




































 

Ecology, Economy and Society–the INSEE Journal 2 (2): 45–74, July 2019 

 
RESEARCH PAPER 
 

Groundwater Extraction, Agriculture and Poverty in 
Godavari River Basin 
 

Jeena T Srinivasan  and Chandra Sekhara Rao Nuthalapati  
 
Abstract: This study analyses the relationship between access to groundwater 
irrigation, agricultural development, and poverty in Godavari river basin with 
heterogeneous hydrogeological resource conditions, and their implications for 
resource governance, using primary data from 825 farm households. The analysis 
showed that households with access to groundwater earn relatively higher 
household and per capita incomes. The large farmers seemed to have better access 
to water resources. Although poverty headcount ratio is high among the upper 
reach farmers, depth, and severity of poverty is more among middle reach farmers. 
Land size per se is not a significant variable in determining access, but the access is 
conditioned by sources of non-farm income, credit facility, education and caste 
hierarchy. While small farmers used more water per acre for irrigation without 
commensurate economic productivity, the differences in their economic 
productivity in comparison to large farmers was found to be associated with 
inequality.  

Keywords: Well Ownership, Groundwater Access, Per Capita Income, Poverty 
and Inequality, Water Use and Productivity 

 

1. INTRODUCTION 

Groundwater irrigation, driven by both demand side and supply side 
factors, has experienced explosive growth during the past few decades, and 
now it plays an important role in the agricultural development of India 

                                                           
 Centre for Economic and Social Studies, N. O. Campus, Begumpet, Hyderabad-500 016, 

Telangana; jeena@cess.ac.in.  

 Institute of Economic Growth, University of  Delhi  Enclave,  North  Campus,  Delhi-
110 007; chandra@iegindia.org 

Copyright © Srinivasan and Nuthalapati 2019. Released under Creative Commons 
Attribution-NonCommercial 4.0 International licence (CC BY-NC 4.0) by the author.  

Published by Indian Society for Ecological Economics (INSEE), c/o Institute of Economic 
Growth, University Enclave, North Campus, Delhi 110007.  

ISSN: 2581-6152 (print); 2581-6101 (web). 

DOI: https://doi.org/10.37773/ees.v2i2.77  

https://doi.org/10.37773/ees.v2i2.77


Ecology, Economy and Society–the INSEE Journal [46] 

 

(Shah, Singh and Mukherji 2006; Shah 2009; World Bank 2010; GoI 2014; 
Kulkarni and Shankar 2014; Zaveri et al. 2016). Availability of better drilling 
and pumping technologies, subsidised energy for extraction, flexibility and 
timeliness of supply, and poor delivery of public water supply system have 
contributed to raising groundwater consumption in farming. This helped 
the farmers optimise input use, diversify to high value crops, and achieve 
higher water productivity (Shah et al. 2007; World Bank 2010). While this 
led to reduction in rural poverty through various pathways (Hussain and 
Hanjra 2004; Narayanamoorthy 2007), the highly intensive development of 
groundwater also resulted in over exploitation, decline in groundwater 
levels in certain areas and sea water intrusion in coastal areas (Dhawan 
1989; Chopra 2003; Bhandhopadhyay 2007). According to recent estimates, 
out of 6584 groundwater assessment units, 1034 are over exploited, 253 
units are critical, 681 units are semi-critical, 4520 units are safe, and 96 units 
are saline in the country. The overall stage of groundwater development in 
the country is 62% and in several states like Punjab, Rajasthan, Haryana, 
and Delhi it is over 100% — implying that annual groundwater 
consumption is more than annual groundwater recharge and is 
unsustainable (GoI 2017).  

The implications of groundwater depletion are many and serious, as there is 
already an inherent inequality in groundwater accessibility due to skewed 
distribution (Sarkar 2011). With falling water tables and reduced well yields, 
users must pump water from greater depths, often using high capacity 
pumps, thus increasing the cost of extraction while making the resource out 
of reach for small and marginal farmers (Saleth 1996; Kulkarni et al. 2015). 
This, together with heterogeneous hydrogeological conditions, implies that 
resource endowment conditions are inequitable even within a typical village 
(Kulkarni and Thakkar 2012). The situation perpetuates with further 
scarcity, leading to unequal economic returns and finally taking the most 
exploitative form where the ‗large landlords‘ emerge as ‗water lords‘ 
through surplus accumulation forcing the small and marginal landholders to 
become landless agricultural labourers (Ballabh 2003; Sarkar 2011). This 
leads to water conflicts, which are becoming endemic at all levels in India in 
its various social, economic, and ecological dimensions (Joy et al. 2008). 
Studies in India have shown that the small landowners are also likely to go 
deeper and poach water from their neighbouring farmers by lowering the 
water table under their own lands (Foster and Rosenzweig 2008). 

The pathways through which irrigation can impact poverty reduction are 
not only direct and indirect but are often also complex and diverse. Not all 
the poor are water-poor and not all water-poor are poor. The influence of 
water and agricultural activity on poverty is strongly dependent on the 



[47] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

overall development trajectory within the basin1 (Woolley et al. 2009; Cook 
et al. 2009). The effect of water stress on livelihoods, hence poverty, can be 
felt in a variety of ways (Cook and Gichuki 2009). A lack of consensus on 
the linkages between irrigation and poverty reduction is evident: studies 
based on macro data tend not to find significant links between investment 
in irrigation and poverty, while micro-data tends to establish a robust 
relationship between access to irrigation and poverty reduction 
(Gebregziabher et al. 2009).  

Contrary to the long-prevalent consensus on Gisser-Sanchez effect (GSE) 
that states that benefits of optimally managed groundwater are insignificant, 
Koundouri (2004) found that groundwater management significantly 
increases welfare in many situations. An inverse relation between access to 
irrigation and poverty has been observed in India (Narayanamoorthy 2007; 
Sekhri 2014). Evidence from Sub-Saharan Africa shows that equitable 
distribution of income is critical for poverty reduction and that nationwide 
policy may inadvertently overlook high inequality at sub-regional and 
commune levels (Manero 2017). Participation in small-scale irrigation was 
found to reduce poverty by 50% among users relative to non-users in 
Northern Ethiopia (Zeweld et al. 2015). While irrigation for small 
landowning farmers is seen as an important vehicle to promote (especially 
rural) poverty alleviation, food security (at various scales, from local to 
global) and land and labour productivity, as well as rural employment and 
general economic development, adaptation to climate variability (Ngigi 
2009 cited in Villoth 2013), the implications of spatially heterogeneous 
resource availability are important, too. Research also needs to focus on 
differential impact on various categories of farmers as farm size is a 
significant determinant of both groundwater-irrigated farm acreage and 
groundwater irrigation application rates per unit of land area; besides, 
spatial variability in groundwater use has important implications for policies 
to correct common property externalities and for welfare maximisation 
(Koundouri 2004; Saak and Perterson 2012).  

On one hand groundwater use by millions of small landholders of 
developing countries has reduced poverty in the rural side (Shah 2005), on 
the other, its depletion and degradation have increased poverty (Janakrajan 
and Moench 2006). Studies have found a significant contribution of 

                                                           
1 The Basin Focal Projects (BFP) of the CGIAR Challenge Programme on Water and Food 
conducted large number of studies in ten river basins, including the Indo-Ganges, in various 
countries to study these relationships. 



Ecology, Economy and Society–the INSEE Journal [48] 

 

groundwater irrigation in production2 and productivity3 (Grogan et al. 2015; 
Srivastava et al. 2014). However, the direct positive impacts hinge on the 
availability of other factors like availability of hydrogeological, institutional, 
technological, and policy frameworks (Koundouri and Xepapadeas 2004; 
Janakrajan and Moench 2006; Zeweld et al. 2015; Srivastava et al. 2014; 
Woolley et al. 2009). Therefore, this is an empirical issue that needs to be 
analysed in a specific context.  

Further, the literature has focused more on the river basins in the Indo-
Gangetic Plain; there is little analysis of the peninsular region. While the 
former covers 1.09 million km2 of area, Godavari river basin covers 0.31 
million km2 of area forming 10 per cent of India‘s geographical area. 
Interestingly, despite the surface irrigation made available in these river 
basins, large share of irrigated land in all the basins depend on groundwater 
(Shah et al. 2009). In the arid and semi-arid regions, where livelihood is 
based on groundwater socio-ecologies (GwSEs), it is the demand-pull that 
leads to unsustainable extraction from relatively poor groundwater 
endowments (Shah 2005). There is a knowledge gap in how groundwater 
systems are exploited in the drier parts of South Asia and how they impact 
livelihoods (Bandyopadhyay 2007). The Basin Focal Projects of the CGIAR 
that analyse different basins indicate specific linkages (Cook et al. 2009; 
Woolley et al. 2009), which again point to the need for basin-wise studies on 
the nature of linkages. 

Against this background, this study aims to contribute to understanding of 
the relationship between access to groundwater irrigation, agricultural 
development, poverty and inequality in a multi-crop farming system in a 
river basin with heterogeneous hydrogeological and resource availability 
conditions and its implications for the resource governance. Based on the 
user recall, we estimate groundwater use at farm level from irrigation hours 
for each crop and the type of pumps used for irrigation. We define access 
to groundwater irrigation as the farming households‘ ability to irrigate their 
crops, fully or partially, either from their own irrigation wells or through 
water markets or their social and kinship networks. Here, the focus on 
‗ability‘ is intended to include a wide range of social relationships that can 
constrain or enable people to benefit from resources without focusing on 
property relations alone (Ribot and Peluso 2004). Those who have been 
identified as having access to groundwater irrigation during the reference 

                                                           
2 Grogan et al. (2015) found, using hydrological model, that mined groundwater contributed 
to 15-27% of national total crop production in China during 1981-2000.  
3 Using panel data analysis from India, Srivastava et al. (2014) concluded that irrigation 
increased yields by 15% in cereals and 3% in oilseeds.  



[49] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

agricultural year 2012-13 are referred to as ‗users‘ and others as ‗non-users‘.4 
As the past studies pointed out inequity in groundwater access by small and 
marginal farmers and those belonging to the marginalised communities 
(Shah 2009; Prakash 2005), we first confirmed whether their access had 
improved or not using our data. For analytical purposes, farmers have been 
categorized into two groups based on land size: small (less than 5 acres) and 
large (above 5 acres).  

While small farmers among the users are found to use more water per acre 
for irrigation without commensurate economic productivity, the difference 
in economic productivity among large farmers was found to exist along 
with higher inequality. While poverty headcount ratios were high among the 
upper reach farmers, depth, and severity of poverty was more among the 
middle reach farmers who were becoming agricultural labourers. Improving 
access to groundwater irrigation and adopting measures to address 
problems associated with groundwater depletion are necessary to reduce the 
gap in economic productivity of small and large farmers as well as inequality 
and poverty in the study area.  

The rest of the paper is divided into the following sections: Section 2 briefly 
describes the study area, data sources and analytical tools. Household 
income, inequality and poverty are analysed in Section 3. Groundwater 
access, utilisation and determinants are examined in Section 4.The final 
Section concludes with policy implications. 

 

2. STUDY AREA, DATA AND METHODOLOGY 

2.1. Study Area 

The study sample is drawn from the Godavari river basin (Figure 1), which 
extends over 3,12,800 km2 across the states of Maharashtra, Andhra 
Pradesh, Telangana, Chhattisgarh, Madhya Pradesh, Odisha, and small parts 
of Karnataka. Divided into 12 sub-basins by the Central Water Commission 
(CWC), the basin is characterised by high variability in the average annual 
rainfall—from 400 to 2500 mm; in the spatial distribution, nearly 15% of 
the basin area received less than 800 mm and another 12% received more 
than 1600 mm of rainfall during 1971-2005 (NRSC 2011). The estimated 
average water resource potential of the basin is 110.540 billion cubic meters 
(BCM) of which the utilisable surface water is 76.3 BCM and replenishable 
groundwater is about 40.65 BCM (CWC, 2011). Hydrogeologically, this 

                                                           
4 Since farming is considered the major groundwater-based livelihood, landless who depend 
on other groundwater-based livelihood, if any, are not considered as ‗users‘ in the study. 



Ecology, Economy and Society–the INSEE Journal [50] 

 

basin is predominantly under basalt (42.54%) and banded gneissic complex 
(17.38%) aquifer system. The upper reaches of the basin are mostly 
occupied by the Deccan Traps, the middle part is principally Archean 
granites and the Eastern Ghats dominate the lower part of the drainage 
basin and are formed mainly from the Khondalites (GoI 2011). The basin 
has agriculture (59.7%) and forests (30%) as major land uses. As much as 
46% of the net irrigated area of the basin is irrigated by ground water 
sources, 37 % by canals, 11% by tank and the rest 6% by other sources 
(CWC 2011). Other socio-economic characteristics of the basin show that 
nearly 14.3% and 16.6% of the estimated population of 75.83 million 
belongs to the socio economically marginalised Scheduled Caste (SC) and 
Scheduled Tribe (ST) communities, respectively, according to the Census of 
India 2011. The basin is also characterised by severe district level rural 
poverty, ranging from 36% to 70% (Chaudhari and Gupta 2009; Mohanty et 
al. 2016). In the typology of Shah et al. (2003), the upper reaches are marked 
by rise in the use of ground water, while the middle reaches in the study 
area fall in the early symptoms of groundwater overexploitation. 

Primary data collected from farming households in the Godavari river basin 
includes both users and non-users of groundwater. It covers Nashik, 
Karimnagar and West and East Godavari districts from Maharashtra, 
Telangana and Andhra Pradesh that belong to G1, G6, and G10 sub-basins 
or river systems corresponding to the upper, middle, and lower reaches 
(GoI 2013). as most of the socio-economic information is available at 
administrative boundary level only (Figure 1). From each of these districts, 
three mandals or sub-districts were selected based on the stage of ground 
water development, viz. overexploited (>100%), critical (70-100%), and 
safe (<70%)5. In a similar manner, three to four villages were selected from 
each of these sub-districts and a survey was conducted to list out all 
groundwater user and non-user households for in-depth survey. The 
groundwater users and non-users were randomly selected from these 
villages in the ratio of 60:40 from the census list till a pre-decided sample 
size of 100 households from each mandal and 300 for each district was 
reached. However, for the lower reaches the sample had to be limited to 
225. The final sample consisted of 825 households (494 users and 331 non-
users) covering a total population of 10,064. The household surveys were 
carried out by trained investigators using structured questionnaires and 
administered to the head of the households. User and non-user weights 

                                                           
5 The stage of groundwater development has been 58% in Nashik, 70% in Karimnagar and 
25 and 45% respectively in East and West Godavari districts with considerable variations 
across mandals in 2011 (CGWB 2014).  



[51] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

Figure 1: Location Map of Study Areas in the Godavari Basin 

 

Source: The figure is prepared based on the map given by CWC (2011) 

were used in the calculation of averages. STATA and SPSS software were 
used for data analysis. 

2.2. Methodology 

To assess poverty and inequality in the region, we estimated total household 
income, comprising both agricultural and non-agricultural incomes. Crop 
and livestock were important components of household agricultural 
income. Net returns from agriculture was estimated by taking the difference 
between gross value of output produced and the total paid out costs of 
variable inputs for each crop separately and then aggregated for all crops 
for the household. Similarly, expenses incurred for the upkeep of animals 
were deducted from gross returns (asset value of livestock not included) to 
arrive at net returns from livestock. Non-agricultural incomes from salary, 
wage or self-employment for each member of the household, social security 
allowances, and remittances from migrated family members were included 
for the reference year of the study. In addition to these, sale of water for 
drinking and irrigation purposes was also found to be a source of income 
for some of the households. Total annual household income was used to 
estimate the monthly per capita income of the household.  

Inequality in the distribution of land and incomes was measured using Gini 
coefficient for the users and non-users from small and large farmers both 
reach-wise and for the entire basin. The Gini coefficient measures the 



Ecology, Economy and Society–the INSEE Journal [52] 

 

extent to which the distribution of wealth within a group deviates from a 
perfectly equal distribution, with values from 0 to 1 (Foster et al. 2013). 
Poverty was estimated using Foster Greer and Thorbecke (1984) measures 
such as the head count ratio, poverty gap and squared poverty gap.6  

 




  






 


q

i

i

z

yz

N
P

1

1
 (1) 

In the above equation α is a measure of the sensitivity of the index to 
poverty; the poverty line is z; the value of per capita monthly income for 
the ith households is yi ; and q is the number of people who were poor or 
below poverty line. When parameter α = 0, P0 is simply the head count 
index, when α=1, the index is the poverty gap index P1, and when α=2, P2 
is the poverty severity index. This study measured poverty using all three 
measures of poverty separately for the users and non-users and by land size 
classes.7 The study reference for the poverty line was Rs 860 per capita per 
month for the state of undivided Andhra Pradesh for middle and lower 
reaches and Rs 967 per capita per month for Maharashtra for the upper 
reaches, as per the erstwhile Planning Commission of India8 for 2011-12 for 
identifying poor households (Planning Commission 2014).  

We analysed the sampled households by size of landholding and examined 
their access to groundwater across different reaches using descriptive 
statistics. This was followed by an analysis of the determinants of sample 
households‘ access to ground water using a logit model as specified here. 

ze
Y




1

1
 (2) 

where the dummy dependent variable Y represents whether farmers have 
access to groundwater or not and on the right hand side, Z is the linear 
combination of the explanatory variables X1, X2, X3,….Xk; Z=a+ b1X1+ 
b2X2+…+ bkXk; and a, b1, b2, b3…bk are the coefficients. The independent 
variables included those related to socio economic and demographic 
characteristics of the households such as family size, age, gender social 

                                                           
6 While estimating poverty and inequality, we also faced problems due to negative net 
incomes. We adopted the ‗equalisation‘ method in which individual income components 
with negative values were set to zero before computing the total income of each household 
(Sandoval and Urzua 2009). 
7 Sandoval and Urzua (2009) show that almost all poverty measures cannot cope with the 
possibility of some incomes being negative. While the problem is less relevant in the case of 
head count ratio, other measures could behave abnormally in extreme cases.  
8The Planning Commission of India was replaced by NITI Aayog by the BJP-led NDA 
government in 2014. 



[53] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

category, farm size, occupation, and literacy status of the head of the 
household, whether they own livestock, whether household head owns the 
property right to land or not, have access to credit, knowledge of aquifer, 
rainfall in the region etc. 

Water use was measured in terms of the quantity of water applied for each 
crop at various stages of crop growth and was estimated using the number 
of irrigation hours, type and horsepower of pump used, depth of the well 
etc. (Srivastava et al. 2015).9 Crop-water productivity is defined as the ratio 
of crop yield or crop value, to groundwater applied in the process of 
growing a crop. If the same crop is grown in more than one cropping 
season, then all the seasons were taken into account. In this study, we 
mostly used average economic water productivity of crop which is defined 
for individual farmer household as: 

 


N

ic i

ii
C

WA

PY
WP  (3) 

Where WPC represent he average economic water productivity of crops and 
C represents the crops from i to N. Yi is the physical yield of crop i, and Pi is 
the price realized by the farmer household, while WAi represents the water 
applied for crop i in cubic meter per acre.10 

The determinants of water use and productivity are then estimated using 
log-linear model that was found to be suitable based on the distribution. 
The independent variables represent the socio-economic characteristics of 
farm households and other factors that may influence their water use (or 
productivity), which included land, crop and irrigation related factors as well 
as other factors (Table 1). 

 

 

                                                           
9 It was assumed that one horsepower pump can pump 30 litres of water per second, and the 
efficiency of pump was taken as 83.5 %. Based on this assumption, we first estimated the 
water pumped in litres per second for a pump of a certain horsepower from a certain depth 
at the above efficiency. This was converted to cubic meter per hour depending on the 
irrigation hours. 
10 We are not estimating marginal productivity of water and therefore not addressing the 
question of optimality. 



 

Table 1: Description of Variables Included in the Regression Models (N=825) 

Variable Definition Mean Std. Dev Min Max 

User_nonuser 1 if farmer household (HH) is user, 
else ‗0‘ 

0.61 0.49 0 1 

Age_hh Age of the HH head (years) 49.11 12.84 20 90 

SC ‗1‘ if household belong to Schedule 
Caste and Schedule Tribe, else ‗0‘ 

0.29 0.46 0 1 

Gender ‗1‘ if gender of the head of the HH 
is male, else ‗0‘ 

0.93 0.26 0 1 

Fam size Number of family members 4.63 1.98 1 16 

Illiterate  ‗1‘ if HH head is illiterate, else ‗0‘ 0.33 0.47 0 1 

Land_size_class ‗1‘ if HH is a small farmer, else ‗0‘ 0.55 0.49 0 1 

Livestock_owned ‗1‘ if HH has livestock assets; else 
‗0‘  

0.65 0.48 0 1 

No_plots Number of plots owned by HH 1.45 0.79 0 6 

No_crops No. of crops cultivated by HH 1.69 0.82 0 5 

Occupation  ‗1‘ if main occupation of HH head 
is own farm agriculture; else ‗0‘ 

0.82 0.38 0 1 

Mobile_ph ‗1‘ if HH has mobile, else ‗0‘ 0.88 0.33 0 1 

Credit ‗1‘ if HH has access to credit, else ‗0‘ 0.74 0.44 0 1 

Aquifer ‗1‘ if HH has knowledge of aquifer, 
else ‗0‘ 

0.52 0.49 0 1 

Depth_well Depth of open or bore well in metres 25 35 0 201 

Dripsprinkler ‗1‘ if HH use drip or sprinkler or 
drip irrigation, else ‗0‘ 

0.07 0.26 0 1 

Energy_cost Annual energy (diesel/electricity) 
cost in Rupees 

1987 5304 0 63600 

Water use per 
acre 

Amount of water used for irrigation 
in cubic meters per acre 

2379 5375 0 73671 

Water 
productivity per 
acre 

Water productivity in Rupee per 
cubic meter of water used 

83 921 0 23900 

Dis_out_mkt Distance to nearest output market 
in kms 

16.49 8.13 1 35 

Hi_value_crops ‗1‘ if HH cultivate high value crops, 
else ‗0‘ 

0.52 0.49 0 1 

Property right ‗1‘ if farmland is in the name of 
head of HH, ‗0‘ otherwise 

0.75 0.43 0 1 

Rainfall Rainfall in millimetres per year 566 122 431 795 

Source_borewell ‗1‘ if the main source of irrigation is 
borewell, else ‗0‘ 

0.31 0.46 0 1 

Water mkts ‗1‘ if HH participates in water 
markets, else ‗0‘ 

0.08 0.28 0 1 

Source: Field Surveys  

 



[55] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

3. INEQUALITY AND POVERTY IN THE RIVER BASIN 

The households‘ average annual and monthly per capita incomes show wide 
variations in the share of agricultural and non-agricultural incomes in 
relation to the relative abundance or scarcity of groundwater resources and 
its access across the reaches (Figure 2). Higher average annual and per 
capita incomes as well as higher share of agricultural incomes have been 
observed among all users in the upper reaches and among large farmers 
with access to groundwater across other reaches. While the average annual 
household income for non-users was less than one lakh rupees, it was over 
two lakh rupees for users in the basin. Though the average incomes of non- 
user small and large farmers did not differ considerably, the same was not 
true in the case of users where the incomes of large farmers were much 
higher than that of small farmers. Statistically, the users in all the reaches of 
the river earned higher farm income as well as total household and per 
capita income vis-à-vis those who did not have access to ground water. The 
large farmers among the users also had significantly higher household and 
per capita income by virtue of better farm earnings. 

The upper reaches reported the highest average annual farm household 
income for groundwater users small (Rs 1.64 lakhs) and large (Rs 4.15 
lakhs). Around 70% of their total income came from agricultural sources. 
However, farm household income for the non-users was very low along 
with low share of agricultural income in total income for both small (25%) 
and large (38%) farmers. Interestingly, middle reaches reported the lowest 
levels of average annual household as well as per capita incomes for all, 
including large farmers for whom share of non-agricultural income in total 
income was over 40%. This part of the river basin is in the fourth stage of 
groundwater socio-ecologies (GwSE) with agrarian distress.11 This naturally 
increases the cost of water extraction as well as borewell failures, as was 
found by Narayanamoorthy (2015). In our sample, about 20% of the 
households in the middle reaches reported well failure as compared to 9% 
and 5% respectively in the upper and lower reaches. However, considerable 
difference in the average incomes between users and non-users and 
between small and large farmers was observed in the lower reaches, where 
the users had about 75% of their income from agricultural sources and non-
users had just 35% share from agricultural sources as they were not able to 
cultivate high value crops like tobacco. 

 

                                                           
11 Shah et al. (2003) explained the stages of groundwater socio-ecologies in four different 
stages. 



 Figure 2: Average Annual Household Income and Monthly Per Capita Income in the Study Area 

 

Source: Field surveys 
Note: I indicates standard errors  

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 



 

When all farmers are considered, the income inequality was found to be 
higher than the inequality in farm size in general among farmers in the 
upper reaches. Similar observation was also made in the case of users as 
compared to non-users across all reaches. While the inequality in the farm 
size averaged around 40%, per capita income inequality was 54% (Table 2). 
Contrary to what was observed in the study by Foster and Rosenzweig 
(2008), farm-size inequality was associated with higher income inequality in 
the Godavari river basin, perhaps due to the differences in the economic 
productivity of water, which we discuss in the latter section of this paper. 
The Gini coefficients of small and large farmers within the user and non-
user subgroups were at comparable levels except for per capita income of 
large farmers among the users in the upper reaches who had higher levels 
of inequality as compared to others. This could also be due to the 
difference in the age of grapevines grown by large farmers, some of which 
may not have started giving yields. The overall inequality of the users and 
non-users in the middle and lower reaches was comparable, though there 
were differences between the small and large farmers in both cases.  

Coming to the poverty levels, the estimated head count ratio shows higher 
levels of poverty in the upper reaches (33%), followed by middle (25%) and 
lower reaches (7%) with discernible difference across the user and non-user 
groups and farm size wise. These figures are comparable with district level 
poverty estimates based on NSSO data by (Mohanty et al. 2016). The head 
count ratio was relatively high for small farmers in the upper reaches 
followed by the large farmers in the middle reaches among both the users 
and non-users. The upper and middle reaches with the highest levels of 
poverty had the contrasting characteristics with respect to groundwater 
availability and access. While the availability of groundwater was better in 
the upper reaches, the status of groundwater has been far from being safe 
in most parts of the middle reaches rendering them to seek non-agricultural 
activities, mostly wage labour as their supplementary source of income. 
Moreover, the poverty gap index and squared poverty gap shows that 
poverty is both deeper and severe in the middle reaches as compared to the 
other two reaches.  

Interestingly, poverty is more severe among the users in the middle reaches, 
where farm households depend on agriculture as their major source of 
income. From the estimates of poverty gap index, it was seen that while 
14% of the poverty line amount is required to take the poor out of poverty, 
for large farmers among the users in the middle and upper reaches it was 
over 26% and 22%, respectively. Depth of poverty was also found to be 
very high among the large user farmers in the middle reaches. On the other 
hand, poverty levels were found to be very low in the lower reaches.  



Ecology, Economy and Society–the INSEE Journal [58] 

 

Table 2: Inequality in the Distribution of Land and Income (Gini Coefficients) 

Reach 
of river 

User/ 
Non-user 

Category of 
farmer 

Gini coefficient 
(in percentage) 

Poverty measures 
(in percentage) 

Land Per 
capita 
Income 

Head 
count 
ratio 

Poverty 
gap 
index  

Squared 
poverty 
gap index 

Upper  

Non-user 
Small (85) 19 36 35.3 10.1 3.9 

Large (36) 22 34 27.8 12.2 6.8 

All (121) 34 35 33.1 10.7 4.8 

User  
Small (95) 20 56 36.8 22.6 17.1 
Large (84) 29 70 28.6 18.1 14.7 

All (179) 42 65 33 20.5 16 

All 
Small (180) 39 63 36.1 16.7 10.9 

Large (120) 39 71 28.3 16.3 12.3 

All (300) 39 65 33 16.5 11.5 

Middle  

Non-
user  

Small (71) 21 30 23.9 8.5 5.7 
Large (41) 18 38 29.3 20.1 33.8 

All (112) 35 33 25.9 12.8 16 

User  
Small (94) 24 34 21.3 18 35.5 

Large (94) 24 45 28.7 26.3 35.6 

All (188) 40 40 25 22.1 35.6 

All  
Small (165) 25 37 22.4 13.9 22.7 
Large (135) 25 44 28.9 24.4 35 

All (300) 40 42 25.3 18.6 28.3 

Lower  

Non-
user  

Small (41) 18 40 2.4 1 0.4 

Large (45) 16 26 17.8 7.9 9.6 

All (86) 29 36 10.5 4.6 5.2 

User 
Small (71) 25 46 5.6 6.9 21.2 
Large (68) 24 49 4.4 2.7 1.8 

All (139) 40 44 5 4.9 11.7 

All 
Small (112) 25 46 4.5 4.8 13.6 

Large (113) 24 39 9.7 4.8 4.9 

All (225) 40 44 7.1 4.8 9.2 

All  

Non-
user  

Small (197) 43 35.5 24.4 7.6 3.8 
Large (122) 40 35.3 24.6 13.3 16.9 

All (319) 41 35.5 24.5 9.8 8.8 

User  
Small (260) 32 57.6 22.7 16.7 24.9 

Large (246) 37 57.4 22 17 19.1 

All (506) 34 57.6 22.3 16.8 22.1 

All  
Small (457) 39 54 23.4 12.8 15.8 
Large (368) 39 54 22.8 15.7 18.4 

All (825) 39 54 23.2 14.1 17 

Source: Field Survey 
Note: Figures in the parentheses are respective sample sizes. 

 

 

 



[59] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

4. GROUNDWATER ACCESS, UTILIZATION AND 
DETERMINANTS 

4.1. Water use Profile of Farmers in Selected Villages 

We will have a look at the groundwater access from the census of the 
selected villages before turning to the detailed analysis of the sampled 
households. Among the 10,064 households enumerated in the census 
survey, 37% were landless, 47%smallandthe rest large farmers (Figure 3).  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

Landlessness ranged from 20% in the upper reaches to as high as 55% in 
the lower reaches. Even when access to groundwater was primarily tied to 
land ownership in India, only 34% of the surveyed land-owning farmers 
reported access to it with considerable heterogeneity across reaches and 
categories of farm sizes. While the large farmers reported better access, 
ranging from 46% to 84% from the lower to upper reaches, small farmers 

Figure 3: Distribution of Households by Land Size-class and Access to 
Groundwater in Godavari River Basin   

 

Source: Census survey of field surveys 
Note: I indicates standard errors 

 
 
 



Ecology, Economy and Society–the INSEE Journal [60] 

 

Figure 4: Well Ownership in the Study Area by Land Size-class in Godavari River 
Basin 

Source: Field surveys 

Note:  1. I indicates standard errors 

    2. Figures in the brackets represents the following:  

First: Number of households reporting total number of wells and no. of 

wells per acre;  

Second: Number of households reporting bore well and its depth;  

Third: Number of households reporting open well and its depth.  

in the lower reaches (20%) and middle reaches (45%) had relatively poor 
access to groundwater. 

4.2. Land and Well Ownership 

The sampled households on an average owned 4.68 acres of land and 1.3 
wells (bore plus open wells) in the Godavari river basin. In other words, 
there existed one well for approximately every three and half acres of land. 
Surprisingly, each well serviced 2.236 acres in case of small farmers, while 
the same was 6.12 acres for large farmers. Across all reaches, it was seen 
that the average number of bore wells, open wells and both put together 
were higher among large farmers as compared to small farmers (Figure 4). 



[61] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

A t-test confirmed that these differences were statistically significant at one 
per cent level of significance, except in the case of open wells in the lower 
reaches. The average depth of bore wells was found to be around 220 m, 
whereas it was around 48 m for open wells in the study area. The depth of 
open wells was more for the large farmers in the upper and middle reaches, 
whereas the depth of bore wells was more only for the large farmers in the 
middle reaches where groundwater scarcity was found to be more severe. 
These differences were statistically significant. It was also seen that about 
15%of all non-users in the study area attempted digging wells at some point 
of time, and encountered problems of well failure. 

More than 80% of such well failures were reported from the middle reaches 
at an average depth of 200 m or so for bore wells. Almost 60 % of those 
who reported well failures were small farmers. This implied that as water 
levels go down, it was mostly the large farmers who are able to construct 
deeper wells and extract groundwater. In other words, the larger the land 
area owned, greater was the possibility of striking groundwater and that the 
heterogeneity in access to groundwater irrigation observed in some of the 
earlier studies (Shah 2009; Prakash 2005; Janakarajan and Moench 2006) 
exists even today. While the analysis of census data using descriptive 
statistics seem to show that the large farmers are better placed than small 
farmers in terms of their access to groundwater and ownership of wells, it 
was not clear what factors determined a household‘s access to groundwater. 

4.3. Determinants of Access to Groundwater 

An analysis of the sample survey data using the earlier specified logit model 
shows that access to groundwater is determined by a host of social, 
economic, climatic and hydrogeological factors along with the households‘ 
access to information and communication technologies (Table 3). 
Interestingly, farm size per se was not a significant variable in determining 
the access, but the access was conditioned by the factors like sources of 
non-farm income, credit facility, education, and social and caste hierarchy 
with better access among the forward communities. Households belonging 
to scheduled castes and scheduled tribes and illiterate household heads were 
less likely to have access to groundwater as compared to others. The 
number of individual plots owned by the households was a significant 
factor influencing access across all reaches. This could be due to the 
possibility that the households owning a greater number of plots might be 
owning them in diverse local hydrogeological conditions and groundwater 
endowment thus increasing the prospects of striking water when a well is 
dug. The  coefficient and  marginal effects   of the variable  for awareness of 



Ecology, Economy and Society–the INSEE Journal [62] 

 

Table 3: Determinants of Access to Groundwater in Godavari River Basin 

 Upper reaches 
(N=300) 

Middle reaches 
(N=300) 

Lower reaches 
(N=225) 

All  
(N=825) 

Variable Co-eff Marginal Co-eff Marginal Co-eff Marginal Co-eff Marginal 

Constant -5.757** 
(-1.839) 

-  -3.168** 
(-1.232) 

-  -0.41 
(-2.04) 

-  -4.10 
(-0.82) 

-  

Fam_ size 0.083 
(-0.079) 

0.193 
(-0.018) 

0.174* 
(-0.096) 

0.038* 
(-0.021) 

0.008 
(-0.110) 

0.002 
(-0.025) 

0.078** 
(-0.046) 

0.180** 
(-0.011) 

Occupa-
tion  

-0.612 
(-0.477) 

-0.134 
(-0.097) 

0.501 
(-0.352) 

0.113 
(-0.082) 

1.741*** 
(-0.636) 

.409*** 
(-0.129) 

0.332 
(-0.212) 

0.079 
(-0.051) 

Illiterate  -0.468 
(-0.430) 

-0.112 
(-0.105) 

-0.363 
(-0.301) 

-0.080 
(-0.067) 

-0.563 
(-0.373) 

-0.129 
(-0.086) 

-0.414** 
(-0.191) 

-0.097** 
(-0.045) 

SC -1.625*** 
(-0.457) 

-0.384*** 
(-0.099) 

-0.687** 
(-0.322) 

-0.156** 
(-0.075) 

-1.797*** 
(-0.538) 

-0.343*** 
(-0.078) 

-0.898*** 
(-0.185) 

-0.21*** 
(-0.043) 

Age_hh -0.002 
(-0.015) 

-0.001 
(-0.004) 

0.003 
(-0.012) 

0.001 
(-0.003) 

0.020 
(-0.015) 

0.005 
(-0.003) 

0.009 
(-0.007) 

0.002 
(-0.002) 

Gender 0.682 
(-0.728) 

0.167 
(-0.181) 

0.394 
(-0.599) 

0.090 
(-0.143) 

-0.523 
(-0.573) 

-0.110 
(-0.109) 

-0.058 
(-0.325) 

-0.013 
(-0.074) 

Land_size
_class 

0.018 
(-0.326) 

0.004 
(-0.076) 

0.251 
(-0.314) 

0.546 
(-0.069) 

0.714** 
(-0.359) 

0.169** 
(-0.079) 

0.105 
(-0.172) 

0.024 
(-0.040) 

No_plots 
.431* 
(.257) 

.100* 
(.059) 

1.413*** 
(.313) 

.306*** 
(.062) 

.344 
(.197) 

.0778546* 
(.04431) 

.530*** 
(.125) 

.122*** 
(.028) 

Property 
right 

0.397 
(-0.453) 

0.095 
(-0.111) 

-0.047 
(-0.337) 

-0.010 
(-0.072) 

0.180 
(-0.378) 

0.041 
(-0.086) 

-0.038 
(-0.198) 

-0.009 
(-0.045) 

Aquifer 1.326*** 
(-0.439) 

0.264*** 
(-0.071) 

-0.406 
(-0.309) 

-0.856 
(-0.063) 

0.328 
(-0.412) 

0.076 
(-0.097) 

0.284 
(-0.179) 

0.065 
(-0.041) 

Rainfall 0.004** 
(-0.002) 

0.001** 
(0.000) 

- - -0.002 
(-0.002) 

-0.001 
(0.000) 

0.002*** 
(-0.001) 

0.001*** 
(0.000) 

Livestock 
owned 

2.124*** 
(-0.359) 

0.486*** 
(-0.070) 

0.64** 
(-0.310) 

0.142** 
(-0.070) 

0.291 
(-0.349) 

0.066 
(-0.080) 

0.851*** 
(-0.170) 

0.200*** 
(-0.040) 

Credit 1.314*** 
(-0.304) 

0.298*** 
(-0.065) 

-0.814* 
(-0.480) 

-0.153** 
(-0.076) 

0.133 
(-0.447) 

0.031 
(-0.104) 

0.779*** 
(-0.185) 

0.186*** 
(-0.045) 

Mobile 
ph 

0.924 
(-0.744) 

0.226 
(-0.180) 

1.169*** 
(-0.456) 

0.278*** 
(-0.110) 

0.812** 
(-0.407) 

.192** 
(-0.091) 

1.076*** 
(-0.255) 

0.261*** 
(-0.061) 

 LR chi2 120.42; 
Prob > chi2=0.000  
Pseudo R =.298 
Probability (user) 
0.631 

LR chi2 79.43; 
Prob > chi2=0.000 
Pseudo R.200 
Probability (user) 
0.683 

LR chi2 61.16;  
Prob > chi2=0.000 
Pseudo R 0.204 
Probability (user) 
0.655 

LR chi2 =170.930;  
Prob > chi2=0.000 
Pseudo R 0.155 
Probability (user) 
0.639 

Source: Estimated from the field survey data 
Notes: 
1. Figures in the parenthesis are standard errors; ***, ** and * are significance levels at 
1%, 5% and 10% respectively. 
2. Co-eff indicates co-efficient; Marginal indicates Marginal effects. 

 
 



[63] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

the type of aquifer and rainfall were found to be positive and significant in 
the upper reaches. Rainfall was found to be a determinant of access in the 
upper reaches as well as in the entire basin. While access to credit was 
found to influence access to groundwater positively in the upper reaches as 
well as when all the reaches were taken together, it showed an unexpected 
negative relation in the middle reaches. It was noted that the highest 
groundwater usage in the Godavari river basin was recorded in the middle 
reaches. The access to mobile phones, which might be helping households 
to reach out to different sources of information and other support services 
like technical and credit services, was also found to be a significant factor in 
determining the household access to groundwater in the study area except 
in the upper reaches.  

4.4. Water use, Productivity and Determinants:  

The average groundwater consumption per acre of farm was the lowest in 
the upper reaches of the river basin with the highest economic water 
productivity (Figure 5). 

It is worth mentioning here that grapes, onion, soybean, maize, wheat, and 
tomato are cultivated in the upper reaches, while paddy, cotton, maize are 
the mainstay in both middle and lower reaches. Tobacco was the major 
crop for the users in the lower reaches and cotton for the non-users.12 
Owing to water intensive crops, the highest per acre water usage was found 
in the middle reaches followed by the lower reaches. Overall, the water 
usage by the small farmers was found to be significantly higher without 
significant difference in productivity. 

Before analysing the results of econometric models on the determinants, we 
present the descriptive statistics of the variables. The farm households on 
average had 4.82 members per family. The average age of the head of the 
household was 49 years (Table 1). The average farm size per household was 
4.68 acres that utilised 2379 cubic meters of water per acre. The average 
rainfall was 566mm. The depth of average well was 25m (201 m- 
maximum). A household on an average incurred an amount of Rs 1987 per 
year as cost of energy for irrigation. The sample households cultivated two 
crops on average, and they had to trudge 17km to reach the nearest market 
for selling their agricultural produce. The formulation of several binary 
variables is explained in the table. 

                                                           
12 Crop-wise water use calculated was comparable to those of Srivastava et al. (2015) with 
minor variations. 



Ecology, Economy and Society–the INSEE Journal [64] 

 

Figure 5: Water use and Economic Productivity of Groundwater for Irrigation 
in Godavari River Basin 

(a) Upper Reaches  

21
60

.3
 

45
7.

0 

12
35

.5
 

99
6.

8 

52
4.

6 75
8.

8 

44
36

.1
 

14
18

.0
 

27
98

.1
 

60
9.

3 

23
0.

3 

38
6.

6 

31
81

.5
 

15
91

.9
 

21
98

.4
 

79
7.

4 

45
7.

8 

61
2.

0 

37
57

.5
 

98
2.

6 

25
66

.1
 

14
52

.9
 

69
7.

5 98
1.

0 

62
1.

7 

19
7.

8 

32
6.

6 55
8.

6 

44
0.

9 

50
2.

0 

28.6 39.8 
125.6 101.4 

3999.6 

2531.6 

18.7 
150.7 

64.9 40.5 34.5 

-250

0

250

500

750

1000

1250

1500

1750

2000

2250

2500

2750

3000

3250

3500

3750

4000

0

1000

2000

3000

4000

5000

6000

Sm
al

l (
11

)

L
ar

ge
 (

13
)

A
ll 

(2
4)

Sm
al

l (
39

)

L
ar

ge
 (

39
)

A
ll 

(7
8)

Sm
al

l (
11

)

L
ar

ge
 (

13
)

A
ll 

(2
4)

Sm
al

l (
3)

L
ar

ge
 (

4)

A
ll 

(7
)

Sm
al

l (
10

)

L
ar

ge
 (

16
)

A
ll 

(2
6)

Sm
al

l (
36

)

L
ar

ge
 (

44
)

A
ll 

(8
0)

Sm
al

l (
25

)

L
ar

ge
 (

19
)

A
ll 

(4
4)

Sm
al

l 
(2

3
)

L
ar

ge
 (

38
)

A
ll 

(6
1)

Sm
al

l (
6)

L
ar

ge
 (

14
)

A
ll 

(2
1)

Sm
al

l (
21

)

L
ar

ge
 (

21
)

A
ll 

(4
2)

Paddy Maize Grapes Groundnut Onion Soyabean Tomato Wheat Bengal gram Bajra

E
co

n
o

m
ic

 p
ro

d
u

ct
iv

it
y 

o
f 

w
a
te

r 
p

er
 c

u
b

ic
 

W
at

er
 u

se
 p

er
 a

cr
e 

in
 c

u
b

ic
 m

et
er

s 
 

Upper reaches 

Water applied per acre Economic productivity of water per acre per cubmt

 

(b) Middle Reaches 

 

54
39

.1
 

36
92

.5
 44

58
.9

 

20
99

.7
 

38
7.

4 

12
48

.6
 

35
2.

5 

16
8.

4 

24
5.

2 6.3 

13.4 

10.3 

5.2 

14.5 

9.8 

15.9 

29.3 

23.7 

-5

0

5

10

15

20

25

30

35

40

-1000

0

1000

2000

3000

4000

5000

6000

7000

Sm
al

l (
37

)

L
ar

ge
 (

47
)

A
ll 

(8
4)

Sm
al

l (
16

)

L
ar

ge
 (

16
)

A
ll 

(3
2)

Sm
al

l (
36

)

L
ar

ge
 (

50
)

A
ll 

(8
6)

Paddy Maize Cotton

E
co

n
o

m
ic

 p
ro

d
u

ct
iv

ty
 o

f 
w

at
er

 p
er

 c
u

b
ic

 m
et

er
 

W
at

er
 u

se
 p

er
 a

cr
e 

in
 c

u
b

ic
 m

et
er

s 

Middle reaches   

Water applied per acre Economic productivity of water per acre per cubmt   

 



[65] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

(c) Lower Reaches 
 

5
3
8
3
.8

 

4
6
9
7
.4

 

5
0
4
4
.5

 

1
9
4
4
.6

 

1
3
2
6
.6

 

1
6
7
6
.1

 

1
0
4
6
.0

 

7
7
4
.6

 

8
8
7
.0

 

4
.7

 

1
1
.2

 

7
.9

 2
8
.4

 

3
7
.9

 

3
2
.5

 

1
5
5
.6

 

1
3
4
.7

 

1
4
3
.4

 

-50

0

50

100

150

200

250

0

1000

2000

3000

4000

5000

6000

7000

8000

S
m

al
l 
(8

)

L
ar

ge
 (

8
)

A
ll
 (

1
6
)

S
m

al
l 
(1

2
)

L
ar

ge
 (

9
)

A
ll
 (

2
1
)

S
m

al
l 
(3

)

L
ar

ge
 (

4
)

A
ll
 (

7
)

Paddy Maize Tobaco E
c
o

n
o

m
ic

 p
ro

d
u

c
ti

vi
ty

 o
f 

w
a
te

r 
p

e
r 

c
u

b
ic

 m
e
te

r 
in

 
ru

p
e
e
s 

w
a
te

r 
u

se
 p

e
r 

a
c
re

 i
n

 c
u

b
ic

 m
e
te

rs
 

Lower reaches   

Water applied per acre Economic productivity of water per acre per cubmt

 

(d) All Reaches 
 

3
4
5
2
.4

 

1
7
7
5
.2

 2
6
6
9
.1

 

4
7
8
2
.5

 

3
6
0
3
.6

 

4
1
8
5
.3

 

4
4
2
7
.0

 

3
8
3
4
.0

 

4
1
4
7
.9

 

3
9
6
2
.9

 

2
6
0
1
.7

 

3
3
0
9
.0

 

1
5
2
.4

 

4
3
8
.8

 

2
8
6
.1

 

2
0
.2

 

4
0
.2

 

3
0
.3

 

2
9
.2

 

4
5
.4

 

3
6
.8

 9
9
.1

 

2
6
3
.7

 

1
7
8
.2

 

0

100

200

300

400

500

600

700

800

0

1000

2000

3000

4000

5000

6000

7000

S
m

al
l 
(9

4
)

L
ar

g
e 

(8
3
)

A
ll
 (

1
7
7
)

S
m

al
l 
(9

4
)

L
ar

g
e 

(9
4
)

A
ll
 (

1
8
8
)

S
m

al
l 
(7

1
)

L
ar

g
e 

(6
7
)

A
ll
 (

1
3
8
)

S
m

al
l 
(2

5
9
)

L
ar

g
e 

(2
4
4
)

A
ll
 (

5
0
3
)

Upper Middle Lower All

E
c
o

n
o

m
ic

 p
ro

d
u

c
ti

v
it

y
 o

f 
w

a
te

r 
in

 R
s 

p
e
r 

c
u

b
ic

 m
e
te

r 

W
a
te

r 
u

se
 p

e
r 

a
c
re

 i
n

 c
u

b
ic

 m
e
te

rs
 

 All reaches  

water use per acre Economic productivity of water per cubic meter

 

Source: Field surveys 
Note: I indicates standard errors 

 



Ecology, Economy and Society–the INSEE Journal [66] 

 

Using the regression model specified earlier, the determinants of water use 
and productivity showed that farm size was inversely related with water use 
per acre and directly with economic productivity of water in the sample as a 
whole and also in all the reaches of the river basin (Table 4). In other 
words, large farmers apply significantly lesser amount of water per acre and 
attain higher levels of economic productivity as compared to small farmers. 
This was a very important result to observe when we consider the fact that 
majority of the farms in India are small and marginal. The sign and 
significance of the coefficient of the variable for number of crops cultivated 
indicates that while  addition of  crops results in  additional water  usage,  it 

Table 4: Determinants of Per Acre Water Use and Productivity in the Study Area 
(contd.) 

 Upper reaches (N=172) Middle reaches (N=170) 

Variable Water use Water 
productivity 

Water use Water 
productivity 

Constant 7.146***  
(0.909) 

1.679*  
(0.966) 

7.270*** 
(0.628) 

2.368***  
(0.649) 

Fam_ size 0.005 (0.043) -0.026 (0.045) -0.044 (0.047) 0.020 (0.049) 

Occupation  -0.248 (0.277) 0.022 (0.294) -0.266 (0.253) 0.397 (0.262) 

Illiterate  0.073 (0.261) -0.049 (0.277) 0.108 (0.183) 0.049 (0.190) 

SC 0.166 (0.345) -0.573 (0.367) 0.155 (0.208) -0.273 (0.215) 

Age_hh -0.016* (0.009) 0.020** (0.010) -0.009 (0.007) 0.001 (0.007) 

Gender -  - - 

Land_size_class 0.628*** (0.204) -0.660*** (0.216) 0.196 (0.177) -0.325* (0.183) 

No_plots 0.006 (0.117) -0.105 (0.124) 0.118 (0.097) -0.140 (0.100) 

Rainfall 0.000 (0.001) 0.001 (0.001) - - 

Livestockowned -0.222 (0.265) 0.198 (0.282) -0.102 (0.208) 0.0451 (0.215) 

Credit 0.235 (0.188) -0.061 (0.199) -0.487** (0.250) 0.423* (0.262) 

Water_mkts 0.472 (0.393) -0.231 (0.418) -0.326 (0.345) 0.326 (0.356) 

No_crops 0.609*** 
(0.102) 

-0.190*  
(0.108) 

1.170*** 
(0.163) 

-0.803*** 
 (0.169) 

Drip sprinkler 0.380 (0.324) 0.268 (0.344) - - 

Source_borewell 0.563 (0.322) -1.155*** (0.342) 0.397** (0.199) -0.325 (0.206) 

Energy_ cost 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) -0.000 (0.000) 

Depth_ well -0.030***  
(0.004) 

0.035***  
(0.005) 

-0.014*** 
(0.002) 

0.013***  
(0.002) 

Dis_out_mkt -0.008 (0.010) -0.023** (0.011) 0.014 (0.011) 0.000 (0.011) 

Hi_value_crops 0.083 (0.248) 0.845*** (0.264) -0.572 (0.223)** 1.048*** (0.231) 

 
F (18, 153) = 9.42  
Prob>F = 0.000 
Adj R2  = 0.470 

F (18, 153) = 10.46;  
Prob>F = 0.000 
Adj R2  = 0.499 

F (16, 153) = 7.62  
Prob>F = 0.000 
Adj R2  = 0.385 

F (16, 153) = 6.81  
Prob>F = 0.000 
Adj R2   = 0.355 

 
 
 



[67] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

Table 4: Determinants of Per Acre Water Use and Productivity in the Study Area 
(from pre-page) 

 Lower reaches (N=127) All (N=469) 

Variable Water use Water 
productivity 

Water use Water 
productivity 

Constant 7.895*** (0.930) 2.396** (1.061) 6.975*** (0.518) 1.739*** (0.559) 

Fam_ size -0.002 (0.043) -0.011 (0.049) -074*** (0.028) 0.054* (0.030) 

Occupation  - - -0.113 (0.164) 0.058 (0.177) 

Illiterate  0.299* (0.160) -0.220 (0.182) 0.347*** (0.130) -0.260* (0.141) 

SC -0.099 (0.167) -0.163 (0.190) 0.146 (0.128) -0.253* (0.138) 

Age_hh -0.002 (0.006) 0.008 (0.006) -0.013*** (0.005) 0.013** (0.005) 

Gender 0.240 (0.231) -0.140 (0.263) 0.033 (0.236) -0.150 (0.254) 

Land_size_class 0.516*** (0.139) -0.605*** (0.158) 0.377*** 0.114) -0.441*** (0.123) 

No_plots 0.084 (0.072) -0.261*** (0.082) 0.152** (0.062) -0.280*** (0.067) 

Rainfall -0.004*** (0.002) 0.004*** (0.002) 0.001 (0.001) 0.002*** (0.001) 

Livestockowned 0.016 (0.144) 0.065 (0.164) -0.113 (0.131) -0.032 (0.141) 

Credit -0.353* (0.211) 0.261 (0.241) 0.231* (0.135) -0.083 (0.145) 

Water_mkts -0.276* (0.171) 0.364* (0.195) 0.089 (0.184) 0.106 (0.199) 

No_crops 1.093* (0.124) -0.804*** (0.141) 0.609*** (0.070) -0.280*** (0.075) 

Drip sprinkler .0191 (0.232) -0.093 (0.265) 0.186 (0.184) 0.181 (0.198) 

Source_borewell 0.334 (0.251) -0.186 (0.286) 0.371*** (0.137) -0.338** (0.148) 

Energy_ cost 0.000 (0.000) 0.000 (0.000) -0.000*** (0.000) 0.000*** (0.000) 

Depth_ well -0.035*** (0.005) 0.038*** (0.006) -0.019*** (0.001) 0.0186*** (0.002) 

Dis_out_mkt 0.095** (0.030) -0.115*** (0.034) -0.002 (0.007) -0.009 (0.007) 

Hi_value_crops -0.404 (0.172) 0.873*** (0.197) 0.109 (0.123) 0.623*** (.0133) 

 
F (18, 108) = 12.87 

Prob>F = 0.000 
Adj R2 = 0.629 

F (18, 108) = 11.97  
Prob>F = 0.000  
Adj R2  = 0.611 

F (19, 449) = 15.14  
Prob>F = 0.000  

AdjR2 =0.365 

F (19, 449) =14.55  
Prob>F = 0.000 

Adj R2 = 0.355 

Source: Estimated from the field survey data  
Note: Figures in the parenthesis are standard errors; ***, ** and * are significance 
levels at 1%, 5% and 10% respectively 

 

results in decline of productivity. This finding was consistent across all 
reaches of the basin. It implies that specialization might help in reducing 
water usage per acre and increasing economic productivity of water. 

The economic productivity of those who cultivated high value crops 
including vegetables and fruits was found to be better in all reaches. Access 
to borewell and subsidized power for water lifting devices encouraged 
elevated water use with lower productivity. In other words, higher the user 
cost of power, lower the water use and better utilization. Participation in 
water markets did not impact water use or water productivity in the sample. 
The coefficient for water productivity for modern irrigation technologies 
was insignificant. This could possibly be because these technologies are 



Ecology, Economy and Society–the INSEE Journal [68] 

 

used mostly in the upper reaches for grape gardens and many of them are 
still in the vegetative phase. The difficulty in accessing the markets in the 
lower reaches of the river basin seems to have resulted in higher water 
usage and lower water productivity. The vibrant water markets in the lower 
reaches resulted in lower water use with higher economic productivity. 

As expected, those with deeper wells in all the reaches of the river basin 
used less water and achieved higher productivity. As water extraction from 
deeper depths becomes more difficult, the users tend to extract less and use 
it better. In other words, the scarcity of the resource might be driving them 
to efficient resource utilisation and better returns. 

 

5. CONCLUSIONS AND POLICY IMPLICATIONS 

This study examined the access to groundwater for irrigation and its impact 
on agriculture, poverty, and inequality against the backdrop of 
heterogeneous conditions of groundwater availability in the Godavari basin 
and its implications for resource governance. It uses data collected from 
825 farm households that included 494 groundwater users in the upper, 
middle and lower reaches of the basin. Data was analysed using various 
statistical and econometric methods that yielded important results at both 
aggregate (basin level) and disaggregate levels. At the aggregate level, large 
farmers were found to have better access to groundwater vis-à-vis small 
farmers. Households with access to groundwater earned higher household 
and per capita income compared to those without access. However, 
reaches-wise disaggregate level analysis, including those done separately for 
farm size as well as user-non-user categories, shows that in addition to 
access to groundwater, the amount of value that the farmers tend to 
generate out of groundwater irrigation is an important factor in determining 
household income or poverty levels. It may be noted that farmers with 
access to groundwater expectedly had a higher share of agricultural income 
to total income but also had higher levels of inequality. The poverty 
headcount ratios were at higher levels in the upper reaches, which had 
better groundwater access and availability conditions. However, the poverty 
levels were both severe and deeper in the middle reaches where 
groundwater scarcity conditions were more widespread as compared to 
others. The experience of the middle reaches also shows that as depletion 
or scarcity of groundwater increases, it is mostly the large farmers who tend 
to reap better productivity. The small farmers drift towards non-agricultural 
sources of income including wage labour that can be termed distress-driven 
diversification. 



[69] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

A detailed analysis of the determinants of access to groundwater showed 
that access was determined by a host of social, economic, climatic and 
hydrogeological factors along with households‘ access to information and 
communication technologies. Even though farm size per se was not found to 
be a significant variable in determining access, it was a significant variable in 
determining per acre groundwater usage as well as economic productivity of 
water. Basin level analysis showed that farm size was inversely related to 
water use per acre and directly related to economic productivity. The higher 
per acre usage of water by small farms without commensurate economic 
productivity was perhaps a major concern for a country like India where 
majority of farms are small and marginal. Reaches-wise disaggregate level 
analysis pointed to the lowest groundwater usage per acre by the farmers in 
the upper reaches with the highest levels of economic productivity in 
contrast to the middle reaches with high per acre water usage irrespective of 
farm size and low levels of economic productivity of water. These 
differences could be due to the choices of crops adopted by farmers, 
including the number and types of crops grown. The upper reaches were 
characterized by a diverse variety of crops, which include high-value and 
less water consuming crops in contrast to the high-water consuming, less 
diverse and low-value crops in the middle reaches. It was observed that in 
the upper reaches mere addition of a number of crops led to higher per acre 
water usage and a decline in economic productivity, thus pointing towards 
the need for crop specialization to reduce per acre water usage to spur 
economic productivity of water. While ownership of borewells had a 
differential impact on water use and productivity across reaches, the depth 
of borewells, an indicator of groundwater depletion, was found to be 
significant in determining water use and productivity. Farmers seem to 
adopt measures to economize water use and improvements in productivity 
when they face actual scarcity situation rather than when they just foresee it 
happening. The case of lower reaches showed that difficulty in accessing 
output markets often resulted in higher water usage and lower water 
productivity, whereas vibrant water markets tend to help in reducing water 
use and increasing economic productivity of water.  

To sum up, this study attempted to throw some light on the complex and 
diverse relationship between groundwater irrigation, agriculture, poverty 
and inequality under heterogeneous conditions of groundwater availability 
in the Godavari river basin. Understanding and incorporating them in 
policy are of paramount importance not only for the conservation of 
groundwater resources but also to reduce the gap in economic productivity 
of water and reduction of poverty and inequality across different size classes 
of farmers in the study region. While the overall policy should be directed 



Ecology, Economy and Society–the INSEE Journal [70] 

 

towards more equitable access to groundwater resources, measures to 
economise groundwater use per farm acre as well as improvement in 
economic productivity across regions and size classes of farmers, 
considering regional specificities, are equally important. Otherwise, 
situations like middle reaches will be more widespread, and as Manero 
(2017) points out, national policy may inadvertently overlook the complex 
relationship of agriculture, water use, economic farm productivity, and 
poverty, which is critical for the reduction of poverty and inequality and for 
the promotion of agricultural growth and development at the sub-regional 
levels.  

 

ACKNOWLEDGEMENTS 

This study was funded by Jamsetji Tata Trust through erstwhile Research 
Unit for Livelihoods and Natural Resources (RULNR) of the Centre for 
Economic and Social Studies, Hyderabad. The suggestions by the two 
anonymous reviewers of the journal helped in improving the analysis, and 
the authors sincerely thank them. The authors are also grateful to 
Himanshu Kulkarni for his inputs on hygrogeological data for the selection 
of field sites. The assistance by V Srinivas is gratefully acknowledged. The 
authors alone are responsible for any errors in the paper. 

 

REFERENCES 

Ballabh,V. 2003. ―Politics of Water Management and Sustainable Water Use.‖ 
Indian Journal of Agricultural Economics 58 (3): 467-77. 

Bandyopadhyay, J. 2007. ―Water Systems Management in South Asia: Need for a 
Research Framework.‖ Economic and Political Weekly 42 (10): 863-873. 

Chaudhuri, S., and N. Gupta. 2009. ―Levels of living and poverty patterns: A 
district-wise analysis for India.‖ Economic and Political Weekly 44 (9): 94-110. 

Chopra, K. 2003. ―Sustainable Use of Water: The Next Two Decades.‖ Economic 
and Political Weekly 38 (32): 3360-3365. 

Cook, S.E., M. J. Fisher, M. S. Andersson, J. Rubiano, and M. Giordano.2009. 
―Water, food and livelihoods in river basins.‖ Water International 34 (1): 13-29. 
https://doi.org/10.1080/02508060802673860 

Cook, S. and F. Gichuki. 2006. ―Mapping Water Poverty: Water, Agriculture and 
Poverty Linkages.‖ BFP Working Paper 3. Colombo: Challenge Program on Water 
and Food.  

https://doi.org/10.1080/02508060802673860


[71] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

Central Water Commission. 2011. Water and Related Statistics. Water Planning and 
Project Wing. New Delhi: Central Water Commission. 

CGWB. 2014. Dynamic Groundwater Resources in India. New Delhi: Government of 
India, Ministry of Water Resources, River Development and Ganga Rejuvenation.  

Dhawan, B. D. 1989. Studies in Irrigation and Water Management. New Delhi: 
Commonwealth Publishers. 

Foster, A., and M. Rosenzweig. 2008. ―Inequality and the Sustainability of 
Agricultural Productivity Growth: Groundwater and the Green Revolution in Rural 
India.‖ Paper presented at the India Policy Conference at Stanford University, 
California, June 1–3, 2005. 

Foster, J., J. Greer, and E. Thorbecke. 1984. ―A Class of Decomposable Poverty 
Measures.‖ Econometrica 52 (3): 761-65. https://doi.org/10.2307/1913475 

Foster, J., S. Seth, M. Lokshin, and Z. Sajaia. 2013. A Unified Approach to Measuring 
Poverty and Inequality: Theory and Practice. Washington DC: The World Bank. 
https://doi.org/10.1596/978-0-8213-8461-9 

Gebreegziabher, T., J. Nyssen, B. Govaerts, F. Getnet, M. Behailu, M. Haile, and J. 
Deckers. 2009. ―Contour Furrows for in Situ Soil and Water Conservation, Tigray, 
Northern Ethiopia.‖ Soil and Tillage Research 103 (2): 257-64. 
https://doi.org/10.1016/j.still.2008.05.021 

GoI. 2011. Godavari River system. Ministry of Water Resources, River Development, 
and Ganga Rejuvenation. New Delhi: Government of India. Available at: 
http://india-wris.gov.in/wrpinfo/index.php?title=Godavari. Accessed on 
September 3, 2013. 

GoI. 2013. Godavari Basin Report. Ministry of Water Resources, River Development, 
and Ganga Rejuvenation. New Delhi: Government of India. Available at: 
http://www.india-
wris.nrsc.gov.in/Publications/BasinReports/Godavari%20Basin.pdf. Accessed on 
October 1, 2013. 

GoI. 2014. Ground Water Scenario of India 2013-14. Central Ground Water Board, 
Ministry of Water Resources. Faridabad: Government of India. 

GoI. 2017. Ground Water Year Book India 2016-17. Central Ground Water Board, 
Ministry of Water Resources. Faridabad: Government of India. 

Grogan, D.S., F. Zhang, A. Prusevich, R.B.Lammers, D. Wisser, S. Glidden, C. Li 
and S. Frolking. 2015. ―Quantifying the link between crop production and mined 
groundwater irrigation in China.‖ Science of the Total Environment 511: 161-175. 
https://doi.org/10.1016/j.scitotenv.2014.11.076 

Hussain, I., and M. Hanjra. 2004. ―Irrigation and Poverty Alleviation: Review of the 
Empirical Evidence.‖ Irrigation and Drainage 53 (1): 1-15. 
https://doi.org/10.1002/ird.114 

https://doi.org/10.2307/1913475
https://doi.org/10.1596/978-0-8213-8461-9
https://doi.org/10.1016/j.still.2008.05.021
http://india-wris.gov.in/wrpinfo/index.php?title=Godavari
http://www.india-wris.nrsc.gov.in/Publications/BasinReports/Godavari%20Basin.pdf
http://www.india-wris.nrsc.gov.in/Publications/BasinReports/Godavari%20Basin.pdf
https://doi.org/10.1016/j.scitotenv.2014.11.076
https://doi.org/10.1002/ird.114


Ecology, Economy and Society–the INSEE Journal [72] 

 

Janakarajan, S., and M. Moench. 2006. ―Are Wells a Potential Threat to Farmers‘ 
Well-being? Case of Deteriorating Groundwater Irrigation in Tamil Nadu.‖ 
Economic and Political Weekly 41 (37): 3977-3987. 

Joy, K.J., B. Gujja, S. Paranjape, V.V.Goud, and S. Vispute, ed. 2008. Water Conflicts 
in India: A Million Revolts in the Making. New Delhi: Routledge. 

Koundouri, P. 2004. ―Current Issues in the Economics of Groundwater Resource 
Management‖. Journal of Economic Surveys 18 (5): 703-740. 
https://doi.org/10.1111/j.1467-6419.2004.00234.x 

Koundouri, P. and A. Xepapadeas. 2004. ―Estimating Accounting Prices for 
Common Pool Natural Resources: A Distance Function Approach.‖ Water Resource 
Research 40 (6). https://doi.org/10.1029/2003WR002170 

Kulkarni, H., M. Shah, and P.S. Vijay. 2015. ―Shaping the Contours of 
Groundwater Governance in India.‖ Journal of Hydrology 4 (Part A): 172-192. 
https://doi.org/10.1016/j.ejrh.2014.11.004 

Kulkarni, H. and P.S.V. Shankar. 2014. ―Groundwater resources in India: an arena 
for diverse competition.‖ Local Environment 19 (9): 990-1011. 
https://doi.org/10.1080/13549839.2014.964192 

Kulkarni, H., and H.Thakkar. 2012. ―Framework for India‘s strategic water 
resource management under a changing climate.‖ In Handbook on Climate Change and 
India: Development, Politics and Governance. edited by N. K. Dubash, 328–340. New 
Delhi: Oxford University Press. 

Manero, A. 2017. ―Income Inequality within Smallholder Irrigation Schemes in 
Sub-Saharan Africa.‖ International Journal of Water Resources Development 33 (5): 770–
87. https://doi.org/10.1080/07900627.2016.1152461 

Mohanty, S. K., D. Govil, R. K. Chauhan, and S.V. Subramanian. 2016. ―Estimates 
of Poverty and Inequality in the Districts of India 2011–2012.‖ Journal of Development 
Policy and Practice 1 (2): 1–61. https://doi.org/10.1177/2455133316642338 

Narayanamoorthy, A. 2007. ―Does Groundwater Irrigation Reduce Rural Poverty?: 
Evidence from Indian States.‖ Irrigation and Drainage 56 (2-3): 349-362. 
https://doi.org/10.1002/ird.307 

Narayanamoorthy, A. 2015. ―Groundwater Depletion and Water Extraction Cost: 
Some Evidence from South India.‖ International Journal of Water Resources Development 
31 (4): 604-617. https://doi.org/10.1080/07900627.2014.935302 

NRSC. 2011. Assessment of Water Resources at Basin Scale using Space Inputs: A Pilot 
Study. Hyderabad: National Remote Sensing Centre. 

Planning Commission. 2014. Report of the Expert Group to Review the Methodology for 
Measurement of Poverty. New Delhi: Government of India. 

https://doi.org/10.1111/j.1467-6419.2004.00234.x
https://doi.org/10.1029/2003WR002170
https://doi.org/10.1016/j.ejrh.2014.11.004
https://doi.org/10.1080/13549839.2014.964192
https://doi.org/10.1080/07900627.2016.1152461
https://doi.org/10.1177/2455133316642338
https://doi.org/10.1002/ird.307
https://doi.org/10.1080/07900627.2014.935302


[73] Jeena T Srinivasan and Chandra Sekhara Rao Nuthalapati 

 

Prakash, A. 2005. The Dark Zone: the Groundwater Irrigation, Politics and Social Power in 
North Gujarat. Wageningen University Water Resources Series, Hyderabad: Orient 
Longman. 

Ribot, J.C., and N.L. Peluso. 2009. ―A Theory of Access‖. Rural Sociology 68 (2): 
153–181. https://doi.org/10.1111/j.1549-0831.2003.tb00133.x 

Saak, A.E., and J.M. Peterson. 2012. ―Groundwater Pumping by Heterogeneous 
users.‖ Hydrogeology Journal 20 (5): 835–849. https://doi.org/10.1007/s10040-012-
0854-2 

Saleth, R. M. 1996. Water Institutions in India: Economics, Law and Policy. New Delhi: 
Commonwealth Publishers. 

Sandoval, H.H., and C.M. Urzua. 2009. ―Negative Net Incomes and the 
Measurement of Poverty: A Note.‖ Journal of Managerial Financial Economics  3 (1): 
29–36. 

Sarkar, A. 2011. ―Socio economic Implications of Depleting Groundwater 
Resources in Punjab: A Comparative Analysis of Different Irrigation Systems.‖ 
Economic and Political Weekly 12 (7): 59-66. 

Sekhri, S. 2014.―Wells, Water and Welfare: Impact of Access to Groundwater on 
Rural Poverty and Conflict.‖ American Economic Journal:  Applied  Economics   6 (3): 
76-102. https://doi.org/10.1257/app.6.3.76 

Shah, T., A.D. Roy, A.S. Qureshi, and J.Wang. 2003. ―Sustaining Asia‘s 
Groundwater boom: An Overview of Issues and Evidence.‖ Natural Resources Forum 
27 (2): 130-141. https://doi.org/10.1111/1477-8947.00048 

Shah, T., J Burke, K. Villholth, M. Angelica, E. Custodio, F. Daibes, J. 
Hoogesteger, et al. 2007. ―Groundwater: A Global Assessment of Scale and 
Significance.‖ In Water for food, Water for life: A Comprehensive Assessment of Water 
Management in Agriculture, edited by D Molden. London, UK/Colombo, Sri Lanka: 
Earthscan.  

Shah, T. 2009. ―Climate Change and Groundwater: India‘s Opportunities for 
Mitigation and Adaptation.‖ Environmental Research Letters 4 (3):1-13. 
https://doi.org/10.1088/1748-9326/4/3/035005 

Shah, T., O.P. Singh, and A. Mukherji. 2006. ―Groundwater Irrigation and South 
Asian Agriculture: Empirical Analyses from a Large-scale Survey of India, Pakistan, 
Nepal Terai and Bangladesh.‖ Hydrogeology Journal 14 (3): 286-309. 
https://doi.org/10.1007/s10040-005-0004-1 

Shah, T. 2005. ―Groundwater and Human Development: Challenges and 
Opportunities in Livelihoods and Environment.‖ Water Science and Technology 51 (8): 
27-37. https://doi.org/10.2166/wst.2005.0217 

https://doi.org/10.1111/j.1549-0831.2003.tb00133.x
https://doi.org/10.1007/s10040-012-0854-2
https://doi.org/10.1007/s10040-012-0854-2
https://doi.org/10.1257/app.6.3.76
https://doi.org/10.1111/1477-8947.00048
https://doi.org/10.1088/1748-9326/4/3/035005
https://doi.org/10.1007/s10040-005-0004-1
https://doi.org/10.2166/wst.2005.0217


Ecology, Economy and Society–the INSEE Journal [74] 

 

Shah, T., M. Hassan, M.Z.Khattak, P.S. Banerjee and O.P.Singh. 2009. ―Is 
Irrigation Water Free? A Reality Check in the Indo-Gangetic Basin.‖ World 
Development 37 (2): 422-434. https://doi.org/10.1016/j.worlddev.2008.05.008 

Srivastava, S.K., R. Chand, S. S. Raju, R. Jain, I. Kingsly, J. Sachdeva, J. Singh and 
A. P. Kaur. 2015. ―Unsustainable Groundwater Use in Punjab Agriculture: Insights 
from Cost of Cultivation Survey‖.  Indian  Journal of Agricultural Economics 70 (3): 
365-378.  

Srivastava, S.K., S. Ghosh, A. Kumar and P. S. Brahmanand. 2014. ―Unravelling 
the Spatio-Temporal Pattern of Irrigation Development and Its Implication on 
Indian Agriculture.‖ Irrigation and Drainage 63 (1): 1-11. 
https://doi.org/10.1002/ird.1779 

Villholth, K.G. 2013. ―Groundwater Irrigation for Smallholders in Sub-Saharan 
Africa – a Synthesis of Current Knowledge to Guide Sustainable Outcomes.‖ Water 
International 38 (4): 369-391. https://doi.org/10.1080/02508060.2013.821644 

Woolley, J., S. E. Cook, D. Molden, and L. Harrignton. 2009. ―Water, Food and 
Development: the CGIAR Challenge Program on Water and Food‖. Water 
International 34 (1): 4-12. https://doi.org/10.1080/02508060802666179 

World Bank. 2010. Deep Wells and Prudence: Towards Pragmatic Action for Addressing 
Groundwater Overexploitation in India. Washington DC: The World Bank.  

Zaveri, E., D. S. Grogan, K. Fisher-Vanden, S. Frolking, R. B. Lammers, D. H. 

Wrenn, A. Prusevich, and R. E. Nicholas. 2016. ―Invisible Water, Visible Impact : 
Groundwater use and Indian Agriculture under Climate Change.‖ Environment 
Research Letters 11 (8):1-13. https://doi.org/10.1088/1748-9326/11/8/084005 

Zeweld, W., G. van Huylendbroeck, A. Hidgot, M.G. Chandrakanth and S. 
Speelman. 2015. ―Adoption of Small-Scale Irrigation and Its Livelihoods Impacts 
in Northern Ethiopia.‖ Irrigation and Drainage 64 (5): 655-668. 

https://doi.org/10.1016/j.worlddev.2008.05.008
https://doi.org/10.1002/ird.1779
https://doi.org/10.1080/02508060.2013.821644
https://doi.org/10.1080/02508060802666179
https://doi.org/10.1088/1748-9326/11/8/084005

