Content uploaded by Jeetendra Prakash Aryal
Author content
All content in this area was uploaded by Jeetendra Prakash Aryal on Sep 07, 2022
Content may be subject to copyright.
Page 1 of 37
Cite as:
Aryal J. P., Holden S. T., Bhattarai M., and Rahut D. B., (2022) Exploration of farm
mechanization among smallholder farmers: Evidence from Indian states of Haryana and Bihar.
Paper presented at the Asian Econmic Development Conference 2022, under the theme
‘Inclusive, sustainable recovery in Asia and the Pacific, 14-15 July 2022, organized by the Asian
Development Bank, in Tokyo, Japan. https://www.adb.org/news/events/asian-economic-
development-conference-2022#accordion-1-0
Page 2 of 37
Exploration of farm mechanization among smallholder
farmers: Evidence from Indian states of Haryana and Bihar
Jeetendra Prakash Aryal1*, Stein T. Holden2, Madhusudan Bhattarai3, and Dil Bahadur Rahut4
1International Maize and Wheat Improvement Center (CIMMYT), El Batan, Mexico
2Norwegian University of Life Sciences (UMB), Aas, Norway
3International Food Policy Research Institute (IFPRI), New Delhi, India
4Asian Development Bank Institute (ADBI), Tokyo, Japan
Abstract
Intensification of agriculture is crucial for food security in India, and farm mechanization offers
a way to achieve it. However, farm mechanization depends on several factors, and thus,
understanding those factors is crucial for enhancing its uptake to improve the farmers’ wellbeing
and also, for designing appropriate policies. Using data collected from 1267 farm households of
Haryana (626 households) and Bihar (641 households) states of India, this study assessed factors
associated with the adoption of farm machineries, including irrigation pump, tractor, harvester,
thresher, and other farm machines. Considering that farmers’ mostly use combinations of these
machines in farm operations, we applied a multivariate probit model for the analysis. Results
show that adoption decisions of these farm machines are interrelated and mostly complementary.
Economic assets such as land, livestock, and improved output/input market, and credit accesses
are positively associated with the likelihood to adopt farm machines. Farmers with off-farm
income are more likely to adopt harvester and other machines compared to those without off-
farm income, confirming that high opportunity costs of labor enhance the adoption of machinery.
Another key finding is that machine ownership is not necessary for the adoption of any of these
machines as farmers mostly hire these services, thus reducing the capital requirement for
Page 3 of 37
adoption and enhancing rural transformation. Therefore, promoting farm machinery hiring
services and enhancing farmers access to credit are key areas of policy reform.
Keywords: Farm mechanization; technology adoption; irrigation pump; tractor; thresher;
harvester; rental service; India.
JEL codes: Q13, Q16, Q18
1. Introduction
Agriculture production in India needs to feed 17.5% of the global population that is holding only
2.4% of the land and 4% of the water resources at its disposal (NAAS, 2013). By 2050, the
population of India will be 1.75 billion (http://www.populstat.info/Asia/indiac.htm), and given
the current shift in consumption trend and economic growth, food demand in India is expected to
double by 2050 (FAO, 2014).Per capita arable land in India declinedto 0.12 ha in 2013 from 0.34
ha in 1961 (World Bank, 2015). With a decline in per capita arable land, there is almost no
possibility to extend the agricultural land frontier in India. Still, agriculture is the leading source
of livelihood; 70% of the rural households depend primarily on agriculture for livelihood, with
82% of farmers being small and marginal(FAO, 2018a). Intensification of agricultural systems,
therefore, remains the only option viable to increase production with no additional land because
the competition for land and water is increasing from non-farm sectors (NAAS, 2013). Increase
in population coupled with economic growth can further add pressure on food production
system, indicating a need for intensification of agriculture to secure food security.
Intensification of the agricultural production system is mostly power-intensive and is amplified
by the timeliness of the agricultural operations. By reducing the power bottlenecks in agricultural
Page 4 of 37
operations, mechanization contributes to intensify the land use and to increase overall food
production (Pingali, 2007).However, mechanization does not necessarily imply intensification in
terms of increasing productivity. In many cases, it has been associated with substituting capital
for labor due to increasing wage rates (Ma et al., 2018; Wang et al., 2016).
Rural labor wages have been increasing across India over the last decade. Out-migration of the
population from the rural to the urban sector hasbeen taking place across India at a much faster
pace than before due to higher wages and secure income (Chander and Thangavelu, 2004). Such
a situation has given impetus to Indian farmers for the use oflabor-saving farm machinery and to
improving efficiency and profit from crop production activities. These changes in the national
and sectoral economy have produced other far-reaching changes in the agricultural sector, which
have also affected the level and process of farmers’ demand for farm machinery.
On this pretext, farm mechanization in India needs a careful assessment and documentation,
given the institutional constraints such as the caste system and skewed land distribution in India.
As stated by (FAO, 2018b), mechanization contributes to farming in developing countries in
several ways: firstly, it helps farm households cope with the labour shortage resulting from
migration and demand from other non-farm sectors; secondly, it helps in adaptation/mitigation to
climate change when combined with conservation agriculture practices, and thirdly it helps in
poverty reduction and ensuring food security. Hence, mechanization can play an important role
in enhancing farm productivity and income, thereby ensuring food security and reducing poverty
in developing countries. Therefore, understanding the drivers of adoption/use of farm machinery
in agriculture by poor smallholder farmers in developing countries may help policymakers in
scaling up the mechanization. Most of the previous studies have focused on the ownership of
Page 5 of 37
farm machinery as the mechanism for adoption (Mottaleb et al., 2016). In this study we have
investigated the existence and extent of hiring machinery from service providers; thus, the
current research explores the factors that drive the use of machinery for farming, including own
machinery as well hired services. Rental services and markets may reduce the capital constraint
associated with adoption of machinery and thereby enhance rural transformation.
2. Review of literature
There is no uniform definition of the mechanization; it varies across the region and the context.
In operational perspective, farm mechanization is considered as a process of improving farm
labor productivity or land productivity through the uses of improved agricultural machinery, or
implements for various operations in the agricultural field to enhance efficiency in operation
(Singh, 2015). These machines can be driven either manually, or through animal power, or
through machine power. As the farming system in a place evolves, the power sources use in
farming also gradually shift from human muscle power to animal power to machine power. This
process is affected by several internal and external factors of farm households. Not all of them
are within the control of farm households, but the farmers have to adapt as per the changing
context.
As in other land-scare economies, the irrigation water pumping operation is the first agricultural
operation mechanized in India. Irrigation pumps contribute to regulate the supply of water
needed for double or triple cropping and also increase the possibility of cultivating the marginal
lands (Pingali, 2007).
Mechanization is an appropriate way to enhance land productivity and to acquire intensification
of agriculture (Sims et al., 2016). Since mid-sixties, mechanization and overall technological
Page 6 of 37
improvements have massively increased agricultural production in India, transforming the nation
from a situation of acute food shortage to food surplus economy (Singh, 2006). However,
different countries may have different pathways to farm mechanization and thus, is an uneven
process (Binswanger, 1986). The incidence of mechanization varies largely across regions or
states in India (Biggs and Justice, 2011). Level of mechanization is higher in the states like
Punjab and Haryana, while it is less in eastern Uttar Pradesh (UP), Bihar, West Bengal, and
Orissa (Biggs and Justice, 2011). Past policies in India emphasized on large equipment, although
it was not appropriate for all farmers, especially the smallholders. Contrary to the general
hypothesis that labor scarcity cannot be a driving force for mechanization in a country with a
large proportion of labor (Binswanger, 1986), there is now evidences that increasing rural to
urban migration is the major cause of labor shortage in peak agricultural seasons, increasing the
scope for mechanization (Zhang et al., 2014). Increasing cost of manual labor and peak time
labor shortages are the major reasons for using tractors, laser land leveler, and zero-tillage drill
machines by farmers in Haryana and Punjab states of India (Aryal et al., 2015b; Aryal et al.,
2015c). Use of tractors and other machineries by Indian farmers exhibit this. In addition,
mechanization reduces the drudgery of the farm laborers, relax peak-season labor constraints,
reduce costs of production, and in some cases, save crucial resources such as energy and water
(Aryal et al., 2016; Biggs and Justice, 2015) and help adapt to climate change (Aryal et al. 2016).
Enhancing farm mechanization may also attract youth entrepreneurs in the agriculture sector and
also reduce the burden of women farmers (Baudron et al., 2015). Some forms of mechanization
contribute to mitigating greenhouse gas emissions from agriculture such as the use of the zero-till
drill for wheat production (Aryal et al., 2015c), while it depends on the type of mechanization
(Aguilera et al., 2019).
Page 7 of 37
Increasing use of machine power, especially tractor and water pumps, is considered a key factor
in wide-spread adoption of the rice-wheat system in the irrigated region of IGP of South Asia.
This is further facilitated by the custom hiring services as it improved smallholder farmers’
access to agricultural machinery services (Aryal et al., 2015a; Aryal et al., 2019; Aryal et al.,
2021; Biggs and Justice, 2015; Erenstein and Farooq, 2009). In India, many farmers or farm
cooperatives hire out the machines for extra income (Erenstein and Farooq, 2009).This leads to a
situation that farmers do not need to invest in purchasing farm machinery. As a result, ownership
of farm machines is no longer necessary for their adoption. Another major issue is that the
adoption of a machine requires not onlycost but also the skill to operate it. Custom hiring
services is thus a better alternative because it not only provides farmers with the appropriate
machine but also provides the skilled operator, thereby relieving a major constraint to adoption.
It is because farmers need to learn the way to operate the machine properly. In this context,
defining the level of adoption of these machines in terms of their ownership largely
underestimates their level of adoption by farmers. Therefore, using data collected from 1267
farm households in India, this paper assesses the factors determining the adoption of the most
common small-scale machineries used in farming, including power tillers, irrigation pumps, and
threshers. In addition, we also tested the hypothesis that defining adoption in terms of machine
ownership underestimates the adoption of machine by farmers.
This paper contributes in three ways: first, unlike the previous studies (for example, (Mottaleb et
al., 2016; Mottaleb et al., 2017), we focused on the use of machines as adoption rather than
ownership of the machines. Thus, we do not undermine the role of custom hiring services in the
adoption of agricultural machineries and underestimate the rate of adoption. Second, comparing
the percentage of households that have used agricultural machinery and households that own
Page 8 of 37
them, we tested whether or not ownership is really required for its adoption. Third, we
acknowledge the interdependence across the decisions to adopt farm machines and jointly
analyze them. Fourth, we provide comparisons between the two states of India, Bihar, and
Haryana with regard to mechanization.
3. Agricultural mechanization and its pattern in India with a
focus on Haryana and Bihar states
Mechanization in India began before its independence. In the mid-1940s, the country established
a Central Tractor Organization as well as a State Tractor Organization, which for the first time
imported tractors and bulldozers in the country (Singh, 2006). Use of tractors by Indian farmers
increased sharply from 8000 units in 1950 to 37,000 units in 1960. Following the import
substitution policy in manufacturing sector, five tractor manufacturing units were set up in India
in 1960 (Singh, 2011). By 2000, the number of tractors increased to 9 units per 1000 hectares
from 0.19 units per 1000 ha in 1961. Likewise, the number of irrigation pumps reached 430000
units in 1960 to 1950000 units in 1961(Singh, 2011). Green Revolution massively contributed to
promote farm mechanization in India. Table 1 shows the number of tractors, drought animals,
power tillers over time.
[Table 1 here]
With the evolution of farming system, power sources used in agriculture has shifted gradually
from human muscle power to animal power and then to machine power. As a result, the total
number of draft animals (animal power used in farming) has been declining over time in India,
whereas the total number of tractors increased sharply (Table 1). The sharp rise in the use of
tractors, irrigation pumps, and other farm machinery in the early 1970s is due to success in the
Page 9 of 37
adoption and diffusion of Green Revolution technologies (especially short duration high yield
varieties (HYVs) of wheat and Paddy). These HYVs allowed farmers to grow wheat after rice in
the same piece of land. However, it required timely preparation of land for growing wheat
immediately after harvesting paddy, and thus, farmers need to complete all the required farm
operations in a short-span of time between the two crops. Consequently, the use of draft animals
alone is not possible for such activities, which paved the wayto the increased adoption of farm
machines such as tractors and threshers in the early 1970s. This is one of the fundamental factors
for large-scale adoption of the tractor in Punjab, Haryana, and Western UP, the regions where
Green Revolution showed its success in India. Low-interest rate credit policy of Indian
government in 1970-80s also gave impetus to purchase agricultural machinery.
The increase in the use of farm machines in the 1990sis associated with the large investment in
basic infrastructures such as roads. Additionally, liberalization of the Indian economy in
1992/93 has further accelerated farm mechanization. After economic liberalization, power tillers
and tractors with an engine displacement of less than 1800cc were exempted from excise duty,
and the manufacturers were exemptedfrom permit requirements for producing such small
machineries in the country. All these policies contributed to make tractor industry in India one of
the largest in the world, which sells about 700,000 tractors per annum, equivalent to US$ 4
billion (Bhattarai et al., 2016). Table 2 presents the density of tractors and water pumps in
Haryana, Bihar, and India.
[Table 2 here]
Farm machinery such as threshers, tractors, tractor-drawn implements, and combine harvesters
were first adopted in Haryana and Punjab states of India starting in the 1950s and 1960s. Then
these machineriesgradually penetrated to other parts of India. Still, a huge variation on the
Page 10 of 37
density of farm machinery uses (tractor and irrigation pump as a proxy for other machines) is
observed across the states. Agricultural policies of the state government, along with farming
systems and agro-climatic factors, largely explain the variation in farm mechanization across the
states of India. Among the two states (Bihar and Haryana states) selected for study, the density
of tractor in Bihar state is 1/4th of that of Haryana state whilst pump density is nearly 2/3rdof that
of Haryana.
Farmers’ adoption of agricultural machineries in a place evolves in a sequential manner, which is
affected by farming systems adopted in a place, and by several other farming and non-farm
sector factors. Usually, farm mechanization began with high power-intensive operation, which
also requires low-skill of labour work such as the use of mechanical thresher, tractor for
ploughing land, or use of pumps for irrigation purpose (see Table 3). Then, this was followed by
medium power and medium skill requiring farm operations such as seed sowing by machine,
harvesting of grain, etc. Then, in the third stage of evolution, farm mechanization occurred for
those farming operations that require low power but high skill-based farming operation such as
for transplanting of paddy, grinding by quality, harvesting/picking of cotton, harvesting of
vegetables, and harvesting of sugarcane. For details on the sequence of mechanization across
farming systems in Asia and Africa, see (Pingali, 2007; Singh, 2011). The same process and
pattern of adoption of farm machinery technologies are taking place across states of India as
well. Looking at the history, it is found that steam threshers were imported in 1912 in India
(Punjab province), tractors were imported first time in around 1914 and then, first-time irrigation
pumps were imported in the 1940s (Singh, 2015). Over time, these farm machines have
penetrated deeply across the Indian states.
[Table 3 here]
Page 11 of 37
The sequence of adoption of farm machinery is important in our study since we have taken
ranges of farm machinery for assessment (tractor, thresher, water pump, zero tillage, and
harvester), and two different states with different pace (and stage) of farm machinery uses1. The
diffusion process and pattern of farm machinery uses greatly vary across regions and states of
India. Examination of these issues at the farm household level is one of the objectives of this
empirical study and field survey in two regions with the heterogeneous and diverse structure of
farm households in the two states of India. Both the states of Haryana and Bihar are located in
the irrigated region of Indo-Gangetic basin, with a predominantly followed rice and wheat
production system. Moreover, in a team of technology adoption and agrarian development
pathways, each of the survey locations (Karnal district) in Haryana is at a different stage of
development pathway than that of the survey location (Vaishali district) in Bihar state.
4. Conceptual Framework
Figure 1 depicts the conceptual framework on mechanization. Mechanization levels are growing
among smallholders in developing countries, indicating increased incentives and capacity to
adopt scale-appropriate mechanization, particularly through service provision (Adu-Baffour et
al., 2019; Aryal et al., 2019; Aryal et al., 2021; Diao et al., 2014; Mottaleb et al., 2017;
Yamauchi, 2016; Zhang et al., 2014).Farm mechanization is gaining its popularity in developing
countries for three main reasons: (1) growth and increase in productivity of the agricultural
sector; (2) shortage of labour and increasing wages; (3) reduced time for land preparation and
providing surplus labour for other activities; (4) easy to practice conservation agriculture.
1For field survey, we have taken to two different places of India: 13 villages from Haryana state (part of older
Punjab state, and 12 villages from Bihar state. Crop productivity and machine density are much higher in Haryana
state than in Bihar. The rural labor density of crop acreage in Bihar is also almost three times higher than the case in
Haryana. Thus, we expect spatial and over times variation on machinery uses in the study sites. The econometric
models are selected considering this heterogeneity in machine adoptions.
Page 12 of 37
[Figure 1 here]
Mechanization drives the growth and development of the agriculture sector and also enhances
the wellbeing of the farmers dependent on farming for livelihood (Kienzle et al., 2013; Pingali,
2007). Mechanization contributes to improving the livelihood by reducing the cost of production,
improving efficiency and productivity (Benin, 2015; Kienzle et al., 2013; Pingali, 2007; Sims et
al., 2016).
Youth and educated people are leaving agriculture in search of better job opportunities
(Maharjan et al., 2013b), leading to acute labor shortages in the agriculture sector that have
affected timely crop establishment and other crop cultivation practices (Maharjan et al., 2013a,
b), which has contributed to rising rural wage rates (Wang et al., 2016; Wiggins and Keats,
2014). Rising rural wage rates and labour shortage has led to an increase in fallow land, low-crop
productivity, and high food insecurity (Khanal, 2018; Khanal et al., 2015; Maharjan et al., 2013a;
Prabakar et al., 2011; Yamauchi, 2016).
As a result of mechanization, farmers can complete the land preparation and other farming
activities quickly, thereby providing surplus labour, which could be used for leisure and in other
more productive nonfarm activities. Mechanization results intoincrease inlabour supply to the
non-farm sector likelihood of involvement in non-farm activities (Ahmed and Goodwin, 2016).
Althoughconservation agriculture enhances farm productivity, reduces pest and diseases and
water usage, and mitigate the adverse effect of climate change, the small single-axle tractor is
crucial is critical farm machinery for practicing conservation agriculture by the smallholder
farmers (Sims and Kienzle, 2006).
Page 13 of 37
Studies have found that farm size is an important factorthat determines farmer's ability to take up
mechanization (Ghosh, 2010; Jha et al., 2000; Jha and Rhodes, 1999; Van den Berg et al., 2007).
Hence land consolidation and market integration also contribute to mechanization (Fan and
Chan‐Kang, 2005; Hazell, 2005; Rigg et al., 2016; Xiaobing et al., 2016; Yang et al., 2013)(Fan
and Chan-Kang, 2005; Hazell, 2005; Otsuka et al., 2016; Rigg et al., 2016; Xiaobing et al., 2016;
Yang et al., 2013).
5. Study area, data, and empirical methods
5.1 Study area, sampling methods, and data description
We used farm household survey data that was collectedin 2013 from1267 farm households in
Karnal (Haryana state) and Vaishali districts (Bihar state) of India. Individual sample units (i.e.,
household) was selected using amulti-stage sampling method. First, two districts (one from
Haryana and another from Bihar) were purposefully chosen based on climatic vulnerabilities.
Second, 12 villages from Vaishali district and 13 villages from Karnal district were selected as
study villages. Third, individual sample households were selected randomly. In order to
comprehend the general characteristics of the households in each study village, we carried out a
village census of about 75% of the total households of the village. Table 4 presents the
distribution of sample households in the study area, and figure 2 shows the location of the study
area.
[Table 4 here]
[Figure 2 here]
Page 14 of 37
We applied a structured questionnaire to gather the data, which consist of information on
household characteristics (i.e.,age, education, family size, gender of household head, migration),
economic and social capital relatedvariables (i.e., land holdings, livestock owned, ownership and
use of machines in farm operations such as thresher, tractor, harvester, irrigation pump and any
other machines, credit access, membership in farmers group, and food security status of the
household) and market access and training related variables.
5.2 Model specification and estimation issues
We assess the factors influencing the decision to adopt a machine for agricultural operation; our
variable of interest is discrete in nature. As we consider five interrelated farm machineries –
tractor, irrigation pump, thresher, harvester, and other, applying univariate logit and probit
models for each of these machines may become not appropriate due to the possibility of the
dependence of error terms of the different machines used for agricultural operations. Indeed, a
farmer may adopt a mix of these machines under study; and the decision to adopt one machine
could be influenced by the adoption decisions of another. Acknowledging the possibility of such
interrelationships, we applied a multivariateprobit model. Ignoring the potential correlation
among the unobserved disturbances across the decision to adopt equations leads to biased and
inefficient estimates (Greene, 2003; Wooldridge, 2002). In some cases, we might observe limited
adoption instigated by the unavailability of a complementary set of machines. Hence effective
correction of these complementarities in the econometric model is crucial to address such issues
adequately. Given that many farmers in Bihar and Haryana use land intensively and grow two or
Page 15 of 37
more crops on the same field each year. In this case, labor peak for harvesting one crop blends
into the labor peak for land preparation of the next crop, leading to a situation where a farmer
takes decisions to use multiple types of machines simultaneously. In such a situation, access to
the farm machines needed for one operation may affect farmer’s decision to use machine for
subsequent operations. For example, if a farmer has no access to tractor, it may affect his
decisions to use thresher and or harvester. Similarly, when a farmer decides to use a combine
harvester, he does not require thresher. Another key issue is that in the process of collecting
information on one type of technology (here, machines), farmers usually gather information on
other related technologies. Due to this, the decision to adopt one type of technology (machine)
for agricultural operation may affect the decision to adopt another type of technology. Hence,
adoption decisions are intrinsically multivariate. Using univariate techniques in such a case could
largely undermine thevital information about interdependent and simultaneous adoption
decisions (Greene, 2003). The multivariate probit model helps us to determine possible
complementarities (positive correlation) and substitutability (negative correlation) between the
agricultural machineries used by the farmers.
A farmer is more likely to adopt a particular agricultural machine if the benefit from its adoption
is higher than non-adoption. Consider the th
i farm household ),....,2,1( Ni
facing a decision on
whether to adopt the th
jmachinery (where j denotes choice of: tractor (T), Irrigation pump (I),
thresher (S), harvester (H), and others (O). Let 0
Uand j
U represent the benefits to a farmer
without and with the adoption of agricultural machinery. A farmer decides to adopt the th
j
machinery if the net benefit )( *
ij
Bwith its adoption is higher than without its adoption, i.e.,
0
0
** UUB jij . In this case, the net benefit of adopting machines is a latent variable which is
Page 16 of 37
determined by observed household, location characteristics )( i
X, and the error term )( i
as
follows:
),,,,()1 '* OHSITjXB ipjiij
Equation (1) can be presentedin terms of the indicator function. In this case, the unobserved
preferences in equation (1) translate into the observed binary outcome equation for each machine
type as follows:
),,,,(
0
01
)2
*
OHSITj
otherwise
Bif
Bij
ij
In the multivariate probit model with the possibility to adopt multiple agricultural machines, the
error terms jointly follow a multivariate normal distribution (MVN) with zero conditional mean
and variance normalized to unity, i.e., ),0(),,,,( MVN
OHSIT uuuuu and the covariance matrix
)(
is given by:
1
1
1
1
1
)3
OHOSOIOT
HOHSHIHT
SOSHSIST
IOIHISIT
TOTHTSTI
Where
denotes the pairwise correlation coefficient of the error terms corresponding to any
two-machinery type adopted. Non-zero value of off-diagonal elements in the covariance matrix
indicates the correlations, justifying the use of multivariate probit instead of univariate probit for
each individual machine type.
Page 17 of 37
Multiple factors are associated with agricultural technology adoption. In order to obtain unbiased
estimates, we need to control for these factors. However,adding more explanatory variables in
the multivariate probit model can lead to the problem of multicollinearity and insufficient sample
size. To examine whether there is severe multicollinearity among the explanatory variables
included in the analysis, we used the condition index and excluded those variables from the
analysisfor which the value of the condition index was greater than 30 (Belsley, 1991; Belsley et
al., 2005). Estimators in the multivariate probit model are predicted based on asymptotic theory
and require a large sample size. Hence, sample size can be a constraining factor while estimating
a multivariateprobit model with many explanatory variables. If the sample size is less than
required, the asymptotic variance-covariance matrix is unlikely to be positive definite. This leads
to poor estimates of parameter variance-covariance and tobiased inference. Following previous
studies (Behera et al., 2015; Jöreskog and Sörbom, 1993), we applied the criterion that number
of observations (sample size) should be greater than 1.5k(k+1), where k refers to total number of
variables (both dependent and explanatory variables) used in the multivariate probit model.
5.3 Explanation of variables and hypotheses
Table 5 presents the descriptive statistics and definition of dependent and explanatory variables
used in the study.
[Table 5 here]
Page 18 of 37
Dependent variables
We have five dependent variables: tractor, irrigation pump, harvester, thresher, and other
machineries. Each of these is a binary variable that takes value one if a farmer decides to adopt it
and value 0 otherwise.
Explanatory variables
The empirical model and variables to be included in it are based on the theoretical framework
and the previous literature on technology adoption in agriculture (Aryal et al., 2014; Aryal and
Holden, 2012, 2013; Bortamuly and Goswami, 2015; Erenstein and Farooq, 2009; Hatibu et al.,
2013; Holden and Ghebru, 2016; Kassie et al., 2013; Kassie et al., 2014; Mottaleb et al., 2016;
Mottaleb et al., 2017; Pingali, 2007). We included household characteristics, socio-economic,
market access, access to agricultural extension services, and training on agriculture-related issues
as explanatory variables in estimating the multinomial probit model. Description of explanatory
variables and possible hypotheses about their effects on the dependent variable(s) are discussed
below.
Household characteristics
Household characteristics include the major characteristics of the household head (such as
literacy status, age, caste, and gender), household size, literacy status of spouse, and migration.
As rural farm households operate under multiple market imperfections and institutional failure,
household characteristics often influence technology adoption decisions (De Janvry et al., 1991;
Holden et al., 2001). Education increases individual’s ability to acquire, absorb and adopt new
information and technology (Chander and Thangavelu, 2004) and thus, literacy status of
Page 19 of 37
household head and spouse is supposed to increase the likelihood to adopt labor-saving
technologies such as tractor, irrigation pump, harvester, and thresher in agricultural operation.
Farm technology adoption is usually a part of an overall household strategy, indicating that
literacy status of the spouse of the household head is more likely to affect it. It is not sure how
the age of the household head affects the adoption of machines under study. Older household
head may go for labor-saving machines because of declining physical capacity, or they may
continue to prefer what they have been doing for long. Migration reduces household labor
endowment and thus, calls for labor-saving technology in agriculture. It also increases access to
alternative income sources required for hiring machines. Therefore, it is hypothesized that
migration enhances the adoption of labor-saving technologies such as power tiller and thresher.
Economic and social capital
Economic capital consists of land owned, livestock owned, and household labor endowment,
whereas social capital encompasses the membership in village institutions such as farm
cooperatives and caste position of the household. Availability of custom hiring service of
agricultural machines makesfarmersfree from the burden of purchasing machines, but it increases
the demand for immediate cash for hiring these services (Pingali, 2007). In rural communities of
India, caste continues to affect the access to public spheres (Birthal et al., 2015; Yamano et al.,
2015). It either restricts or facilitates a household’s participation in some markets and access to
information (Aryal and Holden, 2012; Birthal et al., 2015). In eastern Indian states, farmers
belonging to a scheduled caste group consider them as low profile in adopting new agricultural
technologies (Yamano et al., 2015). Therefore, we tested the hypothesis that compared to the
general caste group; farmers in the scheduled caste group are less likely to adopt farm machines.
However, this can vary between Bihar and Haryana due to their level of development. Access to
Page 20 of 37
credit and off-farm income can facilitate the adoption of agricultural machines as they reduce the
liquidity constraint of the farmer. Adoption of agricultural machines such as thresher can
improve farmers’ capacity to sell the produce on time and reduce the loss that may occur due to
price fluctuation.
Market, institutional services, and training
Access to input/output market and other institutional services influence transaction costs and
farmers’ knowledge and information about the machines. We consider distance to village market
as a proxy for market access and the distance to extension services as a proxy for access to
institutional services. Access to institutional extension services typically plays a crucial role to
enhance adoption (Chowdhury et al., 2014).Farmers’ participation in training on agriculture-
related issues also influences their decision to adopt labor-saving agricultural technologies. Land
rental market participation also matters in technology adoption (Holden and Ghebru, 2016). Often
farmers apply less of the costly technology in the rented in-land.
We also included number of plots owned as a proxy to land fragmentation. Farmers with more
number of small plots face difficulty while using larger machines such as 4WT and laser land
leveling.
We also included district dummies to control for spatial differences such as rainfall,
infrastructure quality of service delivery, and agro-ecological influences on farm households’
decision to adopt machinery.
Page 21 of 37
6. Results and discussion
6.1 Descriptive statistics
Table 5presents the descriptive statics of the variables used in the analysis. Tractors, irrigation
pumps, harvester, thresher, and other farm machines are the main farm machineries and are
adopted by 96%, 93%, 68%, 34%, and 34% of the total sample households, respectively. Almost
93% of the total sample households are male-headed, and the average age of household heads is
51years. On average, a household consists of 6 members. About 64% of household head and
32% of their spouses are literate. Average farm size of the total sample households is 1.38 ha;
however, this varies a lot between Karnal (3.33 ha) and Vaishali (0.41). Nearly 16% of the total
sample households are involved in some village institutions.
6.2 Testing whether ownership is required for adoption of machines under
study
Our data revealed that ownership of machine is much less than its use (Table 6). Only 20% of the
total sample households own a tractor, while almost 96% of them have used it. Similarly,
irrigation pump is used by almost 93% of the total sample household though only 46% reported
to own it. A large difference between the ownership and the use of machineis observed in the
case of thresher (i.e., 2% owner vs. 68% user). Harvester and other machineriesare used by 34%
of the total sample households and is owned by only 5 and 4% of them, respectively. It implies
that by defining adoption in terms of ownership of machine instead of its use (as in the case of
Page 22 of 37
(Mottaleb et al., 2016)2 largely underestimates the status of agricultural machinery adoption.
This is also supported by the other studies as (Aryal et al., 2015a) stated that the majority of
farmers using laser land leveling services do not own land laser leveler, but have used it through
custom hiring services. Similarly, Erenstein et al. (2008) reported that 60-70% of the zero till
adopting farmers use the machines through custom hiring services. Hence, as the market for
machines required for agricultural operations are emerging(Biggs and Justice, 2011), farmers
rely on custom hire services of these machines.
[Table 6 here]
6.3 Analysis of the determinants of adoption of machines under study
Prior to the analysis, we tested for multicollinearity among the explanatory variables using the
condition index method and dropped those variables with the condition index value of 30 or
more. Also, we checked if we have sufficient number of observations to carry out multivariate
probit in our case. As the total number of variables included in the model is 26, the number of
observations required is 1053 (Jöreskog and Sörbom, 1993). We have 1267 observations, and
thus, we do not face the problem of having such biased inference caused by poor estimates of
parameters of the variance-covariance matrix in the multivariate probit model.
Farmers adopt a number of machines for several types of agricultural operations. Some
operations are consecutive, and thus, decisions to adopt one machine may be correlated to the
decision to adopt another. We tested this using pairwise correlation across the error terms
corresponding to each two pairs of machines under study. Out of the 10 pairs, 6 pairs are found
to have highly significant correlations, thereby supporting the hypothesis that error terms of the
2 Mottaleb et al., (2016) defines ownership of machines as their adoption. Their study found that of the total
sampled households, 2.08% owns irrigation pump, 1.68% owns thresher, and 0.45% owns power tiller.
Page 23 of 37
multiple decision equations are correlated. Use of tractors and irrigation pumps are positively
and significantly associated, implying complementarities between them. Similar
complementarities are found between the use of tractor and harvester, and the use of tractor and
thresher. This is realistic because tractor is required to operate most of these machines under
study. There exists a significant negative association between harvester and thresher, indicating
substitutability between them. It might be due to the use of combined harvester by the farmers in
Haryana. In such a case, when farmers use a combined harvester, they do not require a thresher.
Table 5 presents the pairwise correlation coefficients across the error terms corresponding to any
two machines under study. The likelihood ratio test [chi2 (10) =201.77; Prob > chi2 = 0.000] is
highly significant and rejects the null hypothesis of zero covariance of the error terms across the
technology adoption decisions. Thisjustifies the application of a multivariateprobit model instead
of univariate probit models for each adoption equations.
[Table 7 here]
Table 8 presents the results of the multivariate probit model, which jointly assessed the factors
associated with the adoption of tractor, irrigation pump, thresher, harvester, and others. The
Wald test [Wald chi2 (105) = 884.31; Prob > 0.000) shows that the model fits the data well and
rejects the null hypothesis that all coefficients in each equation are jointly equal to zero. Indeed,
this shows the relevance of acknowledging the unobserved correlations across decisions to adopt
multiple machines for farm operations.
[Table 8 here]
General caste households are found to have significantly higher likelihood to use harvester and
other farm machines. Given that most of the large and informed farmers use harvester and other
Page 24 of 37
farm machines, including zero-tillage machine, general caste arein advantage compared to
backward caste groups, especially due to their better access to public resources and information.
Male-headed households are found to have a higher probability to adopt harvester and other
machines, while no significant difference is observed in the adoption of tractor, irrigation pump,
and thresher by gender of the household head. The results corroborate with the findings of
studies by (Mottaleb et al., 2016) in Bangladesh and (Aryal et al., 2015a) in north-west India.
Existing social norms in Bihar and Haryana limit the mobility of females in public space. As
hiring the machine requires frequent contact and negotiation with the male service providers,
adopting machinery through custom hiring service becomes problematic for female-headed
households compared to male-headed households.
Operated land holding has variable association with the likelihood of adopting machines. It is
significantly and positively associated with the likelihood of adopting harvester and other
machineries, is not associated with the use of tractor and irrigation pump, and is negatively
associated with thresher. One of the reasons behind it is that large farmers in Haryana use
combined harvester for timeliness of agricultural operations and also due to labor shortage
during the peak season. However, small farmers in Bihar usually use thresher for such
operations. Like land, livestock ownership has a variable impact on the likelihood of adopting
machinery: it has a positive and significant impact on the probability to adopt irrigation pump,
thresher, and others while no impact on the decision to adopt tractor and harvester.
Adoption of all the technology under study except tractor has a positive association with the
literacy status of the spouse of the household head, implying that these decisions are made at
household level rather than individually by the household head. Access to credit can significantly
Page 25 of 37
and positively affect the adoption of the harvester and other machines. As other machines
include some of the conservation agriculture-related machines, which farmers with credit access
are usually found to have adopted. Membership in the village agricultural institutions such as
cooperative/clubs is found to significantly increase the likelihood to adopt harvester and all
others while the adoption of the tractor, irrigation pump, and thresher is not affected by it. Farm
households with off-farm income are found to have a higher likelihood to adopt power tiller and
thresher, but not the irrigation pump. Households that work as farm laborers are more likely to
use all of the machines studied. Generally, households that are far from the market and
agricultural extension services are less likely to adopt all of these machines studied. The primary
reason for this is the vicinity of custom hire service. Training on agriculture-related issues
increases farmer's exposure to technology, and this enhances the likelihood to adopt labor-saving
technology in farming.
Mechanization differs significantly across the states under study. Compared to farmers in Bihar
state, farmers in Haryana state are more likely to adopt harvester and other machines while they
are less likely to adopt irrigation pumps and thresher. Farmers in Haryana use a combined
harvester, and thus, it substitutes to thresher. As a result, they are less likely to adopt thresher.
Similarly, canal irrigation also supplements the groundwater irrigation (which usually requires
pumps for irrigating water) in Haryana, leading to less use of irrigation pumps as compared to
Bihar.
7. Conclusion
Mechanization in India is mostly on land preparation and harvesting operation. The main
conclusions of this study are: i) the adoption decisions of farm machines, including tractors,
Page 26 of 37
irrigation pumps, thresher, harvester, and others, are interrelated. Thus, ignoring these
interrelationships can lead to biased estimations, ii) custom hiring services of agricultural
machines has increased the access of farmers to these machines, and farmers no longer need to
own machines. Therefore, defining adoption in terms of ownership of machine rather than its use
largely underestimates adoption and ignore the role of the rental market, and iii) as the decisions
to adopt these technologies are largely taken at the household level, improving the level to
education among female farmers (i.e., spouse of the household head) can significantly enhance
the uptake of agricultural machines. This indicates how gender influence mechanization of
agriculture and it will be a matter of growing importance, given the increasing involvement of
male in off-farm income and migration. This study has two important policy implications: i)
enhancing custom hiring services is essential for promoting the farm mechanization in India ii)
improved market access, provision of extension services, and training on agriculture
technologies and systems positively contribute to the adoption of agricultural machineries.
Acknowledgments
This work was carried out by International Maize and Wheat Improvement Center (CIMMYT)
as part of the CGIAR Research Program on Climate Change, Agriculture, and Food Security
(CCAFS), with support from CGIAR Fund Donors and through bilateral funding agreements.
For details, please visit https://ccafs.cgiar.org/donors. We thank field teams for collecting data in
Haryana and Bihar states of India. The views expressed here are those of the authors and do not
necessarily reflect the views of the author's institutions, CCAFS.
Page 27 of 37
References
Adu-Baffour, F., Daum, T., Birner, R., 2019. Can small farms benefit from big companies’
initiatives to promote mechanization in Africa? A case study from Zambia. Food Policy 84, 133-
145.
Aguilera, E., Guzmán, G.I., González de Molina, M., Soto, D., Infante-Amate, J., 2019. From
animals to machines. The impact of mechanization on the carbon footprint of traction in Spanish
agriculture: 1900–2014. Journal of Cleaner Production 221, 295-305.
Ahmed, M., Goodwin, B., 2016. Agricultural mechanization and non-farm labor supply of farm
households: evidence from Bangladesh, Selected Paper Presented at the 2016 Agricultural &
Applied Economics Association Annual Meeting, Boston, MA.
Aryal, J.P., Farnworth, C.R., Khurana, R., Ray, S., Sapkota, T., 2014. Gender dimensions of
climate change adaptation through climate smart agricultural practices in India. Innovation in
Indian Agriculture: Ways Forward.
Aryal, J.P., Holden, S.T., 2012. Livestock and land share contracts in a Hindu society.
Agricultural Economics 43, 593-606.
Aryal, J.P., Holden, S.T., 2013. Land Reforms, Caste Discrimination and Land Market
Performance in Nepal. Land Tenure Reform in Asia and Africa, 29.
Aryal, J.P., Mehrotra, M.B., Jat, M., Sidhu, H.S., 2015a. Impacts of laser land leveling in rice–
wheat systems of the north–western indo-gangetic plains of India. Food Security 7, 725-738.
Aryal, J.P., Mehrotra, M.B., Jat, M.L., Sidhu, H.S., 2015b. Impacts of laser land leveling in rice–
wheat systems of the north–western indo-gangetic plains of India. Food Security 7, 725-738.
Aryal, J.P., Rahut, D.B., Maharjan, S., Erenstein, O., 2019. Understanding factors associated
with agricultural mechanization: A Bangladesh case. World Development Perspectives 13, 1-9.
Aryal, J.P., Rahut, D.B., Thapa, G., Simtowe, F., 2021. Mechanisation of small-scale farms in
South Asia: Empirical evidence derived from farm households survey. Technology in Society 65,
101591.
Aryal, J.P., Sapkota, T.B., Jat, M.L., Bishnoi, D.K., 2015c. On-Farm Economic and
Environmental Impact of Zero-Tillage Wheat: A Case of North-West India. Experimental
Agriculture 51, 1-16.
Aryal, J.P., Sapkota, T.B., Stirling, C.M., Jat, M.L., Jat, H.S., Rai, M., Mittal, S., Sutaliya, J.M.,
2016. Conservation agriculture-based wheat production better copes with extreme climate events
than conventional tillage-based systems: A case of untimely excess rainfall in Haryana, India.
Agriculture, Ecosystems & Environment 233, 325-335.
Baudron, F., Sims, B., Justice, S., Kahan, D.G., Rose, R., Mkomwa, S., Kaumbutho, P., Sariah,
J., Nazare, R., Moges, G., Gérard, B., 2015. Re-examining appropriate mechanization in Eastern
and Southern Africa: two-wheel tractors, conservation agriculture, and private sector
involvement. Food Security 7, 889-904.
Behera, B., Rahut, D.B., Jeetendra, A., Ali, A., 2015. Household collection and use of biomass
energy sources in South Asia. Energy 85, 468-480.
Belsley, D.A., 1991. A guide to using the collinearity diagnostics. Computer Science in
Economics and Management 4, 33-50.
Belsley, D.A., Kuh, E., Welsch, R.E., 2005. Regression diagnostics: Identifying influential data
and sources of collinearity. John Wiley & Sons.
Benin, S., 2015. Impact of Ghana's agricultural mechanization services center program.
Agricultural economics 46, 103-117.
Page 28 of 37
Bhattarai, M., Shekhawat, R.S., Joshi, P., 2016. Diffusion of Farm Tractor in India from 1960 to
2012: Patterns and Prospects. Agricultural and Applied Economics Association.
Biggs, S., Justice, S., 2011. Rural Development & Energy Policy: Lessons from Agricultural
Mechanisation in South Asia. Observer Research Foundation.
Biggs, S., Justice, S., 2015. Rural and agricultural mechanization: A history of the spread of
small engines in selected Asian countries.
Binswanger, H., 1986. Agricultural Mechanizationa Comparative Historical Perspective. The
World Bank Research Observer 1, 27-56.
Birthal, P.S., Kumar, S., Negi, D.S., Roy, D., 2015. The impacts of information on returns from
farming: evidence from a nationally representative farm survey in India. Agricultural economics
46, 549-561.
Bortamuly, A.B., Goswami, K., 2015. Determinants of the adoption of modern technology in the
handloom industry in Assam. Technological Forecasting and Social Change 90, 400-409.
Chander, P., Thangavelu, S.M., 2004. Technology adoption, education and immigration policy.
Journal of Development Economics 75, 79-94.
Chowdhury, A.H., Hambly Odame, H., Leeuwis, C., 2014. Transforming the roles of a public
extension agency to strengthen innovation: Lessons from the National Agricultural Extension
Project in Bangladesh. The journal of agricultural education and extension 20, 7-25.
De Janvry, A., Fafchamps, M., Sadoulet, E., 1991. Peasant household behaviour with missing
markets: some paradoxes explained. The Economic Journal 101, 1400-1417.
Diao, X., Cossar, F., Houssou, N., Kolavalli, S., 2014. Mechanization in Ghana: Emerging
demand, and the search for alternative supply models. Food Policy 48, 168-181.
Erenstein, O., Farooq, U., 2009. A survey of factors associated with the adoption of zero tillage
wheat in the irrigated plains of South Asia. Experimental Agriculture 45, 133-147.
Fan, S., Chan‐Kang, C., 2005. Is small beautiful? Farm size, productivity, and poverty in Asian
agriculture. Agricultural Economics 32, 135-146.
FAO, 2014. Building a common vision for sustainable food and agriculture: Principles and
Approaches Food and Agriculture Organization of The United Nations, Rome, Italy.
FAO, 2018a. India at a glance [http://www.fao.org/india/fao-in-india/india-at-a-glance/en/],
Rome, Italy.
FAO, 2018b. Sustainable Agricultural Mechanization [http://www.fao.org/sustainable-
agricultural-mechanization/overview/why-mechanization-is-important/en/], Rome, Italy.
Ghosh, B.K., 2010. Determinants of farm mechanization in modern agriculture: A case study of
Burdwan districts of west Bengal. International Journal of Agricultural Research 5, 1107-1115.
Greene, W.H., 2003. Econometric Analysis, Journal of the American Statistical Association. ,
Prentice Hall. doi:10.1198/jasa.2002.s458.
Hatibu, N., Kienzle, J., Ashburner, J.E., Sims, B.G., 2013. Mechanization for rural development:
A review of patterns and progress from around the world.
Hazell, P.B., 2005. Is there a future for small farms? Agricultural Economics 32, 93-101.
Holden, S., Shiferaw, B., Pender, J., 2001. Market imperfections and land productivity in the
Ethiopian highlands. Journal of Agricultural economics 52, 53-70.
Holden, S.T., Ghebru, H., 2016. Land tenure reforms, tenure security and food security in poor
agrarian economies: Causal linkages and research gaps. Global Food Security 10, 21-28.
Jha, R., Chitkara, P., Gupta, S., 2000. Productivity, technical and allocative efficiency and farm
size in wheat farming in India: a DEA approach. Applied Economics Letters 7, 1-5.
Page 29 of 37
Jha, R., Rhodes, M.J., 1999. Some imperatives of the Green Revolution: technical efficiency and
ownership of inputs in Indian agriculture. Agricultural and Resource Economics Review 28, 57-
64.
Jöreskog, K.G., Sörbom, D., 1993. LISREL 8: Structural equation modeling with the SIMPLIS
command language. Scientific Software International.
Kassie, M., Jaleta, M., Shiferaw, B., Mmbando, F., Mekuria, M., 2013. Adoption of interrelated
sustainable agricultural practices in smallholder systems: Evidence from rural Tanzania.
Technological forecasting and social change 80, 525-540.
Kassie, M., Ndiritu, S.W., Stage, J., 2014. What determines gender inequality in household food
security in Kenya? Application of exogenous switching treatment regression. World
Development 56, 153-171.
Khanal, U., 2018. Why are farmers keeping cultivatable lands fallow even though there is food
scarcity in Nepal? Food Security 10, 603-614.
Khanal, U., Alam, K., Khanal, R.C., Regmi, P.P., 2015. Implications of out-migration in rural
agriculture: a case study of Manapang village, Tanahun, Nepal. The Journal of Developing Areas
49, 331-352.
Kienzle, J., Ashburner, J.E., Sims, B.G., 2013. Mechanization for rural development: A review
of patterns and progress from around the world.
Ma, W., Renwick, A., Grafton, Q., 2018. Farm machinery use, off‐farm employment and farm
performance in China. Australian Journal of Agricultural and Resource Economics 62, 279-298.
Maharjan, A., Bauer, S., Knerr, B., 2013a. International migration, remittances and subsistence
farming: Evidence from Nepal. International Migration 51, e249-e263.
Maharjan, A., Bauer, S., Knerr, B., 2013b. Migration for labour and its impact on farm
production in Nepal. Centre for the Study of Labour and Mobility Kathmandu, Nepal.
Mottaleb, K.A., Krupnik, T.J., Erenstein, O., 2016. Factors associated with small-scale
agricultural machinery adoption in Bangladesh: Census findings. Journal of Rural Studies 46,
155-168.
Mottaleb, K.A., Rahut, D.B., Ali, A., Gérard, B., Erenstein, O., 2017. Enhancing Smallholder
Access to Agricultural Machinery Services: Lessons from Bangladesh. The Journal of
Development Studies 53, 1502-1517.
NAAS, 2013. Climate Resilient Agriculture in India [ttp://www.naasindia.org], Policy Paper No.
65. National Academy of Agricultural Sciences, New Delhi, India.
Pingali, P., 2007. Chapter 54 Agricultural Mechanization: Adoption Patterns and Economic
Impact, in: Evenson, R., Pingali, P. (Eds.), Handbook of Agricultural Economics. Elsevier, pp.
2779-2805.
Prabakar, C., Sita Devi, K., Selvam, S., 2011. Labour scarcity–Its immensity and impact on
agriculture. Agricultural Economics Research Review 24, 373-380.
Rigg, J., Salamanca, A., Thompson, E.C., 2016. The puzzle of East and Southeast Asia's
persistent smallholder. Journal of Rural Studies 43, 118-133.
Sims, B., Hilmi, M., Kienzle, J., 2016. Agricultural mechanization: a key input for sub-Saharan
Africa smallholders. Integrated Crop Management (FAO) eng v. 23 (2016).
Sims, B.G., Kienzle, J., 2006. Farm power and mechanization for small farms in sub-Saharan
Africa. Agricultural and Food Engineering Technical Report (FAO).
Singh, G., 2006. Agricultural Machinery Industry in India (Manufacturing, marketing and
mechanization promotion). Status of Farm Mechanization in India.
Page 30 of 37
Singh, G., 2011. Mechanization Investments in India and Lessons from Africa, in: Ashburner,
J.E., Kienzle, J. (Eds.), Investment in Agriculture Mechanization in Africa. Food and Agriculture
Organization of The United Nations, Food and Agriculture Organization of the United Nations,
Rome, Italy, pp. pp. 26–31.
Singh, G., 2015. Agricultural mechanisation development in India. Indian journal of agricultural
economics 70, 64-82.
Van den Berg, M.M., Hengsdijk, H., Wolf, J., Van Ittersum, M.K., Guanghuo, W., Roetter, R.P.,
2007. The impact of increasing farm size and mechanization on rural income and rice production
in Zhejiang province, China. Agricultural Systems 94, 841-850.
Wang, X., Yamauchi, F., Otsuka, K., Huang, J., 2016. Wage Growth, Landholding, and
Mechanization in Chinese Agriculture. World Development 86, 30-45.
Wiggins, S., Keats, S., 2014. Rural wages in Asia. . Overseas Development Institute 203
Blackfriars Road London SE1 8NJ. United Kingdom.
Wooldridge, J.M., 2002. Econometric Analysis of Cross Section and Panel Data, Third. ed, MIT
Press Books. MIT Press Cambridge, Massachusetts London, England.
doi:10.1515/humr.2003.021.
World Bank, 2015. World Development Indicators, 2015.
[http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators#].
World Bank Washington DC, USA.
Xiaobing, W., Yamauchi, F., Huang, J., 2016. Rising wages, mechanization, and the substitution
between capital and labor: evidence from small scale farm system in China. Agricultural
economics 47, 309-317.
Yamano, T., Rajendran, S., Malabayabas, M.L., 2015. Farmers’ self-perception toward
agricultural technology adoption: evidence on adoption of submergence-tolerant rice in Eastern
India. Journal of Social and Economic Development 17, 260-274.
Yamauchi, F., 2016. Rising real wages, mechanization and growing advantage of large farms:
Evidence from Indonesia. Food Policy 58, 62-69.
Yang, J., Huang, Z., Zhang, X., Reardon, T., 2013. The rapid rise of cross-regional agricultural
mechanization services in China. American Journal of Agricultural Economics 95, 1245-1251.
Zhang, X., Rashid, S., Ahmad, K., Ahmed, A., 2014. Escalation of Real Wages in Bangladesh: Is
it the Beginning of Structural Transformation? World Development 64, 273-285.
Tables
Page 31 of 37
Table 1: Use of agriculture power sources in India
Machinery 1960 1970 1980 1990 2000 2003 2008
Drought Animals (million)1 80.4 82.6 73.4 70.9 60.3 na 50
Tractor (thousands) 1 37 146 531 1200 2600 na 3600
Irrigation pumps (million) 1 0.43 3.3 6.23 12.87 19.5 na 25
Combine Harvester (million)2 0.0058 0.067 0.158 0.498 2.85 4.04 na
Source: 1( Singh, 2011) ;2http://faostat3.fao.org : Retrieved on 31 Aug. 2016; na refers to non-availability of data
Table 2: Tractors and irrigation pump density over time in Haryana, Bihar and India
State/Years
Tractor density (number per 1000 ha) Pump density (number per 1000 ha)
1962-65 1982-83 1992-93 2005-08 1962-65 1982-83 1992-93 2005-08
Haryana 0.7 17 44.4 55 2 71 143 155
Bihar 0.2 1.8 1.9 13 1 47 89 117
India 0.3 3.7 8.6 17 5 49 79 111
Source: Bhalla and Singh (2012)
Page 32 of 37
Table 3: The sequence of mechanization followed in a place as per complexity in design and
power needs
Sequence type of
operation
Operation
I. High power low
skill
II. Medium power
medium skill
III. Low power high
Skill
Stationary
Example: Grinding,
milling, crushing,
pumping, threshing
Example: Grinding by
size, cleaning
Example: Grinding by
quality
Mobile Example: Land
preparation, Transport
( Tractor, Power tiller)
Example: Seeding of
grain, harvesting of
grain
Example:
Transplanting,
Harvesting of cotton,
fruits and vegetables,
sugarcane
Adapted from (Pingali 2007; and Singh, 2011)
Table 4: Distribution of sample households by village in the Karnal and Vaishali districts
Karnal District (Haryana State) Vaishali District (Bihar State)
Village Name Sample Size Village Name Sample Size
Anjanthali 67 Bhatha Dari 63
Bir Narayana 49 Bilandpur 68
Pakhana 80 Mukundpore 69
Sandhir 64 Raja Pakar 70
Barthal 41 Dedhpore 46
Churni Jagir 20 Dhabhaich 46
Darar 25 Panapur Camp 56
Garghi Jattan 46 Rasalpore 48
Hathlana 46 Varishpore 31
Mohri Jagir 40 Laxminarayanpore 44
Nanhara 43 Mirpur 55
Sanwat 45 Rampur Ratnagar 45
Sounkra 60
Total sample size 626 641
Page 33 of 37
Table 5: Description of Variables used in the analysis
Variables Haryana Bihar overall Variable Description
Mean S.D Mean S.D Mean S.D
Dependent Variables
Use Tractor (D:
Dummy) 0.96 0.20 0.96 0.20 0.96 0.20 1 if used tractor and 0 otherwise
Use Pump (D) 0.91 0.29 0.96 0.19 0.93 0.25 1 if used pump and 0 otherwise
Use Thresher (D) 0.41 0.49 0.94 0.24 0.68 0.47 1 if used thresher and 0 otherwise
Use Harvester (D) 0.68 0.47 0.01 0.12 0.34 0.47 1 if used harvester and 0 otherwise
Use other Machinery
(D) 0.67 0.47 0.02 0.15 0.34 0.47 1 if used tiller/ generator/ others and
0 otherwise
Independent Variables
Demographic variables
Gender of HH (D) 0.97 0.17 0.91 0.29 0.94 0.24 1 if male headed house and 0 if
female
Age of HH (C:
Continuous) 49 13 51 14 50 13 Age of household in years
Education of HH (D) 0.67 0.47 0.62 0.49 0.64 0.48 1 if HH went to school and 0
otherwise
Education of Spouse
(D) 0.51 0.50 0.30 0.46 0.40 0.49 1 if HH's spouse went to school and
0 otherwise
AEC (C) 4.46 1.79 0.89 0.17 2.65 2.19 Adult equivalent member in house
Family Size (C) 6.03 2.47 6.05 2.65 6.04 2.56 Total members in family
Food Security status
(D) 1.00 0.07 0.69 0.46 0.84 0.36 1 if HH is food secure and 0
otherwise
Migration (D) 0.11 0.31 0.28 0.45 0.19 0.40 1 if HH has migrant member and 0
no
Training (D) 0.22 0.41 0.52 0.50 0.37 0.48
1 if HH has received training on
improved seeds, soil & water
management, crop rotation,
minimum tillage, 0 otherwise
Economic And Social Capital
Land operated (C) 3.33 3.81 0.39 0.41 1.84 3.06 Total land operated in ha
TLU (C) 3.81 5.70 0.58 0.74 2.18 4.35 Livestock owned in tropical
livestock unit
Loan Taken (D) 0.40 0.49 0.34 0.48 0.37 0.48 1 if taken loan in last 24 months and
0 otherwise
Non-Agricultural
income (D) 0.15 0.36 0.47 0.50 0.31 0.46 1 if HH has income from non-
agriculture source, 0 otherwise
Farm labor income
(D) 0.01 0.09 0.04 0.19 0.02 0.15 1 if works as on farm labor, 0
otherwise
Non-Farm labor
income (D) 0.03 0.18 0.25 0.43 0.14 0.35 1 if works as non-farm labor, 0
otherwise
Membership in
Group (D) 0.35 0.48 0.09 0.28 0.22 0.41 1 if any family member is member
in any group and 0 otherwise
Access to market and agriculture extension service
Page 34 of 37
Distance To Nearest
Main Market (C) 6.43 3.07 4.69 4.67 5.55 4.06 Distance to nearest main market
from house (in km)
Distance To
Agriculture
Extension Office (C)
5.25 2.84 5.04 3.66 5.15 3.28 Distance to agriculture extension
service from house (in km)
Table 6: Distribution of sample households by the type of machine used and owned
Machine types
Percentage of farm households
Bihar Haryana Overall
user owner user owner user owner
Tractor 96 2 96 37 96 20
Irrigation pump 96 8 91 86 93 46
Thresher 94 1 41 3 68 2
Harvester 1 0 68 10 34 5
Other machinery 2 0 67 7 34 4
Table 7: Pairwise correlation coefficients across farm machine adoption decisions
Farm machinery Coefficient Standard error P-value
Tractor × irrigation pump 0.678*** 0.053 0.000
Tractor ×thresher 0.415*** 0.662 0.000
Tractor × harvester 0.222*** 0.082 0.007
Tractor ×other 0.093 0.085 0.227
Irrigation pump × thresher 0.341*** 0.064 0.000
Irrigation pump × harvester 0.275*** 0.077 0.000
Irrigation pump × other -0.041 0.076 0.594
Thresher × harvester -0.338*** 0.065 0.000
Thresher× others -0.066 0.061 0.351
Harvester × others 0.028 0.071 0.669
Note: *, **, *** refer to 10, 5 and 1 % level of significance.
Likelihood ratio test of rho21 = rho31 = rho41 = rho51 = rho32 = rho42 = rho52 = rho43 = rho53 = rho54 = 0:
chi2(10) =201.77 Prob> chi2 = 0.000
Page 35 of 37
Table 8: Factors associated with the adoption of farm machineries in the study area
(Results of multivariate probit model)
Tractor Pump Thresher Harvester Others
General caste HH 0.05 0.22 -0.14 0.11*** 0.09***
(0.20) (0.15) (0.11) (0.04) (0.03)
Male headed HH -0.27 -0.12 -0.15 0.21** 0.12***
(0.57) (0.73) (0.38) (0.10) (0.03)
Farm experience 0.01 0.02*** -0.00 0.17*** 0.10***
(0.01) (0.01) (0.01) (0.05) (0.03)
Literate HH head -0.07 0.14 -0.03 0.13 -0.00
(0.20) (0.15) (0.12) (0.14) (0.13)
Literate spouse 0.29 0.44*** 0.22* 0.31** 0.29**
(0.21) (0.17) (0.12) (0.13) (0.13)
Family size 0.04 -0.03 0.03 -0.04* -0.02
(0.04) (0.03) (0.02) (0.02) (0.02)
Operated land 0.03 0.13*** -0.08*** 0.07*** 0.04***
(0.03) (0.03) (0.02) (0.02) (0.01)
Livestock owned -0.02 0.09** 0.05* 0.01 0.07**
(0.05) (0.04) (0.03) (0.03) (0.03)
Member institution 0.07 0.13 -0.14 0.45*** 0.37**
(0.26) (0.02) (0.13) (0.16) (0.15)
Credit access 0.19*** 0.26*** 0.16 0.06*** 0.30**
(0.07) (0.06) (0.11) (0.02) (0.12)
Market distance -0.17*** 0.03 0.02 -0.04** -0.04**
(0.02) (0.02) (0.02) (0.02) (0.02)
Distance to extension service 0.01 -0.00 -0.0 -0.04 -0.05***
(0.03) (0.02) (0.02) (0.03) (0.02)
Off-farm income 0.68*** 0.04 0.13*** 0.15** 0.17***
(0.24) (0.20) (0.05) (0.07) (0.06)
Work as farm labor 0.68 0.14 0.30 -0.17*** -0.16**
(2.31) (0.50) (0.48) (0.04) (0.08)
Work as non-farm labor -0.78* -0.26 0.06 -0.13 -0.15
(0.45) (0.26) (0.22) (0.30) (0.25)
Training 0.38* 0.01 0.05 0.46*** 0.22***
(0.22) (0.19) (0.13) (0.17) (0.05)
Number of farm plots -0.09*** -0.11** 0.04 0.03 -0.13***
(0.03) (0.04) (0.05) (0.07) (0.05)
Share of rented land -0.21* -0.08 -0.14 -0.24*** 0.23
(0.13) (0.11) (0.13) (0.07) (0.16)
State (Haryana) -0.00 -0.32 -1.48*** 2.25*** 2.22***
(0.31) (0.26) (0.20) (0.29) (0.24)
Constant 1.90*** 5.25 1.45*** -2.17*** -2.73***
(0.72) (87.03) (0.50) (0.66) (0.66)
Number of observations 1267 1267 1267 1267 1267
Page 36 of 37
Wald chi2(95) 884.31
Prob > chi2 0.000
Log likelihood -1393.51
Figure 1: Conceptual framework for mechanization
Page 37 of 37
Figure 2: Map showing the study area