CHAPTER 3

RESEARCH METHODOLOGY

3.1 INTRODUCTION

The broad methodological outline presented in chapter one is discussed here in detail.
The chapter opens up with an introduction to the study area, starting from Rajasthan State to
the district and finally to the villages. Following it, the research methodological framework
within which the study is conducted is presented. It covers description on pilot survey
undertaken to ascertain study feasibility, sampling strategy, data collection method, the
variables that have been collected from the field survey of farm households and method used
for analysing those variables.

The remaining chapter is organised in the following order. Section 3.2 begins with a
background of the study area. It consists of three sub-sections. The first sub-section – 3.2.1
presents a broad overview of the State of Rajasthan and increasing commercial orientation of
agriculture in the state making it an ideal choice to study commercial agriculture. Second
sub-section – 3.2.2 lays out an outline of Jhunjhunu district, followed by the third sub-section
3.2.3 on the study villages and criterion using which they are selected. Section 3.3 lays out
the research methodology followed in the study. It includes pilot survey in section 3.3.1,
sampling design in section 3.3.2 and method of data collection in section 3.3.3. Following it,
section 3.3.4 list out the key variables identified using structured questionnaire. How the
variables have been analysed is explained in section 3.3.5. Finally, the chapter ends with
conclusion given in section 3.4.

3.2 BACKGROUND OF THE STUDY AREA

3.2.1 An Overview of Rajasthan State

As mentioned in the introductory chapter, the present study employs macro data and
also uses micro data collected from three villages of a district in Rajasthan. The following
five aspects are covered in this section: rationale behind choosing Rajasthan State for the
44
study, its geo-physiological profile, agriculture sector contribution, types of crops grown and
human development index (HDI).

In recent years, subsistence agriculture in the State of Rajasthan has substantially

turned commercial. To be specific, Rajasthan emerged as the highest contributor of
agricultural export of any item in 2012-13. Guar gum derived from guar (cluster bean), which
was a subsistence crop, surpassed basmati rice and was exported to the tune of Rs 2127.8
million in 2012-13 (Planning Commission, 2014). Given the dynamic agrarian conditions
which the State is undergoing makes it a perfect place to study the transition from subsistence
to commercial agriculture. Moreover, the desert and rain-fed agriculture of Rajasthan has not
attracted much scholarly attention unlike the irrigated States of India, which makes it even
more essential to fill the knowledge gap by bringing out the details of agriculture scenario of
the state of Rajasthan.

The north-western state of Rajasthan is the largest State of India covering 10 per cent
of the land mass. As the state is endowed with diverse rainfall level and weather conditions, it
is divided into ten agro-climatic zones, covering three different water regimes – rain-fed area,
area irrigated with ground water (tube well and open well), and canal command area. The
Aravali hill range runs across the state and divides it into two regions, the western region
which comprises Thar desert covering 60 percent of the state area and the eastern region
which is fertile plain. These two regions come under the arid and semi-arid zone,
respectively, experience frequent droughts, a short monsoon season between June to
September, an average rainfall of 567 mm (during 2015-16) and wide variation in summer
and winter temperature (GoR, 2017). Owing to scanty rainfall, there is limited availability of
ground water, and as a result of heavy extraction the water level has fallen sharply (GoR,
2010).

The state is overwhelmingly rural. Agriculture and allied activities continue to be the
mainstay of about 65 per cent of population. Agriculture sector contributed around 22.6 per
cent to the net state domestic product in 2008-2009 (Swain, Kalamkar, and Ojha, 2012). In
2015-16, the gross irrigated area accounted for only 30 per cent (10.56 million ha) of the
gross cropped area (25.01 million ha). Tube wells and open wells owned by farmers dominate
around 60 per cent of the gross irrigated area. As a corollary, agriculture in Rajasthan is
primarily rain-fed and continues to be susceptible to the vagaries of monsoon. Around 65 per
45
cent of the total cultivation is under kharif season which is mostly dependent on rainfall
(GoR, 2017). Due to the wide fluctuation in rainfall, variation in agriculture production,
particularly of kharif crops, is a common pattern (Human Development Report Rajasthan,
2008). Therefore, livestock rearing is practiced by most rural households to prevent sharp fall
in income due to rainfall failure and fall in production. Rajasthan is rich in livestock wealth,
namely cow, buffalo, sheep and goat, along with poultry and camel. The livestock sector
contributes over 9 per cent to the state GDP. It also contributes about 11 percent of the
country’s milk production, 30 percent of mutton production and over 40 per cent of wool
production (GoR, nd ).

The staple food grain crops grown are bajra (pearl millet) in kharif season and wheat
in rabi season. Besides this, Rajasthan is India’s largest producer of mustard oilseed, bajra,
three spices: coriander, cumin, and methi (fenugreek), guar (cluster bean), isabgol and the
second largest producer of maize (Swain, Kalamkar, and Ojha, 2012).

In terms of Human Development Index (HDI) which covers health, education and
income, Rajasthan ranks low as it occupied 17th rank among 23 states in 2007-08 (Planning
Commission, 2011). Especially in education, according to 2011 census, the literacy rates
among males was 79.19 per cent, while among females it was 52.12 per cent, reflecting a
wide gap between male and female literacy level and also low literacy among females
(Census of India, 2011).

3.2.2 An Overview of Jhunjhunu District

Within Rajasthan State, Jhunjhunu District is selected for the study. The reason
behind selecting Jhunjhunu district and its broad agriculture profile is presented here. The
district under study is chosen for its closeness to national capital - Delhi (at a distance of 200
km), wherein lie three large government regulated food-grain wholesale agricultural markets
(krishi mandis) at Narela, Shahdara and Nazafgarh. In these markets, farmers from Rajasthan,
Punjab, Haryana, Uttar Pradesh and Madhya Pradesh come to sell their crop to obtain
maximum price. For study area selection, we have used the criterion of proximity to an urban
agriculture market as it is suggested to have significant implications on cropping pattern, in
particular, motivating farmers to cultivate high value commercial crops over subsistence
crops in the hope of getting a high price in national market. Further, better transport facility
46
and all weather roads become available in an area with proximity to an urban centre of
national importance which may make commercial farming – buying input and selling output
easier.

Jhunjhunu district lies in the north-eastern part of Rajasthan state. The district
receives an annual average rainfall of 444 mm, which is less than the average rainfall
received in the State. Around 35 per cent of the cultivated area is under irrigation by tube
wells and open wells. Jhunjhunu does not come under any canal command area. The cropping
intensity is low in this region because most farms are rain-fed wherein cropping is done for
half of the year in kharif season, and for the remaining half of the year, many fields remain
fallow due to dearth of assured irrigation facility. The major crops grown in Jhunjhunu are
bajra, guar, moong (green gram), moth bean, chana (bengal gram), wheat, mustard and jau
(barley) (Census of India, 2001). The marginal and small cultivators largely dominate
agriculture in Jhunjhunu district, comprising around 37 per cent and 33per cent, respectively,
while the large farmers comprise just 0.47 per cent of all farmers. The average size of land
holding is 1.77 ha. (Department of Agriculture, 2011).

3.2.3 Study Villages

The three study villages are located in Jhunjhunu district of Rajasthan State. How did
we arrive at the selection of the villages is described here. In line with district selection, of all
the tehsils, we identified three tehsils of Jhunjhunu district – Khetri, Surajgarh, and Buhana -
which are closest to Delhi. A tehsil as an administrative division denotes a sub-district and
comprises of several villages and few towns. In Rajasthan state, as an administrative unit,
tehsil, block and Panchayat Samiti are same. Further, from each tehsil Besarda, Khudaniya,
and Shahpur village were selected using the following criteria: (a) proximity to the nearest
town; (b) all weather road connectivity to the town; (c) diverse caste groups; (d) both
irrigated and dry land in the village; and (e) presence of bank in close vicinity. The area
selected for study is encapsulated in figure 3.1. According to census 2011, the three villages
have a total of 1353 households (table 3.1). The three villages are chosen from three different
tehsils of the district to add diversity in the data than selecting adjoining villages. We also did
not choose three different districts of the state to avoid too much of diversification which is
bound to come in cropping pattern of different agro-climatic zones. The primary data was

47
collected from March to December 2017 about rabi 2016 (Nov 2015 to May 2016) and kharif
2016 (June to Oct 2016) cropping season.

Table 3.1: Study Villages

Village Name Number of Households Total Population

A Besarda 565 3186

B Khudaniya 557 2745

C Shahpur 231 1173

Grand Total 1353 7104

Source: Census of India 2011 (2014a)

Change in Cultivators and Agricultural Labourers

After obtaining an overview of the study area, what is the proportion of those
involved in farming is examined. For this, using census data, the changes that has come in the
percentage of cultivators and agricultural labourers in the 21 st century is assessed. Table 3.2
reflects that from 2001 to 2011 the per cent of persons involved as cultivators has come down
at all the three levels - State, district and the villages under study. Likewise, has been the
downfall in per cent of both male and female cultivators. It indicates that people are moving
out of farming and are joining other sectors. However, the picture of agricultural labourers is
different.

In both the State and district level, the per cent of agricultural labourers has gone up
and within agricultural labourers, more females are involved as labourers than males. It
points towards feminisation of agriculture work because the males move out and join other
worker category while females are left behind in the village to earn as agriculture labour,
along with looking after family and children. Further, at the village level, the per cent of
agricultural labourers (persons, males and females) has come down as they have moved out
of the agricultural worker category and have joined the other workers category. The probable
reason behind moving out of agricultural worker category is lack of work for labour in the
agriculture sector because of mechanisation and greater demand, less back breaking work and
more wages on working outside agriculture.

48
 Table 3.2 Change in Cultivators and Agricultural Labourers
2001 2011
Rajasthan State Jhunjhunu District Total Villages* Rajasthan State Jhunjhunu District Total Villages*
Main & Marginal
Workers Number Percent Number Percent Number Percent Number Percent Number Percent Number Percent
(i) Cultivators Persons 13140066 55.3 472232 62.1 1268 68.4 13618870 45.6 521692 58.3 1304 63.5
Males 7062726 48.1 232018 50.7 920 71.8 7518486 41.1 252595 46.7 879 61.7
Females 6077340 67 240214 79.2 348 60.8 6100384 52.6 269097 76.2 425 67.4

(ii) Agricultural
Labourers Persons 2523719 10.6 55257 7.3 135 7.3 4939664 16.5 68878 7.7 20 0.9
Males 1055332 7.2 33376 7.3 90 7 2132669 11.7 41286 7.6 18 1.3
Females 1468387 16.2 21881 7.2 45 7.9 2806995 24.2 27592 7.8 2 0.3

(iii) Other Workers** Persons 8102870 34.1 233421 30.7 450 24.3 11327721 37.9 304079 34 731 35.6
Males 6577744 44.8 192229 42.1 271 21.2 8645921 47.3 247511 45.7 527 37
Females 1525126 16.8 41192 13.6 179 31.3 2681800 23.1 56568 16.0 204 32.3

Source: Census of India 2001.(2006); Census of India 2011.(2014b)

* Besarda Village+ Khudaniya Village+Shahpur Village
** Other workers include household industry workers and other workers

49
3.3 RESEARCH METHODOLOGY

3.3.1 Pilot Survey

The purpose of the pilot survey was to ascertain the feasibility of the study. During the
pilot survey, the possible villages to conduct the study were identified and an understanding of
the broad socio-economic profile of the farm households was obtained. In the identified villages,
the questionnaire was pre-tested to examine the appropriateness of the questions being asked and
to ascertain the reliability and validity of the responses. It helped in refining the questions, the
codes, the serial order, and making the questions more focused in order to meet the study
objectives.

3.3.2 Sampling Design

Sampling Method
Farm households1 are the unit of enquiry in this study, while the farmer is the unit of
measurement. The farmers within the farm household are the key agents in the process of
intensification of commercialisation of agriculture. We consider one farmer per farm household
because the whole family when live together in one household, they cultivate the farm as one
unit. So, the number of farm households selected is same as the number of farmers studied.

To select the study unit, sampling was done in two phases. First, a complete listing of the
households in the three villages was carried out. Each household was given a unique number. A
house listing form was filled containing details of all the households about whether they cultivate
any land or not, whether they cultivate own land or leased–in land or have leased-out land, caste,
and occupation of households if not involved in farming. While doing the house listing, non-farm
households were excluded as they are not required for the study. From the house list, we selected
the sample using simple random sampling. The RANDBETWEEN formula in Microsoft excel

1
By farm household or agriculture household we imply (1) those households who own land and are cultivating it,
and (2) households who don’t own land but are cultivating leased- in land during last 365 days. Households who
own land but are not cultivating it have been excluded because cropping details about such land is obtained from
those who have leased-in the land. Moreover, the one who is tilling the land can give precise and complete details
about seeds-fertilizers and water applied than the owner who is not involved in day to day farm work. Our farm
household concept is in line with 59th NSS Round (NSSO, 2006).

50
was used to identify the farm households from the house list who were interviewed in the second
phase. The households getting repeated in generating a number with RANDBETWEEN formula
were replaced by generating more numbers. Simple random sampling is used for the study
because it gives equal chance to every farm household to get covered without bias towards any
group of household residing in any specific part of the village.

Besarda
Village

Rajasthan Jhunjhunu Khudaniya

State District Village

Shahpur
Village

Figure 3.1: Study Sample

Sample Size

To arrive at an appropriate sample size for this quantitative study, we use Daniel’s (1999)
NZ 2 P(1  P)
n! 
formula: d ( N  1)  Z 2 P(1  P) ,
2
where n! delineates sample size derived from the finite population
N, Z represents Z statistic for 95 per cent confidence level and 5 per cent margin of error (also
called confidence interval), P is the expected proportion, and d indicates precision. Using this,
the researcher arrived at a sample size of 300 farm households. An equal sample of 100 farm
households from all the three villages (3 x 100= 300) were selected.

51
 In Daniel’s (1999) sample size formula, the population size N 1353 is arrived at by
adding the number of households in the three villages as per census 2011 data (see table 3.1). Z
indicates the confidence level, which can be 90 per cent or 95 per cent or 99 per cent. We
employ the most commonly used Z value which is 95 per cent and is set as 1.96. The expected
proportion or expected prevalence, P, varies between 0 and 1, and the sample size is a variant of
P. It is important to note that P is taken in proportion of one, i.e., if expected proportion or
prevalence is 40 per cent, then P is equal to 0.4. In the formula, d signifies precision. Smaller d
implies good precision or smaller error of estimate, and it should be in proportion of one (Naing
et al., 2006). Interestingly, although there is no precise rule to choose an appropriate d, Naing et
al. (2006) suggest that if P is less than 10 per cent (0.1), then d should be half of P, i.e., 0.05. On
the other hand, if P is greater than 90 per cent (0.9), d would be 0.5 (1-P). Of course, a larger or
smaller d can be set depending on the availability of resources. If P is between 0.1 and 0.9, then
it is appropriate to choose 5 per cent precision (0.05). In this study, P is set at 50 per cent (0.5)
because it results in getting the largest sample size and d is set at 5 per cent. Moreover, when
precision is set at 5 per cent, by default confidence interval chosen is 95 per cent.

Thus,
Population Size (N) = 1353

Confidence Level (Z) = 95 % or 1.96

Expected Prevalence (P) = 50 % or 0.5

Precision (d) = 5 % or 0.05

(1353*1.96^2*0.5*0.5)/(0.05^2*(1353)+1.96^2*0.5*(0.5)) = 300

The advantage behind using Daniel’s sample size calculation formula is that it helps in
arriving at an optimum sample size, which is an essential requirement of any research, in order to
arrive at correct results. Before using a formula for sample size calculation, other aspects of the
study has to be taken into account. For instance, the present study uses simple random sampling
method and Daniel’s formula is appropriate in such case because it can be used only when
simple random or systematic random sampling method is employed. Further, Daniel sample size

52
formula fulfil the finite population correction assumption. Under the finite population correction,
a sample size formula is valid if the calculated sample size (300 farm households) is larger than
five per cent of the household population size (1353 households), which is the case in the present
study.

3.3.3 Method of Data Collection

Secondary Data

Using the time series data from 1957 to 2017, the cropping pattern of the state of
Rajasthan is analysed. The data concerning area, yield, production, irrigation, seed, and fertiliser
use trend are obtained from various published sources of Commissionerate of Agriculture and
Department of Planning, Government of Rajasthan, Jaipur and Ministry of Agriculture and
Farmers Welfare, New Delhi. The reference period for crop data collected by the agriculture
department is from July to June. The state level data presents a macro picture, while in order to
obtain a micro-view, farm level data is collected.

Primary Data

Personal interview of farmers were conducted using pre-coded and pre-tested structured
questionnaire to obtain a micro-view. The researcher applied recall-based interview method to
collect details on farming practice in the previous kharif and rabi season. Within each household,
the family member who looks after most farm work was interviewed.

3.3.4 Key Variables

The survey covers a wide range of variables which can be broadly classified into farmer’s
personal characteristics and agriculture characteristics. To be precise, the key variables enquired
about the farmer are age, level of education attained, highest education attained in the household,
caste, religion and household size. Subsequently, to obtain a holistic understanding about
agriculture practice of a farm household, we enquired details on landholding, irrigation, crops
grown in kharif and rabi season, seeds and fertilizers used in various crop, crops held back,
consumed and sold, food items consumed by the household from own farm and that from market
purchase, livestock holding and particulars about loan .

53
3.3.5 Data Analysis

The first level analysis concerning cropping pattern is done using simple statistical
measures such as the mean of secondary data concerning crop area and yield in six decades, seed
replacement rate (per cent), and per cent of crop area treated with fertiliser. While examining the
cropping pattern, it is ascertained that which crops have dominated over the years, and the
cropping pattern has tilted in favour of which crops in recent years. The primary data is analysed
using frequency distribution, percentage, cross tabulation, bar chart, and pie chart.

Subsequently, the second level of analysis is done wherein the standard econometric
methods such as gini coefficient, chi-square statistical test, correlation and logistic regression
technique has been used to ascertain the factors pushing farmers for crop sale.

3.4 CONCLUSION

This chapter presents the methodological aspects of the study. It begins with an outline of
the area chosen for study, explaining why Rajasthan state is most suitable for studying
subsistence and commercial agriculture and its main features such as topography, rainfall,
irrigation, livestock, crops grown and HDI status. Following it, a brief overview of the district is
presented. Consequently, the profile of the three villages where the study is conducted is given.
The sampling framework followed in the study is discussed explaining why and how simple
random sampling framework is used in the study. A structured questionnaire is employed to
interview randomly selected 300 farm households. The primary data obtained about farmer’s
characteristics, crops cultivated, seeds and fertilisers purchased and used for the same, crops sold
and kept back for household use, livestock and loan details are analysed using simple statistical
measures and regression method. In addition, the secondary data on the broad agriculture pattern
in the State of Rajasthan analysed using basic tools of analysis such as mean and percentage is
presented in the next chapter.

54
 Jhunjhunu District Map

Study Villages Besarda, Khudaniya, and Shahpur in Jhunjhunu District

Figure 3.2 Study Location

55