technology: current issues for the international
community and the World Bank. Summary of Conference held october 21-23,
1991. Agricultural and Rural Development Department, World Bank.

Cline, PL. and Y.C. Lu (1971) Efficiency aspects of the spatial allocation of public
sector agricultural research and extension in the United States. Reg. Sci.
Perspectives, 6:1-16.

Dagg, M. (1992) Research Program Formulation. Paper presented at the International

Seminar on "Management of Agpicultural Research". May 25-27, Beijing
Agricultural University, Beijing.

Department of Research and Specialist Services, (1979) A Description of Crop Varieties
Released by the Department of Research and Specialist Services. Supplement to
Zimbabwe Agricultural Journal Vol. 76, Harare.

Department of Research and Specialist Services, (1974) Report of the Director of
Research and Specialist Services for the Year Ending 30'” September, 1984.
Government Printers, Harare.

Duncan, R.C. ( 1972) Evaluating returns to research in pasture improvement. Aust. J.
Agr. Econ. 16: 153-168.

Donovan, PA, (1969) The Objectives and Organization of Agricultural Research. Paper
presented at the Conference on ”Research and the Farmer". September, 18-19,
1969; Harare.

Easter, K.W. and G.W. Norton, (1977) Potential returns from increased research budget
for the Land grant Universities. Agr. Econ. Res. 29:127-33.

Echeverria, R.G. (1991) Impact of Research and Seed Trade on Maize Productivity. (In

Agpicultural Research Polig: Intemation_al Qflntitgtive Perspectives; Eds,
Pardey, P.G.; Roseboom, J. and Anderson, J.R.) Cambridge University Press:

Cambridge, England.

 

Echeverria, R. G. (1990) Returns to investments in the generation and transfer of rice
technology in Uruguay. ISNAR Working Paper, No. 30, The Hague, ISNAR.

Echeverria, R.G., G. Ferreira, and M.Dabezies (1988) Retornos a la Inversion en
Generacion y Transferencia de Tecnologia Agropecuaria en Uruguay: el Caso del
Arroz. Montevideo: CIAAB.

Edwards, G.W. and J.W. Freebairn, (1984) The Gains From Research into Tradable
Commodities. Amer. J. Agr. Econ. 71 43-49.

Eicher, CK. (1989) Sustainable Institutions for African Agricultural Development.
Working Paper No. 19, ISNAR, The Hague.

113

Eicher, CK. (1991) Building African Scientific Capacity for Agricultural Development.
Agric. Econ, 4:2; 117-143.

Eicher, CK (1992) Revitalizing the CGIAR System and NARS in the Third World. Staff
paper No. 92-73, Department of Agricultural Economics, East Lansing, Michigan.

Eicher, GK. and M. Rukuni (1994) Zimbabwe’s Agricultural Revolution: Lessons for
Southern Africa. (In "Beneath the Surface of Zimbabwe’s Aggicultural
Revolution” Eds: Rukuni, M. and CK. Eicher), University of Zimbabwe Press,
Harare.

Elliott, M., (1991) An Economic Analysis of the Market for Carbohydrates in South
Africa With Specific Reference to White Maize. Working Paper, Maize Board,
Pretoria.

Elliott, H. and S. Kang, (1985) Analyzing Conditions of Service for Agricultural
Researchers: An Experiment Using Earning Functions. Working Paper No. 3,
ISNAR, The Hague.

Evenson, RE. (1993) Personal Communications at the AAEA 1993 Annual Meetings.

Evenson, RE. (1991) Notes on the Measurement of the Economic Consequences of
Agricultural Research Investments. Paper presented at the Conference on
”International Agpicultural Research", May 17-19, 1991, Cornell University,
Ithaca, New York.

Evenson, RE. (1989) Technological Opportunities and International Technology
Transfer in Agriculture. Center Paper, No. 420. Economic Growth Center, Yale
University, New Haven, Connecticut.

Evenson, RE. (1987) The International Agricultural Research Centers: Their Impact on
Spending for National Agricultural Research and Extension. CGIAR,
Washington, DC.

Evenson, RE. (1967) The contribution of Agricultural research to production. J. of Farm
Econ. 49:1415-1425.

Evenson, RE. and HP. Binswanger (1978) Technology transfer and research resource

allocation. (In Induced Innovation, Technology, Institutions and Development.
Binswangwer, H.P. e_t a_l Editors), John Hopkins University Press, Baltimore,

Maryland.

Evenson, RE. and P. Flores-Moya (1979) Economic Consequences of New Rice
Technology in Asia. IRR] Working Paper, Los Banos, Laguna, Philippines.

Evenson, RE. and Y. Kislev (1975) Agricultural Research and Productivity. New Haven,
Yale University Press, Connecticut.

114

Evenson, RE. and Y. Kislev (1973) Research productivity in wheat and maize. J. Polit.
Econ. 81:1309-29.

Evenson, RE, RE. Waggoner and V.W. Ruttan (1979) Economic Benefits From
Research: An Example From Agriculture. Science 205:1101-1107.

The Farmer, (1988) Research Dominates Debate:- Zimbabwe Cereals Producers
Association Chairman’s report. The Farmer including Tobacco News July 21,
Harare.

 

Fenner, R.J. (1990a) Budgetary Constraints and Their Impact on On-Farm Projects: Do
we Have any Options? (In "Proceedings, Workshop to assess the performance of
the Committee for On-Farm research and Extension”. Eds. Shumba, EM. and
ST. Waddington), May, 7-9, Kadoma, Zimbabwe.

Fenner, R.J. (1990b) Research Funding, Present and Future. (In Proceedings of regional

workshop on "Agpicultural research and extension and their interaction"), Nairobi,
December 2-9, 1990.

Fenner, R]. (1992) Personal Communications. (R.J. Fenner was the director of DR&SS
in 1992).

Fenner, RJ. and EM. Shumba, (1989) Linking Research and Extension Through On-
Farm Research and Demonstrations: The Zimbabwe Experience. Paper presented
at the ISNAR international workshop on "Making the link between agrjcultural
research and technology users". November, 19-25; The Hague.

Fishel, W.L. (1971) The Minnesota agricultural research resource allocation information

system and experiment. (In Resource allocation in aggicultural research. Fishel,
W.L. Ed.), University of Minnesota Press, Minneapolis.

Fishelson, G. (1971) Returns to human and research capital and the non-south
agricultural sector of the United States, 1949-1964. Amer. J. Agr. Econ. 53:129-
131.

Flores-Moya, P.; R.E. Evenson and Y. Hayami, (1978) Social returns to rice research in
the Philippines: Domestic benefits and foreign spillover. Econ. Dev. and Cul.
Change, 26:591-607.

Goldsmith, AA. (1991) Institutional Development for Agricultural Research: Concepts,
Models and Measures. Staff notes No. 91-115, ISNAR, The Hague.

Grant, PM. (1981) The Fertilization of Sandy Soils in Peasant Agriculture. Zimbabwe
Agric. J. 78:169-175.

Griliches, Z. (1958) Research Costs and Social Returns: Hybrid Corn and Related
Innovations.J. of Pol. Econ. Vol. 66; 419431.

115

Griliches, Z. (1964) Research Expenditures, Education, and the Aggregate Agricultural
Production Function. Amer: Econ. Rev. 54:961-974.

Hayami, Y. and R.W. Herdt (1977) Market Price Effects of Technological Change on
Income Distribution in Semi-Subsistence Agriculture. Amer. J. Agr. Econ,
59:245-56.

Hayami, Y. and V. Ruttan, (1985) Agricultural development: An international
perspective. Second Edition, John Hopkins University Press, Baltimore,
Maryland.

Herbst, J. (1990) State Politics in Zimbabwe; Perspectives on Southern Africa, 45:
University of California Press, Berkeley, California.

Hussein, J. (1987) Agroclimatological analysis of growing season in Natural Regions III,

IV and V of Zimbabwe. (In Cropping in the semi-arid areas of Zimbabwe. Vol. I
Edited by H.R.R. Metelerkamp and

ISNAR, (1988) Report to the Government of Zimbabwe; A Review of the Department of
Research and Specialist Services, Zimbabwe. ISNAR R32e, The Hague.

Jaffe, S.; D. Umali and L. Schwartz, (1991) Private Sector Provision of Agricultural
Services: Opportunities and Limitations. Working Paper, Agricultural Policies
Division, Agricultural and Rural Development Department, World Bank,
Washington DC.

Jayne, T.S.; M. Chisvo; and M. Rukuni (1994) Zimbabwe’s food insecurity paradox:
Hunger amid abundance. (In "Beneath the Surface of Zimbabwe’s Agpicultural
Revolution: Implications for Southern Africa” Eds Eicher, GK. and M. Rukuni),
University of Zimbabwe Press, Harare.

Jayne, TS; and M. Chisvo, (1991) Unravelling Zimbabwe’s Food Insecurity Paradox.
Food Policy. Vol 16, (5).

Kahlon, A.S.; H.K. Bal; P.N. Saxena and D. Jha (1977) Returns to investments in

research in India. (In Resource Allocation and Productivity in National and
international Agpicultural Research. T. M. Arndt, D.G. Dalrymple and V.W.

Ruttan; Editors), University of Minnesota Press, Minneapolis.

Karanja, DD. (1990) The rate of return to maize research in Kenya: 1955-88.
Unpublished M.S. thesis, Department of Agricultural Economics, Michigan State
University, East Lansing, Michigan.

Kupfuma, B. (1992) Fertilizer Use in the Semi-arid Areas of Zimbabwe. Proceedings of
the "First Poligr Analysis Initiative Conference, March 16-20, Harare.

Kupfuma, B.; A. Nyengerai; R.M. Mupawose and H. Ncube (1992) Marketing and
Contribution of Field Crops to the National Economy. (In "Small-Scale

Aggiculture in Zimbabwe: Book One:- Farming Systems, Polig and

 

116

Infrastructural Development" Eds. Whingwiri, E.E.; M. Rukuni; K. Mashingaidze
and CM. Matanyaire) Rockwood Publishers, Harare.

Kupfuma, B.; S. Machado; S.T.O. Musendo and NA. Mashiringwani, (1990) Summer
Wheat in Zimbabwe: Is it Economically Viable? Proceedings of the "Second
Smposium on Science and Technology", Research Council of Zimbabwe, Harare.

Lindner, R.K. and F.G. Jarrett (1978) Supply Shifts and the Size of the Research
Benefits. Amer. J. Agr. Econ, 60:48-58.

Lindner, R.K. and F.G. Jarrett, (1980) Supply Shifts and the Size of the Research
Benefits: A Reply. Amer: J. Agr. Econ, 62:841-44.

Lipton, M. (1988) The Place of Agricultural Research in the Development of Sub-
Saharan Africa. World development, Vol. 16, No 10, pp. 1231-1257.

Lowveld Research Stations (1990) Personal Communications with Research Officers.

Machida, L. (1992) Personal Communications (Mr Machida was a maize breeder at
Harare Research Station in 1992)

MacMillan, J.; G.D. Mudirnu; J. MacRoberts; L. Rugube; E. Guveya; L. Mutemeri; and
K. Chakanyuka, (1991) Ex ante B/C Analysis of Small Farm Maize Research and
Demonstrations, Zimbabwe. Working Paper, Department of Agricultural
Economics and Extension, University of Zimbabwe, Harare.

Mahlstede, J. P. (1971) Long range planning at the Iowa Agricultural and home
Economics Experiment Station. (In Resource allocation in agrjcultural research.
Fishel, W.L. Ed.), University of Minnesota Press, Minneapolis.

Makau, BR (1984) Measurement of economic returns to wheat research in Kenya.
Unpublished M.A. thesis, University of Nairobi, Kenya.

Mashingaidze, K. (1994) Maize research and development.(ln ”Beneath the Surface of

Zimbabwe’s Agpicultural Revolution: Implications for Southern Africa" Eds
Eicher, GK. and M. Rukuni), University of Zimbabwe Press, Harare.

Mashingaidze, K. and D. F. Mataruka (1992) Maize (In "Small-Scale Aggiculture in

Zimbabwe: Book Two:- Field Crop Production" Eds. Whingwiri, E.E. ;
K.Mashingaidze and M. Rukuni.) Rockwood Publishers, Harare.

Mashirinwgani, NA. (1989) Trends in Production and Consumption of Wheat and the
Role of Variety Improvement in Zimbabwe. Zimbabwe Agr. J. (In print).

Mazzucato, V. (1991) Non-research policy effects on the rate of return of maize research

in Kenya:1955-88. Unpublished M.S. thesis, Department of Agricultural
Economics, Michigan State University, East Lansing, Michigan.

117

Mombeshora, B.G.; P.J. Kunjeku; and G. Makombe, (1989) The Economics of Fertilizer
Use for Small-Scale Farmers Under Rainfed Conditions in Zimbabwe. Paper

presented at the "FAQ/FIAC Working Party on the economics of fertilizer use".
April 10-15 1989, Rome.

Mudimu, GD. (1989) The evolution and performance of the Zimbabwe agricultural
research system; 1900-1980: A case study in institution building for agricultural
development. Working Paper AEE 3/89. Department of Agricultural Economics
and Extension, University of Zimbabwe, Harare, Zimbabwe.

Mugwira, L.M. (1984) Relative Effectiveness of Fertilizer and Communal Area Manures
as Plant Nutrient Sources. Zimbabwe Agric. J. 81:85-90.

Muir-Leresche, K and T. Takavarasha (1988) Pan-territorial and pan-seasonal pricing
for maize in Zimbabwe. (In Household and nutritional food securigy for southern
Africa. Edited by G.D. Mudirnu and RH. Bernsten) UZ/MSU Food Security
Research Project, Department of Agricultural Economics and Extension,
University of Zimbabwe, Harare.

Ndoro, R. (1989) Personal Communications (Mr Ndoro was the Chief Economist for the
National Farmers Association of Zimbabwe).

Nelson, RR. (1990) What is "commercial" and what is "public" about technology, and
what should be? (In Technology and the wealth of Nations. Edited by N.
Rosenberg; R. Landau and DC. Mowery), Stanford University Press, Stanford,
California.

Norgaard, RB. (1988) The biological control of cassava mealybug in Africa. Amer. J. Agr.
Econ. 70:(2) 366-71.

North, DC. (1990) Institutions, institutional change and economic performance.
Cambridge University Press, New York.

Norton, G.W. and J.S. Davis (1981) Evaluating returns to agricultural research: A
review. Amer. J. Agr. Econ. 60:685-699.

Oehmke, J.F. and E.W. Crawford (1992) Evaluating the impacts of selected investments
in agricultural technology in Sub-Saharan Africa. A.I.D. evaluation News, Vol.
4(4), Washington DC.

Oehmke, J.F. and IA. Sterns, (1992) Assessing Returns to Research: Implications for
Sub-Sahara Africa. MS U Staff Paper No. 92-43, Department of Agricultural
Economics, Michigan State University, East Lansing, Michigan.

Oehmke, J.F.; L. Daniels; J. Howard; M. Maredia; and R. Bernsten (1991) The impact of
agricultural research: A review of the ex-post assessment literature with
implications for Africa. MS U Staff Paper No. 92-38, Department of Agricultural
Economics, Michigan State University, East Lansing, Michigan.

118

Pardey, PG. and J. Roseboom, (1989) ISNAR Agricultural Research Indicators: A
Global Data Base on National Agricultural Research Systems. Cambridge
University Press, Cambridge, England.

Paulsen, W.L. and DR. Kaldor (1968) Evaluation and planning of research in the
experiment station. J. Farm Econ. 50:1149-62.

Pazvakavambwa, S. (1994) An Analysis of Agricultural Extension in Zimbabwe.(1n
"Beneath the Surface of Zimbabwe’s Agpicultural Revolution: Implications for
Southern Africa" Eds Eicher, OK. and M. Rukuni), University of Zimbabwe
Press, Harare.

Peterson, W.L. (1992) Is the Demand for Extension Station Personnel Declining? Review
ongric. Econ 14:1:55-64.

Peterson, W.L. (1967) Returns to poultry research in the United States. J. of Farm Econ
49:656-669.

Pineiro, M. (1986) The Development of the Private Sector in Agricultural Research:
Implications for Public Research Institutions. Proagro Paper, No. 10, ISNAR,
The Hague.

Pinstrup-Andersen, P. (1977) Decision-making on food and agricultural research policy:
the distribution of benefits from new agricultural technology among consumer
income strata. Agr. Admin. 4:13-28.

Pinstrup-Andersen, P. and D. Franklin (1977) A systems approach to agricultural
research resource allocation in developing countries. (In Resource Allocation and
Productivig in National and international Aggicultural Research. T. M. Arndt,
D.G. Dalrymple and V.W. Ruttan; Editors), University of Minnesota Press,
Minneapolis.

Pinstrup-Andersen, P.; N.R. de Londono and E. Hoover (1976) The impact of increasing
food supply on human nutrition: Implications of commodity priorities in
agricultural research and policy. Amer. J. Agr. Econ. 58:131-42.

Ramalho de Castro, J.P. an G.E. Schuh (1977) An empirical test of an economic model
for establishing research priorities: A Brazil Case Study. (In Resource Allocation

and Productivity in National and international Aggicultural Research. T. M.
Arndt, D.G. Dalrymple and V.W. Ruttan; Editors), University of Minnesota

Press, Minneapolis.

Rattray, A.G.H. (1988) Maize Breeding and Seed Production in Zimbabwe up to 1970.
Paper presented at the ”8‘” South Afri_c_an maize breeding symposium. March 15-
17, Potchefstroom, South Africa.

Rattray, A.G.H. (1969) Advances and Achievements in Crop Research. Paper presented

at the Conference on "Research and the Farmer". September, 18-19, 1969;
Harare.

119

Reid, J. (1977) Crop Production in Mashonaland and Matebeleland. Rhodesia Agric. J.
74, 94-102.

Rohrbach, DD. (1989) The Economics of Smallholder Maize Production in Zimbabwe:
Implications for Food Security. MS U International Development Paper; No. 11.

Department of Agricultural Economics, Michigan State University, East Lansing,
Michigan.

Rohrbach, DD. (1988) The Growth of Smallholder Maize Production in Zimbabwe:
Causes and Implications for Food Security. Unpublished Ph.D. dissertation,

Department of Agricultural Economics, Michigan State University, East Lansing,
Michigan.

Rose, RN. (1980) Supply Shifts and Research Benefits: Comment. American J. Agr.
Econ, 62:834-37.

Roth, M.J. (1990) Analysis of Agrarian Structure and Land Use Patterns. Background
paper: Zimbabwe agriculture sector memorandum, The World bank.

Rukuni, M. (1994) The Evolution of Zimbabwe’s Agricultural Policy: 1890 to 1990. (In

”Beneath the Surface of Zimbabwe’s Agpicultural Revolution: Implications for
Southern Africa" Eds Eicher, GK. and M. Rukuni), University of Zimbabwe

Press, Harare.

Rukuni, M. (1992a) Introduction. (In "Small-Scale Agpiculture in Zimbabwe: Book One:-

Farming Systems, Policy and Infrastructural Development" Eds. Whingwiri, E.E.;
M. Rukuni; K. Mashingaidze and CM. Matanyaire) Rockwood Publishers,

Harare.

Rukuni, M. (1992b) Personal Communications (Prof. M. Rukuni was the Dean, Faculty
of Agriculture and Chairman of The ARC in 1992).

Rukweza, T. (1992) Personal Communications. (Mr Rukweza was a Maize breeder with
DR&SS in 1992)

Russell, D.G. (1977) Resource allocation in agricultural research using socio-economic
evaluation and mathematical models. Can. J. Agr. Econ. 23:29-52.

Saunders, (1978) The Organization and Management of Agricultural Research. The
Rhodesia Science News. Vol. 12, (1).

Schmitz, A. and Seckler, D. (1970) Mechanized agriculture and social welfare: The case
of the tomato harvester. Amer. J. Agr. Econ. 52:569-577.

Schuh, GE. and H. Tollini (1979) Costs and benefits of agricultural research: State of
the arts. World Bank Staff Work Pap. No. 360, World Bank, Washington DC.

120

Schultz, T.W. (1960) A Policy to Redistribute Losses From Economic Progress. Paper
presented at the Conference on "Qbour Mobility and Population in Agpiculture".
November, 9, Iowa State University, Ames, Iowa.

Schultz, T.W. (1953) The economic organization of agriculture. McGraw-Hill Book Co.
New York, New York.

Schwartz, L.A., J .A. Sterns, and J .F. Oehmke (1992) The economic payoff to agricultural
research and extension: The case of cowpeas in Senegal. Forthcoming, Agric.
Econ

Scobie, GM. (1979) Investment in international agricultural research: Some economic
dimensions. World Bank Stafl Working Paper, No. 360, World Bank, Washington
DC.

Scobie, GM. and RT. Posada (1978) Impact of Technical Change on Income
Distribution: The Case of Rice in Colombia. American J. Agr. Econ. 60:85-92.

Shapouri, S. and M. Missiaen, (1991) Economic Performance and Policy Adjustment:
The Experience of Zimbabwe. ERS Stafi Paper, No. AGES 9127, Economic
Research Service, USDA, Washington.

Shapouri, S. and M. Missiaen, S. Rosen (1992) Food Strategies and Market

Liberalization in Africa: Case Studies of Kenya , Tanzania and Zimbabwe. ERS
Staff Report, No. AGES 9220, Economic Research Service, USDA, Washington.

Shumba, EM. (1992) The Farming Systems. (In "Small-Scale Aggiculture in Zimbabwe:

Book One:- Farming Systems, Poligy and Infrastructural Development” Eds.
Whingwiri, E.E.; M. Rukuni; K. Mashingaidze and CM. Matanyaire) Rockwood

Publishers, Harare.

Shumba, EM. (1990) Zimbabwe’s Experience in Agricultural Research Priority Setting
for Communal Area Households. (In "Food Security in Southern Africa” Eds.
Rukuni, M. and J .B. Wyckoff), UZ/MSU Food Security project, Harare.

Shumba, EM. (1985) On-Farm Research Priorities Resulting From a Diagnosis of the
Farming Systems in Mangwende, a High Potential Area in Zimbabwe. Zimbabwe
Agric. J. Special Report, N o. 5.

Shumway, CR. and RJ. McCracken (1975) Use of Scoring Models in evaluating
research programs. Amer. J. Agr. Econ. 57:714-718.

Stack, J.L. (1994) The Peasant Maize Revolution and its Distributional

Consequences.(ln "Beneath the Surface of Zimbabwe’s Aggicultural Revolution:
Implications for Southern Africa" Eds Eicher, GK. and M. Rukuni), University of

Zimbabwe Press, Harare.

Tattesfields, J .R. and Havazvidi (1994) The Evolution of the Seed Industry in Zimbabwe.
(In ”Beneath the Surface of Zimbabwe’s Agpicultural Revolution: Implications for

121

 

 

Southern Africa" Eds Eicher, GK. and M. Rukuni), University of Zimbabwe
Press, Harare.

Tattersfields, JR. (1992) Personal Communications. (Mr Tattersfields was the Head of
Research at the Seed Co-op’s Rattray-Arnold Research Station)

Tattersfields, J .R. (1982) The Role of Research in Increasing Food Crop Potential in
Zimbabwe. Zimbabwe Science News, Vol. 16:6-10.

Tawonezvi, H.R.P. (1994) Evolution of Agricultural Research. (In "Beneath the surface

of Zimbabwe’s agpicultural revolution: Implications for Southern Africa” Eds
Eicher, GK. and M. Rukuni), University of Zimbabwe Press, Harare.

Traxler, G. and D. Byerlee (1991) The economic returns to crop management research
in a post- -Green Revolution setting. CIMMYT Economics Working Paper 91/7.
Mexico, D. F. CIMMYT.

Trigo, E. (1985) Working to Develop Support at the Political Level for National
Agricultural Research: The Case of the Dominican Republic. Country Progress
Report, No. R14/P1, ISNAR, The Hague.

Tweeten, L. G. and F. K. Hines (1965) Contribution of agricultural productivity to
national economic growth. Agr. Sci. Rev., 3:40-45.

Van Zyl, J. (1986) A Statistical Analysis of the Demand for Maize in South Africa.
Agrekon, Vol. 25, (3).

Weinmann, H. (1972) Agricultural Research and Development in Southern Rhodesia,

1890-1923. Occasional Paper, No.4, Department of Agriculture, University of
Rhodesia, Salisbury.

Weinmann, H. (1975) Agricultural Research and Development in Southern Rhodesia,
1924-1950. Series in Science, No.2, Department of Agriculture, University of
Rhodesia, Salisbury.

White, RC. and J. Havlicek, (1978) Rates of return to agricultural research and
extension in the Southern Region. Dept. Agr. Econ. Stafi Paper, 79-2, Viginia
Polytechnic Institute and State University, Viginia.

World Bank, (1990) World Development Report. Washington DC, World Bank.

122

ANNEX A: THE DATA

Table a.1: Total maize area and yields in Zimbabwe: 1945-1990.

Year Year

 

Note: Area stands for area under maize in thouSand of hectares, and Yld
stands for yield in tonnes per hectare.

123

Table a.2: Hybrid maize area and yields in Zimbabwe: 1949-1990.

Year Year

 

Note: Area stands for area under maize in thousand of hectares, and Yld
stands for yield in tonnes per hectare.

124

Table 3.3: Non-hybrid maize area and yields in Zimbabwe: 1949-1990.

Year Year

 

Note; Area stands for area under' maize in thousand of
hectares, and Yld stands for yield in tonnes per hectare.

125

Table a.4: Maize produced and marketed through GMB in
Zimbabwe: 1949-1990

Year Year

 

Note: Prod. represents amount of maize produced in thousand metric
tonnes while mkt represents the amount of maize delivered to the GMB.

126

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table I936 Research and extension costs and other financial data: 1931-
Year Budgets Maize CPI Year Budgets Maize CPI
R&SS EXTE tripe 1970 R&SS EXTE Brice 1970=
1931 422 na ”hi “it? 19611220 1202 134m “192
1932 352 na na 38 ".1962 1326 1258 32.36 83
1933 242 na na 37 a.1?963, 1422 1298 34.75 84
1934 298 na na 37 ”1964 1390 1358 38.39 87
1935 326 na na 37 1771965 1416 1320 37.18 88
1936 380 na na 36 "1966 1518 1361 32.27 90
1937 524 na na 38 1967 1634 1490 32.64 92
1938 524 na na 38 1968 1816 1547 34.76 94
1939 44 na na 41 "2.1969" 2082 1688 33.55 97
1940 438 na na 42 1970 2290 1856 39.16 100
1941 420 na na 44 .1971 2569 1860 34.32 103
1942 402 na na 46 51972 2860 2068 29.59 108
1943 411 na na 49 1973" 3276 2198 39.71 112
1944 403 na na 50 1974 3445 2317 39.32 120
1945 400 na 14.65 53 1975 3995 2522 41.61 129
1946 394 na 17.95 54 1976 4231 2819 47.34 139
1947 389 na 25.30 55 :1977 4629 3119 51.70 148
1948 384 68 26.05 56 ”1978 4775 3089 51.99 158
1949 544 188 30.75 57 1979’; 5478 3509 60.32 175
1950 564 464 38.45 59 {$19803 6052 3671 89.04 193
1951 628 556 41.20 60 2351981 8074 4270 119.07 219
1952 618 648 50.00 63 1982' 8411 4481 119.63 258
1953 898 830 45.60 65 2.511983 8246 14705 119.59 295
1954 1212 966 44.50 67 1984 8978 16965 152.07 332
1955 1050 770 44.50 67 11985? 10852 19988 178.91 369
1956 1503 674 45.60 71 “198612513 23406 177.86 422
1957 10761190 43.95 73 321987, 15040 30345 179.35 471
1958 10961164 45.60 75 2198815602 31824 193.50 503
1959 10761080 38.80 77 “1989 16940 36525 213.50 553
1960 11301094 39.60 79 1990:: 19554 42417 283.50 633

 

Note: R&SS represents the Department of Research and Specialist Services
and EXTE represents Departments of extension services such as Agritex
after 1981 and CONEX, DEVAG, and Ministry of Internal Affairs before
1981.

127

flNEX B: THE RESULTS

Table b.1: The incidence of benefits, 1949-1990. ('000 Z$)

 

Note; The total maize research extension benefits are
denoted GBen which is equal to TB in equation 3.2 is
obtained from the following equation;

(’7 + 1)2k:] (11.1)

=59:qu +-(—.—.-)-

The research and extension benefits captured by consumers
are donated CBen and are calculated from the following
equation;

128

 

 

1 1 c
CBen! = PuQx/(rn " 5k!“ + CHE Z 77 '1 T c : 12]] (ha)

The benefits captured by producers are denoted PBen and are
obtained by subtracting the CBen from GBen.

129

Table b.2: Undiscounted net maize research and extension
benefits, 1941-1990. ('000 Z$)

 

Note: NB@HCL, NB@MCL and NB@LCL refer to the net benefit
after subtracting the high cost level, medium cost level and
low cost level, respectively, from the gross benefit.

130

Table b.3: Undiscounted annual
income by sub-sector due
extension, 1949-90. ('000 Z$)

131

change in

to

maize

farmers'
research

 

cash
and

 

I

W

1111“”

i

I

v.14.

 

HI

| l

l

HICHIGQN STQTE UNIV

 

1:33,}... .. ......
. .....
.. .

, ....1 ....
:5 :3...