Productivity growth in food crop production in Imo State, Nigeria

Abstract

Agriculture plays pivotal roles in Nigeria including food security, employment, foreign exchange earnings and poverty reduction. This study examined the growth in food crop productivity in Imo State in Nigeria with emphasis on the decomposition of total factor productivity (TFP) into technical progress, changes in technical and allocative efficiency and scale effects. A panel data set comprising 210 observations drawn over 2001 – 2007 periods was used. Using the translog stochastic frontier production function, the decomposition components were computed. The results showed that TFP decreased through time, while technical change was negative, implying downward shift of the production frontier. As a major component, technical change was the main constraint to the achievement of high levels of TFP during the study period. The scale effect, which is generally bigger than technical change component shows that the sampled farms had not taken advantage of scale economies. Furthermore, the allocative efficiency had an average magnitude closer to the scale effect and points towards decreases in the efficiency with which production factors are allocated. This is an indication of a decline in technical efficiency. We suggest reforms in the Agricultural Development Programmes (ADPs) geared towards enhancing their capacity in extending novel technologies and innovations to farmers. RÉSUMÉ Agriculture plays pivotal roles in Nigeria including food security, employment, foreign exchange earnings and poverty reduction. This study examined the growth in food crop productivity in Imo State in Nigeria with emphasis on the decomposition of total factor productivity (TFP) into technical progress, changes in technical and allocative efficiency and scale effects. A panel data set comprising 210 observations drawn over 2001 – 2007 periods was used. Using the translog stochastic frontier production function, the decomposition components were computed. The results showed that TFP decreased through time, while technical change was negative, implying downward shift of the production frontier. As a major component, technical change was the main constraint to the achievement of high levels of TFP during the study period. The scale effect, which is generally bigger than technical change component shows that the sampled farms had not taken advantage of scale economies. Furthermore, the allocative efficiency had an average magnitude closer to the scale effect and points towards decreases in the efficiency with which production factors are allocated. This is an indication of a decline in technical efficiency. We suggest reforms in the Agricultural Development Programmes (ADPs) geared towards enhancing their capacity in extending novel technologies and innovations to farmers.

Full text

Page 1

African Crop Science Journal, Vol. 18, No. 3, pp. 119 - 131
Printed in Uganda. All rights reserved
ISSN 1021-9730/2010 $4.00
©2010, African Crop Science Society
PRODUCTIVITY GROWTH IN FOOD CROP PRODUCTION IN IMO STATE,
NIGERIA
C.E. ONYENWEAKU, I.N. NWACHUKWU and T.C. OPARA
Department of Agricultural Economics, Michael Okpara University of Agriculture, Umudike,
Abia State, Nigeria
Corresponding authour: ifnwachukwu@yahoo.com; nwachukwu.ifeanyi@mouau.edu.ng
(Received 4 May, 2010; accepted 2 September, 2010)
ABSTRACT
Agriculture plays pivotal roles in Nigeria including food security, employment, foreign exchange earnings and poverty
reduction. This study examined the growth in food crop productivity in Imo State in Nigeria with emphasis on the
decomposition of total factor productivity (TFP) into technical progress, changes in technical and allocative efficiency
and scale effects. A panel data set comprising 210 observations drawn over 2001 – 2007 periods was used. Using
the translog stochastic frontier production function, the decomposition components were computed. The results
showed that TFP decreased through time, while technical change was negative, implying downward shift of the
production frontier. As a major component, technical change was the main constraint to the achievement of high
levels of TFP during the study period. The scale effect, which is generally bigger than technical change component
shows that the sampled farms had not taken advantage of scale economies. Furthermore, the allocative efficiency
had an average magnitude closer to the scale effect and points towards decreases in the efficiency with which
production factors are allocated. This is an indication of a decline in technical efficiency. We suggest reforms in the
Agricultural Development Programmes (ADPs) geared towards enhancing their capacity in extending novel
technologies and innovations to farmers.
Key Words: Productivity decomposition, scale effect, allocative efficiency
RÉSUMÉ
Agriculture plays pivotal roles in Nigeria including food security, employment, foreign exchange earnings and poverty
reduction. This study examined the growth in food crop productivity in Imo State in Nigeria with emphasis on the
decomposition of total factor productivity (TFP) into technical progress, changes in technical and allocative efficiency
and scale effects. A panel data set comprising 210 observations drawn over 2001 – 2007 periods was used. Using
the translog stochastic frontier production function, the decomposition components were computed. The results
showed that TFP decreased through time, while technical change was negative, implying downward shift of the
production frontier. As a major component, technical change was the main constraint to the achievement of high
levels of TFP during the study period. The scale effect, which is generally bigger than technical change component
shows that the sampled farms had not taken advantage of scale economies. Furthermore, the allocative efficiency
had an average magnitude closer to the scale effect and points towards decreases in the efficiency with which
production factors are allocated. This is an indication of a decline in technical efficiency. We suggest reforms in the
Agricultural Development Programmes (ADPs) geared towards enhancing their capacity in extending novel
technologies and innovations to farmers.
Mots Clés: Productivity decomposition, scale effect, allocative efficiency

Page 2

C.E. ONYENWEAKU et al.
2
INTRODUCTION
Agriculture plays pivotal roles in economic
development over the past several decades.
These include food security, employment,
foreign exchange earnings and poverty
reduction (CBN, 1998; NPC, 2004). Despite the
enormous contributions of agriculture to the
Nigerian economy over the years, the sector has
slipped into a system decline, particularly in the
past three decades since the petroleum industry
replaced the sector as the main earner of
government revenue (NISER, 2004). As a result,
productivity has been the major focus of
agricultural research over the last century. With
improvements in plant and animal genetics,
research has paid off with major productivity
gains such as the tripling of corn yield over the
last 50 years (Njoku, 2005).
Ideally, productivity growth in the agricultural
sector is considered important if the sector is
to improve at a rate equal to or greater than the
population growth rate to meet the demand for
food and raw materials. Also, productivity
performance in the agricultural sector is critical
to improvement in the economic well being of
the entire country (Alabi, 2005).
The general decline in agricultural
productivity has translated into gross
incapacitation of the sector in meeting the rising
food demand and by extension led to perennial
for instance food shortages, soaring food prices
and massive importation ( Imodu, 2005;
Onyenweaku and Nwaru, 2005). Tanko et al.
(2006) averred that Nigeria’s food deficient
situation has been worsened by declining farm
productivity owing to inefficient production
techniques, poor resource base and declining
soil productivity among others.
Although several reports (FAO, 2004; Dayo
et al., 2008; Fakayode et al., 2008; Ebong et
al., 2009) have attributed the low rates of
agricultural production to low rates of
technologies adoption and dependence on
indigenous knowledge, such knowledge is
considered inadequate given the increasing
demand from agriculture. The need, therefore,
for improved strategies becomes imperative.
To this end, a number of strategies have been
advocated to address this situation.
Combinations of farm enterprises with the aim of
increasing the level of farm resource and making
efficient use of resources already committed to
the food sub-sector was advocated by Tanko et
al. (2006). Stabilisation policies to reduce inflation
and subsidies in the form of cheap credit was
suggested by Okoye (2006) to assist farmers in
acquiring inputs. Onyenweaku and Nwaru (2005)
opined the efficient utilisation of productive
resources as ways of increasing productivity.
This study examined productivity growth, its
decomposition and rate of change in food crop
production in Imo State in Nigeria.
MATERIALS AND METHODS
Study area. The study was conducted in Imo
State in Nigeria which lies between latitude 50
10’ and 60 35’ North of the equator and between
longitude 60 35’ and 70 31’ East of the Greenwich
meridian. This is a tropical rainforest zone,
located in the Southeastern zone of Nigeria. Imo
State, one of the 36 states in Nigeria, has a
population of about 3.934 million people
disaggregated into 2.032 males and 1.903
females in 2006 (NPC, 2007). The state is
divided into 27 administrative units called local
government areas which are grouped into 3
agricultural zones of Owerri, Okigwe and Orlu.
Agriculture is the predominant occupation
of the people, for almost all the farm families
either as primary or secondary occupation. The
ecological zone favours the growing of tree
crops, roots and tubers, cereals, vegetables and
nuts. These crops are grown in small holder plots
usually in mixtures of at least two simultaneous
crops (Imo ADP, 1994).
Data collection. The study employed farm
level data elicited from the ADP’s yearly survey
for the period, 2001-2007 using past
questionnaires. The ADP annually collects
micro-economic data from a sample of
agricultural holdings in Imo State. Although the
collected data were based on individual
responses, state and zonal level aggregates were
used to define some variables such as land,
labour and capital. The sample comprised 210

Page 3

Growth in food crop production
3
observations that constituted the panel data used
by the study.
Data analysis. This study followed the primal
approach proposed by Kumbhakar (1990) in
which total factor productivity (TFP) growth was
decomposed into technical progress, changes in
technical and allocative efficiency and scale
effects. In order to estimate decomposition of
productivity growth, a stochastic frontier
production function approach was employed. A
deterministic frontier production function is
specified as:
Yit = f(Xit, t; β). exp(-uit) ...................………….
(1)
where γ is the scalar output of the ith farm (i =
1,….,N) in period t (t = 1,…..,T),
f(Xit, t; β) is the deterministic kernel of a
stochastic production frontier with technology
parameter vector â to be estimated, X = (X1,…,
XN ) > 0 is an input vector, t is a time trend
serving as a proxy for technical change, and u >
0 represents output – oriented technical
inefficiency. Technical efficiency is then
defined by
yit = exp (-uit) < 1.
(Xit, t; β)
The technical efficiency term can be either fixed
or random and the specification in Equation (1)
allows technical efficiency to vary over time. A
primal measure of the rate of technical change
is provided by
∆TC = In f(Xit, t; β) ……………………...
(2)
t
where ∆TC is positive if the exogenous
technical change shifts the production frontier
upward, given the inputs. A primal measure of
the rate of change in technical efficiency is given
as
∆TE = - “u
“t
where ∆TE is negative if technical efficiency
declines through time. ∆TE can be interpreted
as the rate at which a producer moves toward or
away from the production frontier, other things
kept constant.
To operationalise the model, the following was
specified:
TFP = ∆TC + (ε – 1) Σ (εn/ ε) Xn + Σ [(εn/ ε) - Sn]
Xn
+
…………………...................................... (3)
∆TE
where
TFP = Total Factor Productivity;
∆TC = Technical Change Component;
∆TE = Technical efficiency change component;
(ε – 1) Σ (εn/ ε) Xn = Scale effect component;
Σ[(εn/ ε) - Sn] Xn = Allocative efficiency change
component; and
Sn = Observed expenditure share of input.
In estimating the technical efficiency change,
the following Translog stochastic frontier model
was employed in line with Fan (1991);
Mazvimavi (2002); and Nwachukwu and
Onyenweaku (2007). It was specified as:
In Yit = β0 +β1 In X1 + β2 In X2 +β3 In X3 +β4 In
X4 +β5 In t +0.5 β6 In (X1)2 + 0.5 β7 In (X2)2 +
0.5 β8 In (X3)2+ 0.5 β9 In (X4)2 + 0.5 β10 In (t)2
+β11 (X1) In (X2) + β12 In (X1) In (X3) +β13 In
(X1) In (X4 ) + β14 In (X1) In (t ) +β15 In (X2) In
(X3) +β16 In (X2) In (X4) + β17 In (X2) In (t) +β18
In (X3) In (X4) + β19 In (X3) In (t) + β20 In (X4)
(t-) + Vi - Ui …......................... (4)
where In = Natural logarithm
The subscripts i and t represent the ith sample
farmer in the period t.
Yit = Total value of output in Naira of the ith
farmer in the period t;
X1 = Farm size measured as total land area in
hectares;
X2 = Quantity of fertiliser used in kg;
X3 = Labour in man days used in production;
X4 = Capital Inputs (values of farm implements
measured in Nigerian Naira);

Page 4

C.E. ONYENWEAKU et al.
4
t = The period of time in years the study was
conducted (t = 1….7);
β0 = Intercept;
β1 - β20 = Coefficients estimated;
Vit = Random noise error component; and
Uit = Technical efficiency error component.
Given that the study interest centered on the
estimation of productivity change, the estimates
of ∆TC, ∆TE, εn and ε were derived by means of:
∆TC = βt + βitt + Σβnt InXnit
.......……………… (5)
∆TE = Ui.ç.exp (-ç[t – T]) ……………………
(6)
εn = βn + 0.5Σβnk InXkit + 0.5βnt n =
1…N…..… (7)
ε = Σ[βn + 0.5Σβnk InXkit + 0.5 βnt t ]
..…………(8)
Empirically, technical change (∆TC), technical
efficiency (∆TE), elasticity of n output with
respect to each input (εn ) and the scale elasticity
(β) can be obtained empirically as follows:
∆TC = β5 + β10 + β1 In X1 + β2 In X2 + β3 In X3 +
β4
In
……………………..................................... (9)
X4
∆TE = Û.ç.e-ç(t –T) ……………………………..
(10)
εX1 = β1 + 0.5β6 In X1 + 0.5β11 InX2 + 0.5β12 In
X3 + 0.5β13 In X4 + 0.5 β14 t
.................................…(11)
εX2 = β2 + 0.5β7 In X2 + 0.5β11 InX1 + 0.5β15
InX3 + 0.5β16 In X4 + 0.5 β17 t
..................................…(12)
εX3 = β3 + 0.5β8 In X3 + 0.5β12 InX1 + 0.5β15 In
X2 + 0.5β18 In X4 + 0.5 β19 t
…...............................(13)
εX4 = β1 + 0.5β6 In X1 + 0.5β11 InX2 + 0.5β12 In
X3 + 0.5β13 In X4 + 0.5 β14 t
.................................…..(14)
ε = εX1 + εX2 +εX3 +εX4 ..………….…(15)
where:
εX1 = elasticity of output with respect to farm
size;
εX2 = elasticity of output with respect to
fertiliser;
εX3 = elasticity of output with respect to labour
input;
εX4 = elasticity of output with respect to capital
input; and
ε = scale effect.
RESULTS AND DISCUSSION
The mean TFP growth rates decreased through
time at an increasing rate from 3.5% in 2001 to
1.7% in 2007 (Table 1). This conforms to the
findings of several workers (Imodu, 2005;
Njoku, 2005; Onyenweaku and Nwaru, 2005;
Tanko et al., 2006) which confirmed that
agricultural performance has declined in
Nigeria. It has been noted also that agricultural
productivity has slipped into a system decline,
particularly in the past three decades since the
petroleum industry replaced the sector as the
main earner of government revenue and foreign
exchange earnings (NISER, 2004).
TFP growth decomposed into technical
change, scale effect, technical efficiency and
allocative efficiency (Table 1) revealing that the
technical change is negative for the period
studied. This implies that the exogenous
technical change shifts the production frontier
downward given the inputs. This could be as a
result of inefficient production techniques, poor
resource base, declining soil productivity,
predominance of primitive techniques of
agricultural production, inadequate supply of
credit, low capital investment, use of crude
implements to mention but a few (Dayo et al.,
2008; Onyenweaku and Nwaru, 2005). It further
implies that technical change has been the main
constraint to the achievement of high levels of
TFP during the reference period. This result is
consistent with the findings of Nkamleu (2004),
Ogunyinka and Langemeier (2004) who had a
similar outcome. Technical efficiency was
drawn from Table 2.

Page 5

Growth in food crop production
5
TABLE 1. Estim ation of the total factor productivity changes
Year ∆TC ε– 1Σ(εn/ ε)Xn Σ[(εn/ ε) - Sn]Xn ∆TE TFP
2001
2002
2003
2004
2005
2006
-2.669
-2.693
-2.986
-3.132
-3.150
-2.536
-1.370
-0.017
0.640
0.018
0.044
0.203
-0.015
-0.027
-0.267
0.095
-0.046
0.015
0.469
0.460
0.448
0.416
0.356
0.530
-3.585
-2.223
-1.872
-2.798
-3.398
-1.788
Source: Com puted from field survey Data and Frontier 4.1 Output
TABLE 2. Maxim um likelihood estim ates of the Stochastic Translog Production Function
Production factor Param eter Coefficient Standard error t-value
Constant term
Farm size
Fertiliser
Labour
Capital
Tim e
Farm size2
Fertiliser2
Labour 2
Capital2
Tim e2
Farm size x fertiliser
Farm size x labour
Farm size x capital
Farm size x tim e
Fertiliser x labour
Fertiliser x capital
Fertiliser x tim e
Labour x capital
Labour x tim e
Capital x tim e
Diagnostic statistics
Log – likelihood function
Total variance (s2)
Variance ratio (g)
LR Test
β0
β1
β2
β3
β5
β5
β6
β7
β8
β9
β10
β11
β12
β13
β14
β15
β16
β17
β18
β19
β20
1 1.09
0.525
-0.185
0.017
-0.614
0.697
-0.282
-0.022
0.035
0.058
-0.999
0.03
0.044
-0.105
0.1 19
-0.077
0.032
0.236
-0.069
-0.511
0.235
0.718
0.209
0.065
0.218
0.255
0.495
0.037
0.017
0.006
0.022
0.262
0.021
0.032
0.026
0.082
0.024
0.055
0.307
0.179
0.122
0.033
15.438* * *
2.504* * *
-2.839* * *
7.915* * *
-2.410* *
1.409
-7.530* * *
-1.288
5.710* * *
2.678* * *
-3.815* * *
1.437
1.381
-4.063* * *
1.46
-3.233* * *
2.075* *
7.682* * *
-3.883* * *
-4.197* * *
7.100* * *
-223.041
2.945
0.999
43.76
0.169
0
17.592* * *
1126637.900* * *
Source: Com puted from Frontier 4.1 MLE/ Survey data
Note: * * * , * * , * indicates statistically significant at 1.0, 5.0 and 10.0% probability respectively
The scale effect, which was generally bigger
than technical change, showed that the sampled
farms on the average had not taken advantage of
scale economies, though it fluctuated
throughout the period of analysis. By
implication, larger farms had no advantage over
the small ones in the study area. This
consolidates the findings of Owualah (1999) and
Sanusi (2003) which admitted that small scale
enterprises enjoy a competitive advantage over
large scale enterprises and the existence of an
inverse relationship between firm size and firm
growth in developing countries.