Diversification and Livelihood Options A Study of
Diversification and Livelihood Options A Study of
Diversification and Livelihood Options A Study of
net/publication/228797754
CITATIONS READS
34 268
4 authors, including:
Rachel Slater
Centre for International Development and Training (CIDT), University of Wolverhampton
40 PUBLICATIONS 1,755 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Rachel Slater on 02 December 2014.
Y. Mohan Rao
Rachel Slater
December 2002
This report is based on research funded by the UK Department for International Development
(DFID). The opinions expressed in the paper are those of the authors and do not necessarily reflect
the views of DFID.
Uttam Kumar Deb, G.D. Nageswara Rao and Y. Mohan Rao are Scientific Officers on the
Socioeconomic and Policy Programme at the International Crop Research Institute for the Semi-
Arid Tropics. They can be contacted at:
Email: sepp@ICRISAT.org
Rachel Slater is Research Officer in the Rural Policy and Environment Group at the Overseas
Development Institute.
Email: r.slater@odi.org.uk
ii
Contents
Glossary of Terms v
Acronyms v
Abstract vi
1 Introduction 1
1.1 Understanding the livelihoods of the rural poor: analytical concepts 1
1.2 Objectives of the study 2
2 Methods, Context and Location of Study 4
2.1 Participatory Rapid Appraisal 4
2.2 Household census 4
2.3 Household survey 4
2.4 Longitudinal panel study 6
2.5 Aurepalle and Dokur: the study villages 6
3 Dynamics of Livelihood Options 11
3.1 Quantifying diversity in Livelihood Options: who diversifies? How much? 11
3.2 Explaining the driving forces that are behind diversification 16
4 Income, Inequality and Economic Mobility in Aurepalle and Dokur 26
5 Concluding Remarks 33
References 35
Appendix 1 Household Census Schedule for ODI/ICRISAT Collaborative Project on
Livelihood Options 37
Appendix 2 Household Survey Questionnaire for ODI/ICRISAT Collaborative Project
on Livelihood Options 39
Appendix 3 Summary of Government Programmes on Rural Livelihood in Andhra
Pradesh 47
List of Tables
Table 1 Farmsize classification based on operational holding (ha) in the study villages 5
Table 2 Distribution of sample households covered in the household survey, 2001 6
Table 3 Basic features of Aurepalle and Dokur villages, 1975-2001 8
Table 4 Distribution of all households (by caste) in Aurepalle and Dokur villages of Andhra
Pradesh, 2001 9
Table 5 Agro-climatic, socio-economic and technological features of the study villages
1975–8, 1989–90 and 2000–1 10
Table 6 Distribution of households of different farm size categories, according to number of
sources of income in Aurepalle, 1975 and 2001 12
Table 7 Distribution of households of different farm size categories according to number of
sources of income in Dokur, 1975 and 2001 12
Table 8 Distribution of households of different caste categories according to number of
sources of income in Aurepalle and Dokur, 2001 12
Table 9 Percentage of agricultural income and non-farm income to net household income,
1975–6 to 2001 13
Table 10 Primary occupation wise distribution of households (HH) in Aurepalle and Dokur,
1975, 1989 and 2001 14
iii
Table 11 Secondary occupation wise distribution of households (HH) in Aurepalle and Dokur,
1975, 1989 and 2001 15
Table 12 Dependency level of households to different sources of livelihood in Aurepalle
and Dokur, 2000–1 15
Table 13 Farmer’s perception of rainfall, climate and irrigation, 1989–2001 17
Table 14 Wells in Aurepalle and Dokur, 1989 to 2001 18
Table 15 Livestock in Aurepalle and Dokur, 1989 to 2001 18
Table 16 Numbers of tractors in Aurepalle and Dokur between 1989 and 2001 18
Table 17 Average household size 19
Table 18 Level of educational attainment by landholding class in Aurepalle, 1975 19
Table 19 Level of educational attainment by landholding class in Dokur, 1975 and 2001 20
Table 20 Average outstanding loans per household in Rupees 21
Table 21 Seasonal migration by caste in Aurepalle and Dokur, 2000–1 25
Table 22 Total Income from all sources over cropping year (Rs) in 1975–8, 1989–90 and
2000–1 26
Table 23 Distribution (%) of households across net income group in the study villages,
1975–8, 1989–90 and 2000–1 27
Table 24 Degree of inequality in the distribution of per capita income in Aurepalle and
Dokur, 2001 28
List of Boxes
Box 1 Labourer turned large farmer 29
Box 2 Debts and deterioration 30
Box 3 Treading water 31
iv
Glossary of Terms
Acronyms
v
Abstract
The diversification of rural livelihoods is the subject of a growing amount of conceptual and policy-
based research. This paper reports on the findings from a re-survey and longitudinal panel survey
carried out in the villages of Aurepalle and Dokur in Mahbubnagar District in Andhra Pradesh,
India. This is a particularly valuable data source since these villages have been surveyed at intervals
by ICRISAT since 1975 and have enabled an analysis of changing rural livelihoods over time.
Agriculture remains the most important source of livelihood in both villages, though the relative
importance of crop cultivation has decreased, as has real income from crops. Agriculture has
become an increasingly risky pursuit and households have sought other sources of income, most
notably through migration for agricultural labour in other villages or for wage labour in urban areas
such as Hyderabad.
Whilst there are a small number of cases where diversification has enabled households to lift
themselves significantly above the poverty line, the overwhelming experience of diversification is
as a coping strategy. Mahbubnagar District experienced drought in 1997–8 and between 1999 and
2001. The intervening years were characterised by only average rainfall. It remains to be seen,
therefore, whether the diversification into non-farm activities is a short-term response to adverse
agricultural terms of trade and ecological uncertainty brought about as a result of extended drought
or whether diversification represents a long-term move away from agricultural livelihoods in rural
areas that will be sustained. The prospects for a return to agriculture in the future will be diminished
if population density continues to rise and limited by the gradual erosion of agricultural assets, such
as land and large livestock like cattle and buffalo.
The findings from this re-survey of two villages raise important policy challenges for government
and other stakeholders in Mahbubnagar District, in Andhra Pradesh and in the semi-arid tropics of
India more generally. Whilst government policy and state interventions are made along sectoral
lines, household livelihoods are highly diverse. Policy-makers need to reflect on the most suitable
ways of supporting this diversity, for example by facilitating access to the assets that people draw
on to diversify or by ensuring that agriculture is less risky and agricultural assets are not eroded
during periods of uncertainty. Only with more appropriate policies that recognise the importance of
diversity will it be possible for more people to make positive exits from poverty through
diversification.
vi
1
1 Introduction
The broad aims of the Livelihood Options Project are to understand how rural livelihoods have
changed and diversified, to identify the conditions under which poor people have been able to
obtain access to new and more productive livelihood opportunities, to understand how this process
has affected the well-being of poor households and to identify the role of the state in these
processes.
This paper reports on findings from one aspect of the research that has taken place in Andhra
Pradesh, namely a re-survey of two villages, Aurepalle and Dokur, in the Andhra Pradesh district of
Mahbubnagar. The re-survey of villages adds value to the study’s analysis of livelihoods by
contributing a strong temporal dimension to the analysis of livelihoods. Whilst the main study in
Andhra Pradesh focuses on synchronic censuses of villages, year-long surveys and in-depth studies
in various villages, Aurepalle and Dokur have been the subject of a substantial amount of high
quality research by ICRISAT for three decades and offer, therefore, an opportunity to learn more
about how livelihoods have changed over time, to see the impacts of policy change and economic
transformation on livelihoods and, finally, to think about the types of livelihood trajectories that
have enabled people to exit poverty.
More recently, there has been an increasing preoccupation with more holistic views of poverty and
a recognition of the fact that a plethora of activities make up the livelihoods of the rural poor. This
can, and often does, involve acknowledgement of the many non-agricultural activities that are
carried out by poor people in rural areas. Whilst in sub-Saharan Africa, this diversification is
frequently interpreted as a response to the difficulties that poor (and richer) households face in the
context of structural adjustment and liberalisation (Bryceson, 1999; Ellis, 1998, 2000), the ways in
which households in India are responding to the processes of liberalisation that began in the early
1 www.livelihoodoptions.info
2
1990s are the subject of increasing debate and continue to be disputed (Meenakshi and Ray, 2002;
Datt and Ravallion, 2002).
Research elsewhere has shown that diversification is not necessarily a strategy pursued by poor
people, nor is it just about coping. For some people it can help in mitigating risk or coping with
vulnerability where risk remains high and in setting poor people on a cumulative path towards
greater livelihood success (Davies, 1996). In addition to reducing the risk of livelihood failure (Gill,
1991; Alderman and Paxson, 1992), diversified livelihoods can also help to reduce seasonality in
labour demands and consumption (Morduch, 1995), offset the impacts of natural risk factors on
staple food availability (Reardon et al., 1992), add activities with higher returns to the household
livelihood portfolio (von Braun and Pandya-Lorch, 1991), provide cash resources that enable
household assets to be built up, and help people to hold onto the assets they already possess
(Netting, 1993). Diversification across income sources helps households to combat instability in
income and thereby increases the probability of their maintaining livelihood security. Poor people
build diversification strategies sensitive to their context and livelihood strategies. A significant issue
raised when studying diversification in the context of rural Andhra Pradesh is caste and its
occupational categories. People in rural Andhra Pradesh may depend for their living and livelihood
on various activities but the options that they can explore are limited by caste. In addition to
changes in the availability of natural resources and other sources of livelihood, policy and the
institutional environment may also affect peoples’ livelihoods and livelihood security.
Documentation of such changes provides scope for an improved understanding of household
livelihoods and presents an opportunity to provide important inputs into the policy-making process.
Only via effective policy can the best action be taken to support the diverse livelihoods of the rural
poor and to enable them to cope with uncertainty and adversity.
The next section of the paper outlines the methodological approach that was used in the re-survey of
the villages of Aurepalle and Dokur. Whilst a longitudinal study offers a valuable opportunity to
think about changing livelihoods and diversification over time, there are a number of constraints
that arise when carrying out a re-survey, not least the fact that the original survey that began in 1975
had very different analytical objectives. Next, the two villages in which the study took place are
introduced and their socio-economic and agro-ecological characteristics are reviewed.
The main section of the paper analyses the changes in livelihood and processes of diversification
that have taken place in the two villages. The analysis is broken up into two sections. The first and
larger section begins with a sketch of the main sources of livelihood in the villages and shows how
these changed between 1975, 1989 and the re-survey in 2001. The next section of the analysis
attempts to place these changes in their institutional, political and economic context and thereby
uncover some of the driving forces behind the changes that have been identified. The final section
of the analysis considers the impact of these changing livelihoods and interrogates changing poverty
and inequality levels within the villages. The conclusion draws on some of the main findings to
identify some potential policy contributions that arise out of the re-survey.
4
Longitudinal research methods have great analytical strength in that they allow processes of change
in households to be tracked. Whilst year-on-year surveys that sample a proportion of the population
can provide a series of snapshots showing what proportion of the population is unemployed or has
no land, in longitudinal studies it is possible to see who has become unemployed and who has lost
or gained land. In the case of the Village Level Studies (VLS) carried out at ICRISAT, they were
part longitudinal study, in so much as that they involved a census that covered every household in
each village and households could therefore be traced from one round of the study to the next, and
part year-on-year survey since they included a survey that did not cover the same households at
every round (Singh et al, 1985).
The study in 2001 was based on information gathered through Participatory Rapid Appraisal
(PRA), a household census and household survey and panel interviews in Aurepalle and Dokur
villages. The data gathered in 2001 was then compared to data from the VLS in 1975 and 1989. In
2001, the household census was conducted for all households in each village, with the objective of
providing a broad overview of the villages, land holdings, household sizes, castes and major sources
of livelihood. This and the subsequent survey laid the foundations for a later in-depth panel study of
households from the two villages.
The development of this sample for the household survey was based on the original survey
sampling technique that was used in 1975 and requires some further explanation. At the time of the
census round (May 1975), the total number of households was 476 in Aurepalle and 313 in Dokur.
Of these households, a sample of 40 respondent (30 cultivator and 10 labour) households was
selected in each village to ensure representation of all categories of households – labour, small
farmers, medium farmers and large farmers. The small, medium and large farm sizes were derived
by ranking all census households by size of operational land holding and dividing them into three
equally numerous terciles. Ten households were selected at random from each tercile. In 1989 a
new sample (of 36 cultivator and 12 labour households) was derived in the same way and there are,
therefore, different farm sizes for 1989 and 1975 (Table 1). So what farm sizes were to be used in
the 2001 sample? The research team decided that, since understanding change was the primary goal
of the research, the most appropriate method would be to use the same farm size categories that had
been used in 1975 in order to construct the 2001 sample. Statistical representativeness gave way to
a more direct comparison of the experiences of small, medium and large farmers, and of landless
households.
Table 1 Farmsize classification based on operational holding (ha) in the study villages
Farmsize class Operational holding (ha)
1975 1989 2001
Aurepalle Dokur Aurepalle Dokur Aurepalle Dokur
Small 0.20–2.50 0.20–1.01 0.20–1.01 0.20–0.81 0.20–2.50 0.20–1.01
Medium 2.51–5.26 1.02–3.04 1.02–2.43 0.81–1.62 2.51–5.26 1.02–3.04
Large >5.26 >3.04 >2.43 >1.62 >5.26 >3.04
Note: Operational holding was calculated as: owned land minus land leased-out/share cropped-out plus land leased-
in/share cropped-in. Operational holdings for 1975 are taken from the ICRISAT Village Level Studies and not from
Walker and Ryan (1990), in which different operational holding sizes are quoted.
The next potential stumbling block was that non-farm households were not studied under the
Village Level Studies in 1975 and 1989. Any households from the census that were not involved in
agriculture (either as farmers or labourers) were not included in the sample. This left the research
team in 2001 with a problem. How was it possible to retain a level of consistency in the sampling
between 1975, 1989 and 2001, whilst enabling a focus on non-farm activities? How could the
sample be constructed to ensure that households dependent on non-agricultural activities were not
ignored? The research team selected ten households from each of the categories of landless
labourers and small, medium and large farms in each village following the 1975 method. Then, an
additional 41 households (21 from Aurepalle and 20 from Dokur) were selected on the basis of their
involvement in non-farm livelihoods (Table 2). These were sampled from the remaining census
households. A range of different livelihoods had been recorded in the census and a similarly broad
range was used in the sample. Some of these were placed under the category of ‘livestock’ rather
than ‘non-farm’, since those who made their living from, for example, shepherding goats, could not
be classified under ‘non-farm’. Therefore, all of the households in the non-farm group could also
have been part of the small, medium, large and landless labour groups. Whilst there are a number of
drawbacks to the sampling approach taken, especially that statistical comparisons between 1975,
1989 and 2001 are not strictly reliable, it was felt to be the most appropriate way of allowing some
compatibility with the 1975 and 1989 samples whilst enabling an analysis of non-farm and diverse
livelihoods.
6
The present Mahbubnagar district was part of the dominions of the Nizam of Hyderabad from the
later part of 17th century, when the dynasty of this feudal ruler was established in this part of south
central India, until 1949, when Hyderabad State was absorbed by independent India. In general,
land tenure in Mahbubnagar was freehold (ryotwari). However, in 1901 half of the district was not
under the direct administration (khalsa) of the Nizam but was granted to office holders as payment
during the period they served the Nizam (jagir).
Although average rainfall was around 750 mm per annum, the district supported a considerable
amount of rice cultivation with the help of irrigation from numerous runoff collection reservoirs,
tanks and wells. Tank building was one of the important activities of kings and rulers for centuries
in the uplands and semi-arid granitic areas (now western Andhra Pradesh and western Tamil Nadu),
mainly to assure water for rice cultivation. Around 1998, under a government programme (see
Appendix 3), new arrangements for getting potable water came to both villages. Before 1998, water
was taken from wells but the supply was unreliable in terms of both quantity and quality. Around
1998 water tanks were built into which water was pumped from more reliable and safer boreholes
some distance away. Thirteen drinking water wells were supplied to Aurepalle and eight to Dokur.
In 2001, drinking water for village households was supplied in two ways and was controlled by the
7
gram panchayat (village council). First, there were taps for communal use at various locations
around the village. Second, individual households could also have a tap fitted in their own yard for
which an initial charge and then monthly consumption charges were payable.
Aurepalle was electrified in 1962 and Dokur in 1967. Whilst for some time electricity was used
only for lift irrigation and relatively few houses were electrified, in 2001 at least 90% of households
had a domestic electricity supply that was used for lighting and for powering radios and televisions.
Villagers paid a standing charge of Rs 50 per month and were charged according to their
consumption, though most of the meters were broken. When people were disconnected after failing
to pay their standing charge, they sometimes made illegal connections to the power supply. Despite
widespread availability of electricity, there was frequently only electricity supply for about ten
hours each day.
There was a village panchayat in both the villages. The villagers elected the panchayat members
and president every five years. The village president was responsible for collecting house taxes and
getting funds from governments to fund education, sanitation, drinking water, roads and streetlights.
In each village there was a village patwari (revenue official), a Malipatel and a police Patel. The
patwari was responsible for maintaining land records, the Malipatel for the collection of revenues,
and the police Patel for maintaining law and order in the village. In each village a Development
Officer from the state government gave advice to the farmers and worked as an extension agent of
the block development office. A television set was installed in the gram panchayat office in
Aurepalle in November 1978. It was used for educating farmers in the use of new agricultural
technologies.
In 2001, there were more than 10 small shops in each village that sold basic consumer goods. Most
of the labourers and small farmers sold their in-kind wages and farm produce to and purchased
provisions from these shops. Each village had a post office, a fair price shop, and flourmills. There
were private medical practitioners in both the villages. Primary health centres were located in the
respective block head quarters.
In 1975 there were 476 households in Aurepalle and 313 in Dokur (Table 3). By 1989, there were
664 households in Aurepalle and 444 in Dokur. By 2001, the total number of households in
Aurepalle and Dokur was 649 and 515, respectively. Between 1976 and 2001 the number of
households grew by over a third in Aurepalle and by two-thirds in Dokur. Total population in
Aurepalle increased only marginally from 2,711 people in 1975 to 2,960 in 2001. Dokur saw a
much greater population increase from 1,783 people in 1975 to 2,737 in 2001. Thus, whilst
Aurepalle saw a population increase of less than 10%, in Dokur the increase was more than 50%.
Twelve households are known to have migrated permanently from Aurepalle between 1989 and
2001, though this does not fully explain why population in Aurepalle has decreased in that time
period. The overall rise in population density (Table 3) raises questions about the decreasing
viability of agriculture as population pressure and fragmentation through inheritance lead to smaller
and smaller holdings (Table 5). The implications of this for household livelihoods will be
considered in a later section.
8
In the villages there were households belonging to forward, backward and scheduled castes. As
many as 24 castes existed in both villages, among which the Brahmins, Reddys (Kapus) Vaisyas
(Komati) and Velamas were generally more influential and rich. Mala and Madiga caste people
(Harijans) were ranked as low caste people in the social hierarchy. In Aurepalle, more than 51% of
households belonged to a backward caste followed by scheduled castes (36%), forward castes
(11.4%) and about 1.4% households were Muslims. In Dokur, 63% of households were backward
caste, 19.4% were forward caste and 16.4% were scheduled caste. About 1.5% households were
Muslim (Table 4).
9
Table 4 Distribution of all households (by caste) in Aurepalle and Dokur villages of Andhra
Pradesh, 2001
Caste Aurepalle Dokur
No of HH % of HH No of HH % of HH
Forward caste 74 11.40 100 19.42
Brahmin 2 0.31 1 0.19
Reddy 54 8.32 90 17.47
Velma 9 1.38 0 0
Vysya 9 1.38 9 1.74
Backward caste 333 51.31 324 62.91
Balija 0 0 1 0.19
Battu 0 0 3 0.58
Bichhagalla 0 0 1 0.19
Boya 2 0.31 43 8.34
Chakali 20 3.08 10 1.94
Gowda 164 7.73 20 3.88
Hamsala 5 0.77 1 0.19
Jogi 0 0 4 0.77
Katika 5 0.77 0 0
Kamsali 0 0 1 0.19
Kummari 10 1.54 3 0.58
Kurma/Golla 69 10.63 43 8.34
Mangali 12 1.84 7 1.35
Medari 0 0 9 1.74
Munnuru Kapu 4 0.61 0 0
Sevaka
Musti 0 0 34 6.6
Padmasali 9 1.38 3 0.58
Telaga 12 1.84 131 25.43
Vadla 21 3.23 9 1.74
Vasishta 0 0 2 0.38
Scheduled caste 233 35.91 83 16.12
Madiga 141 27.12 73 14.17
Mala 88 13.55 7 1.35
Yerukula 4 0.61 3 0.58
Muslim 9 1.38 8 1.55
Total 649 100 515 100
Source: Household Census, 2001 and 1989
Table 5 presents the broad agro-ecological and technological features of the study villages. Average
operational holdings are difficult to compare because the basis on which holdings were calculated
changed between 1975 and 2001. However, all the evidence does point towards a decrease in the
size of operational holdings, largely as a result of fragmentation through land inheritance. Another
significant change was the prevalence of new cropping patterns. Irrigation increased in both villages
between 1975 until the early 1990s. Improved water supply enabled farmers to switch to
commercial crops like cotton, paddy and castor. By 2001, difficulties with irrigation meant that
farmers were experiencing problems with irrigated crops and some were turning back to coarse
cereals that grew drawing on residual soil-water moisture. In a later section, the reasons behind the
changing cropping patterns will be explored, with reference to the local institutions and policy
environments.
10
This section considers how livelihood sources have changed between 1975, 1989 and 2001. The
starting point is an analysis of the changing number of sources of income in the two villages
between 1975 and 2001. The sources of income are then analysed to try and understand the relative
importance of different activities within household livelihood repertoires. Thus, the changing
proportion of agricultural income in total household incomes and the relative dependence by
households on different activities are calculated. The data shows a decrease in the proportion of
household income that is derived from agriculture. In the last part of this section the forces that have
driven the diversification process in terms of assets and capital portfolios, agrarian change,
migratory labour movements and social change are identified.
Referring to the 1975 data, Jodha et al (1977) argued that small farm households were more likely
to have more than one source of income. They suggested that, where land holdings were small,
households were more vulnerable to the exigencies of drought and unreliable yields. Diversification
of resource use, particularly family labour use, was one of the ways in which the risky returns from
land could be supplemented. In terms of operational land holdings, households from all land
holding groups, including the landless, diversified between 1975 and 2001. Beyond this broad
change, it is difficult to discern any other pattern regarding land holding groups and levels of
diversification. However, if Jodha et al were correct that diversification was a response to risk then,
in the context of the changing conditions under which agriculture was carried out in 2001, it
becomes apparent that all households, not just small farm holdings, faced risk in agriculture and
diversified in order to reduce their vulnerability to shocks and trends within the agricultural sector.
12
Whilst Tables 6 and 7 quantify diversification by farm size, Table 8 shows levels of diversification
by caste in the two villages. In Aurepalle, backward and scheduled caste households depended on
more sources of income than forward caste households. About 56% of forward caste households
had two sources of income and 11% of households had only one source of income. In backward and
scheduled caste groups, no households had only one source of income. The majority of the
backward caste households (30.8%) had four sources of income and 23% of households had five
sources of income. Among scheduled caste households, 42% had three sources of income, while
17% of households had four sources of income and 8% had five sources of income. Like Aurepalle,
in Dokur, backward caste households also depended on more sources of income than forward
castes. In the case of forward and backward castes, the majority of households had between two and
four sources of income but all scheduled caste households had only two sources of income. A much
smaller number of forward and scheduled caste households had 4–6 sources of income. Diversity of
income sources for different castes were not dealt with in previous studies, so comparison cannot be
made between 1975 or 1989 and 2001.
In Table 9, income is expressed per capita and not per household. No attempt was made to convert
to equivalence scales to adjust for the age and gender composition of the household. Not using
equivalence scales should lead to under estimating welfare for households with more members and
more children because of potential economics of scale in consumption and because of children
costing less than adults (Deaton and Mullbauer, 1982). However, this method allowed comparison
with previous studies conducted in these two villages using the same computation method (Singh et
al, 1982; Walker and Ryan, 1990).
Between 1976 and 2001, the level of dependence on agriculture as a source of income changed. In
1975, the major source of income in both villages was agriculture. More than 87% of the net
income of Aurepalle villagers in 1975 and more than 96% of Dokur villagers’ income was from
agriculture (Table 9). In contrast, only 32% and 27% of the net income of Aurepalle and Dokur
villagers in 2001 was from agriculture. Non-agricultural income accounted for 68% of net income
in Aurepalle and 73% of net income in Dokur in 2001.
Table 9 Percentage of agricultural income and non-farm income to net household income,
1975–6 to 2001
Agricultual income 2001 1975
Aurepalle Dokur Aurepalle Dokur
Net crop income 21.17 10.40 29.8 46.1
Net livestock income 4.57 9.25 25.5 2.0
Farm/casual labour 6.02 6.52 32.8 46.3
Regular Farm Servant (RFS) 0.57 1.21 - -
Rental - - -0.8* 2.2
Total agricultural income 32.27 27.38 87.30 96.6
Non-farm income
Non-farm wages 1.71 1.33 - -
Net migration labour 4.97 25.34 - -
Remittances 1.34 0.20 - -
Salaried jobs 16.85 4.75 - -
Caste occupation 12.62 6.15 - -
Business/trade and handicraft 9.59 7.58 11.60 1.10
Others 20.65 27.27 1.10 2.30
Total non-farm income 67.73 72.62 12.70 3.40
* The negative figure here is ascribed to losses from the rental of family-owned assets (Singh et al, 1982)
Source: For 1975–8, Singh et al (1982); for 2000–1, Household Survey.
In 2001, the predominant source of livelihood in Aurepalle was still agriculture and related
activities (28% agriculture and 21% farm work) but in Dokur, income from seasonal migration
(37% households) and income from agriculture and related work (18.3% agriculture, and 16.5%
farm work) had equal importance. Compared to the situation of 1975–8, this was a significant
change (Tables 10, 11 and 12).
14
Table 10 Primary occupation wise distribution of households (HH) in Aurepalle and Dokur,
1975, 1989 and 2001
Main 1975 1989 2001
occupation Aurepalle Dokur Aurepalle Dokur Aurepalle Dokur
No % of No % of No % of No % of No % of No % of
of HH of HH of HH of HH of HH of HH
HH HH HH HH HH HH
Agriculture 201 42.2 167 53.4 123 21.47 165 44.47 179 27.58 94 18.25
Business 70 14.7 14 4.5 140 24.43 17 4.58 21 3.24 8 1.55
Carpentry - - - - - - - - 20 3.08 8 1.55
Farmwork 132 27.9 75 24 180 31.41 127 34.23 134 20.64 85 16.5
Govt job 13 2.7 10 3.2 13 2.27 9 2.43 8 1.23 9 1.74
Migration - - - - - - - - 51 7.85 191 37.08
earning
Milk sale - - - - - - - - 3 0.46 4 0.77
Private job - - - - - - - - 3 0.46 11 2.13
Regular job - - - - 14 2.44 2 0.54 24 3.69 2 0.39
Sheep rearing 10 2.1 16 5.1 6 1.05 0 0 38 5.86 20 3.88
Toddy sale - - - - - - - - 105 16.18 11 2.13
Washing clothes - - - - - - - - 8 1.23 5 0.97
Contract labour - - - - 1 0.17 - - 0 0 8 1.55
Others* 50 10.4 31 9.8 96 16.75 51 13.75 55 8.47 59 11.45
Total 476 100 313 100 573 100 371 100 649 100 515 100
Note: * For Aurepalle village in 2001 other occupation includes cart, commission agent, flour mill, money lending,
permanent servant, pot maker, vegetable sale; and for 1989 includes rural crafts, caste occupations and others. For
Dokur village in 2001, other occupation includes auto driver, bangle sale, broomstick making, cable operator, carpentry,
electrician, filling air, grinding chillies, hiring out bullocks, jewellery making, lawyer, line man, lorry cleaner, mason
work, mechanic, post master, priest, regular job, renting land, rice mill, saw mill, std booth; and in 1989 includes rural
crafts, caste occupations and other activities
15
‘Land holdings are more scattered and fragmented these days. Many land transactions
were reported after 1990. Labourers and farmers belonging to small and medium size
land holding groups had purchased land from large landlords. Many of these
transactions were distress sales (a coping mechanism during drought years). At present
it appears that marginal and small farmers leased in more land in order to use their
excess human and bullock labour more productively. Large landowners were not in a
position to cultivate their entire land holding due to the non-availability of regular farm
servants and the increase in the maintenance cost of bullocks.’
Unpublished PRA fieldnotes, G.D.N. Rao.
The key reasons for these changes are as follows. First, in the late 1970s large farmers lost land
under the 1977 Land Ceiling Act which set an upper level for land holdings in both rural and urban
areas in India. Second, between 1975 and 2001 the number of households grew by 36.3% in
Aurepalle and 65.5% in Dokur. The modest increase in the total cultivatable area in both villages
(Table 3) was not sufficient to soak up the growing population and family land holdings became
fragmented through inheritance. Another reason for the decrease in operational land holdings is that
many irrigation systems failed to provide sufficient water, so the amount of irrigable land declined
in Dokur (Table 3). Land was left fallow and was, for the time being, out of operation. PRA
exercises carried out by ICRISAT field researchers showed that between 1989 and 2001, irrigation
water availability declined (see Table 13) leading, in Dokur, to the dramatic reduction in irrigated
land that was shown in Table 3.
17
The numbers of bore wells and dug wells increased in Aurepalle between 1989 and 2001 but
farmers faced problems with the availability of water as the water level in many of the wells was
precariously low. In Dokur, fewer bore wells and dug wells were actually reported in 2001 than had
been recorded in 1989. (Again, this is reflected in Table 5 which shows a sharp reduction in the
irrigated area as a proportion of the gross cropped area in the same time period.) In the PRA focus
groups, householders argued that three out of five years were drought years in the villages. Annual
rainfall statistics from the Andhra Pradesh Directorate of Economics and Statistics show that
drought was experienced in Mahbubnagar District in 1997–8, 1999–2000 and 2000–1. Rainfall in
1995–6 and 1996–7 was average and there was slightly more rain in 1998–9
(www.andhrapradesh.com/apwebsite/tables). In the years immediately preceding the ICRISAT re-
survey, the increase in the number of wells and tube wells, coupled with low rainfall, led to a
lowering of the water table (see Table 14).
18
The changing livestock assets of households are shown in Table 15. The number of large livestock
in Aurepalle decreased between 1989 and 2001, whilst in Dokur, numbers of bullocks and cows
decreased but numbers of buffaloes increased marginally. The reasons for this are discussed below
but are linked to changing crop patterns and the decreasing availability of both fodder and
communal grazing land. The marginal increase in Dokur of buffaloes can be attributed to preference
in villages for milk production from buffaloes rather than more expensive varieties of cows that are
difficult to maintain.
The increased mechanisation of the villages was another factor behind the reduction of livestock in
both villages. The number of tractors increased to 8 in Aurepalle and 9 in Dokur. Tractors were
used in place of draught livestock for ploughing and threshing.
Table 16 Numbers of tractors in Aurepalle and Dokur between 1989 and 2001
Farm size 1989 2001
Aurepalle Dokur Aurepalle Dokur
Landless 0 0 1 0
Small 0 0 2 0
Medium 0 0 3 4
Large 0 4 2 5
Total 0 4 8 9
Threshers were also introduced in Dokur. Whereas in 1975 landless labourers with livestock had
rented out their bullock pairs for ploughing, in 2001 a rental market for tractors, threshers and
sprayers had been established. In 2001, a greater proportion of farm produce and inputs was
transported by tractor. New forms of transportation opened up markets for both labour and
agricultural crops. A large portion of the road between Aurepalle and the small market town of
Amangal was tarred in 2002. From Amangal the road is tarred through to Hyderabad. Privately
operated jeeps (funded by subsidies from the government aimed at supporting private individual
enterprise), state-run and private buses all operated between Aurepalle and Amangal.
There were significant changes in household portfolios of human and financial capital between
1975, 1989 and 2001. The availability of household labour was largely dependent on household size
19
and stage in the household development cycle. In both Aurepalle and Dokur, the average household
size decreased gradually between 1975 and 2001 (Table 17). Should this decline continue, there
would be fewer people available in the household for agricultural work (either on household land
holdings or as agricultural wage labour).
Educational levels have also improved (Table 18), with significant increases in the education level
of people within the villages and varying impacts on livelihoods in both villages. Education
facilities existed within the villages to study up to high school level (tenth standard) in Aurepalle
and up to seventh standard in Dokur village. There were some private (convent) schools within the
villages and nearby villages that provided additional opportunities to study. The supply of free
textbooks to school-going children and a midday meal programme encouraged low-income
households to send their children to school (see Appendix 3). As a result, there were large increases
in the number of years of schooling in all landholding classes (Table 18 and 19). In terms of gender,
the average number of years of schooling of males increased from 1.92 years in 1975 to 5.57 years
in 2001. Whilst the schooling of women increased from 0.78 in 1975 to 4.03 years in 2001, the
education of girls and women still lags behind that of boys and men. A large increase in education
level amongst boys was seen amongst landless labour households but within these same
households, the education of girls remained very low. In 2001, for both the male and female
members, the level of education within the non-farm category was the highest (5.5 years for male
and 4.03 years for female). Increasing levels of education may be a positive impact of cash income
on human poverty. Alternatively, it could be the case that education is an important precursor for
entering the non-farm economy.
Table 19 Level of educational attainment by landholding class in Dokur, 1975 and 2001
Age 1975 2001
Land Small Medium Large All Land Small Medium Large Non- All
less less farm
6–10 0.00 0.00 2.00 2.30 1.43 3.55 1.80 2.00 2.58 2.05 2.53
11–15 0.00 1.50 2.67 3.36 2.31 5.43 2.50 4.00 6.86 5.80 5.10
16–20 0.00 0.50 4.50 5.61 3.29 9.75 2.20 14.00 6.76 11.25 7.23
21–25 0.00 0.00 2.21 5.81 2.82 7.75 0.11 10.00 5.79 5.67 4.62
26–35 1.00 0.00 2.81 1.20 1.33 6.67 0.00 5.00 3.45 6.67 4.22
36–45 0.40 0.36 0.67 0.40 0.43 2.00 0.83 3.90 4.27 5.20 3.51
46–60 0.00 0.00 1.36 0.40 0.63 1.40 0.00 4.00 1.92 1.00 1.76
60+ 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.38 1.11 2.37 1.25
All 0.31 0.40 1.92 2.31 1.45 4.33 1.10 3.55 4.33 4.70 3.71
In terms of healthcare facilities, the number of private medical practitioners increased. Health
workers appointed by the Government were providing services to the villagers and a primary health
centre (30-bed hospital) was located within 10 km of one of the villages. The supply of protected
drinking water through a pipeline was available, though the hours during which drinking water
could be collected were restricted in both villages.
Financial capital came from varied sources in both villages. Access to a cash income, rather than
payment in kind, had grown since 1975. For those who did not have access to their own land in
1975 the mode of wage payment for agricultural work in Aurepalle was mainly in kind. 85% of
labourers received payment in kind whilst only 15% received payment in cash. In contrast, the
mode of payment in Dokur was predominantly in cash. In 2001, almost all labourers in both
villages received their wages in cash. This change in traditional farm-labour arrangements resulted
in a change in relationship between employers and labourers. The new system of cash payment
provided more freedom to labourers than the traditional farm servant arrangement. Labourers no
longer worked as Regular Farm Servants (RFS) in the villages. In Aurepalle and Dokur, the shift
from payment in kind to wage labour had implications for the choices that were available to
labourers in terms of how they disposed of their incomes. It enabled labourers to invest in
alternative sources of livelihood and to take loans under the terms of which they provided half the
funds themselves.
In the PRA focus groups, respondents commented that credit markets were more efficient than they
had been in the past. The timing of these improvements corresponded with an increased government
intervention and the formalisation of credit markets. A broader range of loans was available from a
wider range of institutions than previously and the importance of village moneylenders was
reported to be decreasing. The majority of farmers had borrowed from the Primary Agriculture
Credit Co-operative Society (PACS) that had offices in both villages. The tendency was for
households to have long-term loans with PACS and short-term loans with the village moneylenders.
The size of outstanding loans increased between 1989 and 2001 (Table 20). The increases were
greater amongst large and medium farm size households and smallest amongst landless and small
farm size households. Whilst outstanding loans had increased, the number of households who were
saving had fallen. In 1989, people from all land holding groups had savings. Most people kept their
savings in a bank but others kept the money at home so they could access it more easily in a crisis.
In comparison, in the 2001 survey, only 5 households reported savings. These were both in
Aurepalle village and were from the landless and small farm classes. No medium or large farmers
reported saving and no households reported lending money to others.
21
Thus far, changes in the capital assets that people have drawn on to develop different livelihood
activities have been reviewed. Some assets have improved or increased, though the benefits have
not be felt across all land holding groups, by both men and women or across different caste groups.
Elsewhere, assets, particularly physical and natural capital, were eroded with consequence for the
livelihoods of all. In the next section some of the processes by which these changing asset portfolios
resulted in diversification are considered, with particular reference to agrarian change, agricultural
policy and migration.
Until 1982, the minimum support prices for sorghum and millet were the same as that of paddy, but
after 1982 the government announced support prices for coarse paddy that were much higher than
those for the coarse cereals (Rao, 1999). This was the first factor that led farmers to produce more
paddy and to decrease their cultivation of coarse cereals. In 1985, faced with the lack of purchasing
power held by many rural and urban households and with growing reserves of wheat and paddy, the
Indian government established a Public Distribution System (PDS) through which poor households
could access subsidised food (see Appendix 3). In Andhra Pradesh, the level of subsidy was higher
than the national level. With cheaper food available to buy, farmers were able to decrease their
dependence on staple food crops, such as pearl millet and sorghum, and move into the production of
cash crops such as cotton or irrigated paddy. The shift to castor was part of India’s ‘yellow
revolution’ or the rapid spread of the cultivation of oilseeds resulting from government support
(Gulati and Kelley, 1999). In the mid-1980s, India had been importing about 30% of its edible oils
and sought to become self-sufficient in edible oils in order to improve its balance of payments. With
the imposition of import tariffs, the domestic market grew steadily. The implication for cropping
patterns in Aurepalle and Dokur was manifested in a dramatic increase in the cultivation of castor.
Both the villages experienced labour abundance in lean season and a labour shortage in peak
season. Growing highly labour-intensive crops such as cotton and paddy created more demand for
labour, and hence the bargaining power of labourers increased considerably. Households that
acquired more land had to tap into the labour market in which wage rates had increased by 8–10
times between 1975 and 2001. The shortage of labour that pushed up labour costs was also driven
by an increase in migrant labour. This will be discussed in a later section.
22
The outcome of the factors outlined above meant that crop cultivation became an increasingly risky
activity. In Aurepalle, for example, 31% of households had either negative or no income from
cultivation. As Jodha et al argued as early as 1975, adopting additional sources of livelihoods (or
changing cropping patterns) reduced the vulnerability of households to shocks and trends in
agriculture. In the context of the risks faced by those gaining their livelihood from agriculture,
seasonal labour migration became an increasingly common phenomenon in Mahbubnagar district
between 1975 and 2001. Whilst the 1989 census and survey did not include labour migration as an
income category, it is clear from the findings that, with the exception of those involved in business,
the numbers of people leaving the villages to seek non-farm work were much lower than in 2001.
This change was largely due to diminishing and increasingly unreliable returns from cropping and a
lack of local employment opportunities throughout the year. Thus, many households in the two
villages depended mainly on labour earnings despite owning some land (Tables 9 and 12).
Household members, and occasionally entire households, periodically migrated to other cities for
their livelihood. Some people travelled to Hyderabad whilst others ventured as far as Pune, Goa,
and Mumbai in Maharashtra, and Surat, Baroda and Ahmadabad in Gujarat. There they sought non-
farm work, e.g. driving, mud work, construction, watchmen, canal digging, or found employment in
their caste occupations as washers of clothes, carpenters, goldsmiths and toddy tappers.
Villagers received information about migration mainly from migrants who visited the village for
festivals and from relatives who were staying in the urban areas. Up-to-date information about the
chances of employment opportunities, nature of work, terms and conditions and wage rate for
different work for male and female workers at Hyderabad and other towns was important for
successful migration. Those educated up to 10th standard or more worked in monthly salaried jobs
(part- and full-time) and others worked as day labourers. Beyond the broad findings that scheduled
and backwards castes were more likely to migrate than people of forward caste and the importance
of social networks within villages for accessing migrant labour opportunities, there were some
important differences between the two villages. For this reason, the discussion of migration will
deal with each village in turn.
Around 350 people (12% of the population) from Aurepalle village, including men, women and
children, migrated to cities and towns such as Hyderabad, Kalwakurthi, Mal and Mahbubnagar to
seek employment opportunities. Around 300 of all the migrants went to Hyderabad because it had
relatively more employment opportunities and better transport facilities than other nearby
destinations. Seasonal out-migration from Aurepalle began in the early 1980s (though it wasn’t
counted in the 1989 data) and increased gradually over time. The main reasons for migration
reported by the migrants were:
(a) not getting employment throughout the year within the village;
(b) negligible alternative employment opportunities locally;
(c) high population pressure, and
(d) low wage rates for farm and non-farm activities.
23
Some migrants also reported a lack of interest in working as labourers within the village, a decline
in the importance of and remuneration for their for caste occupation within the village, a decline in
the area under irrigated crops which had provided employment opportunities, a lack of employment
opportunities for educated persons in the village, a surplus of family labour compared to family land
holdings, and the desire to lead an enjoyable life in an urban area. A smaller number (10 or 15) of
households had left the village permanently to take advantage of larger markets in towns.
Landless households and people participating in the Aurepalle labour market generally migrated for
the whole year and visited the village for festivals and family functions. Small and marginal farmers
migrated in the month of August after completion of the major farm operations. Old people
(parents) took over responsibility for housekeeping, childcare and agricultural activities during
migrants’ absence. Migrants received a monthly salary of around Rs 1,500 for part-time work or Rs
3,000 for full-time work. Daily-rated work earned them around Rs 60–80 per day. Migrants
reported that they got an average 22–25 days employment in one calendar month. They received no
benefits such as bonuses, medical and educational allowances and food except for a few cases
where the employment was regular in nature. Migration helped to improve the conditions (both
economical and social) of households in terms of standard of living, assets position, awareness of
livelihood opportunities, education of their children, and their ability to buy food and clothing.
Seasonal migrants’ families did not face any negative attitudes in the village and the children of
those who migrated were more eligible for marriage than those who had never left the village for
work.
Seasonal out-migration from Dokur village began in the 1970s but on a very small scale. Out-
migration increased more rapidly after 1992–3 because of the increase in population (leading to
fragmentation of land holdings), the lack of work within the village throughout the year, the higher
wage rates that were offered outside the village and the evolution of a young generation that were
attracted towards urban life. Around 910 people out of 2,737 (more than 30% of Dokur’s
population) were seasonal migrants to Hyderabad, Nizamabad, Pochampadu and Mahbubnagar
within the state, and to Gujarat and Maharashtra outside Andhra Pradesh.
Irrigated paddy, the most labour-intensive crop, was grown in both rainy and post-rainy seasons in
Dokur. Farmers faced labour shortages in peak season from around 1995. At that time the village
labour force (both male and female) could find work within the village throughout the year. By
2001, the situation had changed dramatically. Drought and uneven distribution of rainfall at critical
stages of crop growth led to a decline in the productivity of both irrigated and rainfed crops. The
area under paddy crop decreased drastically due to the non-availability of water in tanks and wells
and the failure of borewells. In the face of this decline, villagers sought alternative employment
opportunities elsewhere. About 30 servicing caste households (washermen and barbers) migrated
permanently to Goa and Pune. The majority of labourers migrated to Hyderabad for mud work,
construction, hamali (loading and unloading) and private monthly salaried jobs such as watchmen,
telephone booth operators, drivers and waiters at hotels and lodges. Labourers received Rs 60–75
per day depending upon the type of work and their gender. Monthly salaries varied between Rs
1,500 and Rs 3,000.
Out-migration to Maharashtra and Gujarat increased in Dokur from around 1998 when a local
labour contractor began offering advance payments of between Rs 7,000–10,000 for migrant labour
contracts. Advances were useful for clearing old debts, repairing or reconstructing houses and for
meeting marriage expenses. Workers were employed for 9–10 months with a monthly salary of Rs
750–800 with free accommodation and food. Monthly salaries were adjusted against advances.
24
Other information sources came from the television sets and radios brought back from urban areas.
In 1978, there was one television set in Aurepalle and no set in Dokur. The television set in
Aurepalle was frequently used to show programmes reviewing new agricultural techniques and the
most efficient use of inputs. By 2001, there were 69 television sets and 180 radios in Aurepalle and
86 television sets and 106 radios in Dokur. In Aurepalle 35 households had a telephone whilst in
Dokur the figure was 11.
As a result of migration, and of government food distribution policy, the perceived consumption
needs of village households changed. As rice became available more cheaply through public food
distribution programmes, preference for sorghum or millet declined. Children that grew up eating
government-subsidised paddy became reluctant to eat sorghum or pearl millet (pers. comm. P.
Parthasarathy Rao, 3/09/02). The shifts in lifestyle and consumption that resulted from public
distribution programmes and from migrant labour had implications for achieving food security and
ensuring the livelihoods of the rural poor (Government of India Planning Commission, 2001).
A final issue relating to social and cultural change is that of caste. As is shown in Table 8,
households of all caste groups adopted increasingly diverse livelihoods. For some households this
required a move away from their caste occupation. Backward caste households had diversified the
most. The options for diversification amongst forward caste households were more limited,
especially where women were not able to leave the house to work. Scheduled caste households
appeared to have found it more difficult to move away from their caste occupation whilst remaining
in the village. In Aurepalle, scheduled caste households still lived at the edge of the village, away
from the centre of the village where trade was at its greatest and where landowners found labour on
a daily basis. A combination of different types of exclusion limited their participation in
entrepreneurial activities and their access to the resources that would be required in order for them
to pursue new non-farm livelihoods.
The capacity of different caste groups to migrate also influenced the level and nature of
diversification. In 2001, when more than 21% of households in Aurepalle and 48% households in
Dokur had at least one household member involved in seasonal out-migration as a source of
livelihood, most of the migrant households belonged to the backward castes (BC) and scheduled
castes (SC) (Table 21). In Aurepalle, more than 50% of the migrant households belonged to the
Gouda (BC) and Madiga (SC) castes, whilst in Dokur about 60% of the migrant households
belonged to the Telaga (BC) and Madiga castes.
25
Scheduled and backward castes were better placed to migrate for a number of reasons. First, it was
socially acceptable for the women of scheduled and backward caste households to carry out various
labour roles, whilst women of forward caste households were expected to occupy themselves only
with household work. Even if their activities in the village were limited, scheduled and backward
caste women could seek out migrant labour opportunities for themselves, or take over the
agricultural and other work usually done by men in the village when men themselves migrated.
Second, whilst for forward caste households, involvement in many of the labour opportunities
available would represented a step down the social ladder, for scheduled (and sometimes backward)
caste households, labour opportunities were often either commensurate with their current social
status or represents a step up the social hierarchy. Finally, there were certain caste occupations that
were particularly valued and required special skills (for example blacksmiths or teachers). These
activities tended to be those of forward or backward castes and were forcefully protected by
households to prevent people of other castes entering the occupation. Thus, for some forward or
backward castes, there was an advantage to be had by focusing on a particular niche activity. A
small number of these households (belonging to weaving, business, goldsmith and service castes)
migrated permanently to towns where they could access larger markets.
26
Thus far, we have seen how rural livelihoods became increasingly diversified in the context of
increasing risks within agriculture and new opportunities in the non-farm sector such as migration
and social change. What are the outcomes of this process of diversification? What changes in
income levels, poverty and inequality have accompanied the shift to more diverse household
livelihoods? Has diversification enabled households to make positive exits from poverty?
Table 22 shows the change in actual and real incomes in Aurepalle and Dokur between 1975, 1989
and 2001. The table demonstrates that gross and net household incomes and net per capita incomes
all increased since the first study. In both villages, net real incomes grew, marginally in Aurepalle
between 1975 and 1989, but significantly in both villages between 1989 and 2001. In 1975 and
1989, the large gap between gross and net household income resulted from the increasing costs
associated with buying inputs for agricultural production. The increase in the price of inputs was
greater than the increase in crop prices, thus eroding the profitability of agriculture. The reason for
the decrease in the difference between gross and net household income between 1989 and 2001 is
that people are increasingly involved in activities outside agriculture that do not require such high
input costs.
Table 22 Total Income from all sources over cropping year (Rs) in 1975–8, 1989–90 and
2000–1
Villages 1975–8 1989–90 2000–1
Aurepalle Dokur Aurepalle Dokur Aurepalle Dokur
Average gross household 21,759 28,753 53,510 75,416 59,397 72,371
(4,564) (6,031) (21,315) (30,041)
Average net household 11,256 14,145 12,640 31,060 39,928 58,417
(2,361) (2,967) (5,035) (12,371)
Average per capita 2,012 2,270 2,782 5,786 8,284 9,577
(422) (560) (1,108) (2,305)
Note: Main figures indicate adjusted real income in 2001 in Rs. Figures in the parentheses indicate actual income in Rs.
Real income is calculated from the Andhra Pradesh Consumer Price Index for Agricultural Labour based on 1960–1 =
100 (www.ap.gov.in/apbudget/tab17_2.htm) and 1986–7 = 100 (Andhra Pradesh Economic and Statistical Bulletin,
various issues)
Source: For 1975–8, Singh et al (1982); for 1989–90 and 2000–1, Household Survey.
Analysis of inequality in income and in productive assets is very important since it provides
meaningful insights for understanding and taking necessary policy actions. To understand whether
diversification contributes to poverty reduction, it is useful to know whether only poor people
benefit from diversification, or whether successful diversification is something that is only achieved
largely by those who already have stable incomes that allow them to invest in other activities. Given
that a sampling strategy was drawn on that allows only partially representative comparisons to be
made between 1975, 1989 and 2001, only basic income distributions were estimated in this study.
Household income distributions are given in Table 23 for Aurepalle and Dokur villages between
1975, 1989 and 2001.
27
Table 23 Distribution (%) of households across net income group in the study villages,
1975–8, 1989–90 and 2000–1
Net (real) income 1975–8 1989–90 2000–1
group (in 2001 Aurepalle Dokur Aurepalle Dokur Aurepalle Dokur
Rupees)
Negative 5 3 28 19 0 0
0–6,810 29 16 30 35 11 2
6,811–13,619 33 38 20 13 18 7
13,620–20,429 11 16 4 15 15 12
20,430–27,238 8 9 2 8 10 13
27,239–34,048 2 8 4 0 11 17
34,049–68,096 7 6 7 6 21 30
68,097–102,144 3 3 2 4 7 12
102,145–136,192 3 0 2 0 2 5
>136,192 0 2 2 0 5 3
Total 101 101 101 100 100 101
Note: Equivalent income in 2001 is calculated by the author using Consumer Price Index (CPI) for rural areas in
Andhra Pradesh.
Source: For 1975–8, Singh et al (1982). Totals equal 101 because of rounding.
Results of the income distribution analysis are provided in Table 24. The table shows that in
Aurepalle the income share of the poorest 40% of households increased from 4% to 11% between
1975 and 2001, whilst the income share of the richest 5% of households decreased slightly from
29% to 28%. In Dokur, the income share of the poorest 40% of households also increased from
13% to 16% whilst the share of the richest 5% decreased more sharply from 27% to 20%. The
income shares of households in 1989 have not been included in the table as the proportion of
household who experienced negative or no income was very high in 1989 (see Table 23). This
increase in the prevalence of negative incomes would suggest increasing inequality of income
between 1975 and 1989, and decreasing inequality of income between 1989 and 2001.
It has already been demonstrated that diversification is important for all categories of farm size and
this analysis of the distribution of income further supports the finding that diversification is
important in both resource-rich and resource-poor households. Furthermore, the declining inequality
of income suggests that poorer households may even be diversifying more successfully than their
richer neighbours. Diversification by those with large operational landholdings has not enabled
them, in general, to accumulate wealth and capital. In fact, given that households with large
operational landholdings that have invested heavily in irrigation systems that are now failing,
diversification appears to be a coping strategy for both richer and poorer households in the two
villages. Richer households may even be less adaptable than poorer ones where they are dependent
on large amounts of investment in irrigation and on a high proportion of agricultural labour, and
when caste limits their options for diversifying.
28
Table 24 Degree of inequality in the distribution of per capita income in Aurepalle and
Dokur, 2001
Income group % Share of total income
1975–8 2000–1
Aurepalle Dokur Aurepalle Dokur
First quartile (poorest) - - 5.44 7.88
Second quartile - - 12.19 14.50
Third quartile - - 21.34 24.03
Fourth quartile (richest) - - 61.03 53.59
Poorest 40% 4 13 11 16
Richest 5% 29 27 28 20
Note: Inter-quartile comparisons of income are not made with 1975 and 1989 because, in the case of 1989, the first
quartile is wholly made up of households within negative net income (see Table 21).
The relative successes and failures of households’ attempts to diversify can be considered by a
longitudinal comparison of households in Aurepalle and Dokur that were interviewed in the 1975
VLS. Walker and Ryan (1990) documented economic mobility in Aurepalle between 1975 and
1984. They observed considerable reshuffling of households in relative income positions and
showed how the economic conditions of different households had improved, declined or remained
static over time. These changes were due to household behaviour, social actions and transformation,
government economic policy and changing asset bases.
Based on the information gathered from visits to 1975 panel households, it was found that 75% of
the respondents of Aurepalle village improved their economic status while no changes were
observed in 10% of respondents and remaining 15% of respondents’ status declined over time.
During the same period in Dokur village, 60% of respondents’ condition was improved, no change
was observed in 20% of households and the remainder of households experienced a decline. The
findings from the panel study tally with those from the 2001 census and survey and demonstrate
that income and well-being increased in both villages and that inequality decreased. What follows is
an attempt to identify some of the key factors that enabled some households to improve their
economic situations and the factors that prevented upward mobility amongst others.
In both the study villages, cases of landless households were documented where there had been a
steady rise in economic status until the household had become one of the wealthiest in the village.
This extreme upward mobility tended to follow a particular trajectory, though it typically took two
or more generations for upward mobility to manifest itself. An example is given in Box 1. In each
example a pseudonym is used to protect the identity of the respondent. Frequently, one generation
took on a low-paying but secure post as a regular farm servant, benefited from government land
redistribution programmes and relied on family labour to cultivate household land holdings. Other
household members migrated and sent remittances through which agriculture could be funded.
Another source of upward mobility was being selected to participate in government agricultural
programmes or projects. The household of VMR had fairly large landholdings (5.48 ha) but these
were rainfed and poor quality and VMR was dependent on local agricultural technology. In 1979,
the VMR household was selected as part of farming systems research programme based at
ICRISAT and they received free inputs and technical advice. The following years saw bumper
harvests that enabled the repayment of outstanding debts, investment in irrigation and bore wells
and the establishment of paddy cultivation in two seasons. In the 1990s, VMR became one of the
village’s key money lenders. In Dokur, the household of NCHR provides a good example of where
diversification, enabled by strong kin relations and co-operation, has lead to the accumulation of
wealth and capital. In 1975, the household had a single income from agriculture and four
dependants. As the children grew up, they began to contribute to the household. The eldest daughter
learnt tailoring skills (an example of people moving into the caste occupations of others in order to
29
gain additional income), and the eldest son leased a telephone booth in Hyderabad. Another son
began to trade in second-hand electric pump sets, whilst the youngest son sold milk in the nearby
town. Profits from the business and income from the dowries received at the marriages of each son
enabled the household to drill borewells and purchase additional land. The household also opened a
hardware shop and began to act as village moneylenders.
Therefore, there were both structural and lifecycle factors that enabled households to establish
themselves on an upward trajectory of accumulation and to exit poverty. Changing markets for
agricultural produce, access to knowledge and information and timely involvement in agricultural
projects were all important. A common factor in all cases where households managed to lift
themselves out of poverty was reaching a later stage in the household developmental cycle. It is no
coincidence that households that made positive exits from poverty over time had gone from having
30
young dependants in 1975, to having large amount of (free) adult household labour and skills to
draw upon in 2001. Thus, within households, co-operation between household members and the
pooling of labour and resources were crucial.
There were also cases of extreme downward mobility in which households with large landholdings
were forced to sell their land due to the failure of sharecropping contracts and increasing
indebtedness. An example is given in Box 2. In these cases, the caste relations that originally
enabled households to acquire land and establish a position of status in the village often became a
constraint as households faced increasingly risky conditions of agricultural production. Another
common factor in all the cases of those who had experienced severe downward mobility was the
failure or breakdown of kin relations and co-operation within the household. Whilst households that
had moved out of poverty were effectively managing and co-ordinating the livelihood activities of
all household members, those whose economic position had deteriorated had often experienced
fragmentation, out-migration without subsequent remittances and a lack of co-operation between
household members.
Information about gender relations was not explicitly considered in interviews with panel
households so it was not completely clear how changing power relations in households, particularly
the empowerment of women, might have affected household livelihoods. However, out of the four
households where in-depth interviews were carried out about deteriorating economic position, two
were headed by women. Women were exceptionally vulnerable to the death of their husbands,
31
especially if they did not inherit any land or assets. KG was the female head of a household in
Dokur when she lost her husband (and the land and assets that he would have inherited) and then
faced the debts incurred through spending on medical treatment and funeral costs. Her son was
accused of stealing gold from the owner of land on which he was labouring and the household’s
relations with other villagers soured. It became more difficult to get agricultural labour and so the
son migrated to Gujarat for non-farm work. He sent no remittances.
Whilst the precise dynamics of gender relations were not clear, it was possible to detect growing
generational conflict, especially in households which had to sell land and assets to survive. In such
households, young people were often not interested in staying in Aurepalle and Dokur and sought
alternative livelihoods in Hyderabad. When households that were straddled over long distances
ceased to co-operate in the allocation of labour and resources, this was one way in which income
and assets were gradually eroded. Other households on downward trajectories found themselves at
the sharp end of household developmental cycles because they had many dependants or were
saddled with payments for marriages or funerals that left a heavy debt burden. Some households
had spent savings and income on medical treatment, especially for tuberculosis, and were then
unable to cope when drought came and their crops failed. Other reasons for downward mobility
were failed investments, particularly in irrigation, that led to increasing indebtedness.
There were households that, between 1975 and 2001, were ‘treading water’ and experienced little or
no change in their economic position. In most of these cases, households had attempted to invest in
agriculture or other activities, with a view to accumulating wealth and assets, but this had been
constrained by various shocks. A more detailed example is given in Box 3. In these cases, the gains
made through investments were eroded but households were coping and holding on to their
productive assets. The main shocks that households faced were related to agricultural production
and health. The drought in Aurepalle and Dokur and the failure of irrigation systems led to
widespread crop failure. Households sought alternative income within the village but demand for
agricultural labour had also declined because many large landowners had left a large proportion of
their land fallow. Thus, the most important coping strategy became migration (either for farm or
non-farm work) to other villages or to towns and cities. Here migrant labour enabled households to
avoid findings themselves on a downward trajectory.
There was a disproportionate number of female-headed households amongst those that were
‘treading water’. VP was the head of a household in Aurepalle. Compared to the household of KG,
VP was better placed following the death of her husband because she inherited his 5 palm trees. Her
income from toddy tapping was modest because she had to pay a male neighbour to carry out the
work for her but she could sell the palm juice in the village. In the slack season for tapping she
participated in the village labour market. Her modest income enabled her to continue to tread water,
though her position was highly precarious.
The re-survey of the 80-household panel from the 1975 to 1989 round of the VLS lends weight to
the findings from the 2001 survey and census. The panel demonstrates the ways in which
dependency on livelihoods that focus on agriculture became increasingly precarious. The strongest
testimonies came from households which had seen their land and other productive assets (especially
livestock) decline between 1975 and 2001 and had sought sources of income from non-farm
livelihoods. There was evidence that land distribution programmes offered a route out of poverty
but that certain poor households were not able to take advantage of the opportunity because of non-
routine expenditure, especially for medical treatments and rite of passage ceremonies.
33
5 Concluding Remarks
The story of agriculture in the villages of Aurepalle and Dokur between 1975 and 2001 paints a
rather depressing picture for agricultural livelihoods in the Mahbubnagar District of Andhra
Pradesh. Whilst agriculture remained the most important source of income for the majority of
households in Aurepalle and Dokur, the proportion of income that was derived from agricultural
activity decreased and there was a growing dependence on migration and non-farm livelihoods.
Alongside a decline in the relative proportion of income derived from agricultural activity, real
income from cultivation has also decreased. This results largely from disproportionately low price
increases for agricultural crops, especially coarse cereals such as pearl millet and sorghum,
compared to other goods and from the lower yields resulting from drought and the failure of
irrigation systems.
In order to cope with the loss of real income from cultivation, households have developed an
increasingly broad repertoire of livelihood activities. There has been both a change in cropping
patterns (increasingly towards commercial crops in the context of liberalisation, infrastructure
development and government food distribution policies) that represents diversification within
agriculture, and diversification into non-farm activities, especially labour migration in the non-farm
sector. Opportunities to migrate for non-farm work are mediated by caste rules that are more
constraining for some castes than others, and by social networks and kin relations. Migration,
however, is no ‘magic bullet’. For most households, migration required some investment, for
example to pay for transport costs or accommodation. This eroded the returns and remittances from
the activity. Furthermore, maintaining household relationships and co-operation across long
distances is difficult. Migration sometimes raised the expectations of younger members of the
households and, in extreme cases, led to a breakdown in household relations.
Diversification was a strategy taken up by landless households and by small, medium and large
farmers. Those with large land holdings and productive assets were not immune to the risks faced in
agriculture. In fact, there was only limited evidence of diversification enabling households in
Aurepalle and Dokur to accumulate wealth and assets in significant measures. The story of GL (in
Box 1), for example, was in stark contrast to the testimonies of the majority of households for
whom diversification was solely a coping mechanism. Those who experienced an erosion of income
and assets were then forced into the non-farm sector because there were no opportunities for them
in agriculture, except perhaps as very low-paid regular farm servants. The investments made by
others in irrigation and machinery, or the benefits accruing to people who received land under
distribution programmes offered a life-line to many households in the context of drought and crop
failure. Whilst some of the diversification strategies within and outside agriculture appear to have
increased incomes in real terms in the villages, diversification strategies are not themselves free of
risk and, in the prevailing agro-economic climate, often offered little more than an opportunity to
cope and mitigate risk or to tread water and hold on to productive assets for the future.
The findings beg an important question about the process of diversification in Aurepalle and Dokur
and in the semi-arid tropics of India more generally. Whilst both villages faced drought and a
subsequent dearth of water for irrigation, it was not clear whether years of drought, and only
average rainfall in intervening years, had brought about short-term or intermediate coping strategies
or a more meaningful and long-term change in the livelihood strategies of households. Given that
very few households accumulated significant wealth through diversification, it may well be that, if
future rainfall is both plentiful and timely, then there will be a return to an overwhelming
dependence on agriculture and agricultural labour, and a parallel decline in migrant labour and other
non-farm activities. However, even if there is a will to return to agriculture when improved rainfall
conditions prevail, it also remains to be seen whether households have, during the drought, disposed
34
of too many of their agricultural assets to make a serious return to farming. Similarly, whilst in
Aurepalle, population density declined between 1989 and 2001, population pressure in Dokur
continues to increase. The population of both villages increased between 1975 and 2001. This also
has implications for the future of agriculture since a continued rise in population pressure
diminishes the possibility of households gaining a livelihood from cultivation.
The diversification process, coupled with uncertainty over availability of agricultural assets in the
future, also raises important policy questions. Above all, there remains a challenge for the structure
in which government policy is made and state interventions are carried out. Whilst policy and
interventions are implemented largely along sectoral lines, household livelihoods are highly diverse.
How might the linkages between farm and non-farm livelihoods be exploited within existing policy
channels to help generate new sources of livelihood? One appropriate strategy here might be to
encourage forward and backward linkages to agriculture by supporting enterprises that either enable
better agricultural production (for example village repair services for agricultural machinery and
implements) or the process of adding value to agricultural production before it leaves the village
(for example milling, food processing, packaging and transportation).
35
References
Asokan, M., Bhaskar Rao, V. and Mohan Rao, Y. (1991) Agro-economic Profiles of Selected
Villages in Mahbubnagar District of Andhra Pradesh, India. Unpublished ICRISAT Paper.
Patancheru: ICRISAT.
Alderman, H. and Paxson, C. (1992) ‘Do the Poor Insure? A Synthesis of Literature on Risk and
Consumption in Developing Countries’. Policy Research Working Papers. Washington DC:
World Bank.
Andhra Pradesh Directorate of Economics and Statistics (2002) ‘Average Annual Rain, District
Wise 1995–2000’ www.andhrapradesh.com/apwebsite/tables. Accessed 14 November 2002.
Bryceson, D.F. (1999) ‘African Rural Labour, Income Diversification and Livelihood Approaches:
A Long-term Development Perspective’. Review of African Political Economy 26 (80): 171–89.
Davies, S. (1996) Adaptable Livelihoods: Coping with Food Insecurity in the Malian Sahel.
London: Macmillan.
Datt, G. and Ravallion, M. (2002) ‘Is India’s Economic Growth Leaving the Poor Behind?’ Journal
of Economic Perspectives 16 (3): 89–108.
Deaton, A. and Mullbauer, J. (1982) ‘On Measuring Child Costs: With Applications to Poor
Countries’. Journal of Political Economy 94 (August): 720–41.
Ellis, F. (2000) Rural Livelihoods and Diversity in Developing Countries. Oxford: OUP.
Ellis, F. (1998) Survey Article: Household Strategies and Rural Livelihood Diversification. Journal
of Development Studies 35: 1–38.
Gill, G.J. (1991) Seasonality and Agriculture in the Developing World: A Problem of the Poor and
Powerless. Cambridge: CUP.
Government of Andhra Pradesh (2001) Consumer Price Index Numbers for Agricultural Labourers
– Andhra Pradesh and All India www.ap.gov.in/apbudget/tab17_2.html. Accessed 2nd
September 2002.
Government of Andhra Pradesh, Directorate of Economics and Statistics (2000) Andhra Pradesh
Economic and Statistical Bulletin Vol 48 No2
Government of India Planning Commission (2001) Report of the Working Group on Public
Distribution System and Food Security for the Tenth Five Year Plan (2002–2007),
http://www.planningcommission.nic.in/wrkgrp/wg_pds.pdf, Accessed 10/09/02.
Gulati, A. and Kelley, T.G. (1999) Trade Liberalization and Indian Agriculture. New Delhi: OUP.
Jodha, N.S., Asokan, M. and Ryan, J.G. (1977) ‘Village Study Methodology and Resource
Endowments of the Selected Villages in ICRISAT’s Village Level Studies’. Economics
Programme Occasional Paper 16 (Village Level Studies Series 1.2).
Meenakshi, J.V. and Ray, R. (forthcoming, 2002) ‘How Have the Disadvantaged Fared in India? An
Analysis of Poverty and Inequality in the 1990s’ in Sharma, K. (ed.), Trade Policy Regime,
Growth and Poverty in Asian Developing Countries, London: Routledge.
Morduch, J. (1995) ‘Income Smoothing and Consumption Smoothing’. Journal of Economic
Perspectives 9 (3): 103–14.
Netting, R.M. (1993) Smallholders, Householders: Farm Families and the Ecology of Intensive,
Sustainable Agriculture. Stanford: Stanford UP.
Rao, K.P.C. (1999) ‘Development of Dryland Agriculture: Policy Issues’ in Singh, H.P.,
Ramakrishna, Y.S., Sharma, K.L. and Benkateswarlu, B. (eds.), Fifty Years of Dryland
Agricultural Research in India, Central Research Institute for Dryland Agriculture, Hyderabad:
565–72.
Reardon, T., Delgado, C. and Matlon, P. (1992) ‘Determinants and Effects of Income
Diversification Amongst Farm Households in Burkina Faso’. Journal of Development Studies 28
(2): 264–96.
Singh, R.P., Binswanger, H.P and Jodha, N.S. (1985) Manual of Instructions for Economic
Investigators in ICRISAT’s Village Level Studies. Unpublished ICRISAT Manual.
36
Singh, R.P., Asokan, M. and Walker, T.S. (1982) ‘Size, Composition and Other Aspects of Rural
Income in the Semi-arid Tropics of India’. ICRISAT Economics Program Progress Report 33.
Singh, R.P. and Asokan, M. (1981) ‘Concepts and Methods for Estimating Rural Income in
ICRISAT Village Level Studies’. ICRISAT Economics Program Progress Report 28.
von Braun, J. and R. Pandya-Lorch (eds.) (1991) Income Sources of Malnourished People in Rural
Areas: Micro level Information and Policy Implications, Washington DC: IFPRI.
Walker, T.S. and Ryan, J.G. (1990) Village and Household Economies in India’s Semi-arid
Tropics. Baltimore: Johns Hopkins University Press.
37
1. Village Information
1.1. Village: Aurepalle/Dokur 1.2. Mandal: Madgul/Devarkadra
1.3. District: Mahbubnagar 1.4. State: A. P. 1.5: Census code No.-------------
2. Household Information
2.1. Name: ------------------------------------ 2.2. Fathers’ Name:-----------------------------
2.3.Caste: ------------- 2.4. Education: -------------- 2.5. Main occupation: ------------------
2.5. Secondary occupation: ----------------
3. Family Structure
Family members Working Caste Farm Off-farm Migration Others
on own occupation wages work
farm
Male
Female
Children (<15)
Total family size
4. Assets
4.1. Land (Acres)
Particulars Dryland area Irrigated area Total area
Own land
Leased/shared in
Leased/shared out
Operated area (owned area +
leased/shared in – leased/shared out
area)
4.2. Livestock
Particulars Number In milk
Bullocks
Cows
He Buffaloes
She Buffaloes
Young Stock
Goat
Sheep
Pigs
Others
38
4.3. Machinery
Particulars Number
Tractor
Thresher
Oil engine
Power sprayer/duster
Iron/wooden plows
Seed drill
Blade harrows
Bullock cart
Manual sprayer/duster
Rice/floor mills
Others
6. Sources of income
% share to total HH income % time spent
6.1. Agriculture --------------- ----------
6.2. Livestock --------------- ----------
6.3. Caste occupation --------------- ----------
6.4. Regular job --------------- ----------
6.5. Farm work --------------- ----------
6.6. Non-farm work --------------- ----------
6.7. Business (trade) --------------- ----------
6.8. Migration --------------- ----------
6.9. Sale of CPR products --------------- ----------
6.10. Others --------------- ----------
39
1. Family Information:
Leased/shared-in
Leased/shared-out
4. Plot wise Input/output details (For one representative plot for each crop/crop combination)
Crop (s): __________________ Plot size (Ac): ___________ Plot No. ________
Season: __________________ Ratio/Proportion: ________________
5. Livestock
7. Seasonal migration
Sex Place of Distance Type of work Work period Wage Amount Amount
work (Days/Months) (day/month) Received spent
Thresher
Flour mills
Electric motors
Power sprayers
Bullock cart
Minor implements
Unproductive assets
Residential House
Pucca, Semi pucca, Kacha etc.
Residential plot
Household articles (TV, Fridge, Two
wheeler, fan etc.)
Jewellery
Consumer durable
Household utensils
Others
44
10. Expenditure
Type Quantity (Kg/L) Month/Year Price/Unit Total Amount
Cereals
1. Rice
2. Sorghum
3. Pearl millet
4. Wheat
5. Others
Pulses
1. Pigeonpea
2. Chickpea
3. Greengram
4. Cowpea
5. Others
Oils
1. Groundnut
2. Ghee
3. Others
Vegetables
Education
Travel
Ceremonial
Entertainment
Clothing
Medical
Farm inputs
1. Material inputs
2. Labour wages
3. Farm electricity
Inputs purchased for
handicrafts
Other expenses
1.
2.
45
12. Does any member of the household a Position/s of responsibility in any of the village organisations
(either formal or informal)?
Yes / No
Yes/No
What type of ration card do you have? White: _____ Pink: ______
Which of the Govt.schemes and programmes do you benefit from these programmes?
Yes/No
Programme Benefits
Yes/No
Self Help Groups
Janma Bhoomi
Other Progrms
Prog1.
Prog2.
Prog3.
46
13. What benefits did you receive from the following Government programmes during 2000-01?
Educational allowance
Other
Prog 1.
Prog 2.
Prog 3.
Jatiya Prasuti Supply of nutritious Provides Rs 1,000/- to pregnant women before and after
Sahaya Programme food for pregnant delivery to get nutritious food.
women
Construction of Maintain cleanliness of Provides Rs. 2,000/- for the construction of individual
toilets for the the village toilets to maintain cleanliness of the village
public
Banning child To eliminate child Enforce the ban on child labour and prevent the practice by
labour labour addressing this problem. The state will enforce the child
labour abolition act and ensure that all vulnerable children
have access to education and parents will be made
responsible for ensuring that their children go to school.