1

EFFECTS OF OFF-FARM WORK ON TECHNICAL

EFFICIENCY AND PRODUCTION RISK AMONG
RICE FARMERS IN ENUGU STATE, NIGERIA.

BY

ONYENEKWE, SYLVIA CHINASA

PG/M.Sc/09/52072

DEPARTMENT OF AGRICULTURAL ECONOMICS

UNIVERSITY OF NIGERIA, NSUKKA

FEBRUARY, 2014
 2

TITLE

EFFECTS OF OFF-FARM WORK ON TECHNICAL

EFFICIENCY AND PRODUCTION RISK AMONG RICE
FARMERS IN ENUGU STATE, NIGERIA.

A DISSERTATION SUBMITTED TO THE DEPARTMENT OF

AGRICULTURAL ECONOMICS, UNIVERSITY OF NIGERIA,
NSUKKA IN PARTIAL FULFILMENT OF THE
REQUIREMENTS FOR THE AWARD OF MASTER OF
SCIENCE DEGREE (M.Sc) IN AGRICULTURAL
ECONOMICS

BY
ONYENEKWE, SYLVIA CHINASA
PG/M.Sc/09/52072

PROFESSOR E.C. OKORJI

(SUPERVISOR)

FEBRUARY, 2014
 3
ii

CERTIFICATION

This is to certify that this research work is an original work undertaken by ONYENEKWE,
SYLVIA CHINASA, a postgraduate student of the Department of Agricultural Economics,
with registration number PG/M.Sc/09/52072, and has been prepared in accordance with the
regulations governing the preparation of project work in the University of Nigeria, Nsukka.

......................................... .............................................

Prof. E.C. Okorji Prof. E.C. Okorji

PROJECT SUPERVISOR HEAD OF DEPARTMENT

Date.................................. Date.....................................

......................................

External Examiner
 4

iii

DEDICATION

I dedicate this work to the Almighty God who is the reason for my living and to my husband
Mr. Onyenekwe, Jude Okechukwu.
 5
iv

ACKNOWLEDGEMENT

I want to first of all express my profound gratitude to the Almighty God who has
been my inspiration, guide and strength. Also, I wish to appreciate all those who have in one
way or the other contributed in making this research work a success. First in the litany of
contributors is my first supervisor, Prof. E.C. Nwagbo of the blessed memories, who despite
his health challenges took pains and time to go through my work up-to the proposal stage

My gratitude also goes to the Head of Department, Prof. E.C. Okorji who incidentally
is my new supervisor who supervised the work through to completion. He is a mentor indeed
and always attended to me promptly. I also acknowledge other lecturers in the Department
namely: Prof. S.A.N.D. Chidebelu, Prof. C.J. Arene, Prof. N.J. Nweze, Prof. E.C. Eboh, Prof.
C.U. Okoye, Prof. E.O. Arua, Prof. A.I. Achike, Dr. A.A. Enete, Dr. F.U. Agbo, Dr. N.A.
Chukwuone, Dr. Ben Okpukpara, Amaechina, E.C., to mention but a few for their
contributions in fine tuning the work.

I am highly indebted to my darling husband Mr. Onyenekwe, Okechukwu for his

immense assistance, encouragement, understanding and moral support throughout the period
of this work. I thank my parents Mr & Mrs Linus Oguike and my siblings for their prayers,
care and support.

I will not fail to appreciate some of my colleagues and friends who contributed in
different ways. They are Mr. K.P. Adeosun who assisted in data collection, Mrs C.U. Ike, Mr.
U.T. Okpara, Mrs R.N. Arua and Mrs C.J. Ayogu. I also want to thank my spiritual mentors,
Pastor Obeta Amos, Pastor Ebia Chris, Pastor Tony Njoku and the other brethren in the
Watchman Campus Fellowship, Nsukka for their prayers and words of encouragement. I will
not forget to thank Dr. Ogundari Kolawole for his immense assistance. I thank others who are
too numerous to be mentioned, who contributed in making this work a huge success.

Finally, I want to appreciate the financial support given to me by the African Network
for Agriculture, Agroforestry and Natural Resources Education (ANAFE).

Onyenekwe, Sylvia Chinasa

 6
v

ABSTRACT

The study analysed the effects of off-farm work on technical efficiency and production risks
among rice farmers in Enugu State, Nigeria. Ninety respondents were selected using multi-
stage sampling technique. Data for the study were collected by the use of well structured
questionnaire. Descriptive statistics and stochastic production frontier model were used to
analyse the data. Results showed that the two groups of rice farmers had similar socio-
economic characteristics. Technical efficiency scores for the farmers ranged from 0.579 to
1.000 and 0.0606 to 1.000 for the rice farmers without and with off-farm work respectively.
The average efficiencies are 0.964 and 0.871 for rice farmers without and with off-farm
work, respectively. This suggests that off-farm work has a negative effect on farmers’
technical efficiency. The result of the student t-test conducted at 5% significance level
showed that there is a significant difference in the mean technical efficiency of the two
groups of rice farmers. For rice farmers without off-farm work average number of farmers
associations (0.646), age (0.328), education (3.838) and extension access (3.144) significantly
and positively influenced technical inefficiency effects while for their counterpart age (0.159)
and extension access (4.727) significantly and positively influenced technical inefficiency
effects and household size (-0.970) was significant but negative. For farmers without off-farm
work, family labour (1.287) has a positive and significant effect on production risk, meaning
that family labour is a risk increasing factor. Depreciated value of equipment (-12.255) used
has a negative and significant effect on production risk which indicates that investment on
equipment will decrease the production risk in rice production. For rice farmers with off-farm
work none of the factors was significant even though they all had negative sign. The
constraints faced by the farmers were inaccessible road, high cost of transportation,
inadequate credit, birds’ invasion, inadequate extension support, inaccessibility to cheap farm
inputs ranked in ascending order of importance.
 7

vi

TABLE OF CONTENTS
Title.............................................................................................................................................i
Certification...............................................................................................................................ii
Dedication.................................................................................................................................iii
Acknowledgement.....................................................................................................................iv
Abstract.....................................................................................................................................v
Table of contents......................................................................................................................vi
List of tables............................................................................................................................viii

CHAPTER ONE: INTRODUCTION

1.1 Background of the Study.....................................................................................................1
1.2 Problem Statement...............................................................................................................5
1.3 Objectives of the Study........................................................................................................8
1.4 Hypotheses of the study.......................................................................................................8
1.5 Justification of the Study.....................................................................................................9

CHAPTER TWO: REVIEW OF RELATED LITERATURE

2.1 Characteristics of Nigeria’s Rice Production Systems ......................................................11

2.2. Off-Farm Work by Farm Households...............................................................................12

2.3 Determinants of choice of off farm work...........................................................................13

2.4 Interrelationship between off-farm work and efficiency....................................................14
2.5 Agricultural risk.................................................................................................................15
2.5.1 Classifications of Agricultural Risks...............................................................................16
2.6 Efficiency..........................................................................................................................17
2.7 Determinants of production efficiency...............................................................................20
2.8 Measures of Efficiency.......................................................................................................21
2.8.1 The non-parametric programming approach ..................................................................22
2.8.2 Deterministic production frontier models and technical efficiency................................22
2.8.3 Stochastic production frontier models and technical efficiency.....................................24
2.8.4 Stochastic Frontier model estimation–The Maximum Likelihood Estimation
Method.....................................................................................................................................27
 8

vii
2.9 Estimating Production Risk and Inefficiency Simultaneously.........................................28
2.10 Review of Related Empirical Studies...............................................................................29
2.11 Theoretical Framework....................................................................................................33
2.11.1 The Production Function, Risk, and Technology..........................................................33
2.11.2 Farm Household’s Maximization Problem...................................................................33
2.11.3 Production Economics Theory..................................................................................... 36
2.12 Analytical Framework......................................................................................................38
2.12. 1. Estimating the stochastic frontier model with risk......................................................38

CHAPTER THREE: RESEARCH METHODOLOGY

3.1 The Study Area...................................................................................................................41
3.2 Sampling Procedure...........................................................................................................42
3.3 Data Collection...................................................................................................................42
3.4 Data Analysis.....................................................................................................................43
3.4.1 Model Specification........................................................................................................43
3.4.1.1 Stochastic frontier production function........................................................................43
3.5.1.2 Likert Scale Rating Technique.....................................................................................47

CHAPTER FOUR: RESULTS AND DISCUSSION

4.1 Socio-economic Characteristics of Respondents...............................................................48
4.2 Rice production in the study area.......................................................................................54
4.3 Comparing the distributions of the technical efficiency scores between the two groups of
rice farmers........................................................................................................................55
4.4 Estimation of the Stochastic Production Frontier Model...................................................57
4.5 Constraints faced by rice farmers.......................................................................................60
4.6 Test of hypotheses..............................................................................................................61

CHAPTER FIVE: SUMMARY, CONCLUSION AND RECOMMENDATION

5.1 Summary............................................................................................................................63
5.2 Conclusion..........................................................................................................................64
5.3 Recommendations..............................................................................................................64

REFERENCES.......................................................................................................................67
 9
viii

LIST OF TABLES
Tables Pages
4.1 Socio-economic characteristics of respondents................................................................52
4.2 Stochastic Production Frontier Estimates........................................................................57
4.3 Distributions of technical efficiency.................................................................................58
4.4 Constraints faced by rice farmers.....................................................................................59
4.5 Generalized likelihood ratio test.......................................................................................60
4.6 T-test result......................................................................................................................60
 1

CHAPTER ONE

INTRODUCTION

1.1 Background of the Study

The struggle for food is desperate for the 240 million people of West Africa: one of

every three of who is a Nigerian (West Africa Rice Development Association (WARDA),

2002). Rice (Oryza sativa) is the staple food of approximately half of the world’s population

(International Rice Research Institute (IRRI), 1997). It has traditionally been an important

basic food commodity for certain populations in sub-Saharan Africa (SSA), and West Africa

in particular. Recent important and major changes have led to a structural increase in rice

consumption in the sub-region. Since 1973, regional demand has grown at an annual rate of

6%, driven by a combination of population growth and substitution away from traditional

coarse grains (WARDA, 2001). The consumption of traditional cereals, mainly sorghum and

millet, has fallen by 12 kg per capita, and their share in cereals used as food decreased from

61% in the early 1970s to 49% in the early 1990s. In contrast, the share of rice in cereals

consumed has grown from 15% to 26% over the same period (Akpokodje, Lancon and

Erenstein, 2001). Also it was observed that between 1961and 2005, the annual increase in

rice consumption was 4.52% in SSA (WARDA, 2007). Growth in regional rice consumption

remains high.

The demand for rice has been increasing at a much faster rate in Nigeria than in other

West African countries since the mid 1970s (WARDA, 2001). For example, during the

1960’s Nigeria had the lowest per-capita annual consumption of rice in the sub-region

(average of 3 kg). Since then, Nigeria’s per capita consumption levels have grown

significantly at 7.3% per annum (Ogundele and Okoruwa, 2006). Consequently, per-capita

consumption during the 1980s averaged 18 kg and reached 22 kg in 1995-1999. In an

 2

apparent move to respond to the increased per capita consumption of rice in Nigeria, local

production boomed, averaging 9.3% per annum (Ogundele and Okoruwa, 2006). These

increase in local production have been traced to vast expansion of rice area at an annual

average of 7.9% and to a lesser extent to increase in rice yield of 1.4% per annum. The reality

is that Nigeria has not been able to attain self-sufficiency in rice production despite increasing

hectares put into production annually (CBN, 2000). Consequently, Nigeria has depended

heavily on imported rice to meet her consumption needs and has become the World’s largest

importer of rice (WARDA, 2003). According to Okorji and Onwuka (1994) the rice import

bill for Nigeria was N123.61 million in 1980 and has since continued to rise. That Nigeria has

remained a net importer of rice with well over 150.15 billion spent annually (FOS, 2000) is

indicative of the declining self-sufficiency. This constitutes a huge drain on Nigeria’s foreign

reserve and a major bottleneck in the balance of payments (Egbuna, 2003).

In a bid to address the demand-supply gap for rice, governments have at various times

come up with policies and programmes. However, these policies according to WARDA

(2001) have not been consistent due mostly to oscillating import tariffs and import

restrictions. For instance, from 1986 to the mid-1990s imports were illegal. In 1995 imports

were allowed at a 100% tariff. In 1996 the tariff was reduced to 50% but increased to 85% in

2001. The erratic rice import policy reflects the dilemma of securing cheap rice for

consumers and a fair price for producers. Notwithstanding the various policy measures,

domestic rice production has not increased sufficiently to meet the increased demand. Even

during the rice import ban period, Nigeria was still importing several hundred thousand tons

of rice per year through illegal trade. With the removal of the rice import ban, consumption

resumed its rapid growth taking advantage of the downward trends of rice price on the world

market.
 3

This fluctuation and limited capacity of the Nigerian rice economy to match the

domestic demand has raised a number of pertinent questions both in the policy circle and

amongst researchers. Such questions include: what are the factors explaining the lag in

domestic rice production vis-a-vis demand for the commodity in Nigeria? Which strategy

could lead to a sustainable contribution of the Nigerian rice economy to the national food

security within a competitive and open economy? Central to this explanation is the issue of

efficiency of the rice farmers in the use of resources. West African Rice Development

Association (WARDA) in collaboration with the Nigerian Institute for Social and Economic

Research (NISER) suggested this strategy of increasing efficiency at producer level as one of

the key components to revitalize the rice sector in Nigeria (WARDA, 2003).

Efficiency is a very important factor of productivity growth in an economy where

resources are scarce and opportunities for new technologies are lacking. Inefficiency studies

have shown that it is possible to raise productivity by improving efficiency without

increasing the resource base or developing new technology (Adedapo, 2008). Farrel (1957)

decomposed economic efficiency into technical efficiency and allocative efficiency.

Technical efficiency is the ability to produce a given level of output with a minimum quantity

of inputs under a given technology (Tijani, 2006). Allocative efficiency refers to the ability to

choose optimal input levels for given factor prices.

All production is subject to uncertainty, but the risks associated with agricultural

production are particularly salient. Crop yields may be affected by the amount and timing of

rainfall, temperatures during the growing season, pests, diseases, hailstorms and fire among

many other factors (Nauges, O’Donnell and Quiggin, 2011). However, studies have shown

that the effect of these uncertainties on production can be investigated through the choice of

inputs on the output variance, otherwise known as production risk in inputs (Jaenicke,

Frechette and Larson 2003; Bokusheva and Hockmann 2006; Villano and Fleming 2006).
 4

This is because some inputs increase while others reduce the level of output variance

(production risk). Assessment of the efficiency of biological production sectors (such as rice

production) distils important policy implications, provided such assessment takes into

account the effects of input use on output variance (production risk), as this is vital for

agricultural development planning (Jaenicke, Frechette and Larson 2003; Villano and

Fleming 2006). The implication of this is that it is important to incorporate production risk in

the measurement of the efficiency of rice farmers.

Off-farm work by farm households is a persistent phenomenon throughout the world,

both in less developed and developing countries, and the dependence of farm families on the

income from off-farm work has increased steadily over the years (Chang and Wen, 2011) .

The importance of off-farm work has also been acknowledged in many countries. For

example, by using a random farmer survey in rural Ghana, Jolliffe (2004) reported that

approximately 74% of the farm households engaged in some form of off-farm work. Similar

evidence has also been found in Taiwan. Based on the statistics summarized from the

Agricultural Census data in 2001, approximately 75% of the farm households have reported

off-farm income.

In light of the increasing importance of off-farm income as a crucial determinant of

farm household well-being, a considerable body of literature has examined the roles of

household characteristics, the human capital of the farm operator and spouse, as well as farm

programs related to off-farm labour participation (e.g., El-Osta and Morehart, 2008; El-Osta,

Mishra and Morehart, 2007, 2008; Huffman and Lange, 1989; Lass, Findeis and Hallberg,

1991; Mishra and Goodwin, 1997). Attention has also been paid to the interaction between

the farm practice and the off-farm work of the farm household (e.g., Phimister and Roberts,

2006). It is expected that the increased reliance on off-farm employment affects the allocation

of family labour, and thus exerts an influence on farm productivity. On the other hand, off-
 5

farm work provides an opportunity for farm households to stabilize household income and

reduce the uncertainty associated with agricultural production. It is a generally held belief

that off-farm employment provides a risk management tool to reduce the income variability

associated with the farm household (El-Osta and Morehart, 2008; El-Osta et al., 2007).

Hence, this study becomes crucial in examining the interconnection among off farm work,

technical efficiency and rice production risk since increased output and productivity are

directly related to production efficiency and production risk.

1.2 Problem Statement

Rice is an important food crop in Nigeria and its consumption is growing particularly

among urban dwellers. Domestic production of this commodity has been inadequate and

unable to bridge the increasing demand-supply gap. Government efforts of making the

country self sufficient has not yielded the required results and thus the resort to importation

of the commodity. This constitutes a great drain in the country’s foreign reserve.

Government’s goal of achieving self sufficiency in rice production to a large extent will

depend on the level of farmers’ productivity which can be determined by rates of adoption of

improved technologies and efficiency of resource use. However, with the difficulties

encountered by farmers in developing countries for developing and adopting improved

technologies due to resource poverty, efficiency has become a very significant factor in

increasing productivity (Ali and Chandry, 1990).

Numerous studies (Obwona, 2006; Ogundele and Okoruwa 2006; Tijani, 2006; Al-

hassan, 2008; Nwaru and Iheke, 2010, Onoja, and Achike, 2010;) have attempted to

determine technical efficiencies of farmers in developing countries including Nigeria because

determining the efficiency status of farmers is important for policy purposes. However, few

of these studies took account of the presence of risk and the farmers’ responses to it,
 6

considering the common knowledge that economic units make their decisions under

uncertainty. This presents a research gap that needs to be filled. The presence of risk

influences not only production output but also producers’ behaviour, primarily with regard to

input use. If risk mitigation plays a principal role in decision-making, then a farm's technical

efficiency score may alter significantly. Therefore, technical efficiency assessment

considering a producer's response to uncertainty is not the same in a setting where no effect

of risk on input-use decisions is taken into account. Thus, when uncertainty is pervasive, the

theoretical framework for studying technical efficiency is to be extended with respect to risk

and producers’ responses to risk. In this study, production risk is assumed to be an important

factor that influences production decisions of farmers. Hence, the present study aims to

estimate the magnitudes of both technical inefficiency and production risk in rice production

in Enugu State.

In recent times farm families in general and small farm operators in particular have

been seeking off-farm employment in increasing numbers. A number of studies exist which

try to explain the reasons for this phenomenon. Some of the reasons observed by Lanjouw

(1999) include declining farm incomes and desire to insure against agricultural production

risk. In the beginning, increasing off-farm work by farm operators was viewed only as a

transitional phenomenon, but in recent years researchers have concluded that this change is

rather permanent. Economists have since turned their attention towards developing models of

multiple job holding (Bollman, 1979; Shiskhko and Bernard, 1976). Little attention has been

given to estimating the relative efficiency of rice farmers who take up off-farm work viz a viz

their counterpart who do not especially in Nigeria. While there is a growing realization that

part-time farming is becoming an integral part of Nigerian agriculture, these farms are also

likely to have the most staying power since they do not depend solely on farm income for

their viability. Some studies have also documented the impacts of off-farm work on farm
 7

productivity (Bagi, 1984; Kumbhakar et al., 1989; Sherlund et al., 2002; Smith, 2002). On

the other hand, studies on production risk in agricultural production have been extensive over

several decades and research interest in agricultural risk continues to grow. The relationship

between farmers’ off-farm work and production risks has also been examined. Mishra &

Goodwin (1997), for example, have demonstrated that farm households may treat off-farm

work as a vehicle to stabilize their income due to farm commodity prices being more variable

than off-farm wages. As predicted by production theory, a risk-averse farmer will allocate

labour and other resources to the less risky income sources (i.e., the off-farm work) until the

expected marginal returns between available activities are equal to each other. As a result, the

reduction in farm production risk may lead farmers to participate in the off-farm labour

market. In spite of this inter-relationship between off- farm work, production risk and

technical efficiency available literature reviewed showed that little or no study have been

done in Nigeria with particular emphasis on Enugu State to examine these connections. This

study intends to fill this gap. Also, there is need to assess the current levels of technical

efficiency of rice farmers in Enugu State and identify factors that affect their technical

efficiency. More importantly there is need to determine whether off- farm work improves

efficiency, whether there are differences between rice farmers with and without off- farm

work, and why.

This study will contribute to previous studies on off farm work by accommodating

technical efficiency and production risk simultaneously. The joint consideration of risk,

efficiency and off-farm work is important since the technical efficiency of each farmer, as

well as production risk, may affect the output response of crop production.

This study in a bid to fill the above-mentioned gaps will provide answers to the

following research questions:

 8

i. How does off-farm work influence technical efficiency and production risk of rice

farmers in Enugu State;

ii. Are there factors that affect the technical efficiency and production risk of rice

farmers with and without off-farm work and

iii. What are the constraints faced by rice farmers in the study area.

1.3 Objective of the Study

The broad objective of this study is to examine the effects of off-farm work on

technical efficiency and production risk among rice farmers in Enugu State, Nigeria. The

specific objectives are to:

i. describe the socio-economic characteristics of rice farmers with and without off-farm

work;

ii. describe rice production in the study area;

iii. determine the technical efficiency of rice farmers with and without off-farm work;

iv. compare the technical efficiency of rice farmers with and without off-farm work;

v. ascertain factors that affect technical efficiency and production risk of the two groups

of rice farmers;

vi. identify the constraints faced by the rice farmers in the study area and

vii. make policy recommendations based on the findings.

1.4 Hypotheses of the Study

i. farmers in the study are not technically efficient;

ii. production risk effects are not present in the model;

 9

iii. there is no significant difference in the technical efficiency of farmers with

and without off-farm work and

iv. socio-economic characteristics do not affect the technical efficiency of the two

groups of rice farmers.

1.5 Justification of the Study

Efficiency studies help countries to determine the extent to which they can raise

productivity by improving the neglected source, i.e., efficiency, with the existing resource

base and available technology (Al- Hassan, 2008). Such studies could support decisions on

whether to improve efficiency first or to develop a new technology in the short run. More

importantly, enhanced technical efficiency will not only enable farmers to increase the use of

productive resources, it will also give direction for the adjustments required in the long run to

achieve food sustainability.

It is now widely accepted that to meet the challenges of future agricultural policy

design in the future, we must understand farm household behaviour in a more complex policy

context that recognizes the diverse nature of farms with respect to farm size, farm production

and business organization. We must also recognize the increasing interconnection between

decisions made by farm businesses and farm households, and their effects on the well-being

of farm families (Kuhn and Offutt 1999; Offutt, 2002). Recommendations suggested from

such would boost the effectiveness and success of policies and programmes targeted at

increasing productivity. Given the increasing number of farm families taking up off-farm

work, the increasing rice demand-supply gap and the risk associated with rice production the

need for this study is justified. The recommendations from this study are expected to give

direction to policy makers in designing appropriate policies to increase rice productivity in

Enugu State and Nigeria as a whole. This is because more knowledge about whether farm
 10

productivity and efficiency are affected by part-time farming could help policy makers

introduce better rural development policies.

Also, farmers will find this work very useful as it will help increase their productivity

as they gain understanding of the interconnection between off-farm work, technical efficiency

and rice production risk. Moreover, the results from this study will provide baseline

information to researchers who want to undertake further studies in this area. Other

stakeholders who can benefit from the results of this study are donor agencies, research

stations and even extension workers.

 11

CHAPTER TWO

REVIEW OF RELATED LITERATURE

Literature was reviewed under the following major sub-headings:

Characteristics of Nigeria’s rice production systems

Off-farm work by farm households

Determinants of choice of off farm work

Interrelationship between off-farm work and efficiency

Agricultural risk

Classifications of agricultural risks

Efficiency

Determinants of production efficiency

Measures of efficiency

Estimating production risk and inefficiency simultaneously

Review of Related Empirical Studies

Theoretical Framework

Analytical Framework

Estimating the stochastic frontier model with risk

2.1 Characteristics of Nigeria’s Rice Production Systems

Rice can be grown over a wide range of edaphic and ecological conditions. In order

to formulate a national strategy and action plan for increasing rice production, due

cognizance must be made of these widely varying conditions. The prevalent types of rice

production systems in Nigeria include rainfed upland, rainfed lowland and irrigated lowland

(WARDA, 2001). Other less common rice production systems include deep water and

mangrove rice (Singh et al., 1997). Rice farmers tend to be small-scale, with farms of 1-2 ha.
 12

2.2 Off-Farm Work by Farm Households

Part time and multiple job holdings among the farm household members are not new,

and have been evident for United States farms for over fifty years. Many farm households

combine farming with a variety of other pursuits, since off farm work by farm operators and

their spouses has traditionally been viewed as an action necessary to save the farm by

providing resources to pay farm bills or to repay debt (Ahearn and Lee 1991). In past years,

off-farm employment was also considered temporary and as an income supplement (Mishra,

El-Osta, Morehart, Johnson, and Hopkins, 2002).

In south eastern Nigeria where population pressure is a problem of accelerating

importance, the agricultural land is in short supply thereby necessitating off-farm jobs to

supplement farm income. There is therefore, a high incidence of part-time farming in the

region (Okafor, 1982). For most Nigerian farmers, farming is a seasonal occupation except in

areas where some form of irrigation is practiced; farmers therefore endeavour to supplement

their income with petty jobs outside farming (Ekong, 2003). According to Amao (2008),

early in the 20th century, farming household did little off-farm work because the costs of

such participation were prohibitive. Farm households relied on farming as their primary and

usually sole source of income (Brewster, 1979).

Some widely known off–farm occupations include the following; saw milling,

pottery, weaving, carving, leather works, carpentry, bicycle repairing, black smiting, knitting

and dressmaking, dyeing, retailed trading, barbing, hair dressing, entertainment, drinking

parlour operation, teaching, bricklaying and house construction, midwifery native doctoring,

preaching, transport operation etc. (Ekong, 2003). The growing uncertainties of farming

together with increased opportunities for off–farm work have led to a new arrangement of

combining off–farm work with farming (Godwin and Marlowe, 1990). This now makes the

farm household members to increase their participation in off– farm work at the expense of
 13

farming when the marginal returns to the former become larger than the marginal returns to

the latter (USDA, 2001).

2.3 Determinants of Choice of Off-Farm Work

Earlier findings typically show that farmers’ characteristics (such as farmer’s age,

experience, marital status, education) affects share of farmer’s time off farm (Huffman, 1980;

Sumner, 1982; Serra, Goodwin, & Featherstone, 2004, 2005; Ahituv and Kimhi, 2006: Lien,

Flaten, Jervell, Ebbesvik, Koesling, and Valle, 2006). Previous studies also typically show a

negative association between farm size and likelihood of off-farm work (Mishra & Goodwin,

1997; Ahituv and Kimhi, 2002; Serra et al., 2005; Benjamin and Kimhi, 2006). Farm

characteristics such as yields, region, and urbanization may also influence the household time

allocation decision (Sumner, 1982; Mishra and Goodwin, 1997; Goodwin and Mishra, 2004;

Serra et al., 2005). Lieu, Kumbhakar and Hardaker (2007) found out that farmers who are

single generally worked less frequently off the farm than non-single farmers. The presence of

children in the farm household was found to have a significant negative effect on the off-farm

activities of farmers and their spouses, supporting some earlier findings (Mishra and

Goodwin, 1997; Goodwin and Mishra, 2004) but contradicting the results of Ahituv and

Kimhi (2006) who found that the presence of children did not significantly affect the supply

of off farm labour. In addition to these established factors, it is also assumed in this study that

farmer and farm characteristics will exogenously affect the extent of off-farm work. Hence,

as shown in Figure 2.1, in stage 1 we assume that farmer and farm characteristics affect the

share of farmers’ time devoted to off-farm work.

 14

Farm other input use

characteristics

Farm productivity

and efficiency

Share of farmers’ time

Farmer characteristics off farm (stage 1)

Figure 2.1: Assumed relationships between farmer and farm characteristics, share of farmer’s

time off farm, other input use, and the farm productivity and efficiency (adapted from Lieu,

Kumbhakar and Hardaker, 2007)

2.4 Interrelationship between off-farm work and efficiency

Much research has investigated the economic efficiency of farm households. In

general, previous investigations have focused on the efficiency of farm activities. Yet, off

farm activities can contribute to significant improvements in the welfare of agricultural

households (Hill, 2000). This is true in developed as well as developing countries. For

example, Gardner (2002) documented how the growth of off-farm income in the US over the

last 40 years reduced income inequality in agriculture and contributed to the catch-up of

farmers’ incomes with those of the off-farm population. Phimister and Roberts (2002) found

evidence of significant linkages between off-farm work and farm decisions in Scotland. In the

context of Africa, Reardon, Delgado and Matlon (1992) and Reardon (1997) documented the

importance of non-farm earnings for African rural households. For example, Reardon (1997)

reports estimates of non-farm income as a share of total household income ranging from 22 to

93 percent, with an average of 45 percent. In Africa, considerable income diversification

 15

between farm and off-farm activities may be seen as a response to poorly functioning capital

markets: the cash from non-farm earnings can help stimulate farm investments and improve

agricultural productivity (Haggblade, Hazell and Brown, 1989; Hazell and Hojjati, 1995).

Given that very poor households often lack access to non-farm income (Reardon et al., 1992),

imperfections in the labor market can contribute both to inefficient labor allocation in rural

households and to a more unequal distribution of income. This stresses the need to include

off-farm activities in the analysis of farm household efficiency. This appears particularly

important for poor African rural households where incomes are low and even small amounts

of inefficiency can have large impacts on incomes and welfare.

2.5 Agricultural risks

It is often said that agriculture production is a risky business, that is, it is subject to

risk. This means that due to complexities of physical and economic systems, the outcomes of

farmers’ actions and production decisions are uncertain, and many possible outcomes are

usually associated with a single action or production plan. The uncertainty concerning

outcomes that involve some adversity or loss that negatively affects individual well-being is

normally associated with the idea of risk (Organisation for Economic Cooperation and

Development (OECD), 2009). Some authors have made a distinction between uncertainty as

imperfect knowledge and risk as exposure to uncertain unfavourable economic consequences

(Hardaker, Huirne, Anderson and Lien 2004). In practice both concepts are very much related

and are used interchangeably, one with more emphasis on ―probabilities as the description

of the environment, and the other with more emphasis on the ―potential negative impact

on welfare. There is no risk without some uncertainty and most uncertainties typically imply

some risk.
 16

2.5.1 Classifications of Agricultural Risks

OECD (2000) differentiated between risks that are common to all businesses (family

situation, health, personal accidents, macroeconomic risks etc.) and risks that affect

agriculture more specifically: production risk (weather conditions, pests, diseases and

technological change), ecological risks (production, climate change, management of natural

resources such as water), market risks (output and input price variability, relationships with

the food chain with respect to quality, safety, new product etc.) and finally regulatory or

institutional risk (agriculture policies, food safety and environmental regulations).

Huirne, Meuwissen, Hardaker, and Anderson (2000) and Hardaker et al. (2004)

distinguish two major types of risk in agriculture. First, business risk includes production,

market, institutional and personal risks. Production risk is due to unpredictable weather and

performance of crops and livestock. Market risk is related to uncertainty about the price of

outputs and, sometimes also inputs, at the time production decisions are taken. Institutional

risk is due to government actions and rules such as laws governing disposal of animal manure

or the use of pesticides, tax provisions and payments. Personal risks are due to uncertain life

events such as death, divorce, or illness. Second, financial risks result from different methods

of financing the farm business. The use of borrowed funds means that interest charges have to

be met before equity is rewarded which may create risk due to leverage. Additionally there is

financial risk when interest rates rise or loans are unavailable.

Musser and Patrick (2001) follow Baquet, Hambleton, and Jose (1997) and define five

major sources of risk in agriculture. Production risk concerns variations in crop yields and in

livestock production due to weather conditions, diseases and pests. Marketing risk is related

to the variations in commodity prices and quantities that can be marketed. Financial risk

relates to the ability to pay bills when due, to have money to continue farming and to avoid

bankruptcy. Legal and environmental risk concerns the possibility of lawsuits initiated by
 17

other businesses or individuals and changes in government regulation related to environment

and farming practices. Finally, human resources risk concerning the possibility that family or

employees will not be available to provide labour or management.

2.6 Efficiency

Efficiency is one of the most important topics of economic theory. Arene (2008)

defined efficiency as the relative performance of the processes used in transforming a set of

inputs into output. del Hoyo, Espino, and Toribio (2004) defined efficiency as the

relationship between what an organization (producer, production unit, or any decision making

unit) produces and what it could feasibly produce, under the assumption of full utilization of

the resources available. Within the latter conceptual framework, as stated by Kumbhakar and

Lovell (2000) efficiency represents the degree of success which producers achieve in

allocating the available inputs and the outputs they produce, in order to achieve their goals

namely; to attain a high degree of efficiency in cost, revenue, or profit. As stated in del Hoyo

et al., (2004) and Kumbhakar and Lovell (2000), efficiency is the ability of a decision making

unit to obtain the maximum output from a set of inputs (output orientation) or to produce an

output using the lowest possible amount of inputs (input orientation).

Olayide and Heady (1982) defined agricultural production efficiency as the index of

the ratio of the total farm output to the value of the total inputs used in farm production. A

production frontier refers to the maximum output attainable by given sets of inputs and

existing production technologies. The production frontier defines the technical efficiency in

terms of a minimum set of inputs in order to produce a given output or a maximum output

produced by a given set of inputs. This approach involves selecting the mix of inputs which

produces a given quantity of output at a minimum cost, namely the production frontier. If

what a producer actually produces is less than what it could feasibly produce then it will lie
 18

below the frontier. The distance by which a firm lies below its production frontier is a

measure of the firm’s inefficiency (Bera and Sharma, 1999). A graphical illustration of a

production efficiency frontier is shown in figure 2.2.

Output

Z X

Input bundle

Figure 2.2: Production Frontier

Source: (Ogunyinka and Ajibefun, 2003)

A farm, for example, at point X refers to the inefficient farm, while points Y and Z are

both efficient because they are on the frontier. The farm at point X should therefore move

upward to point Y or backward to point Z in order to be efficient. If its movement is toward

Y, more output is obtained with the same amount of inputs or if it is toward Z, fewer amounts

of inputs yield the same output. Both cases depict more technical efficiency than the initial

position X.

Farrell (1957) was the first to empirically measure productive efficiency in terms of

deviations from an ideal frontier. He also proposed a decomposition of economic efficiency

into:
 19

a) technical efficiency (TE), which measures the ability of a firm to obtain the maximum

output from given inputs, and

b) allocative efficiency (AE), which measures the ability of a firm to use inputs in optimal

proportions given their prices:

Economic efficiency is the product of technical efficiency and allocative efficiency

(Technical efficiency * allocative efficiency)

If the only information available are input and output quantities, and there is no information

on input or output prices, then the type of efficiency that can be measured is technical

efficiency. If price information on inputs and outputs is available, in addition to input and

output quantities, then the type of efficiency that can be measured is allocative efficiency.

Profit maximisation requires a firm to be both technically efficient (by producing the

maximum output given the level of inputs employed), as well as allocative efficient (by using

the right mix of inputs, or producing the right mix of outputs given their relative prices,

respectively (Kumbhakar and Lovell, 2000). Nevertheless, in real economic life, producers

are hardly fully productive efficient. The difference can be explained in terms of technical

and allocative inefficiencies, as well as a range of unforeseen exogenous shocks, making it

unlike all (or even any) producers, firms or, even, economies operate at the full efficiency

frontier (Reifschneider and Stevenson, 1991). However, one of the main related questions is

whether inefficiency occurs randomly, or whether some economic agents (producers, firms or

economies) have predictably higher levels of inefficiency than others. That is the reason why

estimating efficiency is one of the core tools of economic analysis. Firstly, efficiency

estimation provides an indication of the percentage by which potential output could be

increased, or potential cost could be decreased, in relation to the corresponding production

frontier. The further below the frontier a producer lies, the more inefficient it is.
 20

The type of efficiency that can be measured using a production possibility frontier is

technical efficiency. The technical efficiency of an individual farm is defined in terms of the

ratio of the observed output to the corresponding frontier output, conditioned on the level of

inputs used by the farm. Technical inefficiency is therefore defined as the amount by which

the level of production for the farm is less than the frontier output.

TE = Yi/Yi* where Yi* = f (xi; β ) is highest predicted value for the ith farm

Technical inefficiency = 1- TE

2.7 Determinants of production efficiency

Hayami and Ruttan (1970) found that educational level was an important determinant

of differences in agricultural productivity among countries. In a survey of research on

education and farmer productivity, Lockhead, Jamison and Lan, (1980) confirmed that

education had a positive effect on farmers’ efficiency in all 37 data sets included in their

review. Kalirajan and Shand (1985), in their study of high-yielding varieties (HYVs) of

paddy in India, found that although schooling is productive for the individual, a farmer’s

education is not necessarily significantly related to yield. They argue that a farmer may gain

improved knowledge of the technology with experience of using it or by observing others.

Kalirajan (1981) estimated a stochastic frontier Cobb–Douglas production function using data

from 70 rice farmers for the rabi season in a district in India. The variance of farm effects was

found to be a highly significant component in describing the variability of rice yields.

Kalirajan (1981) proceeded to investigate the relationship between the difference between the

estimated maximum yield function and the observed rice yields and such variables as

farmer’s experience, educational level, number of visits by extension workers, etc. In this

second-stage analysis, Kalirajan (1981) noted the policy implications of these findings for

improving crop yields of farmers. Ali and Flinn (1989) estimated a stochastic profit frontier
 21

of modified translog type for Basmati rice farmers in Pakistan’s Punjab. After estimating the

technical efficiency of individual farmers, the losses in profit due to technical inefficiency

were obtained and regressed on various farmer- and farm-specific variables. Factors that were

significant in describing the variability in profit losses were level of education, off-farm

employment, unavailability of credit, and various constraints associated with irrigation and

fertilizer application. Kalirajan and Shand (1989) estimated the time-invariant panel-data

model using data for Indian rice farmers over five consecutive harvest periods. The farm

effects were found to be a highly significant component of the variability of rice output. A

regression of the estimated technical efficiencies on the farm- or farmer-specific variables

indicated that farming experience, level of education, access to credit and extension contacts

had significant influences on the variation of farm efficiencies.

2.8 Measures of Efficiency

Regarding that the production frontier cannot be observed directly, several techniques

have been developed in order to estimate technical efficiency or technical inefficiency. If one

collects farm-level data that can be used in linear programming, then one can use Data

Envelopment Analysis (DEA) to estimate technical efficiency (Makombe, Namara, Hagos,

Awulachew, Ayana, Bossio, 2011). If one collects data that can be used for regression

analysis, then one can use the stochastic frontier production function and use the residuals to

estimate technical inefficiency. As broadly described in del Hoyo et al (2004) and

Kortelainen (2008), the main methods of production frontiers and efficiency estimation may

be classified into two core groups:

a) non-parametric models, regarding Data Envelopment Analysis, developed by Farrell

(1957) and Charnes, Cooper and Rhodes, (1978), and

 22

b) parametric models, regarding Deterministic Frontier Analysis and Stochastic Frontier

Analysis, developed by Aigner, Lovell and Schmidt, (1977) and Meeusen and Van den

Broeck (1977).

2.8.1 The non-parametric programming approach

The programming approach (Farrell, 1957; Afriat, 1972; Hanoch and Rothchild,

1972; Diewert and Parkan, 1983) requires one to construct a free disposal convex hull in the

input-output space from a given sample of observations of inputs and output. The convex

hull, which is generated from a subset of the given sample, serves as an estimate of the

production frontier, depicting the maximum possible output. Now, a measure of production

efficiency of an economic unit (farm) is measured as the ratio of the actual output to the

maximum possible output on the convex hull, corresponding to the given set of inputs. The

non parametric approach has the advantage of imposing no a priori parametric restrictions on

the underlying technology (e.g., Fare, Grosskopf and Lovell, 1985; Seiford and Thrall, 1990).

A major criticism of this approach is that the convex hull, representing the maximum

possible output, is derived using only marginal data rather than all the observations in the

sample. Thus the technical efficiency measures are susceptible to outliers and measurement

errors (Forsund, Lovell and Schmidt, 1980). Second, the method has very demanding data

needs. Finally, being non-parametric, no statistical inferences on the estimates can be carried

out.

2.8.2 Deterministic production frontier models and technical efficiency

Aigner and Chu (1968) were the first researchers to estimate a deterministic frontier

production function using Cobb-Douglas production function. They argued that, within a

given industry, firms might differ from each other in their production processes, due to

certain technical parameters in the industry, due to differences in scales of operation or due to
 23

organizational structures. Under this assumption, they considered a Cobb - Douglas

production function, with an empirical frontier production model such as:

qit < f(xit) (2.1)

This equation defines a production relationship between inputs, x, and output qit, in

which for any given x, the observed value of qit must be less or equal to f(xit). Since the

theoretical production function is an ideal (the frontier of efficient production), any non-zero

disturbance is considered to be the result of inefficiency, which must have a negative effect

on production function:

qit = f(xit) - uit, i = 1, 2, 3, 4,....... N, t = 1....., N (2.2)

Taking natural logarithms, the model becomes:

ln qit = β 0 + ln xit β - uit (2.3)

where:

1. ln qit is the natural logarithm of the output of the ith firm;

2 ln xit is the natural logarithms of inputs;

3. β is a column vector of the unknown parameters to be estimated;

4. uit is a non negative random variable associated with technical inefficiency, representing

the shortfall of actual output from its maximum possible value.

According to Aigner and Chu (1968) technical efficiency for the ith firm is defined as

the ratio of the observed output for the ith firm relative to the potential output (frontier

function):

TEit = = = (2.4)

= exp(- uit), 0 < TEit < 1 (2.5)

And
 24

uit = ln(TEit) (2.6)

Technical efficiency measure takes a value between zero and one:

 TEit = 1 shows that the producer is fully productive efficient and, correspondingly, the

observed output qi reaches its maximum obtainable value,

 TEit < 1 provides a measure of the shortfall of the observed output from maximum

feasible output.

Letting:

TEit = exp(-uit), 0 < TEit < 1 (2.7)

will ensure that the observed output lies below the frontier, that is:

qit < f(xit β ) (2.8)

Nevertheless, in this case, the model is deterministic, and all deviations from the

frontier are assumed to be the result of technical inefficiency and no account is taken of any

measurement errors (i.e. errors associated with the choice of functional form) or any

statistical noise (i.e. omission of relevant variables from the vector xit).This approach is dealt

by the Stochastic Production Frontier models (Aigner and Chu, 1968).

2.8.3 Stochastic production frontier models and technical efficiency

In the decade of 1970, deterministic production frontier model was extended by Afriat

(1972), and more systematically by Aigner et al., (1977) and Meeusen and Van den Broeck

(1977). Based on the literature commencing with theoretical work by Debreu (1951) and

Farrell (1957), Aigner et al., (1977) and Meeusen and van den Broeck (1977) extended the

deterministic frontier approach in order to account not only for technical efficiency, but also

for any measurement errors or any statistical noise. They developed a statistically and

theoretically sound method for measuring efficiency, different in the sense that it allows

random events to contribute to variations in producer output. Since the introduction of

stochastic frontier analysis, it has been widely accepted that frontier models provide a number
 25

of advantages over non frontier models (Forsund et al., 1980; Bravo-Ureta and Pinherio,

1993). The economic literature on efficiency and stochastic frontier analysis has been rather

extensive with numerous studies. To name just a few, there are influential research papers by

Forsund et al. (1980), Greene (1993, 1997), Bauer (1990), Battese (1992), Schmidt (1985),

Cornwell and Schmidt (1996), Kalirajan and Shand (1999), and Murillo-Zamorano (2004), as

well as book length approaches, including Coelli et al. (1995), Coelli et al. (1998),

Kumbhakar and Lovell (2000) and Fried, Lovell and Schmidt (2008).

Aigner et al. (1977) and Meeusen and Van den Broeck (1977) proposed, almost

simultaneously, but independently, a formulation within which observed deviations from the

production function could arise from two sources: a) productive inefficiency, that would

necessarily be negative, and b) effects specific to the firm, that could be of either sign. In

order to incorporate this feature, there is need to introduce another random variable

representing any statistical noise or measurement errors. In order to capture this, the

stochastic model includes a composite error term that sums a two-sided error term, measuring

all effects outside the firm’s control, and a one-sided, non-negative error term, measuring

technical inefficiency. The resulting frontier is presented in terms of a general production

function, known as a stochastic production frontier.

The general model is specified as:

Yi = f ( X i β ) exp(vi − ui ) (2.9)

Where Yi = output,

X i = input,

β = vector of other factors affecting output, and

vi , ui = error components.
 26

The error component vi is a random error with zero mean, which reflects random

deviations due to factors outside the control of the production unit. Influence of weather

factors and economic conditions set exogenously as well as measurement errors could be

included in this error component. The second component of the disturbance term, ui, is

restricted to be non-negative, and is attributed to firm effects, resulting from inefficiencies

due to factors within the control of each firm. While the random errors, vi, were assumed to

be independently and identically distributed (i.i.d.) as N(0, ) random variables, symmetric,

and distributed independently of ui. The technical efficiency, effects, ui, follows distributions

such as half normal, truncated normal, exponential, and gamma distributions. Decomposition

of the disturbance term allows for better identification of actual technical

efficiency/inefficiency.

The stochastic econometric approach enables to attempt to distinguish the effects of

noise and inefficiency, thereby providing the basis for statistical inference. Providing

estimates of producer - specific technical efficiency, which is the ultimate objective of the

estimation process in addition to obtaining estimates of the production technology parameters

β in f ( X i β ) , requires an extraction of separate estimates of statistical noise vi and technical

inefficiency ui, form the estimates of i ( i = vi-ui) for each producer. Therefore, the

distributional assumptions of the inefficiency term are required to estimate the technical

inefficiency of each producer.

The assumption of a certain inefficiency distribution as well as a normal noise

distribution suggests the use of maximum likelihood estimation method (Behr and Tente,

2008), one of the commonly used methods of estimating the parameters of a stochastic

frontier.
 27

2.8.4 Stochastic Frontier model estimation – The Maximum Likelihood Estimation

Method
Aigner et al. (1977) were the first researchers to estimate the unknown parameters of

the stochastic frontier model using the method of maximum likelihood followed also widely

in later decades by Greene (1982) and Coelli (1995), among others. Maximum likelihood

estimation is a popular statistical method used for fitting a mathematical model to real world

data. The concept of maximum likelihood estimation is based on the idea that a particular

sample of observations is more likely to have been generated from some distributions than

from others. Consequently, the maximum likelihood estimate of an unknown parameter is

defined to be the value of the parameter that maximizes the probability (or likelihood) of

randomly drawing a particular sample of observations. The maximum likelihood estimates

can be computed using the statistical package STATA 11.

The stochastic frontier model can be estimated using one-step or two-step approaches.

In the two-step procedure, the frontier production function and the firm’s efficiency levels

(TE) are estimated first [ Yi = xi β + (vi − ui ) ], ignoring a set of variables ( z ) that affect

technical efficiency. In the second step, efficiency levels (TE) are regressed against the

variables ( z ) to see how efficiency levels vary with these variables ( TEi = z i δ ). However,

the two-stage procedure has long been recognised to provide biased results because the model

estimated at the first step is misspecified (Coelli et al., 1998; Wang and Schimidt, 2002). A

one-step procedure is suggested to solve the bias problem, in which the relationship between

technical efficiency and the variables is imposed directly in estimating the frontier production

function and the firm’s efficiency levels (Kumbhakar et al., 1991; Wang and Schimidt, 2002).
 28

2.9 Estimating Production Risk and Inefficiency Simultaneously

The specification of residuals from a production model's deterministic portion plays a

central role in two generally separate analytic frameworks, each dating from the late 1970s.

Just and Pope (JP) focused on production risk, measured by the variance of output, and

suggested use of production function specifications satisfying some desirable properties. Risk

analysis in a Just and Pope (1978, 1979) framework involves recovering the residuals and

using them to investigate the marginal effects of inputs on production risk, or noise. One of

the specification requirements of the Just-Pope framework is that there be no a priori

restrictions on the marginal risk effects so that an input to production could be either risk-

increasing or risk-decreasing. Recent empirical applications include Asche and Tveterbs

(1999); Smale et al., (1998); Traxler et al., (1995); and Tveterbs, (1999) among others.

In contrast, inefficiency analysis in a stochastic frontier framework (Aigner, Lovell,

and Schmidt, (1977); Battese and Corra, (1977); Meeusen and van den Broeck, (1977))

involves specifying the residuals with both a two-sided white noise component and a one-

sided inefficiency component. Recent empirical applications (including Morrison-Paul,

Johnston, and Frengley (2000); Goaied and Ayed-Mouelhi (2000); Cuesta; Reinhard, Lovell,

and Thijssen; Battese and Broca (1997); and Ahmad and Bravo-Ureta (1996), which

generally focused only on the inefficiency component, estimate either technical inefficiency

or the marginal effects of inputs or exogenous factors on inefficiency.

Recently, however, researchers have noted that the stochastic frontier model is

compatible with the Just-Pope model and have begun combining risk and inefficiency

analysis in a single framework [Wang (2002); Kumbhakar (1993, 2002); Kumbhakar &

Lovell (2000); Battese, Rambaldi, & Wan (1997)]. By conducting a risk analysis within a

stochastic frontier framework, one can investigate whether inputs remain risk-increasing (or -

decreasing) even after accounting for inefficiency. Use of a combined framework, which
 29

simultaneously accounts for inefficiency and marginal impacts on production risk, may affect

empirical results based on a Just-Pope model, which addresses only risk. For instance, using a

Just-Pope framework, Tveterds (1999) found labor to be risk-reducing but capital to be risk-

increasing in Norwegian salmon farming. However, while investigating a similar application

but using a stochastic frontier framework, Kumbhakar (2002) found partially conflicting

results: both capital and labor were risk-reducing.' This partial reversal of Tveterds' results

therefore points toward a likely mutuality between risk and inefficiency effects associated

with input use that is revealed only through Kumbhakar's systematic accounting of two

components of the residuals-noise and inefficiency.

2.10 Review of Related Empirical Studies

Some studies have documented the impacts of off-farm work on farm productivity

(e.g., Bagi, 1984; Kumbhakar et al., 1989; Sherlund et al., 2002; Smith, 2002). For instance,

by estimating a stochastic production frontier model, Kumbhakar et al. (1989) examined the

effects of off-farm work on the farm level efficiency of dairy farms in the state of Utah. Their

results show that off-farm work is negatively associated with technical efficiency. Using a

similar approach to the vegetable farm survey in Florida, Fernandez-Cornejo (1992) obtained

similar results. In addition, Goodwin and Mishra (2004) used the gross cash income over the

total variable costs as a simple proxy for the farm’s efficiency and analyzed the relationship

between the off-farm labour supply and farm productivity. Their results showed that those

farm households that work off the farm are less efficient.

Some studies also have been carried out to ascertain the determinants of technical

efficiency. For instance, a study by Battese and Coelli (1995) on paddy rice farms in

Aurepalle India used panel data for 10 years and concluded that older farmers were less

efficient than the younger ones. Farmers with more years of schooling were also found to be
 30

more efficient but declined over the time period. Battese, et al., (1996) used a single stage

stochastic frontier model to estimate technical efficiencies in the production of wheat farmers

in four districts of Pakistan ranging between 57 and 79 percent. The older farmers had

smaller technical inefficiencies. Bedassa and Krishnamoorthy (1997) used a two-step

approach to estimate technical efficiency in paddy farms of Tamil Nadu in India. They

concluded that the mean technical efficiency was 83.3 percent, showing potential for

increasing paddy production by 17 percent using present technology. Small and medium-

scale-farmers were more efficient than the large-scale farms. In addition, the study concluded

that animal power was over utilized and therefore suggested reduction. However, the paddy

rice farmers could still benefit by increasing the fertilizer use and expansion of land.

In measuring technical efficiency of maize producers in Eastern Ethiopia for farmers

within and outside the Sawakawa–Global 2000 project, Seyoum, et al., (1998) used a translog

stochastic production frontier and a Cobb-Douglas production function. Some of the key

conclusions from this study were that younger farmers are more technically efficient than the

older farmers. In addition, farmers with more years of school tended to be more technically

efficient. On the other hand, those that obtained information from extension advisers tended

to reduce the technical inefficiency. The mean technical efficiency of farmers within the SG

2000 project was estimated to be 0.937 while the estimate of the farmers outside the project

was 0.794. However, this study should have squared the age to address the linear relationship

of the age variable.

A study by Wilson, et al., (1998) on technical efficiency in UK potato production

used a stochastic frontier production function to explain technical efficiency through

managerial and farm characteristics. Mean technical efficiency across regions ranged from 33

to 97 percent. There was high correlation between irrigation of the potato crop and technical

efficiency. The number of years of experience in potato production and small-scale farming
 31

were negatively correlated with technical efficiency. A study by Liu and Zhuang (2000) on

technical efficiency in post-collective Chinese Agriculture concluded that 76 and 48 percent

of technical inefficiency in Sichuan and Jiangsu, respectively, could be explained by

inefficiency variables. They used a joint estimation of the stochastic frontier model. Awudu

and Huffman (2000) studied economic efficiency of rice farmers in Northern Ghana. Using a

normalized stochastic profit function frontier, they concluded that the average measure of

inefficiency was 27 percent, which suggested that about 27 percent of potential maximum

profits were lost due to inefficiency. This corresponds to a mean loss of 38,555 cedis per

hectare. The discrepancy between observed profit and frontier profit was due to both

technical and allocative efficiency. Higher levels of education reduced profit inefficiency

while engagement in off-farm income earning activities and lack of access to credit

experience higher profit inefficiency. The study also found significant differences in

inefficiencies across regions.

Awudu and Eberlin (2001) used a translog stochastic frontier model to examine

technical efficiency in maize and beans in Nicaragua. The average efficiency levels were 69.8

and 74.2 percent for maize and beans, respectively. In addition, the level of schooling

represented human capital, access to formal credit and farming experience (represented by

age) contribute positively to production efficiency, while farmers’ participation in off-farm

employment tended to reduce production efficiency. Large families appeared to be more

efficient than small families. Although a larger family size puts extra pressure on farm

income for food and clothing, it does ensure availability of enough family labour for farming

operations to be performed on time. Positive correlation between inefficiency and

participation in non-farm employment suggests that farmers reallocate time away from farm-

related activities, such as adoption of new technologies and gathering of technical

information that is essential for enhancing production efficiency. The result indicated that
 32

efficiency increased with age until a maximum efficiency was reached when the household

head was 38 years old. The age variable probably picks up the effect of physical strength as

well as farming experience for the household head.

In a study by Wilson, et al, (2001) a translog stochastic frontier and joint estimate

technical efficiency approach was used to assess efficiency. The estimated technical

efficiency among wheat producers in Eastern England ranged between 62 and 98 percent and

found farmers who sought information, and had more years of managerial experiences and

had large farm, were associated with higher levels of technical efficiency. A study by

Mochebelele and Winter-Nelson (2002) on smallholder farmers in Lesotho used a stochastic

production frontier to compare technical inefficiencies of farmers who sent migrant labour to

the South African mines and those who did not. They concluded that farmers who send

migrant labour to South African are closer to their production frontier than those who do not.

Belen, et al., (2003) made an assessment of technical efficiency of horticultural

production in Navarra, Spain. They estimated that tomato producing farms were 80 percent

efficient while those that raised asparagus were 90 percent efficient. Therefore, they

concluded that there exists a potential for improving farm incomes by improving efficiency.

Gautam and Jeffrey (2003) used a stochastic cost function to measure efficiency among

smallholder tobacco cultivators in Malawi. Their study revealed that larger tobacco farms are

less cost inefficient. The paper uncovered evidence that access to credit retards the gain in

cost efficiency from an increase in tobacco acreage. This suggested that the method of credit

disbursement was faulty.

Bravo-Ureta, et al., (1994) concluded that Paraguan cotton had 40.1 percent average

economic efficiency while cassava producers were 52.3 percent efficient. They concluded

that there was room for improvement in productivity for these basic crops. However they did

not find a relationship between economic efficiency and socioeconomic characteristics. This
 33

observation was explained by the possibility of existence of a stage of development threshold

below which this type of relationship is not observed. In this case the sampled Paraguan

farmers were yet to reach the threshold.

2.11 Theoretical Framework

2.11.1 The Production Function, Risk, and Technology

A simple theoretical framework is built on the conventional agricultural household

production model that examines the labor allocation decisions (e.g., El-Osta et al., 2007,

2008; Hallberg et al., 1991) by accommodating production risk and technology. There are

fixed endowments of operator time (Ē), and time was allocated to leisure (l), farm production

(L), and off-farm work (Lm). The household receives income from agricultural sales as well

as the paid salaries from the off-farm jobs. Following Kumbhakar (2002), the production

function is a function of farm labor and is specified as: F(L) = f (L)+g(L)ε −h(L)u. The

functions f (.) and g(.) specify the effects of inputs on the mean level of output and

production risk, respectively. The error associated with output risks, ε, is assumed to follow

an arbitrary distribution of ε ~ i.i.d. (0, ). An input is regarded as risk increasing

(decreasing) if g (.) is positive (negative). Production efficiency is reflected in h(.)u, where u

~ i.i.d.( ū, ) is the random noise on a stochastic production frontier function.

2.11.2 Farm Household’s Maximization Problem

According to Chang and Wen (2011) utility of the farm household depends on the

consumption (C) and leisure (l), and the farm households maximize their expected utility

subject to the total available income and a time constraint

Max = EU(C, l) (2.10)

C,l

s.t.
 34

C = P * [f (L) + g(L)ε − h(L)u] + w * Lm. (2.11)

Ē = L + l + Lm (2.12)

where EU(.) is the expected utility of each farm, and P and w represent the price of the

agricultural product and the equilibrium off-farm wage rate, respectively. To solve the model,

we first substitute Eqs. (11) and (12) into Eq. (10), and this yields

Max = EU{{P * [f (L) + g(L)ε − h(L)u] +w * Lm}, { Ē − L − Lm}} (2.13)

L,Lm

The first-order necessary Kuhn-Tucker conditions for the optimal locations of time

allocated in farm production and off farm work are

∂EU(.) = ∂EU(.) * P * (fL + gLε − hLu) − ∂EU(.) = 0; (2.14)

∂L ∂C ∂l

∂EU(.) = ∂EU(.)* w − ∂EU(.) ≤ 0; Lm ≥ 0;

∂Lm ∂C ∂l

∂EU(.) * Lm = 0. (2.15)

∂Lm

Equation (15) determines the optimal time allocation of the farm household to off-

farm work. Two optimal conditions possibly occur: the inequality constraint holds if the

farmers do not work off the farm, while the equality constraint holds if the farmers participate

in the off-farm labor market. Solving Eqs. (14) and (15) simultaneously yields two possible

optimal labor allocations: (L1*, L m*) for farmers who work off the farm, and (L0*, 0) for

farmers who do not engage in off-farm work. If the optimal use of labor is further plugged

into the production function, this yields two possible optimal agricultural supply functions

F(L1*) = f (L1*) + g(L1*)ε − h(L1*)u if L m*> 0, (2.16)

F(L0*) = f (L0*) + g(L0*)ε − h(L0*)u if L m* = 0 (2.17)

 35

Equations (16) and (17) guide our empirical analysis. To link the theoretical

framework to the empirical analysis, several econometric issues have to be addressed. First, it

is likely that the off-farm work decision may be correlated with the farm productivity due to

some unobservable characteristics (such as pride, the preference of the farm operator to work,

etc.), which may cause the potential endogeneity or self-selection bias problem (Chang and

Wen, 2011)

A conventional way of correcting the endogeneity problem is to apply Heckman’s

method by adding the Inverse Mills Ratio (IMR) to the production function (Heckman,

1979). While this approach is theoretically sound, deriving the symbolic forms of the

correction terms is not obvious because the production functions contain a composited error

in this case. To derive the IMR (i.e., the truncated mean function) given the complicated error

structure in this case is not as straightforward as that in the original Heckman model.

Moreover, even if the symbolic forms of the IMR can be derived with any luck, one still

needs to empirically find some valid instruments, which are assumed to directly affect the

off-farm work decision of the operator and which affect the farm production in an indirect

way, to econometrically identify the model (i.e., the exclusion condition, as in Wooldridge,

2002). Finding a good instrument is also empirically challenging and using invalid

instruments can lead to worse estimations when compared to the case where there is no

correction for self-selection bias (Wooldridge, 2002). Therefore, due to the lack of a tractable

empirical method to handle the endogeneity between a binary choice model and the

stochastic production frontier model with production risk, in this empirical analysis, the

farmers will simply be separated into two groups (those with and those without off farm

work), and estimate the production function for each group of farmers.
 36

2.11.3 Production Economics Theory

This is part of microeconomic theory that deals with production of goods using a set

of inputs (Doll and Orazem, 1984). A production function is a model used to formalize this

relationship. Below is a specification of a production function

Q=f {L, S, F…} (2.18)

Where Q represent a firms output, L may represent the amount of labour, S represents

quantity of seeds used in production of Q while F represent the amount of fertilizers applied.

The objective of the producer is to maximize profit either by increasing the quantity of Q

produced or by reducing the cost of producing Q. The production function shows the

maximum amount of the good that can be produced using alternative combinations of labour

(L), seed (S) and fertilizer (F). Q is also referred to as the total physical product (TPP). This

production relationship can be expressed in several forms such as: linear functional forms,

polynomial functional forms and Cobb-Douglas functional form. The later is modified into

the transcendental and translog functional forms.

Production functions are widely used to analyze efficiency in terms of output for a

given level of inputs. In most microeconomic analyses, production functions are estimated

under the assumption that producers are rational profit maximizers that operate on their

production frontiers. However, Farrell (1957), Aigner et al (1977), Meeusen and Van den

Broeck (1977), and Battese and Coelli (1995) support the view that producers differ in the

measured output that they produce from a given bundle of measured input, or, alternatively,

in the input requirements to produce a given output.

The marginal physical product (MPP) of an input is the additional output that can be

produced by employing one more unit of that input while holding all other inputs constant

(Doll and Orazem, 1984).

Example, the MPP of labour, MPL = , (2.19)

 37

This is derived from the first derivative of the production function. However, if labour

is employed indefinitely while holding all the other inputs of production indefinitely, this

results into diminishing marginal productivity where the rapid increase in use of additional

input results to lower productivity. Therefore the second derivative is less than zero:

(2.20) (Doll and Orazem, 1984)

The average physical product (APP) is a measure of efficiency. The APP depends on

the level of other inputs employed.

(2.21) (Doll and Orazem, 1984)

The concept of returns to scale shows how output responds to increase in all inputs

together. Returns to scale can either be constant, decreasing or increasing.

The elasticity of supply of an input measures how an output responds to changes in

inputs. This is derived by dividing the MPP by the APP (i.e. MPP/ APP). In addition, the

total variable product (TVP) is derived by multiplying TPP by the output price (i.e.

TPP*output price). Given the output price (Py), its marginal value product (MVP) can be

computed by multiplying (MPP*Py). Given the above economic concepts from a production

function, a profit function can be generated as follows: Profit ( )=TVP-TVC, applying the

first order condition (FOC) we get a change in profit with respect to change in input, for

example, labor (L) is =MVP-MVC=0. Therefore, at profit maximization, MVP

(MPP*Py) =MVC=w (unit of input), (Ingosi, 2005). To determine if the inputs are used at

optimum level, the MVP is equated to the unit factor price. It is important to note that in the

traditional production function, social economic characteristics and management are not

considered as explanatory variables and are thus lumped together in the error term. The
 38

stochastic frontier production functions deals with the analysis of socio-economic

characteristics of the household that are assumed to be in the composed error

2.12 Analytical Framework

In this study the empirical strategy adopted by Chan and Wen (2011) was used. This

involved estimating two stochastic production frontier functions that account for the

production risk of two groups of farmers: those who report off-farm income, and those who

do not. With the consistent estimates of the production parameters, the technical efficiencies

for these two groups of farmers are then calculated and compared.

2.12. 1. Estimating the stochastic frontier model with risk

The estimation model proposed for this study is a production function commonly

associated with the stochastic frontier framework and is an extension of the standard frontier

model that allows heterogeneous risk terms (Battese et al., 1997; Wang, 2002). Following

Wang (2002), the econometric specification of the production function can be shown as

yi = xi β + vi − ui ; (2.22)

where yi and xi are the logarithm of the production yield and inputs, respectively, and β is a

vector of coefficients that characterize the production frontier. The notations vi and ui are the

random error (noise) and inefficiency terms, respectively which according to Jaenicke,

Frechette, and Larson (2003) can take a number of forms, depending on specific assumptions.

Following the conventional specification in the stochastic production frontier model, the

random error vi follows a normal distribution with zero mean and variance but the

normality assumption is not required in a Just-Pope framework.

A stochastic frontier model that not only allows for heterogeneity in the mean of the

inefficiency term to investigate inefficiency effects but also allows for heterogeneity in the
 39

noise component to investigate risk effects is given by equation (22), with the following

assumptions:

where

; ui ~ N + ( ūi, ) (2.23)

= exp (zi ); ūi = wiα

the inefficiency term ui follows a truncated-normal distribution with mean ūi and variance

. To capture the heterogeneity of the efficiency and risk terms, the mean efficiency and

risk functions are determined by exogenous factors (Chang and Wen, 2011). The vector wi

denotes exogenous variables that have influences on the mean value of production

inefficiency.

The risk function is assumed to have an exponential functional form with the vector of

the exogenous determinants zi (Battese et al., 1997; Just and Pope, 1979). The notation α is a

vector of parameters associated with the mean of the production inefficiency while the

notation is the vector of parameters associated with the production risk. The consistent

estimators of Eq. (23) can be obtained by using the maximum likelihood estimation method

on the following log-likelihood function

ln L = constant

(24)

where = ; = yi xi β; and = exp (zi )]0.5;

 40

The general specification of Eq. (24) is testable for several special cases. Testing the

null hypotheses H0:α = 0, and H0: = 0 provides the statistical justification if the technical

inefficiency and risk functions are heteroscedastic. If the parameters α = 0, then Eq. (22) is

simply a Just-Pope type of production risk function (Just and Pope, 1979). By contrast, Eq.

(22) becomes the conventional stochastic production frontier model without the consideration

of risk if the parameter = 0 (Aigner et al., 1977). The likelihood ratio test can be conducted

for each null hypothesis. The technical efficiency of each farmer can be then calculated as

TEi = E[exp(−ui )|ε i ] (Battese and Coelli, 1988), and the risk term of each farmer is the

exponential function of the vector specified in the risk functions, that is, exp(ziγ ).
 41

CHAPTER THREE

METHODOLOGY

3.1 The Study Area

The study area was Enugu State, Nigeria. The state lies approximately between

latitudes 5056 N and 7005 N of the equator and longitudes 6053 E and 7055 E of the

Greenwich Meridian (Anyadike, 2002). The state is bounded to the north by Benue and Kogi

States, to the south by Abia State, to the east by Ebonyi state and to the west by Anambra

State. The state occupies an area of about 8,022.95km2 with seventeen (17) local government

areas (LGAs) (Enugu State Agricultural Development Programme (ENADEP, 2008) and a

population of about 3,257,298 persons (NPC, 2006).

The State has tropical climate marked by two distinct seasons, which includes rainy

and dry seasons. The dry season occurs between November and April, while the rainy season

begins in April and lasts until October. The mean annual rainfall decreases from about 200cm

at its southern extremities to about 150cm in the north (Ukwu et al., 1998). The predominant

agricultural practice in the state is crop farming. However, animals are reared in all parts of

the state though in small numbers. Food crops grown in the state include yam, cassava, rice,

maize, melon, vegetables, sweet potato, cocoyam, groundnut and cowpea. Among the tree

crops grown are oil palm, Citrus spp, mango, pears and cashew (NAERLS and PCU, 2006).

According to ENADEP (2008) Enugu State is delineated into three (3) major

agricultural zones, namely:

North Zone comprising Nsukka, Igbo-Eze North, Igbo-Eze South, Igbo-Etiti, Udenu

and Uzo-Uwani LGAs

East Zone comprising Isi- Uzo, Enugu East, Enugu North, Enugu South, Nkanu East

and Nkanu West LGAs

 42

West Zone comprising Awgu, Aninri, Oji River, Ezeagu, and Udi LGAs

3.2 Sampling Procedure

A combination of purposive, multi-stage and stratified random sampling technique

were used in selecting ninety respondents. Firstly, two agricultural zones (North zone and

West zone) were purposively selected due to the predominance of rice production in the area.

The second stage involved a purposive selection of two local government areas Aninri (in

west zone) and Uzo-Uwani (in north zone) also due to the predominance of rice production in

these areas. Thirdly, one community was randomly selected from each of the selected local

government areas giving a total of two communities, Oduma (in Aninri) and Adani (in Uzo-

Uwani). The fourth stage involved stratification of the farmers (in the two communities) into

rice farmers with and without off farm work. The last stage involved the random selection of

rice farmers from each stratum in a ratio proportional to the size of the population of rice

farmers in each stratum, 59 for rice farmers without off-farm work and 31 for rice farmers

with off-farm work. The sample frame was the list of rice farmers obtained from ENADEP.

3.3 Data Collection

Data for the study was collected from primary source. Primary data was collected by

the use of structured questionnaire which was administered by trained enumerators. Both

qualitative and quantitative information on the relevant variables such as the production and

cost data in rice production, socio-economic characteristics of the farmers, engagement in

off-farm work and income sources were collected. Information collected were for the period

of 12months (August, 2011- July, 2012).

 43

3.4 Data Analysis

Data collected were analysed using descriptive and relevant statistical and

econometric tools in order to achieve the specific objectives.

Objective (i) and (ii) was realized using descriptive statistics such as mean, percentages and

frequency distribution. Objective (iii) and (v) was realized using stochastic production

frontier model. Objective (iv) was realized using student t-test. Likert scale rating technique

was used to realize objective (vi)

Generalized likelihood ratio test was used to test hypothesis (i) and (ii). Student t-test was

used to test hypothesis (iii). Hypothesis (iv) was tested using t-test embedded in the

inefficiency model component of the stochastic production frontier function.

3.4.1 Model Specification

3.4.1.1 Stochastic frontier production function

Following the specification proposed by (Battese et al., 1997 and Wang, 2002) this

study employed a stochastic frontier production function to measure the technical efficiency

of rice farmers with and without off- farm work in Enugu state. A stochastic frontier model

that not only allows for heterogeneity in the mean of the inefficiency term to investigate

inefficiency effects but also allows for heterogeneity in the noise component to investigate

risk effects is given by the equation:

yi = xi β + vi − ui ; (3.1)

where,

yi is the logarithm of the production yield

Xiß is a suitable production function such as the Cobb-Douglas or translog;

xi is the logarithm of the production inputs

 44

β is a (k x 1) vector of unknown parameter (coefficients) that characterize the production

frontier

ui is the inefficiency term which follows a truncated-normal distribution with mean ūi and

variance as specified below:

ui ~ N + ( ūi, ) (3.2)

ūi = w i α

where:

wi = vector of exogenous(explanatory) variables like farm household characteristics and socio

economic characteristics that have influences on the mean value of production inefficiency.

They include:

w1= average number of farmers in each cooperative team (number)

w2 = family size (number)

w3 = age (years)

w4 = education (number of years)

w5 = extension services (number of visits).

α = vector of unknown parameters (coefficients) associated with the mean of the production

inefficiency

vi is the noise component that investigate risk effects which follows a normal distribution

with zero mean and variance

; (3.3)

= exp (zi );

Where:
 45

zi = vector of the exogenous determinants which include z1 = family labour used in rice

production per ha (person-days),

z2= hired labour used in rice production per ha (person-days),

z3= depreciated value of farm implements used (naira)

= vector of unknown parameters (coefficients) associated with the production risk

The risk function is assumed to have an exponential functional form.

The technical efficiency of each farmer can be then calculated as:

TEi = E[exp(−ui )|ε i ]

(3.4)

The method of maximum likelihood will be applied for simultaneous estimation of the

parameters of the stochastic frontier and the model for the technical inefficiency and

production risk effects (Battese and Coelli, 1993). This is available in the statistical software

STATA 11. The likelihood function is expressed as:

ln L = constant (3.5)

where = ; = yi xi β; and = exp (zi )]0.5

A number of previous studies specified a Cobb-Douglas production function to

represent the frontier function. This study will fit the data into Cobb-Douglas. The model is

specified as follows:

Cobb-Douglas production function form

ln = (3.6)

Where:

ln= natural logarithms

Yi = value of output of rice per ha (naira)

 46

X1= total labour used in rice production per ha (person-days)

X2= fertilizer cost per ha (naira)

X3= total area of land used in rice production (hectares)

X4= total value of agrochemicals (i.e pesticides and herbicides) used in rice production in

naira

X5=cost of seed per ha (naira)

X6= depreciated value of farm implements used (naira)

β = parameters to be estimated

To choose the functional form that best describes the inefficiency and production risk effect,

the following null hypothesis will be tested:

1. H0: α = 0, this hypothesis specifies that the technical inefficiency effects are not present in

the model. If this hypothesis is accepted, then the rice farmers are fully technically

efficient and we can say that the stochastic production frontier model specified in Eq (1)

is simply a Just-Pope type of production risk function.

2. H0: = 0, this hypothesis specifies that the production risk effects are not present in the

model. If this hypothesis is accepted, then rice farmers are faced with production risk and

so equation (1) becomes the conventional stochastic production function without the

consideration of risk.

The generalised likelihood ratio test was conducted for each null hypothesis and is specified

as:

λ = -2ln[L(H0)/L(H1)] = -2ln[L(H0)-L(H1)] (3.7)

(Kodde and Palm, 1986)

where L(H0) and L(H1) are the maximum values of the log likelihood functions for the

frontier model under the null and alternative hypotheses, respectively. The null hypothesis is
 47

accepted when λ has approximately a chi-square (χ2) or a mixed χ2 distribution with degrees

of freedom equal to the difference between the parameters involved in the null and alternative

hypotheses. The critical values for the generalized likelihood ratio test would be obtained at

5% level of significance.

To compare the estimated technical efficiencies between these two groups of farmers

(i.e., with and without off-farm work), a statistical test was conducted. The test is based on

the conventional method of moments to test if the means between these two groups of

farmers are statistically equal by applying the t-test.

t= X1 - X2

√ S12 + S22

N1 N2 (3.8)

Where t = calculated t – statistic,

X1 = mean technical efficiency of farmers with off-farm work

X2 = mean technical efficiency of farmers without off-farm work

S12 and S22 are the variance of the two groups of rice farmers

N1 and N2 are the number of rice farmers with and without off farm work respectively

This was tested at 5% level of significance.

Likert Scale Rating Technique.

Likert scale of a 4-point rating was used in this study to measure the level of

constraint faced by rice farmers in the area. The grading was done in this order: strongly

agree=4; agree= 3; disagree= 2; strongly disagree=1. The values of the likert scale was added

to get 10, which was later divided by 4 to get a mean score of 2.5 (i.e 4+3+2+1 = 10/4 =2.5)

cut off point. The respondents mean score will be obtained from each response item such that
 48

anyone higher or equal to 2.5 will be categorized as agree, while anyone less than 2.5 will be

categorized as disagree (Ndagi, 1984).

CHAPTER FOUR

RESULTS AND DISCUSSION

4.1 Socio-economic Characteristics of Respondents

The socio-economic characteristics considered in this study include: sex, age, level of

education, household size, marital status, membership of farmers association, access to

extension services, and number of years of experience in rice farming and participation in

off-farm work. Table 4.1 presents a summary of the socioeconomic characteristics of the rice

farmers in the study area.

4.1.1 Sex of Respondents

Table 4.1.1 shows that majority of the rice farmers sampled (51%) were female;

While male dominate (55%) in the group of rice farmers with off-farm work, female

dominate (46%) in the other group of rice farmers without off-farm work. This could be

attributed to the fact that males seem to take up other off-farm work to augment for the

income from the rice production since in most African families men are the bread winners

while the women folk support.

4.1.2 Age of Respondents

One of the importance of age distribution of rice farmers is that most of the farm

operations in rice cultivation (especially in developing countries like Nigeria that rely to a

large extent on manual labour) require physical energy. Thus, only those farmers within the

productive age group are likely to possess the required energy to carry out the various farm

operations. Table 4.1.1 shows that majority of the sampled rice farmers (86%), about 90%

and 83% of rice farmers with and without off –farm work respectively fall within the age
 49

bracket of 20-60years. The mean age of the sampled rice farmers was 49%; 45% and 51% for

rice farmers with and without off-farm work respectively. In all the groups the mean age was

tending towards the declining productivity class of greater than 50years. This result agrees

with the study of Ogundele & Okoruwa (2006) who found the mean age of traditional

technology and improved technology rice farmers to be 42 and 45 respectively. The mean

ages of the rice farmers suggest that unless there is injection of young people in rice farming,

rice production may suffer some setback in the country in the near future as the existing

farmers would have reached the declining productivity level.

Table 4.1.1: Socio-economic characteristics of respondents

Full sample Rice farmers with Rice farmers without

(pooled) off-farm work off-farm work
Variable Frequency Percentage Frequency Percentage Frequency Percentage
Sex
Male 44 48.9 17 54.8 27 45.8
Female 46 51.1 14 45.2 32 54.2
Total 90 100 31 100 59 100

Age(years) Mean =49.08 Mean = 45.35 Mean =51.03

20-40 23 25.6 10 32.3 13 22
41-60 54 60 18 58 36 61.1
61-80 13 14.4 3 9.7 10 16.9
Total 90 100 31 100 59 100

Level of education Mean =6.43 Mean = 9.68 Mean =4.73

No formal
education 36 40 4 12.9 32 54.2
Primary 21 23.3 10 32.3 11 18.6
Secondary 20 22.2 10 32.3 10 16.9
Tertiary 13 14.4 7 22.6 6 10.2
Total 90 100 31 100 59 100

Source: Field survey, 2012

4.1.3 Level of Education of Respondents

Generally, education broadens our horizon. It enhances technology adoption, skill

acquisition and increase efficiency of farmers. Table 4.1.1 shows that 40% of the rice farmers

sampled had no formal education, the remaining 60% had one form of formal education and
 50

the average years of schooling was 6years. In the group of rice farmers with off-farm work

only few of the farmers had no formal education (13%), the remaining (87%) had one form of

formal education while in the other group of rice farmers without off-farm work more than

half of them (54%) had no formal education, the remaining (46%) had one form of formal

education. The mean of the number of years spent in school by rice farmers with and without

off-farm work was 10years and 5years respectively. This result is not surprising because

education no doubt helps unlock the inherent enterprising abilities of farmers. This finding

agrees with the study of Nwaru and Onuoha (2010) who found that the average years of

schooling was 10years.

4.1.4 Household Size of the Respondents

A household unit comprises the immediate family (father, mother and children) and

other dependents living with them. A large household size is important as it influences the

supply and availability of unpaid labour services especially in Nigeria where there is

increasing cost of hired labour due to migration of the young generation to urban areas.

Although, where most members of the household are young children of school age and old

family members, a large household size may not contribute much to rice production.

Table 4.1.2 shows that majority (56%) of the rice farmers sampled; about 63% and

52% of rice farmers with and without off-farm work respectively had household size of 6 to

10. This is followed by household size of 1to5where 36%, 26%, and 41% of the sampled rice

farmers, rice farmers with and without off-farm work respectively fell into. The average

household size was seven for the full sample and rice farmers with off-farm work while rice

farmers without off-farm work had average household size of six.

4.1.5 Marital Status of Respondents

Table 4.1.2 shows that majority (76%) of the sampled rice farmers, 81% and 74% of

rice farmers with and without off-farm work respectively were married. This was followed by
 51

the class widowed where 15%, 13%, 16% of the pooled sample, rice farmers with and

without off-farm work respectively fell into. The greatest percentages which belong to the

class of married may be in an effort to contribute to the upkeep of the family.

Table 4.1.2: Socio-economic characteristics of respondents (cont’d)

Full sample Rice farmers with Rice farmers without

(pooled) off-farm work off-farm work
Variable Frequency Percentage Frequency Percentage Frequency Percentage

Household Size Mean =6.58 Mean = 7.07 Mean =6.33

1-5 29 35.8 7 25.9 22 40.7
6-10 45 55.6 17 63 28 51.9
11-15 7 8.6 3 11.1 4 7.4
Total 81 100 27 100 54 100

Marital Status
Single 7 7.9 2 6.5 5 8.6
Married 68 76.4 25 80.6 43 74.1
Divorced 1 1.1 - - 1 1.7
Widowed 13 14.6 4 12.9 9 15.5
Total 89 100 31 100 58 100

Social Participation
0 67 74.4 21 67.7 46 78
1 16 17.8 7 22.6 9 15.3
2 6 6.7 2 6.5 4 6.8
3 1 1.1 1 3.2 - -
Total 90 100 31 100 59 100

Source: Field survey, 2012

4.1.6 Social Participation of the Respondents

Social participation in the form of farmers belonging to farmers association plays a

significant role in enhancing farmers’ access to credit and other resources required for

agricultural production, information transfer and access to developmental initiatives through

their ability to organise themselves. Table 4.1.2 shows that majority (74%) of the rice farmers

sampled, 68% and 78% of rice farmers with and without off-farm work respectively do not

belong to any association. Eighteen percent (18%), twenty three percent (23%) and fifteen

percent (15%) of the pooled sample, rice farmers with and without off-farm work
 52

respectively belonged to only one association; Seven (7%) of the sampled farmers belong to

two associations; while only one (1%) and three percent (3%) of the sampled rice farmers and

rice farmers with off-farm work respectively belong to three associations.

4.1.7 Access to Extension Services by Respondents

The Agricultural Development Project was introduced in Nigeria to help facilitate

extension activities. The extension agent help to disseminate latest information, innovations

and technologies to the farmers and this enhances their productivity. Table 4.1.3 shows that

majority (76%) of the sampled rice farmers, 81% and 73% of the rice farmers with and

without off-farm work respectively had no access to extension services. The remaining 24%,

19%, 27% of the pooled sample, rice farmers with and without off-farm work respectively

had access to extension services. The average number of visits during the cropping season

was twice yearly. This poor access to extension services could have serious implication on

the productivity of the farmers.

4.1.8 Participation in Off-farm work

Table 4.1.3 shows that majority of the rice farmers sampled (65.6%) do not engage in

off-farm work while the remaining percent (34.4) do.

4.1.9 Rice farming experience

Table 4.1.3 shows that 71.1% of the sampled rice farmers had rice farming experience

of 11-30years. In the group of rice farmers with off-farm work 80.6% had rice farming

experience of 11-30years while the remaining 19.4% had rice farming experience of 1-

10years. In the other group of rice farmers without off-farm work 66.1%, 20.3% and 13.6%

had rice farming experience of 11-30, 1-10 and 31-50 years respectively. The mean rice

farming experience were 21, 17 and 23 years for the sampled rice farmers, rice farmers with

and without off-farm work respectively.

 53

4.1.10 Primary occupation

Table 4.1.3 shows that the primary occupation of the majority (82.2%) of the sampled

farmers was farming, 10% were traders and 7% were civil servants. In the group of rice

farmers with off-farm work 48.4% were farmers, 29% were traders and 22.6% were civil

servants.

4.1.11 Secondary occupation

Table 4.1.3 shows that the secondary occupation of the majority (51.6%) of the

sampled farmers was farming, 35.5% were traders and 13% were civil servants. In the group

of rice farmers with off-farm work 51.6% were farmers, 29% were traders and 19.4% were

civil servants.

4.1.12 Farm Size

Table 4.1.3 shows that majority of the sampled rice farmers had farm size of between

0.5-5 hectares, 6.7% and 5.6% had farm size of less than 0.5hectares and 6-10hectares

respectively. In the group of rice farmers with off-farm work majority (96.3%) had farm size

of 0.5-5hectares and the remaining 3.2% had farm size of between 6-10hectares. In the other

group 83%, 10.2% and 6.8% had farm size of between 0.5-5, less than 0.5 and 6-10 hectares

respectively. The average farm size was 1.8, 2 and 1.7 hectares for the sampled rice farmers,

rice farmers with and without off-farm work respectively

 54

Table 4.1.3: Socio-economic characteristics of respondents (cont’d)

Full sample Rice farmers with Rice farmers without

(pooled) off-farm work off-farm work
Variable

Access to extension service

Yes 22 24.4 6 19.4 16 27.1
No 68 75.6 25 80.6 43 72.9
Total 90 100 31 100 59 100
Participation in Off-farm work
Yes 31 34.4 31 100 - -
No 59 65.6 - - 59 100
Total 90 100 31 100 59 100
Rice farming experience Mean =20.91 Mean = 17.19 Mean =22.86
1-10 18 20 6 19.4 12 20.3
11-20 30 33.3 18 58 12 20.4
21-30 34 37.8 7 22.6 27 45.7
31-40 6 6.7 6 10.2
41-50 2 2.2 2 3.4
Total 90 100 31 100 59 100
Primary occupation
Farming 74 82.2 15 48.4 59 100
Trading 9 10 9 29 - -
Civil Service 7 7.8 7 22.6 - -
Total 90 100 31 100 59 100
Secondary occupation
Farming 16 51.6 16 51.6 - -
Trading 11 35.5 9 29 - -
Civil Service 4 13 6 19.4 - -
Total 90 100 31 100 - -
Farm Size (hectares) Mean =1.8 Mean = 1.95 Mean =1.72
<0.5 6 6.7 - - 6 10.2
0.5-5 79 87.7 30 96.8 49 83
6-10 5 5.6 1 3.2 4 6.8
Total 90 100 31 100 59 100

Source: Field survey, 2012

4.2 Rice Production in the Study Area.

Rice farmers tend to be small-scale, with average farm size of 1.8ha. The land is

cleared between December and March. With the onset of the rain in early April, the land is

prepared either manually or with tractor in areas that are not swampy. Two rice production
 55

systems, rainfed upland and rainfed lowland rice cultivation were predominant in the study

area. This agrees with the study of Akpokodje, Lancon, and Erenstein (2001) which found

these two systems as the common practice in southern Nigeria where rainfall is better than

other areas. Under the rainfed upland rice cultivation which was the common practice in

Adani, rice is broadcasted in non-flooded, well drained soil on level to steeply sloping fields.

Some farmers prepare rice nursery and transplant from the nursery. Under rainfed lowland

rice production which was a common practice in Oduma, rice is broadcasted or transplanted

on level to slightly sloping fields with variable depth and duration of flooding depending on

rainfall. After 3-4weeks of planting (depending on the variety planted) weeding is done.

Some of the farmers use manual weeding while some use herbicides after which they hand

pick the remaining weed that was not removed by the herbicides. After two weeks of weeding

fertilizer is applied. Most of the farmers usually employ the services of children to help scare

away birds which is a serious problem in the study area. The maturity period of rice ranges

from 2-5months depending on the variety planted. When the rice is matured harvesting is

done. Most of the farmers process the harvested rice with the help of hired labour after which

they sell. In Adani, there existed an irrigation facility which is non-functional now and the

farmers have been appealing to the government to fix it.

4.3 Comparing the distributions of the technical efficiency scores between the two
groups of rice farmers

Table 4.2 reports the sample statistics of technical efficiency in terms of percentiles

for the two groups of rice farmers. Technical efficiency scores for the farmers ranged from

0.579 to 1.000 and 0.606 to 1.000 for the rice farmers without and with off-farm work

respectively. A negative impact of off-farm work on farm efficiency was found in previous

studies, such as Kumbhakar et.al (1986), Fernandez-Cornejo (1992), Goodwin and Mishra

(2004) and Chang (2011). Our empirical findings support this conclusion since the average
 56

efficiencies are 0.964 and 0.871 for rice farmers without and with off-farm work, respectively.

The average efficiency scores of 0.964 and 0.871 for rice farmers without and with off-farm

work imply a technical inefficiency of 0.037 (1-0.964/0.964) and 0.148 (1-0.871/0.871)

respectively. The economic interpretation of these figures is that an average farmer in the study

area requires approximately 4% (for rice farmers without off-farm work) and 15% (for rice

farmers with off-farm work) more resources to produce same output (or meet the same

objectives) as an efficient rice farmer on the frontier.

A student t-test was conducted to test the equality of the sample mean between the two

groups of rice farmers. The tcal value (3.423) was greater the ttab value (1.289) at 5% level of

significance and hence the null hypothesis which states that there is no significant difference in

the mean of the two groups was rejected.

Table 4.2: Distributions of technical efficiency

Without off-farm work off-farm work
Mean 0.964 0.871
Minimum 0.579 0.606
Maximum 1.000 1.000
Std. Deviation 0.092 0.124
Percentile (%)
1 0.579 0.606
5 0.668 0.621
10 0.869 0.727
25 0.994 0.759
50 0.999 0.889
75 0.999 0.999
95 1.000 1.000
Source: Field survey, 2012
 57

4.4 Estimation of the Stochastic Production Frontier Model

The parameters of the stochastic production frontier model with a flexible risk

specification were estimated simultaneously using the linear estimation procedure of the

maximum likelihood estimation available in the statistical software STATA 11 and the result

is presented in Table 4.3. The upper section of the table represents coefficients of the

production function, while the middle and lower sections represents coefficients of the

production risk and production inefficiency function respectively.

4.4.1 Estimation results for Production Function

The result of the maximum likelihood estimates (MLE) of the Cobb-Douglas model is

presented in Table 4.3. Estimated output elasticities for all the inputs all differed from zero at

the 1% significance level for the two groups of rice farmers except hired labour which has

significance level of 10% for rice farmers without off-farm work. For the group of rice

farmers without off-farm work the elasticity for depreciated value of equipment is the largest

(0.172). This means a 10% increase in the depreciated value of equipment used will give rise

to a 1.72% increase in output. This is followed by the use of agrochemical (-0.113). The

relationship seem to be negative, it could be that the farmers are not applying it in the right

quantity required. This is followed by farm size (-0.107). This means that a 10% increase in

farm size will decrease output by 1.07%.The negative influence of farm size could be as a

result of poor or lack of education among the rice farmers, a condition necessary to bring out

the efficiency of land use and other resources normally employed in rice farming. The next

on the row is family labour (-0.027) which has a negative relationship with output. A possible

explanation of this may be that the quality of family labour used is not good enough for

example using children to do the work that adults should effectively handle. This is followed

by seed (0.017) and fertilizer (0.011) which had a positive influence on the output as
 58

expected. Hired labour has the least elasticity of 0.002. The higher elasticity of family labour

than that of hired labour for rice farmers without off-farm income is consistent with the

findings of Chang (2011) and Audibert (1997).

For the other group of rice farmers with off-farm work depreciated value of

equipment has the largest elasticity (0.265) just like their counterpart. The negative influence

of this variable could be that this group because of their engagement in off-farm work pay

little attention to farm management and lack good knowledge regarding the use of inputs.

This is followed by seed with elasticity of 0.162. The reason for seed coming second instead

of fifth as in the case their counterpart could be that this group use the additional income

from off-farm work to purchase very high quality seeds. The third is hired labour with

elasticity of 0.058. This is not surprising since this group engage in off-farm work they will

need to engage the services of very competent hired labour to take care of most of their

production activities. The fourth on the row is agrochemical having elasticity of -0.043. The

negative sign as have earlier been stated could be that the agrochemical is not being applied

correctly. The next on the row is farm size (0.033) and family labour (0.018). The variable

with the least elasticity for this group of rice farmers with off-farm work is fertilizer (0.014).

4.4.2 Estimation results for production risk function

This result is presented in the middle of Table 4.3. For farmers without off-farm work,

family labour has a positive and significant effect on production risk, meaning that family

labour is a risk increasing factor. Depreciated value of equipment used has a negative and

significant effect on production risk which indicates that investment on equipment will

decrease the production risk in rice production. This agrees with the findings of Just and Pope

(1979); Gardebroek, Chavez and Lansink (2010) and Chang and Wen (2011).
 59

For the group of rice farmers with off-farm work none of the factors were significant

even though they all had negative signs.

Table 4.3: Stochastic Production Frontier Estimates

Variables Without off-farm work off-farm work

Coefficient Std. Err. Coefficient Std. Err.

Lnseed 0.017*** 0.013 0.162*** 0.000

Lnfertilizer 0.011*** 0.000 0.014*** 0.000
Lnfarmsize -0.107*** 0.012 0.033*** 0.000
Lnfamilylabour -0.027*** 0.000 0.018*** 0.000
Lnhiredlabour 0.002* 0.000 0.058*** 0.000
Lnagrochemical -0.113*** 0.006 -0.043*** 0.000
Lnequipment 0.172*** 0.002 -0.265*** 0.000
Constant 2.095*** 0.111 9.964*** 0.000
Risk function
Lnfamilylabour 1.287*** 0.191 -1.198 0.999
Lnhiredlabour -0.052 0.093 -0.087 1.000
Lnequipment -12.255*** 1.442 -0.005 1.000
Constant 107.161*** 13.032 -29.649 0.995
Inefficiency function
AvgNoAsso 0.646*** 0.111 0.007 0.007
Age 0.328*** 0.641 0.159*** 0.602
Household size 0.092 0.152 -0.970*** 0.340
Educational level 3.838*** 0.648 0.391 0.374
Extension access 3.144*** 0.988 4.727*** 1.438
Constant -41.256*** 5.830 -14.458*** 3.041

Log-likelihood 82.634 34.722

*, ***indicate significance level of 10% and 1% respectively.
Note: A negative sign of the parameters in the inefficiency function means that the associated
variable has a positive effect on technical efficiency, and vice versa.
Source: Field survey, 2012

4.4.3 Estimation results for technical inefficiency function

The results of technical inefficiency effects are also presented in the lower part of

Table 4.3. For rice farmers without off-farm work average number of associations, age,

education and extension access significantly and positively influenced technical inefficiency

effects. This is surprising. The explanation may be that the extension agents and the

association they belong to are not bringing relevant and up-to-date information to the farmers

or the farmers are not making use of the information provided to them. This is similar with
 60

the result of Tijani (2006) who found extension service to have negative relationship with

efficiency.

For rice farmers with off-farm work age and extension access significantly and

positively influenced technical inefficiency effects. The explanation may be that the older

farmers lack the strength to carry out some of the activities and may tend to be less open to

innovative technologies that could boost their efficiency. This result agrees with the findings

of Khai and Yabe (2011) who found that age had negative relationship with technical

efficiency. The variable household size significantly and negatively influenced technical

inefficiency effects.

4.5 Constraints faced by rice farmers

Table 4.4 indicate the serious constraint faced by the rice farmers in the study area.

Inaccessible roads ranked first, followed by high cost of transportation. This problem makes

some of them sell their product at very cheap price within the community instead of

transporting it to markets where they will sell at high price. Inadequate credit ranked third as

some of them complained that the interest rate charged on credit is too high and sometimes

the credit is not even available. They complained that the government loan was not easy to

access. Birds’ invasion ranked fourth. Quelea birds’ invasion posed serious threats to their

production especially for delayed planting. Most of the farmers engaged the services of

children to scare away the birds. Inadequate extension support ranked fifth. The farmers

complained of infrequent visits by the extension agents. The last on the rank is

inaccessibility to cheap farm inputs. They complained that the cost of improved seeds,

fertilizer and pesticides was high and this militates against effective performance of farmers.

They complained that fertilizer was not made available on time when they are needed by the
 61

ADP. And middlemen most time hijacked the commodity and sell it to them in the open

market at exorbitant prices.

Table 4.4 Constraints faced by rice farmers.

S/N Constraints Mean Standard dev. Rank

1. Inaccessible road 3.68 0.84 1

2. High cost of transportation 3.54 0.90 2
3. Inadequate credit 3.49 0.10 3
4. Birds invasion 3.30 0.34 4
5. Inadequate extension support 3.17 0.21 5
6. Inaccessibility to cheap farm inputs 2.78 0.41 6
Source: Field survey, 2012

4.6 Test of hypotheses

Hypothesis 1: Rice farmers in the study area are not technically efficient

The result of the test of the hypothesis that farmers are fully technically efficient is

presented in Table 4.5. The null hypothesis (H0: α = 0), specifies that the technical

inefficiency effects are not present in the model. This null hypothesis is rejected. Thus,

indicating that the rice farmers are not fully technically efficient.

Hypothesis 2: Production risk effects are not present in the model

The null hypothesis (H0: = 0), specifies that the production risk effects are not present in the

model. This null hypothesis is rejected and the result of the test is presented in Table 4.5

below.

Table 4.5.Generalized likelihood ratio test

 62

Null hypothesis χ2-calculated χ2-tabulated (0.05) decision

H0 : α = 0 47.085 11.07 Reject H0

H0 : =0 116.965 7.81 Reject H0

Source: Field survey, 2012

Hypothesis 3: There is no significant difference in the technical efficiency of rice farmers

with and without off-farm work.

The result of the student t-test conducted to test the equality of the sample mean

between the two groups of rice farmers is presented in Table 4.6. The tcal value (3.423) was

greater the ttab value (1.289) at 5% level of significance and hence the null hypothesis was

rejected.

Table 4.6: T-test result

Null hypothesis t-calculated t-tabulated (0.05) decision

H0: TE1= TE2 3.423 1.289 Reject H0

Subscripts 1 and 2 refers to rice farmers with and without off-farm work respectively.

Hypothesis 4: Socio-economic characteristics do not affect the technical efficiency of the

two groups of rice farmers.

As indicated in the lower section of Table 4.2 socio-economic characteristics had

significant influence (p<0.01) on the technical efficiency of the two groups of farmers and so

the null hypothesis was rejected.

 63

CHAPTER FIVE

SUMMARY, CONCLUSION AND RECOMMENDATION

5.1 Summary

This study was designed to empirically investigate the effect of off-farm work on

technical efficiency and production risk among rice farmers in Enugu State, Nigeria.

Specifically the study sought to identify the socio-economic characteristics of rice farmers

with and without off-farm work; determine and compare the technical efficiency of rice

farmers with and without off-farm work; ascertain factors that affect technical efficiency and

production risk of the two groups of rice farmers; identify the constraints faced by the rice

farmers in the study area. Data for the study was collected from a sample of 90 rice farmers.

Descriptive statistics and stochastic production frontier model were used in analysing the

data.

Results showed that the two groups of rice farmers had similar socio-economic

characteristics. Technical efficiency scores for the farmers ranged from 0.579 to 1.000 and

0.0606 to 1.000 for the rice farmers without and with off-farm work, respectively. The

average efficiencies are 0.964 and 0.871 for rice farmers without and with off-farm work,

respectively. This suggests that off-farm work has a negative effect on farmers’ technical

efficiency. The result of the student t-test conducted at 5% significance level showed that

there was significant difference in the mean technical efficiency of the two groups of rice

farmers. For rice farmers without off-farm work average number of associations (0.646;

p<0.01), age (0.328; p<0.01), education (3.838; p<0.01) and extension access (3.144; p<0.01)

significantly and positively influenced technical inefficiency effects while for their

counterpart age (0.159; p<0.01) and extension access (4.727; p<0.01) significantly and
 64

positively influenced technical inefficiency effects and household size (-0.970) was

significant at 1% but negative. For farmers without off-farm work, family labour (1.287) has

a positive and significant effect (p<0.01) on production risk, meaning that family labour is a

risk increasing factor. Depreciated value of equipment (-12.255) used has a negative and

significant effect (p<0.01) on production risk which indicates that investment on equipment

will decrease the production risk in rice production. For rice farmers with off-farm work none

of the factors were significant even though they all had negative sign. The constraints faced

by the farmers were inaccessible road, high cost of transportation, inadequate credit, birds’

invasion, inadequate extension support, inaccessibility to cheap farm inputs ranked in

ascending order of importance. A number of hypotheses were tested. The first hypothesis

which states that rice farmers in the study area are technically efficient was rejected at 5%

probability level. The second hypothesis which states that there is no significant difference in

the technical efficiency of rice farmers with and without off-farm work was also rejected at

5% probability level.

5.2 Conclusion

This study was carried out to ascertain the effect of off-farm work on technical efficiency

and production risk among rice farmers in Enugu State, Nigeria. Comparisons were made

between the technical efficiency of rice farmers who engaged in off-farm work and rice

farmers who do not engage in off-farm work. The following conclusions were drawn:

(i) there is substantial difference in the mean levels of technical efficiency between rice

farmers with and without off-farm work. Hence, off-farm work can be said to

negatively affect technical efficiency of the rice farmers in the study area.

(ii) investment on equipment will decrease the production risk in rice production.
 65

(iii) certain factors such as inaccessible road, high cost of transportation, inadequate

credit, birds’ invasion, inadequate extension support, inaccessibility to cheap farm

inputs among others have constrained successful rice production in the study area.

Hence, policies that will help address these issues should be pursued if Nigeria is to

achieve self-sufficiency in rice production.

5.3 Recommendations

From the findings of this study, it is evident that a lot of factors militate against the

productivity of rice farmers. To address these constraints and subsequently increase

farmers’ efficiency the following recommendations are proffered:

i. Government, community leaders and philanthropist should build rural infrastructure

such as roads, market structures etc. to enable the farmers easily move their product to

markets where they will be offered fair price for their products. This will also help to reduce

transportation cost to urban areas where they dispose their products.

ii. Genuine efforts should be made to avoid playing politics with agriculture in Nigeria.

Fertilizer, credit and other production inputs should be given direct to the real farmers rather

than giving it to political stakeholders and middlemen and this at affordable price and at the

right time. Also the interest rate should be reduced to encourage farmers to borrow. The

farmers can also organise themselves into cooperatives to access credit and some other

production resources they need.

iii. The activities of extension agents should be investigated further stemming from the

fact that extension contact unexpectedly had a negative relationship with the technical

efficiency. Government support in terms of revitalizing and priority funding of the extension

activities of the states’ Agricultural Development Programmes ADPs is required. This will
 66

help to mobilize the extension workers to reach the farmers with relevant information that

could boost their production.

iv. To curtail the problem of bird invasion modern bird scarers should be purchased or

distributed to the farmers’ free by government or other stakeholders. Alternatively the

farmers could adopt the culture of early planting.

 67

REFERENCES
Adedapo, K.D. (2008). Technical efficiency of maize farmers in Ogbomoso agricultural zone
of Oyo State. International Journal of Agricultural Economics & Rural Development,
1(2): 102-107

Afriat, S.N. (1972). Efficiency estimation of production function. International Economic

Review, 13:569-98.

African Rice Centre (WARDA), 2007. Africa Rice Trends: Overview of recent developments
in the sub-Saharan Africa rice sector. Africa Rice Centre Brief. Cotonou, Benin:
Africa Rice Centre (WARDA).

Ahearn. M.C., & Lee, J.E. (1991). “Multiple Job-Holding among Farm Operator
Households in the United States”. In Multiple-Job Holding Among Farm Families,
edited by M.C Hallberg., J.L Findeis., and D.A. Lass. Iowa State University Press,
Ames, IA.

Ahituv, A. & Kimhi, A. (2002). Off-farm work and capital accumulation decisions of
farmers over the lifecycle: the role of heterogeneity and state dependence. Journal of
Development Economics 68: 329-353.

Ahituv, A. and Kimhi, A. (2006). Simultaneous estimation of work choices and the level of
farm activity using panel data. European Review of Agricultural Economics 33: 49-
71.

Ahmad, M. & Bravo-Ureta, B. E. (1996). Technical efficiency measures for dairy farms
using panel data: A comparison of alternative model specifications. Journal of
Productivity Analysis, 7: 399-415.

Aigner D.J. & Chu S.F. (1968). On Estimation the Industry Production Function. American
Economic Review, 58: 826-839.

Aigner D.J., Lovell C.A.K., & Schmidt P. (1977). Formulation and estimation of stochastic
frontier production function models. Journal of Econometrics, 6:21-37.

Akpokodje, G., Lancon, F., & Erenstein, O. (2001). The Nigerian rice economy in a
competitive world: constraints, opportunities and strategic choices. Nigeria’s rice
economy: state of the art. Final Report presented to West Africa Rice Development
Association (WARDA) Bouake, Côte d’Ivoire.
 68

Al-hassan, S. (2008). Technical efficiency of rice farmers in Northern Ghana. African

Economic Research Consortium (AERC) Research Paper 178, Nairobi.

Ali, M. & Chaudry, M. A. (1990). Inter regional farm efficiency in Pakistan’s Punjab: A
frontier production study. Journal of Agricultural Economics, 14: 45 – 60.

Ali, M. & Flinn, J.C. (1989). Profit efficiency among Basmati rice producers in Pakistan
Punjab. American Journal of Agricultural Economics, 71:303-10.

Amao, J.O. (2008). Analysis of off-farm work among farming households in Oyo State.
International Journal of Agricultural Economics & Rural Development, 1(2).

Anyadike, R.N.C. (2002). Climate and Vegetation. In: G.E.K. Offormata (Eds). A
survey of the Igbo Nation. Onitsha: African First Publishers Limited.

Arene, C.J. (1996). Agricultural Economics: A functional Approach, Enugu: Prize Publishers
Ltd, Nigeria. pp. 21-34

Asche, F. & Tveterds, R.(1999). Modeling production risk with a two-Step procedure.
Journal of Agriculture and Resource Economics, 24: 424-39.

Audibert, M. (1997). Technical inefficiency effects among paddy farmers in the villages of
the ‘Office du Niger’, Mali, West Africa. J. Product. Anal. 8: 379–394.

Awudu, A. & Huffman, W. (2000). Structure adjustment and economic efficiency of rice
farmers in northern Ghana.” Economic Development and Cultural Change, 503-519.

Awudu, A. & Eberlin, R. (2001). Technical efficiency during economic reform in Nicaragua:
Evidence from farm household survey data. Economic Systems, 25: 113-125.

Bagi, F.S. (1984). Stochastic frontier production function and farm-level technical efficiency
of full-time and part-time farms in west Tennessee. North Central Journal of
Agricultural Economics 6: 48-55.

Baquet,A., Hambleton, R. & Jose, D. (1997). Introduction to Risk Management. USDA

Risk Management Agency.

Battese, G.E. (1992). Frontier production functions and technical efficiency: A survey of
empirical applications in agricultural economics. Agricultural Economics 7:185-208.
 69

Battese, G. E & Corra, G. S. (1977). Estimation of a production frontier model: With

application to the pastoral zone of eastern Australia. Australian Journal of
Agricultural Economics 21:169-79.

Battese, G.E. & Coelli, T.J. (1988). A stochastic frontier production function incorporating a
model for technical inefficiency effects: a comparative study of wheat farmers in
Pakistan. Journal of Econometrics, 38: 387-399.

Battese, G.E. & Coelli, T.J. (1993). A stochastic frontier production function incorporating a
model for technical inefficiency effects. Working Paper No. 68 in Econometrics and
Applied Statistics. Department of Econometrics, University of New England,
Armidale.

Battese, G.E. & Coelli, T.J. (1995). A Model for Technical Inefficiency Effects in a
Stochastic Frontier Production Function for Panel Data. Empirical Economics, 20:
325-32.

Battese, G.E., Malik, S. J. & Gill, M. A. (1996). An investigation of technical inefficiencies

of production of wheat farmers in four districts of Pakistan. Journal of Agricultural
Economics, 47 (1): 37-49.

Battese, G.E., Rambaldi, A.N. & Wan, G.H. (1997). A stochastic production function with
flexible risk properties. Journal of Productivity Analysis 8.

Battese, G. E. & Broca, S. S. (1997). Functional forms of stochastic frontier production

functions and models for technical inefficiency effects: A comparative study for
wheat farmers in Pakistan. Journal of Productivity Analysis 8: 395-414.

Battese, G., Rambaldi, A., & Wan, G. 1997. A stochastic frontier production function with
flexible risk properties. Journal of Productivity Analysis, 8: 269–280.

Bauer, P.W. (1990). Some recent developments in the econometric estimation of frontiers.
Journal of Econometrics 46, 39-56.
Bedassa, T. & Krishnamoorthy, S. (1997). Technical efficiency in paddy farms of Tamil
Nadu: An analysis based on farm size and ecological zone.” Agricultural Economics,
16:185-192.

Behr, A. & Tente, S. (2008). Stochastic Frontier Analysis by means of maximum likelihood
and the method of moments. Deutsche Bundesbank, Discussion Paper, Series 2:
Banking and Financial Studies, No. 19/2008.

Belen, I., Rapun, M. & Zabaleta, I. (2003). Assessing the technical efficiency of horticultural
production in Navarra, Spain. Agricultural Systems, 78:387-403.
 70

Benjamin, C. & Kimhi, A. (2006). Farm work, off-farm work, and the hired labour:
estimating a discrete choice model of French farm couples’ labour decisions.
European Review of Agricultural Economics 33:149-171.

Bera, A.K., &. Sharma,S. C. (1999). Estimating production uncertainty in stochastic Frontier
production function models. Journal of Productivity Analysis 12, pp. 187-210.

Bokusheva, R. & Hockmann, H. (2006). Production Risk and Technical Inefficiency in

Russian Agriculture. European Review of Agricultural Economics 33(1):93–118.

Bollman, R. D. (1979). "Off-Farm Work by Farmers: An Application of the Kinked Demand

Curve for Labour. Canadian Journal of Agricultural Economics, 27: 37-60.

Bravo-Ureta, B.E. & Pinherio, A.E. (1993). Efficiency analysis of developing country
agriculture: A review of the frontier function literature. Agricultural and Resource
Economics Review, 22: 88-101

Bravo-Ureta, B.E. & Evenson, R. E. (1994). “Efficiency in Agricultural Production: A Case

of Peasant Farms in Eastern Paraguay.” Agricultural Economics 10:27-37.

Brewster, D.E. (1979).“The Family Farm: A Changing Concept,” Structure Issues of

American Agriculture. U.S. Department of Agriculture, Economics, Statistics, and
Cooperative Service, AER-438.

Central Bank of Nigeria (CBN) (998). Annual Report and Statement of Accounts, Central
Bank of Nigeria, Lagos, Pp. 78 – 79.

Central bank of Nigeria (CBN) (2000). Statistical Bulletin Vol. 2 No.2

Chang, H. &Wen, F. (2011). Off-farm work, technical efficiency, and rice production risk in
Taiwan. Agricultural Economics, 42: 269–278.

Charnes, A., Cooper, W. W. & Rhodes, E. (1978). Measuring the efficiency of decision
making units. European Journal of Operational Research. 2 (4) 429 - 444.

Chavas, J.P. (2004). Risk Analysis in Theory and Practice. Amsterdam: Elsevier Butterworth-
Heinimann Publishers.

Coelli, T. J. (1995). Recent developments in the frontier modelling and efficiency

measurement. Australian Journal of Agricultural Economics 39, 219-245.
 71

Coelli, T., Rao, D. S. & Battese, G. E. (1998). An Introduction to Efficiency and

Productivity Analysis. Kluwer Academic Publishers, Norwell, MA.

Cuesta, R. A. (2000). "A Production Model with Firm-Specific Temporal Variation in

Technical Inefficiency: With Application to Spanish Dairy Farms." J. Productivity
Analy. 13: 139-58.

Dayton, C.M. (1992). Logistic Regression Analysis. Retrieved from http://bus.utk.edu/sta. on

30/05/201
Debreu, G. (1951), ¨The coefficient of Resource Utilization¨, Econometrica 19, 273-292.

del Hoyo, J.J.G., Espino, D.C., & Toribio, R. J. (2004). Determination of technical efficiency
of fisheries by stochastic frontier models: A case on the Gulf of Cadiz (Spain).
Journal of Marine Science, 61: 416-421.

Diewert, W.E. & Parkan, C. (1983). Linear Programming tests of regularity conditions for
production function. In W. Eichorn, R. Henn, K. Neumann & R.W. Sherpherd (Eds).
Quantitative studies on production and prices. Vienna: Physica-Verlag.

Doll, J.P & Orazem, F. (1984). Production Economics: Theory with applications. John
Wiley& Sons Inc, Canada

Egbuna, E.N. (2003). A Review of the Activities of the National Strategic Grains Reserve;
Issue Arising C.B.N. Bullion, April-June, vol. 12pp. 42-45.

Ekong, E.E. (2003). Rural sociology: An Introduction and Analysis of Rural Nigeria. Dove
Publishers Uyo. Pp. 305.

El-Osta, H., Mishra, A., & Morehart, M. (2007). Determinants of economic well-being
among U.S. farm operator households. Agric. Econ. 36, 291–304.

El-Osta, H. & Morehart, M. (2008). Determinants of poverty among U.S. farm

households. J. Agric. Appl. Econ. 40, 1–20.

El-Osta, H., Mishra, A. & Morehart, M. (2008). Off-farm labour participation decisions of
married farm couples and the role of government payments. Rev. Agric. Econ. 36,
291-304

Enugu State Agricultural Development Programme (ENADEP) (2008). A publication of

Enugu State Government of Nigeria.

Fare, R., Grosskopf, S. & Lovell, C. A.K. (1985). The Measurement of Efficiency of
Production. Boston: Kluwer-Nijhoff Publishers.
 72

Farrell, M.J. (1957). The measurement of productive efficiency. Journal of the Royal
Statistical Society, Series A (General), 120: 253-81.

Fernandez-Cornejo, J. (1992). Technical efficiency and chemical input use in

agriculture. Agric. Resour. Econ. Rev. 21, 11–21.

Forsund, F.R., Lovell, C.A.K. & Schmidt, P. (1980). A survey of frontier production
functions and their relationship to efficiency measurement. Journal of Econometrics,
13: 5-25.

Federal Office of Statistics (FOS) (2000). Annual Book of Abstract, Federal Office of
Statistics, Lagos, Nigeria.

Fried, H. O., Knox Lovell, C.A., & Schmidt, S.S. (2008). The measurement of productive
efficiency and productivity growth. New York: Oxford University Press.

Fuller, A.M. (1991). Multiple Job-Holding among Farm Families in Canada”, Multiple job
holding among farm families, M.C. Hallberg, J.L. Findeis and D.A. Lass (Eds), Ames,
IA: Iowa State University Press.

Gardebroek, C., Chavez, M. & Lansink, A. (2010). Analysing production technology and risk
in organic and conventional Dutch arable farming using panel data. Journal of
Agricultural Economics 61: 60-75.

Gardner, B.L. (2002). American Agriculture in the Twentieth Century. Harvard University
Press, Cambridge, MA.

Gautam, H. & Jeffrey A. (2003). Access to credit, plot size and cost inefficiency among
smallholder tobacco cultivators in Malawi.” Agricultural Economics, 29: 99-109.

Goaied, M. & Ayed-Mouelhi, R. B. (2000). Efficiency measurement with unbalanced panel

data: Evidence from Tunisian textile, clothing, and leather industries. Journal of
Productivity Analysis 13: 249-62.

Godwin, D. & Marlowe, J. (1990). Farm wives labour force participation and earning. Rural
Sociology, 55: 25-43.

Goodwin, B.K. and Mishra, A.K. (2004). Farming efficiency and the determinants of
multiple job holding by farm operators. American Journal of Agricultural Economics
86: 722-729.
 73

Greene, W.H. (1982). Maximum likelihood estimation of stochastic frontier production

models. Journal of Econometrics, 18 285-289.

Green, W.H. (1993). The Econometric Approach to Efficiency Analysis, In H.O. Fried,
C.A.K. Lovell & S.S. Schmidt (Eds). The measurement of productive efficiency:
Techniques and applications (pp. 68-119). New York: Oxford University Press.

Greene, W.H. (1997). Frontier production functions. In: Pesaran, M.H., Schmidt, P. (Eds.),
Handbook of Applied Econometrics, Microeconomics, vol. II. Blackwell, Oxford
(UK), Cambridge, MA (USA), pp. 81– 166.

Haggblade, S., Hazell, P. & Brown J. (1989). “Farm-Nonfarm Linkages in Rural Sub-
Saharan Africa.” World Development 17: 1173-1201.

Hallberg, M., Findeis, J., & Lass, D. (1991). Multiple job-holding among farm families.
Iowa State University Press, Ames, IA.

Hanoch, G. & Rothchild, M. (1972). Testing the assumptions of production theory: a non
parametric approach. Journal of Political Economy, 80: 256-75.

Hardaker, J.B., Huirne, R., Anderson J.R. & Lien G. (2004). Coping with risk in agriculture.
CABI Publishing.

Hayami, Y. & Ruttan, V.W. (1970). Agricultural productivity differences among countries.
American Economic Review, 60: 895-911.

Hazell, P.B. & Hojjati, B. (1995). “Farm/Non-Farm Growth Linkages in Zambia.” Journal of
African Economies 4, 406-435.

Heckman, J. (1979). Sample selection bias as a specification error. Econometrica: 47, 153–
161.

Hill, B. (2000). Farm Incomes, Wealth and Agricultural Policy. Third Edition, Ashgate
Publishing, Aldershot

Huffman, W. & Lange, M. (1989). Off-farm work decisions of husbands and wives: Joint
decision making. Rev. Econ. Stat. 71, 471–480

Huffman, W.E. (1980). Farm and off-farm work decisions: The role of human capital.
Review of Economics and Statistics 62: 14-23.
 74

Huirne, R.B.M., Meuwissen, M., Hardaker, J.B. & Anderson J.R. (2000). Risk and risk
management in agriculture: An overview and empirical results. International Journal
of Risk Assessment and Management, 1: 125-136.

Ingosi, A. (2005). Economic Evaluation of Factor Influencing Maize Yield in the North Rift
Region of Kenya. Masters of Science Thesis, Colorado State University.

International Rice Research Institute (IRRI), (1997). IRRI rice facts. Los Baños, Philippines,
IRRI.

Jaenicke, E.C., Frechette, D.I. & Larson, J.A. (2003). Estimating production risk and
inefficiency simultaneously: An application to cotton cropping systems. Journal of
Agricultural and Resource Economics 28(3): 540-557.

Jolliffe, D. (2004). The impact of education in rural Ghana: Examining household labour
allocation and returns on and off the farm. J. Dev. Econ. 73, 287–314.

Just, R.E., & Pope, R.D. (1978). Stochastic Specificationof Production Functions and
Economic Implications. Journal of Econometrics 7:67– 86.

Just, R. & Pope, R. (1979). Production function estimation and related risk considerations.
American Journal Agricultural Economics, 61(2): 276–284.

Kalirajan, K.P. (1981). An econometric analysis of yield variability in paddy production.

Canadian Journal of Agricultural Economics, 29: 283-94

Kalirajan, K.P. & Shand, R.T. (1985). Types of education and agricultural productivity: A
quantitative analysis of Tamil Nadu rice farming. Journal of International
Development Studies, 21: 232-43.

Kalirajan, K.P. & Shand, R.T. (1989). A generalised measure of technical efficiency.
Applied Economics, 21: 25–34.

Kalirajan, K. P. & Shand, R. T. (1999). Frontier production functions and technical efficiency
measures. Journal of Economic Survey, 13 (2): 149 - 172.

Khai, H.V. & Yabe, M. (2011). Technical efficiency analysis of rice production in Vietnam.
Journal of International Society for Southeast Asian Agricultural Sciences, 17(1):135-
146.

Kodde, D.A & Palm, F.C. (1986). Wald Criteria for Jointly Testing Equality and Inequality
Restrictions. Econometrica 54:1243-1246.
 75

Kortelainen, M. (2008). Estimation of semi parametric stochastic frontiers under shape

constraints with application to pollution generating technologies, MPRA, Paper No
9257.

rd
Kuhn, & Offutt, S. (1999). Farm Policy in an Era of Farm Diversity. Choices, 3 Quarter: 37-
38.

Kumbhakar, S., Biswas, B., & Bailey, D. (1989). A study of economic efficiency of Utah
dairy farmers: A system approach. Rev. Econ. Stat. 71, 595–604.

Kumbhakar, S.C., Ghosh, S. & McGuckin, J. T. (1991). A generalized production frontier

approach for estimating determinants of inefficiency in U.S. dairy farms. Journal of
Business and Economic statistics 9, pp. 279-86.

Kumbhakar, S. C. (1993). Production risk, Technical efficiency, and panel data. Econ. Letters
41:11-16.

kumbhakar, S., and C. Lovell. 2000. Stochastic Frontier Analysis. Cambridge University
Press.

Kumbhakar, S.C. (2002). Specification and estimation of production risk, risk preferences
and technical efficiency. American Journal of Agricultural Economics, 84: 8-22.

Lanjouw, P. (1999). Rural Non Agricultural Employment and Poverty in Ecuador. Economic
Development and Cultural Change, 48(1):91-122.

Lass, D., Findeis, J., & Hallberg, M. (1991). Factors affecting the supply of off-farm labour:
A review of empirical evidence. In M.C. Hallberg, J.L. Findeis, A.

Lien, G., Flaten, O., Jervell, A.M., Ebbesvik, M., Koesling, M. and Valle, P.S. (2006).
Management and risk characteristics of part-time and full-time farmers in Norway.
Review of Agricultural Economics 28:111-131.

Lien, G., Kumbhakar, S.C. & Hardaker, J.B. (2008). Determinants of part-time farming and
its effect on farm productivity and efficiency. Paper prepared for presentation at the
107th EAAE Seminar "Modelling of Agricultural and Rural Development Policies".
Sevilla, Spain, January 29th -February 1st, 2008.

Liu, Z. & Zhuang, J. (2000). Determinants of Technical Efficiency in Post-Collective

Chinese Agriculture: Evidence from Farm-Level Data. Journal of Comparative
Economics 28: 545-564.
 76

Lockhead, M.E., Jamison, D.T. & Lan, L.J. (1980). Farmer education and farmer efficiency:
A survey. Economic Development and Cultural Change, 29: 37–76.

Makombe, G., Namara, R., Hagos, F., Awulachew, S.B., Ayana, M & Bossio, D. (2011). A
comparative analysis of the technical efficiency of rain-fed and smallholder irrigation
in Ethopia. International Water Management Institute (IWMI) Working Paper 143.

Meeusen, W. & van den Broeck, J. (1977). Efficiency estimation from Cobb-Douglas
production functions with composed error. International Economic Review, 18: 435–
44.

Mishra, A.K. and Goodwin, B.K. (1997). Farm income variability and the supply of off-farm
labor. American Journal of Agricultural Economics 79: 880-887.

Mishra, A.K., H.S. El-Osta., M.J. Morehart., J.D. Johnson., & J.W Hopkins. (2002). Income,
Wealth, and the Economic Well-Being of Farm Households. Agricultural Economic
Report 812, Economic Research Service, United State Department of Agriculture,
Washington D.C.

Mochebelele, M.T. & Winter-Nelson, A. (2002). Migrant labour and farm technical
efficiency in Lesotho. World Development, 28(1): 143-53.

Morrison-Paul, C. J., Johnston, W. E. & Frengley, G. A. G. (2000). Efficiency in New

Zealand Sheep and Beef Farming: The Impacts of Regulatory Reform. Rev. Econ. and
Statis. 82: 325-37.

Musser, W.N. & Patrick, G.F. (2001). How much does risk really matter to farmers? Chapter
24 in Just & Pope (2002).

National Agricultural Extension and Research Liaison Services (NAERLS) & Projects
Coordinating Unit (PCU), (2006). Field situation assessment of 2006 wet season
agricultural production in Nigeria. Report of a study conducted by NAERLS and
PCU.

Nauges, C., O’Donnell, C. J. & Quiggin, J. (2011). Uncertainty and technical efficiency in
Finnish agriculture: a state-contingent approach. European Review of Agricultural
Economics 38 (4): 449–467

National Population Census (NPC) (2006). Census Report. Abuja FCT: NPC.

Ndagi, J.O. (1984). Essentials of Research Methodology for Nigeria Educators. University
Press Limited, Ibadan.
 77

Nwaru, J. C. & Iheke, O. R. (2010). Comparative analysis of resource use efficiency in rice
production systems in Abia State of Nigeria. Journal of American Science 6(11).

Obwona, M. (2006). Determinants of technical efficiency differentials amongst small and

medium scale farmers in Uganda: A case of tobacco growers. African Economic
Research Consortium, (AERC) Research Paper 152.

Offutt, S. (2002). The Future of farm policy analysis: A household perspective. American
journal of Agricultural Economics 84(5): 1189-1200.

Ogundele, O.O. & Okoruwa, V. O. (2006). Technical efficiency differentials in rice

production technologies in Nigeria. African Economic Research Consortium (AERC)
Research Paper 154, Nairobi.

Ogunyinka, E. O. & Ajibefun, I. A. (2003). Determinants of Technical Inefficiency in Farm

Production: The Case of NDE Farmers in Ondo State, Nigeria. Selected Paper
prepared for presentation at the Western Agricultural Economics Association Annual
Meeting at The Denver Adam’s Mark Hotel, Denver, Colorado, July 11-15, 2003.

Okafor, C.F. (1982). Environmental constraints and part-time farming in south eastern
Nigeria. Springer, 6(4): 359-366. Retrieved from
http://www.jstor.org/stable/41142734 on 14/05/2012.

Okorji, E.C. & Onwuka, K.O. (1994). A Comparative Analysis of Costs and Returns of Non-
irrigated and Irrigated Rice Production Systems in Uzo-Uwani Local Government
Area of Enugu State, Nigeria. Journal of the West African Farming Systems Research
Network, 4(2), 12-19.

Olayide, S.O. & Heady, E.O. (1982). Introduction to Agricultural Production Economics.
University Press, University of Ibadan, Nigeria.11 – 17, 25 – 52,108 – 116.

Onoja, A. O. & Achike, A.I. (2010). Comparative analyses of farm resource efficiencies in
rice production under small scale farmer-managed irrigation schemes and rainfed
systems in Kogi State, Nigeria. Research Journal of Agriculture and Biological
Sciences, 6(6): 1047-1054.

Organisation for Economic Cooperation and Development (OECD), (2000). Income Risk
Management in Agriculture. OECD Paris.

Organisation for Economic Cooperation and Development (OECD), (2009). Managing risk in
Agriculture: A Holistic View. OECD Paris.
 78

Peng, C.J., Lee, K.L. & Ingersoll G. (2002). An introduction to logistic regression analysis
and reporting. Journal of Educational Research, 96(1): 1-14.

Phimister, E. & Roberts D. (2002). “The Effects of Off-Farm Work on Production Intensity
and Output Structure.” Workshop on the Importance of the Household-Firm Unit in
Agriculture, University of London.

Phimmister, E. & Roberts, D. (2006). The effect of off-farm work on the intensity of
agricultural production. Environmental & Resource Economics, 34: 493-515.

Reardon, T. (1997). “Using Evidence of Household Income Diversification to Inform Study

of the Rural Non farm Labour Market in Africa.” World Development 25, 735-747.

Reardon, T., Delgado, C. & Matlon P. (1992). “Determinants and Effects of Income
Diversification amongst Farm Households in Burkina Faso.” Journal of Development
Studies 28, 264-296.

Reifschneider, D. & Stevenson, R. (1991). Systematic departures from the frontier: A

framework for the analysis of firm inefficiency. International Economic Review, 32:
715-23.

Reinhard, S., Lovell, C. A. K. & Thijssen, G. (1999). "Econometric Estimation of Technical

and Environmental Efficiency: An Application to Dutch Dairy Farms." Amer. J. Agr.
Econ. 81:44-60.

Seiford, L.M. and Thrall, R.M. (1990). Recent developments in DEA: The mathematical
programming approach to frontier analysis. Journal of Econometrics 46: 7-38.

Serra, T., Goodwin, B.K. and Featherstone, A.M. (2004). Determinants of investments in
non-farm assets by farm households. Agricultural Finance Review 64: 17-32.

Serra, T., Goodwin, B.K. and Featherstone, A.M. (2005). Agricultural policy reform and off-
farm labour decisions. Journal of Agricultural Economics 56: 271-285.

Seyoum, E.T., Battese, G.E & Fleming, E.M. (1998). Technical efficiency and productivity
of maize producers in eastern Ethopia: A study of farmers within and outside the
Sasakawa Global 2000 Project. Agricultural Economics 19 (3): 341-348.

Sherlund, S., Barrett, C. & Adesina, A. (2002). Smallholder technical efficiency controlling
for environmental production conditions. J. Dev. Econ. 69, 85–101.
 79

Shiskhko, R. & Bernard R. (1976). "The Economics of Multiple Job Holding." Amer. Econ.
Rev. 66: 298-308.

Singh , B.N., Fagada, S. , Ukwungwu, M.N., William, C., Jagtap, S.S., Oladimeji, O., Effisue
A., & Okhidievbie, O. (1997). Rice growing environments and biophysical constraints
in different agroecological zones of Nigeria. Meterol. J. 2:35-44.

Smale, M., Hartell, J., Heisey, P. W. & Senauer, B. (1998). The contribution of genetic
resources and diversity to wheat production in the Punjab of Pakistan. American
Journal of Agricultural Economics, 80: 482-93.

Smith, K.R. (2002). Does off-farm work hinder smart farming? Agricultural Outlook.
Economic Research Service/USDA, September, 28-30.

Sumner, D.A. (1982). The off-farm labour supply of farmers. American Journal of
Agricultural Economics, 64: 499-509.

Tijani, A. A. (2006). Analysis of the technical efficiency of rice farms in Ijesha Land of Osun
State, Nigeria. Agrekon, 45 (2): 126.

Traxler, G., Falck-Zepeda, J., Ortiz-Monasterio, R. J. I. & Sayre, K. (1995). Production risk
and the evolution of varietal technology. American Journal of Agricultural
Economics, 77: 50-64.

Tveterbs, R. (1999). Production risk and productivity growth: Some findings for Norwegian
salmon aquaculture. Journal of Productivity Analysis, 12: 161-79.

Ukwu, I.U., Anyanwu, E.A., Imaga, E.U.L., Akwuba, M., Ofong, I., Nwakoby, F., Omeje,
K., Onyeukwu, O.E., Umoh, B.D. & Ugwuonah, G. (1998). Enugu State Poverty
Report, Institute for Development Studies University of Nigeria Enugu Campus,
Enugu.

United State Department of Agriculture (USDA) (2001). Income wealth and the economic
well being of farm households.

Villano, R. & Fleming, E. (2006). Technical Inefficiency and Production Risk in Rice
Farming: Evidence from Central Luzon Philippines. Asia Economic Journal
20(1):29–46.
 80

Wang, H. J. & Schimidt, P. (2002). One-step and two-step estimation of the effects of
exogenous variables on technical efficiency levels. Journal of Productivity Analysis,
18(2): 129-144.

Wang, H. (2002). Heteroskedasticity and non-monotonic efficiency effects of a stochastic

frontier model. J. Product. Anal. 18, 241–253.

West Africa Rice Development Association (WARDA) (2001). Report of the Stakeholders
Workshop. Ibadan, Nigeria, 8-9 November.

West Africa Rice Development Association (WARDA) (2002). News Release, Buoake, Cote
D’Ivoire; May.

West Africa Rice Development Association (WARDA) (2003). The Nigerian rice economy
in a competitive world: constraints, opportunities and strategic choices: strategy for
rice sector revitalization in Nigeria. A paper on rice sector study implemented by
WARDA; Presented at a Technical Workshop 20-21 August, Ibadan, pp: 1- 10.

West Africa Rice Development Association (WARDA) (2007). Overview of recent

developments in the sub-Saharan African rice sector. Cotonoou, Benin.

Wilson, P., Hadley, D., Ramsden, S. & Kaltas, I. (1998). Measuring and explaining technical
efficiency in UK potato production. Journal of Agricultural Economics 49: 294-305.

Wooldridge, J. (2002). Econometric Analysis of Cross Section and Panel Data. The MIT
Press, Cambridge, MA.