Research Article

Examining Technical South Asia Economic Journal

19(1) 22–42
Efficiency in Indian ©2018 Research and Information
System for Developing Countries &
Agricultural Production Institute of Policy Studies of Sri Lanka
SAGE Publications
Using Production sagepub.in/home.nav
DOI: 10.1177/1391561418761073
Frontier Model http://journals.sagepub.com/home/sae

Kailash Chandra Pradhan1

Shrabani Mukherjee2

Abstract
The study estimates the technical efficiency (TE) of agricultural production in
India using production frontier model for both cross-section and panel data for
the years 1999 and 2007. Given the persistent low productivity of agriculture in
India, there is a serious need to identify the determining factors for technological
lock-in for agricultural production in order to accelerate sustainable productivity
and technical efficiency. The model claims that farmer’s education, household’s
production process, proportion of irrigated area covered by canals, availability
of wells, yielding variety of lands, government services, agricultural expenditure
by local government and women reservations in local government significantly
contribute to efficiency in resource utilization in farm production. Traditional tech-
niques such as ‘learning by doing’ is, generally, preferred than the adoption of new
technologies for agricultural production. It makes a constraint of technological
lock-in.

JEL: C23, C33, D20, O13

Keywords
Production function, agricultural farmers, technical efficiency, panel data

1
Joint Director, National Institute of Labour Economic Research and Development (NILERD), Delhi,
India.
2
Associate Professor, Symbiosis School of Economics, Symbiosis International University, Pune, India.

Corresponding author:
Shrabani Mukherjee, Associate Professor, Symbiosis School of Economics, Symbiosis International
(Deemed University), 3rd Floor, SCHC Building, Senapati Bapat Road, Pune, Maharashtra 411004, India
E-mails: shrabani.mukherjee@sse.ac.in; shrabani0808@gmail.com
Pradhan and Mukherjee 23

Introduction
Agricultural productivity is extremely important for long-run growth of a devel-
oping country like India where agriculture continues to lead the economic scene
as a second largest producer for agriculture product which is 7.68 per cent of total
global agricultural output (Government of India [GoI], 2017). Agriculture is the
spine of the Indian economy and agricultural productivity has acquired a key posi-
tion in developmental policy to reduce poverty and also to increase food security
in rural areas. Although physical contribution from agriculture is only about 14
per cent to the overall gross domestic product (GDP) but it affects all other sectors
directly and indirectly through the consumption demand from rural population.
Around 500–550 metric ton (mt) of crop residues are produced per year in India.
However, farm production is expected to be two times more with dwindling cul-
tivable land and resources to increase farmers’ income and environmental sustain-
ability given its potential (GoI, 2015). India became the fastest growing major
economy of the world since last few years. However, the accelerating economy-
wide growth has not been resulted through an acceleration of agricultural output
growth.1 The real per capita GDP (1999–2000 prices) averaged less than 5 per
cent per year during the 1980s and 1990s, and it has been increased to more than
7 per cent per year during the period 2003–2007 (Planning Commission, 2008).
Whereas, in terms of the share in GDP, the agricultural output growth fell from
37.9 per cent to 17 per cent between 1980–1981 and 2008–2009. Further, the
share of agriculture in total employment has declined as well. Further, the agricul-
ture sector used to employ 73.9 per cent of the economically active population in
1973–1974, which has declined to 56.5 per cent in 2004–2005. However, it has
been estimated that this sector has huge potential to engage millions of the coun-
try’s poor to improve their standard of living. Moreover, rural Indian women are
widely involved in agriculture. Around 84 per cent women depend on agriculture
for their livelihood (Rao, 2006). Given women participation and their contribu-
tion in agricultural production, agricultural growth is necessary for building up
women workforce for long-run inclusive growth and rural development (Shiva,
1991).
Although agriculture and allied sectors like forestry and fisheries still domi-
nates Indian economy by acquiring around 14 per cent GDP, half of work force
and 10 per cent of the economy’s exports which challenged by underutilization of
capacity, low productivity and limited revenue benefits for the farmers. The effi-
ciency of output generation from proper input utilization given technological con-
straints is extremely important for sectoral growth in endogenous growth
frameworks. Input quality, access to input markets, economies of scale and use of
technology are also influencing factors for agricultural productivity. In order to
assess the productivity of agricultural sector, we need to examine the total factor
productivity (TFP, henceforth), which tracks the changes in inputs to outputs.
Overall agricultural production has been influenced by the rise in prices in com-
modity market as well as by input prices. We can’t ignore the impact of inflation
in commodity market. Specifically, price rise of food items and price upswing in
factor markets through demand side and supply side factors affect agricultural
24 South Asia Economic Journal 19(1)

productivity. Thus, variations in TFP infers as a share of the contribution of

conventional and non-conventional inputs in agriculture. In India, agricultural
commodity prices are largely determined by TFP clubbed with domestic price
policy. Therefore, the overall price rise and the sequential impact on demand for
food items become a serious concern for agricultural policy practitioners in devel-
oping countries and predominantly for India. India has been experiencing, for last
37 years, an agricultural growth based on rising conventional factors where,
practically, average TFP was negative (Tripathi, 2008).2 As recorded, it is mainly
because of the declining proportional contribution of public investments in the
agricultural sector. To mention, for only initial periods of reforms, agricultural
TFP growth was positive. Otherwise, TFP growth has experienced large variation
over time and it became negative after the mid-1980s. In the initial periods of
90’s, TFP growth has accelerated and then it has again shown a downward trend
during mid-1990s. The reason behind this is also the swing in public and private
investment in agricultural sector. The TFP in Indian agriculture has been rising at
a faster rate after the mid-2000s. The rising trend in TFP growth after 2004–2005
suggested that the recent growth in agriculture is accompanied with improved
technology and efficiency and thus became sustainable (GoI, 2015). Several insti-
tutional factors have been considered as imperative for the sustainability of agri-
cultural growth in empirical studies. Huge reforms implemented in the different
other sectors have contributed to the agricultural growth as well through fluctua-
tions in agricultural prices. There is no doubt that agriculture remains priority in
India for long-run growth plans, as given India’s poverty level and large propor-
tion of population based on agriculture, a higher priority to agriculture will help
to achieve poverty eradication through inclusive growth. The outcome of 5-year
plans has recorded an outstanding progress in agricultural productivity. The high-
est priority was rendered to increase agricultural production during the first 5-year
plan (1951–1956). The third 5-year plan onwards growth plans emphasized on
self-sufficiency in food grain production. During the sixth plan (1980–1985),
India moved towards long-run growth plans for agriculture through land reforms,
use of high yielding varieties (HYV) seeds and groundwater resources. Vast tech-
nological improvement as well as up gradation in marketing and storage facilities
helped agricultural growth to reach to highest level ever during seventh plan
(1985–1990), which is considered as green revolution in Indian economy. Further,
it may be noted that in the last two 5-year plans, it is clearly mentioned that for the
economy to grow at 9 per cent, it is important that agriculture should grow at least
by 4 per cent per annum. A National Agricultural policy has been announced in
2000 to achieve more than 4 per cent per annum growth rate in agriculture through
privatization, price protection, land leasing and contract farming by private com-
panies. However, since 2008–2009, the world economy has been witnessing the
slow growth rate which has resulted in sluggish growth across the sectors in India.
Eleventh 5-year plan (2007–2008 to 2011–2012) made a target to reverse the
deceleration in agriculture growth and productivity, which continued to be main
focus in Twelfth 5-year plan. In order to achieve 8–9 per cent rate of growth in
overall GDP and to fulfil the target in terms of poverty reduction, agricultural
growth needs to be accelerated. Further, another policy target, namely, ‘inclusive
Pradhan and Mukherjee 25

growth’ can be reached only if agriculture productivity raises and reaches to its
optimum utilization of its potential. It has been widely viewed across regions of
the country. Given this, agriculture has always been kept at the centre of planning
agenda to ensure long-term food security for the people.
Despite having number of reforms in agricultural sector and huge policy level
initiatives to increase productivity through technological improvement, number
of problems persists, such as, fragmented land holding, irrigational constraints,
extreme dependence on traditional technique. It has been recorded that large land
holdings helps to implement modern agricultural techniques and thus enhance
productivity, whereas small land holdings confine the farmers to use traditional
techniques that limit yielding efficiency. Moreover, most of the farming in India
is still monsoon-dependent due to lack of proper irrigation. Huge debts, heavy use
of fertilizers, low productivity and water mismanagement are the added problems.
Therefore, the main question pursued in this article is why the farm economy has
not responded to the demand side forces that have been unleashed on it from the
rapid economy-wide growth. One more issue is that before making one after
another policy to enhance productivity, we need to know whether the farmers
grow crops efficiently with available technologies or they use resources with
excess capacity. This article finds some of important factors that may influence
the technical efficiency TE, hereafter, of the farmers so that development of agri-
cultural policies follows the right track. This article examines the TE of Indian
agricultural farmers in both cross-section and panel year of 1999 and 2007 using
production frontier model. Then, it determines the factors which may affect the
production efficiencies. The study has taken the factors such as household age,
education level of the households, household size, proportion of family supervi-
sion cost to total labour cost, proportion of HYV area, distance to ‘pucca’ road,
proportion of irrigated area covered by canals, tanks as well as wells, the govern-
ment agriculture extension services, agricultural expenditure by local government
and women reservations. These factors may influence the production process and
TE of the farmers.
This article is organized in five sections including introduction. Second section
covers review of literature which focuses on theoretical explanation of relation
between determinants of sectoral growth and long-run inclusive growth of an
economy following different schools of thought and need of measuring TE in
agricultural production. Third section describes methodology and data base.
Fourth section presents the empirical estimation and analysis. Fifth section pro-
vides brief conclusion of the study.

Review of Literature
An emerging economy needs a strong progression in agricultural growth in transi-
tional phase as the contribution of primary sector on overall economic growth
remains throughout the process of structural transformation. Agricultural surplus
mobilizes investment towards industrial development and eventually lowers the
food prices during the transformation process.3 Further, it stimulates industrial
26 South Asia Economic Journal 19(1)

sector by generating demand for intermediate inputs, as increase in agricultural

incomes are spent, basically on domestically produced goods which are not
tradable.
At theoretical level, at the beginning, Solow (1956) presented a fundamental
model of economic growth followed by Harrod (1939) and Domar (1946) on
fixed-factor proportions model or Lewis (1954) and Fei–Ranis (1961) on dual
sector migration model. Solivian concept specifies neoclassical production pro-
cess, where physical capital, labour and an exogenous technology influence the
level of output. Growth accounting analysis generally assumes that the Solow
residual captures TFP growth. It serves as the point of departure for later growth
theories. Various endogenous growth models also specified product innovation
for sustainable long-run growth and sectoral productivity through technological
development and physical capital and human capital accumulation for inclusive
growth (Barro, 1990; Lucas, 1988; Romer, 1986, 1990). Endogenous growth
models modified the basic concept by considering the role of constant and increas-
ing returns to capital. Another argument comes against the neoclassical growth
model, which is that it cannot explain country-wide disparity in per capita growth
rates. Therefore, there should be something other than physical capital and human
capital, which creates the differences in growth rates of across countries. The
‘Learning by Doing Model’ of Arrow (1962) provides a clear idea that labours
become efficient at producing output by undergoing the training process by doing
the same over time, and hence more productive. They can then make others more
productive by teaching the same skills. Passing the same skill from worker to
worker and across generations, the economy gets productive through learning by
doing the same method. Romer’s (1986) model and Robert Lucas’s (1988) human
capital models get around the diminishing marginal returns to ‘capital’ accumula-
tion by expanding it through knowledge or human capital, both of which may
have positive externalities. Recent description of growth models opened another
side of the growth theories, which direct the sources of TFP growth. By incorpo-
rating technological change, those models consider the diffusion of technology
between countries, and the ability of developing countries to adopt and implement
foreign technology. Rivera-Batiz and Romer (1991) showed three determining
factors for long-run growth of an open economy which are mentioned as: first,
more efficient utilization of sectoral resources; second, labour productivity tends
to boost sectoral growth through production externality that spill overs interna-
tionally; and third, proper reallocation of factors of production across sectors.
Some economists think that physical capital accumulation plays the key part in
growth process while others claim that TFP drives the output growth. Moreover,
in order to make a steady well-behaved investment, the models introduced con-
cept of returns from private and social investment. The diverse inferences of these
two different schools of thoughts in terms of exogenous and endogenous growth
models have led to many experiential studies in recent times focusing on TFP.
Studies also considered role of institutions in determining where and how the
economy will move in the long run. For example, favourable government policies
focusing on greater public investment in productive sectors direct the economy
Pradhan and Mukherjee 27

towards faster growth in long run (Barro, 1995). Further, studies recorded that
agricultural productivity growth has the maximum poverty-reducing effect for a
developing economy with the lowest levels of development, such as in Sub-
Saharan Africa and South Asia (Hanmer & Nashchold, 2000). On the other hand,
in developed regions of the world, such as East Asia and Latin America, sectoral
progress other than agricultural production appears to have higher poverty reduc-
tion impact (Hasan & Quibria, 2004). Agricultural productivity through forward
and backward linkages with other sectors of the economy escalates pace of devel-
opment and has been established in literature.
In any case, there is no doubt that primary sector, at least for the developing
countries, has enormous ability not only to reduce poverty but also to determine
overall economic growth through strong linkages with the rest of the economy.
Empirical studies experience time to time that agricultural productivity slows
down poverty and livelihood security for developing countries. Further, studies
on growth models exhibited the evidence that capital stock in the agricultural
sector, labour productivity in agricultural sector and incidence of starvation in
developing countries are highly connected. Economic growth is caused mainly
by three direct factors, namely, cost of labour and capital which came through
basic systematic quest for the drivers of long-run economic growth by Solow
(1956, 1957), Romer (1986, 1987, 1989), Lucas (1988), Human Capital by
Schultz (1961), Becker (1964); as well as learning by doing by Arrow (1962),
knowledge capital and social capital generated through investment by Romer
(1990), and TFP by Senhadji (2000), Hughes et al. (2008). There are recent
empirical studies found out the determinants of agricultural productivity for dif-
ferent countries. These socio-economic differences affect agricultural productiv-
ity through the change in land-use pattern (Baumann et al., 2011; Kuemmerle et
al., 2008). It has been seen that agricultural production in developing countries is
subsistence-oriented. In order to reach the goal to reduce poverty through food
security and nutrition for all, it is required to efficiently convert the farm sector
away from household-level production towards an integrated economy driven by
growth in agricultural yield. Across the globe wherever the agricultural transfor-
mation progression has been examined, it has been established that agricultural
productivity growth has been determined by enhanced technologies, improved
seeds, quality fertilizer and water control (Gabre-Madhin & Johnston, 2002;
Johnston & Kilby, 1975; Mellor, 1976). Foster and Rosenzweig (2003) showed
in an empirical study with similar data set for Indian agriculture experienced
over the past three decades huge yield improvements through the impact of green
revolution. It has eventually improved rural non-farm activities, incomes of
enterprises through the non-local capital. Further, it has seen that India has expe-
rienced substantial growth in both rural per-capita incomes and in rural diversi-
fication, with now almost 50 per cent of rural incomes channelizing from outside
the agricultural sector. Therefore, agricultural productivity complements other
sectors progress as well. Based on this empirical evidences on the theoretical
framework, Indian developmental policies continued to get implemented focus-
ing on agricultural productivity over this period has been continued.
28 South Asia Economic Journal 19(1)

Therefore, efficiency measurement is useful in determining the magnitude of

the gains that could be achieved by adopting improved practices in agriculture
production with a given technology (Armagan, 2008; Rahman, 2003; Tauer, 2001;
Zhu, 2000). There are a number of studies that have estimated TE in different sec-
tors. Thiam, Bravo-Ureta and Rivas (2001) have used a meta-analysis to review
empirical estimates of TE in developing country for agricultural sector. A data set
of 51 observations of TE from 32 studies has been used to examine whether spe-
cific characteristics of the data and econometric specifications account for sys-
tematic differences in the efficiency estimates. Shanmugam (2000) projected the
efficiency of rice farms in Bihar. He has also estimated TE for rice production in
Karnataka in 2002 and rice, cotton and groundnut production in Tamil Nadu, in
2003, using same methodology, the results of these studies were beneficial for
policymakers to rationalize the improvement in suitable plans for a particular crop
production in the region. Shanmugam and Venkataramani (2006) have shown that
health, education and infrastructure can be powerful drivers of efficiency.
O’Donnell, Rao and Battese (2008) developed same meta-frontier (MF) model,
which enables the estimation of technology gaps for producers under different
technologies relative to the potential technology available to the industry as a
whole. The model facilitated the interpretation of grand TE scores by decompos-
ing them into group-specific efficiency and technology differences as well. Rao,
Brümmer and Qaim (2012) have taken a MF approach and combined this with
propensity score matching to estimate treatment effects among vegetable farmers
in Kenya. Limam and Miller (2004) examined cross-country configurations of
economic growth through assessing a stochastic frontier production function for
80 developed and developing countries. They have used decomposition analysis
on output change into factor accumulation and production efficiency improve-
ment via TFP growth. Battese and Coelli (1988) analysed dairy farms in New
South Wales and Victoria for the 3 years 1978–1979, 1979–1980 and 1980–1981
and using a generalized-likelihood-ratio test for the stochastic frontier Cobb–
Douglas production functions for both states. Battese, Coelli and Colby (1989)
projected a stochastic frontier production function for Indian farms using data
over 10 years and placed evidence that the non-negative farm effects had half-
normal distribution and measured TE.

Methodology and Data

TE is considered as the ability to produce a given level of output with a minimum
quantity inputs under certain technology. Studies primarily focused on TE using a
deterministic production function with parameters computed using mathematical
programming techniques. However, with inadequate characteristics of the assumed
error term, this approach has an inherent limitation on the statistical inference on
the parameters and resulting efficiency. Aigner, Lovell and Schmidt (1992) and
Meeusen and Van den Broeck (1997) independently developed the stochastic
frontier production function to overcome this deficiency. The stochastic frontier
Pradhan and Mukherjee 29

production function for the panel data for estimating household level TE is speci-
fied as:

Parametric Frontier Model

Yit = f ( X it ;� β ) ; where, I = 1, 2,…n and t = 1, 2,...T (1)

Where i is the observation and t is time period. Yit represents the possible produc-
tion level for the ith sample firm (Cobb–Douglas production function) of the vec-
tor, Xit denotes the actual input vector, β is vector of production function. In order
to estimate β of ith firm in a particular industry by mathematical programming
methods based on a cross-section of N firms we need to minimize
n
min ∑ | Yit − f ( X it ; β ) | subject to Yit ≤ f ( X it ; β ) if is f ( X it ; β ) is linear in β
1

n
min ∑ [Yit − f ( X it ; β )]2 subject to Yit ≤ f ( X it ; β ) if is f ( X it ; β ) is nonlinear
1

In order to depict variation in output among firms with identical input vectors, a
disturbance term has been implicitly assumed. In order to estimate with statistical
basis, Schmidt (1976) explicitly added a one-sided disturbance to the aforemen-
tioned conditions, which produces the modification as follows.

Stochastic Frontier Model

Yit = f ( X it ;� β )+ εit (2)

A stochastic factor that refers random shocks disturbing the production process.
Here, e is the error term that is composed of two elements. Given a distribution
assumption for the disturbance term, the model can then be estimated by maxi-
mum-likelihood techniques.

ε = Vit −U it (3)

Where Vit is the symmetric disturbances assumed to be identically, independently

and normally distributed as N (0, ev 2 ) given the stochastic structure of the frontier.
The second component Uit is a one-sided error term that is independent of Vit
and is normally distributed as, N (0,  u 2 ) , allowing the actual production to short
fall below the frontier but without attributing all short falls in output from the
frontier as inefficiency. The reason behind this specification is, according to
literature there exist two different types of economically distinguishable random
disturbances for production process. However, our perspective of analysis and
30 South Asia Economic Journal 19(1)

interpretation is obviously new with respect to theoretical and practical view-

points. The negative disturbance term Uit refers the detail of each farm’s output
which necessarily lies on or below its frontier [ f ( xit ; β) +Vit ]. This deviation is
the considered as the outcome of features under the farm’s control, which is even-
tually considered as technical and economic inefficiency. However, the frontier
itself can vary randomly across farms, or over time for the same farm. This model
collapses to a deterministic frontier model when σ v 2 = 0 , and it collapses to the
stochastic production function model when σ u 2 = 0. This model is evidently sto-
chastic. Following the logic, we can measure the variances of Vit and Uit, in order
to get indication on their relative sizes. Further, we can estimate TE which is
measured by the ratio. The TE has two types of measures: output-oriented and
input-oriented. Our study has considered an output-oriented measure, then it is
defined as observed output (Yit) to the corresponding frontier output (Y*it), which
is maximum feasible output using the available technology derived which is
defined as follows:

Yit
TEit  (4)
Y *it
where, 0 ≤ TE ≤ 1,
The TE takes values within the interval (0,1), where 1 indicates a fully efficient
firm.

Yit = Y *it .TE = ’ ( xit ; β ) (5)

Yit = f ( xit ; β ) exp (−uit ) (6)

where u  0
The frontier becomes deterministic. Every deviations from extreme output are
attributed to inefficiency. However, occasionally extreme output itself might be
lesser (higher) due to some impact of exogenous random shocks. That time pro-
duction frontier will shift.

Yit = f ( xit ; β ) exp(vit ) exp (−uit ) (7)

with v ≤ 0 and u ≥ 0
Where, ƒ(x; β) refers to deterministic trend, exp(v) identifies effect of exogenous
shocks and exp(–u) refers to inefficiency. Finally, ƒ(x; β).exp(v) refers to stochas-
tic frontier.

Yit f ( xit ; β ) exp(vit ) exp (−uit )

TE = = (8)
f ( xit ; β ) exp(vit ) f ( xit ; β ) exp(vit )

TE = exp(–uit)
If Ui = 0, the farm were 100 per cent efficient.
Pradhan and Mukherjee 31

The Cobb–Douglas production function is used in this article. Here, the model
is estimated using STATA11.2. The article used the two-stage estimation proce-
dure in which first the stochastic production function is estimated, from which
efficiency scores are derived, then in the second stage, the derived efficiency
scores are regressed on explanatory variables using ordinary least square (OLS)
methods. The TE of ith farm on tth period is defined as

TEit = Z it δ (9)

Where z is a vector of observable explanatory variables and  is a vector of

unknown parameters. We will estimate first, production function as
n
ln (Yit ) = β 0 + ∑ ln ( xit ) β j + (Vit −U it ) (10)
j=1

We will measure TE of each firm (household). The functional form of the Cobb–
Douglas stochastic frontier production model is defined as
n
ln (TEit ) = δ0 + ∑ δ jln( X it ) +Vit −U it (11)
j=1

Where the subscripts i and t represent the ith farm and tth year of observation and
jth refers to number of explanatory variables specified in the model, and TEit rep-
resents the TE in agriculture. ‘ln’ refers to the natural logarithm.
A two-step process has been used to bring out the TE in Indian agriculture
using maximum-likelihood method. The OLS approximations of the parameters
were achieved by grid search in the primary stage, and then these measures were
used to estimate the maximum-likelihood estimates of the parameters considered
as the frontier estimates of Cobb–Douglas stochastic frontier production model.
The study has used the last two rounds (1999 and 2006)4 of Rural Economic
and Demographic (REDS) household survey data organized by the National
Council of Applied Economic Research (NCAER) providing consistent informa-
tion on 242 villages spread across 16 states in India. The original sampling design
has been framed as stratified form. Study districts in each state were selected on
the basis of the following characteristics. First, two districts from each state got
included on the basis of high potential in crop intensity with input provision pro-
gramme placed in areas; second, one random sample from rest of the districts.
There were 100 districts represented in the 1971 in first round of ARIS data
set. In 1982, 250 of the original 259 villages were revisited (the state of Assam
was excluded due to local political disturbances rendering survey activity impos-
sible) and 4,979 households have been surveyed approximately. Two-thirds of
which were the same in 1999. Almost all the villages were surveyed. Only eight
sample villages from Jammu and Kashmir got excluded from survey due to prob-
lems of local insurgency. In the last two survey rounds, all of the surviving
households in the 1982 survey were surveyed again, including all split-off
households residing in the same village clubbed with a random sample of new
32 South Asia Economic Journal 19(1)

households. Because of household division and the new sample design incorpo-
rating all village-resident, the number of households surveyed in the 1999 round
increased to 7,474 compare to number recorded in 1980 round. The latest round
in 2006 has a sample size of 8,659 households selected from 242 villages and it
includes all of the households surveyed in 1999 and the split-off households
residing within these villages. Each village has approximately eight new ran-
domly selected households. The panel data set encompassing 1999 and 2007
rounds of the survey includes 5,885 households. Each round of the survey has
three components. A listing of the village households is first done for better sam-
pling. Household-level questionnaires were canvassed over three agricultural
periods (basically, information for most of the households recorded for the agri-
cultural year 2007–2008).

Results and Discussions

This article has considered a stochastic frontier approach for estimating produc-
tion functions from panel data for two crop seasons. We assess these for each
crop, such as paddy, wheat, cereals, pulses, oilseeds and other crops, according to
their yield season at farm household level. The data description for the 1999 and
2006 rounds are summarized as shown in Table 1. We have taken the variables,
such as labour income, cropped area, price of fertilizer, manure, land use, etc. The
stochastic production frontier framework brings out TE and identifies the fact that
random shocks outside the control of farmers, may influence the production out-
put (Aigner, Lovell & Schmidt, 1977; Meeusen, & van den Broeck, 1977; Battese
& Coelli, 1992, 1995). Explanatory variables are selected based on literature.
Proportion of irrigated area, value of seeds per acre, value of fertilizer and manure
per acre, pesticides per acre, value of total labour cost per acre, proportion of hired
labour and fixed cost per acre by the farms would have an important role in

Table 1. Descriptive Statistics

Variables 1999 Round 2006 Round

Value of crop 7,940.52 9,020.04
Cropped area 6.79 6.55
Proportion of irrigated area 0.57 0.60
Value of seeds per acre 440.52 568.32
Value of fertilizer and manure per acre 1,012.58 1,243.82
Proportion of manure to total value of fertilizer and 0.24 0.15
manure
Value of pesticides per acre 150.82 191.33
Value of total labour cost per acre 1,044.31 1,821.82
Prop. of hired labour 0.21 0.22
(Table 1 Continued)
Pradhan and Mukherjee 33

(Table 1 Continued)

Variables 1999 Round 2006 Round

Value of fixed cost per acre 3,083.99 6,528.18
Value of bullock cost per acre 73.28 59.61
HH head age 50.19 51.63
Household size 6.55 5.68
Mean education of the HH 3.93 4.74
Family supervision/labour cost 0.10 0.04
Distance to Pucca road 2.41 1.67
No. of times AES activities 2.51 3.97
Proportion of government canal irrigated 0.25 0.65
Proportion of tank irrigated 0.05 0.07
Proportion of well irrigated 0.08 0.08
Proportion HYV area 0.61 0.69
Proportion of agricultural expenditure 0.05 0.03
Women reservation 0.25 0.32
No. of households 4,487 4,869
No. of villages 242 242
Source: Authors’ calculation based on REDS survey.

increasing total output. Irrigation technic is another key factor for farm production
and thus it affects the TE. We have measured number of times agricultural exten-
sion services (AES) activities held in production process, proportion irrigated
land from government canals, proportion of area under tank irrigated, proportion
of well-irrigated and proportion of HYV covered area as independent variables.
Apart from that, it is required to improve the human resource capability and
access to information for the improvement of productivity. Keeping the impor-
tance of the aspect, we have considered mean education of the owner of the farms.
Given women participation rate in Indian agriculture and their association with
local-level political party, we have considered a variable to capture that impact as
well. It is evidently proved that political reservations for women generates posi-
tive impact on India’s progress through its diversity in decision-making process
and creating positive externality through reducing scope of corruption and asym-
metric information with efficient governance benefits (Afridi, 2017). Table 1
describes sample means of specified variables.
The present study used the state-level consumer price indices for mean wage of
rural agricultural labour. Since the survey was rolled out over more than 2 years,
the survey period is matched in each state with the average of the respective
months of the CPI for rural agricultural workers. The article used the consumer
price index to convert all values and prices in real term. The agricultural house-
holds are 4,487 in 1999 round of survey and they are 4,869 in 2006 round of survey.
The total panel agricultural households are 2,273. The 2007 values are deflated
34 South Asia Economic Journal 19(1)

using 1999 as a base period. The crop value has increased by 12 per cent in 2007,
which can be attributed to the high base effect of 1999. The data show that the
aggregate cost of agricultural production has been rising over the periods. It is
evidently present in the data distribution. The proportion of irrigated land and the
value of seed per acre both have been increased, respectively, by 5 per cent and
22.5 per cent. Value of fertilizer and manure per acre while has also been increased
from `1,013 to `1,244 from 1999 to 2007. However, the proportion of manure to
total value of fertilizer has declined. This perhaps reflects declining demand for
manures. Value of pesticides per acre has increased modestly from `151 in 1999
to `191 in 2006. Labour cost in 1999 was `1,044 which has climbed up to `1,821
in 2007. Surprisingly, the price of hired labour per acre has been declined from
`480 in 1999 to `322 in 2006. Fixed cost per acre was `3,084 in 1999 which has
gone up to `6,528 in 2006 reflecting the increased fixed cost component in farm-
ing or might be because of increase in plant size or other long-run investment to
improve the efficiency of farms. This pattern seems that Indian agriculture is
shifting towards capital investment though Indian agriculture is more labour
intensive. Another major point is that value of bullock cost has been declined
which signifies that manual ploughing of land has been decreasing over time.
Average age of household head has remained stable for all the periods while mean
year of education has risen from 3.93 in 1999 to 4.74 in 2007. The proportion of
family supervision to total labour cost has been declined from 1999 to 2007. Mean
distance to pucca road has increased. The number of times AES activities by the
government has increased from 2.5 times in 1999 to 4 times in 2007. Proportion
of government canals irrigated has increased in 2007 to 61 per cent. A proportion
of tanks irrigated have also increased in 2007. It is interesting to see that the pro-
portions of wells irrigated area have constant between periods. Proportions of
HYV area have increased to 12 per cent. This data shows that the irrigated area
from different sources and HYV area has increased over the periods which is very
good sign for Indian agriculture. Proportion of agriculture expenditure has
declined in 2007 when compared to 1999. In 1999, women reservation was at 25
per cent which rose to 32 per cent in 2007. A recent study has provided the evi-
dence of influence of political reservations for women in village councils on the
governance of India’s one of the ambitious anti-poverty programmes, namely, the
National Rural Employment Guarantee Scheme (Afridi, Iversen, & Sharan, 2013).
Our observation supports that there is a positive impact of such political reserva-
tion in agricultural performance as well.
Further, a Cobb–Douglas production function is estimated for an entire sample
for each year separately and for pooled and panel data households as well as
shown in Table 2. The different models are estimated to understand the consist-
ency of the results and it will help to provide the confirmed wide-ranging conclu-
sions. The results showed that there is an inverse relationship between cropped
area and output is perhaps due to the diseconomies of scale in Indian agricultural
production process which was evidently proved by Constantin, Martin and Rivera
(2009) for Brazilian agriculture. The proportion of irrigated area was positively
related to crop output. This attributes that to higher yields on the larger irrigated
cropped area of the households which supports the observation provided by Kyi
Pradhan and Mukherjee 35

Table 2. Estimated Results

1999 1999–2006 1999–2006

Round 2006 Round (Pooled) (Panel)
Variables Ln (Value of Crop Per Acre)
Ln (Cropped area) −0.0813*** −0.0401*** −0.0916*** −0.0454***
(0.00808) (0.00899) (0.00651) (0.00893)
Proportion of irrigated area 0.242*** 0.223*** 0.120*** 0.297***
(0.0171) (0.0191) (0.0158) (0.0193)
Ln (Value of seeds per acre) 0.297*** 0.0137*** 0.0290*** 0.0413***
(0.00841) (0.00296) (0.00243) (0.00393)
Ln (Value of fertilizer and 0.0327*** 0.0316*** 0.0380*** 0.0331***
manure per acre) (0.00340) (0.00335) (0.00231) (0.00333)
Prop. Of manure to total value −0.0640** −0.188*** −0.194*** −0.139***
of fertilizer and manure (0.0323) (0.0313) (0.0232) (0.0325)
Ln (Value of pesticides per 0.0106*** 0.0128*** 0.0107*** 0.0106***
acre) (0.000841) (0.000901) (0.000655) (0.000890)
Ln (Value of total labour cost 0.0665*** 0.125*** 0.0770*** 0.111***
per acre) (0.00869) (0.00896) (0.00723) (0.00898)
Prop. Of hired labour 0.437*** 0.268*** 0.281*** 0.377***
(0.0345) (0.0443) (0.0281) (0.0395)
Ln (Value of fixed cost per 0.0135*** 0.0194*** 0.00788*** 0.0161***
acre)
(0.00211) (0.00187) (0.00139) (0.00204)
Ln (Value of bullock cost per 0.00416*** −0.00265*** 0.00114** 0.00115
acre) (0.000758) (0.000800) (0.000567) (0.000810)
Constant 6.636*** 7.861*** 8.239*** 7.613***
(0.0768) (0.0804) (0.170) (0.0729)
Number of households 4,487 4,869 9,356 4,531
Panel households – – – 2,273
Wald chi2 4,159.46 1,533.36 9,993.25 1,618.08
Technical efficiency 0.751 0.755 0.798 0.855
Source: Authors’ calculation based on REDS survey.

and Matthias (1999) for rice in Myanmar that for large as well as small farmer
irrigation does matter for farm production. The seed is an important factor in a
production process and it has a positive impact on the output. The fertilizers com-
pare to manures have a more significant contribution to output. The increased
proportion of manure has declined the output. The uses of pesticide increase the
production of the output. The results show that the hired labour is still the impor-
tant factor in farm production as the coefficients of the hired labour are larger
36 South Asia Economic Journal 19(1)

compared to other inputs. The fixed cost such as the value of mechanical assets,
non-mechanical assets and other assets has a positive impact on the output. The
bullock cost per acre has declined the output in 2007 and it has positively influ-
enced the output in 1999, pooled and panel periods. Overall, the results have
shown that the coefficients of the labour inputs have a larger impact than other
inputs and here the results conclude that mostly Indian agriculture depends upon
the labour supply.
The results from cross-sectional data show that the TE of the agricultural farm-
ing households is 75.1 per cent in 1999 and it has marginally increased to 75.5 per
cent in 2007. The TE of the pooled sample is 79.8 per cent and it is 85.5 per cent
for panel households. These results show that on an average the panel households
are more technically efficient than whole sample households and the TE has been
increased over periods. Studies (Shanmugam & Venkataramani, 2006) have
measured Indian districts that have a mean TE of 79 per cent and nearly half of the
sample districts (123 out of 248), TE values lie below 80 per cent. This apparently
signifies that Indian agriculture has been improved in TE.
In Table 3, the results of determinants of TE are presented. At first place, TE has
been estimated using production function and then it is considered as the

Table 3. Determinants of Technical Efficiency

1999–2006 1999–2006
Variables 1999 Round 2006 Round (Pooled) (Panel)
HH head age −0.00285 0.00313 0.00168 0.0116***
(0.00432) (0.00398) (0.00213) (0.00133)
Household size 0.00423* 0.0107*** 0.00785*** 0.00768***
(0.00245) (0.00228) (0.00121) (0.000688)
Mean education of 0.00183*** 0.000415* 0.0001 0.000385***
the HH (0.000700) (0.000237) (0.000168) (0.0001)
Family supervision/ 0.0163** 0.0282** 0.00282 0.0184***
labour cost (0.00668) (0.0117) (0.00397) (0.00168)
Distance to Pucca −0.000658*** −0.00159*** −0.000102 −0.000456***
road (0.000149) (0.000127) (0.0001) (0.00004)
No. of times AES 0.00130*** 0.000261** 0.000376*** 0.0001***
activities (0.000131) (0.000132) (0.0001) (0.00003)
Proportion of govt. 0.000352 0.000864*** 0.000614*** 0.00003
canal irrigated (0.00276) (0.000193) (0.000149) (0.000139)
Proportion of tank 0.0119 0.00120 −0.00280 0.0108***
irrigated (0.00880) (0.00504) (0.00323) (0.00277)
Proportion of well 0.000609 0.0163*** 0.00263 0.0114***
irrigated (0.00623) (0.00480) (0.00278) (0.00321)
(Table 3 Continued)
Pradhan and Mukherjee 37

(Table 3 Continued)

1999–2006 1999–2006
Variables 1999 Round 2006 Round (Pooled) (Panel)
Prop. HYV area 0.0274*** 0.00542 0.00360* 0.0326***
(0.00432) (0.00387) (0.00205) (0.00171)
Proportion 0.00907 0.0445*** −0.00296 0.0119***
of agricultural (0.0122) (0.0110) (0.00591) (0.00351)
expenditure
Women reservation 0.000865 0.0215*** 0.00289** 0.00277***
(0.00283) (0.00225) (0.00128) (0.000524)
Constant 0.740*** 0.684*** 0.776*** 0.797***
(0.0164) (0.0161) (0.00835) (0.00522)
No. of households 4,487 4,868 9,355 4,531
No. of panel – – – 2,273
households
Hausman Fixed effect – – – Yes
F-stat 20.07*** 27.35*** 9.42*** 13.92***
Source: Authors’ calculation based on REDS survey.

dependent variable in the second stage of regression framework to estimate its

determinants as mentioned in methodology part. The household characteristics of
the agricultural household (farmer’s age, yearly income, number of dependents in
family, etc.) are the important factors to influence the TE. The results show that the
aged head of household is more efficient to produce the output efficiently and it is
significant for the panel households. This ensures that aged households have more
experience and using their past learning in the production process to produce more
output with given level of inputs efficiently. Famers mostly use the traditional
method of farming which directs that Indian farm sector has been restricted in old
process of production. There is less scope formed for new techniques and low
technological improvement has happened in the agricultural sector. The larger
household size has a positive impact on TE. The average education of household is
positive and significant with TE in most of the regression models. We can refer to
endogenous growth models mentioned in literature review portion (Barro, 1990;
Romer, 1986, 1990, etc.) that education and knowledge through learning by doing
method are the factors that positively influence the economic performance through
enhancing labour productivity. However, it questions on adoption of technological
innovations as farm sector suffers from loss of technological ‘lock-in’ due to lack
of information about the technology and proper infrastructure to adopt the tools in
due time. Technological ‘lock-in’ has been examined and discussed in empirical
studies since the mid-1980s (Arthur, 1989; Cowan, 1990; David, 1985; Liebowitz
& Margolis, 1995). Our results also revealed such kind of technological ‘lock-in’
for Indian agricultural production. Policy-level technological trajectories with
proper design of implementation can help the technological progress in agriculture.
38 South Asia Economic Journal 19(1)

The coefficient of the proportion of family supervision to total labour cost is posi-
tively significant. This ensures that the family supervision is technically efficient
and the family supervisions have a greater role in the production process. The
results find that the distance to ‘pucca’ road negatively influences the TE.
Infrastructure development can make a positive impact on productivity. That
means if the household is situated in a remote area then the TE will be declined.
The AES, such as demonstration, film, exhibition, and lecture about the agriculture
production, have significantly increased the household efficiencies. This reveals
the local government should undertake more AES activities in villages.
The most important determinant of TE is the different irrigation sources. The
models have taken the proportion of area irrigated by government canal, tank
waters and open well waters as determinates of TE. The results find that all the
sources have made a significant contribution to agricultural production and the TE
have increased. The results have shown that if the proportion of HYV area of a
village increases then it tends to increase the TE of farmers. The expenditure on
agricultural programmes by the government shows that TE has increased due to
more agricultural expenditures by the government. These results suggest that the
government should increase the agricultural expenditures to push up the farmers
to increase their efficiencies. The results find that the women headed panchayats
are doing better and the TE of these villages has increased over the periods.

Conclusion
The conclusion of the empirical study has clearly emerged with some interesting
outcomes along with some usual relationships following the norms. There exists
an inverse relationship between cropped area and output. The share of irrigated
area is positively associated with crop output. This outcome indicates that higher
yields come from a larger irrigated cropped area of the households. Use of proper
seeds is an important factor in a production process and it increases the output.
The fertilizers are more important determinant compare to manure. Further, the
results showed that the hired labour is more productive compared to other inputs.
The results have shown that the coefficients of the labour inputs are larger than
other inputs and this reveals that mostly Indian agriculture is labour intensive.
On an average, the panel households are more technically efficient than all
sample households. It suggests that mostly the panel households have the ability
to achieve the maximum output with given and obtainable technology and also
they use the inputs optimally. Infrastructure is an exogenous factor for agricultural
productivity. The remoteness of villages or distance from the ‘pucca’ road is
inversely related to the TE. The irrigation facilities are important factors in the
production which increase the technical efficiencies of the agricultural farmers.
There are some other household-based characteristics which are also contributing
significantly to farm productivities, such as family size, age of the household,
education in terms of years of schooling of the respondent, etc. It has been clear
that aged households are more productive. It is mainly because they have more
Pradhan and Mukherjee 39

experience in the production process to produce more output using optimal inputs
efficiently as learning by doing method is the option rather than adopting new
technologies, given the infrastructure. Technological lock-in is pervasive in Indian
agricultural sector. Further, education of the household is an important factor in
the production process that is why TE has been increased for educated farmers.
More informed farmers are less locked in older process. Therefore, it suggests that
education enhances skills and innovations which are useful in terms of the alloca-
tion of inputs in a rapidly changing technological environment and then it
increases the farm efficiency. The larger family sizes are doing significantly better
off and they are more technically efficient. Governance has some role in this
regard as we cannot ignore the positive influence of welfare-based programmes in
rural development. Farmers avail the benefit provided by government through
local rural bodies. Our results have provided evidence that participation in such
welfare-based programmes have a positive influence in determining agricultural
progress. The local government expenditures on agricultural programmes have a
positive impact on the productivity of the farmers. Last but not the least, results
found that the women reserved panchayats have a positive impact on TE of the
farmers in that areas. It indicates that women-reserved panchayats not only handle
the delivery of public programmes efficiency but also generates governance divi-
dends through making proper welfare generation.

Acknowledgements
Authors are greatly indebted to H. K. Nagarajan, RBI Chair Professor, IRMA, Gujarat and
Shashanka Bhide, Director-MIDS, Chennai for the support and supervision during the study.
Authors are also thankful to NCAER, New Delhi and IDRC, Canada for giving opportunity
to get involved in ‘Decentralization and Rural Governance in India’ project and for providing
the detailed primary data for the study. Authors are grateful to anonymous referee of the
journal for comments on the earlier version of the article. Usual disclaimers apply.

Notes
1. In view of the structural change in the economy, there has been a continuous decline in
the share of agriculture and allied sector in the GVA from 18.5 per cent in 2011–2012
to 17.4 per cent in 2014–2015 at current prices. A fall in the share of the agriculture and
allied sector in GVA is an expected outcome in a fast growing and structurally changing
economy. Refer, Annual Report (2015–2016), Department of Agriculture, Cooperation
& Farmers Welfare, Government of India.
2. Tripathi (2008) has showed after examining the performance of agricultural productiv-
ity in India for the period over 1969 and 2005 that agricultural growth relied almost
entirely on increased in conventional factors while growth in productivity was negative.
Further, the study suggested that the relative decline in public investments in the agri-
cultural sector is one of the prominent causes of slowdown of agricultural productivity
growth.
3. Rostow’s defined stages of growth clearly refer a dynamic role for the agricultural sec-
tor in the transition process. Agriculture must (a) provide food for a rapidly increasing
population; (b) provide a mass market for the products of the emerging industrial sec-
tors and (e) generate the capital investment for new leading sectors outside of agricul-
ture (Rostow, 1960).
40 South Asia Economic Journal 19(1)

4. The data collection for the last round of the REDS survey started in 2006, which is why
it is normally referred to as the 2006 round. However, for most states the household
schedule that contains the agricultural data was collected in 2007, with the exception of
Kerala, where it was collected in 2008.

References
Afridi, F. (2017). Governance and public service delivery in India (IZA Discussion Paper
No. 10856). Bonn: IZA Institute for Labor Economics.
Afridi, F., Iversen, V., & Sharan, M. R. (2013). Women political leaders, corruption and
learning: Evidence from a large public program in India (IZA Discussion Paper No.
7212). Bonn: IZA Institute for Labor Economics.
Aigner, D. J., Lovell, C. A. K., & Schmidt, P. (1992). Formulation and estimation of
stochastic frontier production models. Journal of Econometrics, 6(1), 21–32.
Armagan, G. (2008). Determining the factors affecting efficiency scores in agriculture.
International Journal of Agricultural Research, 3(4), 325–330.
Arrow, K. J. (1962). The economic implications of learning by doing. Review of Economic
Studies, 29(3), 155–173.
Arthur, W. B. (1989). Competing technologies, increasing returns and lock-in by historical
events. The Economic Journal, 99(394), 116–131.
Barro, R. J. (1990). Government spending in a simple model of endogenous growth.
Journal of Political Economy, 98(5), S103–S125.
———. (1995). Determinants of economic growth. London: The MIT Press.
Battese, G. E., & Coelli, T. J. (1988). Prediction of firm-level technical efficiencies: With
a generalized frontier production function and panel data. Journal of Econometrics,
38(3), 387–399.
———. (1992). Frontier Production Functions, Technical Efficiency and Panel Data with
Application to Paddy Fanners In India. Journal of Productivity Analysis 3, 153–169.
———. (1995). A Model for Technical Inefficiency Effects in a Stochastic Frontier
Production Function for Panel Data Empirical Economics, 20, 325–332.
Battese, G. E., Coelli, T. J., & Colby, T. C. (1989). Estimation of frontier production
functions and the efficiencies of Indian farms using panel data from ICRISAT’s village
level studies. Journal of Quantitative Economics, 5(2), 327–348.
Baumann, M., Kuemmerle, T., Elbakidze, M., Ozdogana, M., Radelof, V. C., Keuler,
N., & Prishchepova, A. V. (2011). Patterns and drivers of post-socialist farmland
abandonment in Western Ukraine, Land Use Policy, 28(3), 552–562.
Becker, G. (1964). Human capital (2nd ed.). New York, NY: Columbia University Press.
Constantin, P. D., Martin, D. L., & Rivera, E. B. B. R. (2009). Cobb–Douglas, translog
stochastic production function and data envelopment analysis in total factor productivity
in Brazilian agribusiness. Journal of Operations and Supply Chain Management, 2(2),
20–34.
Cowan, R. (1990). Nuclear power reactors: A study in technological lock-in. The Journal
of Economic History, 50(3), 541–567.
David, P. A. (1985). Clio and the economics of QWERTY. American Economic Review,
Papers and Proceedings, 75(2), 332–337.
Domar, E. D. (1946). Capital expansion, rate of growth, and employment. Econometrica,
14(2), 137–147.
Foster, A. D., & Rosenzweig, M. R. (2003). Agricultural productivity growth, rural
economic diversity, and economic reforms: India, 1970–2000. Economic Development
and Cultural Change, 52(3), 509–542.
Pradhan and Mukherjee 41

Gabre-Madhin, E., & Johnston, B. (2002). Accelerating Africa’s structural transformation:

Lessons from Asia. In T. S. Jayne, M. Isaac, & A-K. Gem (Eds), Perspectives on
agricultural transformation: A view from Africa. New York, NY: Nova Science.
Government of India (GoI). (2015). Annual report Indian Council of Agricultural Research
(ICAR) 2014–15. New Delhi: Department of Agricultural Research and Education
(DARE), Ministry of Agriculture.
———. (2017). Economic survey 2016–17. New Delhi: Ministry of Finance, Department
of Economic Affairs, Economic Division.
Hanmer, L., & Nashchold, F. (2000). Attaining the international development targets: Will
growth be enough? Development Policy Review, 18(1), 11–36.
Harrod, R. F. (1939). An essay in dynamic theory. The Economic Journal, 49(193), 14–33.
Hasan, R., & Quibria, M. G. (2004). Industry matters for poverty: A critique of agricultural
fundamentalism. Kyklos, 57(2), 253–264.
Hughes, B. B., Dickson, J. R., & Irfan, M. T. (2008). Patterns of potential human progress.
New Delhi: Oxford University Press.
Johnston, B. F., & Kilby, P. (1975). Agriculture and structural transformation: Economic
strategies in late-developing countries. New York, NY: Oxford University Press.
Kuemmerle, T., Hostert, P., Radeloff, V. C., van der Linden, S., Perzanowski, K., &
Kruhlov, I. (2008). Cross-border comparison of post-socialist farmland abandonment
in the Carpathian. Ecosystems, 11(11), 614–628.
Kyi, T., & Matthias von, O. (1999). Stochastic frontier production function and technical
efficiency estimation: A case study on irrigated rice in Myanmar. Retrieved from ftp://
ftp.gwdg.de/pub/tropentag/proceedings/1999/referate/STD_C6.pdf
Lewis, W. A. (1954). Economic development with unlimited supplies of labour. The
Manchester School, 22(2), 139–191.
Liebowitz, S., & Margolis, S. E. (1995). Path dependence, lock-in and history. Journal of
Law, Economics and Organization, 11(1), 205–226.
Limam, Y. R., & Miller, S. M. (2004). Explaining economic growth: Factor accumulation,
total factor productivity growth, and production efficiency improvement (Working Paper
No. 2004–20). Mansfield, CT: Department of Economics, University of Connecticut.
Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary
Economics, 22(1), 3–42.
Meeusen, W., & Van den Broeck, J. (1997). Efficiency estimation from Cobb–Douglas
production functions with composed error. International Economic Review, 18, 435–
444.
Mellor, J. W. (1976). The new economics of growth: A strategy for India and the developing
world. Ithaca, NY: Cornell University Press.
O’Donnell, C., Rao, D., & Battese, G. (2008). Metafrontier frameworks for the study of
firm-level efficiencies and technology ratios. Empirical Economics, 34(2), 231–255.
Planning Commission. (2008). Agriculture, Rural Development, Industry, Services and
Physical Infrastructure. Eleventh Five Year Plan 2007–2012, Vol. 3. New Delhi:
Government of India.
Rahman, S. (2003). Profit efficiency among Bangladesh rice farmers. Food Policy, 28(5–
6), 487–503.
Ranis, G., & Fei, J. C. H. (1961). A theory of economic development. American Economic
Review, 51(4), 533–565.
Rao, E. J. O., Brümmer, B., & Qaim, M. (2012). Farmer participation in supermarket
channels, production technology, and efficiency: The case of vegetables in Kenya.
American Journal of Agricultural Economics, 94(4), 891–912.
42 South Asia Economic Journal 19(1)

Rao, K. N. (2006). Risk management as a pillar in agriculture and food security policies:
India case study (Policy Brief, EASYPol Module 209 Louis BOCKEL Agriculture
Policy Support Service, Policy Assistance and Resource Mobilization Division).
Rome: FAO.
Rivera-Batiz, L. A., & Romer, P. M. (1991). Economic integration and endogenous growth.
Quarterly Journal of Economics, 106(2), 531–555.
Romer, P. (1986). Increasing returns and long-run growth. Journal of Political Economy,
94(5), 1002–1037.
———. (1987). Crazy explanations for the productivity slowdown. NBER Macroeconomics
Annual, 2, 163–210.
———. (1989). Human capital and growth: Theory and evidence (NBER Working Paper
No. 3173). Cambridge. MA: NBER.
———. (1990). Endogenous technological change. Journal of Political Economy, 98(5),
S71–S102.
Rostow, W. W. (1960). The stages of economic growth: A non-communistic manifesto.
Cambridge: Cambridge University Press.
Schmidt, P. (1976). On the Statistical Estimation of Parametric Frontier Production
Functions. Review of Economics and Statistics, 58, 238–239.
Schultz, T. W. (1961). Investment in human capital. American Economic Review, 51(1),
1–17.
Senhadji, A. (2000). Sources of economic growth: An extensive growth accountancy
exercise. IMF Staff Papers, 47(1), 129–158.
Shanmugam, K. R. (2000). Technical efficiency of rice growers in Bihar. Indian Journal of
Applied Economics, 8, 377–389.
———. (2002). Technical efficiency of growing rice crop in Karnataka: A panel data
study. Artha Vijnana, 44(3, 4), 213–224.
Shanmugam, K. R., & Venkataramani, A. (2006). Technical efficiency in agricultural
production and its determinants: An exploratory study at the district level. Indian
Journal of Agricultural Economics, 61(2), 169–184.
Shiva, V. (1991). The green revolution in the Punjab. The Ecologist, 21(2), 57–60.
Solow, R. M. (1956). A contribution to the theory of economic growth. Quarterly Journal
of Economics, 70(1), 65–94.
———. (1957). Technical change and the aggregate production function. The Review of
Economics and Statistics, 39(3), 312–320.
Tauer, L. (2001). Efficiency and competitiveness of small New York dairy farm. Journal of
Dairy Science, 84(11), 2573–2576.
Thiam, A., Bravo-Ureta, B. E., & Rivas, T. E. (2001). Technical efficiency in developing
country agriculture: A meta-analysis. Agricultural Economics, 25(2–3), 235–243.
Tripathi, A. K. (2008). Agricultural price policy, output, and farm profitability—
Examining linkages during post-reform period in India. Asian Journal of Agriculture
and Development, 10(1), 91–111.
Zhu, J. (2000). Multi-factor performance measure model with an application to fortune 500
companies. European Journal of Operational Research, 123, 105–124.