See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311699620

Adoption of technology and economic efficiency of the small-holder

mango farmers in Indonesia

Article · January 2015

CITATIONS READS

7 199

3 authors:

Lies Sulistyowati Ronnie Susman Natawidjaja

Universitas Padjadjaran Universitas Padjadjaran
42 PUBLICATIONS   74 CITATIONS    34 PUBLICATIONS   145 CITATIONS   

SEE PROFILE SEE PROFILE

Bobby Rachmat Saefudin

Ma'soem University
54 PUBLICATIONS   151 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Coffee View project

Horticulture View project

All content following this page was uploaded by Bobby Rachmat Saefudin on 18 January 2020.

The user has requested enhancement of the downloaded file.

I J A B E R, Vol. 13, No. 7 (2015): 4621-4645

ADOPTION OF TECHNOLOGY AND ECONOMIC

EFFICIENCY OF THE SMALL-HOLDER MANGO
FARMERS IN INDONESIA
Lies Sulistyowati1, Ronnie S. Natawidjaja2 and Boby Rahmat3

Abstract: The demand of mangoes for both domestic and export markets have been
continuously increased in the last two decades. Indonesia as the sixth mango producer in
the world is just contributed to a very small proportion (0.07 per cent). This was happened
due to only small number of mango farmers who are applying cultivation technology, so
the productivity and the quality of mangoes produced are very low. On the other hand,
number of imported mangoes tend to steadily increase.
This paper is intended to explain the response of mango farmers to mangoes market
demand related to factors affected on the application of mango cultivation technology, as
well as its economic efficiency respectively. This study was conducted in East and West
Java as two main mango producer provinces in Indonesia, through an explanatory survey.
The data was analyzed using Logit model and stochastic frontier analysis method in the
Cobb-Douglass production function.
This research indicated that the percentage of Indonesian farmers applying this
technology is still very low. The decision of mango farmers in applying off-season
technology were affected by the status of mango cultivation, number of trees, accessibility
upon information, market and capital, and level of education and number of followed
training. Mangoes production is significantly affected by numbers of tree, application of
off-season technology, human resources allocation, the use of pesticide and pest control
tools. Technical efficiency, allocation efficiency dan economic efficiency of Indonesian
mangoes farmers are respectively: 70,37%, 15,05%, and 11,19%. West Java has a lower
efficiency rate compared to East Java due to lack of experience and access to funding. The
major contributing factors for technical in-efficiency were the experience of the farmer,
credit and information accessibility.
Keywords: Economic-efficiency, Indonesia, Mango, Smallholder farmers, Technology-
adoption.

1,2.
Socio-Economic Department, Faculty of Agriculture, Padjadjaran University, Indonesia, Campus
Unpad-Jatinangor, Bandung 40600, Indonesia.
3.
Post graduate from Socio-Economic Department, Faculty of Agriculture, Padjadjaran University,
Indonesia, liesindra@yahoo.com, lies.s@unpad.ac.id
 4622 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

INTRODUCTION

Research Background
As the second largest agricultural subsector after food crops,horticulture in
Indonesia had involved 9.3 million number of farmers, or 23 per cent of 41.5
million total Indonesian farmers. (BPS, 2012). One of the top national horticultural
commodities is mangoes (Mangifera indica L.)
Aside from highly demanded and consumed by local market, this commodity
was also very potential to improve well-being of the farmers due to its high
economic value. Unfortunately, the economic potential of mango was not being
explored yet because of many obstacles, either technical or non-techical. These
were resulted in instability and lower quality of national mangoes production.
The BPS data had shown that mangoes production in Indonesia increased on
the average of 4.7 per cent a year (Figure 1). Natural harvest cycle on year of 1997,
2002, 2004, and 2009, indicating production of mangoes exponentially increases.
However,
The BPSthedata
increasing trendthat
had shown didmangoes
not supported by the
production stability of
in Indonesia production
increased on the
average of 4.7 per cent a year (Figure 1). Natural harvest cycle on year of 1997,mangoes
throughout years. On years of 1998, 2005, and 2010 for example, 2002, 2004,
andproduction had drastically
2009, indicating production ofdecreased
mangoes due to climateincreases.
exponentially and pests. Meanwhile,
However, with
the increasing
gradually better climate which supporting harvest cycle, mangoes production
trend did not supported by the stability of production throughout years.On years of 1998,
were
2005, andincreased
2010 fortoexample,
more than 2 tonsproduction
mangoes as on the year
had of 2011 anddecreased
drastically 2012. due to climate
and pests. Meanwhile, with gradually better climate which supporting harvest cycle, mangoes
production were increased
Figureto
1: more than 2 Mango
Indonesian tons asProduction
on the yearTrend
of 2011(Ton)
and 2012.

2500000
2000000
1500000
1000000
500000
0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013*)

Source : BPS, 2013

Figure
Source: BPS, 20131. Indonesian Mango Production Trend (Ton)

Unstable productioncondition
Unstable production condition andand
lowlow mango
mango quality
quality leadsleads to a dilemmatic
to a dilemmatic situation.
situation.
While While
Indonesia Indonesia
as one as one
of the five of the
largest five largest mango-producing
mango-producing nations
nations beside India, China,
Kenya, and Thailand (FAOSTAT, 2012), however, mangoes that were able to meet export
standard was very limited to only 1 per cent on the average from the total Indonesian
mangoes production (BPS, 2010-2011). On the other hand, Indonesian import data indicating
steadily increased of mangoes import during the last 10 years. (BPS, 2014)

Table 1.Production, Export, and Import Data of Indonesia

 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4623

beside India, China, Kenya, and Thailand (FAOSTAT, 2012), however, mangoes
that were able to meet export standard was very limited to only 1 per cent on
the average from the total Indonesian mangoes production (BPS, 2010-2011). On
the other hand, Indonesian import data indicating steadily increased of mangoes
import during the last 10 years. (BPS, 2014)

Table 1
Production, Export, and Import Data of Indonesia

Export
Year Production (Ton) Import (Ton)
Ton % dari Prod
2000 876.027 430 0,05 64
2001 923.294 425 0,05 186
2002 1.402.906 1.573 0,11 254
2003 1.526.474 559 0,04 348
2004 1.437.665 1.880 0,13 689
2005 1.412.884 941 0,07 869
2006 1.621.997 1.182 0,07 966
2007 1.818.619 1.198 0,07 1.088
2008 2.105.085 1.908 0,09 969
2009 2.243.440 1.415 0,06 821
2010 1.287.287 999 0,08 1.129
Average 1.514.153 1.137 0,07 671
Sumber: BPS, 2011

Natawidjaja et al. (2009) had mentioned that Indonesian mango production

were dominated by lowly-cared and not intensively productive farm, being grown
at the home yard and mixed farming, causing low productivity and its dependency
on weather and rainfall condition. Although the Ministry of Horticulture had
already introduced modern agriculture method and off-season technology 15 years
ago, however, only small number of farmers implemented the those technology.
This condition leads to both low quality as well as low quantity of mangoes with
uncertain continuity of its production supply. Low quality of Indonesian mangoes
was indicated by unattractive physical appearance and highly-varied product
(form, size, colour, and ripeness). Moreover, supply continuity problem due to
low level of off-season technology implementation that may cause harvest season
only once a year, starting from November up until February. This condition may
lead to a low price during high supply of harvest season, and the rest season of
the year mangoes supply were very low. This condition was also mentioned by
4624 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

Wandschneider, Baker, and Natawidjaja (2012), that only small percentage of

mangoes were being cultured optimally with low level of technology. Therefore,
studies regarding factors affecting off-season technology implementation and
its efficiency on technical, allocation, and economical view are necessary to be
conducted for mango development program in Indonesia.
This research aims to: 1) describe current off-season technology adoption
level and identify factors that are affecting mango farmers’ decision regarding
its adaption, 2) analyze technical, allocation, and economic efficiency level of
Indonesian mango farming, and 3) identify factors that may cause technical
inefficiency.

METHODOLOGY
This study was conducted by using explanatory survey method. Econometric
model was used to test farmers’ decision model on the off-season technology
implementation, while efficiency level is analysed using stochastic frontier method
with Cobb-Douglas production function.

Locations and Sampling Method

The study was conducted in the Provinces of East and West Java as the first and the
second highest mango producers in Indonesia. Meanwhile, location of the study
and sampling were determined by using Multi Stage Cluster Random Sampling
Method. Mango farmers population data was obtained from the data provided by
BPS. From each province, 3 sub-provinces were chosen based on each production
quantity.

Table 2
Research Location in Indonesia

West Java Province East Java Province

1. Cirebon 1. Probolinggo
2. Majalengka 2. Pasuruan
3. Kuningan 3. Bondowoso

Moreover, 3 districts were chosen from each of the sub-province where two of
the sub-provinces with high production, and another one from the sub-province
with low production. Therefore, throughout these process location of the study
areas consisted of 2 provinces, 6 sub-provinces, and 18 districts. Mango farmers
were then randomly sorted out based on the data provided by BPS that were
updated every 10 years. From each of the sub-district, the sample was randomly
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4625

selected so that 633 total mango farmers were chosen, 319 from East Java and 314
from West Java.

Data Analysis Method

Data Analysis Method
1.1).Off-season Technology
Off-season Technology Implementation
Implementation ModelModelused : used:
��� � �� � �� ���� � �� �� � �� ��� � �� ���� � �� ��� � �� ��� � �� ����� � �� ����
� �� ���� � ��� ���� � ��� ���� � ��� ����
Where :
Where:
Yos = Implementation of Mangoes Off-season Technology (yes / no)
SphY i = Tree = ownership
Implementation status (rent of /Mangoes
owner-operator)
Off-season Technology (yes / no)
�� os= Farmer’s age
���Sph = i Farmer’s
= Tree ownership
formal education status
level (rent / owner-operator)
���U � = Farming Data
= Farmer’s Analysis
experience Method
age(year)
i
��� ` = Number 1). Off-season
of family memberTechnology Implementation Model used :
Pd
��� =i Number��� = Farmer’s formal
� �tree education level
of total � � �� ���� � �� �� � �� ��� � �� ���� � �� ��� � �� ��� � �� ����� � �� ����
���� � =i Information
Pgl = Farming accessibility
experience �level (year)� ��� ���� � ��� ���� � ��� ����
�� ����
���� = Market Where accessibility
: level
Ak = Number of family member
���� =i Capital Yos accessibility
= Implementation level of Mangoes Off-season Technology (yes / no)
����Jpi= Water = Sph Number
resources of total tree
i = Tree ownership status (rent / owner-operator)
����Info = Tool = � /Information
machine resources
= Farmer’s age
accessibility level
�
���� = i Training �� = Farmer’s formal education level
b o….,nPaS =i Regression
= Market
� accessibility level
coefficient
���� = Farming experience (year)
Modi = ��Capital accessibility level
� ` = Number of family
2). Stochastic FrontierProduction Function member Analysis
Sar
Analysis = ��
i method
Water = Number
resources of total tree
� used to decide production factors that directly affects production output is
the Sal = ����
Stochastic Tool � = /Information
Frontier Production
machine accessibility level
Function Analysis.
resources Mangoes production in West Java,
i
East Java, and ��� � = Market
Indonesia, eachaccessibility
was analysedlevel by using production function model.
Peli = Training
���� = Capital � accessibility level
�� � � � �= Water
b o….,n = Regression
��� � � � � � � resources
coefficient
�� �� � � �� � �� (1)
2. Stochastic ����Frontier= Tool��� / machine resources
Production Function Analysis
Where : ���� = Training
Analysis
�� = output method
bofo….,nmangoused to decide
production
= Regression production
from
coefficientfarmer i(kg) factors that directly affects production
output
�� � = quantityis the Stochastic
of mango Frontier
tree on farmerProductioni (unit) Function Analysis. Mangoes production
in��West
�
= Java,
quantity East
of Java,
organic and
fertilizer
2). Stochastic FrontierProductionIndonesia,
used eachFunction
by farmer was analysed
i (kg) Analysis by using production
function
�� � = quantity model. of inorganic
Analysis method fertilizer
used used by farmer
to decide i(kg) factors that directly affects production output
production
= quantity
�� � Where: theofStochastic
pesticide used by farmer
Frontier i(liter) Function Analysis. Mangoes production in West Jav
Production
�� � = quantity East of Java, and Indonesia,
additional labor used by each was i(HOK)
farmer analysed by using production function model.
�
�� � = quantity of fly trapper used by farmer i(unit)
�� � = quantity �� �� � ��used
of ZPT fertilizer �� by � � �� ��i(litre)
farmer �
� �� � �� (
(1)
�� = constant (parameter vector) ���
�� = parameter Where : of production function (� � 1��� � � � ��)
vector
Y = output
�� =i random �� errorof mango production from farmer i(kg)
term of mango production from farmer i(kg)
= output
�� = nonnegative error termof mango tree on farmer i (unit)
�� � = quantity
�� � = quantity of organic fertilizer used by farmer i (kg)
�� is the random �� � = error component
quantity that represent
of inorganic fertilizerexternal
used by factor
farmer (weather,
i(kg) pests, etc.) which
distribution is symmetrical and normal (�� ~�(0� ��� )). �� is the non-negative random variable
that representing
�� � = internal
quantity factor
of pesticide used by farmer i(liter)
affecting technical inefficiency level with half-normal
distribution. �(� = quantity
� �� ~|�(0� ��� )|).of Error
additionalcomponentlabor used by farmer
�� represent i(HOK)
internal factors controllable by
 b o….,n = Regression coefficient
2). Stochastic FrontierProduction Function Analysis
2). Stochastic FrontierProduction Function Analysis
Analysis method used to decide production factors that directly affects production ou
Analysis method used to decide production factors that directly
the Stochastic Frontier Production Function Analysis. Mangoes production in Wes
the Stochastic Frontier Production Function Analysis. Mangoe
East Java, and Indonesia, each was analysed by using production function model.
East � Java, and Indonesia, each was analysed by using production
4626 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat �
�� �� � �� � � �� �� �� � � �� � ��
�� �� � �� � � �� �� �� � � �� � ��
���
X1i = quantity of mango tree on farmer i (unit) ���
Where :
Where :
X2i = quantity
�� = of organic
output fertilizer
of mango used byfrom
production farmerfarmeri (kg)
i(kg)
�� = output of mango production from farmer i(kg)
�� � = of
X3i = quantity quantity of mango tree used
on farmer i (unit)
inorganic fertilizer
�� � = quantity mango i(kg)
byoffarmer tree on farmer i (unit)
�� � = quantity of organic fertilizer used by farmer i (kg)
X4i = quantity of pesticide used �� �by = quantity of organic fertilizer used by farmer i (kg)
farmer i(liter)
�� � = quantity of inorganic fertilizer used by farmer i(kg)
X5i = quantity �� � =used quantityfarmer
of inorganic
i(HOK) fertilizer used by farmer i(kg)
�� � = of additional
quantity labor
of pesticide used by by farmer i(liter)
�� � = quantity of pesticide used by farmer i(liter)
�� � = of
X6i = quantity quantity of additional
fly trapper used bylabor
farmer by farmer i(HOK)
usedi(unit)
�� � = quantity of additional labor used by farmer i(HOK)
�
X7i = quantity = quantity of
of ZPT fertilizerfly trapper used by farmer i(unit)
i(litre)
�� ��used�
by farmer
= quantity of fly trapper used by farmer i(unit)
�� � = quantity of ZPT fertilizer used by farmer i(litre)
β0 = constant (parameter vector) �� � = quantity of ZPT fertilizer used by farmer i(litre)
�� = constant (parameter vector)
βj = parameter vector ofvector �� = constant
production function(parameter
() vector)
�� = parameter of production function (� � 1��� � � � ��)
�� = parameter vector of production function (� � 1��� � � � ��)
�� =
νi = random random
error term error term
�� = random error term
�� = nonnegative error term
ui = onnegative error term �� = nonnegative error term
νi is the random��error
is thecomponent
random errorthat representthat
component external
represent factor (weather,
external factorpests, (weather, pests, etc.)
�� is the random error component � that represent external factor
distribution
etc.) which distribution is symmetricaland
is symmetrical normal (�� ~�(0� �� )). �� .isisthe
andnormal the non-negative
non- random v
distribution is symmetrical and normal (�� ~�(0� ��� )). �� is the n
that representing internal
negative random variable that representing factor affecting
internal technical
factor factor affecting inefficiency
technical level with half-
that �representing internal affecting technical inefficien
distribution. (�� ~|�(0� �� )|). Error component � represent internal factors controlla
inefficiency level with half-normal distribution. (�� ~|�(0� ��� )|).� Error
distribution. Error component
component �� represent inte
farmers, and is the value reflecting farmer managerial capability with asymmetric distr
farmers, and is the value reflecting
represent internal factors controllable by farmers, and is the value reflecting farmer managerial capability w
with�� > 0. If production process running efficiently, its output will tends to be
farmer managerial capability with with� � > 0. Ifdistribution
asymmetric production with. process running efficiently, its out
If production
process running efficiently, its output will tends to be close maximum potential ui
= 0. Otherwise, ui > 0maximum
means it ispotential = 0.orOtherwise,
nowhereui near far from its ui maximum
> 0 meanspotential. it is nowhere near or f
maximum potential.
3. Technical Efficiency / TE Analysis
Technical efficiency 3).
canTechnical Efficiency
be measured / TE Analysis
using formula:
Technical efficiency can be measured using formula :
��� = exp(�� ��� |�� �) � �1� � � �
(2)

where TEi is the technical

where efficiency of farmer,
��� is the technical i, exp (–E[u
efficiency |εi]) is�,mean
of farmer
i exp(���� ui mean value of
value� |�of� �)is
based on known, εi hence 0 ≤ TEi ≤ 1.
known�� , hence� ≤ ��� ≤ 1.
Technical
Technical efficiency method efficiency
used method
on thisused on thisisresearch
research based on is based on technical ineffici
technical
inefficiency model developed by Battese and Coelli (1995) and Coelli (1996), ininwhich value of
developed by Battese and Coelli (1995) and Coelli (1996),
computed
which value of TEi will using software
be computed FRONTIER
using software 4.1d from
FRONTIER 4.1d Centre for Efficiency and P
from Centre
Analysis (CEPA)
for Efficiency and Productivity Analysisfounded
(CEPA) by founded
Coelli. As
by shown
Coelli. on As Aigner
shown on et al. (1977), Jon
(1982), Greene (1993)
Aigner et al. (1977), Jondrow et al. (1982), inCoelli
Greene (1996), �� will
(1993) in �Coelli
� give estimated
(1996), TEi will variance value in
� � � �
parameterization � = � � �
give estimated variance value in the form of�parameterization
� dan � = � ⁄
�σ =� � σ 2 + σ 2 dan γ = σ of
. This parameter 2 variance will t
2 v u u
+ σv2. This parametertooflook for� will
variance value,
thenandtherefore�
be able to look � �for� 1. Value
value, andoftherefore
parameter� is the cont
technical efficiency on its total residual effect.
0 < γ < 1. Value of parameteris the contribution of technical efficiency on its total
residual effect.
4). Economic and Allocation Efficiency Analysis
Measuring economic and allocation efficiency can be conducted by modifyin
function from homogenous Cobb-Douglas production function (Debertin, 1986). I
dual frontier cost function need to be modified from equation (1) to cost func
simplify exposition, notation i is eliminated from equation (1) that will be tran
below:
�
 ncy can be measured using formula
Technical : where ��
��efficiency method is theused technical
on this efficiency
researchofis farmer based on �, exp(���� |�� �)is meanmodel
technical �inefficiency value of
��� = exp(�� � |�� �) � �1� � � �� (2)
��� |�� �) � �1� � � � (2)
developed where by�� � is the� ,and
known�
Battese technical
hence� Coelli ≤efficiency
�� � ≤ 1.
(1995) of farmer
and Coelli �,(1996),
exp(���� � |�� �)isvalue
in which meanofvalue of ��be
��� will based
computed
where ��� is the technical known� using Technical
, hence�
software ≤ efficiency
�� ≤
FRONTIER 1. method 4.1d usedfrom on
� efficiency of� farmer �, exp(����� |�� �)is mean value of �� based on this
Centre research
for is based
Efficiency on
and technical
Productivity ineffici
technical efficiency known� of farmer
Analysis
, hence� �, exp(����
Technical
(CEPA)
≤ �� developed � |�� �)is
efficiency
founded bymean
method
byBattese value
Coelli. used
and of
Ason � based
�shown
Coellithis (1995)on
research
on Aigner and is basedet al.on
Coelli technical
(1996),
(1977), inefficiency
inJondrow
which value
et al. of m
� � ≤ 1.
� ≤ ��� ≤ 1. Technical (1982), developed
efficiency Greene computedby used
(1993)
method Battese using
inCoellion and
thissoftware
Coelli
(1996),
research FRONTIER
��(1995)will
� is based andon
give 4.1d
Coelli
estimated from
(1996),
technical Centre
variance in which
inefficiency for Efficiency
value model value
in the formof ��of and� wilP
ncy method used on this
developed research
computed
parameterization
by Battese is Analysis
based
and�using
�
=on�(CEPA)
Coelli technical
software
�
� �
� dan
�Economic
(1995) inefficiency
founded
FRONTIER
and �= ��by
Coelli
� �model
Coelli.
4.1d from
� . This
⁄�(1996), As shown
parameter Centre
in Small-Holder…
which ofonfor
value Aigner
Efficiency
variance
of et al.
will then (1977),
and ableJon
beProducti
Adoption of Technology � and Efficiency of the • �� � will be
4627
attese and Coelli (1995)
computed and
to using Coelli
Analysis
look for� software (1996),
(CEPA)
(1982),
value, in
Greene which
founded
andtherefore�
FRONTIER (1993)value
4.1d by of
Coelli.
inCoelli
�from ��
� � 1. will
As
(1996),
Centre
� Value be
shown �� on
will
forofEfficiency Aigner
give
� parameter� is et
estimatedthe contribution of inet
al. (1977),
and Productivity variance Jondrow
value
software FRONTIER Analysistechnical4.1d from
(1982),
(CEPA)efficiency Centre
Greene
founded by for
parameterization
on its Efficiency
(1993) � � and� Productivity
inCoelli
total residual
Coelli. =
As shown � (1996),
�
�effect.� �
dan
��
on� Aigner � will
� = � �⁄ �
give �estimated
et �al. �(1977), Jondrow of
. This variance
parameter value
et al. in thewill
variance form t
A) founded by (1982), 4. Economic
Coelli. As shownand Allocation
on Aigner Efficiency
et � Analysis
�al. (1977), �Jondrow et� ⁄ al. �
Greeneparameterization
(1993) to look (1996),
inCoelli for�
� =value, ����� � � dan
andtherefore�
� will give �= �� ����. variance
estimated This
� 1. parameter
Valuevalueofinofparameter� is
variance
the form of will the thencont be
1993) inCoelli parameterization
(1996),
Measuring �� will
4). �Economicto give
economic � estimated
look technical
and for�
� Allocation
and variance
� efficiency
value,
allocation value
on
andtherefore�
Efficiency
efficiency
� its
� in the
total
can �
Analysisform
residual
�
be
� = �� � �� dan � = �� ⁄�� . This parameter of variance will then be able � of
1.
conducted effect.
Value byof parameter� is
modifying dual the contribution
� � = ��� � ���todan cost
look ��� ⁄�value,
� function
=Measuring
for�
�
.from
This
�technical parameter
economic efficiency
homogenous
andtherefore� andofon variance
allocation � 1.will
its� total
�Cobb-Douglas then production
residual
efficiency
Value be
of able
effect.
can be conducted
parameter� is function the by modifying ofdual cost
(Debertin,
contribution
alue, andtherefore� �
1986). In
technical � �
function1.
this case,
efficiency Value
from of
4). parameter� is
Economic
homogenous
ondual frontier
its total residual and the
Cobb-Douglas
cost contribution
Allocation
function need
effect. Efficiency
production
to be of modified
Analysis
function (Debertin,
from equation 1986). In this case,
ncy on its total residual effect.
dual 4).
frontier EconomicMeasuring
cost and
function economic
Allocation
need to and
be
(1) to cost function C. To simplify exposition, notation i is eliminated from equation allocation
Efficiency
modified Analysis
fromefficiency
equation can(1) beto conducted
cost function by modifyin
C. To
4).(1)Economicsimplify
that will be Measuring function
exposition,
andtransformed economic
Allocation Efficiency from
notation homogenous
and
i
as below: Analysisis allocation
eliminated Cobb-Douglas
efficiency
from can
equation production
be
(1) conducted
that function
will by
be (Debertin,
modifying
transformed 1986).
dual
as
nd Allocation Efficiency
Measuring Analysis
below: function dual
from frontier
homogenous cost function
Cobb-Douglas
economic and allocation efficiency can be conducted by modifying dual cost need to be modified
production from
function equation
(Debertin, (1)
1986).to cost
In func
this c
omic and allocation efficiency dual can be
frontier
function from homogenous Cobb-Douglas �productionconducted
simplify cost by
exposition, � modifying
function notation
need to dual
bei is cost
eliminated
modified from from equation
equation
function (Debertin, 1986). In this case, (1) (1)
to that
cost will
functionbe tran
C.
mogenous Cobb-Douglas production
simplify function
below:
exposition,
dual frontier cost function need to be modified � (Debertin,
= � notation
� � 1986).
�
i
exp is (Infromthis
eliminated
(3) case,from(1)
equation equation
to cost(1) that will
function C. be Totransforme
�) (3)
� � C. To
st function need to
simplify be modified below: from equation
exposition, notation i is eliminated (1) to cost function
from equation
��� �� (1) that will be transformed as
ion, notation i below:is eliminated
Wherefrom : equation (1) that will be�� transformed = � � �� exp as (�)
��
� = � � � exp ( �)
Where: � = exp (�� ), dan � = � � � (4)
�
� ���
�
��
Where : ��
� = � � �� exp (�) (3)
���
� = � � �� exp (�) (3)
Cost Where : � =���
function exp (�� ), dan � = � � �
(4)

���Where : � = exp (�� ), dan � � = � � �
Cost
� =function:
exp (�� ), dan �Cost = ��function � �:� �� (5)
�=� (4)
dan � = � � � (4)
Cost function : ��� �
� � = � � �
Cost function : � �
Where�� is the input � price ��. = � �� �����
(5)
�
Form of dual � = frontier
� � � function ���cost is modified under cost minimization assumption bya
(5)
� �
� = � �� �� (5)
constraint� = Where� � . Firstly,
���� � is the input price
expansion path�.value had to be formulated using Lagrange function
��� Whereas Where�
is the inputForm
Pj follow: is the
� priceof input price �.
j. dual frontier function cost is modified under cost minimization assu
Where� is the Form
input of dual
constraint�
price �. frontier
� = ��function
. Firstly, cost expansionis� modifiedpath value under hadcost to be minimization
formulated using assumptionLagran
put price �. Form� of dual frontier function cost is modified��under cost minimization
Form of dual by constraint�
frontier as follow:
=
function � . Firstly,
cost expansion path value had to be formulated using Lagrange func
assumption
ontier functionconstraint�
cost is modified aunder
constraint � =��
cost Y �=� �isY
minimization
�� modified
. � �� � �
Firstly,
assumption
under
� �cost
expansion �
bya
minimization
exp path value had assumptionto be bya
(�)� (�)
=usingas follow:
�� . Firstly, expansion ��� pathas
0
value �
had to��� �
be formulated using��Lagrange function
. Firstly, expansion formulated
path value had Lagrange
to be formulated function using follow:
� � Lagrange function �
as follow: = � � � � � � �� � �
� � �� exp (�)�
��
�
To get�the value of �� , the first order� condition � = � � � �
����� � �� � � �
of L can be � exp (�)�
� calculated
��� as follow :
� ��
� = � �� �� � � �� � � � �� exp (�)� (�)
� ��� ��� �
� = � �� �� � � �� � � � �� � exp ( �)� (�) ��
(6) �� �
To���get the value of �� ,��� the first order condition=of��L�can ��be calculated
� �� � exp(�) as follow (7) :
��� ��� To get the value of �� , the first order condition of�� L �can be calculated �� as follow : �
�� ��� �
�� ��
To get Tothe getvalue
the valueof �� , of theXfirst,j theorder first condition
order condition of L canofbeL calculatedcan be calculatedas�� followas : = ��
follow: � � �� � � ��
of �� , the first order condition of L can be calculated as follow : �� � �� � � �
=� �� � � ��� ��� exp(�)
� �� ����� �� ��
�� �� � = � � �� � �� exp(�) (7) ���
= �� � �� � �� � exp(�) ���(7) � ��
��� �� (9) ��� 5
���

�
�� �� ��
(8)
= �� � �� � �� exp(�) (8)
��� �� 5
��� 5
�
�� �� �
= �� � �� � �� � exp(�) (9)
��� ��
���
�
�
� �� � exp (�)
���
 �
�� Ronnie S. Natawidjaja
4628 • Lies Sulistyowati, ��
= �� � �� � �� and Boby(8)
Rahmat
��
exp(�) �
��� �� �� ��
���
� � = �� � �� �
��� ��
�� �� � �� �� � �� ���
= �� � �� � �� � exp(�) = � (9) � �� � � exp(�) (8)
��� �� ����� = (9) �
�� � �� ��� � �
�� � ��
� �� exp(�) �� (8) ��
�

��� �� � � � � = �� � �� �
� �� � ���� �� ��� ��
�� �� = �� � ���� � � � � �� exp(�) (8) ���
� �� exp (�) � �� ��=
� �� � �� � ���� �� ���exp(�) (9)
�� �� � = � � � �� � � ���� � exp(�) (9)
� ��� ����� �
���
� = � � � � �� � exp ��(�)
��� ��� ��
�� �� ���
�� = �� � �� � �� exp(�) (9)
), (9) and (8), �� the = � �=
value ���
of� �
(expansion exp path) can also be
���
obtained by: ���
��(�) ��
�� � � � �� exp (�) ���
�� �� �� � � ��� � � �
= From
�� �equation
� ��� (7), �=�
(9) and (8), the value of Xj
From equation (7), (9) and (8), the value of�� (expansion
� � (expansion path) can also path)be can also be obta
�� obtained = �
by: �
� � � � exp ( �)
�
�� �� exp(�) �����
From equation
�
∏���(7), �
�� (9) exp(�)
and
�
���value �
∏���of� �
�� (expansion
exp(�) �� �� path) can also be �obtained � �� �� ��
From equation (7),��� (9)(8),
andthe(8), the � =value �of�� (expansion
�
path) = can also be obtained �
by:
by: = �
�� �� �� �� �� �� �������∏ ������
�
�� �� exp(�) �� ��� ��∏
�
�
�����
� �� � exp(�) ��� ∏���� �� � exp(�
�� = � ���= � (9)
� equation
= = = = =path) can also
�� �� From ��� ∏�� � ∏�
(7),
��
(9)
��and (8), the value
∏
of� (expansion
� ∏� �� � �� �� � �∏� ����
be obtained by:
� �
��� ���� exp(�)
�
�� ��� exp(�) �� � �� �
��� ���� �exp(�) exp(�)
��� ��� �� = �
�� ��∏
��
������ ����� exp(�)
� exp(�)

���
�� �� �= = =
� � � � �
�� = � �� �� �� ��
� � (10)
�� � � ��
�� �� � ��� ∏����� �� ��=
� = ���
exp(�) � ��� �
� ∏��� �� exp(�) � ��
�
� ∏��� �� exp(�)
�� �� (9)
� (9)
�
�
�� �� ���� � � �
�� = � � �
� �
�� �� �
����=��� �� ��� �� (9)
9) to (3), and (10) to (3). The results �� = ��would� = �� � ��be: � �� (10)
�� (10)
� �� �� �� ���������
�� �� �� = � Substitution of (9) to (3), and (10) to (3). The results would be:
� �� (10)
� = � � �� Substitution � �� � exp of
(�) (11)
(9) to (3), ��(3),
and �(10) to (3). The results would� be:
� �
Substitution
� � Substitution of
of (9) to (3), and (10) (9) to � and (10)
to (3). The�results to (3). The
would results
be: ��would
�� be:
��
��� �
� �� of (9) to � �� �� �� � � = � � ��� � � �� � exp (�)
��Substitution
�� � =(3), �� and �(10)
�� �� � to�(3).
�� �Theexp results would be:� �
(�) (11)
���
� = � � �� � �� � exp � �
� ������ exp �� (�) (11)
� =(�) (12)
� � �� �
�� �� � � �� ��
���
� ���� �� ��
��
��� � ���
= �� � �� � ��exp (�) (11)
� �������� �� � = � � �� � � � exp (�)
� = � ��� � � �
��
� �
�� �� �
� �� � exp (�) (12)
� ∗ �� �� ���
ion (11) and (12), demand input � =function
� � �� (� � ���be � presented
exp
�� (�) (12)
��
���
�� �) �
= � �����
can
� � as follow:
� � � exp (�) (12)
���
Next,��from
� �� �equation (11) and (12), demand input function (� ∗ ) can be presente
Next, from equation (11) ��� (12), demand input function (��∗ ) can be presented as follow: �
and
� Next, from�equation (11) and (12), demand input∗ function (Xj*) can be � presented
Next, from�equation (11) and (12), demand input function
�
(� � ) can be presented as� follow:
���� )� ∏������� �as��follow: � �∗ ) can be presented as follow:
� �
Next, from equation (11) and�(12),∗ demand input
� � ∗exp (�) � function (�
�� = � �
�� = � � �� ��
�
�� � ∏
�(�� �� ) ����� � � �� �� �(� � ���)� ∏� �� � �� � exp (�)
∗
� �� = � ∗
�
� � �� � exp (� �)
� � ��� � �
���(�= ���� )� ∏� �� ��� ��� exp (�) � � �
� � �� �����
� �(� ) � ∏�� ��� � �� � � exp
�
( �) � �
�
���� )� ∏������� ���� � ∗exp (�) � � ��� � � �
∗ � �
� =� �� =�� � �
� �
� � �� � �� ��
1∗ = � ∗�� �� �(� �
� �1
�� � ∏
) � �
��
���� � � � � exp
�(� ( �)
�
� �
��� � ∏�
) ��
��� � � � � exp ( �)
��∗ = �� ���� �� � ���� = �exp ��
�� � �� �� � �
�
�� (13)
� �
�� � )� �∏
�(�� ��(� � �)� ∏
�� ������
�� �����1���exp � (��)
� �exp � ��(�) 1 1 1
�� ∗���
= � � ��� �� exp �� ∗ � ��
�� (13)
��� � � �� � � �� =� �� � ���
� �� � exp �� ��
1 1 ��� �
� �� �� �� 1 1 �� �
���
with ��∗ =���∗� �=���
�� ���� �� �� � �� exp �� exp ���� (13)
�� (13)
� �with
� �� � �
��� ��� �
with � = � ��
with with
� � � = � ��
���
���� and�
�= �=
��� �� ���
�
��� �
and
����
���
� �� �� � and �
����
and �� = �� �� � �� �� �����
���
�
�
���
���� �� = �� �� � �� �� �
��
� � = �� �� ��� � �
� ���
ier cost function = ��be
�� will ��presented
� �� �
���by
Lastly, dual frontier
substituting equation (13) to (5):
cost function will be presented by substituting equation (13) to (5):
 ��∗ = � ��
� �
�(�� ���� )� ∏������� ���� � exp (�) �
�
��∗ = � �
�
�� ��
�
�� � ∏�
�(�� �� �) ������ �� � exp (�)
��∗ = � �� �
�
�(�� ���� )� ∏������� ���� � exp
1 1 (�) �� ��
��∗ = ������� � �� exp �� ��
Adoption of Technology and Economic∗ Efficiency 1of the Small-Holder…
�����
� 1 • 4629�
�� = �� ���� � �� exp ��
��� ��
with �� �
���
�
with
� � = � ��
� = � �� ���
and
���
and and ����
�
����
�
�� = �� �� � �� �� �
�� = �� �� � �� �� �
���
���

Lastly, dual frontier cost Lastly, will

function dual frontier cost function will be presented by substituting eq
Lastly, dual frontier be
costpresented bybe
function will substituting
presented byequation
substituting equation (13) t
(13) to (5):
�

� � � ��∗ ��
∗
�
�
� ∗ ∗
���
� � � �� �
� ∗ � � ��∗ ��
� �
∗ ∗ �
� � � �� ��
� �� 1
�∗ � �∗� � �� ���
� �� � ��� �� ��� �� (14)
���
��∗ ��
��� � � �
1
� ���
���
� ∗
� �� ���
� �
�� ��
��� �� �� �� 1
�� (14
� �� 1
with �
∗
� � ��
� ���
��
� � ��� �� ��
� ∗ � ����� � ��with �
��� �� �� (14) ����
� �� 1 ��� �
� �
�∗ � ����� �
� �� � ��� �� �� (14) ���
��� with �� �
� � � �� � � �� � ���� �� � with �
����

�
���� with ��� ���
� � � �� � �
����
��� ���� � � � � �� �
�� � �� �� � �
Equation (14) will achieve�minimal
�
��� � � � � � � �� � �
value on ��� its
��
�
� lower limit so its � minimal dual frontier cost
��� � � � � � �
function can be calculated as follow:
� �� � � �
�
�
��� ���
Equation ��� (14) will achieve Equation
��� (14) will value
minimal achieveon minimal
its lower valuelimit on its so lower � limit so its minimal dual frontier co
its minimal dual 1
achieve minimal value on its lower limit so its function dual
minimal can be calculated
frontier cost as follow: � ∗ � ����� � ����� ��� �� ��
frontier cost function can be calculated as follow: Equation (14) will achieve minimal value on its lower limit so its m � � �
culated as follow: Equation (14) will achieve minimal value on its lower limit so its minimal ��� dual frontier cost � �� 1
function can be calculated as follow: � � function can be calculated as follow:
1
� ∗
� �� ���
� �� � ��� �� �
�
∗ )��� �� �� �� ��
�∗ � � ∗�� � ��� �(�� �� � � ���
� �� �
(15)
�� 1 ���
� � ∗ ��� �� ��
�� �� � �� � �� ��� �� �� ∗ �
��� ���
� � � ��
� whereas � ∗ � �(� ∗ )��� � �� � (15
���
� ��
� ∗ � � ��� (��)
�(� ∗ )��� � �� � (15) � ���
�
�� ��
∗
� � �(� whereas
∗ )���
� �� (15)
�� ��
��� ∗ ∗ ���
Bravo-Ureta and Rieger ∗
� (1991)����
� ��� ∗ � � �(� )
(��) farmer production output that� was already set to its
defined� ��
whereas statistical disturbance, as v can be obtained by using equation (4)��� with value u from equation
)
whereas� ∗(2) � and� ��� � from
(��)equation Bravo-Ureta
(3). whereas and Rieger (1991) defined� ∗ farmer production output that was already set to i
As studied on Jondrow et al. disturbance,
statistical (1982) ∗ on Ogundari
as v canand Ojo (2006),
be obtained economic
by using equationefficiency
(4) with (EE) value u from equatio
eger (1991) defined� ∗ farmer was production output that of
defined as the ratio
was already
observed �equation
�set� ���
minimal (��) total cost (C*) with actual production
to itsproduction
Bravo-Ureta and Rieger (2) and
(1991) � from ∗ (3).
ce, as v can be obtained by using
totalequation
cost (C) as (4)follow:
with valuedefined�
u from equation farmer production output that was already set to its
As studied on Jondrow et al. (1982) on Ogundari and Ojo (2006), economic efficiency (EE
statistical disturbance,
Bravo-Ureta and Rieger as v can (1991)be obtained
defined by Y* using equation
farmer (4) with value
production outputu from
∗ that equation
tion (3). was defined Bravo-Ureta
as the ratio of and Rieger
observed (1991)
minimal defined�
production total cost (Cwas
farmer production
*
) with actual output productio th
row et al. (1982) on (2)
Ogundari and and� from
already set to its statistical Ojo equation
(2006), (3).
economic
disturbance,
� ∗ efficiency as (EE)
v can be obtained by using equation (4)
atio of observed minimalAs studied on
production totalJondrow
cost (C
total
�� * �
et) al.
cost
with (1982)statistical
(C) as follow: disturbance, as v
on Ogundari and Ojo (2006), economic efficiency (EE)can be obtained by using
(16) equation (4) wit
with value u from equation � actual
(2) and εproduction
(2) minimal
and from � from equationequation (3).(3).
ow: was defined as the ratio of observed production
�∗ total cost (C*) with actual production
As total cost (C)on
studied as follow:
Jondrow etbeal. As studied
(1982) � �on�on Jondrow etand al. (1982) on Ogundari and Ojo (2006), ec
�Ogundari Ojo (2006), economic
where ��� ����� �� ��will between �� �� (16
� 1.
�∗ was defined
efficiency (EE) was defined∗ as the ratio of observed minimal production total cost as the ratio of observed minimal production total cost (C*
�� � (16)
�is the combined
� (C*) with Economic actual production efficiency total
�� �where ��� ����� �� ��will
total costcost (C)value (C)of
as asbe
follow: follow:
technical
between and � allocation
� �� � 1.efficiency. Allocation
(16)
efficiency itself can be�obtained by using the following formulation:
��will be between � � �� � 1. Economic efficiency is the combined value � ∗ of technical and allocation efficiency. Allocatio
where ��� ����� �� ��will be��
efficiency between
itself � �be
can ��obtained
� 1. �� by �
using
the following formulation:
�� � (17)
�
y is the combined value of technical and allocation efficiency. �� Allocation
be obtained by using theEconomic following efficiency
formulation: is the combined value of technical �� and allocation efficiency. Allocation
where ��will
efficiency itself canhave its value between�
be obtained by using �� � ��
the �
where ��� ����� �� ��will � 1.
following
��
(17
formulation: be between � � �� � 1.
��
�� � (17)
5). Factors affecting�� Technical Inefficiency
 ∗
Bravo-Ureta and Rieger (1991) defined� ∗ farmer production output that was already se
Bravo-Ureta and
Bravo-Ureta Rieger
and (1991)
Rieger defined�
(1991)
statistical disturbance, as v can be obtained farmer
defined� ∗production output that was already set to its
farmer
by using production
equationoutput (4) with thatvalue
was ualreadyfrom eqse
statistical disturbance,
statistical as v
disturbance,
(2) and � from equation (3).can be
as v obtained
can be by
obtainedusing by equation
using (4)
equation with value
(4) with u from
value equation
u from eq
(2) and As
�(2)from
andequation
� from
studied (3). et(3).
equation
on Jondrow al. (1982) on Ogundari and Ojo (2006), economic efficiency
As studiedAs on Jondrow
studied on et al.
Jondrow (1982)
et al. on(1982)Ogundari and
on Ogundari Ojoand (2006),
Ojo economic
total(2006), * efficiency
economic (EE)
efficienc
was defined as the ratio of observed minimal production *cost (C *) with actual prod
was defined as the
was defined
total cost (C)ratio
as of ratio
asthe observed
follow: minimalminimal
of observed production total costtotal
production ) with
(C cost (Cactual
) withproduction
actual prod
4630 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat
total costtotal
(C)cost
as follow:
(C) as follow:
∗ �∗
��� � � ∗
�� � �� � �
(16)
(16)
� �
where ��� ����� �� ��will be between � � �� � 1.
where thewhere ��� ����� �� ��will
of EE will be between
valuewhere ��� ����� �� ��will be between0 ≤beEE �≤ �
between1. �� ���1.�� � 1.
Economicisefficiency
Economic efficiency the combined is thevaluecombined value of
of technical and technical
allocation andefficiency.
allocation efficiency. Allo
Economic efficiency
Economic is the be
efficiency combined
isobtained
the combinedvalueusing of technical and allocation efficiency. Allocation
efficiency
Allocation efficiency itselfitself
can can
be obtained by by usingvalue thethe of technical
following
following and allocation
formulation:
formulation: efficiency. Allo
efficiency itself canitself
efficiency be obtained
can be obtainedby usingbytheusing following formulation:
the following formulation:
��
�� (17)
�� � ��
� ��
�� � �� (17)
�� ��
where AE will have its valuehave
where ��will betweenits value0 ≤ AE ≤ 1. � �� � 1.
between�
where ��will have its have
where ��will valueitsbetween�
value between� � �� ��1.�� � 1.
5. Factors affecting Technical Inefficiency
5). Factors affecting Technical Inefficiency
To obtain5).value
Factors
5).
To affecting
Factors
ofobtain
distribution
value Technical
affecting
ofparameter Inefficiency
Technical
distribution Inefficiency
ofparameter
technical ofinefficiency, this research
technical inefficiency, this research will u
will use To
theobtain To
followingvalue
obtainof distribution
value
equation:
following equation: of parameter
distribution of
parameter technicalof inefficiency,
technical this research
inefficiency, will usewill
this research the
following equation:
following equation: �
� �
�� � �� � � �� ��� � ��
�� � �� ��� � ��� ��� ���� �� � �� (18)
�� � ��
��� � (18)
Where: ��� ���
Where: Where:
Where:
Z1 = dummy value of the tree age (0 = not productive age; 1 = productive age (20- 7
40 y. o.))
Z2 = dummy value of fruit caring (0 = not conducted; 1 = conducted)
Z3 = farmerage (year)
Z4 = farmer education (year)
Z5 = farming experience (year)
Z6 = membership on farmers group (0 = non-member; 1 = member)
Z7 = credit access (score)
Z9 = information access (score)
For consistency purposes, the estimation of inefficiency function parameter
will be simultaneously arranged with equation (1) using software FRONTIER 4.1d
(Coelli, 1996).

RESULTS AND DISCUSSION

Mango Production Centers in Indonesia

The BPS data on 2012 indicated that mango production contributed 1.627.997 tons
or 10.07 per cent from the total Indonesian fruit production, ranked as the third
ܼ଺ = membership on farmers group (0 = non-member; 1 = member)
ܼ଻ = credit access (score)
଼ܼ = information access (score)

For consistency purposes, the estimation of inefficiency function parameter will be

simultaneously arranged with equation (1) using software FRONTIER 4.1d (Coelli, 1996).
Adoption of Technology and Economic Efficiency of the Small-Holder… • 4631
RESULTS AND DISCUSSION
largest national fruit after banana and orange (BPS, 2012). Therefore, Indonesia
is a potential
Mango Production country
Centers toin
develop mango crop, and furthermore to be leading as
Indonesia
mango producer
TheBPS country
data on 2012 either domestically
indicated that mangoor for export.
production This optimism
contributed couldtons
1.627.997 be or
realized
10.07 if itfrom
per cent is supported by sufficient
the total Indonesian fruitland resources,
production, stable
ranked asweather
the thirdand climate,
largest national
fruithuman resources
after banana andand technology.
orange (BPS, 2012). Therefore, Indonesia is a potential country to
developThere
mango arecrop,
3 mainand furthermore
provinces to beproducers
of mango leading asin mango producer
Indonesia, that iscountry either
West Java,
domestically or for export. This optimism could be realized if it is supported by sufficient
Central Java, and East Java, with its contribution to national mango production are
land resources, stable weather and climate, human resources and technology.
21.1, 19.6, and 35.5 per cent respectively.
There are 3 main provinces of mango producersin Indonesia, that is West Java,
Central Java, and East Java, with its contribution to national mango productionare 21.1, 19.6,
Figure
and 35.5 per cent2:respectively.
Comparation of mango production in West, Central, and East Java

1,000,000

800,000

600,000 West Java

400,000 Central Java

East Java
200,000

-
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Source : BPS
Source: BPS
Figure 2 . Comparation of mango production in West, Central, and East Java

Performance
Performance of Mango
of mango Farming
farming in Indonesia
inIndonesia
Mangoes as natural crop mostly grow in
Mangoes as natural crop mostly grow in the
the forest.
forest. Since
Since the
the sweet
sweetfruits
fruitsof
ofmango
mangohad
been
hadfond by the
been fond people
by theas people
local consumers, it hadmotivated
as local consumers, the motivated
it had farmers to grow mango on
the farmers
their home yard, even without any cultivation technology treatment.
to grow mango on their home yard, even without any cultivation technology Later on, by the
significant
treatment. Later on, by the significant increase of demand for mango along with the of
increase of demand for mango along with the demand of global market, some
mango farmers started to apply a certain kind of cultivation technology to improve its
demand of global market, some of mango farmers started to apply a certain kind
production quantity as well as its quality. And now more and more mango farmers had
of cultivation technology to improve its production quantity as well as its quality.
becoming consumers oriented.
And now more and more mango farmers had becoming consumers oriented.
There is a challenge for local small farmers of mango due to market competition
There is
againsthigher a challenge
quality for local
of mangoes small
that can farmers of mango
be imported due toscale.
in a larger market
Thecompetition
challenge not
against higher quality of mangoes that can be imported in a larger scale. The
challenge not only for local farmers to implement necessary technology, but also
to change their perception to see how new chances can be acquired from mango
8
trading.
In Indonesia, the most well-known variety of mango are Arumanis and
Gedong Gincu. Both are very popular not only in a domestic market but also on
4632 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

the overseas global market. Arumanis mangoes are produced mostly in East Java,
while Gedong Gincu are produced in West Java. Besides of these two, however,
there are still a lot of other varieties such as Dermayu, Golek, Gadung, Podang,
Manalagi, and others.
Numbers of productive mango trees cultivated by the farmers will become a
factor that might determine how much quantity of production and earning the
farmers obtained. On the average, there were 43 trees per farm owner in East Java
out of 720 mango tree owners sampled. For comparison, mango farmers in West
Java cultivated only 23 trees per farmer out of the total 300 mango tree farmers
sampled. It can be concluded that scale of business mango farmers in East Java
were larger compared to the farmers in West Java. Sulistyowati et al. (2013) found
that 94.97 per cent of mango farmers were self-cultivating their own mangoes,
while only 5.03 per cent sharing through rental arrangement.

Table 3
Total Productive Mango Trees

Description Total East Java West Java

(tree) (tree) (tree)

N 633 319 314

Average 32.75 42.49 22.88

Std. Dev. 63.87 83.07 32.27

Min 0 0 0

Max 720 720 300

Adoption of Off-season Technology by the Mango Farmers

ZPT (plant regulator)on crops is anon-nutrient organic compound that in
appropriate dose is able to promote, inhibit, and change the physiological process
of crops, so that mangoes can grow and produce fruits out of its natural harvest
season. As shown on Table 4, there were more mango farmers in West Java that had
used off-season technology (23.42 per cent) compared to those in East Java (12.50
per cent). Off-season technology had been used by the farmers in both provinces
intended to boost mango production out of its natural harvest season. In West Java
mango farmers indicated their ability to learn the technology faster than those in
East Java. Even though, off-season technology implementation in both provinces
was still relatively low, no more than 17.92 per cent.
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4633

Table 4
Percentage of Off-season Technology Implementation

Total Jatim West Java

Y7
Freq. % Freq. % Freq. %
0 519 82.08 279 87.5 240 76.58
1 114 17.92 40 12.5 74 23.42
Total 633 100 319 100 314 100

Note: 1 = using off-season technology,

0 = not using off-season technology

Farmers are free to decide whether or not they will adopt new idea or knowledge
about mango farming. The farmers should be able to pass by several milestones in
adopting new technology, (Mardikanto, 2009), consisted of the awareness, interest,
assessment, evaluation and trial upo na new necessary method. Adoption of new
technology will also be greatly influenced by the type of innovation, its media,
current socio-culture characteristic, and eagerness of the of innovation agents them
self (Hubeis, 2007). With respect to innovation, as a comparison, Gundari and Ojo
(2007) concluded that in Nigeria, innovation technology remains very important
factor to increase productivity and revenue of sweet potato farmers, besides it also
had contributed to economic growth of the country.

Factors Affecting Off-season Technology Adaptation

Dixon (2007) mentioned that technology adaptation is caused by the changes in
demand and growth. Meanwhile, Induced Technological Innovation theory stated
that relative change on production factor prices pushed innovation directed to a
cheaper production factor used. (Hayami and Ruttan, 1985)
Farmers motivation for technology adaptation itself is affected by: internal
factors such as farm scale and information access and external factors such as
infrastructure and innovation type related to complexity of the technology itself.
(Syamsyiah 2009) Masianus (2011) stated that other internal factors may include
formal and non-formal education, as well as years of experience, age, number of
relatives, and participation on the farmers group. As for external factors, it may
also include capital access, infrastructure availability, counselling intensity, and
market accessibility. On the other hand, Andi and Afrizaon (2011) mentioned that
age, education level, farm scale, and income level had insignificant effects.
Structural equation model used to analyze off season technology imple-
mentation on mangoes (Yos) was significantly affected by numbers of trees (Jph),
4634 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

information access (Info), market access (Pas), capital access (Mod), tooling facility
(Sal), and numbers of related training (Pel)were positively correlated, while tree
management (Pgl)was negatively correlated (Table 5).Structural equation model
then can be formulated as follow:
Yos = 0.24613753 - 0.4697997 Sph + 0.0011022 Jph + 0.029477 Info + 0.0385959 Pas +
0.0227872 Mod+ 0.0201667X10

Table 5
Analysis results of factors affecting off-season technology implementation (Yos)

Yos Coef. Std. Err. Z P>|z|

Sphi -0.4697997 0.0925911 -5.07 0*
Ui -0.0001334 0.0011121 -0.12 0.904
Pdi -0.001567 0.0038772 -0.4 0.686
Pgli 0.0017543 0.0015362 1.14 0.253
Aki -0.0108483 0.0083342 -1.3 0.193
Jphi 0.0011022 0.0004354 2.53 0.011*
Infoi 0.029477 0.0120924 2.44 0.015*
Pasi 0.0385959 0.0181764 2.12 0.034*
Modi 0.0227872 0.0064606 3.53 0*
Sari 0.0027223 0.0159372 0.17 0.864
Sali 0.0003994 0.0062515 0.06 0.949
Peli 0.0201667 0.0102757 1.96 0.05*
Cons 0.2461375 0.149826 1.64 0.100
Note: Yos= 1, uses off-season technology
Yos= 0, not using off-season technology

The results indicated that tree ownership status, number of trees, information,
market, capital accessibilities, and number of training affected farmers decision
on the implementation off-season technology. The farmers developed their mango
trees by themselves and tend not to implement off-season technology, however,
farmers who had more than 10 mango trees tend to do otherwise.
The farmers with high market accessibility were preferred to implement
off-season technology since they are capable of selling their own products, and
sometimes directly to end customers. The decision of mango farmers to implement
this technology was really motivated by a high selling price of their products.
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4635

Capital is one of the most important factors in farming business, either it is to

invest in farming tools or even production resources. Sufficient capital through
sponsorship can effectively help the farmers to reach optimal business scale,
adoption of technology, and post-production management. The farmers with high
capital accessibility will be able to implement off-season technology and to purchase
expensive growth-enhancing compound to boost their mango production.
The training for mango farmers included training about agriculture, harvesting
methods, post-production and marketing, and also innovation including off-season
technology. Through number of trainings the farmers would receive additional
knowledge of how to use growth enhancing compound such as phyto-hormone
that made mango produce its fruit not in her natural season. The analysis indicated
that the farmers who get training were more likely to implement off-season
technology. This is in line with Daadi et. Al. (2014) observing mango farmers in
northern Ghana with conclusion that alternative actions, such as specific training
was needed to help mango farmers to overcome a decline both in socio-culture
and effective mango farming.

Technical, Allocation, and Economic Efficiency of Mango Farming in Indonesia

Parameter estimation on stochastic frontier production function is the production
elasticity on the used of production factors. Table 6 shows that there are 5 input
variables that can be used to increase mango production : number of trees, the
use of pesticide, additional labor force, use of fly-trap, and off-season technology
implementation. Two variables had significant effects i.e. the additional labor force
and the use of pesticide, as also suggested by Mar et. al. (2013). From all of the five
inputs, number of mango trees is also a significant variable with ɑ= 5% either for
Indonesia as a whole or for the two provinces of East and West Java, with score
0.7810, 0.8113, and 0.7755 respectively. This production elasticity coefficient means
that by the increase of the mango tree number would result in the increasing of
mango production, while other inputs kept constant. This variable is the most
responsive compared to other variable as it is also suggested by Nadapdap (2014).
However, Daadi et. al. (2014) had shown that productivity were largely affected
by the type of mango care management. Group out-growers turns to have a lower
productivity compared with family out-growers, due to lack of commitment.
For East Java mango farmers the variable of pesticide had also a significant
effect with production elasticity coefficient of 0.0714. This means that increase in
production can still be realized by increasing the number of trees and pesticides.
Meanwhile for the farmers in West Java, there were off-season technology and
additional labor variables that had a large effect beside number of trees, with each
of the coefficient is 0.0842 and 0.0521 respectively. It means that mango production
in West Java can be significantly improved by the application of off-season
technology and additional labors.
 Table 6
Parameter Estimation of Stochastic Production Frontier (SPF) with Maximum Likelihood Estimation Method

East Java West Java Indonesia

Coef. Std. Error P>|z| Coef. Std. Error P>|z| Coef. Std. Error P>|z|
Production Function
–cons 2.585868** 0.285655 0.000 2.879765** 0.185521 0.000 2.8486** 0.15185 0.000
No. of Trees (x1) 0.811319** 0.057555 0.000 0.775535** 0.071808 0.000 0.7810** 0.042709 0.000
Use of organic fertilizer 0.008561 0.013418 0.523 -0.00042 0.01265 0.974 0.004183 0.009198 0.649
(x2)
Use of inorganic 0.0207 0.017527 0.238 -0.01203 0.01691 0.477 0.007109 0.012025 0.554
fertilizer (x3)
Pesticide (x4) 0.071383** 0.02956 0.016 0.047999 0.03254 0.104 0.0662** 0.021661 0.002
Additional labour (x5) -0.00336 0.016674 0.840 0.052149** 0.01983 0.009 0.019369* 0.01256 0.0123
Flytrap (x6) 0.061376 0.079304 0.439 0.098998 0.060297 0.101 0.092437* 0.047636 0.052
Off-season technology / 0.02238 0.035442 0.528 0.084167** 0.029366 0.004 0.0596** 0.021794 0.006
ZPT(x7)

* sig. at 0.10 level; ** sig. at 0.05 level

4636 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4637

Analysis of Technical Efficiency / TE

The results on Table 7 shows that average value of technical efficiency of
Indonesian mango farmers is 70.37 with a wide range of 26.93 and 90.29. This
indicates that technical efficiency of mango farmers are quite varied due to the
difference of understanding and application of technology. This findings are in
line with Ogundari and Ojo (2006), Prayoga (2010), and Fatwiwati et. al. (2013) on
the other case of commodities. Table 5 shows that technology implementation is
affected by : status of trees ownership, number of trees, information accessibility,
market accessibility, capital accessibility, and number of trainings attended.

Table 7
Frequency Distribution of Technical Efficiency (TE) on Mango Farmers at in East Java,
West Java, and Indonesia

East Java West Java Indonesia

Efficiency Level
No. of Farm- No. of Farm- No. of
(%) % % %
ers ers Farmers
<=40 0 0.00 13 4.14 10 1.58
40,01-50 3 0.94 15 4.78 37 5.85
50,01-60 13 4.08 47 14.97 66 10.43
60,01-70 54 16.93 68 21.66 145 22.91
70,01-80 157 49.22 101 32.17 242 38.23
80,01-90 92 28.84 69 21.97 131 20.70
90,01-100 0 0.00 1 0.32 2 0.32
Total 319 100.00 314 100.00 633 100.00
Maximum (%) 89.35 91.17 90.29
Minimum (%) 44.86 20.48 26.93
Average (%) 75.08 68.73 70.37

The results also shows that 59.25 per cent of mango farmers reach more than
0.70 technically efficient, while the rest of 40.75 per cent remain on technical
inefficiency condition. This means there are still chances for further increase on
technical efficiency of mango farming through the improvement on hercultivation
management. This condition had also previously stated by Mar, Sein et. al. (2013)
that analyzes on technical efficiency of mango farmers on Central Myanmar i.e.
71.01.
4638 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

Comparatively, the farmers in East Java had a higher technical efficiency than
the farmers in West Java, i.e. 75.08 and 68.73 respectively. This condition was in
line with the secondary data that East Java was the highest contributor of mango
production. Analysis for each province had shown that 78.06 per cent of the
farmers in East Java had a technical efficiency level above 0.70, while the rest of
21.94 per cent still in-efficient. Meanwhile, in West Java the farmers had slightly
lower condition,54.45 per cent of their efficiency level were above 0.70, and the rest
of 45.54 per cent still in-efficient. However, both lead to the similar conclusion, that
in these two provinces there were opportunities for the improvement of technical
efficiency to increase their mango production.

Analysis Results of Economic and Allocation Efficiency

Economic and allocation efficiency were calculated with dual cost frontier function,
modified from stochastic frontier function, as presented on Table 8.
Alocation Efficiency (AE) shows a minimum price to produce certain output
on the isoquant (Coelli et al. 2005, Debertin, 1986, Bravo, Pinheiro 1997). Mango
farmers in Indonesia had an average value of 15.05 with the range vary between
0.05 up to 81.46. By comparing between provinces, the result had shown a similar
condition to technical efficiency analysis that the farmers in East Java had a higher
allocation efficiency than in West Java, 20.09 in average with range 0.20 ~ 81.33
to 12.84 in average with range 0.04 ~ 76.71. It shows that allocation efficiency of
mango farming in Indonesia is still considerably low. This condition is caused by
lack of accessibility to technology information. The farmers were more likely to
use inorganic fertilizer and excessive labor. Both will increase cost of production
inputs, leading to allocation inefficiency.
Most important factor for farmers welfare is the economic efficiency (EE).
Economic efficiency is the cross-product of technical efficiency and allocation
efficiency with the value between 0 to 1. Score of 1 on economic efficiency means
the mango farming had already reach optimal economic efficiency, while the score
below 1 means economically it still inefficient. (Coelli et al. 1998, Debertin, 1986,
Bravo, Pinheiro 1997)
Table 8 shows overall value of economic efficiency (EE) for Indonesia mango
farmers by the average of 11.09 per cent and the range is between 0.03 to 67.13 per
cent. It is also shown that the farmers were able to save up to 83.5 per cent their
cost if they were able to achieve maximum economic efficiency. As a comparation,
the farmers in East Java are economically better than the farmers in West Java.
Economical efficiency for East Java mango farmers averages in 15.44 per cent with
the range between 0.16 to 67.36 per cent. It was shown that on the average, East
Java farmers would able to save up to 77.10 per cent(1-15,44/67,36) if they were able
 Table 8
Allocation Efficiency (AE) and Economic Efficiency (EE) Frequency Distribution of Mango Farmers in East Java, West
Java, and Indonesia

East Java West Java Indonesia

Economic Efficiency Allocation Efficiency Economic Efficiency
Allocation Efficiency Allocation Efficiency Economic Efficiency
Efficiency (AE) (AE) (EE)
Level (%) (EE) (AE) (EE)

No. of Farm- No. of No. of Farm- No. of Farm- No. of Farm- No. of
% % % % % %
ers Farmers ers ers ers Farmers

<=40 265 83.07 290 90.91 286 91.08 301 95.86 567 89.57 598 94.47
40,01-50 19 5.96 11 3.45 12 3.82 5 1.59 27 4.27 19 3.00
50,01-60 15 4.70 14 4.39 7 2.23 7 2.23 17 2.69 11 1.74
60,01-70 13 4.08 4 1.25 6 1.91 1 0.32 17 2.69 5 0.79
70,01-80 5 1.57 0 0.00 3 0.96 0 0.00 3 0.47 0 0.00
80,01-90 2 0.63 0 0.00 0 0.00 0 0.00 2 0.32 0 0.00
90,01-100 0 0.00 0 0.00 0 0.00 0 0.00 0 0.00 0 0.00
Total 319 100.0 319 100.0 314 100.00 314 100.00 633 100.00 633 100.00
Maxi-
81.33 67.36 76.71 67.02 81.46 67.13
mum (%)
Mini-
0.20 0.16 0.04 0.02 0.05 0.03
mum (%)
Average
20.09 15.44 12.84 9.36 15.05 11.09
(%)
Adoption of Technology and Economic Efficiency of the Small-Holder… • 4639
4640 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

to achieve maximum economic efficiency. Meanwhile, economic efficiency (EE)

of mango farmers in West Java on the average was 9.36 per cent with the range
between 0.02 to 67.02 per cent. This was shown that West Java farmers would able
to save up to 86.10 per cent (1-9,36/67,06) if they were able to achieve maximum
economic efficiency. From these results, it can be valued that Indonesian mango
farming were still had a low economic efficiency level, hence there were wide
room for improvements. The efforts that can be conducted to increase economic
efficiency may include: to restrict the use of fertilizer in West Java, both of organic
and inorganic; and to recalculate additional labor used in East Java.

Factors Affecting Technical Inefficiency of Indonesian Mango Farming

The factors that had affect technical inefficiency can be observed from the value
of σ2andγfrom the stochastic frontier model, and all significant atɑ value of 0.05
(Table 9). Value of γ= 0.5841 means that 58 per cent variance of farmers production
quantity is caused by the difference of technical differences,and 42 per cent
caused by external variable such as rainfall, pests and crops disease, soil fertility,
mango variety, and model error. From the results of analysis, it can be concluded
that significant variables affecting technical inefficiency are : experiences, credit
access, and information access. While other variables such as number of trees,
fruit treatment, farmer age, education, and membership in farmer group have no
significant effect. While the insignificancies of age and education is similar with
the study of Daadi et. al. (2014) results, and on the other hand contradicted with
the study of Msuya et. al.(2008) and Fadwiwati et. al. (2014) with their results on
corns farming.
Table 9 shows that experience affects upon technical inefficiency on level of
a = 0.15, specifically scored -0.0215 for Indonesia and -0.0259 for West Java. This
means that the longer the farmers have experiences on mango farming, the more
inefficiency can be reduced. This finding is not the same as what Daadi et. al. (2014)
found where the farmers learn about farming knowledge from their parents right
since childhood. Furthermore it is not so often Indonesia farmers get formal lessons
about mango farming.
Credit access variable had also affect significantly on inefficiency on level of
a = 0.15, with negative estimated parameter (-0.0248) for Indonesia and (-0.0526)
for West Java. This explains that the higher credit access available for the farmers,
the lower inefficiency they will have. Since credit access, as also shown by Mosher
(1965) is the accelerating factor for agriculture development, it also important
factors for the farmers. It can also theoretically increase efficiency. World Bank
(2007) had stated that farmers access to formal credit market remains one of the
highest priority for developing countries.
 Table 9:
Parameter Estimation Results for Production Technical Inefficiency Effect on East Java, West Java, and
Indonesia
Variable East Java West Java Indonesia
Coef. Std. Err t-ratio Coef. Std. Err t-ratio Coef. Std. Err. t-ratio
_cons 0.5012 0.4031 1.2434 0.9165* 0.5764 1.5900 0.7882*** 0.3256 2.4211
Dummy age of 0.1473 0.1904 0.7735 0.1115 0.2566 0.4346 0.1519 0.1727 0.8798
trees(z1)
Dummy fruit 0.0456 0.1156 0.3941 -0.0117 0.1384 -0.0844 0.0210 0.0896 0.2340
treatment (z2)
Farmer age (z3) 0.0014 0.0042 0.3314 0.0089 0.0075 1.1830 0.0032 0.0037 0.8472
Farmer education -0.0049 0.0179 -0.2726 -0.0019 0.0166 -0.1147 -0.0026 0.0120 -0.2206
(z4)
Experiences (z5) -0.0153 0.0138 -1.1103 -0.0259* 0.0205 -1.4618 -0.0215* 0.0144 -1.4935
Groupmember 0.2359 0.2457 0.9601 -0.0833 0.2306 -0.3610 0.0922 0.1216 0.7582
-ship (z6)
Credit access (z7) -0.0033 0.0217 -0.1530 -0.0526* 0.0339 -1.5512 -0.0248* 0.0191 -1.4967
Information -0.1167* 0.1202 -1.9708 -0.0217 0.0508 -0.4261 -0.0685* 0.0588 -1.4650
access(z8)
Sigma-square 0.2509*** 0.0781 3.2129 0.3102*** 0.1160 2.6737 0.3073*** 0.0878 3.4999
Gamma 0.3131 0.4048 0.7735 0.6877 0.1221 5.6342 0.5841 0.1082 5.3970
LR Test 15.6516 30.5750 30.2204
* sig. at 0.15 level; ** sig. at 0.10 level; *** sig. at 0.05 level
Adoption of Technology and Economic Efficiency of the Small-Holder… • 4641
4642 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

Table 9 shows that information access also have significant effect upon
inefficiency on level ɑ = 0.15, with negative estimated parameter (-0.0068) for
Indonesia and (-0.1167) for East Java. Information in this case includes inputs,
market, and technology. The higher the farmers access upon information, then
the lower inefficiency level would be. This results is different with what Daadi et.
al. (2014) found that concludes that gender, demonstration frequency, and mango
development were those which affect technical inefficiency of mango farmers in
Ghana.

CONCLUSIONS AND IMPLICATIONS

Conclusions
This research studies regarding adoption and innovation of off-season technology
in Indonesia, estimation of technical, allocation, and economic efficiency using
stochastic frontier function, and factors affecting technical inefficiency. From the
analysis, it can be concluded that :
zz Off-season technology adoption by Indonesian mango farmers is still low
(17.92 per cent). By comparing between the two biggest mango centres
in Indonesia, West Java have higher implementation level than East Java,
23.42 per cent compared to 12.50 per cent. Farmers decision to implement
off-season technology is affected by tree ownership status, number of trees,
information access, market acess, capital access, and number of trainings
conducted.
zz Technical efficiency of Indonesia mango farmers averages in 70.37 with
wide range between 26.93 and 90.29. Meanwhile, allocative efficiency
averages in 15.05 with range of 0.05 to 81.46. Economic efficiency, being
a valuable indicator for farmers welfare, averages in 11.09 per cent with
range between 0.03 up to 67.13 per cent. This indicated that Indonesia
had a low economic efficiency level and still have potential to increase her
mango production.
zz Inefficient condition for Indonesia mango farming is affected significantly
by the variables of: farmer experience, credit access, and information access.

Implications
zz Area that agro-ecologically fit for mango crops need further treatment and
development, such as : Pasuruan, Probolinggo, Situbondo, and Bondowoso
on East Java, also Cirebon, Indramayu, Majalengka, and Kuningan on West
Java. Off-season technology is feasible for more intense socialization, since
it has been proven to ensure mango availability all season continuously.
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4643

To increase off-season technology adoption, it is necessary to improve

information access, market accessibility, capital access, and number of
trainings.
zz For farmers on West Java, it is recommended to reduce excessive input
usage to increase mango farmers economic efficiency. On the other
hand, East Java farmers need to optimize additional labor usage without
sacrificing productivity.
zz Government should prepare sufficient resources and infrastructure to
support mango farming development, for example : market, transportation,
water resource, and other production inputs such as seeds, fertilizer,
insecticide and farming tools.
zz Mango farmers groups leading to associations or unions that can act
as source of market information, technology and capital should be
strengthened. This will then leads to increase of off-season technology
implementation, causing a major increase on mango farmers revenue.

ACKNOWLEDGEMENT
Field survey and data collection of this work were supported by the USAID
Prime Award [grant number EDH-A-00-06-00003-00] issued to the University of
Wisconsin under sub-agreement with the Michigan State University No. 61-3939B.
Data processing, analysis and writing of this paper were funded by the Universitas
Padjadjaran Research Funds [grant number 023.04.2/189726/2013]. Researchers are
very grateful to the award received.

References
Aigner, D.J., C. A. K. Lovell and P. Schmidt. 1977. Formulation and Estimation of Stochastic
Frontier Production Function Models. Journal of Econometrics, 6 (1) : 21-37.
AndiIsaac and Afrizon. 2011. Perception and Adoption Levels Paddy Farmers Against
Application System of Rice Intensification(SRI) in the village of Bukit PeninjauI, District
of Sukaraja, Seluma. Journal of Agricultural Informatics. Vol 20.20. (version Indonesian)
Battese, G.E and Coelli, T.J. 1995. A Model for Technical Inefficiency Effects in a Stochastic
Frontier Production Function for Panel data. Empirical Economics. 20 (2), 325-332.
Bravo-Ureta, Boris E., and Laszlo Rieger. 1991. “Dairy Farm Efficiency Measurement
Using Stochastic Frontiers and Neoclassical Duality.” American Journal of Agriculture
Economics 73, no. 2: 421–28.
Bravo-Ureta, Pinheiro (1997). “Technical, Economic and Allocative Efficiency in Peasant
Farming: Evidence from the Dominican Republic.”The Developing Economies 35(1):
48-67.
4644 • Lies Sulistyowati, Ronnie S. Natawidjaja and Boby Rahmat

Coelli, T. 1996. A Guide to Frontier Version 4.1: A Computer Program for Stochastic Frontier
Production and Cost Function Estimation. Centre for Efficiency and Productivity
Analysis, University of New England, Armidale.
Coelli, T.J., Rao, D.S.P, O’ Donnell, C.J. and Battese, G.E. 2005. An Introduction to Efficiency
and Productivity Analysis. 2nd Edition. Publisher: Science and Business Media, Inc. 233
Spring Street, New York, NY 10013. USA.
Daadi, B.E, Issahaku Gazali and Joseph Amikuzuno. 2014. Technical Efficiency Analysis of
Organic Mango Out-grower Farm Management Types : The Case of Integrated Tamale
Fruit Company (ITFC) Out-growers in Northern Region. African Journal of Agriculture-
Economics and Rural Development. ISSN 2141-5091. Vol. 2(3). Pp. 129-137. May 2014.
Debertin, David L. 1986. Agricultural Production Economics. Macmillan Publishing
Company. New York.
Dixon, Huw. 1997. “Contraversy : The Source and measurement of Technical Change”. The
Economics Journal 107. September 1997. Pp. 1518
Fadwiwati, Andi Yulyani, Sri Hartayo, Sri Utami KuncoroandI. Wayan Rusastra. 2014.
Analysis of Technical Efficiency, allocative efficiency and Economic Efficiency Based
Farm Corn Varietiesin Gorontalo province. Agro Economic Journal Vol. 32. No. 01.
(version Indonesian)
Greene, W. H. 1993. The Econometrics Approach to Efficiency Analaysis. In: Fried, H.O.,
C. A. K. Lovell and S. S. Schmidt (Eds). The Measurement of Productive Efficiency. Oxford
University Press, New York.
Hayami, Yujiro and Vernon W Ruttan. 1985. Agricultural Development, An International
Perspective. Tehe Johns Hopkin University Press. Baltimore and London.
Hubeis AVS. 2007. Innovation-Comunication. Universitas Terbuka. Jakarta. (version
Indonesian)
Jondrow, J., C. A. K. Lovell, I. S. Materov and P. Schmidt. 1982. On Estimation of
Technical Inefficiency in the Stochastic Frontier Production Function Model. Journal of
Econometrics, 19 (1) : 233-238.
Mar, Sein. Mitsuyasu Yabe and Kazuo Ogata. 2013. Technical Efficiency Analysis of Mango
Production in Central Myanmar. Journal ISSAAS Vol. 19, No.1: 49-62.
World Bank (2007) World Development Report 2008: Agriculture and Development
(Washington DC: The World Bank).
Mardikanto T. 2009. The agricultural extension system. First Edition. Sebelas Maret University
Press. Surakarta. (version Indonesian)
Marsianus Falo. 2011. Technology Adoption Rate of Corn Hybridsby Dryland Farmers in
North Central Timor East Nusa Tenggara province. [Thesis]. Graduate School ofIPB.
Bogor. (version Indonesian)
Msuya, E.E., S. Hisano, and T. Narlu. 2008. Explaining Productivity Variation among
Smallholder Maize Farmers in Tanzania. MPRA Paper no. 14626. Online at http://
mpra.ub.uni-muenchen.de/14626
Nadapdap, Hendrik Johanes. 2014. The Study of Production Technology Adoption of
 Adoption of Technology and Economic Efficiency of the Small-Holder… • 4645

Mango Commodity (An Example of Mango Farmers in West Java Province). Thesis
Magister-Padjadjaran University-Bandung.
Natawidjaja et al., 2009. Mango Value Chain Key Informant Interview Synthesis. Access to
Modernizing Value Chains by Small Farmers in Indonesia. USAID AMA CRSP Project.
Ogundari, K. and S. O. Ojo. 2006. An Examination of Technical, Economic and Allocative
Efficiency of Small Farms: The Case Study of Cassava Farmers in Osun State of Nigeria.
Journal of Central European Agriculture, 7 (3) : 423-432.
Prayoga, A. 2010. Productivity and Technical Efficiency of Organic Paddy farming in Wet
Land. Agro Ekonomi Journal 28 (1); 1-19. (version Indonesian)
Sulistyowati, Lies; R.S. Natawidjaja, Z. Saidah. 2013. Socio –Economic Factors Affecting The
Decision Involved in Mango Farming System with Informal Trader. Sosiohumaniora
Journal, 15(3): 284-292. Padjadjaran University. (version Indonesian)
Syamsyiah Gafur., 2009. Motivation Implementing Farmers in Cocoa Production
Technology (Case Sirenja District of Donggala, Central Sulawesi). [Thesis]. Bogor
Agricultural Institute. (version Indonesian)
Wandschneider, T, I. Baker, and R. Natawidjaja, 2012. Indonesia Agribusiness Development
Opportunities – Analysis of Mango Value Chains, Aus AID-IPD Rural.

View publication stats