CAUSES AND IMPACTS. OF TURNOVER 0F RESEAC’H.
SCIENTISTS WITHIN THE SENEGALESE ;
AGRTCUITIIRAI RESEARCH INSTITUTE .

1981-1987

‘ BY
PAUL DAVID TITSSSII , '

CAUSES AND IMPACTS OF TURNOVER OF RESEARCH
SCIENTISTS WITHIN THE SENEGALESE
AGRICULTURAL RESEARCH INSTITUTE:

1981 - 1987

By

Paul David Wessen

Plan B Paper

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Agricultural Economics

1988

12W

1.1) Introduction

1.2) Brief Overview of Senegalese Agriculture
And Agricultural Policies

1.2.1) An Overview of Senegalese Agriculture
1.2.2) Agricultural Policy in Senegal

1.3) The Role of Agricultural Research in
Economic Development

1.3.1) Agriculture in Economic Development

1.3.2) The Role Agricultural Research In
Agricultural Development: The Induced
Innovation Model

1.4) Trends in African Agricultural Research

QflA£I§§_Z: AGRICULTURAL RESEARCH ORGANIZATION AND

POLICIES AFFECTING RESEARCH STAFF
IN SENEGAL

2.1) Historical Overview

N

.1.1) The Colonial Era
2.1.2) Post Independence (1960 - 1975)

2.2) Research Planning and Development in
Senegal (1975 - 1987)

.2.1) The Establishment and Evolution of ISRA
.2.2) Legal Status of ISRA
.2.3) Research Planning at ISRA

NNN

2.3) Organizational Structure for Research
at ISRA

2.4) Profile of ISRA Personnel

111

QHAEI§E_1: INTRODUCTION: THE ROLE OF AGRICULTURAL RESEARCH
IN THE DEVELOPMENT OF SENEGALESE AGRICULTURE

page 1

12

19

21

21
21
23
24
24
27
28
3O

33

h)h)h>h)h3h>h)
bbbbbbb
p~
V

.1) ISRA Personnel

.2) Educational Background of Researchers

.3) Qualifications of Research Support Staff
Experience Profile of National Researchers
.5) Age Profile of Senegalese Researchers

.6) Reorganization of Personnel-- 1987

.7) Recruitment and Training

QEAIIER_§: FACTORS AFFECTING RESEARCHER PRODUCTIVITY,

U)

www 09

WW

JOB SATISFACTION, AND TURNOVER

The Process of Agricultural Research

.1.1) Research Continuity
.1.2) Types of Agricultural Research
.1.3) Building Research Staff Capacity

Factors Affecting Researcher Performance
And Job Satisfaction

.2.1) Research Funding and the Role of Donors

The Planning and Management of Research

Turnover in Research Staff

.4.1) Types of Turnover
.4.2) Causes of Research Staff Turnover

The Incentive Stucture at ISRA

.5.1) Professional Rewards
.5.2) ISRA Salary Structure
.5.3) Work Environment

The Impacts of TurnoVer

.6.1) Erosion of Research Capacity
.6.2) Lost Returns to Investment

A Conceptual Framework for Measuring
The Impacts of Turnover

iv

33
35
37
37
38
4O
41

43

43
43
45
47
SO
50
53

57

57
59

61
61
63
67
69
69
7O

72

QHAEIEB_&: RESEARCH STAFF TURNOVER AT THE SENEGALESE
NATIONAL AGRICULTURAL RESEARCH INSTITUTE
(1981 - 1987): AN ANALYSIS OF THE PROCESS,
CAUSES, AND IMPACTS

4.1)
4.2)

4.3)

4-‘4-‘9

91>

4.4)

99“

4.5)

4-‘4-‘4‘
U1U'IUI

Plan of Analysis

The Data

Analysis

Calculation of Turnover Rates

Overall Turnover Rates (1981 - 1987)
Yearly Turnover Rates Based on
Nationality of Researchers

Turnover Rates By Research Department
Research Staff Instability: Internal
Transfers, Promotion into Administration
And Turnover Combined

Characteristics of Researchers Who Left ISRA

Status of Researchers Who Left
Educational Level of Researchers

Who Left

Turnover By Length of Service at ISRA
Experience and Status

Turnover By Specialization of
Researchers

Turnover By Research Program

Salary Structure

Conditions of Service

Results of the Earnings Function
Analysis .

Real Earnings of ISRA Researchers
Between 1981 and 1987

ISRA Salaries Relative to Alternative
Research Organizations

Salaries at International and Regional
Organizations

Government Salaries

Present Employment of Leavers

A Comparison of ISRA's Actual Income
Profile with an Ideal Income Profile

75

75

81

83

83
85

88
91
93
97
97
102
105
106

108
110

111
111
115
120
123
124
125
128

131

4.6) Conclusions about Turnover in Research
Staff at ISRA

4.6.1) Causes of Turnover at ISRA
4.6.2) The Impacts of Turnover

QBAEIEB_§: CONCLUSIONS AND RECOMMENDATIONS

5.1) Personnel Turnover

5.1.1) General Conclusions
5.1.2) Recommendations

5.2) Recommendations for Future Study

vi

134

134
138

141

141

141
141

151

153

Table 2.

2.

.10:

W

Research Personnel at ISRA (1974 - 1987)

Educational Profile of Senegalese
Researchers at ISRA in 1987

Educational Characteristics of Senegalese
Researchers in Comparison to other
African Countries

Experience Profile of National Researchers
At ISRA: 1983 versus 1987

Turnover Rates for all ISRA Researchers

Yearly Turnover Rates amongst Expatriate
And Senegalese Researchers at ISRA
(1981-87)

Yearly Turnover Rates in ISRA Research
Departments

Instability in Senegalese Research Staff
At ISRA (1981 - 1987)

Characteristics of Leavers: Research
Status

Education of Researchers Who Left ISRA
Between 1981 and 1987

Proportion of Researchers Who Left
Versus Proportion of Researchers in
The Population: By Degree Group

Turnover Rates for Senegalese Researchers:
Experience of Leavers

Relationships between Research Experience
And Research Status at ISRA (1981 - 1987)

Turnover in Senegalese Research Staff
(1981 - 1987): Specialization of
Researcher

vii

34

35

36

38

86

88

92

95

99

103

104

105

107

109

4.11: Results of Earnings Function Analysis:
Determinants of Researchers' Earnings
at ISRA (1987) 116

viii

Figure 1.

W

Interrelationships between the Four
Elements in the Induced Innovation
Model

Ages of Senegalese Researchers at
ISRA in 1987

A Conceptual Framework for Measuring
Losses to Investment Resulting from

Turnover in Research Staff

Yearly Turnover Rates for all ISRA
Researchers (1981 - 1987)

Yearly Turnover Rates: Turnover Rates
By Nationality

Yearly Turnover Rates: Turnover Rates
By Research Department (1981 - 1987)

Declining Real Salaries at ISRA:
Two Examples

Salary Scales: ISRA and Competing
Organizations

A Comparison of ISRA's Income Profile
With an Ideal Income Profile

ix

16

39

73

87

9O

94

121

126

133

mu: INTRODUCTION: THE ROLE OF AGRICULTURAL RESEARCH IN
THE DEVELOPMENT OF SENEGALESE AGRICULTURE

1.1) W

Turnover of research scientists is a commonly recognized
constraint on research performance. However, there have been few
detailed empirical analyses of the nature of the turnover process
within national agricultural research systems in developing
countries. This study analyzes turnover of research scientists
within the Senegalese Agricultural Research Institute (ISRA)
between 1981 and 1987. Particular emphasis is given to the causes
of turnover and the impacts of turnover on research performance.

The goal of agricultural research is to develop new or
improved technologies, knowledge or techniques which yield
sustained increases in agricultural output. As such, agricultural
research is of critical importance in reversing the trend of
declining per capita food production in Africa (Falcon et a1.,
1984). Building institutional capacity to perform research
effectively 9 is of central importance in achieving long-term
development goals.

Researchers are the most important input in conducting
agricultural research. The development of a staff of researchers
of sufficient size, qualifications, and experience is a long-term
and expensive process. The ability of a national agricultural
research system to produce meaningful research results is in

1

large part dependent upon its ability to retain and motivate its
scientific staff. High rates of turnover of research scientists
is indicative of weaknesses in a national agricultural research
institute's ability to manage its scientific personnel.

Several observers at ISRA have noticed high numbers of
departures of researchers in recent years. Data available from
personnel files between 1981 and 1987 permits the preliminary
analysis of the nature of the turnover process at ISRA.

In this paper, thntg;_1 contains background information on
Senegalese agriculture, and describes the role of agriculture and
agricultural research in economic development. §h§p§g1_2 contains
background information on the history, structure, and policies of
ISRA” thp;g;__1 presents the theoretical framework 'used to
analyze turnover in research staff. thp;g1__4 contains the

analysis of turnover in research staff at ISRA between 1981 and

1987, and gh§p§g1_§ contains recommendations.

1.2) WWW
£91.19!

1.2.1) w a e u

The World Bank (1987a) estimates that 65 per cent of the
Senegalese population lives in rural areas. In 1980, 81 per cent
of the Senegalese labor force was either directly or indirectly
employed in agriculture. The share of agriculture in Senegal's

Gross Domestic Product declined from 25 to 19 per cent between

1965 and 1985. However, given frequent droughts these figures
often fluctuate. As in most of Africa, Senegalese agricultural
production has failed to keep pace with rapid population growth.
Per capita food production declined by 1.1 per cent a year
between 1960 and 1985.

Cereal imports to Senegal (primarily wheat and rice) rose
from 341,000 metric tons in 1974 to 510,000 metric tons in 1984.
Over the same period, food aid in cereals rose from 27,000 metric
tons to 130,000 metric tons. The low productivity of Senegalese
agriculture has contributed to problems of widespread
malnutrition, rural poverty, high rates of rural-to-urban
migration, and urban unemployment.

Due to limited natural resources, Senegal is heavily
dependent upon its agricultural resources. Peanut products, fish
products, and to a lesser extent, cotton are important Senegalese
exports. Peanut exports declined steadily between 1977 and 1982,
from 46 to 27 per cent of all exports. Over the same period
exports of fish products have increased from 8.5 to 25.4 per cent
of all exports. Since 1978, agricultural imports have far
exceeded agricultural exports. As a result Senegal has a large
trade deficit. In 1973, foreign debt represented 16.4 per cent
of Senegal's GDP. By 1984, this figure had risen to 75 per cent
of GDP.

The productivity and the diversity of regional agriculture

in Senegal is largely determined by rainfall. The average yearly
amount of rainfall increases progressively as one moves from
north to south. Agricultural production in the northern regions

of Senegal is severely limited by erratic rainfall which averages
only 300 to 500 mm. per year during a one and a half month rainy
season. .Agricultural production. is concentrated in southern
Senegal, where yearly rainfall averages from 700 to over 1000 mm.
during a four month rainy season. Average yearly rainfall has
been declining throughout Senegal during the past decade.
Frequent droughts and very erratic rainfall often cause dramatic
fluctuations in agricultural production. Soils in Senegal are
characteristically marginal or fragile.

Since low and erratic rainfall adversely affects the yields
and diversity of crops in the relatively sparsely populated
northern sections of Senegal, livestock raising is an important
activity. Desertification of lands, and food shortages during the
dry season are common problems. Droughts are very commonplace in
the North and often wipe out entire harvests. Sixty per cent of
the Senegalese population lives in the central and southern
portions of Senegal, where sufficient rainfall permits relatively
diverse agricultural production. In jyears with good rainfall
these regions produce small surpluses.

Dryland farming accounts for 90 per cent of total

agricultural production. Approximately 55 per cent of all farms

are l to 5 hectares in size and 38 per cent of all farms are 6 to
10 hectares in size. Relying, predominantly on family labor,
farmers typically grow cereal crops for home consumption and cash
crops such as cotton and peanuts for income. Millet, sorghum,
and groundnuts comprise 90 per cent of cultivated land, and 75
per cent of the value of agricultural production. Other important
crops include rice, maize, cotton and cowpeas.

Improving the standards of living for the rural and urban
poor, and generating sustained economic growth in Senegal hinges
upon increasing agricultural productivity and output. The
development of new agricultural technologies, and farming and
management techniques are crucial if increases in the
productivity of land and labor are to occur in Senegal. The
major challenges facing Senegalese agricultural researth are to
ease the rainfall constraint on agriculture, and to both increase
and stabilize crop yields. More specifically, important research
needs in Senegal include the development of crop varieties with
shorter growing cycles, higher yields,-and/or increased drought
and pest resistance, and improved soil and forest conservation

techniques.

1.2.2) Ag:1ggl£u£§l_£21121—13—5292331

Between Senegalese independence in 1960 until 1984,

Senegalese agricultural policy was defined by the Agricultural

Program (W M). The policy's objectives were to:

organize agricultural production and marketing, increase
agricultural growth, diversify production, and to supply
agricultural inputs. A number of parastatal organizations were
established to accomplish this goal, including ISRA. Regional
development agencies were established in each agricultural
region. Farmers were organized into cooperatives. Parastatal
marketing organizations maintained fixed prices for agricultural
products, despite wide fluctuations in output. The National
Bureau for Cooperation and Development Assistance (ONCAD) was
established as part of a large-scale program to provide producers
credit with which to purchase agricultural inputs.

The impact of the Agricultural Program was dubious. The
policy was overly ambitious and was plagued by cost overruns
which contributed to the government's budget deficit and exterior
debt. Agricultural inputs were often supplied to producers in
untimely fashion. Some observers believe the Agricultural Program
made little contribution to increasing agricultural growth or
productivity in Senegal (see Faye and Bingen, 1988).

The Agricultural Program was halted in 1979 and replaced in
1984 by the New Agricultural Policy (Nouvelle Politigue
Agr_i_9_o_1_e). The objectives of this policy were to reduce the state
role in agriculture, which started with the dismantling of ONCAD

in 1980, and to liberalize marketing policies for agricultural

products. Included in the New Agricultural Policy were the
following objectives:
1) To achieve food security through agricultural growth and
diversification, intensifying and expanding rain-fed. and
irrigated crops, and redirecting food consumption patterns
towards locally produced products.
2) To raise rural standards of living by paying
remunerative prices for agricultural products, increasing
agricultural productivity, and improving the organization of
product markets and rural institutions.

3) To reduce the role of the state in production activities
and let farmers manage their own activities.

4) To liberalize and. privatize agricultural input and
product markets. The government's role would be limited to
market interventions in times of poor harvest, and setting
floor prices for products.

5) limiting the role of governmental regional development
institutions to training and extension functions.

6) To conserve and rehabilitate the natural environment.

7) To reduce the foreign trade deficit through import

substitution strategies, and improved marketing of

Senegalese exports, with an emphasis on peanuts.

In 1986, the New Agricultural Policy was expanded to include
the Cereals Plan. In this plan, the Senegalese government set a
target of achieving 80% self-sufficiency in cereal production by
the year 2000. l/ A number of studies have shown the
governmental. target to be unrealistic. It should be noted that

neither the New Agricultural Policy nor the Cereal plan

explicitly expressly defined a role for agricultural research

 

1 At the time that the Cereals Plan was formulated, Senegal
was only 52 per cent self-sufficient in cereals production.

within their objectives (Faye and Bingen, 1988).

1.3) Ihe Role of Agricultural Research in Economic Development

Increasing agricultural productivity and production through
the generation of new technologies which are adapted to a
country's agro-climatic and socioeconomic environment is a
critical element in economic development strategies for
developing countries. The goal of agricultural research is to
develop, promote and disseminate new knowledge, technologies, and
techniques which enhance agricultural production and
productivity. Hence, building a strong national agricultural
research system is a crucial element in a country's long-run

economic development strategy.

1.3.1) AW

In low-income developing countries, 60-80 per cent of the
population typically lives in rural areas. The agricultural
sector is generally the primary source of income and employment.
However, traditional, or technologically static, agricultural
production systems are unable to meet expanding demand for food
caused by rapid population growth and rising incomes. High food
prices are a leading cause of widespread. rural poverty and
malnutrition. Low' productivity agricultural systems thus ‘have

large negative impacts on rural peoples' livlihood, and constrain

economic development.

A model developed by Mellor (1973, 1978; Mellor and
Johnston, 1984) describes the process of agriculturally based
economic development. In the model, the key to economic
development is the transfer of resources from the agricultural to
the non-agricultural sector of an economy. This is accomplished
by creating a technologically dynamic, science-based agricultural
sector.

New' or improved. agricultural technologies and. techniques
developed through agricultural research reduce costs by either
reducing the cost or quantities of inputs used in production, or
by increasing jyields for' a. given level of input ‘use. Since
farmers who adopt new technologies can produce more food at lower
cost, they expand production. This increases the supply of food
in society, lowering food prices.

Studies show that low-income consumers in developing
countries typically spend 70 to 90 per cent of their income on
staple foods (Mellor, 1978). Thus, lower fbod prices translate
into increased real incomes for consumers. Mellor believes that
higher real incomes for small farmers and urban workers increases
the demand for labor-intensive, domestically produced
agricultural and non-agricultural goods sudh as chemical and/or
organic fertilizers, textiles, and processed foods.

Of particular importance is increased demand for the

10

products of rural non-farm industries. Chuta and Liedholm (1984)
have demonstrated that rural, small-scale industries which
typically employ five workers or less are a major source of
income in rural areas. These industries usually are either a
primary or secondary source of employment for 30-50 per cent of
the rural labor force in developing countries. Thus, increased
demand for labor-intensive agricultural and non-agricultural
products generates employment for low-income consumers in both
rural and urban areas.

Since the income elasticity of demand for food in developing
countries typically is in the range of 0.7 to 0.8 (Stevens and
Jabara, 1988), demand for food rises as per capita income rises.
However, since agricultural research institutions continue to
produce a flow of new technologies which are diffused to farmers,
the suppLy of food to society continues to increase. Increased
food supplies offset rising demand for food, resulting in stable
food prices. If food demand grows more rapidly than food supply,
the resulting rise in food prices lowers the real incomes of
rural and urban wage earners and thus curtails rising demand for
products produced in the non-agricultural sector. Employment
generation in the non-agricultural sector is slowed as resources
are transferred back into the agricultural sector. ,

The increased employment and income generated through

agricultural development improves nutritional standards.

11

Reutlinger (1985) concludes that the major cause of malnutrition
in developing countries is limited access to food supplies due to
low incomes. Thus, lower food prices improve the access or the
rural and urban poor to food markets. Higher rural incomes slows
rural-to-urban migration. Todaro (1981) has demonstrated that
rural-to-urban migration is stimulated by the perception that
urban wages are higher than rural wages, regardless of actual
opportunities for employment. Higher rural incomes alter this
perception.

Thus, agriculture can be viewed as the locomotive for growth
in the non-agricultural sector. Rising ~rural incomes and
employment increases demand for products produced in the non-
agricultural sector, permitting the net transfer of resources out
of the agricultural sector that is necessary for economic growth
in the non-agricultural sector. In. principle, the increased
public sector revenues generated through economic growth creates
a source of reliable funding for expanding agricultural research
capacity and activities. This would increase the flow of new
technologies and knowledge that increase agricultural output and
productivity. However, this is contingent on agricultural
research being made a priority in development policy.

The key to the model is its unimodal, small farmer focus.
Research policies and agricultural development strategies must be

oriented towards the large proportion of small and medium sized

12

farmers in developing countries. Also, price policy incentives
are needed that encourage farmers to adopt new technologies and
increase production of food crops.

By contrast, development strategies focusing on
industrialization often fail to generate employment opportunities
or rising incomes for the rural and urban poor. Whereas urban
employment initially rises in such schemes, a stagnant or low
productivity agricultural sector often slows economic growth. If
increased food supplies do not offset rising urban demand for
food, food prices rise. The real incomes of the urban poor
decrease. Thus, demand for locally produced products, and the
resulting expansions in employment opportunities are not
generated. If rising demand for food is offset by increasing food
imports, the resulting foreign exchange outflows might slow
growth in the non-agricultural sector.

Underlying this process of agriculturally based economic
growth is the creation of a science, or research, based
agricultural production system. Thus, agricultural research plays
a crucial role in the long-term development strategies of

developing countries.

1.3.2) WWW
eve 0 me ' u ed ovati Model

Hayami and Ruttan (1985) believe that there are multiple

paths of technological development. The common basis for

l3

achieving rapid growth in agricultural production is a country's
capacity to develop ecologically sound and economically viable
technologies that respond to local resource and cultural
endowments. Technological change involves the substitution of
relatively abundant and cheap factors of production for
relatively scarce and expensive factors of 'productionu As a
result, the constraints on agricultural production imposed by
inelastic resource supplies are relaxed.

As an example, Hayami and Ruttan contrasted the agricultural
development patterns of Japan and the United States during the
past century. Japan had a limited supply of land and a relatively
abundant supply of labor. Thus, land was expensive with respect
to labor. Sustained agricultural growth in Japan was accomplished
by increasing the productivity of land through the introduction
of high yielding, fertilizer responsive crop varieties. Labor was
substituted for land.

The 'United States 'had. a limited supply of labor and a
relatively abundant supply of land. Thus, labor was expensive
with respect to land. Sustained agricultural growth was
accomplished by increasing the productivity of labor through the
introduction of animal power and then mechanical power. Land was
substituted for labor. Thus, in both Japan and the United States
the direction of technological change reflected the factor

endowments of each country.

l4

Hayami and Ruttan conclude that technical change is induced
by changes in the relative prices of factors of production. If
these prices accurately reflect the conditions of supply and
demand for resources, technical change is guided along an
efficient path by price signals in the market. Policies which
distort prices such as export subsidies, input or product
subsidies, or fixed prices for agricultural products can induce
technical change which does not accurately reflect conditions of
supply and demand for resources in society. Thus, technological
innovation proceeds along an inefficient path.

The process of induced technological innovation operates as
follows. Faced with changing relative prices of factors of
production, farmers demand technical alternatives from public
research‘ institutions that reduce production costs. These
technical alternatives either reduce the use of expensive and
scarce factors of production, or increase the productivity of
scarce inputs. Research scientists and administrators respond by
developing and making available cost reducing technologies.

The response of research scientists and administrators to
farmers' demands for new technologies is critical to the induced
innovation model. Hayami and Ruttan believe that incentive
structures for scientists and administrators in research
institutions must be designed to reward research on problems of

social and economic significance. These rewards can be based on

15

status or material gain.

Agricultural development also involves institutional change.
Institutions, or social structures, are defined as "persistent
patterns of social interactions with associated statuses, roles,
norms of behavior, and sanctions" (Stevens and .Jabara 1988,
p.100). Economic institutions define rights to resource use and
the distribution of income streams resulting from economic
activity (Hayami and Ruttan, 1985). By nature, institutions
remain stable over extended periods of time. Institutions are
influenced by resource endowments, available technologies, and
cultural endowments.

Hayami and Ruttan (1985) believe institutional change is
induced by changes in resource endowments as signalled through
prices, and by new opportunities for productive uses of
resources.

Figure 1.1 shows the interrelationships between the four
elements included in the induced innovation. model. Cultural
endowments refer to non-economic variables in society. Religion,
ideology, and social ethics are cultural variables. The four
elements in the model interact and influence each other. Changes
in one element in the model induce changes in the other elements
of the model. Agricultural development can be accelerated or
impeded by any of the four elements in the model.

In the induced innovation model, agricultural research is of

l6

Eiggrg_1*1; Interrelationships between the Four Elements in the
Induced Innovation Model

 

 

 

 

 

 

 

 

IOIWCC /— . \ 'flm .
Endowment;

 

 

Culluul
Endeamonu

In human

 

 

 

 

 

 

 

Source: Hayami and Ruttan (1985).

 

central importance to agricultural development. Agricultural
research represents a mechanism by which to accelerate the
process of technological and institutional change 'that is
neccessary to generate long-term increases in productivity in the
agricultural sector (Stevens and Jabara, 1988).

The development of new or improved biological, chemical, and
mechanical technologies that are consistent ‘with. a. country's
resource endowments can ease the constraints imposed upon
agriculture by inelastic resource supplies. Institutional
constraints on agriculture can be identified through research on

agricultural institutions such as land tenure systems, marketing

17

institutions, credit policy, and ‘price policy. As a result,
policy recommendations and reforms may be devised which create an
institutional environment that induces long-term. increases in
productivity in the agricultural sector. Stevens and. Jabara
(1988) believe that advances in social science research lower the
costs of institutional change and that advances in agricultural
and basic sciences lower the costs of technological change.

If agricultural research systems are to generate productive
inputs for the agricultural sector, the resources allocated to
them must be effectively used. Hayami and Ruttan believe that
scientific manpower constraints severely limit the capacity of
low-income countries to shift towards a science-based
agriculture. Thus, developing scientific and technical manpower
through training and education is of crucial importance to
agricultural development.

It is not sufficient, however, to simply expand research
staff. Stevens and Jabara (1988) believe that certain conditions
for productive scientific work are essential if maximum returns
are to be obtained from the resources allocated to agricultural
research. These conditions include: stable and flexible
administration of research, the provision of adequate financial
and logistical support to research staff, adequate access of
research staff to scientific equipment and up-to-date

information, and clear definition and coordination of research

18

tasks. Attractive salaries and career structures are important in
attracting, retaining and motivating researchers.

Hayami and Ruttan (1985) and Stevens and Jabara (1988)
believe that if the induced innovation model is to work, the
development of a publicly funded research system is essential.
This is especially true of agricultural. research. since
agricultural commodities characteristically display low ‘price
elasticities of' demand” Thus, social ‘benefits resulting from
research are largely received by consumers in the form of lower
food prices. As a result, Hayami and Ruttan believe that the
general public should pay for agricultural research.

Private agricultural research generally focuses on the
development. of' mechanical and. chemical technologies. Property
rights such as patents can be created to exclude non- investors
from sharing in the returns generated by these technologies.
Thus, private firms find it profitable to develop new mechanical
and chemical technologies.

However, biological technologies generally have public good
attributes. As such, it is next to impossible to exclude others
from the stream of benefits generated by new biological
technologies. The small size of farms and limited incomes of most
farmers in developing countries prevent individual farmers from
organizing to fund research or to produce new technologies. In

the early stages of economic development, private firms are often

19

too weak to fund research.

Thus, public funded. agricultural research institutes are
crucial in producing biological technologies and in responding to
the needs of small farmers in developing countries. Hayami and
Ruttan believe that if agricultural research was left in the
hands of private firms, the path of technological change would be
seriously biased towards large farmers. Mechanical and chemical
technologies are often too expensive for the majority of small

farmers in developing countries to afford.

1.4) ends 1 fr can i u ura
Judd, Boyce and Evenson (1986) estimate that expenditures
(in constant 1980 dollars) for agricultural research in Africa
rose from $119,149,000 in 1959 to $251,572,000 in 1970. By 1980,
these expenditures totalled $424,757,000. Between 1959 and 1980,'
expenditures for agricultural research in West Africa rose from
$44,333,000 to $205,737,000. Bennell (1986) indicates that this
trend will continue. USAID alone is expected to channel over $1
billion in direct assistance to African national agricultural
research systems over the next 15 years.
Judd et a1. estimate that manpower in Africa, measured in
scientific person years, rose from 1,919 in 1959 to 8,088 in
1980. For West Africa the figures were 412 in 1959, and 2,466 in

1980. Bennell (1986) estimates that the number of full-time

20

agricultural researchers in Senegal rose 272 per cent between
1962 and 1985, from 55 to 205.

Despite these major increases in agricultural research
capacity, research at many African agricultural research
institutes has been largely unproductive (World Bank, 1987b).
Although exceptions do exist (see Jahnke et al., 1987), African
agricultural research has largely failed to make a significant
impact on increasing agricultural productivity. Mellor and
Johnston (1984) conclude that most national agricultural research
systems in Africa have failed to meet the "organizational
requirements of agricultural development". Ineffective national
research institutions limit the ability of a country to create a
science based agricultural production system.

Commonly cited problems plaguing African national
agricultural research institutions include: poorly matched
organizational objectives and organizational capacities, poorly
defined research priorities, weak management, a shortage of
qualified scientific personnel, high rates of transfer and
turnover in scientific staff, and severe revenue-expenditure
squeezes. In sum, the resources allocated to African agricultural
research institutes, and to scientific personnel in particular,

have not been efficiently employed in most countries.

21

thn;g;;2; AGRICULTURAL RESEARCH ORGANIZATION AND POLICIES
AFFECTING RESEARCH STAFF IN SENEGAL
This chapter provides background information about the
evolution of agricultural research in Senegal, and the Senegalese
Agricultural Research Institute (ISRA) in particular. For more
detailed information see Faye and Bingen (1988), ISNAR (1988),

and Sene (1985).

2.1) W

The history of agricultural research in Senegal can ‘be
divided into three periods: the Colonial era prior to 1960, the
post-independence era between 1960 and 1975, and the are from the

creation of ISRA in 1975 until the present.

2.1.1) 92W

Senegal was the administrative center for French West Africa.
In 1921, the French established a research experiment station in
Bambey, Senegal in order to increase peanut production. This
early research was primarily intended to benefit French vegetable
oil industry interests and trading firms.
In 1933, the research focus at Bambey was expanded to
include crops grown in rotation with peanuts such as millet,
sorghum, and cowpeas, as well as research on farming techniques

such as animal traction. In 1935 France created a veterinary and

22

livestock research institute in Hann, near Dakar, to conduct
veterinary studies for the entire semi-arid zone of French West
Africa. A chemistry laboratory was built at Bambey in 1936 for
soil science research. By 1938, Bambey had become the primary
site of agricultural research in French West Africa.

After World War II, great strides were made to develop a
coherent agricultural research policy in French West Africa. The
entire research structure was expanded and reorganized on the
basis of ecological zones. The Bambey research station, and the
veterinary institute at Hann became the research centers for
agronomic and animal research respectively for the Semi-Arid zone
of French West Africa (which included Chad, Niger, Burkina Faso,
Mali, Mauritania and Senegal).

Secondary research stations, and. 'numerous experiment
stations were created throughout the Semi-Arid zone. Thus, before
Senegal became independent in 1960, the French had created a
well-developed public research structure that had a major focus
on food crops and cropping techniques, as well as cash crops.

Also integrated. into the colonial research structure in
Senegal were a number of commodity based French institutes. These
commodity institutes concentrated solely on cash and export
crops. These institutes were funded from a variety of private and
public sources. As such, they were more independent and flexible

than the regional research centers which were funded by the

23

general French overseas administration. Jahnke et al._ (1987)
believe the research performance of these private commodity
institutes tended to be better than that of the regional research
centers.

Sene (1985) concluded that French colonial research greatly
enhanced the body of available knowledge about tropical
agriculture, and contributed to growth in agricultural production
and productivity by supplying improved crop varieties, cropping
techniques and fertilization guidelines. Great strides were made
in controlling livestock epidemics and improving animal health
and sanitation. Finally, the introduction and adoption of animal
traction techniques in many areas of Senegal was largely a result

of French research.

2.1.2) ost- e ende ce 9 0 - 19
Upon gaining independence in. 1960, Senegal inherited. the
Bambey and Hann research stations, and those research sub-
stations in Senegalese territory. The management and supervision
of these research installations were entrusted to the French
tropical agriculture research institutes.
The French institutes provided scientists and managers for
their research activities. Senegal provided agricultural
technicians and support staff. The funding of agricultural

research was divided evenly between France and Senegal. Mixed

24

Senegalese and French commissions were responsible for research
planning and for setting research agendas. In theory, Senegalese
authorities had the final say in these commissions. However,
given the French role in financing research and. ability to
manipulate funds, the French research institutes had considerable
influence over research planning (Jahnke et a1, 1987).

During this period, few Senegalese researchers were trained.
In 1974, only eight of fifty agricultural researchers were
Senegalese (Faye and. Bingen, 1988). Furthermore, high
administration costs and poor coordination among the French
institutes forced Senegal into providing more financial support
than had been planned. France opposed Senegalese efforts to seek
other external sources of funding, or to engage in research

collaboration with other countries.

2.2) EEWWW
2.2.1) WW

ISRA was founded in January, 1975 to permit an increased
Senegalese role in defining and conducting agricultural research.
The goals of ISRA as stipulated in the empowering legislation
are:

1) To conduct and develop crop, animal and fisheries

research which contributes to the economic and social

development of Senegal, and to formulate research programs

consistent with governmental objectives.

2) To implement research programs set by the
Interministerial Council on Scientific Research.

25

3) To collect, preserve, and protect all Senegalese
agricultural research.

4) To manage all agricultural research stations, and sub-
stations.

5) To promote and assist the creation of a competent
national research staff.

6) To develop inter-African and. international scientific
cooperation.

The period between 1975 and 1982 was a transition period for
ISRA. Until 1982, much of the inherited research structure
remained unchanged. Research was organized and oriented along
the strict disciplinary lines inherited from the French.
Gradually, the French role in both managing and conducting
research was lessened as more and more Senegalese administrators
and researchers were trained. Although ISRA's links with France
remained strong, this period witnessed an expansion of
relationships with non-French institutions and organizations.
The role of the World Bank and USAID in funding and supporting
research projects became particularly strong. Other support came
from Canada, Belgium and the Food and Agriculture Organization
(FAO), and Great Britain.

Between 1978 and 1982 the World Bank planned and designed
the Agricultural Research Project, a strategy designed to: l)
strengthen ISRA's research infrastructure, 2) reorganize the

institute, 3) improve scientific capacity, and 4) reorient

26

research towards addressing problems that face the rural
population. In 1982, the Agricultural Research Project was put
into effect. Key features of the Agricultural Research Project
were the establishment of five regional farming systems research
programs in Senegal, and six nationally coordinated and
multidisciplinary commodity programs on peanuts, cowpeas, millet,
sorghum, maize and rice.

In 1985, a multinational project review and evaluation team
concluded that: 1) ISRA's financial management was inadequate and
unable to mobilize funds in timely fashion, 2) there ‘was a
chronic over-expenditure, especially for salaries, 3) there was a
lack of motivation among researchers due to limited long-term
career prospects, and 4) ISRA's performance and efficiency were
sub-par. As a result of this report, ISRA decided in July 1986
to reorganize its management and administration, to dramatically
reduce its total staff, and to improve coordination of research
programs.

Thus, since 1975 ISRA has largely pursued the slow process
of building institutional and scientific capacity. There have
been two major structural reorganizations. The training and
development of a competent national. research staff has been a
major priority. ISRA has also attempted to redefine and reorient
its research objectives. The Director General of ISRA considers

ISRA to be in a "semi-developed" state (World Bank, 1988).

27

2.2.2) W

ISRA's enabling legislation gives ISRA the status as a
"public establishment with industrial or commercial
characteristics (EPIC). There are 16 EPIC institutions in
Senegal. There are both benefits and disadvantages to classifying
an agricultural research institute as EPIC. One advantage of EPIC
status is that ISRA has a large degree of autonomy in managing
its affairs. EPIC statutes give ISRA complete autonomy in
recruiting personnel.

Most EPIC institutions produce goods or provide services
that generate income. As such, these government organizations are
self-sufficient, generating many of the financial resources they
need for their operations. However, EPIC status is not completely
appropriate for an agricultural research institution.
Agricultural research is a long-term process. The benefits of any
given research project are difficult to measure. Successful
research is never guaranteed. Many research results have public
good attributes, and cannot be bought and sold. Thus,
agricultural research does not, by nature, generate goods or
services that can be sold to produce income.

Although EPIC status permits ISRA to sell vaccines and the
harvests from field trials, it also creates major financial
constraints. As a result of its limited ability to generate

income, and poorly conceived personnel hiring practices,

28

salaries have absorbed the Senegalese government's financial
contribution to ISRA. Thus, ISRA must rely on donor sources to
finance the bulk of its research activities.

In 1986, to help ease the pressure of a long-accumulating
public debt the Senegalese government decided to reduce its
financial contribution to all EPIC organizations by 5 per cent a
year for 5 years. This decision has had particularly adverse
consequences for ISRA given the fact that ISRA has limited

opportunities to generate financial resources of it's own.

2.2.3) W

The institutional model and actual procedure for defining
agricultural research policy and priorities in Senegal differ. On
the governmental level, the Ministry of Rural Development (to
which ISRA belongs) allocates financial resources to ISRA from
the national budget. Donor agencies provide project funds.
ISRA's General Direction then develops a research agenda and
budget that is based upon the government's agricultural policy.
The research plan is examined by the Technical and Scientific
Committee, which is composed of the directors of state
agricultural technical service agencies and academic research
institutions.

The Technical and Scientific Committee makes recommendations

to ISRA's Administrative Council, which is composed of all

29

ministers who are involved with agricultural research. This
council approves ISRA's budget, and sends it to the Ministry of
Finance for final approval.

Internally, ISRA's Director General allocates financial
resources and identifies research priorities to research
departments. In turn, the director of each research department
allocates resources to research projects within that department,
and identifies research priorities to program coordinators.
Program coordinators plan research activities and project budgets
in collaboration with researchers. Once finalized, program
agendas and budgets are sent to research department directors,
and then to the Director General for approval.

Faye and Bingen (1988) report that in practice, the
government's agricultural development plan is often vague and not
directly stated. There is little coherent planning of
agricultural research at the governmental level. Also, there are
weak mechanisms within ISRA to ensure that government policy and
research, and the research activities of research departments are
well coordinated. Since the government's budgetary contribution
to ISRA does not cover salaries, much less research activities,
ISRA has very little flexibility in allocating resources. Donor
agencies provide operating and research funds for all ISRA
research projects, and thus wield a major influence over research

policies and research programs. The different policies and

30

research agendas of donor agencies complicate the management,
planning and coordination of research activities at ISRA.

In 1987, as part of the plan to reorganize ISRA, a new
research orientation and strategy was conceived which stresses:
1) a collaborative research planning process at the regional
level that emphasizes cohesiveness between the government's
regional development policies and programs, and ISRA's research
program priorities and objectives, 2) research programs with a
multidisciplinary, farming systems approach, and 3) improved
links between research and extension, as well as international

research centers.

2.3) MW

Prior to the implementation of the Agricultural Research
Project in 1982, the organization and structure of ISRA followed
the model inherited from the French. Management of research was
structured on a dual basis-- by research department and by
research center. The research departments were: agronomy and
bioclimatology, agricultural machinery and equipment, soil
science and hydrology, economics and sociology, veterinary
medicine and animal health, forestry and oceanography. There
were ten research centers.

The directors of each research department planned research

programs and budgets. Directors of each center were responsible

31

for the management of their respective centers, the allocation of
resources for research, and the implementation of research
programs. There were no defined authority links between research
departments and centers, and resources were often allocated on an
arbitrary basis by research center directors. Planned research
and actual research were not always consistent.

At the same time that the World Bank designed a
reorganization plan for _ISRA as part of it's Agricultural
Research Project, ISRA designed a reorganization plan of its own.
Key differences existed between the plans. In the World Bank plan
the previously independent research centers were to be placed
under the authority of a Research Support Department. Research
departments were to remain under the authority of the Director
General.

In the ISRA plan, research departments were to be placed
under the authority of a Scientific Director, whereas the
independent research centers would remain under the authority of
the Director General. Both the ISRA and World Bank plans retained
a dual structure between research departments and research
centers. Common to both plans was a reorganization of research
departments organized by discipline into more multidisciplinary,
systems oriented departments.

In 1983, a compromise reorganization plan put into effect.

Research departments were reorganized along more multi-

32

disciplinary and less rigid lines. Five new research departments
were created: Crop Production, .Animal Production. and. Health,
Agricultural Production Systems, Forestry, and Oceanography. The
research departments were put under the authority of a
coordinating body, the Scientific Director. However, the
research centers remained. under the control of' the Director
General's office. Also, at the administrative level three offices
were created to enhance research and personnel planning: the
Office of Human Resources, the Office of Macroeconomic Analysis,
and the Office of Planning and Evaluation. This organizational
plan remained in effect until 1987.

Thus, between 1974 and 1987 ISRA had a largely decentralized,
and dualistic organizational plan. The research centers had a
large degree of planning authority and retained a spirit of
independence (Faye and Bingen, 1988). There was no real
coordination between research planning and actual research.
Organizational instability has been a recurrent problem for ISRA.

In 1987, ISRA streamlined its organizational plan. In the new
organizational structure research became organized on a strictly
departmental basis, with research centers becoming attached to
the relevant research department.

An office was created within the Crop Production Research
Department to coordinate the activities of the previously

independent Bambey, Djibelor, St. Louis, CDH (horticulture), and

33

newly created Tambacounda agricultural research centers. Also,
the Bambey and Kaolack research departments were combined. The
LNERV (veterinary institute), Kolda, and Dahra research centers
were placed within the Animal Production and Health Research
Department. The CNRF (forestry) center was placed within the
Forestry Research Department, and the CRODT oceanography
institute placed within the Fisheries department. The former
Office of Macroeconomic Analysis and Production Systems research
department were merged to become the Agrarian Systems and
Agricultural Economics department, which oversaw' the farming

research teams .

2.4) Weiss
2.4-1) We].

In 1960, only four people in Francophone Africa had received
university degrees in agricultural disciplines compared with 150
Anglophone Africans (Eicher, 1984). The first college-level
agricultural school (for W) in Francophone Africa was
established at Thies, Senegal in 1979. As recently as 1974, there
were only eight Senegalese researchers among the fifty
agricultural researchers in Senegal.

Since its creation in 1974, ISRA has undergone a rapid
expansion in research personnel. In particular, ISRA has been

able to train numerous national researchers and "nationalize" its

34

Table 2.1: Research Personnel at ISRA (1974 - l987)'

 

 

 

Year Total Number of Expatriate Senegalese
Researghgrs

1974 50 42 (84%) 8 (16%)

1979 85 46 (54.1%) 39 (45.9%)

1985 198 68 (34.3%) 130 (65.7%)

1987 148 41 (27.7%) 107 (72.3%)

 

Source: Faye and Bingen (1988) and ISRA Personnel Files.

cadres of researchers. Table 2.1 shows both the increases in
research staff at ISRA, and the declining proportion of
expatriate researchers at ISRA since 1974.

The majority of the expatriates are French researchers
working with CIRAD and ORSTOM, two major French overseas research
institutions. Despite their decreasing proportion at ISRA,
expatriate researchers continue to play a pivotal role in the
sorghum, cotton, and peanut crop research programs. Expatriate
researchers supply expertise in forestry, animal health,
agronomy, and agricultural machinery-- areas in which the
qualifications of Senegalese scientists are weak. ORSTOM

scientists tend to focus their efforts on more basic research.

35

2.4.2) Edugational Background of Researchers

The educational qualifications of ISRA's Senegalese
researchers are comparatively higher than most African
agricultural research institutions. Expatriate researchers have
slightly lower levels of training than Senegalese researchers
(Faye and Bingen, 1988). Table 2.2 shows the educational profile

of all national researchers at ISRA in 1987. ISRA researchers

Table 2.2: Educational Profile of Senegalese Researchers at ISRA
in 1987.1/

 

a) University Diploma/ Limited Specialization:

 

 

 

De v Ember Percentage
Ingenieur 21 19.3%
Dr. Veterinaire 8 7.3%
Sub-total: 29 26.6%

b) Intermediate Post-Graduate Education:

 

 

 

DEA/DAA 35 32.1%
MSc. 26 23.9%
sub-Total 61 56.0%

c) Advanced Post-Graduate Education:

 

 

PhD/ Dr. d'etat 8 7.3%
3eme Cycle 11 10.1%
sub-total 19 17.4%

 

1 - includes Fisheries Department and Research Administration.

 

Source: ISRA Personnel Files.

36

Table 2. 3: Educational Characteristics of Senegalese Researchers
In Comparison to Other African Countries.

 

Mount Lara LS; 1.1.19 ELL.
Senegal 1987 26.6% 56.0% 17.4%
Ghana 1981 21.4 50.4 28.2
Kenya 1982 52.5 44.7 2.7
Malawi 1981 66.7 26.9 6.4
Nigeria 1984 27.9 43.7 26.4
Somalia 1983 84.8 12.1 3.0
Uganda 1981 71.9 24.5 3.5
Zambia 1984 49.0 41.3 9.4
Zimbabwe 1984 58.7 29.8 11.4
Swaziland 1984 55.5 44.0 0.0
Mozambique 1984 92.9 7.1 0.0
Lesotho 1984 66.6 33.3 0.0
Botswana 1984 62.0 38.0 0.0

 

Source: Bennell (1986) and ISRA Personnel Files.

have received post-graduate education in both the United States
and. in France. Thus, degrees have been. grouped. by level of
specialization.

Table 2.3 shows that the educational characteristics of
Senegalese researchers at ISRA compares favorably to those of
national researchers at other African national research
institutions. The majority of the countries listed in Table 2.3
are English-speaking. Although exact figures are unavailable, the
educational characteristics of ISRA researchers are considerably
higher than those of researchers in other French-speaking

countries in Africa (Faye and Bingen, 1988).

37

2.4.3) c 0 Su t Sta

Although Senegalese researchers at ISRA are better educated
than many of their African counterparts, the qualifications of
technical support staff and administrative staff are limited.
Faye and Bingen (1988) report that two years of education beyond
high school is the normal educational background for technical
support staff and administrative staff. In 1986, of 1354 total
personnel employed at ISRA, 445 (33 per cent) were illiterate,
and 490 (36 per cent) had received only general, primary school
level training. Thus, 69 per cent of ISRA personnel in 1986 had
not received specialized training in research.

Faye and Bingen conclude that although in. numbers the
research support staff at ISRA appears satisfactory for a country
the size of Senegal, their professional qualifications and skills
are insufficient for an agricultural research institute. It may
be inferred that this weakness constrains the research activities

of ISRA scientists.

2.4.4) ‘x e P e o ona Re ear he

As with many African countries, Senegal has tried to
develop an experienced corps of research personnel. Table 2.4
compares the experience levels of Senegalese researchers in 1983
and 1987. Table 2.4 shows that while the experience level of many

researchers is gradually increasing, large numbers of researchers

38

Table 2.4: Experience Profile of National Researchers at ISRA:
1983 vs. 1987.

 

Years of

Impatience 1283. m
0 - 5 years 68.7% 42.0%
6 - 10 years 21.4% 37.0%
11 + years 9.8% 21.0%

Segzee: Bennell (1986) for 1983 and ISRA Personnel Files for
1987.

remain relatively inexperienced. Seventy nine per cent of ISRA
researchers have less than 10 years of experience. A large
proportion of researchers with 5 years of experience or less
indicates that many researchers have yet to fully develop the
skills needed. to effectively' perform research. Experience in
conducting research plays a major role in forming research skills

(see Section 3.1.1).

2.4.5) Age Profile ef Segegalese Reseagchers at ISRA

A by product of the limited training of researchers
undertaken by the French prior to the creation of ISRA in 1985 is
a youthful group of researchers. Figure 2.1 (following page)
shows that the majority of Senegalese researchers at ISRA were in
their 30's. The mean age of Senegalese researchers was 36.1 years

in 1987. Eighty per cent of Senegalese researchers were between

39

 

I?!“ {I q u 151.! n 1.3: y

 

Figure 2.1

negotese Researchers at ISRA
198.7

33*
CA.)
\ 1]
0 J
C!
---n
U]
(D

 

 

 

 

 

20 LEGEND
I... + Number
113T
II-
I3+
1.3+
7

4+ g 2

? " 1
TII 1 I
" I 35 ~
.1; f “a . 3
\hi Ii V I IVT- IT f

 

Source: {SBA Personne! F‘Iies

 

 

40

the ages of 30 and 40. The median age of Senegalese researchers
was between 34 and 35 years. Generally, the "elder statesmen" at
ISRA were in their low to middle 40's.

The fact most ISRA researchers are in the same general age
group has large implications for research turnover, as will be
seen in Chapter 4. Faye and Bingen (1988) believe that unless
changes are made, there will be blockages in ISRA's career

advancement system.

2.4.6) t --

The total number of ISRA employees rose by 24.4 per cent
between 1979 and 1985, from 1,200 to 1,493. Between 1983 and
1985, the total salary budget rose from 1,593,064,466 FCFA to
2,020,167,444 FCFA. However, during the same period governmental
financial support for salaries remained constant at 934,731,000
FCFA. To reduce this budgetary deficit, ISRA decided to lay off
473 employees in July, 1986. Despite these layoffs, ISRA still
has a personnel budget deficit. Thus, further cutbacks in
personnel are anticipated. Only two researchers were included in
the 1987 layoffs.

The continuing burden of salary payments severely constrains
ISRA finances, and thus limits research productivity. Shortfalls
in operating funds, untimely procurement of needed research

resources, and erratic funds for day-to-day expenses interfere

41

with the smooth operation of research projects.

2.4.7) geeruigmeng egg Ireiging

ISRA has well-established recruiting procedures which are
not always followed in practice. Recruitment entails the
creation and budgeting of a research position by ISRA's General
Direction and the Administrative Council. The position is then
listed, and a candidate is selected from a pool of applicants by
the research department director. The candidate must then be
approved by a recruiting commission and the Director General.
Prior to being appointed to full researcher status, a recruit
must complete a major research project during a probationary
period. If successful, a researcher is hired and given a contract
of indeterminate length.

Given limited numbers of qualified Senegalese researchers,
recruitment in the past has frequently followed the path of
choosing a candidate and immediately sending that candidate
overseas for specialized training. Thus, upon completion of
graduate education ISRA must again recruit the candidate. Faye
and Bingen (1988) report that recruitment in the past has largely
been based more on the availability of governmental and donor
agency scholarships than on systematic personnel planning based
on research needs.

Given the lack of systematic planning, jobs are often

42

unplanned for researchers who have completed their overseas
training. The probation period for a recruit can last up to two
years due to bureaucratic delays, financial shortages, and slow
planning of a recruit's duties. Thus, ISRA's recruiting practices
is inadequate and often demoralizes beginning researchers.

ISRA has assigned a high priority to training Senegalese
researchers since 1975. However, Faye and Bingen (1988) report
that ISRA lacks a well-defined manpower planning policy. The
recruitment and training of scientists in the past has largely
been guided by the goal of replacing each French researcher with
a Senegalese researcher. Manpower development has occurred
without reference to anticipated financial resources, or a medium

or long term plan for agriculture research.

43

W: FACTORS AFFECTING RESEARCHER PRODUCTIVITY, JOB
SATISFACTION AND TURNOVER

3.1) The Preeees of Agricultegal geseerch
3.1.1) Researeh Cogginuity

Agricultural research is a long-term process. Ruttan (1982)
believes that major innovations from agricultural research are
the ”cumulative synthesis" of' a. number of' relatively simple
innovations. Pinstrup-Anderson (1982) sees agricultural research
as a chain, the individual links of 'which. are agricultural
research projects. Research is a highly uncertain process since
in conducting research one is essentially exploring the unknown.
For any given research project there are no guarantees of
producing tangible results. Thus, research is a process which
builds upon the results of previous research.

The results of a research effort are dependent to a large
degree on the training, experience, intelligence, abilities and
motivation of individual researchers (Pinstrup-Anderson, 1982).
Ruttan (1982) belieVes that all research innovations are the
result of "acts of insight" by individual researchers with high
degrees of skill. Skill is perceived to be a function of both
education and experience.

Continuity in research efforts and staff is a fundamental
requirement for producing effective research results.

Agricultural research projects often must cover several years

44

before tangible research results are obtained. The World Bank
(1988) estimates that on average it takes ten years to breed and
introduce a new crop variety, and twenty years to breed and
introduce new tree and animal varieties. The Special Program on
African Agricultural Research--SPAAR (World Bank, 1987b) uses as
a rule of thumb estimates of 4 to 5 years to complete adaptive or
maintenance research projects, and ten years to complete basic
research programs.

Much of a researcher's skill is acquired on the job, or
learned through doing research. The definition of research
problems, the identification. of' research. priorities, and. the
designing of research projects, are scientific skills that are
enhanced with experience. Wortman (1977) believes that career
scientists represent a store of knowledge that is often lost when
research efforts are of short duration and participants scatter.
Experienced, career scientists are especially important in
providing leadership in research and in training and advising
inexperienced researchers. Thus, continuity in research staff is
essential.

ISNAR (1984) and Schultz (1979) believe that high rates of
return to research are generated by cadres of competent
scientists who remain with a research institution over the course
of their careers. Given the interdisciplinary nature of

agricultural research, successful research efforts are critically

45

dependent upon the interaction of scientists with different
training and points of view. Nestel and Trigo (1984) believe that
research continuity is essential in the development of productive

and complementary work styles between scientific personnel.

3.1.2) WW

There are two types of research-- basic and applied. Arndt
(1977) describes applied agricultural research as a search for
new technology and knowledge, using the results of basic research
programs. The World Bank (1988) also distinguishes between
adaptive and maintenance research. Adaptive research molds new
technologies to specific needs and agro-ecological conditions.
Maintenance research adjusts and improves technology in use by
producers to changes in the agro-ecological milieu such as
disease.

Basic research establishes the fundamental knowledge and
research methods, or frontiers, within which innovation is
possible. Thus, there is an interaction between basic and applied
research. In general, basic research is considered to demand more
highly developed research skills. Experience at the "scientific
frontier" demands at least Ph.D. level training.

A number of authors refer to the "small country" research
problem. Ruttan (1982) and the World Bank (1981) see small

countries as having limited population bases, and hence limited

46

financial and human resources. As a result, these authors think
small developing countries are unable to afford basic research.
Instead small countries should rely on the basic research that
has been done by larger countries or international agricultural
research centers. Small countries should devote the bulk of their
resources to applied research on commodities of major importance
to the economy and to farming systems.

Von der Osten (World Bank, 1988) of the International
Service for National Agricultural Research (ISNAR) warns African
national agricultural research institutions against focusing too
much of their efforts on applied research. Though von der Osten
agrees that applied and adaptive research are of central
importance to Africa, he believes that much of the basic research
required for successful technological development in Africa has
not yet been undertaken. In particular, many staple food crops
and dryland farming systems in Africa have been poorly
researched" Jahnke et a1. (1987) support this view; .African
agricultural research institutions must strike a balance between
basic and applied research.

An emphasis on adaptive or maintenance research does not
imply that the qualifications of researchers may be lessened.
Ruttan (1982) argues that adaptive and. maintenance research
require the same research skills and capabilities that are

required in generating new technology. A highly educated and

47

trained research staff, especially doctoral level training, is
crucial for agricultural research institutions in developing
countries.

This study assumes that given Senegal's tradition of being
the colonial center for agricultural research in semi-arid French
West Africa, and given the limited basic knowledge base
applicable to Africa, basic research and applied research are

both important areas of inquiry at ISRA.

3.1.3) W

The most critical element in a building an agricultural
research institutes's capacity to conduct research is the
establishment of a well-qualified professional research staff.
Evenson (1977) presents a typology of researcher skills that is
instructive in gauging the lost skills resulting from turnover in
research staff. Ieehnieel skills are learned by researchers in
most undergraduate level programs in agriculture. Evenson
believes that undergraduate university education establishes the
basic ability to apply "textbook" principles to technical
problems.

Ieehnieeleeeiegeifiie skills generally are received in
Master's programs. Evenson believes these programs impart high-
level technical skills, but limited conceptual skills.

Cogcepgual-scientifiie skills generally are learned in PhD level

48

graduate programs. In this skill category conceptual skills have
become highly developed, and scientists have had experience at
the "scientific frontier".

There is widespread agreement that high level post-graduate
training in specific disciplines is a prerequisite for acquiring
basic research methodology skills (Bennell, 1986). Evenson
believes that without such training, researchers engage in
relatively low level, unproductive research. Often this entails
repeating work that has been done elsewhere. Thus, assembling a
staff of researchers who possess higher order research skills is
important in manpower development plans.

A key issue in building research capacity is the achievement
of "critical mass" in national scientific personnel. However,
there is no set criteria for determining "critical mass" since
staff requirements vary by commodity, agro-ecological
circumstances, and agricultural policy. It is beyond the scope of
this study to define "critical mass". However, given the relative
youth of ISRA as an agricultural research institution, and given
the large role currently played by expatriate researchers in
research programs, it is believed that ”critical mass" in
national research staff has not yet been achieved at ISRA and
remains a long—term institutional goal.

Despite dramatic growth in numbers of researchers at most

African national agricultural research institutions over the last

III

49

25 years, understaffing of research projects is a commonly cited
constraint on the productivity of African agricultural research
(Bennell,l986). Researchers at African agricultural research
institutes typically are very young and inexperienced. Table 2.4
(see p.35) shows that a large number Senegalese researchers
remain inexperienced. Seventy nine per cent of Senegalese
researchers at ISRA have less than 10 years of experience. Forty
two per cent have 5 years of experience or less. A shortage of
experienced researchers hampers a country's ability to both
formulate and implement research projects, as well as to train
and supervise junior researchers.

Expatriate researchers have often been relied upon to
overcome Senegal's limited numbers of experienced researchers
(see Table 2.1, p.34). Bennell argues that there is an
increasingly serious shortage of highly educated and experienced
expatriate researchers who are willing to work longer than three
years in Africa. Thus, the recent trendchas been for African
agricultural research institutes and donor projects to rely upon
expatriate consultants with contracts of three years or less.

Some authors are worried by a heavy reliance on short-term
expatriate researchers. Jahnke et al. (1987) believe that short
term stays by expatriates disrupt research continuity in those
national research projects that are dependent upon expatriate

researchers. SPAAR (World Bank, 1987b) worries that the

50

structural and managerial autonomy enjoyed by some expatriate
researchers complicates and impedes national research programming
and budgeting. Bennell (1986) believes that short-term, and less-
well qualified expatriate researchers are inappropriate for
serving managerial and training functions at African agricultural
research centers. He also believes that pressures upon
expatriates to publish in order to further their careers at home
may distort their research objectives and further impede their
contribution to African agricultural research.

Finally, rapid expansion of research facilities and/or
research programs combined with limited numbers of professional
researchers can lead to understaffing of research programs and
projects. In these circumstances existing scientific manpower at
agricultural research institutions gets spread too thin over a
number of research projects. In determining the scope of a
national agricultural research institution's research effort,

planners should account for limitations in scientific manpower.

3.2) W
3.2-1) W

A common problem in African national agricultural research
systems is insufficient financial support from governmental
sources. The FAO and UNDP (1984) recommend that countries should

spend at least the equivalent of 1 per cent of agricultural Gross

51

Domestic Product (GDP) on agricultural research. Most African
countries spend between 0.4 and 0.75 per cent of their
agricultural GDP on agricultural research (de Janvry, 1985), well
below the target level of l per cent.

Senegal has traditionally demonstrated a greater financial
commitment to agricultural research than most African countries.
De Janvry (1985) shows that in 1980 the Senegalese government's
$9,797,000 contribution to agricultural research represented 1.21
per cent of agricultural GDP. Only two countries in Africa,
Zimbabwe and Mali, exceeded this figure.

Due to budgetary constraints the Senegalese government's
financial contribution to ISRA, has decreased significantly in
both current francs and constant real francs since 1981 (Faye and
Bingen, 1988). Despite this decreasing financial support ISRA has
rapidly expanded its facilities and staff since 1981, fueled in
large part by donor initiatives. The result has placed a severe
strain on ISRA's resources. By 1987, Senegal contributed the
equivalent of only 0.9 per cent of agricultural GDP to
agricultural research. Although these figures should be treated
with caution since many external factors such as weather
influence agricultural GDP, they remain a commonly accepted
measure of a country's commitment to agricultural research.

The Senegalese gpvernment's financial contribution to ISRA

is entirely absorbed by the salaries of ISRA personnel (Faye and

52

Bingen, 1988). Thus, donors such as France, the World Bank and
USAID play pivotal roles in providing funds with which to conduct
research. The influence this funding provides donors in
formulating ISRA's research agenda was described in Chapter 2.
Donors provide nearly all the operating funds with which ISRA
researchers conduct research.

A commonly cited problem with donor funding of agricultural
research is the 4-6 year time frame of many donor investments.
Donor funding for agricultural research projects in developing
countries is often unstable-- plentiful during the course of
donor projects but scarce as projects wind down or are
renegotiated. The erratic disbursement of donor funds often de-
stabilizes agricultural research programs. Gamble (in Nestel and
Trigo, 1984) believes this de-stabilizing force is compounded by
the inconstancy of domestic support. Erratic and uncertain
funding of research hinders the long-range planning of research
programs, and can create disjointed research programs.

Donor projects designed to expand national research capacity
(facilities and staff) often fail to take into consideration
developing countries' and research institutions' resource bases.
Overly ambitious expansion of research facilities, research
programs, or research staff often severely strains a research
institution's limited financial, administrative, and technical

capacities (Daniels and Nestel, 1981; Nestel and Trigo, 1984). If

53

researchers receive inadequate financial and logistical support,
their ability to both conduct research and achieve professional
goals is hampered. As a result, a researcher's motivation, morale
and job satisfaction might tumble. A researcher's performance is
lowered.

Low operational funding of research contributes to the
turnover process. Research personnel returning from overseas
studies to their national research institutes often find working
incentives and operational support at levels which are far below
the levels they received and learned to expect during their
training abroad (Rohrbach, 1984). Daniels and Nestel (1981)
believe instability in funding keeps salaries low, and results in
erratic cash flows and budget cutbacks, thus aggravating the

turnover problem.

3.3) Wm

The outcome of the research efforts of an agricultural
research institute depends heavily. upon the institutional
environment within which research is performed (Ruttan, 1982).
Thus, issues in research management, organization and planning
are important in describing conditions of service for
agricultural researchers. -.

Clear articulation of national agricultural research

policies and priorities is essential for the formulation of

r1

54

coherent research programs. If weakly stated, research managers
often lack formal criteria upon which to guide their resource
allocation decisions. Furthermore, poorly defined research goals
often result in redundant or irrelevant scientific
investigations, at failure to make use of available technologies,
and a loss of institutional credibility. As a result, recruitment
of researchers becomes more difficult, and currently employed
researchers become demoralized. In sum, the productivity of the
research institute suffers.

A commonly cited weakness of African agricultural research
systems is the lack of systematic, long-run research and manpower
planning (World Bank, 1987b). Staffing and training decisions are
often 'based upon the availability’ of scholarships, or donor
interests in funding or staffing research programs. Governments
often expect a quick payoff to investments in upgrading research
staff capacity, ignoring the fact that such investments yield
only long-term payOffs. Rapid expansion of research staff and
research facilities in the absence of careful analysis of
research needs and resource constraints strains limited research
capacity. Many African national research institutes' employment
strategies are conducted on an ad hoc basis. As seen in Section
2.4.7, ISRA is no exception.

The organization and management of research provides the

framework within which .research priorities are defined and

55

implemented, and personal responsibilities, relationships, and
patterns of communication are formally decided. Thus, continuity
of organizational leadership and structure is essential in
establishing an environment conducive for research. Unstable
organizational leadership often leads to frequent changes or
reversals in research policies. As a result, the research goals
and priorities of an institution often become confused. If the
environment within which- research is conducted is uncertain,
individual research goals often take precedence over
institutional goals. Thus, an institute's research performance
suffers.

Idachaba (1985) believes that a major weakness of most
African agricultural research systems are institutional
arrangements for administering research which are in a constant
state of flux. The period between 1981 and 1987 at ISRA was
characterized by institutional instability. As seen in Chapter
2, ISRA has undergone three major reorganizations since 1975. The
organizational structure at ISRA was very vague between 1982 and
1985. The government ministry to which ISRA is attached changed
five times between 1975 and 1987 (Faye and Bingen, 1988). As a
result, the resource allocation procedures and hierarchical
relationships at ISRA have been unstable.

In investigating continuity in managerial leadership, the

qualifications and experience levels of managerial and

56

administrative personnel must be explored. The lack of qualified
research managers is a commonly cited problem of African national
agricultural research systems. Traditionally, manpower training
and expansion programs have focused 'primarily upon research
staff. The training of management and administrative personnel
has largely been ignored. Limited qualifications in
administrative and managerial staff complicates both research and
manpower planning, and the allocation of resources to research
(World Bank, 1987b).

Frequently, research managers and administrators are
selected from the ranks of experienced researchers employed at an
agricultural research institution. A number of authors consider
this practice to be a misuse of resources. Since management
training is rarely incorporated a researcher's training program,
the promotion of researchers into administrative positions often
makes poor administrators out of good researchers (SPAAR, World
Bank, 1987b). A common problem is that the researchers selected
for administrative positions are often amongst the most
experienced and qualified of researchers at a research
institution. If these researchers spend the bulk of their time
doing purely administrative tasks, scientific leadership and
continuity in research is undermined.

Promotion policies and salary structures which reward

administrative and managerial functions rather than scientific

57

functions often. detract researchers from focusing their full
efforts on research. Researchers may spend their time politicking
and positioning themselves to increase the probability that they
will be promoted into administration.

Daniels and Nestel (1981) see both positive and negative
aspects to promoting scientists into administrative positions.
The benefit is that the scientists involved have a fundamental
understanding of the research process. Conversely, an
institution's overall managerial capacity may be weakened given
most researchers' limited skills in management and
administration. Since the development of management skills and
management style must take place while on-the-job, research
planning and resource allocation suffer while scientists acquire

skills in management .

3.4) lemeve; in fieeeerch Sgeff
3.4.1) Typee e: Turgover

Natural turnover refers to retirements, and departures
resulting from illness, injury or death. Although researchers who
leave an institution due to retirement carry with them a lifetime
of research experience and knowledge, retirements can be
anticipated and accounted for in personnel planning. Natural
occurrences such as illness and death are impossible to

anticipate. The focus of this study is voluntary turnover, or

58

attrition, which refers to researchers who resign to accept
alternative jobs.

Another concept that is important in discussing turnover is
the question of internal movement within a research institution.
Internal movement refers to changes of job function, location or
research project. Promoting scientists into administration is an
example of internal movement. Many' authors agree that large
numbers of researchers changing programs is counterproductive to
research since it disrupts continuity in research staffs. In
changing research roles, a researcher usually must familiarize
himself with a new research problem and new' colleagues. If
replaced by another researcher, the replacement must undergo the
same learning, process. Thus, 'widespread internal. movement of
researchers impedes and disrupts research efforts in the short-
term.

Internal movements of researchers are often associated with
voluntary turnover. Researchers who remain at an institution are
often transferred to replace departed researchers. Thus, the
departure of a researcher can send a ripple of instability
throughout an institution. Some internal movement is the result
of donor projects which demand counterpart national researchers

for their activities.

59

3.4.2) Qaueee of Zergove; i3 Researeh Staff

Most authors believe that the underlying causes of turnover
relate to three factors: conditions of service, salary and
professional rewards, and the work environment.

Ruttan (1982) has developed a model of a typical researcher.
The reward structure for research scientists is comprised of both
financial earnings and professional recognition. Ruttan believes
a typical researcher is prepared to accept a nigh degree of
deferred earnings and professional recognition as long as he or
she develops professionally. Professional development refers to
improvements in research skills, publications, and assuming
greater responsibility. Income is perceived to be a positive
function of a researcher's output. As a researcher acquires
experience and becomes more productive, increased rank and
earnings are expected. Ruttan believes that this reinforces a
researcher's initial research orientation.

If increased rank and earnings fail to accompany increased
productivity, a researcher will shift his research efforts
towards research that has more short-run payoffs, or leave the
institution altogether. Researchers involved in basic research on
the scientific ”frontier" are often prepared to substitute
professional recognition and status for financial earnings. Thus,
the incentive structure is critical in channelling research

efforts, motivating researchers, and in retaining researchers.

60

Human. capital theory' explains the. economic. calculations
involved in an individual's decision to leave a job. Hammermesh
and Rees (1984) believe that the probability that a worker will
leave a job is increased if wage differentials exist between
jobs requiring similar skills. If an employee perceives that
higher relative wages and nonpecuniary advantages are available
by accepting alternative employment, the employee will leave his
current job.

However, leaving a job is not without cost. There are
economic costs such as the cost of seeking another job, of moving
one's family and belongings, and psychic costs involved in
leaving. Psychic costs include leaving one's friends, leaving
familiar work and living surroundings, and uprooting one's
family. Hammermesh and Rees believe that if the present value of
the discounted earnings and benefits in alternate employment are
greater than the discounted costs of leaving a job, it is in the
economic interest of a worker to leave his job.

There is a near unanimous belief that low salaries are a
major contributing cause to high rates of turnover in research
staff at national agricultural research institutes in developing
countries. Compounding the problem of low salaries is the lack of
opportunities to be promoted. Frequently, national agricultural
research institutes do not offer incentives to researchers that

encourage scientists to remain at the institute over the course

61

of their career. Thus, faced with limited opportunities to
increase their rank and earning, scientists often stay at a
national research institute long enough to acquire research
experience before moving to higher paying jobs elsewhere.

In many African agricultural research institutes the real
wages of researchers have been declining over the past 10 to 15
years. Bennell (1986) 'believes declining 'real salaries are a
major factor in causing -low morale and dissatisfaction among

researchers, and thus fuels the turnover process.

3.5) e nce tive t c u e a ISRA
3.5.1) P o essi a R we ds

Professional rewards are a major component of an
agricultural researcher's incentive structure. Being recognized
by one's peers for research accomplishments, participation in
seminars and conferences, and especially publications are
important elements of a scientists's professional reward
structure. Faye and Bingen (1988) report that in general ISRA has
limited publishing facilities. Until the reorganization of 1987,
ISRA did not have a department to edit or screen documents and
assist researchers with papers. Most ISRA publications consist of
research program reports and a large number of mimeographed
reports that have not been well catalogued. Personal

communications with ISRA researchers indicate that the

62

performance of the ISRA publishing office has been poor.

Senegalese researchers rarely publish in international
journals. Until recently, ISRA did. not encourage or promote
international publication. Those Senegalese researchers who have
published internationally have generally had personal contacts in
the international scientific community, or have been participants
in expatriate research teams sponsored by donor agencies. Most
expatriate researchers publish in the journals of the
organizations with which they work.

Bennell and Thorpe (1987) conducted a survey of crop science
research publications by researchers at African agricultural
national research institutes between 1973 and 1983. In 1973, 36
articles authored by ISRA researchers were published in
international journals. Of these 36 publications, 34 articles
were published by expatriates, one published article was authored
by a Senegalese researcher and one article was co-authored by
Senegalese and expatriate authors. In 1978, there were 56
published articles, all in nmn-African journals or book
proceedings. Of these 56 articles, 53 were authored by
expatriates and three were co-authored by Senegalese and
expatriates. In 1982, there were 25 published. articles in
international journals of which 22 were authored by expatriates,
one by a Senegalese, and two were co-authored by Senegalese and

expatriates.

63

It should be noted that since documentation and cataloguing
of publications has been a major weakness at ISRA, it is likely
that many research publications by Senegalese researchers that
were circulated only within Senegal fail to appear in these
figures. Nevertheless, it appears that Senegalese researchers at
ISRA have limited opportunities for professional recognition
through publications, especially in the international scientific
community.

Other professional reward mechanisms for research scientists
include perquisites such as the opportunity to attend domestic
and overseas seminars, conferences, workshops, or short-term
training programs. Faye and Bingen (1988) report that although
not always evenly distributed, there are generally ample

opportunities for ISRA researchers to attend such functions.

3.5.2) W

Since its creation, ISRA's enabling legislation has given it
legal status as a public establishment with industrial and
commercial characteristics (EPIC). Since the implementation of
the EPIC salary statutes in 1977, ISRA employees 'have ‘been
classified on the basis of their job. Advances in rank and salary
are contingent upon changes in classifications.

Between 1976 and 1982, a salary structure applicable to ISRA

was formulated. A proposed structure developed by researchers

64

that emphasized scientific functions over administrative
functions was rejected. The salary structure that was adopted was
developed by managers. Faye and Bingen (1988) report that as a
result, the ISRA salary structure rewards administrative roles
and positions more than scientific positions.

The number of job categories were tightened and restructured
in the salary regulations adopted in 1982. Salary became based
entirely on job title rather than an individual's qualifications.
Qualifications for a given job were not systematically or clearly
stated. Faye and Bingen report that the 1983 reclassification of
jobs that accompanied the adoption of the new ISRA salary
regulations, created great disparities. The salaries of
researchers and agricultural technicians were lowered or
flattened. The salaries of a number of support personnel doubled.

As a result of the 1983 reclassification, total salary costs
at ISRA rose by 38 per cent, far exceeding the Senegalese
government's financial contribution to ISRA (see Chapter 2.4.6).
However, at the same time that the job reclassification took
place, the government decided to reduce its funding of ISRA by 20
per cent over five years. As a result of these funding
constraints, salary bonuses that are contained in EPIC salary
regulations have gone unpaid since 1983.

The EPIC statute assigns workers on the basis of job

function to one of eight job classifications. ISRA researchers

65

are all classified as class 6, which refers to a professional
cadre with superior qualifications and decision making
responsibilities. There are three sub-categories in. grade 6:
grades 6-1, 6-2, and 6-3. With each grade classification is a
base salary. A researcher's base salary is changed only by being
promoted into a higher grade class. If the base salary is raised,
the raise is given to all members of a given job classification.

In 1987, the base salary for researchers classified as 6-1
was 199,350 FCFA per month. The monthly base salaries for grade
classifications 6-2 and 6-3 were 209,050 FCFA and 219,750 FCFA
respectively. A researcher who remains a researcher over the
course of his career at ISRA can thus anticipate only a 10 per
cent increase in base salary if promoted from grade 6-1 to 6-3.
Differences in research experience and education do not lead to
higher base salaries.

EPIC salary regulations include higher job classes. Grade 7
and its three categories are assigned to "directors of an
establishment", or at ISRA, research center directors. Grade 8,
which also has three categories, is assigned to the Director
General of the institution. Analysis of ISRA personnel files
indicates that grade 6-3 is largely linked to administrative
positions such as director of a research department. In the EPIC
salary structure management and. administration positions are

clearly more highly rewarded than research positions.

66

The EPIC salary structure also envisions bonuses for length
of service. These bonuses are calculated on the basis of a
researcher's base salary. After three years of service at ISRA, a
researcher receives an additional 3 per cent of base salary each
month. This 3 per cent bonus based on length of service continues
through the fifth year of service. After the fifth year of
continuous employment at ISRA, the monthly length of service
bonus rises to 5 per cent. The bonus rises an additional 1 per
cent for each year of service at ISRA until the fifteenth year.
After the fifteenth year of service, the length of service bonus
remains constant at 15 per cent until retirement. Thus, under the
EPIC salary regime a researcher who does not change job
classifications can anticipate a l per cent raise each year
during his sixth through fifteenth years of service.

The EPIC salary provisions foresee the payment of a
productivity bonus, the amount of which is determined by an
evaluation. However, ISRA does not pay productivity bonuses. In
part this is the result of difficulties in defining the
productivity of a researcher. However, the most important reason
for productivity bonuses going unpaid is the classification of
ISRA as an EPIC institution. EPIC institutions usually are
industrial or commercial in nature. They produce goods for sale.
As such, data are available regarding total sales and profit, and

the productivity of workers can be determined.

67

The only income ISRA generates to supplement the financial
contributions of the government and donors is obtained through
the sales of vaccines, seeds, livestock and the yields from field
projects. This permits ISRA to cover some, but by no means all,
of the budget shortfall for salaries described in Section 3.2.1.

Thus, there is little difference in salary amongst
researchers at ISRA despite differences in educational status,
seniority, and scientific creativity and productivity. Salaries
are based on seniority and not merit. To be promoted, a
researcher must abandon research for administration. ISRA lacks a

salary structure which encourages a career in research.

3.5.3) Hezk Egvirenment

Many of the elements of a work environment that lead to loss
of researcher motivation and morale have been referred to
previously. To summarize, frequent changes of organization and
management affect the interpersonal relationships within an
institution, the allocation of resources for research, and often
entail changing research priorities and goals. The motivation of
researchers, and hence, the quality of their research suffers as
a result. Shortages of operating funds often hamper and impede
researchers in conducting research. This also reduces the job
satisfaction, motivation levels and the research performance of

researchers.

68

A salary structure which does not foster a cadre of career
researchers, and does not reward a researcher's work undermines
researcher motivation and morale. Limited opportunities to be
recognized professionally or to be rewarded for a job well done
also features of a poor' working environment. Poorly ‘planned
research programs, vaguely defined researcher roles, and lack of
challenging ‘work are other factors ‘which lead. to researcher
frustration and low motivation.

A work environment in which there is a high degree of
turnover often encourages more turnover. Conversations with ISRA
researchers suggest that when associates and colleagues leave an
institution, the morale of remaining researchers is lowered. If
remaining researchers perceive former colleagues as being better
off by leaving the institute, they might anticipate high future
earnings and satisfaction from leaving.

A. work environment and incentive structure that is not
conducive to research leads to poor motivation and
dissatisfaction. in. researchers. This increases the likelihood
that a researcher will leave a research institution. To
underscore this fact, the preliminary results of a 1987 survey of
researchers at ISRA indicate that half of the respondents did not
expect to be working at ISRA in ten years. Over two thirds of the
responding researchers indicated that their earnings and

possibilities for professional advancement were unsatisfactory.

69

3.6) The Impecte of Turnover
3.6.1) Eroeion gr Reseerch Capacity

Most authors agree that high rates of turnover in research
staff decrease an agricultural research institution's capacity to
produce useful research results. Bennell (1986) believes that
turnover in research staff undermines attempts to develop a cadre
of experienced and highly trained researchers. Pinstrup-Anderson
(1982) echoes this belief. In his view, turnover of researchers
leads to continuously low qualifications in research staff. ISNAR
(1984) believes that the leadership, planning, and supervisory
capacities of senior scientists are eroded, or at least strained,
when inexperienced researchers replace departed researchers.
Ruttan (1982) believes that rapid turnover threatens the
viability of research.

As a result of turnover, research activities are often
slowed or disrupted. Some research projects are never completed
when a member of a research team leaves in the middle of a
project, or are substantially delayed while new scientists are
recruited. Shortages of researchers that are caused by turnover
often force the scope of a research program's activities to be
scaled back. Since agricultural research is often a
multidisciplinary effort conducted by teams of researchers,
turnover lowers the productivity of a research program by

disrupting continuity.

70

Thus, rapid rates of turnover in research staff exacerbate
manpower constraints in national agricultural research systems.
Limited capacity to staff research activities often means
research planners must make tradeoffs between national
development goals.

Jahnke et a1. (1987) and ISNAR (1984) believe rapid rates of
researcher turnover severely strains the recruitment process at
agricultural research institutes. In a country like Senegal in
which higher level research skills are obtained through post-
graduate training overseas, this process can be very expensive,
or impossible to accomplish. Since turnover is largely
unanticipated and difficult to predict, high rates of turnover

can seriously complicate research and manpower planning and

management .

3.6.2) Lest Rerume to Inveermenr

Ruttan (1982) believes that rapid turnover compromises an
agricultural research institution's research performance. As a
result, returns to investments in research projects, training
programs, and institution building are reduced.

In a study of turnover in research staff at national
agricultural research institutes in Colombia, Argentina and Peru,
Ardila, Trigo and Piniero (1981) found that the worst impacts of

turnover were felt after training programs were completed. When a

71

flow of replacements was not forthcoming after large-scale,
externally funded training programs were concluded, the "critical
mass" of researchers that had been developed quickly eroded. They
conclude that entire research programs can disintegrate rapidly
if continuing allocation of resources for training and adequate
inducements to retain scientists are not maintained.

The costs of rapid rates of researcher staff turnover are
borne primarily by the national agricultural research institutes
for which the personnel are trained. The consensus is that rapid
turnover of qualified researchers sacrifices the limited
resources of a public research organization, thus compromising
the potential of national agricultural research to contribute to
agricultural development goals.

However, Ruttan (1982) believes there are social benefits
associated with turnover of researchers. If skilled and
experienced researchers leave a national research institute to
accept positions with other in-country agricultural development
projects and organizations (agricultural ministries,
international development agencies, private voluntary
organizations), the human capital of these organizations is often
enhanced.

Thus, Ruttan (1982) believes that investments in national
research capacity have large spillover effects. Turnover

truncates the stream of benefits an agricultural research

72

institute receives from investments designed to improve or expand
its scientific personnel. However, if departed scientists
continue to work in the same country, social benefits continue to
be obtained from these investments since national intellectual
capacity is increased.

From an institutional perspective, Mobley (1982) believes
there are instances in which turnover can be beneficial.
Institutional performance can be enhanced if turnover entails the
displacement of employees with poor performance, creates
promotion opportunities for remaining staff, or if replacements

bring an infusion of new ideas into the institution.

3.7) WW

ove
Costly investments in human capital are undertaken by
individuals and sponsoring organizations since, like investments
in physical capital, more highly skilled individuals are expected
to be more productive. Thus, investments in human capital yield
an increased stream of ‘benefits to an. agricultural research
institution and, by inference, a country's agricultural sector.
The World Bank (1988) estimates that it takes 15 to 20 years
for investments in new staff education and training to bear full
impact. ISNAR (1984) estimates that it requires two to three
years for a researcher to become familiar with a research

problem. Thus, it is expected that the average productivity curve

73

of a researcher slowly increases during the first two to three
years of service, and then rises more rapidly until the fifteenth
year of service. At the fifteenth year the productivity of a
researcher levels off. Furthermore, it is assumed that the higher
one's educational background, the more productive a researcher
will be. Thus, higher levels of education shift the productivity
curve of research scientists upwards.

If a researcher leaves an agricultural research institution,
the stream of benefits produced by investing in a researcher's
level of skills is truncated. By plotting a hypothetical curve of
a researcher's productivity against the number of years a
researcher is actively involved in research, the investment lost
from turnover may be conceptualized. For example, assume a

researcher with the equivalent of a Bachelor's degree is sent

Figure 3.1: A.Conceptual Framework for Measuring Losses to
Investment Resulting from Turnover in Research
Staff.

Increased
Product
from \\
Investment A

 

 

 

Time

74

overseas to receive a Master's degree. The anticipated benefits
from this investment are represented by the area under the
productivity curve in Figure 3.1. The productivity curve
represents the increased research product produced by a
researcher's higher educational qualifications. If a researcher
were to leave a research institution after a given time n, the
shaded area to the right of the line connecting points N and A
represents the lost investment resulting from a researcher
leaving the national research institute.

Although social benefits continue if a researcher is
employed in other research projects in a country after leaving
the agricultural research institute, the costs are likely to be
higher than indicated in Figure 3.1. The departure of a
researcher often lowers the productivity of those researchers who

remain in research programs and projects.

75

§h§n§eI_5: RESEARCH STAFF TURNOVER AT THE SENEGALESE NATIONAL
AGRICULTURAL RESEARCH INSTITUTE (1981-1987): AN
ANALYSIS OF THE PROCESS, CAUSES AND IMPACTS

Inrrodectiog

The Senegalese Agricultural Research Institute (ISRA)
underwent a major expansion in research capacity between 1981 and
1987. In particular, major investments were made in both training
researchers and in increasing their numbers. However, in recent
years ISRA managers and donor representatives have been troubled
by a high degree of turnover in research staff. The problems
underlying the turnover of research scientists at ISRA have never
been thoroughly researched. The goal of this chapter is to
describe the process of research staff turnover' at ISRA, to

identify the factors causing turnover, and to gauge its impacts.

4.1) Migrate

Chapter 3 discussed a number of factors which cause
researchers to leave national agricultural research institutions
for alternative employment opportunities. This discussion
suggests a number of findings that can be anticipated in the

analysis turnover at ISRA.

1) Low relative salaries at ISRA and the lack of a career
incentives in conditions of service are a major contributing
cause to high rates of researcher turnover.

76

Conditions of service at ISRA are believed to be a major

contributing factor in rapid rates of research staff turnover. To
test the influence of conditions of service on research staff
turnover, the following factors will be investigated:
a) Seleryr, Low relative salaries, and declining real salaries
offered by public agricultural research institutes are generally
considered to be a major cause of turnover. This study analyzes
the salary structure for researchers at ISRA, in both nominal and
real terms.

In human capital theory, relative salaries play the crucial
role in an individual's decision to change jobs. Thus, the salary
structure for researchers at ISRA will be compared to the salary
structures of the competing research organizations seeking
skilled agricultural researchers in Senegal. International and
regional research organizations, private voluntary and nonprofit
organizations with development projects in Senegal, and
governmental development agencies are the primary sources of
alternative employment for ISRA researchers.

b) Promorien pelieiee; Conditions of service are evaluated using
an earnings function analysis based on human capital theory and
suggested by ISNAR (Elliott and Kang, 1985). The lack of
incentives to pursue a career in research, and limited rewards
for productivity at ISRA are believed to be further elements

which encourage turnover amongst Senegalese researchers at ISRA.

77

2) Rates of turnover amongst researchers increased after the job
reclassification and the implementation of rigid new salary
provisions by ISRA in 1983.

ISRA tightened and implemented the salary provisions
contained in the enabling legislation which classifies ISRA as a
public establishment with industrial and commercial
characteristics (EPIC) in 1983. Essentially, this action entailed
a restructuring of job classifications and the linking of salary
to job function rather than a researcher's qualifications or
productivity (Faye and Bingen, 1988). This job restructuring
resulted in rigid salary scales and limited promotional prospects
for researchers over the course of their careers. Thus, it is
anticipated that turnover rates increased after 1983 when work

conditions for researchers deteriorated.

3) Younger, more inexperienced researchers are more likely to
leave ISRA than older, more experienced researchers.

Hammermesh and Rees (1984) conclude that a major
implication of human capital theory is that younger employees are
more likely to leave a job than older workers. The reason for
this is threefold. First, younger employees have a longer time
horizon over which to calculate the discounted benefits expected
from alternative employment. Secondly, younger researchers have

lower moving costs since they have not developed as strong

78

psychic ties to their surroundings, or amassed as many personal
belongings as experienced researchers.

Finally, younger, inexperienced researchers have not
received the same degree of on-the-job training as older
researchers. Hammermesh and Rees (1984) believe that most on-
the-job training has both a general and firm-specific content.
Human capital theory predicts that the more firm-specific skills
a worker acquires through work experience, the less likely a that

worker will change jobs.

4) Highly educated researchers are more likely to leave ISRA
than less-well educated researchers.

A number of authors believe that highly productive and
well-qualified researchers are often the most likely to leave
national agricultural research institutes in developing
countries. Data limitations prevent any analysis based on a
scientist's research output. However, a researcher's
qualifications and skills are to a large degree dependent on a
researcher's education. Human capital theory predicts that a
researcher's productivity increases with level of skill. Since
earnings are a function of education and experience, it is
anticipated that the ISRA salary structure which links salary to
job function rather than an individual's skills and productivity

hampers ISRA's ability to retain highly educated researchers.

79

In many developing African countries there is a general
shortage of highly educated and skilled labor. Since such human
capital is in short supply, and demand for such personnel is
high, it is anticipated that researchers with PhD's or equivalent
French degrees are more likely to leave their jobs than Master's

level or Bachelor degree level researchers.

5) Turnover rates among researchers increase when donor projects
are completed.

Low operational support for research and erratically
allocated resources needed to conduct research are commonly cited
reasons for low motivation and high levels of frustration among
researchers. Hence, there is an increased likelihood. that a
researcher will decide to change jobs during periods of low
operational support for research.

Ardila, Trigo and Piniero (1981) and Nestel and Trigo (1984)
both link high rates of turnover in research staffs to decreased
support from donors. The Senegalese government's financial
contribution to ISRA is entirely consumed by salaries. Funds for
ISRA research operations are obtained primarily through donor
agencies or through borrowing. However, donor programs typically
operate within a 4 to 6 year time frame. When donor projects near

completion, operational funds for research become scarce.

80

6) Turnover rates amongst expatriate researchers are higher than
for Senegalese researchers at ISRA.

Expatriate researchers at ISRA play an important role in
implementing donor research projects, and in staffing research
projects for which there are a shortage of Senegalese
researchers. As seen previously, donor projects are often of
relatively short duration. FUrthermore, Bennell (1986) has shown
that increasingly expatriate researchers are hired as short-term
consultants. This trend is most evident at Francophone West
African national research institutes where the experienced cadres
of expatriate researchers that had traditionally been relied upon

for leadership and management roles have largely retired or left.

7) Promoting researchers into research administration positions
undermines the ability of ISRA to develop an experienced,
well-qualified staff of researchers.

Throughout the analysis the promotion of researchers into
research administration positions is considered to be a form of
turnover in researcher staff. As seen in Chapter 3.5, the salary
structure at ISRA rewards administrators at higher rates than
researchers. Chapter 3.5 suggests that more experienced, highly
educated researchers are apt to be promoted into research

administration at ISRA. This practice increases instability in

81

research staffs.

4.2) The Data

The data upon which this study is based consist of yearly
personnel files from ten ISRA research centers: Bambey, Kaolack,
C.D.H., St. Louis, Djibelor, C.N.R.F., L.N.E.R.V., Dahra, Kolda,
and the General Direction. Data are also available for the
Tambacounda research center that was opened in the latter part of
1986. Personnel data were not available for the Oceanographic
research center (CRODT), and thus researchers in the Fisheries
department have been excluded from this analysis.

Complete personnel lists between 1981 and 1987 are
available for only two research. centers-- Bambey' and. Dahra.
Missing personnel information was collected by Michigan State
University faculty members participating in an institution-
building project between M.S.U. and ISRA financed by USAID/Dakar.

Extensive analysis of turnover trends amongst expatriate
researchers was largely precluded by the fact that ISRA does not
compile personnel data for expatriate researchers. Thus, the
available personnel data on expatriates were also compiled by
means of communications with ISRA administrators. However, since
the focus of this study is turnover amongst Senegalese
researchers, it is not believed that the data limitations for

expatriate researchers seriously affect the analysis.

82

Certain anomalies remain in the data. ISRA did not include
age, entry dates at ISRA, and educational data in its personnel
listings prior to 1985. Thus, for researchers who left ISRA prior
to 1986 gaps remain concerning age, education, salary and length
of service variables. Gaps in the data were filled in as
thoroughly as possible through communications with a variety of
individuals familiar with ISRA.

The personnel records that were compiled for this study are
more complete than personnel records kept at ISRA headquarters in
Dakar. Data are available for all years from 1981 until the end
of 1987. Personal communications reveal that two researchers left
ISRA in early 1988. In the analysis of researchers who left
ISRA, these two researchers have been included.

Information is available for the following ‘variables:
nationality, contractual status, job classification, 1987 salary,
research program, education, discipline studied, age, length of
service at ISRA, researcher status (full-time research, research
administration, recruit, or temporary researcher), research
center, and whether a researcher was present at ISRA in a given
year.

Information about job changes was compiled from the
personnel files. However, information about full-time researchers
who studied abroad during the 1981-1987 period. was unevenly

available. Thus, researchers sent abroad in the midst of their

83

careers at ISRA have been considered as full-time ISRA
researchers. Finally, there were at least ten cases of
researchers who were promoted from assistant researcher to
researcher status. In. the turnover statistics and researcher
numbers that follow, assistant researchers turned. researchers
were omitted until they obtained full research status. Length of
service data for salaries was calculated on the basis of the date

of entry listed on ISRA personnel lists.

4.3) We
4.3.1) WW

Numerous studies of the turnover process at firms and
organizations in Ideveloped. countries are available. The 'vast
majority of these studies have compared the responses to surveys
administered to "leavers" and ”stayers". These surveys have
attempted to explain the causes of turnover by measuring
variables such as job satisfaction, job commitment, and
behavioral intentions to leave an institution (McEvoy and Cascio,
1985; Mobley, 1982). Data limitations prevented any such analysis
in this study.

Recently, a number of authors have developed models to
measure the impacts of turnover on institutional performance.
Underlying these studies is the belief that there may be

beneficial aspects to turnover. These studies typically include

84

both attitudinal variables and variables which measure the output
of employees (see Boudreau and. Berger, 1985; Hollenbeck. and
Williams, 1986). Given the lack of data about ISRA researchers'
output, such as publications, this study could not model the
impacts of turnover on research performance at ISRA.

Thus, this study measures turnover by calculating turnover
frequencies and attempts to correlate turnover with individual
and organizational variables that are contained in. the ISRA
personnel files.

Yearly turnover rates were calculated as follows. Based on
the yearly personnel records, researchers were coded as "1" if
they were employed at ISRA in a given year, and as "0" if their
names were not listed in the personnel files for a given year. In
this manner a record of ISRA researchers was established for each
year between 1981 and 1987.

Yearly turnover rates were calculated by first selecting
those researchers wooded as "l” in. a ‘base year n, and then
selecting those researchers coded as "1" in both the base year n
and n + 1. After selecting those researchers who were present in
consecutive years, turnover rates were calculated. using the

following formula (Idachaba, 1985):

 

 

Yearly (Researchers coded __ (Researchers coded
turnover - "l" in yeere n. n+1) ”1” in year n) X 100.
rate (Researchers coded "l"

in year n)

85

The formula was expanded to measure turnover rates over
longer periods of time by selecting researchers who were present
in year n, and years n+1, n+2,...,n+i. This turnover formula was
selected because it is by far the most widely used method of
calculating turnover (Mobley, 1982). Most available measures of
turnover of researchers at national agricultural research
institutes are stated in terms of annual turnover rates (see
Ardila, Trigo, and Piniero, 1981). Thus, annual turnover rates
may be compared between different agricultural research
institutions. Turnover rates were also calculated on the basis

of nationality and research department.

4.3.2) Qgerell Iurnever getee (1281 - 1982)

Of the 227 researchers who were employed at ISRA between
1981 and 1987, 100 (44.1 per cent) left ISRA. Fifty seven of the
90 expatriates researchers (63.3 per cent) at ISRA during the
sample period left the institution. Forty three (31.4 per cent)
of the 137 Senegalese researchers in the population left.

Table 4.1 and Figure 4.1 show the annual turnover rates for
all researchers on a yearly basis between 1981 ~1987. Turnover in
research staff at ISRA increased in 1985 and 1986 relative to the
1981 - 1984 period. The number of researchers who left ISRA
almost doubled beWeen 1984 and 1985. In 1987, the yearly

turnover rate soared dramatically. More than a quarter of the

86

Table 4.1: Turnover Rates for All ISRA Researchers (1981 - 1987)

 

Percent of Leaving

 

e a e
Annual
Year Number of Expatriates Senegalese Turnover
Researchers (% of total) (% of total) Rate
Who Left
81-82 7 71.4 28.6 6.2
82-83 4 75.0 25.0 3.5
83-84 8 75.0 25.0 5.9
84-85 15 60.0 40.0 9.3
85-86 15 46.7 53.3 8.5
86-87 48 56.3 43.7 28L1

 

Seeree: ISRA.Personnel Files

researchers employed by ISRA in 1986 left by the end of 1987.
Although expatriates form the majority of those researchers
who left ISRA between 1985 and 1987, Table 4.1 indicates that
since 1985 turnover increasingly has become a Senegalese problem.
ISRA.has little say in the decisions of expatriate researchers to
leave ISRA, since these researchers are attached to ISRA under
the auspices of various donor agencies. However, the increasing
proportion of' Senegalese researchers leaving ISRA. since 1985
signal increasing problems in fISRA's ability to retain it's

research personnel.

87

 

 

Turnover Rates:
Puree n1:

Figure 4.1: Yearly Turnover Rates
All lSRA Researchers
1981 — 1987

30 LEGEND

_ All Researchers

 

f-D
[.11

r.)
<3-

—-J-
(J1

—_A.
C."
§_

 

 

 

81-2 ' 83-4 85-6
82-3 84-5 86-7
“I ears

Source: ISRA Personnel F1188

 

 

88

4.3.3) Yearly IBIBQVGE Reree hx Nariohality 9f Researchers

The annual turnover rates among expatriate and Senegalese
sub«groups of the researcher population support the conclusion
that turnover in Senegalese research staff increased after 1984.
Figure 4.2 and Table 4.2 show the trends in researcher turnover
rates among expatriate and Senegalese researchers. With the
exception of 1982, the turnover rates for expatriate researchers
remained relatively’ constant ‘between. 1981 and 1986. However,
turnover rates between 11 and 15 per cent are indicative of a
high degree of instability. In 1987, nearly half of the
expatriate researchers present at ISRA in 1986 left ISRA.

The Special Program on African Agricultural Research (SPAAR)
considers yearly turnover rates of 3-4 per cent for researcher
staffs to be the norm at agricultural research institutions in
developing countries (World Bank, 1987b). SPAAR infers that these
rates are acceptable target levels. The turnover rates for
Senegalese researchers ranged between 1.4 and 2.8 per cent
between 1982 and 1984. This indicates that turnover was in an
acceptable range at ISRA during the early stages of this study.

Beginning in 1984-85 turnover rates among Senegalese
researchers began to rise, before exploding in 1986-87. SPAAR
considers turnover rates in excess of 7 per cent to be normal for
researchers at African national agricultural research

institutions, inferring that these rates are indicative of a high

 

89

Table 4.2: Yearly Turnover Rates Amongst Expatriate and
Senegalese Researchers at ISRA (1981 - 1987).

 

 

 

 

What: 8 esea e 5
Year Number Who Turnover Year Number Who Turnover
Left Rats Left Rate——
81-82 5 of 42 11.9 81-82 2 of 71 2.8
82-83 3 of 45 6.7 82-83 1 of 70 1.4
83-84 6 of 51 11.8_ 83-84 2 of 85 2.4
84-85 9 of 60 15.0 84-85 6 of 102 5.9
85-86 7 of 60 11.7 85-86 8 of 116 6.9
86-87 27 of 55 49.1 86-87 21 of 116 18.1

 

Seurce: ISRA Personnel Files.

degree of instability within research staffs. ISNAR (1984)
considers annual turnover rates of 10 per cent to be common in
Africa. Turnover rates among Senegalese researchers at ISRA fall
below the range SPAAR considers to be a normal in Africa prior to
1985, and fall within the normal African range during 1985 and
1986. The turnover rate at ISRA in 1987 exceeds the estimate of
typical turnover rates in Africa. Turnover at ISRA in 1987 was a
major problem.

Bennell (1986) interprets turnover rates as follows: a

turnover rate of 5 per cent a year means that a national research

90

 

 

 

 

'T u I" :1 over I‘ Ra ta
P-—+r c: e nt

Figure 4.2: Yearly Turnover Rates

Turnover Rates by Nationality
1981 - 1987

 

"It

a. it LEGEND

45 .. _ Expatriates
41:! r __ Senegatese

 

 

\ T /

‘v’fi

 

 

"fear

Source: iSRA Personnel Fries

 

 

91

institute must replace half its research staff in 10 years. If
the 1987 turnover rate at ISRA continued at 18.1 per cent for 5
1/2 years, the entire Senegalese research staff would have to be
replaced. The dramatic increase in turnover among Senegalese
researchers in 1987 parallels the dramatic increase in turnover

among expatriate researchers at ISRA.

4.3.4) Turnover Retee by Beeeareh Depertmehr

Table 4.3 and Figure 4.3 show yearly turnover rates on the
basis of research department. Some of the turnover rates in the
table are deceptive. The number of expatriates in the Animal
Health and Production Department was less than 10 between 1981
and 1987. On average, one expatriate left the department each
year. The Forestry Department statistics are also misleading. In
1986, there were only 4 expatriate and 14 Senegalese researchers.
One expatriate researcher left Forestry Department in 1984, 1985,
and 1986.

In 1987, the extremely high turnover rates for expatriate
researchers of 52.2 and 68.4 per cent in the Crops and Systems
Departments, respectively, reflect the fact that a number of
donor projects were terminated. The World Bank's Agricultural
Research Project (PRA) financial contribution to ISRA began to
gradually decline in 1986. This decline in. support gathered

momentum as the 1988 completion date of the project grew closer.

92

Table 4.3: Yearly Turnover Rates in ISRA Research Departments

 

a) Crops Deperrmehr
1982 1983 1984 1985 1986 1987

Expatriates: 14.3 14.3 18.2 20.0 8.3 52.2
Wifififlfliflm
Overall: 8.5 8.7 8.0 10.0 6.3 29.2

b) Systeme Departmehr

Expatriates: 0.0 0.0 0.0 10.5 14.3 68.4
Sehegeleee; ,9 0,9 4,8 2,4 9,4 12y;
Overall: 0.0 0.0 2.7 8.7 11.3 33.3

c) n ma He th d od on
DQRQEEQEDE

Expatriates: 11.1 0.0 11.1 10.0 11.1 25.0
Senegalese; 0,0 Q,Q Q,Q 13,1 5,9 llai

Overall: 3.6 0.0 3.3 12.1 5.9 15.6
d) Eorestry erartmenr

Expatriates: 0.0 0.0 25.0 20.0 20.0 0.0
Sehegalese; _QrQ, _QiQ _QiQ _Q19 _QLQ 1319
Overall: 0.0 0.0 7.7 5.9 5.3 26.3

 

Sohree: ISRA Personnel Files

In the Crops department, a Belgian horticultural project under
the auspices of the Food and Agriculture Organization (FAO) ended
in 1986, as did an ICRISAT project. In the Systems Department,

the first phase of a USAID project ended in December, 1986.

93

Funding continued at a reduced rate in 1987 while the second
phase of the project was being renegotiated.

The most striking feature of Table 4.3 is that turnover of
Senegalese researchers increased dramatically in all research
departments in 1987. This indicates that the problems experienced
at ISRA in 1987 were widespread. Turnover in the Systems
Department has shown a steady increase since 1984. The Crops
Department experienced relatively stable turnover rates for
Senegalese researchers until 1987, when turnover became a major
problem. In the Animal Health and Production Department, turnover
was low amongst Senegalese researchers with the exception of
1985, and 1987.

The Forestry Department characteristically has a very
youthful group of Senegalese researchers. In 1981, there were
only two Senegalese Researchers in the Department. Of the five
researchers who left in 1987, two died and one was a temporary

researcher.,

4.3.5) c nst ' Inte Transfers Pr motion
ov ned

The promotion of researchers into administrative positions,
and movements of researchers within a research institution, like
turnover, disrupt the continuity: of research programs. Table 4.4
shows the total annual movement in Senegalese research staff

between 1981 and 1987.

 

 

 

94

 

Figure 4.3: Yearly Turnover Rates
Turnover Rates by Research Department

1981 — 1987

 

VA Crops Dept.
- Systems Dept.
'0 Livestock Dept.

LEGEND

 

 

 

Forestry Dept.

 

 

H§-~‘§\\\§‘§‘§S‘§§§‘§S§

"gl’agggifiggddglfififig

S‘g
I!“

Ill/glut...

gs:

Vgglllidfld
V§\.§
El,

girl/lt’g

'r.‘
7”””””;

 

-
4i

5 0 Rd 0 5 0

2 “I. all 1....
#C mot...m.u&
WQHUK L®>OCL3F

8.3-4 85-6

81-2

84-5 85-7

year

82-3

Source: lSRA Personnel Files

 

 

95

In 1986 and especially in 1987, movement of researchers
within ISRA increased dramatically. Turnover in research staff
was likely to have been a contributing cause to the increased
movement of researchers between research centers. As seen
previously, turnover in Senegalese research staff began to
increase in 1985. Other factors that may have influenced the
internal movement of researchers at ISRA include donor projects
which began to wind down and end in 1986 and 1987, and the
institutional reorganization in 1987.

The promotion of researchers into administrative positions
appears to have added an element of instability at ISRA

throughout the period under study. As seen in Chapter 3,

Table 4.4: Instability in Senegalese Research Staff at ISRA
(1981 - 1987).

 

 

Year Changed Promoted Left ISRA Total
Research into Research
Center Adninietration
1982 2 (2.9%) 2 (2.9%) 2 (2.9%) 6 (9.6%)
1983 0 l (1.4%) 1 (1.4%) 2 (2.8%)
1984 l (1.2%) 7 (8.2%) 2 (2.4%) 10 (11.8%)
1985 l (1.0%) 2 (1.9%) 6 (5.9%) 9 (8.8%)
1986 4 (3.5%) 3 (2.7%) 8 (6.9%) 15 (13.1%)
1987 13 (13.3%) 4 (3.5%) 21 (18.1%) 38 (34.9%)

 

Seurce: ISRA Personnel Files

 

96

researchers rarely, if ever, have formal training in management
and administration, largely acquiring these skills while on the
job. Thus, the steady flow of researchers into research
administration and management likely constrained ISRA's capacity
to perform management functions throughout the 1981 - 1987
period. The high percentage of researchers promoted into
administrative positions in 1984 reflects ISRA's structural
reorganization in 1983.

Rapid expansion of research staff between 1982 and 1986 is
likely to have further burdened ISRA's research administration
and management capacities. At the same time that researchers were
assuming research and administration positions, demands on
research administrators and managers were increasing as a result
of the expansion.

Finally, Table 4.4 shows that turnover of research personnel
only partially explains the tremendous degree of instability that
ISRA experienced in 1987. 34.9 per cent of the Senegalese
researchers present at ISRA in 1986 left the institution or
experienced a change in their research roles. When expatriates
are included in the analysis, 65 researchers either left ISRA,
changed research centers or were promoted into research
administration in 1987. This represents a severe disruption in
the continuity of research projects and research personnel that

is neccessary for effective research. The productivity and

 

 

97

effectiveness of agricultural research in Senegal appears to have

been seriously undermined in recent years.

4.4) Sheraereristice 9f Reeearehere flho Lefr ISRA
4.4.1) Status of Researehere She Left

Senegalese researchers at ISRA between 1981 and 1987 were
classified as full-time researchers, research administrators,
trainees, and temporary researchers. Full-time researchers refer
to the majority' of researchers at ISRA. who ‘have officially
attained researcher status, and spend the majority of their time
actively participating in research. Researchers classified as
research administrators refer to the General Director, research
department directors, and research center directors who have been
promoted into these positions from research. Research
administrators spend the majority of their time doing 'non-
research functions. Trainees refer to recruits to ISRA who never
were granted full researcher status and left during their
probation period. As seen in Chapter 2.4.7, all researchers must
fulfill a number of requirements before attaining full researcher
status. The probation period lasts up to two years.

Temporary researchers refer to researchers who have signed
short-term contracts of less than a year. They have been included
in the analysis since they are classified as researchers in the

personnel records. Also, personal communications with ISRA

 

 

 

98

researchers indicate that a possibility often exists that
temporary researchers will be retained as full-time researchers.
For this reason, and since temporary researchers fill research
functions, they have been included as researchers in the turnover
statistics. Their inclusion inflates turnover statistics to a
degree. It should be noted, however, that only one temporary
researcher left ISRA in 1986 and 1987.

Table 4.5 shows the breakdown of those researchers who left
ISRA classified on the basis of their research status. Nearly a
quarter of those researchers who left ISRA had trainee status.
This in large part reveals weaknesses in ISRA's recruitment
practices. As seen in Chapter 2, the recruitment of ISRA
researchers is often based upon the availability of scholarships
and donor funds rather than detailed manpower plans. As such, a
well-defined research role for recruits returning from overseas
study is often lacking (Faye and Bingen, 1988). Frustrated by
long time delays, recruits often leave ISRA before becoming
actively involved in research.

The large proportion of trainees who left ISRA between 1981
and 1987 represent a large degree of lost investment in human
capital for ISRA and the donor agencies which have invested in
educating researchers. The benefits anticipated from investments
in education are lost to ISRA when recruits leave before fully

becoming involved in research projects. In ten cases, the costs

 

"n

 

99

Table 4.5: Characteristics of Leavers: Research Status

 

 

 

 

Status Ember LL—Jse centa
Full-time Researcher 16 37.2%
Research Administration 13 30.2%
Trainees 10 23.3%
Temporary 4 9 ' 3%
Total 43 100.0%
Status* Total Number Percent Who
111W Left——
Full-time Researcher 93 17.2%
Research Administration 29 44.8%

 

* Trainees and temporaries not included since these labels are
applicable only to those who left. Trainees and temporaries who
did not leave attained full-time status.

 

Sonrce: ISRA Personnel Files.

of educating trainees were born, yet benefits not received.

The number of trainees no doubt includes some researchers
who were unqualified to be researchers, although there is no way
to measure the degree to which this is true. However, selecting
unqualified recruits for overseas training also would indicate
weaknesses in ISRA's recruiting practices. Presumably, weak

research candidates should be identified prior to receiving

 

 

 

100

overseas training .

Of the 10 trainees who left, 7 left in 1986 or 1987. Thus,
it appears that turnover of trainees is in part associated with
the ending of donor projects and the cutbacks in the Senegalese
government's financial support to ISRA during the same period. As
a result of these actions, research activities slowed and
operational funds for conducting research decreased. Limited
research resources and research roles were thus available for
trainees. Given low productivity and vaguely defined roles, it is
believed that a number of trainees left ISRA as a result of
frustration.

Underlying the turnover of trainees is the ad hoc nature of
personnel planning at ISRA. At the same time that new researchers
were arriving at ISRA, research funds and support for research
activities were declining. Poorly coordinated manpower planning,
and the lack of long-term manpower planning within resource
constraints appear to be major causes of turnover among research
trainees.

Table 4.5 also reveals that a large proportion of research
administrators left ISRA between 1981 and 1987, a period marked
by a large degree of structural instability (see Section 2.3).
These results help to explain the generally acknowledged weakness
in management at ISRA. Research administrators represent the

personnel who plan and manage research programs and projects, as

fl

 

 

101

well as allocate resources to research. High rates of turnover
amongst research administrators indicate a high degree of
organizational instability and inefficiency.

There are 15 key research administrators at ISRA (excluding
the Fisheries Department): the Director General, Assistant
Director General, 4 Research Department Directors, and 9 Research
Center Directors. Between 1981 and 1987, ISRA had three different
Director Generals and two Assistant Director Generals. There were
three Directors at the Animal Health and Production and Forestry
Departments, and two different Directors of the Crops and Systems
Departments. Among research centers only Dj ibelor, which had two
Center Directors, did not have three Research Center Directors.

The impacts of this high instability in research
administration and management staff at ISRA have far-reaching
impacts upon research at ISRA. Research administrators are
selected from the ranks of experienced researchers at ISRA. Thus,
if a research administrator leaves ISRA, he or she is replaced by
a researcher who previously was involved in research activities.
High turnover amongst research administrators thus not only
constrains ISRA's research management and administrative
capabilites, but also undermines efforts to assemble an
experienced group of researchers. Turnover of research
administrators can be seen to send a ripple effect of instability

throughout a research department.

102

Research administrators represent the primary source of
turnover in the early years of this study. Five of the 11
researchers who left ISRA prior to 1985 were classified as
research administrators. By contrast, of the 16 full-time
researchers who left ISRA, only two left prior to 1985. These two
researchers left in 1984. Thus, turnover of full-time research
staff is a relatively new phenomenon at ISRA. The majority of
those Senegalese researchers who left ISRA in 1986 and 1987 were
either full-time researchers or trainees. As such, turnover in
these years was particularly disruptive, and represents a major

loss of benefits anticipated from investment in human capital.

4.4.2) Eeneerignel Eve]. and Sneeielizetign._of.figmmhm___Who
Left

The task of trying to measure the impacts of turnover in
terms of educational qualifications was complicated by
differences between the French and American educational systems.
This study assumes that the French Ingenieur degree and Mr
Vererineire degrees are comparable to receiving a four year
Bachelor of Science (BSc) degree in the American system since
these degrees represent limited degrees of post-graduate
specialization.

The French WW (DEA) is assumed to
be equivalent to a Master of Science (MSc) degree in the American

educational system, since both involve intermediate levels of

 

103

Table 4.6: Education of Researchers Who left ISRA Between
1981 and 1987.

 

a) BSc Equivalent:

9.6.212: AILW MW eft LLsLt
Ingenieur 27 8 29.6%
D V e e l: S 40,03
subtotal: 42 14 33.3%

b) MSc Equivalent:

 

Master's 37 14 37.8%
DEL _2.L _L M
Subtotal: 66 20 30.3%

c) PhD Equivalent:

3eme Cycle 15 6 40.0%
W42— __1_ 4.33.
Subtotal: 27 7 25.9%
TOTAL: 135 41 30.4%

 

Seuree: ISRA Personnel Files.

post-graduate specialization. Finally, the French Doerorer Q'Erar
degree and Sene_§ye1e awards are assumed to be the equivalent of

an American PhD since they represent advanced levels of post
graduate specialization” To avoid. confusion, both French and
American diplomas have been included in Table 4.6, but have been

grouped according to the previously mentioned assumptions.

104

Of the four temporary researchers who left, two had
Master's Degrees and two had lngenienr degrees. Eight of the ten
trainees who left had Master's degrees. This in part reflects
the fact that American trained researchers have English speaking
skills that are much in demand in a French speaking country.

A worrisome trend at ISRA is that the seven researchers with
advanced, PhD equivalent: degrees left ISRA. in '1986-87. Thus,
turnover amongst the most qualified researchers at ISRA is a
relatively recent phenomenon.

There is no clear trend to research staff turnover on the
basis of educational background. Table 4.7 compares the
proportion of researchers who left ISRA with the entire

population of Senegalese researchers -- grouped by degree. The

Table 4.7: Proportion of Researchers Who Left vs. Proportion of
Researchers in the Population: By Degree Group.

 

 

 

Degree Group Proportion of , Proportion of
Leavers With Degree Holders In
ML WM

BSc Equivalent 34.1% 31.1%

MSc Equivalent 48.8% 48.9%

PhD Equivalent 17.1% i 20.0%

missing -2

 

Senree: ISRA Personnel Files.

 

105

results in Table 4.7 are statistically insignificant, thus
indicating that more highly educated researchers are not leaving

ISRA at disproportionally high rates.

4.41.3)me

The relationship between turnover and experience was
statistically insignificant. Table 4.8 shows that when
researchers are grouped according to their levels of experience,
the researchers in different experience categories left in
proportion to the researcher population at large.

Screening trainee and temporary researchers from the
analysis did not significantly change the results. Experience
alone, as with educational background, did not seem to affect a

researcher's decision to leave ISRA. When considering only the

Table 4.8: Turnover Rates for Senegalese Researchers
(1981 - 1987): Experience of Leavers

 

Experience Proportion of Proportion in
Leevere (n—SS) Eonulation (n-lSZ)

0 to 5 years 18 (47.4%) 59 (44.7%)

6 through 10 12 (31.6%) 45 (34.1%)

11 or more 8 (21.1%) 28 (21.2%)

Source: ISRA Personnel Files.

106

education or experience level of ISRA researchers, those who left
did so in direct proportion to the entire researcher population.
Of the 16 full-time researchers who left ISRA, only one had
11 or more years of experience. Six (40.0 per cent) had 0 to 5
years of experience. Eight (53.3 per cent) had 6 through 10 years
of experience. If one assumes that it takes a researcher 15 years
to reach peak productivity and also a couple years to become
productive in research, .the turnover of full-time researchers

represents a loss of benefits to investments in human capital.

4.4.4) Wares

There is a tendency at ISRA to promote experienced
researchers into research administration. positions. Table +4.9
shows that between 1981 and 1987, the majority of researchers at
ISRA with 11 or more years of experience have been promoted into
research administration positions.

Of the 28 researchers with 11 or more years of experience at
ISRA 8 left. Of these 8 researchers who left, 7 had research
administration status, and only one was a full-time researcher.
There seems to be a trend to promote experienced research
personnel into research administration posts from which
researchers leave ISRA. In 1987, of 20 researchers with 11 or
more years of experience, eight (40 per cent) were employed in

research administration positions, and twelve (60 per cent) were

.4

107

Table 4.9: Relationship of Researcher Experience and Status at
ISRA (1981 - 1987).

 

Experience of Number Engaged in Number in Research
W W min—M t tio
0 - 5 years 58 1 (1.7%)
6 - 10 years 45 8 (17.8%)

11 years or
more 28 16 (57.1%)

 

Senree: ISRA Personnel Files.

employed as full time researchers.

Although experience at ISRA seems to be the primary
criterion in selecting researchers for administrative positions,
it appears that highly qualified researchers were promoted into
administrative positions more quickly than less well qualified
researchers. The mean experience of research administrators with
BBC and MSc equivalent degrees was 12.6 and 13.1 years
respectively. The mean experience of research administrators with
PhD equivalent degrees was only 10.1 years. These differences are
statistically significant at the 99 per cent level. The majority
of researchers with BSc. and MSc. equivalent degrees with over 11

or more years of experience were/promoted into administration.

 

108

4.4.5) WW

Table 4.10 shows that social scientists left ISRA in higher
proportions than other researchers. Of the 13 sociologists and
economists who left ISRA, four had trainee status. Three of the
economists had temporary status. One reason for the high turnover
of social scientists is that training in social sciences has a
content which is less-specific to agricultural research than
training in agricultural and animal sciences. Thus, the skills of
social scientists are more highly marketable than other
disciplines.

In addition to the five agronomists who left ISRA, four were
promoted into research administration positions. Thus, of 12
agronomists at ISRA in 1987, only eight were employed as full-
time researchers. Faye and Bingen (1988) indicate that ISRA has a
shortage of researchers with generalized training, such as
agronomists. The analysis indicates that turnover and promotion
into administration have contributed to this shortage.

Three of the four plant pathologists who left had degrees in
entomology. Turnover among researchers with degrees in
agricultural research-specific disciplines (animal pathology and
genetics, plant pathology and genetics, basic sciences and soil
science) exhibit comparatively low rates proportions of
researcher turnover. Of 52 researchers holding such degrees, 10

( 19.2 per cent) left ISRA. If research administrators are

 

109

Table 4.10: Turnover in Senegalese Research Staff (1981-87):
Specialization of Researcher

 

W WWLLI—et

a) Social Sciences:

 

 

 

 

 

 

 

Economics 16 7 44.0%
Sociology 7 6 85.7%
Geo graphv _1_ 1 M
Subtotal: 24 14 58.3%
Excluding

temporaries: 21 11 52.4%
b) Plant Sciences:

Agronomy 19 5 26.3%
Pathology 13 4 30.8%
Basic Sciences 6 1 16.7%
Plant Breeding 10 2 20.0%
Soil Science 11 l 9.1%
cher; _3_ 1 .1143
Subtotal: 62 14 22.6%
c) Animal Sciences:

Animal Science 11 3 27.3%
Veterinary 11 3 27.3%
Pathology 5 1 20.0%
Maegan: 4 _L 11.93
Subtotal: 31 8 25.8%
d) Forest Sciences: 6 2 33.0%
e) Ecology: 10 3 33.0%
f) Computer Science/ 2 1 50.0%

Biometrics

 

Seuree: ISRA Personnel Files.

110

excluded, only 5 of 47 (10.6 per cent) researchers with

agricultural research specific skills left ISRA.

4.4.6) W

There does not appear to be a pattern to turnover among
Senegalese researchers on the basis of research program.
Information on research programs in the ISRA personnel files was
often incomplete or ambiguous. Nine researchers in uniquely
administrative positions were excluded from the analysis on
research programs on the assumption that they were not actively
involved in research. Given the gaps in the personnel files,
specific research program information was available for only 25
of the 34 remaining ISRA researchers who left ISRA between 1981
and 1987.

Twenty different research 'programs 'were affected. by the
departures of these 25 researchers. Two or more researchers left
only four research programs-- one in the Animal Health and
Production Department, and three in the Systems Department.

Due to the limited data on research programs, any
conclusions must be treated with a degree of caution. However,
there are strong indications that turnover in Senegalese research
staff has been a phenomenon that was generalized across research
programs. Nearly half (20) of the 43 total research programs at

ISRA are known to have experienced the loss of at least one

 

111

researcher between 1981 and 1987. This suggests that the effects
of turnover in Senegalese research staff at ISRA have been felt
throughout the institution. Although program data on expatriate
researchers is very incomplete, the high rates of turnover in
expatriate researchers likely compounded the disruptive impacts
that the turnover of Senegalese researchers had upon research
programs.

Given ISRA's limited numbers of qualified and experienced
researchers, turnover appears to have limited the productivity
and efficiency of numerous research projects. Turnover in
research staff, and the resulting strain on research personnel
remaining in research programs is very likely to have diluted
ISRA's overall research performance. Additionally, the high
degree of turnover amongst research managers and administrators
most likely compounded the effects of the turnover in research
staff by constraining and undermining institutional capacity to

plan and allocate resources for research.

4.5) ISRA Selery Srrncture
4.5.1) Qendifione ef Serviee

Human capital theory considers investment in human capital
to be equivalent to investment in physical capital (Hammermesh
and Rees, 1984). That is, individuals, sponsors, or organizations

invest in formal education and on- the- job training in

112

anticipation of receiving increased benefits in the future.
Higher level of skills, and hence higher levels of productivity,
are the benefits of investment in human capital. However,
individuals or sponsors of individuals incur costs in training
workers. These costs include income forgone for individuals
during training, forgone productivity of the workers receiving
training fer the institution, and the direct costs of tuition,
transportation and training.

Sponsors are willing to incur the costs of investing in
human capital in anticipation of a worker producing a greater
stream of benefits resulting from more productive, and highly
skilled workers. Individuals are willing to incur the costs of
training and formal education in anticipation of receiving a
greater stream of earnings as a result of their increased
productivity. Educated labor in this respect is an economic good
that is scarce. Thus, the supply of educated labor is responsive
to price (Hammermesh and Rees, 1984). Higher earnings offered to
workers possessing higher order skills induce an individual to
incur the costs of fermal education and training. The
effectiveness of agricultural research is highly dependent upon a
highly educated and skilled staff of researchers.

Skills are also acquired through experience on a job. On-
the-job training is generally informal. However, on-the-job

training involves costs to an employer in the form of lost time,

 

113

lowered output of individuals during training, and resources used
during the training of workers. Investment in on-the-job training
also yields higher benefits to an employer in the form of more
productive researchers.

In human capital theory earnings are considered to be a
function of an individual's formal educational background and
experience. Since wages are determined by the marginal value
product of a worker, the key feature in the theory is that
investment in human capital yields higher earnings. These
earnings are believed to decrease later in a worker's career for
two reasons. First, there are diminishing returns to formal
education over time. Secondly, a worker usually receives less on-
the-job training towards the end of his career. As an employee
nears retirement, the time horizon over which institutions
receive benefits from investing in on-the-job training is
shortened.

Elliott and Kang (1985) use an earnings function analysis
based upon personnel files to describe the conditions of service
at national agricultural research institutions. In their model,
the amount of human capital possessed by an individual is a
function an individual's formal education and experience on the
job. Education is proxied. by' dummy ‘variables indicating the
highest degree possessed by a researcher. Years of experience

(LOS) are calculated on the basis of the date entry of employees

114

at the institution.

A logged relationship is expected between salary and length
of service. That is, the shape of the reward structure is
expected to rise rapidly in early years and to peak after 15 to
20 years of service (see Figure 3.1). The rapid rise in salary
during early years serves to attract individuals to an
institution, and to retain them at the institution while they
refine their research skills and become more productive. Thus,
the natural log of monthly salary (LNSAL) was calculated for all
ISRA researchers. The rate of salary increase is expected to
decline late in a researcher's career to reflect declining
investment in a researcher's human capital. To capture the
peaking and declining nature of expected salaries in later years,
the length of service variable was squared (LOSSQ). A negative
coefficient for the LOSSQ variable is expected.

Educational background which reflects higher skill levels
and institutional factors such as promotion into research
administration are other factors which influence earnings. Higher
degrees are believed to yield a higher stream of earnings than
lower degrees. That is, the salary curve for PhD holders (DPHD)
is expected to be higher than the curve for MSc holders (DMSC).
BSc holders (DBSC) are believed to have the lowest curves.

To see the degree to which promotion into research

administration (DADMN) or research management (DDIR) affects

 

115

salaries, dummy variables were introduced into the model.
Thus, the earnings function multiple regression equation
used to analyze conditions of service for researchers at ISRA is

as follows:

LNSAL - a + b LOS - c LOSSQ + d DBSc + e DMSc + f DPhD + g DDIR
+ h DADMN + residual

WHERE:

LNSAL - natural log of Monthly Salary (FCFA)

LOS - length of employment at ISRA

LOSSQ - length of employment at ISRA squared

DBSc - dummy variable-- 1 if highest educational degree is
equivalent to Bachelor's degree (no post-graduate study)

DMSC - dummy variable-- 1 if highest degree attained is

equivalent to Master's training (intermediate graduate
study)

DPHD - dummy variable-- 1 if highest degree attained is
equivalent to PhD training (advanced graduate study)

DDIR - dummy variable-- 1 if director of research center or
director general

DADMN - dummy variable-- 1 if researcher involved in
administration of research, or research department
director.

4.5.2) t s Fu c s

In Table 4.11, the coefficient of the length of service (LOS)
variable indicates that a 3.8 per cent increase in a researcher's
salary is associated with an added year of experience. As
expected, the LOSSQ (length of Service Squared) variable has a
negative coefficient. However, the coefficient for LOSSQ is very
small. An increase of one unit of squared experience decreases

the rate of salary increase by only 0.1 per cent. The correlation

116

between log of salary (LNSAL) and length of service (LOS)
variables is 0.85.

Education does not play a significant role in determining a
researcher's income. None of the education variables were
statistically significant. The negative coefficients of the BSc
and MSc variables reflect in part the fact that the majority of
contracted researchers with these degrees have relatively low

levels of experience.

TABLE 4.11: Results of Earnings FUnction.Analysis: Determinants
of Researcher Earnings at ISRA, 1987 (n-81).

 

 

 

yarigblg Co-gfficign; I-Statistic
CONSTANT 12.143 . 251.18 **
L08 0.038 5.24 **
LOSSQ -0.001 -2.25 *
DDIR 0.189 6.09 **
DADMN 0.065 2.80 **
DBSc -0.008 -0.17
DMSc -0.040 . -O.87
DPhD 0.019 0.39

E2 - 0.82

** - denotes statistical significance of t-statistic at 99% level
* - denotes statistical significance of t-statistic at 95% level

 

Source: ISRA Personnel Files.

117

Significant increases in salary are associated with
promotion into research management. An 18.9 per cent salary
increase is associated with becoming a director of a research
center, or being promoted into top management. A salary increase
of 6.5 per cent is associated with promotion into research
administrative posts such as director of a research department.

The R2 of .82 indicates that 82 per cent of the variation of
the dependent variable, log of monthly salary, is associated with
the independent variables in the equation. Some of the
unexplained variation in salary is believed to result from
changes in ISRA's salary structure in 1983.

Salaries reported in the personnel files exhibited a fair
degree of variation prior to the job reclassification and the
tightening of EPIC salary provisions in 1983. The salaries of
many researchers were higher than those imposed by the 1983
salary regulations. Rather than out these researchers' salaries,
their previous base salaries were maintained.

Overall, the results of Table 4.11 show that the main
determinants of earnings at ISRA are length of service and
promotion into research administration or upper management
positions. If education variables do in fact proxy differences in
levels of research productivity, Table 4.11 reveals that
productivity of researchers, or output, is not rewarded in ISRA's

salary structure. A highly productive researcher receives the

118

same salary as an unproductive researcher with similar experience
and classification.

Combined with limited and uneven opportunities to achieve
professional recognition at ISRA (see Chapter 3.5.2), it appears
that the reward structure at ISRA contains disincentives for some
researchers to do research. Length of service bonuses are assured
to researchers who stay at ISRA, regardless of the quality of
research they produce. To significantly increase earnings,
researchers must be promoted into research administration.

'Underlying dhe salary' structure at ISRA. are EPIC salary
provisions. Since salaries are linked to job classifications and
seniority, researchers with similar classifications and
experience at ISRA receive the same salaries regardless of
differences in education or previous work experience. Faye and
Bingen (1988) report that under existing salary policy a
researcher named director of a research center but having limited
formal education and experience could, in theory, receive a much
higher salary than a highly experienced and educated researcher
in a research position.

Finally, in describing the EPIC salary regulations in
Chapter 3.5.2, it was indicated that there are three job
classifications applicable to researchers: 6-1, 6-2, and 6-3.
Changes in job classification also seem to be associated with

length of service. The mean length of service for researchers

119

with 6-1 classification was 4.5 years. For researchers with 6-2,
and 6-3 classifications, the mean length of service at ISRA was
8.1 and 13.4 years respectively.

If one compares mean salaries between job classifications,
there is a 13.3 per cent difference between 6-1 and 6-2 level
salaries. The difference between mean salaries of 6-2 and 6-3
researchers is 13.2 per cent. The difference between the mean
salaries of those with 6-.l and 6-3 classifications is 28.3 per
cent. As reported in Chapter 3.5.2, there are very limited
opportunities to achieve classification of over 6-3. Thus, the
salary structure at ISRA does not offer incentives to researchers
that encourage a career in research.

This analysis does not include outside incomes earned by
ISRA researchers. Personal communications with ISRA researchers
revealed that some researchers do have opportunities to
significantly supplement their incomes as consultants. As a
result, there may be hidden benefits available to ISRA
researchers which counterbalance the disincentives contained in
ISRA's salary structure. However, the availability of these
opportunities to earn outside income is uncertain. It is believed
that the opportunities for an ISRA researcher to earn outside

income are erratic and somewhat unevenly distributed.

120

4.5.3) Real Earnings of ISRA Researchers between 198; and 1982

The analysis of the real earnings of ISRA researchers
between 1981 and 1987 was complicated by the lack of a consistent
salary reporting method in the personnel records. Salary
information was not available for all years at most research
centers. Large fluctuations in the reported salaries of
researchers occurred in the personnel files of' a. number of
research centers during the earlier years of the study. Complete
salary information was available from the personnel files of only
two research centers-- Bambey and Dahra. Thus, the salaries of
eight researchers from either Bambey and Dahra were selected to
represent all researchers.

Real salaries of ISRA researchers were calculated in terms of
constant 1981 francs (FCFA) from a ‘price index for Senegal
reported by the Banque Centrale des Etats de l'Afrique
Occidentale (BCAO) in Dakar. The price index declined
significantly in 1987. It is believed that this decline reflects
changes in the international money market since the FCFA is
pegged to the French Franc.

Figure 4.4 shows the decline in real incomes of two
representative ISRA researchers between 1981 and 1986. All eight
salaries that were tested displayed a similar decline. The tests
reveal that ISRA salaries have failed to keep pace with

inflation. Even researchers promoted into research administration

 

0:!

«woe-3m

121

Figure 4.4: Declining Real Salaries at ISRA-- Two Examples

"HOHIHRL US. REEL SHLRRY: PROflOTED INTO RDflIH. IN 138?"

 

388

275‘

258‘

 

225 -

288«

175‘

158+

 

 

 

125
1981

1982 1983 1984 1985 1988 1987

"HOfllflAL US. REEL SALARY: ISRA RESEARCHER PROHOTID

T0 6'2 STATUS IN 1983"

1988

 

258

2251

 

 

288+”fi

175‘

OHIO“! "C'DF’M

1584

 

125

. .-
....~... "d..
..o.00.... ".-00‘
I. .

 

 

1981

1982 1983 1984 1985 1986 1987

.___HOHIHAL SALARY ....RRAL SALARY (1981 fCFA)

Source: ISRA Personnel Files

 

1988

 

122

experienced. decreasing in real incomes during the 'period. of
study. Promotion into research administration, the most
significant mechanism by which ISRA researchers can increase
their salaries, appears only to have slowed the rate of decrease
in real salaries.

The trend of declining real salaries between 1981 and 1987
is likely' to 'have further reduced. the career incentives for
researchers at ISRA. Researchers at ISRA who perceive a declining
stream of real income in their present employment are more apt to
seek higher paying jobs elsewhere.

Thus, an image of dissatisfied and poorly motivated
researchers emerges from the analysis of ISRA's salary structure.
Presumably, the productivity' of' research throughout ISRA. has
suffered as a result.

Declining real salaries likely were a generalized phenomenon
in Senegal between 1981 and 1987. The BCAO reports the minimum
wage for factory workers in Senegal. In real terms, the minimum
wage declined by 33 per cent between 1981 and 1987. Attempts to
compare the decline in the real wages of ISRA researchers with

other the real wages of other professionals in Senegal failed. 1\

 

1. Although salary data are published from a number of

countries around the world in the Bulletin of Labor Statistics
(ILO, 1988) and Xeaz flock of Labor §ta§isticg (ILO, 1988), no

salaries were reported from Senegal between 1981 and 1988.

123

41.5.4)me
Qrganizations

Figure 4.5 (p.126) shows ISRA salaries in relation to
salaries at other organizations which. compete with ISRA for
personnel with similar skills. Private voluntary research
projects (PVO's) refer to activities funded by donors such as
church groups or the United Nations that have ongoing development
and research projects in Senegal. The agricultural projects of
many of these private voluntary organizations represent potential
sources of employment for ISRA researchers. PVOs pay their
employees in local currency.

The salary scale in Figure 4.5 represents a conservative
estimate of a typical PVO salary scale. It 'was possible to
estimate only the beginning salary offered by a representative
PVO organization. This beginning salary was believed to be about
100,000 FCFA higher than ISRA salaries. Thus, a beginning monthly
salary of FCFA 325,000 was assumed" A 2.5 per cent salary
increase for each year of research experience possessed, by
researchers was also assumed. It is believed that this is a very
conservative estimate. Table 4.11 showed that a 3.5 per cent
increase in salary was associated with an added year of
experience at ISRA. PVO organizations operate in a labor market
which is open to the forces of labor supply and demand.

Thus, the actual salary structure of PVC organizations is

apt to be higher than the curve estimated in Figure 4.5. It

124

should be noted that the estimated entry salary at PVO
organizations is competitive with the base salaries of upper
level management personnel at ISRA.

No attempt has been made to estimate non-salary benefits
such as transportation allowances or housing allowances in
estimating salaries. These benefits are often substantial. PVOs
in general are believed to offer more attractive benefits than
ISRA. However, the average tenure of researchers at PVO
organizations is probably shorter than tenure for researchers who
remain at ISRA. Contracts given to ISRA researchers are of
indeterminate duration. Personal communications indicate that
once a researcher is hired, ISRA has very limited legal

mechanisms for terminating a researcher's employment.

4.5.5)

 

Senegalese researchers employed at international and
regional institutions outside Senegal receive internationally
competitive salaries. The salary structure of one such
international organization was obtained for this study and has
been assumed to be representative of all international
organizations. International salaries are determined by supply
and demand forces in the international labor market for
researchers.

Figure 4.5 shows three levels of international salaries--

125

converted from dollars into FCFA ($1 - 300 FCFA), and from yearly
salaries into monthly salaries. The salary scale labelled
"International 1" represents the salary structure that is most
likely to enter the benefit calculations of ISRA researchers.
This salary structure is applicable to relatively inexperienced
researchers, short-term contracted researchers, and recent
graduates.

The salary scale labelled "International 2" is applicable
to researchers with PhD's who have 7 to 10 years of experience in
research. The "International 3" salary scale is offered to

researchers with PhD's and 10 to 12 years of experience.

4.5.6) government Salaries

The salary structure for civil servants in Senegal has not
been shown due to data limitations. However, ISRA has a number of
civil servant researchers assigned to ISRA by the Senegalese
government. These lifetime civil servants receive government
salaries while at ISRA. EPIC salary provisions require ISRA to
pay civil service researchers the difference between their civil
service salaries and the salaries of contracted researchers at
ISRA with equivalent job classifications (Issa-Sayegh and
LaPéyre, 1981). The average salary paid to the 20 civil servant
researchers at ISRA in 1987 FCFA 60,880. The range of salaries

was FCFA 42,000 for researchers with 6-1 classification, and FCFA

 

126

 

 

Monflfly Sawfly
FCFA

2000000
1800000

1400000
1200000

800000

Figure 4.5: Salary Scales
ISRA and Competing Organizations

1987

 

4-

«1-

d .mOOOMO“OCIIOO~I
O

1 000000 ‘

, /

4p /
1000000 I

—_V

EDOOOO*- ”0'”

400000 ‘*

200000 1’

0

 

 

 

 

. I
I r I I' I I I T

13 5 7 911131517

Researcher Experience
Years

, LEGEND

_ ISRA

_- PVO Projects
...... International 1
._- International 2
......... International 3

Source: ISRA Personnel Study

 

 

 

127

81,000 for researchers with 7-2 classification. ISRA salaries are
thus significantly higher than those salaries paid by the
Senegalese government.

A few of the researchers who left ISRA are currently
employed at other EPIC development agencies. Although the base
salaries of these organizations are similar to ISRA's, personal
communications with ISRA researchers indicate that overall
salaries at other EPIC organizations are higher than those
offered at ISRA. As stated previously, ISRA is something of an
anomaly among EPIC institutions in Senegal in that it does not
pay productivity bonuses. Other EPIC institutions do pay salary
bonuses. The difference in monthly salary between ISRA and other
EPIC institutions is thought to be at least FCFA 50,000 after
monthly and annual productivity ‘bonuses are included. in the
salary calculations.

Finally, salaries offered by competing organizations are
not the only factors which enter a researcher's calculation of
the benefits and costs in seeking alternative employment. Work
environments that are perceived to be more attractive, better
promotion opportunities and other fringe benefits offered by
other organizations, and personal considerations influence a
researcher's decision to change jobs. For example, opportunities
to gain professional recognition through publications or

conferences are apt to be greater at PVO's, and particularly at

128

international organizations. International organizations also
offer very attractive cost of living, vehicle, and. housing
allowances.

Due to data limitations and. the fact that researchers'
perceptions are often subjective and very difficult to quantify,

non-salary benefits have not been included in this analysis.

4.5.7) mm

Figure 4.5 shows that ISRA salaries are low relative to the
alternative organizations in Senegal that demand professional
labor. Low and declining real salaries at ISRA. relative to
alternative organizations would appear to be a major factor
underlying a researcher's decision to leave ISRA.

Information on current employment was obtained for only 23
of the 43 researchers who left ISRA between 1981 and 1987. Three
of these 23 researchers were employed by government ministries,
including two researchers who became Ministers. Two of the 23
departed researchers were employed by other EPIC organizations.
Six departed researchers were employed by either PVOs or
organizations that pay salaries equivalent to PVOs. Eight
researchers who left ISRA 'were employed 'by international or
regional research agencies. Three departed researchers went
overseas, including two who married expatriates. Additionally,

one departed researcher was employed by a multinational company

129

in Dakar.

In. most of the above cases, researchers improved. their
financial earnings by leaving ISRA. Turnover is a process which
may create more turnover. As researchers who stay at ISRA compare
their earnings with those of colleagues who left, they are apt to
perceive themselves to be relatively less well-off. Thus, an
individual's perception of both the benefits in alternative
employment and the costs of leaving ISRA might change during
periods of rapid turnover.

The analysis suggests that low salaries relative to other
organizations influence the decisions of research trainees and
research administrators to leave ISRA. Research administrators
and managers are usually amongst the most experienced members of
ISRA's research staff. The ISRA salary structure has been shown
to reward both experience and research administration and
management roles. The mean age of research administrators in 1987
was 41.1 years, with a range between 33 and 45. Thus, many of
ISRA's research administrators reach or approach their peak
earnings while at a relatively youthful age.

Whereas staying at ISRA yields a declining stream of
benefits in terms of real wages, the higher relative salaries
offered to research administrators by other organizations yield
high anticipated benefits. There is an opportunity cost to

remaining at ISRA for research administrators and experienced

130

researchers. Since limited supplies of experienced research and
administrative personnel are generally much in demand in Africa,
the opportunities for research administrators to find employment
in alternative organizations presumably are relatively good.

As seen previously, eight of the ten researchers with
trainee status who left had Master's Degrees. Two of the
temporary researchers who left had Master's Degrees. Discussions
with West African colleagues strongly suggest that individuals
educated in North America are in high demand in Francophone
Africa. Reasons for this include fluency in English, a greater
emphasis on quantitative skills at American institutions, and the
fact that the supply of American educated individuals in
Francophone Africa is limited in relation to French educated
individuals .

In light of this, American educated recruits to ISRA who
become frustrated with long delays or vaguely defined research
roles during their probationary period have no incentive to
remain at ISRA. Another possible explanation for high turnover
among American educated recruits at ISRA is that these recruits
acquired higher reward (earnings) expectations while overseas,
and thus are less apt to be retained by the limited reward system
at ISRA.

Table 4.10 showed relatively high turnover in researchers

trained in the social sciences. Turnover was relatively low for

 

131

researchers with highly specific training. Senegal has only one
college level or post-secondary agricultural school which offers
Ingenieur (BSc equivalent) degrees. Specialized post-graduate
training in agriculturally related disciplines must be obtained
overseas. Supplies of well-qualified agricultural scientists are
limited. However, agriculture has become the primary focus of
development projects in Senegal. Much agricultural development
activity has been associated with the damming of the Senegal
River in 1986. A number of long-term development projects were
established by a variety of donor groups after the 1984 Sahelian
drought.

Thus, agricultural scientists with relatively applied
research skills in the social and basic agricultural science are
believed to be in high demand in Senegal. Highly specialized
researchers are believed to be less highly demanded by

organizations outside ISRA, since their skills have more narrow

applications.
4.5.8) WM
Inseam

Figure 4.6 compares ISRA's income profile with an ideal
income profile. The actual income profile represents the salary
progression a recruit can expect while at ISRA. It was derived by

applying EPIC salary statutes. The length of service bonuses

132

described in Chapter 3.5.2 were applied over time to the base
salary of 199,350 FCFA that all beginning researchers at ISRA
receive. To represent possible reclassification during the course
of an ISRA researcher's career, promotions to grade 6-2 and 6-3
were assumed after the eighth year and fourteenth. years of
service respectively. This reflects the mean length of service of
researchers classed as 6-2 and 6-3.

The "ideal" or target salary structure was based on the same
assumptions that were used to derive the researcher's
productivity curve in Chapter 3.7. Elliott and Rang (ISNAR, 1985)
describe a normal, or ideal, income profile for researchers based
on human capital theory. During early years, a steep rise in
salary is useful in attracting researchers to a research
institute. As researchers develop their skills by doing research
they 'become more productive on. the job. Ideally, an. income
profile should reward increased productivity. Towards the end of
a researcher's career, the rate of salary increase slows before
peaking in the 20th year. This reflects both the obsolescence of
skills and diminishing investments in on-the- job training and
formal education that occur later in a researcher's career.

Salary increases of 200-250 per cent over the course of a
professional's career are considered desirable in salary
structures (ISNAR, 1988). A 200 per cent salary increase over a

20 year period was assumed in calculating an ideal salary

133

Figure 4.6: A.Comparison of ISRA's Income Profile with an
Ideal Income Profile
13.58

 

13.251 ............................................
13.08‘ 0»: ...........

12.75. .
12.501 ,,,8, I

12.25.1’" .5____,o—-r-r""’_7_d-_Fdfr

12.90 ,

 

 

£2.10me ”'30 not“

 

 

 

2 ‘ 4 8 . 8 ' 18 ' 18 ' 18 g 18 ' 18 ' 28 ' 28 ' 28
Years of Experzenoe

.___Actual lncone Profile .....ldeal lncone Profile
Source: ISRA Turnover Study.

structure at ISRA. A 200 per cent increase in the base salary of
199,350 FCFA results in a salary of 598,050 FCFA. Thus, the ideal
salary structure in Figure 4.6 is competitive with international
and regional salaries (see Figure 4.5, p.126).

Compared to the ideal income profile, the actual ISRA income
profile is relatively flat. Flat income pr0files are
characteristic of low-skilled occupations. In these occupations
there are low training costs, low productivity gains over time,
few incentives for workers to’remain at an institution, and

lgttle incentive for an organization to retain employees

(Elliott, 1988; Elliott and Kang, 1985). Thus, rapid turnover of

134

employees is associated with flat income profiles.

The actual income profile of researchers at ISRA presented
in Figure 4.6 is unsatisfactory for professional personnel. A
major contributing factor to the increasingly high rates of
turnover among research scientists at ISRA since 1985 has been a

salary structure that is ill-designed for a research institution.

4.6) MW
4.6.1) Causes of Iurggvg:

Turnover in Senegalese research staff at ISRA. was not
correlated with the following variables: formal educational
background of a.researcher, length of service at ISRA, research
department of a researcher, and research program. Turnover in
Senegalese research staff was a generalized phenomenon between
1981 and 1987.

The status of researchers was a significant factor in
turnover rates amongst Senegalese researchers. Recruits to ISRA
who left the institution before attaining full-research status,
temporary researchers, and research administrators formed the
majority (62.8 per cent) of those researchers who left ISRA
before 1986. The frequency of full-time researchers leaving ISRA
increased steadily since 1985. The majority of trainees who left
ISRA did so in 1986 and 1987. Thus, turnover at ISRA began to

significantly erode research capacity in the later years for

135

which data are available. The departure of trainees also
represents a weakening of future research capacity.

Indications are that social scientists and researchers with
general agricultural skills were more apt to leave ISRA than
researchers with highly specialized agricultural backgrounds.
This finding echoes human capital theory. Hammermesh and Rees
(1984) believe that individuals with high degrees of firm-
specific training are less apt to change jobs than researchers
with more general skills. Highly specialized, or job-specific,
skills link an employee to a job since there are limited numbers
of jobs to which these skills can be applied. In Senegal, social
science and general agricultural skills are believed to be in
high demand given a plethora of development projects.

Turnover at ISRA appears to have been a manageable problem
until 1985. After 1985 turnover at ISRA became an increasingly
frequent phenomenon. In the 1986-87 period, turnover and
instability in research staff at ISRA was pervasive. Given the
limited explanatory power of many of the variables contained in
ISRA personnel files, other factors are considered to have
influenced turnover of research scientists at ISRA.

In 1986, a major expansionary effort at ISRA ended as a
number of donor projects began nearing completion. The size of
ISRA's Senegalese research staff grew 65.7 per cent between 1982

and 1985. There were 70 Senegalese researchers at ISRA in 1982,

If

 

136

and 116 in 1985 (excluding the Fisheries Department). As numbers
of Senegalese researchers peaked, donor financing of research
projects by the World Bank, FAO, USAID and ICRISAT began to
decline. Accompanying the decline in donor support was a very
high rate of turnover of expatriate researchers. The period also
witnessed declining real financial support from France, and
decreasing financial support from the Senegalese government.

Thus, at the same time that the size of ISRA's national
research staff increased, financial support for research at ISRA
decreased. As demand for research resources increased, the supply
of research resources decreased. Manpower development planning
took place without accounting for ISRA's medium and long-term
financial resources.

Further constraining ISRA's financial and management
capacities were high rates of turnover and instability in
research administration and management personnel, and two changes
in organizational structure. The research administrators and
managers who left ISRA. were usually replaced. by’ experienced
researchers with limited experience and managerial skills.
Reorganizations in 1983 and 1987 changed the framework within
research was organized and managed. Included in the 1987
reorganization plan was the laying off of a third of ISRA's
research support staff. Clearly, 1986 and 1987 were characterized

by a large degree of institutional uncertainty.

137

The 1983 - 1987 period also witnessed an erosion of
researchers' salaries, resulting in declining job satisfaction
and morale. The 1983 reclassification of jobs largely penalized
researchers and benefitted non-technical support staff.
Researchers' salaries and careers became frozen. Inflation in
Senegal significantly lowered the real salaries of ISRA
researchers between 1981 and 1987.

Compounding the salary problem was the availability of more
highly paying jobs in other EPIC institutions, other development
projects and in international and regional development
organizations. Low relative salaries for researchers, declining
real salaries, poor promotional prospects, and a salary structure
which rewards seniority at ISRA and administrative and management
functions more than research functions combined to weaken ISRA's
ability to retain and motivate researchers.

Thus, the primary factors that influenced the process of
turnover of research scientists at ISRA were: declining financial
support for research projects, overly rapid expansion based on
poor planning, unstable and weak management capacity,
institutional instability, and a professional reward structure
that is ill-suited for professional researchers. As a result of
these factors, and their interactions, turnover rates at ISRA
began to increase at ISRA in 1985. In 1987, the combined effects

of the above factors coalesced, leading to high rates of

138

turnover, a high degree instability in research staff, and an

unraveling of institutional cohesiveness and continuity.

4.6.2) The Impacee ef Iugneve;

The fact that turnover at ISRA has been generalized across
research departments and. research. programs suggests that its
impacts have been felt throughout the institution. The
productivity and effectiveness of ISRA's research efforts were
likely to have suffered as a result. Turnover undermined, or at
least delayed, ISRA's efforts to expand its research capacity by
improving the qualifications, experience and skills of its
research personnel. The selection of experienced researchers for
administrative and management posts further undermined research
capacity.

In 1987, the continuity in research and research staff that
is essential in creating effective and productive research
projects was severely disrupted. 34.9 per cent of the Senegalese
researchers present at ISRA in 1986 either left or changed job
functions. Nearly half of the expatriate researchers at ISRA in
1986 left by the end of 1987. The impacts of this instability had
to have reverberated throughout the institution.

To gauge the impacts of turnover in research staff upon
research continuity, turnover rates were calculated over multiple

year periods. All these rates reflect the high turnover in 1987.

139

Over the six year period between 1981 and 1987, half (50.4 per
cent) of the 113 researchers present at ISRA in 1981 left the
institution. 76.2 per cent of the 42 expatriates at ISRA left
during this period and 35.2 per cent of the 71 Senegalese
researchers left. Over a four year period, 1983 - 1987, 45.6 per
cent of the 136 researchers who were present at ISRA in 1983
left. 70.6 per cent of the 51 expatriate researchers departed
ISRA, as did 30.6 per cent of the 85 Senegalese researchers.
Between 1985 and 1987, over a third (34.1 per cent) of the 176
researchers at ISRA in 1985 left. 55 per cent of the 60
expatriate researchers, and 23.3 per cent of the 116 Senegalese
researchers departed ISRA during this two year period.

Given that ‘adaptive or applied. research 'projects take a
minimum of 4 or 5 years to complete, and basic research projects
take at least 10 years, the instability of scientific personnel
resulting from high turnover must have severely disrupted
research continuity.

The high rates of expatriate departures between 1981 and
1987 in part reflect the goal of replacing them with Senegalese
researchers. However, rapid rates of turnover among expatriates
probably lowered the performance level of research at ISRA given
the large role expatriates play in the operations of a number of
research programs.

Thus, in constraining the operations of various research

140

projects, and in lowering research performance, turnover has
lowered the returns to the resources invested into agricultural
research in Senegal. In particular, turnover has truncated the
stream of potential benefits that were anticipated from the
substantial investments were made in the formal education and
training of ISRA researchers (see Chapter 3.7). A weakness of
this study has been its inability to measure the effects that
turnover has had on research performance at ISRA.

A number of the researchers who departed ISRA became
actively involved in other governmental agencies, a number of PVO
projects, and in international and regional development
organizations. This suggests that while turnover was detrimental
to ISRA's institutional capacity' to ‘perform 'research, social
benefits have continued to accrue to investments in the human

capital of the majority of researchers who left ISRA.

141

W: CONCLUSIONS AND RECOMMENDATIONS

5.1) W
5.1.1) Qegeral Conelusioge

The major causes of turnover among research scientists at
the Senegalese Agricultural Research Institute (ISRA) were:
overly rapid expansion, a rigid and inappropriate salary
structure for researchers, haphazard manpower planning
strategies, weak research administration and management capacity,
and declining financial support for research. activities from
donors and the Senegalese government. These factors combined to
create a very high rate (18.1 per cent) of turnover in Senegalese

researchers in 1987.

5.1.2) Reeogmendatiene

The analysis of turnover in research scientists at ISRA
between 1981 and 1987, and the identification of factors
influencing this turnover strongly suggest that ISRA must improve
its ability to retain researchers. In this section, policy
recommendations to reduce turnover in research scientists are

proposed.

1) ISRA should be given special legal status or modified rules
for operation.

142

Underlying the process of turnover is ISRA's legal status
as a "public establishment with industrial and commercial
characteristics (EPIC)". Classifying an agricultural research
institute as EPIC is undesirable. Unlike most EPIC institutions,
ISRA is unable to generate income to supplement the insufficient
funds it receives from the Senegalese government. These funds are
entirely absorbed by salaries. To compensate for funding
shortfalls, ISRA. must rely on external sources to fund its
research activities. As a result, donor agencies exert a large
influence over ISRA's research and institutional development
goals.

ISRA's current salary structure is based upon EPIC statutes.
However, the performance bonuses that are paid to most EPIC
employees are not paid at ISRA as a result its financial
constraints. The salaries paid to researchers declined in real
terms between 1981 and 1987. Researchers' salaries are not
competitive with those offered. by alternative organizations.
Salaries at ISRA are based on seniority and job classification
rather than merit or a researcher's performance.

EPIC salary statutes lack career incentives for researchers.
Promotions and increases in earnings are largely contingent upon
a researcher abandoning research and assuming administrative or
managerial roles. As a result, the professional skills of

researchers are used inefficiently since experienced and skilled

Jr:

 

143

researchers are frequently inexperienced and unskilled
administrators and managers. EPIC salaries are inappropriate for
motivating and retaining researchers.

Classification as an EPIC institution gives ISRA a large
degree of autonomy in managing its affairs and in recruiting
personnel. A number of researchers and managers at ISRA have
argued that ISRA should be given special legal status which
recognizes the scientific character of the institute (Faye and
Bingen, 1988). Such a change in status would allow ISRA to have
the same degree of managerial autonomy, and also would enable
ISRA to devise a salary structure that gives scientists
incentives to pursue careers in research. Special legal status
might also permit changes in the way in that the Senegalese
government allocates financial resources to ISRA.

A modification of ISRA's legal status appears to be
unlikely. Faye and Bingen report that past demands to change

ISRA's status have been refused.

2) Barring a change in legal status, modifications within ISRA's
current salary structure are desirable.

Currently, all research positions at ISRA are classified as
either grade 6-1, 6-2, or 6-3. In 1987, 7 of the 10 researchers
who were classified as 6-3 served primarily administrative

functions. Grades 7 and 8 apply only to research center directors

144

and upper level management. Thus, the salary structure is biased
against research personnel in favor of administration.

An appropriate reward structure for a research institution
offers researchers opportunities to increase their rank and
earnings while practicing research. ISNAR (1987) recommends that
national agricultural research institutes should offer dual
administrative and scientific career ladders for ‘professional
staff. Research staff should receive an income comparable to, if
not greater than, that of management staff. As a result,
researchers are given incentives to pursue a career in research.

It would be relatively easy to modify existing EPIC salary
statutes to equate scientific and management functions. Expanding
the number job classifications available for researchers by
creating research positions graded as 6-3, 7-1, 7-2 and 8-1 would
equate the salaries of researchers with those of administrators.
In offering improved career opportunities in research, ISRA would
enhance its ability to motivate and retain researchers.

Crucial to such a system would be a clear delineation of job
responsibilities and criteria for promotion. Ideally, promotion
criteria should be based upon an evaluation of a researcher's
performance and merit rather than length of service. Currently,
ISRA lacks clearly defined criteria to either evaluate a
researcher's performance or to select researchers for

administrative positions.

145

3) The professional reward structure for ISRA researchers should
be improved.

A reward structure for scientists should include financial
and professional incentives. However, ISRA researchers generally
have limited opportunities for gaining professional recognition
through publication. ISRA should both encourage its researchers
to publish, and actively assist researchers in preparing and
editing articles which are to be submitted for publication.

Although planned, a scientific journal which is to be
published by ISRA has yet to get off the ground (Faye and Bingen,
1988). Such a journal would give ISRA researchers a medium in
which to publish and inform others about their research
activities. Professional recognition would also be enhanced by
increasing opportunities for ISRA researchers to attend national,

regional or international seminars and conferences.

4) Prior to further expansion, existing resources and research
activities should be consolidated.

Major investments were made in expanding ISRA's research
capacity between 1981 and 1987. For example, between 1982 and
1985 the number of Senegalese researchers at ISRA grew by 65.7
per cent, from 70 to 116. Major investments were also made to

expand research facilities, equipment, and the number of research

146

activities. However, there are strong indications that this rapid
expansion outstripped ISRA's limited financial, managerial, and
technical capacities. ISRA's ability to retain research
personnel, to provide sufficient operating funds for research
activities, and to manage research efforts was weakened as a
result.

Von der Osten (World Bank, 1988) believes that rapid
expansions in research capacity that are undertaken. without
consideration of an institution's financial and human resource
base hamper research performance in African agricultural research
institutes. By dispersing scarce resources over a broad range of
activities, resources allocated to research are used
inefficiently. Von der Osten urges research institutes to
increase the productivity and efficiency of existing resources by
consolidating and concentrating research efforts. The size and
complexity of various research activities should be tailored to
the human and financial resources of a research institute.

To increase the amount of' funds available for research
operations, ISRA's budget should be rearranged so that a small
amount of research funds are directly available to the institute
from the Senegalese Government. Although SPAAR. (World. Bank,
1988b) recommends that 30 per cent of all government funds should
be 'budgeted for research operations and. maintenance, such a

figure appears to be unrealistic for Senegal in the foreseeable

147

future given the fact that all government funds are absorbed by
salaries. However, a smaller percentage of funds could be
earmarked for research. If research efforts were tailored in
accordance with the new budget, research performance at ISRA
would be less constrained by limited funds for daily research

operations.

5) Research planning and manpower development strategies should
be coordinated between donors, ISRA, and the Senegalese
government.

A major contributing factor to the turnover process was the
poor coordination of activities between ISRA, the Senegalese
government and donor agencies. At the same time that some donor
groups were financing manpower training and development programs,
financial support for research operations decreased as other
donor programs neared completion. At the same time, the
Senegalese government reduced its financial support for ISRA. The
research efforts of ISRA researchers were hampered as a result.
ISRA lacks a long-term and clearly defined manpower development
policy that could have avoided this funding bottleneck.

The World Bank (1988) and the Special Program on African
Agricultural Research (SPAAR, World Bank, 1987b) recommend that
the time horizon for both research and manpower development

planning should cover 15 years. The planning horizon should

148

include short-term. (0-5 years), medium-term (5-10 jyears) and
long-term (10-15 years) projections of research goals,
activities, specific research needs, and available resources.
Planning should be conducted at both an institutional and
governmental level with the participation of donor agencies.
Donor activities should be well-defined and coordinated in
the long-term research plan. SPAAR argues that donor assistance
should be committed to national agricultural research institutes
on a long-term basis. The 4-6 year time frame of most donor
investments impedes research program planning, and leads to gaps
in project funding. SPAAR also recommends that the number of
short-term expatriate researchers that are employed. in. donor
projects should be limited. Rapid turnover of expatriate
researchers contributes to problems of organizational
instability, and limits the effectiveness of technical assistance

programs .

6) Future manpower development strategies should focus upon
developing administrative and technical support capacity in
addition to research capacity.

Past manpower development efforts at ISRA have focused
primarily on developing the skills of researchers . Training
administrative and technical support staff was not a ma j or

priority. Limited technical support and administrative capacity

149

at ISRA has been strained by the rapid expansion of ISRA's
research facilities and manpower.

The outcome of a research effort depends upon the competence
of administrative and technical support staff, in addition to
scientific staff. The lack of qualified and competent personnel
in large part explains the weakness of ISRA's management,
administrative, and technical capacities (Faye and Bingen, 1988).
Thus, efforts should be made to train and upgrade the skills of

ISRA's non-professional personnel.

7) Personnel record keeping procedures at ISRA.should be
improved in order to provide adequate information on which to
base manpower planning.

To analyze turnover of researchers at ISRA, a data base had
to be created which contained information that should have been
readily available from ISRA's personnel files. The procedure that
was used to determine which researchers left ISRA in a given year
illustrates the weaknesses in ISRA's personnel record keeping
procedures. Personnel files were collected from each research
center for the years 1981-87. By coding and transcribing the
names of employees on the personnel files into a data base, it
could be determined whether an individual had left ISRA or if new
employees had been hired. Employees were classified on the basis

of their jobs. Researchers were then selected from each research

150

center, put into a separate file, and reclassified on the basis
of research department.

Having researchers classified on a departmental basis,
supplemental educational, salary, and disciplinary information
could be entered into the data base. Much of this information was
obtained through communications ‘with ISRA administrators and
researchers. Ambiguities and missing information complicated the
task of compiling an accurate data base for ISRA researchers.
Clearly, ISRA's personnel record keeping procedures are
inadequate.

In addition to basic personal characteristics such as name
and age, a personnel data base should contain information
concerning: the educational background and qualifications of a
researcher, additional training received while on-the-job,
previous work experience, up-to-date job history including
changes in research program or job status, salary, publication
and performance data, and termination of employment.

Personnel data bases contain the' information upon which
personnel management and manpower planning is based. There is a
major need at ISRA to begin compiling a personnel data base which
is accurate and up-to-date. Poor information about personnel has
contributed to inefficiencies in ISRA's manpower development

planning and management of scientific resources.

151

5.2) W

This study originally intended to measure the costs
associated with turnover of researchers by estimating lost
returns to investments in human capital. Proxy measures of a
researcher's output and productivity, such as publications, would
have permitted a cost-benefit analysis which estimated the
benefits forgone as a result of turnover. Severe data limitations
prevented this analysis from being undertaken. However, an
analysis such as this would enhance understanding of the costs
and impacts of the process of turnover in research staff.

This study was limited by its lack of detailed performance
and program information for researchers. Information was also
unavailable on levels of funding in research programs and
projects. Had more detailed personnel and institutional data been
available a model of the turnover process could have ‘been
developed. Such a model could either have more clearly identified
the factors which contribute to the turnover process, or
estimated the impacts of turnover on institutional performance.
Modeling the turnover process of researchers would enhance
research management planning, and inform policies designed to
reduce turnover.

Turnover in research staff increased dramatically in 1987.
ISRA's structural reorganization in 1987 established clear lines

of authority between research departments and research centers.

152

More coordinated procedures for planning research and identifying
priorities were also established. It is believed that the 1987
reorganization will lead to increased stability in research
staff. A follow up study of turnover in research staff at ISRA
would be beneficial in evaluating the effects this new
organizational structure has had on both turnover and
institutional performance.

Finally, this study is somewhat unique. There have been few
similar studies that have analyzed turnover of researchers within
national agricultural research institutions in developing
countries, and in Africa in particular. In many respects, this
study has been limited by its case study approach. Comparative
studies of turnover at other national agricultural research
institutes in Africa would both increase the usefulness of this

study, and enrich understanding of the turnover process.

W

Ardila, J., E. Trigo, and M. Piniero. 1981. In D. Daniels and

B. Nestel, eds. u 1 c t to icultura
es ° oc W e - 0
u e 19 . Ottawa: International Development Research

Centre.

Arndt, T. M. and V. W. Ruttan. 1977. "Valuing the Productivity of
Agricultural Research: Problems and Issues" In T. M. Arndt,

G. Dalrymple, and V.W. Ruttan, eds. Be eeegreee gellecegiog egd

od ct v t Na on a ti al ch.
Minneapolis: University of Minnesota Press.

Arndt, T.M., D. G. Dalrymple, and V.W. Ruttan, eds. 1977.
cc t d v in t d
Ingerneeieeel_3eeee;ehe Minneapolis: University of Minnesota

Press.

Bennell. P. 1986. W
A waggitagive Ovefliew. The Hague: International Service
for National Agricultural Research (ISNAR).

Bennell, P. and P. Thorpe. 1987. "Crop Science Research in
Sub-Saharan Africa: A Bibliometric Overview". AggieeLEQIei
Administration agd Egeensien (25): 99-123.

Boudreau, J.W. and C.J. Berger. 1985. "Decision-Theoretic Utility
Analysis Applied to Employee Separations and Acquisitions“.

W281. 70. 581-612.

CGIAR.1984.

_ Washington
D.C.: The World Bank.

 

Chuta, E. and C. Liedholm. 1984. "Rural Small-Scale Industry:
Empirical Evidence and Policy Issues". In C.K. Eicher and
J.M. Staatz, eds. v m e
Wogie, Baltimore: The Johns Hopkins University Press.

Daniels, D. and B. Nestel, eds. 1981. Beeeg:ee_filleee£ien_§2_
WWW ksho ed n
Singepere 8-10 June 1981. Ottawa: International Development

Research Centre.

De Janvry, A. and J.J. Dethier. 1985. Ieeheeiegiee1_lnneze§ien_ig

A c tur ° P a c It te and

154

CGIAR Working Paper #1. Washington, D.C.: The World Bank.

Eicher, C.K. 1982. "Facing Up to Africa's Food Crisis". In C.K.

Eicher and J.M. Staatz, eds. ASIIQEISEIQL Qevelepment in the
Ih1£Q_EQIl§i Baltimore: The Johns Hopkins University Press.

Eicher, C.K. and J.M. Staatz, eds. 1984. Ag;ieei§e;el_negelenmen§
1n_£h£_1h1IQ_KQIlQi Baltimore: The Johns Hopkins University
Press.

Elliott, H. 1988. "Analyzing Compensation Issues Using Earnings
Functions". Paper Delivered at the International Workshop on
Human Resource Management in National Agricultural Research.
7-11 November, 1988. The Hague.

Elliott H and 3 Kane 1935 Analxzias_§_nsiti_aa_2f_§srxi___

u a ea ch x e Earnin

Eu negion s, The Hague: ISNAR.

Evenson, R. E. 1981. "Benefits and Obstacles in Deve10ping
Appropriate .Agricultural Technology". In. C.Ku Eicher' and
J.M. Staatz, eds. ve th d
World, Baltimore: The Johns Hopkins University Press.

Evenson, R.E. 1977. "Comparative Evidence on Returns to
Investment in National and International Research
Institutions". In T.M. Arndt, G. Dalrymple, and V.W. Ruttan,

eds. Beggars: Allasatian_ans 2r2dastixitx_in_flatienal_and_
lg;ezne§ienel_3eeee1eh. Minneapolis: University of Minnesota

Press.
Faye. J. and R- James Bingen. 1988. QI8ED1§§£19£L££_92§£123_Q2_L§
Bs2hsrshe_§ur_lss_51atsaea_se_£redus£ien_au_§snéseli The

Hague: ISNAR. (Manuscript).

Falcon, W.P., C.T. Kurien, F. Monckberg, A.P. Okeyo, S.O.
Olayide, F. Rabar, and W. Tims. 1984. "The World Food and
Hunger Problem: Changing Perspectives and Possibilities,
1974-84". In J.P. Gittinger, J. Leslie, and C. Hoisington,
eds. E22d__22lisx1__InssxIasins__5unnlxi__nis£rihutian__and
Qeneemeeiene Baltimore: The Johns Hopkins University Press
for the World Bank.

Gittinger J.P., J. Leslie, and C. Hoisington, eds. 1987. £221.
'0» ' Q ‘- :_ - no I t o t o! g 8 o, _110 0
Baltimore: The Johns Hopkins University Press for the World
Bank.

155

Hammermesh, D.S. and A. Rees. 1984. MW

Fey, ”IQ, fig), New York: Harper and Row Publishers.
Hayami. Y. and V.W. Rattan. 1985. W
nte a v . Baltimore: The

Johns Hopkins University Press.

Hollenbeck. J.R. and C.R. Williams. 1986. ”Turnover Functionality
Versus Turnover Frequency: A Note on Work Attitudes and
Organizational Effectiveness". r i d

2022021281. 71, 606-611.

Idachaba, F.S. 1985. "The Evolution of National Agricultural
Research Systems in Sub-Saharan Africa". In Ming”:

mummwmm
W Malaga, Spain. 26 August-4 September 1985.
Aldershot, UK: Grower Publishing Co. 1986.

ILO. 1988. Wee. Geneva: International
Labor Organization.

ILO. 1988. MW. Geneva: International
Labor Organization.

ISNAR. 1984. The National Council for Science and Technology of
Kenya and ISNAR. WW

WWWEQL
The Hague: International Service for National Agricultural
Research.

ISNAR. 1987. Wo u e

W. The Hague: International Service

for National Agricultural Research.

 
   

ISNAR.1988. - - _ __ - .14. -- :
W The Hague:
International Service for National Agricultural Research.
(Manuscript) .

Issa-Sayegh,J . and C. Lapeyre. 1981. W
W
W Dakar:
Institut Universitaire de Technologie.

Jahnke, H.E., D. Kirschke, and J. Lagemann. 1987. W
' d the
bo t e t te a i a o
Reseagch m CGIAR Study Paper #21. Washington, D.C.:

156

The World Bank.

Judd, AWM., J.R. Boyce, and RAE. Evenson. 1986. "Investment in
Agricultural Research and Extension". In V.W. Ruttan and

C.E. Pray eds. Eoiiey fie; Agzieeieuzei geeeageh. Boulder:

Westview Press.

McEvoy, G.M. and W.F. Cascio. 1985. "Turnover and Performance: A

Meta-Analytic Review". Jegmel ef Applied Psyehelegy, 70,
342-353.

Mellor, J.W. 1973. ”Agricultural Development and the
Intersectoral Transfer of Resources". In C.R. Eicher and

J.M. Staatz, eds. WWW
HQIlQi Baltimore: The Johns Hopkins University Press.

Mellor, J.W. 1978. "Food Price Policy and Income Distribution in
Low-Income Countries". In C.R. Eicher and J.M. Staatz, eds.

WWW Baltimore: The

Johns Hopkins University Press.

Mellor, J.W. and B.F. Johnston. 1984. "The World Food Equation:
Interrelations among Development, Employment and Food
Consumption”. In J.P. Gittinger, J. Leslie, and C.
Hoisington, eds. '

W11
WWW Baltimore: The J ohne Hookina
University Press for the World Bank.

Mobley. WH 1982 WWW.
Qo 3:10 01. Reading, MA: Addison- -Wellesley.

Nestel, B. and E.J. Trigo, eds. 1984. te 3 es
AWE—Ammo. The Hague:
International Service for National Agricultural Research
(ISNAR).

Pinstmp- -Anderson. P 1982 AWN
W New York: Lonsman Group Ltd

Price, J.L. 1977. W. Ames, IA: Iowa State

University Press.

Reutlinger, S. 1985. ”The Nutritional Impact of Agricultural
Projects". In J. P. Gittinger, J. Leslie, and C. Hoisington,
eds WWW

Qeneemegieni Baltimore: The Johns Hopkins University Press

for the World Bank.

157

Rohrbach, D.D. 1984. leenee in Beeeegen Syeeen Qrgenizetion end
WWW CARSOM Working

Paper. East Lansing: Michigan State University Institute of
International Development.

Rattan. V.W. 1982. MW Minneapolis: The

University of Minnesota Press.

Schultz, T.W. 1979. "The Economics of Agricultural Research”. In

C.R. Eicher and J.M. Staatz, eds. Agrienlenrel_pezelenmen§
1n_£h2_Ih1IQ_HQIlQi Baltimore: The Johns Hopkins University
Press.

Schultz, T.W. 1977. ”Uneven Prospects for Gains from Agricultural
Research Related to Economic Policy". In T.M. Arndt, G.

Dalrymple. and V.W. Rattan. eds. MAW
Wm.

o v
Minneapolis: University of Minnesota Press.

Sene, D. 1985. "Etude de l'Impact de la Recherche Agronomique sur
le Dévelloppement Agricole du Senenegal". (Unpublished
Manuscript).

Stevens, R.D. and C.L. Jabara. 1988. MW
W.

V
Baltimore: Johns Hopkins University Press.

Todaro, M.P. 1981. "Internal Migration in Developing Countries: A
Survey". In R. Easterlin, ed.

W
in Developing Qenntgiee, New York: National Bureau of

Economic Research.

Tuinenberg , K . 1984 . ”Experience with Food Strategies in Four
African Countries" In J. P. Gittinger, J. Leslie, and C.
Hoisington, eds. e t Su

Distribution Wan... Baltimore: The Johns Hookine
University Press for the World Bank.

UNDP and FAO- 1984. WWW
Wale; Rome: FAO.

World Bank. 1981. Agzienlgnnel Beeeegen; Seeto; Policy Paneg,
Washington, D.C.: The World Bank.

World Bank, 1987a. Wellg Developmen; Reno;§ 1982, New York:

Oxford University Press.

 

158

World Bank. 1987b.

Special Program for African Agricultural
Research. uid e o t e nin atio a A icul ura
Reseagcn Systene in Snb-Sanaran Africa, Washington,

The World Bank.

D.C.:
ec
Meet

World Bank. 1988. D. Pickering, ed. Aggieen Aggicultural Researcn

velo ' es in Hi h- v

Feldalfin ed e u c o Wes German

September 24-27, 1987. Washington, D.C.: The World Bank.
Wortman, S. 1977.

    

"Impact of International Research on the
Performance and Objectives of National Systems". In T.M.
Arndt, G. Dalrymple, and V.W.
i.,°=.. ‘10 'oo
Reseanch

 

Ruttan,

eds.
0!

   
 

Reeougce
. Minneapolis: Universityo

  

'. 1!. -.
Minnesota Press.

 

 

M

II Ill!lllllllllllllllllll llllll.‘lllllll

 

3 1293 02364 9423