You are on page 1of 34

130

CHAPTER VII
DETERMINANTS AND LINKAGES OF RURAL NON-FARM
SECTOR
A number of pre-requisites and compatible socio-economic environment are
must for the development of rural non-farm sector (RNFS). In other words, the
development of social and physical infrastructure is a sine qua non for the
development of this sector. There are also chances that the RNFS may develop
because of stagnant agriculture or an imbalanced growth strategy. Moreover, the
numerous linkages between the rural non-farm sector and other sectors of the
economy, mainly agriculture and urban based industrial sector are needed to be
developed for any sustainable and productive development of this sector. The
literature suggests a number of factors and processes which have a bearing on the
development of this sector. However, the debate regarding their relative importance
and significance has not yet settled. The discussion on a number of important factors
responsible for the development of RNFS is divided into two sections. Section I deals
with the growth linkages between agricultural development and RNFS on one side,
and other determinants influencing the development of RNFS on the other. And, the
empirical evidence generated on RNFS by the present study has been presented in
Section II.
I
7.1.1 Agriculture led growth linkages of RNFS
There is a considerable body of literature which maintains that agriculture is
the prime mover behind the emergence, sustenance and growth of rural non-farm
sector. The processes through which the stimuli of agricultural advancements are
transmitted to RNFS are popularly known as agricultural growth linkages in the
literature.

The dominant view before 1970s was that the agriculture has week

stimulating linkages (Hirschman, 1958). This view was based on the impression that
the traditional agriculture used few capital goods purchased from other sectors. Mellor
and Lele (1973) asserted that even the traditional agricultural systems purchased nonagricultural goods. However, modern agriculture, based on high yielding and cost
reducing techniques, really creates dynamic forward and backward linkages with non-

131
agricultural sectors. They argued that agriculture cannot be denied its role in the
process of economic development on the misplaced belief of its having devoid of or
having weak linkages with other sectors.
The basic premise of Mellor (1976) is that increases in agricultural production
through cost reducing technological innovations normally distribute the initial
benefits largely to the already more prosperous and better endowed farmers. These
incremental increases in the hands of relatively better-off farmers stimulate
consumption of non-food goods which in turn gives rise to numerous non-farm local
occupations; which are local, labour intensive and requiring low capital in the
production process. Increased agricultural production also creates a demand for agro
processing, marketing and transport facilities through forward linkages with non-farm
sector and at the same time increased demand for inputs like pump sets, mechanical
ploughs, seed drillers, etc. through backward linkages give rise to numerous local
firms producing, supplying/distributing and repairing these inputs.
Agricultural growth linkages are normally classified into following five
categories:
(i)

Production linkages
Production linkages of agriculture with rural non-farm sector are of two types:

(a) forward linkages reflect the need to process, transport and trade, distribute and
sale, etc. of agricultural produce. Many agro-processing units like rice-shellers, flour
mills, fruit processing units, and the like develop in the course of development; (b)
backward linkages are generated by the demand of agricultural production process
itself. Johnston and Kilby's analysis (1975) of farm equipment in India, Pakistan and
Taiwan stresses that traditional tools are most often made by rural artisans, while
improved implements and irrigation pumps and motors are likely to be fabricated by
light engineering workshops located in rural towns. Tractors, combine harvesters,
chemical fertilizers and other large items with high performance characteristics, tend
to be produced abroad or in urban areas having little potential for promoting
local/rural non-farm linkages. Consequently, the nature of agricultural technology
adopted has important effects on rural non-farm linkages. Even when the above said
high capital intensive items are produced in metropolitan towns or even in foreign
locations, a wide spread network of dealership, spare-parts and repair shops are

132
generally set up in large villages or in mandi towns where agriculture produce is
normally marketed (Chadha, 1986). The market-led and market-linked agricultural
development in many states of India have experienced small linkages (Gill and
Ghuman, 2001)
(ii)

Consumption linkages
Consumption linkages manifest themselves when the farm households spend

their earnings from agriculture on non-farm goods and services. As their incomes
increase, farm households increase the share of their expenditure devoted to non-food
items (Mellor and Lele, 1973; Hossain, 2004). It has also been found that
consumption linkages are more important than production linkages (Haggblade et al.,
1989; Hazell and Haggblade, 1990; Bell et al., 1982; Hazell and Ramasamy,1986;
Shukla, 1992).
(iii)

Factor market linkages


In rural labour markets, a strong seasonality of demand in agriculture

generates corresponding surges in rural non-farm activity (Anderson and Leiserson,


1980; Fabella, 1986; Leons and Feldman, 1998) Tightening of labour markets caused
by the expansion of RNF employment has been widely acknowledged (Bhalla, 1993).
Similarly, cash surpluses from agricultural sale frequently finance non-farm
investments, while reciprocal reverse flows from rural non-farm activities finance the
purchase of agricultural inputs (Evans and Ngau, 1991). Rural non-agricultural
activities may affect agriculture directly, through a reallocation of resources, and
indirectly though investment (Ho, 1986b).
(iv)

Productivity linkages
Productivity linkages of agriculture with non-farm economy have emerged

only recently in the growth linkage discussions (Haggblade et al., 2007). Increase in
agricultural productivity may lower the price of foodstuffs and the resulting increase
in absorption may increase the productivity of poor manual workers in non-farm
sector.
(v)

Reverse linkages
Rural non-agricultural sector also impacts agricultural production and

productivity. Evans and Ngau (1991) note down that at the level of household, rural
non-farm incomes have a positive impact on agricultural productivity. Others have

133
observed that increased participation of rural households in non-farm occupations led
to a fall in family's incomes from agriculture (Kada, 1986; Ho, 1986b).
7.1.2 Quantitative estimates of agricultural led growth linkages
Quantitative estimates of the production and consumption linkages at the level
of regional economy were pioneered by Bell et al. (1982) in their evaluation of Muda
irrigation project in northern Malaysia. They estimated that $ 1 of the value added in
agricultural income generated from the project stimulated an additional $ 0.83 in the
region's non-agricultural economy. About 40 per cent of it was due to production
linkages (both backward and forward), and 60 per cent to consumption linkages.
Subsequent analysis of consumption linkages by Hazell and Roell (1983) claim that
large farmers exhibit highly multiplicative patterns of demand for regional nontradable commodities and that their consumption behaviour is, therefore, key source
of growth in the regional economy. Hazell and Ramasamy (1991) have estimated
agricultural growths multipliers over the 1970s for the North Arcot district of Tamil
Nadu in India. They estimated that as a result of production and consumption
linkages, every 100 rupees increase in agricultural income induced an additional 82
rupees income in other sectors of the rural economy. Production linkages accounted
for half of the income, the other one-half of the increase was due to consumption
linkages. Haggblade et al., (1989) estimate that African regional growth multipliers
are approximately 60 per cent those in Asia.
Haggblade et al. (2007) provided a much useful survey of the studies on farmnon-farm linkages. A cross sectional econometric study by Hazell and Haggblade
(1990) across 85 different districts in India analyzed the relationship between
agricultural growth and growth in rural non-farm economy. It estimates agriculture to
rural non-farm income multiplier of the order of 0.64, distributed as 0.39 in rural
towns and 0.25 in rural areas. But these multipliers are not invariant. Their values
increase with the level and growth of agricultural development. The value of
multiplier is 0.93 in Punjab and Haryana and 0.46 in low productivity states such as
Bihar and M.P. From this evidence, the study claims that rural non-farm sector is
driven primarily by agricultural growth.
Agriculture influences not only the level but also the composition of rural nonfarm employment. When demand for agricultural labour goes up as it does initially in

134
most green revolution settings, it pulls the agricultural wage rate upwards and the
opportunity cost of labour in non-farm pursuits, thereby rendering low-return nonfarm activities uneconomic. This leads to the demise of many low-productivity craft
and household manufacturing activities and agricultural supporting services
(Haggblade and Liedholm, 1991). The artisans who earlier used to make a living by
producing and selling these products start disappearing from the scene and a large
proportion of these join the ranks of agricultural labor (Gill, 1980). But,
simultaneously, there is a growth of higher-return non-farm activity such as
mechanical milling, transport, commerce, and personal, health and educational
services (Bhalla, 1981; Hossain, 2004). Advancements in agricultural productivity,
thus, change the very nature, structure and organization of rural non-farm enterprises
(Bhalla, 2004).
Many studies from rural India show a positive correlation between earnings
per worker in agriculture and development of rural non-farm activity (Chadha, 1986;
Papola, 1987; Vaidyanathan 1986; Bhatt, 1998; Shukla, 1991, 1992; Rosegrant and
Hazell, 2000). Although cross-section comparisons of agricultural and rural non-farm
income typically yield positive correlation, one cannot necessarily infer causality
from these associations. Favourably endowed zones may simultaneously attract highproductivity agriculture, infrastructure investments, improved agricultural extension
and technology adoption, public investment in schools, roads and development,
offices, as well as external non-farm investment (Haggblade et al., 2007).
Study results from slow-growing agricultural regions suggest that a sluggish
agriculture gives rise to anemic non-farm employment and wage rates. Though rural
non-farm employment may actually increase in lackluster agricultural zones
(Haggblade and Liedholm, 1991; Bhalla, 1994), it normally emerges in low wage-last
resort activities such as basket weaving, embroidery and gathering activities. It means
that level of rural non-farm employment is not always a true indicator of significance
of these activities towards, households' living standards. Productivity or returns from
them certainly is a better criterion for judging their economic importance.
In Latin America, rural linkages appear to be low because of the extreme
inequality in land and income distribution, absentee ownership of large land holdings,
and the consequently feeble rural consumption and input linkages emanating from the

135
often urban-based and urban-supplied Hacienda owners (Haggblade et al., 2007). In
South Africa increasing non-farm activity in some rural communities has been
accompanied by a decline in agricultural production and income (Low, 1981). In
Kenya, regions with higher non-farm incomes tend to have higher agricultural income
levels (Evans and Ngau, 1991). In Bicol, a resource-poor region of the Philippines, a
rapid increase in non-farm income in absolute and relative terms accompanied a
decline in both agricultural and total income in the late 1970s and 1980s (Reyes,
1987). These results prompted the researcher to assert that "the direction and causality
of growth in non-farm and farm activity is far from clear" (Leons and Feldman, 1998).
Some scholars (Vyas and Mathai, 1978; Islam, 1986; Ray, 1994) argue that
unless the structure of rural demand is broad based, compassing the demand generated
by all the sections of rural society and not merely of richer farmers, rural non-farm
sector would not hold much prospect for absorbing the underutilized/unutilized rural
labour force and for pulling the workers participating in it out of the poverty trap.
These scholars therefore recommended that unless rural assets, especially the land are
not equitably distributed the virtuous interaction of farming and non-farming will not
play to its fullest potential.
Forces other than local agriculture may strongly influence the growth of nonagricultural activities (Hariss and Hariss, 1984; Hariss, 1987, Amsden, 1991; Shih,
1983; Hart, 1998; Bhalla, 1993b). In the Indian context, the recent closure orders of
Supreme Court of India of the household-based/small non-household manufacturing
units operated from the residential areas of Delhi, pushed many of these units into the
rural regions of neighbouring UP, Haryana and Rajasthan apart from rural areas of
Delhi itself. Rural industrialization, thus, should not be treated as an adjunct to
agricultural growth; but rather as a development which needs to be fostered
independently (Papola, 1987).
In the Indian Punjab, the epicenter of green revolution, high agricultural
incomes did not result in industrial development in general and rural industrialization
in particular. There was a drain of investible surpluses to far flung regions via banking
mechanisms (Chadha, 1986; Gill, 1994). Policies biased against industrial bourgeois
(Chadha, 1986) or the interactions of majority nationalism and minority nationalism
(Singh, 2008) are said to be important reasons for industrial backwardness of the

136
state. Thus, the assertions of agriculture-led employment strategy of Mellor and
associates, based on a mechanical interpretation of agricultural and non-agricultural
equations, in disregard to social, political and institutional contexts, seem to have a
limited relevance for the policies and programmes of non-farm employment
generation in rural areas of many a developing countries of the world. Agricultural
growth may be a necessary condition for rural diversification, but certainly not a
sufficient condition (Hariss, 1991).
7.1.3 Pull and push factors in RNF employment
The extent, growth and variation in rural non-farm involvement have also
been explained by the conceptual framework of 'pull' and 'push' factors. 'Pull' factors
are those which lure the rural workers towards high productive activities/areas where
relative returns to labour are higher. A vibrant non-farm sector is a major 'pull' factor.
On the other hand, when rural workers are joining the non-farm sector even when
wages here either are equal to or lower than agricultural wages, they are said to be
pushed by the farm sector. It may be due to non-availability/access to land,
fragmentation, marginalization and unviable farm holdings or may be due to
incapacity of agricultural sector to absorb increments of labor force. There is no
unanimity as to whether it is the operation of 'pull' factors which is behind the
observed movement of rural workforce from agriculture to non-agricultural activities
or they are pushed out of agriculture.
Agriculture-led-employment school of thought propagates that the most
important 'pull' factor in luring the rural workers towards non-farm pursuits is the
emergence of a vibrant and productive rural non-farm sector caused by dynamic
agriculture through a variety of production and consumption linkages. The study has
already referred to the studies by Hossain (2004), Chadha (1986), Bhalla (1981), Raj
(1976), Ho (1986b), Oshima (1986b), Haggblade et al. (1989), Hazell and Haggblade
(1990) and various others cited in support of this contention.
There are other studies which emphasize that it is the rather operation of 'push
factors' which is the main reason for the observed increase in rural workers'
engagement in non-farm occupations. Vaidyanathan (1986) emphasizes that there is a
significant and positive relation between rural unemployment rate and the incidence
of non-agricultural employment and this association is much stronger than in the case

137
of all other explanatory variables on which he estimated an importance of rural nonagricultural employment in total employment in rural India. This gives credibility to
the notion that non-agricultural activities in rural areas may be acting to some extent
as a residual sector absorbing labour which cannot find work in agriculture. Similar
findings have been reported by Basu and Kashyap (1992), Simmons and Supri (1995),
Parthasarthy et al. (1998), Ghuman (2005) and Ranjan (2009) which emphasize that it
is mainly the increased helplessness and marginalization of rural workers in the wake
of increased landlessness and fragmentation of land holdings, grossly unequal land
holdings and sinking growth rates of agricultural output and employment. Under these
circumstances, rural non-farm occupations are nothing but refuse activities towards
which workers are pushed to avoid deprivation.
The actual reality always lies in the middle of these two contrasted situations.
Ho (1986a) and Kada (1983) demonstrate that it is the simultaneous and complex
interaction of both push and pull factors which are the reason behind observed
expansion of rural non-farm employment. Urbanization, especially the preponderance
of small and medium typed towns, also pulls the rural workforce for joining the rural
non-farm sector. White (1991) argues that the conceptual frame work of "pull versus
push" factors, in fact, does not explain much and the debate on the relative
importance of pull and push factors is not very productive as different groups of rural
society and households enter into different kinds of rural non-farm pursuits for
different reasons, producing all-together different outcomes.
7.1.4 Other determinants of RNFS
(i)

Infrastructure
Almost all the studies on the theme of rural non-farm employment, in all the

regions, point to the development of infrastructure as one of the basic factor in


promoting rural non-farm employment. Fabella (1985) has emphasized population
density, education, electrification, irrigation and high yielding varieties of rice as the
main modernizing factors determining the development of RNF employment in
Philippines. Oshima (1986a), while contrasting the development of rural non-farm
sector in Japan and Taiwan with South Korea and these countries with that of SouthEast Asian ones, argues that the 'transport indexes' are approximately correlated with
the share of off-farm income in Japan, Taiwan, South Korea and Philippines. Rural

138
electrification and telecommunication are again very important for the growth of zgoods component of rural non-farm sector (Ranis and Steward, 1993; Lanjouw,
1998). Similarly, Park (1986), Kada (1986), Ho (1986b) and Saith (1987). All these
studies have underlined the importance of infrastructural facilities for promoting rural
non-farm employment in East Asian countries. Hazell and Haggblade (1990) have
also identified the importance of rural infrastructure like roads, electrification,
banking services etc. in enhancing the size of the agricultural growth multipliers.
Shukla (1992) has estimated that infrastructural facilities of roads, electricity, posts
and telegraphs have statistically positive influences on the level and density of rural
non-farm employment.
It is not only physical infrastructure such as transport, communication, electric
power which is important; equally important is social infrastructure such as
educational, medical and recreational facilities (Oshima, 1986a). Since services will
be among the most rapidly growing rural non-farm activities, investment in human
capital will be essential for realizing the potential gains. But there is another side to
infrastructural development. The development of transport and communication
facilities tend to lift the barriers between cities and rural areas leading to an easy and
cost-effective availability of goods and services produced at urban locations to rural
consumers. More often than not the result is demise of rural manufacturers and
artisanal products though services have been reported to be less affected by this
increased confrontation. Also, this development affects the tastes, preferences and
attitudes of rural households and orients them towards urban products again hitting
the rurally produced goods from demand side.
(ii)

Education
Education not only improves an individual's qualifications for non-agricultural

jobs, it also increases his ability to allocate his work time efficiently among
alternative income-producing activities (Huffman, 1980). Education contributes to
higher productivity in trading, construction, service and manufacturing activities.
Secondary education stimulates entrepreneurial capacity, whereas primary education
enhances the productivity of the workforce, including foremen, supervisors, and other
middle-level personnel, makes it easier to impart on-the-job training. Education is an
important factor in the choice of non-farm activity and it raises productivity in the

139
non-farm sector (Khandker, 1995). In fact, a non-linear relationship between
educational level and income has been found (Vijverberg, 1995).

Lanjouw (1998)

also notes that returns to education are quite substantial in the non-agricultural sector.
A review by Reardon et al. (2001) pertaining to Latin American countries showed
very vividly that education determines participation and success in RNF employment
and income. More and higher level education tended to mean more non-farm wage
employment in high productivity, well paying jobs. The more educated tend to avoid
farm wage labour and gravitate towards non-farm wage employment and only
secondarily towards non-farm self employment.
(iii)

Urbanization
Urbanization is another major pull force for the expansion of rural non-farm

activity. Ho (1986a) and Kada (1983) argue that proximity to urban areas is an
important determinant of both the extent of rural involvement in non-agricultural
activities and the quality of that involvement (as measured, for example, by their
average income earned from non-agricultural sources). In India, the most dynamic
growth areas in the rural non-farm sectors rely heavily on urban and export demand
and possess negligible linkages with agriculture (Fisher et al. 1997).
The linkages between rural based small scale units and urban based larger
units (in manufacturing as well in trade) have been noted widely (Ho, 1986b). Otsuka
(2007) contends that these linkages are rather the defining characteristics of mature
phase of rural industrialization of many countries like Japan, Taiwan and South
Korea. Rural industry in many countries seems to develop most vigorously near an
urban industrial nexus either on the peri-urban fringes of major cities or within a 2530 mile radius of a major economic centre (Yusuf and Kumar, 1996; Fafchamps and
Shilpi, 2003). There is evidence that productivity is greater for those located near
large towns and cities (Lanjouw and Lanjouw, 1995).
In the Indian context, Jayaraj (1994) found the factor of urbanization
interacting in a complex way in explaining the extent and variation of rural non-farm
activity in the state of Tamil Nadu in India. Micro level studies reveal that temporary
migration of labour force from rural to urban area, particularly of commuting variety,
account for a sizable portion of workforce in various economic activities of the urban
centres. It also forms a major share of off-season employment of agricultural labour

140
and small farmers (Basu and Kashyap, 1992). It is being increasingly recognized that
the employment linkages generated by investment in urban areas could be stronger
than those from agriculture (Basant, 1994). In the mature phase of agricultural
growth, rural demand for non- agricultural goods produced in urban areas may
increase faster than for goods produced locally (Chandrasekhar, 1993). However,
there is no one to one or monotonic relationship between increase in urban population
and activity in an economy and the level of its rural non-farm employment because
the processes of urbanization may have a differential impact on different types and
forms of non-agricultural activities. Different sized towns can influence the structure
of rural non-farm employment in a different manner. Other things being the same, a
more equitable pattern of urbanization biased towards small towns will probably be
more conducive to the growth of rural non-agricultural sector than a less equitable
one (Shukla, 1992; Basant, 1994).
(iv) Inequality in land distribution
Oshima (1986a) lists greater inequality in the distribution of arable land
ownership accompanied by a much larger share of non-cultivating land owners and
tenants (including the landless) as against owner-cultivators as one of the reasons for
much larger share of off-farm employment in the Philippines than in South Korea.
Unni (1991), too, found that unequal land distribution was positively and significantly
associated with the percentage of male non-agricultural workers. This could occur, he
reasoned, because of a positive stimulus of land concentration (via surplus generation)
or due to distress diversification. Dev (1990) also confirmed the factor of inequalities
in asset structure as an important variable explaining rural non-farm employment.
Vaidyanathan (1986), however, found that other things being equal, higher inequality
of operational holdings seem to go with lower incidence of non-agricultural
employment. The importance of redistributive land reforms enhancing access to land
to a large mass of rural populace in promoting localized rural non-farm activities has
been emphasized by several studies (Islam, 1986; Saith, 1991, 1992; Ray, 1994). The
experience of development of rural non-farm sector in West Bengal testifies to it. It is
claimed that after the land reform measures, the demand for agricultural inputs by
marginal and small farmers has greatly increased. With the result that 40 per cent of

141
rural workers in the state are engaged in non-farm activities (West Bengal Human
Development Report, 2004).
(v)

Public Investment
Direct public investment in rural areas is also one of the main sources of rural

non-farm

employment.

Public

investment

in

agricultural

marketing,

rural

infrastructure, roads and bridges, and irrigation facilities has generally spurred RNF
employment (Shukla, 1992). Visaria (1995) highlights the contribution of investment
in public utilities resulting in higher involvement of rural labour force in nonagricultural employment. Men are the main takers of the jobs that arise from such
investment. Fisher et al. (1997) also note that almost 20 per cent of total non-farm
employment in rural India was generated by public sector services. The contraction in
agricultural investment in particular and in rural investment in general in India is said
to be the main reason behind the wide spread decline in the rate of growth of
employment in rural trade which was linked to slowdowns in agricultural growth
through lower agricultural investment since the mid-1980s and specially after the
'reforms' period (Chadha and Sahu, 2002; Sen and Jha, 2005). But contraction in
public investments under the dictates of Structural Adjustment Programmes, followed
by many African Sub-Saharan countries since the early 1990s, have been shown to
have triggered a huge unplanned income diversification in African rural areas
(Bryceson, 1999).
(vi)

Commercialization of Agriculture
The higher percentage of commercial crops in the cropping pattern is said to

create

opportunities

of

tertiary

employment

(Dantwala,

1953).

Increased

commercialization and specialization in agriculture lead to general commercialization


of rural economy also which, other things being equal, is likely to reduce rural nonfarm activity as a part time or secondary occupation but to encourage rural industry as
a specialized activity. Sharma (2005) emphasizes the diversification of agriculture
into commercial off-season vegetables and fruits in Himachal Pradesh as the main
reason for the increase in rural non-farm employment in the state in recent period.
Vaidyanathan (1986) also notes a significant positive association of percentage of
area under non-food grain crops and census estimates of rural non-farm activities in
India.

142
7.1.5 Determinants of household participation in RNFS
(i)

Land holding size


As agricultural land becomes scarce, households must seek compensating

earnings in non-farm economy. For this reason, landless households typically depend
most heavily on non-farm earnings (Anderson and Leiserson, op.cit). However, the
effect of land holding on participation and earning from RNF activity is complex. A
household with a larger land holding may be more committed to agriculture, thus,
exhibiting a negative relationship of land holdings with non-farm activity. But large
land holding may also mean more wealth and contacts which are important for getting
entry into high-end non-farm activities. Households with large land holdings are, thus,
better placed in search of and engaging in rural non-farm endeavours. A study by
Kijima and Lanjouw (2005) of rural India suggests that the probability of employment
in regular non-farm activities and in non-farm self employment (relevant to
agricultural labour) is significantly higher for those with more per capita land
holdings.
(ii)

Education
Less educated households rely on low-paying farm wage employment or very

low productivity non-farm pursuits (Lanjouw and Shariff, 2004). In a study of


Ecuador, Elbers and Lanjouw, (2001) found that the least educated were found to do
low-paying non-farm work in manufacturers or services. Those with basic education
were found managing small enterprises mainly in manufacturers and those with more
education work in the higher paying jobs such as teachers or manage large local
enterprises.
(iii)

Seasonal nature of income streams from agriculture


Risk and uncertainty in agriculture also forces the households into rural non-

farm employment. Ellis (1998) asserts that the prime motive and consequences of
successful household diversification is to reduce vulnerability. He distinguishes
"rational risk-management" from default "coping strategies". Both seasonal
smoothing and risk diversification can be very important factors in environment
where agricultural output varies greatly over the year and across years and where
mechanisms for smoothing income such as credit and transfers are costly or absent
(Lanjouw and Lanjouw, 1995).

143
(iv)

Credit markets
The failures of credit or insurance markets drive households to self-ensured

and self-funded input purchases (Evans and Ngau, 1991). Where credit and insurance
markets are missing, rural non-farm activity becomes a vehicle of self insurance and
for financing agricultural inputs and assets. Migration and remittances tend to help the
households in coping with the missing credit markets. The household left behind
becomes less inclined to engage in rural non-farm activities. Out-migration, thus,
becomes an important factor which discourages the household from joining the nonfarm activities (Simmons and Supri, 1995). Foster and Rosenzweig (2008), however,
found that in rural Pakistan, migration has positive impact on rural non-farm sector
both by altering the size and composition of the local labour force and also by directly
affecting incomes, given the size of the local labour force.
(v)

Household size and structure


Household size and structure is positively and highly significantly related to

the incidence of RNF activities. Where household size is large, it is more likely to
participate in RNF activities (Simmons and Supri, 1995). This study also found
number of male workers in a household to be significantly related with the incidence
of RNF employment, but they termed the 'household size' a superior predictor than the
number of male workers. Reardon (1997) also provides evidence showing positive
association of household size and participation in non-farm occupations.
(vi)

Gender composition
Gender is another important factor influencing participation patterns and

trends in the RNF sector. The role of gender enabling or restricting access to
economic activity also varies from country to country. Women tend to have lower
participation in non-farm sector than men (Coppard, 2001). They earn lower than men
per unit of time of work done. They are the main victims of the process of evolution
of rural non-farm economy (Haggblade and Liedholm, 1991; Som et al., 2002). It is
only in regions where new labour intensive industries are located near farms and farm
steads that a large number of women tend to join non-farm wage work (Ho, 1986b).
(vii)

Ethnicity and caste


Ethnicity is an important determinant of participation in rural non-farm

activities. Scheduled Castes (SCs) in India, for example, are more likely to be found

144
in agricultural wage labour than in other lucrative segments of rural non-farm activity.
They are more in casual wage segment than in regular wage employment as compared
with non-SC sections of the workforce (Lanjouw and Shariff, 2004; Thorat and
Sabharwal, 2005).
II
7.2

Empirical evidence from present study

7.2.1 Education and RNF activities


In order to examine whether education has any influence on household
participation in rural non-farm employment, the study first examined the association
of educational attainments of household heads and the extent of their participation in
non-farm activities. It is evident from Table 7.1 that there is positive relation between
the level of education and the employment in non-farm activities.
Table 7.1: Educational attainment of household heads and household participation in
farming and non-farming occupations in sampled villages
Educational slab
Type of household
Percentage of
total*
RNF
Farming
Total
Illiterate
1943
1821
3764
46.37
(51.62)
(48.38)
(100.00)
Up to Primary
613
371
984
12.12
(62.30)
(37.70)
(100.00)
Beyond primary but
1834
1010
2844
35.03
up to high
(64.49)
(35.51)
(100.00)
Beyond high but up
180
76
256
3.15
to senior secondary
(70.31)
(29.69)
(100.00)
Graduate and above
158
53
211
2.60
(74.88)
(25.12)
(100.00)
Others (skilled)
53
06
59
0.73
(89.83)
(10.17)
(100.00)
Total
4781
3337
8118
100.00
(58.89)
(41.11)
(100.00)
Notes: 1. Due to rounding up percentage shares may not total to 100
2. Figures in brackets denote percentage share.
3. * denotes the percentage of total households in a particular education slab.
Source: Field Survey.

The data show that among 3764 illiterate household heads, 51.62 per cent are
in RNF and 48.38 per cent in farming. In the category of household heads which had
studied up to class fifth, one finds that out of 984 households, 62.30 per cent are in
RNF and 37.70 per cent in farming. Similarly, in the educational slab of more than
primary but up to matriculation, the percentage of RNF households has gone further
to 64.49 and that of farming decreased to 35.51 pr cent. Amongst the rural households

145
whose heads were either graduates or above, about three-fourth (74.88 per cent) were
in RNF. Similarly, of the technically qualified heads as high as 89.93 per cent of the
households had chosen rural non-farm activities as their major source of livelihood (in
terms of employment) and only 10 per cent remained dependent solely in farming.
It is obvious that formal education of household heads is very important for
the engagement of households in non-farm pursuits. Furthermore, even primary
education increases the households participation in RNF by as much as 11 percentage
points. It seems that education empowers household heads in making informed
choices about livelihood sources. It further revealed that though less than 1 per cent of
households were having their heads as technically qualified, yet about 90 per cent of
them preferred RNF as primary activity.
Across the different zones, the less developed zone (zone-III) has a lower
percentage of educated and skilled persons (Table 7.2). However, the relationship
between education and skill level with the employment in non-farm sector is almost
similar to the aggregated data. However, the illiteracy is higher in the least developed
zone as compared to the developed zones. Significantly, the proportion of household
heads, having education above primary level, is lower in the least developed zone
than that in the developed zones.
Table 7.2: Education attainments of household heads and household participation in
non-farming occupations by different zones
(Percentage share)
RNF households
Percentage of total non-farming
Education slab
households in each education slab
Zone I
Zone II
Zone III
Zone I
Zone II
Zone III
Illiterate
806
452
685
40.04
32.22
50.18
(64.27)
(51.89)
(41.79)
Up to primary
243
178
192
12.07
12.69
14.07
(65.15)
(69.80)
(53.93)
Beyond
primary
788
647
399
39.15
46.12
29.23
but up to high
(65.02)
(72.94)
(53.56)
Beyond high but up
91
48
41
4.52
3.42
3.00
to senior secondary
(73.98)
(71.64)
(62.12)
Graduate and above
65
60
33
3.23
4.28
2.42
(77.38)
(76.92)
(67.35)
Others (skilled)
20
18
15
0.99
1.28
1.10
(90.91)
(90.00)
(88.24)
Total
2013
1403
1365
100.00
100.00
100.00
(65.61)
(64.42)
(47.53)
Note: Figures in brackets denote percentage share out of RNF and agricultural households together in
respective categories.
Source: Field survey.

146
Table 7.3 shows that like the household heads, the percentage share of RNF
workers also increases with the increase in their education level. Though 33.58 per
cent of all the workers in study villages are illiterate, only 34.79 per cent are illiterate
in RNF and the rest, i.e., 65.21 per cent are in pure agricultural sector. Of the workers
who are read up to primary level, almost 52 per cent are in RNF and the remaining 48
per cent in farming. Of all the workers who have studied from class 10th up to senior
secondary, graduation and above and had any technical education, 68.25 per cent,
80.00 per cent, and 90.98 per cent respectively are in RNF. It means that educated
workers prefer non-farming over farming as a means of livelihood. Beside, there is a
positive correlation between the level of education and participation in non-farm
employment.
Thus, analysis of household and workers' participation in rural non-farm
occupations and their educational attainments confirms the findings of many other
research studies (Huffman, 1980; Singh, and Prabhakar, 2000) that education enables
a worker to make better choices over livelihood options available to him and also that
with the spread of education more and more of rural workers are joining non-farm
vocations as compared to agriculture. Further, the probability of participation in RNF
is lower if a worker is illiterate; he/she is more likely to be in farming. Of all the
uneducated workers, only about 35 per cent were in RNF (Table 7.3). Nevertheless,
the illiterate and lowly educated workers settle at low-productivity low-earning nonfarm activities.
It has been seen earlier that 58.55 per cent of rural workers in Zone I was in
RNF. Participation of workers in RNF activities was low at 39.78 per cent in zone III
(Table 7.4). Table 7.4 shows that only 27.61 per cent of rural workers in Zone I
(developed zone) are illiterate as compared to 43.16 per cent in Zone III (less
developed zone). Secondly, similar level of education in Zone III as in Zone I is
producing lower shares of RNF workers in the former zone. For example, share of
workers who have education up to primary is 11.93 per cent of the total in Zone III
and 10.57 per cent in Zone I. But in Zone III share of these workers in RNF activities
is only 42.95 per cent as compared to 55.72 per cent in Zone I.

147
Table 7.3: Educational attainment and distribution of workers in agricultural and nonagricultural occupations in study villages
Educational slab
Percentage
RNF
Agri. workers
Total
share of each
workers
in total*
Illiterate
1725
3234
4659
31.55
(37.03)
(69.41)
(100.00)
Up to primary
847
785
1632
11.05
(51.90)
(48.10)
(100.00)
Beyond primary but up
3554
2521
6075
41.14
to high
(58.50)
(41.50)
(100.00)
Beyond high but up to
935
435
1370
9.28
Senior Secondary
(68.25)
(31.75)
(100.00)
Graduate and above
488
122
610
4.13
(80.00)
(20.00)
(100.00)
Others (skilled)
111
11
122
0.83
(90.98)
(9.02)
(100.00)
7660
7108
14768
Total
100.00
(51.87)
(48.13)
(100.00)
Notes: 1. Figures in brackets indicate percentage share.
2. * indicates column wise shares in total
Source: Field survey.

Table 7.4: Educational attainment and RNF workers in different zones


Education slab
RNF workers
Percentage of total workers in
each education slab
Zone I
Zone II
Zone III
Zone III
Zone I Zone II
Illiterate
Up to primary
Beyond primary
but up to high
Beyond high but up
to Senior
Secondary
Graduate and above
Others (skilled)
Total

771
(48.95)
336
(55.72)
1547
(59.50)
435
(69.27)

419
(35.60)
249
(59.43)
1148
(65.75)
271
(73.24)

535
(24.24)
262
(42.95)
859
(49.68)
229
(61.56)

23.08

18.33

26.30

10.06

10.89

12.88

46.32

50.22

42.23

13.02

11.85

11.26

215
(82.38)
36
(94.74)
3340
(58.55)

167
(83.08)
32
(86.49)
2286
(57.87)

106
(71.62)
43
(91.49)
2034
(39.78)

6.44

7.31

5.21

1.08

1.40

2.11

100.00

100.00

100.00

Note: Figures in brackets denote percentage share out of RNF and agricultural households together in
respective categories.
Source: Field Survey.

Educational facilities in all their forms from physical infrastructure to absenteeism


of teachers - are far less developed in these villages than in villages of developed blocks (and
of moderately developed blocks) of which Zones I and II are, respectively, constituted. Thus,

148
it is not only the level of education which is important, equally important, for participation in
rural non-farm activities is the quality of education. The latter assertion is proved from the fact
that though in Zone I, percentage of workers with technical education (0.67 per cent) is less
than that in Zone II (0.94 per cent), 94.74 per cent of these workers are in RNF activities in
Zone I as compared to only 86.49 per cent in Zone II.
7.2.2 Land holding size and RNF activities
Land holding size, too, has an influence on participation of workers and households
in rural non-farm pursuits. Table 7.5 reveals that more than three-fifth of all the rural
households (62.50 per cent) in the study villages are landless. Of all the landless
households, 70.63 per cent participate in RNF activities and only 29.37 per cent are pure
farming households. As the size of land holding increases, the percentage share of RNF
households declines and the decline is very neat. For the farm households' tendency is
otherwise. With the increase in land holding size, the share of pure farm households
increases. The data, thus, confirms the fact of an inverse relationship between land holding
size and household participation in RNF activities.
Table 7.5: Land holdings size, RNF and agricultural households in study villages
Land holding size
Type of household
class
Percentage share of
RNF
Farming
Total
total households*
Landless
3585
1491
5076
62.53
(70.63)
(29.37)
(100.00)
Marginal
491
395
886
10.91
(55.42)
(44.58)
(100.00)
Small
271
397
668
8.23
(40.57)
(59.43)
(100.00)
Semi-Medium
245
550
795
9.79
(30.82)
(69.18)
(100.00)
Medium
160
419
579
7.13
(27.63)
(72.37)
(100.00)
Large
29
85
114
1.40
(25.44)
(74.56)
(100.00)
Total
4781
3337
8118
100.00
(58.89)
(41.11)
(100.00)
Notes: 1. Figures in brackets indicate percentage shares.
2. *indicates column wise shares in total.
Source: Field survey.

By looking at the distribution of rural households in terms of land holding


sizes in three zones, one can easily find that percentage of RNF households in Zone I
is higher (65.61 per cent) as compared to Zone II (64.42 per cent) and Zone III (47.53

149
per cent). One of the reasons is the distribution of land holding in these zones. In Zone
I, about 65 per cent of the households are land less. In zone III, the percentage of
landless households is less (57.21 per cent). This is shown in Table 7.6.
Further, in Zone III, the share of medium and other large holders is about
10.73 per cent as compared to 8.76 per cent in Zone I and 5.32 per cent in Zone II.
Given the inverse relationship between land holding size and household participation
in RNF, the participation of households in RNF is bound to be less in zone III than
that in Zone I and Zone II. Secondly, per household availability of land is more in
Zone III as compared to Zone I and II. It has been held that in regions where per
household land availability is large, household participation in RNF is less (Reardon
et al., 2007).
Table 7.6: Land holdings and RNF households in different zones
Land
RNF households
Percentage share of total households in
holding size
each land holding size class
class
Zone I
Zone II
Zone III
Zone I
Zone II
Zone III
Landless
1600
1014
971
79.48
72.27
71.14
(80.56)
(70.07)
(59.10)
Marginal
132
183
176
6.56
13.04
12.89
(49.44)
(68.80)
(49.86)
Small
92
92
87
4.57
6.56
6.37
(37.55)
(57.5)
(33.08)
Semi102
76
67
5.07
5.42
4.91
Medium
(33.89)
(40.21)
Medium
74
31
55
3.68
2.21
4.03
(31.22)
(35.63)
(21.57)
Large
13
7
9
0.65
0.50
0.66
(40.63)
(24.14)
(16.98)
Total
2013
1403
1365
100.00
100.00
100.00
(65.61)
(64.42)
(47.53)
Note: Figures in brackets denote percentage share out of RNF and agricultural households together in
respective categories.
Source: Field survey.

At the household level, it has been found that size of land holding and
participation of rural households in non-farm occupations is inversely related. Now,
question arises how rural workers react to changes in land holding size classes. An
assessment of detain Table 7.7 shows that out of 14768 rural main workers, 7660
workers (51.87 per cent) were engaged in non-farm occupation. Further, 57.83 per
cent of all rural workers in all the study villages do not own any land. They are
landless. But of all these workers 68.07 per cent are in RNF employment and 31.93

150
per cent in agriculture and related activities. Thus more landless workers tend to join
non-farming than farming occupations. In fact, in the absence of any land, majority of
these workers have joined non-farm employment.
Table 7.7: Land holdings, non-farm and farm workers in study villages
Land
holding RNF workers
Agri. workers
Total workers
size class
Landless
Marginal
Small
Semi-Medium
Medium
Large
Total

5814
(68.07)
798
(49.44)
409
(30.66)
352
(21.42)
243
(18.13)
44
(14.86)
7660
(51.87)

2727
(31.93)
816
(50.56)
925
(69.34)
1291
(78.58)
1097
(81.87)
252
(85.14)
7108
(48.13)

8541
(100.00)
1614
(100.00)
1334
(100.00)
1643
(100.00)
1340
(100.00)
296
(100.00)
14768
(100.00)

Percentage
share of total
workers*
57.83
10.93
9.03
11.13
9.07
2.00
100.00

Notes: 1. Figures in brackets indicate percentage shares.


2. *indicates column wise shares in total.
Source: Field survey.

It is also clear from Table 7.7 that nearly one-tenth of rural workers (10.93 per
cent) belong to the land size class of marginal farmers. In this class, the share of RNF
(49.44 per cent) is slightly less than that of farm activities (50.56 per cent) in
employment. With increase in land holding size, the participation of rural workers in
non-farm operations further declines and that of farming increases. In fact the
percentage share of workers in RNF decreases from 68.07 per cent for landless class
to 14.86 per cent in the size class of 25 acres or more. Consequently, the percentage
share of farming in employment increases from 31.93 per cent for landless workers to
85.14 per cent for workers having 25 acres or more. Thus, the existence of inverse
relationship between land holding size and participation of family workers in nonfarm occupations is confirmed by data. About 18.13 per cent of medium farmers (with
land between 10-25 acres) and slightly less than 15 per cent of large farmers engage
in rural non-farm pursuits. Taking together, 11.13 per cent of rural workers belong to
the land size classes of 10 acres or more. Of these workers, 21.42 per cent work in
non-farm occupations and the remaining 78.58 per cent in farming. From this, it can

151
be inferred that slightly more than one-sixth of workers belonging to better-off
households engage in various non-farm activities. It has been emphasized that this
class of workers join RNF sector not out of any compulsion, but as a matter of a wellthought of strategy of accumulation (White, 1991; Saith, 1992).
With regard to different zones, the landless workers in RNF activities varied
from 54.89 per cent in Zone III to almost 80 per cent in Zone I as is evident from
Table 7.8. Among the land owners, the data show inverse relationship between land
holding size and the workers in RNF sector across the zones.
The higher percentage of RNF workers amongst large farmers in zone I may
be due to higher level of infrastructural development as compared to other zones. It is
clear from the foregoing discussion that the participation of land owning households
in RNFS and that of RNF workers from such households has an inverse relationship
with the size of land holding. A vast majority of landless households and workers
prefer RNF to farming.
Table 7.8: Land holdings and non-farm workers in different zones
Land
holding
RNF workers
Percentage share of total
size class
workers*
Zone I
Zone II
Zone III Zone I Zone II
Zone III
Landless
2733
1634
1447
81.83
71.48
71.14
(79.66)
(66.05)
(54.89)
Marginal
200
325
273
5.99
14.22
13.42
(43.67)
(65.79)
(41.24)
Small
120
163
126
3.59
7.13
6.19
(25.48)
(44.05)
(25.56)
Semi-Medium
144
114
94
4.31
4.99
4.62
(21.88)
(29.46)
(15.72)
Medium
121
41
81
3.62
1.79
3.98
(20.30)
(24.40)
(14.06)
Large
22
9
13
0.66
0.39
0.64
(24.18)
(15.79)
(8.78)
Total
3340
2286
2034
100.00 100.00
100.00
(58.55)
(57.87)
(39.78)
Note: 1. Figures in brackets denote percentage share out of RNF and agricultural households together
in respective categories.
2. * indicates percentage shares of each land holding size in total.
Source: Field survey.

From the above tabular analysis, it is clear that education level of the
household heads as well as of workers exert a positive influence on the participation
in non-farm activities. The size of land holding with the household, however, has a

152
negative relationship with the extent of participation. The study further explored these
relationships with the help of probit analysis.
7.2.3 Village level factors and RNF activities
It has been seen earlier that the availability of and access to land is one of the
important determinants of household participation in RNF activities. If one looks at
the village level engagement of workers in RNF and per capita land holding in the
village, it is found that in those villages where per capita land holding is more, the
extent of participation of workers in RNF is less than in the villages where per capita
land holding is less. In 10 villages out of 24 study villages, the extent of RNF workers
is 68 per cent or more. In all of these villages, per capita land holding is less than 2
acres. At the other extreme, there are four villages where per capita land holding is
between 4 to 5 acres. In 3 of these villages, extent of RNF workers is less than 25 per
cent.
(i) Proximity to towns/mandi towns
The study has also reviewed a number of studies which established that ruralurban linkages are perhaps as important as agriculture and non-agricultural linkages in
the development of rural non-farm employment. In 6 out of 24 study villages the
share of RNF workers in total workforce varied between 71 per cent and 91 per cent.
What is common in these villages? None of these villages, except Dhupsari, is close
to a large town of the state. Four villages i.e. Behnangal, Udassian, Chunni Kalan and
Purkhali, are located either on a state highway or on a main district road. These
villages are situated in such a way that villagers from surrounding 6 to 7 villages
come here for taking a bus to nearby towns. These villages are rather located on midway between two large towns linked by the road. All these villages are connected
with hinterland by a network of all-weather metalled roads. These villages, over the
years and without any conscious effort by the state government, have evolved into
market-villages.
Across 13 villages having RNF share more than 60 per cent of workforce, the
study finds that four villages, namely, Dhupsari (77), Assal (69), Chakmirpur (69) and
Udassian (79) fall within 3 kilometers of an urban town or a rural town (mandi town).
Another set of four other villages, namely, Sidhwan Khurd (67), Randhawa (64),
Dhandour (68) and Gobindpura (71) fall within 8 kilometers of an urban town or a

153
mandi town. The mandi towns have grain markets, offices of the local market
committees, parking lots for trucks and taxies where the operators wait for their turn,
and so on. Round the clock electricity and clean drinking water promote many
manufacturing and repair activities. These small towns are host to many private
schools, hospitals/nursing homes, cinemas and other entertainment sources. During
the marketing seasons of wheat and paddy much of the interaction amongst farmers,
grain cleaners and loaders, local commission agents, workers of procurement agencies
and of transporters takes place here. These towns, especially during the marketing
seasons become very vibrant and attracting manual labour from the hinter land
villages. Taxi stands, barber shops, liquor vends and many other service providers
jostle for space near the mandis. Thus, 8 out of 13 villages having RNF workers more
than 60 per cent can be said to have urban influence. Remaining 5 villages, namely,
Chunni Kalan, Purkhali, Behnangal, Simbro (64) and Ammonangal (68) all fall on the
main roads and are located favourably. One of the factors common in all these 5
villages was their ability to cater to a cluster of villages falling in the hinter land.
At the other extreme are four villages having RNF share between 20 per cent
and 30 per cent. These villages are 10 or more kilometers away from the nearest town.
Surprisingly, the villages which are more than 15 kilometers from the nearest town
show larger participation of workers in RNF than those villages which are located at a
distance of 10 to 15 kilometers. From the above discussion, it can be safely concluded
that closeness to an urban township or a mandi town along with the location of the
village explain much of the variations in the shares of RNF workers in the village
workforce. However, some exceptions are always there.
(ii) Infrastructure facilities
Some of the villages showing high shares of RNF in their total workforce have
all the infrastructural facilities. For example, four villages, namely, Chunni Kalan,
Beh Nangal, Purkhali and Dhadour have regulated purchase centres for marketing
agricultural produce, focal points, telephone exchanges, clean drinking water supply,
primary health and veterinary centres. But there is no one-to-one relationship. Some
of the villages having small shares of RNF like Badalgarh and Chakkar also have
purchase centres. Further more, some of the villages which are closer to urban towns,

154
though have high shares of RNF, but do not have these facilities located in the
villages; they use the facilities located in the towns.
In rural Punjab of today, none of the villages is without all weather metalled
link road. All the houses, even in remotest of the villages have electric power.
However, the quality of service delivery deteriorates as one moves away from villages
on the main roads to ones in the remote locations. In the latter set of villages,
absenteeism amongst teachers, pharmacists, doctors, operators of water supply,
electricity supply etc. is much more.
Table 7.9: Impact of village level factors on RNF employment in village
Profile

RNF workers
As percentage of total
As percentage of population
workers
Mean
S.D.
Mean
S.D.
Connectivity with national/state highway
Connected (5)
20.75
2.38
65.20
7.24
Not Connected (19)
17.46
6.58
54.76
10.54
t-Value
1.08
2.07**
Connectivity with railway station
Connected (5)
20.95
3.19
68.15
12.40
Not Connected (19)
17.41
6.48
53.98
13.35
t-Value
1.17
2.14**
Technical education institutions (TEI)
Villages with TEI (6)
21.76
4.09
69.24
12.86
Villages without TEI (18)
16.95
6.22
52.83
14.10
t-value
1.76*
2.52**
Distance from nearby city/town
Up to 5 km (4)
22.07
2.32
73.68
5.61
5-10 km (10)
17.98
4.59
55.76
11.45
10-15 km (4)
14.07
9.28
41.92
16.82
>15 km (6)
18.54
7.11
57.72
12.73
r-value
-0.170
-0.411*#
-0.200
-0.472**#
Operational area
Up to 500 acres (9)
22.69
2.28
74.10
8.53
500-1000 acres (8)
18.76
5.61
57.78
16.40
> 1000 acres (7)
11.61
4.00
33.88
11.34
r-value
-0.594***
-0.659***
Population
Up to 1000 (2)
20.97
0.22
65.56
2.45
1001-2000 (17)
19.00
6.47
59.76
13.64
Above 2000 (5)
14.11
4.05
43.85
11.60
r-value
-0.320
-0.360
Notes: 1 ***, ** and * denote significance at 1%, 5% and 10%, respectively. # denotes r-values from 1
km to 15 km.
2 Figures in parentheses indicate number of villages.

155
The study has also statistically tested the significance of influence of various

village level factors on RNF employment in the village. RNF employment was taken
in two forms; (i) RNF workers as percentage of total population of the village; and (ii)
RNF workers as percentage of total workers in the village. To compare the average
level of RNF employment between two categories, student's t-test was applied to see
the relationship between two parametric variables i.e. between RNF employment and
distance from the nearest city/town; RNF employment and operational area; and RNF
employment and population size of the village, coefficient of correlation (r-value) was
worked out. The results so obtained are presented in Table 7.9.
(iii) Connectivity with national/state highway
The proportion of RNF workers out of total workers was significantly higher
in connected villages (65.20 per cent) as compared to that in other villages (54.76 per
cent). This shows that village connectivity to a state/national highway has a
significant impact on village level RNF employment. It is also clear from the
statistical significance (Table 7.9)
(iv) Connectivity with railway station
The proportion of RNF workers in total workers in connected villages is
significantly higher than in those villages which are not connected with a railway
station. This indicates that connectivity to railway station also exerts significantly
positive effect on RNF employment.
(v) Presence of technical education institutions
It is clear from Table 7.9 that the RNF workers as proportion of total workers
was 69.24 per cent in the villages which enjoyed the facility of technical education
institutions, while this proportion was 52.83 per cent in other villages. This difference
of proportion of RNF workers' in total workers between the two categories of villages
was significant at 5 per cent level.
(vi) Distance of villages from nearby city/town
The RNF workers as a proportion of the total workers were the highest in the
villages which fall within 5 kms from the nearest town. This proportion declines with
the increase in distance from 5 to 15 kms. After that, it again shows an increase in

156
proportion. Though no definite trends are discernible, yet the results support a study
by Fafchamps and Shilpi (2003) which had showed that rural non-farm wage
employment falls away quickly as one leaves peri-urban areas for the hinterland. But
there was a U-shaped pattern for self employment because some rural non-farm
activities in the hinterland had to come up to support local needs not met by supply
from urban areas. The results also support the contention that the distantly located
regions/villages are insulated from the onslaught of cheap supplies from urban based
industrial units.
(vii)

Size of operational land area

To see the effect of operational area of the villages on RNF employment, the
area was grouped into three categories of up to 500 acres, 500-1000 acres and above
1000 acres. The proportion of RNF workers to total workers declined significantly
from 74.10 per cent in village having up to 500 acres of operational land to 57.78 per
cent with 500 to 1000 acres and further to 33.88 per cent in the village having more
than 1000 acres of operational land. The decline was significant as indicated by the
coefficient of correlation (- 0.659***) significant at one per cent level of probability.
It is significant to note that there is a negative correlation between the land mass
available for cultivation in the village and the proportion of workers in RNF
employment. As the operational area increases the proportion of RNF workers goes
on declining.
(viii) Population size of village
The RNF workers as a proportion of total workers in the village showed a
declining trend with the increase in population of the villages under study. The
proportionate share of RNF workers in total workers was 65.56 per cent in villages
having population up to 1000. It decreased to 59.76 per cent and further to 43.65 per
cent with the increase in village population from up to 1000, to 1001-2000, and above
2000 respectively. In this case, coefficient of correlation estimated at -0.360, is
significant at 10 per cent level. This indicates that proportion of RNF workers out of
total workers was inversely related to the population size of the study villages. This
result is contrary to many studies which claim that the amount of non-farm activity
tends to vary directly and positively with the size of rural settlements.

157
7.2.4 Linkages of RNF sector with other sectors
(i) Linkages with agriculture
A number of studies reviewed have shown that it is mainly the agriculture
sectors' production and consumption linkages with the local rural economy which
cause the emergence and growth of rural non-farm sector. One way to look at the
issue is to examine the nature of RNF activities. In a sample of 300 rural non-farm
households, 65 households were self-employed households, 198 were wage labour
households and another 37 were mixed households. In all 452 workers were found to
be engaged in 86 non-farm activities. Of these, 49 were in self employment and the
rest were wage labour activities.
Of the 49 self employment activities, 13 were closely related with agriculture. Thus,
almost 27 per cent of self employment activities had direct linkages with the agriculture. Out
of these, 9 were found to have forward production linkage with agriculture. These are flour
mill, fruit shop, juice shop, quilt filling, vegetable vending, livestock trading, sweets making,
ration depot and commission agency. All these activities use agricultural produce as input for
transformation into new products or value addition of the same product. Four activities were
found to have backward linkage with agriculture. These are diesel engine repair, land
leveling, combine harvesting and wheat reaper trading. Here rural non-farm activities provide
valuable inputs to agricultural production processes. Thus, out of all the agricultural related
self-employed activities, 69 per cent were found to have grown because of RNF sectors'
forward linkage with agriculture and 31 per cent were the result of backward linkages.
New agricultural technologies by increasing the agricultural productivities not only
release the labour force into the non-farm markets but increased agricultural production also
provides funds for redeployment into non-farm pursuits. From this investment angle, sources
of funding of RNF self employment activities were examined. It was found that as much as
36.71 per cent of start-up capital came from surplus generated in agriculture. Table 7.10 gives
an idea how different RNF activities used the agricultural surplus to start-up their ventures.
Looking at the zonal level view, the study found that it was the most
developed zone I where the contribution of agricultural surplus to start-up capital was
the highest (Table 7.11). In certain lines, e.g., mining and quarrying, manufacturing
related activities, other services, and trade and commerce, the share of agricultural
savings as a part of total initial outlay is more than 40 per cent. RNF sector in the

158
study villages, thus, has become a vehicle to mop up unutilized idle savings and their
re-deployment in productive activities of different kinds.
Table 7.10: Nature of activities and amount of savings generated by agriculture for
setting up different RNF enterprises (per unit)
Activities

No. of units

Amount which came from


Total start-up
agricultural savings
Capital
Mining and quarrying
1
1000,000
2400,000
(41.67)
(100.00)
Manufacturing, processing,
15
25,143
60,084
services and repairs
(41.85)
(100.00)
Construction
7
13,000
44,686
(29.09)
(100.00)
Trade and commerce
33
133,837
310,336
(43.13)
(100.00)
Transport, storage and
21
106,381
417,786
communications
(25.46)
(100.00)
Other services
26
78,926
195,160
(40.44)
(100.00)
Total
103
98,746
268,959
(36.71)
(100.00)
Notes: 1. Besides, agricultural savings, other sources of funding included funds raised from
relatives/friends, loans from institutional sources and loans from non-institutional sources
2. Figures in brackets denote percentage of the total.
3. Amount per unit
Source: Field survey.

Table 7.11: Percentage of funds generated by agriculture to start up different


RNF activities in different zones
No. of
units

Amount
which came
from
agricultural
savings

Mining and
quarrying
Manufacturing,
processing,
services and
repairs

1000,000
(41.67)

2400000
(100.00)

40,000
(51.02)

78,400
(100.00)

Construction

Trade and
commerce
Transport,
storage and
communications

10

29,000
(96.67)
168,333
(74.85)

30,000
(100.00)
224,892
(100.00)

205,200
(29.31)

700,000
(100.00)

Other services

Total

35

253,333
(63.47)
188,613
(46.25)

399,167
(100.00)
407,814
(100.00)

Nature of
activity

Total
start-up
capital

No. of
units

Zone I

Amount
which
came
from
agricultu
ral
savings

Total
start-up
capital

No. of
units

Zone II

Note: Figures in brackets indicate percentage shares.


Source: Field survey.

Amount
which
came
from
agricultu
ral
savings

Total
start-up
capital

Zone III

0
(0.00)

0
(100.00)

0
(0.00)

0
(100.00)

0
(0.00)

0
(100.00)

17143
(31.83)

53,857
(100.00)

1000
(13.16)
211,000
(34.52)

7600
(100.00)
611,280
(100.00)

15,000
(13.04)
47,944
(33.20)

115,000
(100.00)
144,400
(100.00)

58,333
(28.08)

207,750
(100.00)

6200
(3.03)

204,500
(100.00)

13,750
(27.50)
89,198
(33.02)

50,000
(100.00)
270,147
(100.00)

15,500
(8.47)
25,195
(17.56)

183,000
(100.00)
143,463
(100.00)

3
10
6
9
29

2
13
6
10
39

159
(ii) Rural/Urban linkages
(a) Employment linkages
Saith (1992) argued that location approach alone to rural non-farm sector does
not capture it entirely. Enterprises can be located in urban or peri-urban areas. The
only condition is that they must have some linkages with the rural sector. To be able
to generate employment for rural labour is an important criteria which can indicate
whether an enterprise based in urban areas has linkages with rural hinter land or no.
Table 7.12 shows where the rural workers are engaged. Mining and Quarrying
is an activity which is located in rural areas. Employment is completely generated in
rural areas, but the activity has urban links as well because its output is not entirely
used in rural areas. From the employment angle, of all the RNF workers in
manufacturing related activities, 80.51 per cent were working in units located within
the villages. Similarly, in other services related activities, as high as 96.97 per cent of
workers found work in villages. Construction activities are such attracted 71.43 per
cent of workers employed in these have to travel to urban areas for getting
employment.
Table 7.12: Distribution of RNF self-employed workers in different activities in
rural and urban areas
Nature of activity
Rural
Urban
Rural + Urban
Total
Mining and quarrying
1
0
0
1
(100.00)
(0.00)
(0.00)
(100.00)
Manufacturing, processing,
16
4
0
20
services and repairs
(80.00)
(20.00)
(0.00)
(100.00)
Construction
5
2
0
7
(71.43)
(28.57)
(0.00)
(100.00)
Trade and commerce
26
4
4
34
(76.47)
(11.76)
(11.76)
(100.00)
Transport, storage and
15
8
0
23
communications
(65.22)
(34.78)
(0.00)
(100.00)
Other services
32
1
0
33
(96.97)
(3.03)
(0.00)
(100.00)
All
95
19
4
118
(80.51)
(16.10)
(3.39)
(100.00)
Note: Figures in brackets indicate percentage of workers in total workers.
Source: Field survey.

From the employment angle, one can say that 15.53 per cent of rural workers
engaged in self employment activities find non-farm work in urban areas (Table 7.13).

160
2.91 per cent are located in rural and urban area. The remaining 81.55 per cent are
located in rural areas.
In no way, thus, rural areas can be considered as completely bounded spheres.
Rather, rural non-farm activity grows in a continuum of rural-urban space. It calls for
infrastructural development of roads, transport, communication etc.
Table 7.13: Location of RNF self-employed enterprises/activities
Nature of activity
Rural
Urban
Rural + Urban
Mining and quarrying
1
0
0
(100.00)
(0.00)
(0.00)
Manufacturing, processing,
11
4
0
services and repairs
(73.33)
(26.67)
(0.00)
Construction
5
2
0
(71.43)
(28.57)
(0.00)
Trade and commerce
26
4
3
(78.79)
(12.12)
(9.09)
Transport, storage and
15
6
0
communications
(71.43)
(28.57)
(0.00)
Other services
26
0
0
(100.00)
(0.00)
(0.00)
Total
84
16
3
(81.55)
(15.53)
(2.91)

Total
1
(100.00)
15
(100.00)
7
(100.00)
33
(100.00)
21
(100.00)
26
(100.00)
103
(100.00)

Note: Figures in brackets indicate percentage shares.


Source: Field survey.

(b)

Market linkages

In this context, the study calculated from where the self-employed ventures
bought these inputs and where did they sell their end-products. The study makes the
result interesting (Table 7.14). For example, a mining and quarrying unit, after paying
a certain annual license fee, mines sand and gravel in rural areas and end-users are in
urban as well as rural areas. The study found that 70 per cent of sand and gravel was
extracted for use in rural areas. It speaks about the construction boom in rural areas
both by private as well as government development agencies. In trade and commerce,
there are 33 units. In terms of purchase of inputs, 7 units purchase their raw material,
etc. from purely rural areas and remaining 26 were found to depend on urban areas for
the purchase of inputs. For selling the products, 25 units sold their output in rural
areas and the 8 units sold it in urban locations. Thus, as revealed by the data in Table
7.14, rural self employment enterprises have both rural-urban linkages in terms of the
purchase of inputs as well as sale of their products and services.

161
Table 7.14: Input output destinations of RNF self-employed enterprises
Nature of activity
Purchasing from (%)
Selling to (%)
Rural
Urban
Rural
Mining and quarrying
1
0
1
(100.00)
(0.00)
(70.00)
Manufacturing, processing,
3
12
13
services and repairs
(28.57)
(71.43)
(86.67)
Construction
0
7
7
(0.00)
(100.00)
(100.00)
Trade and commerce
7
26
25
(23.33)
(76.67)
(75.76)
Transport, storage and
2
19
13
communications
(9.09)
(90.91)
(61.90)
Other services
10
16
24
(37.50)
(62.50)
(92.31)
Total
23
80
82
(25.88)
(74.12)
(79.61)

Urban
0
(30.00)
2
(13.33)
0
(0.00)
8
(24.24)
8
(38.10)
2
(7.69)
21
(20.39)

Note: Figures in brackets indicate percentage shares.


Source: Field survey.

7.2.5 Probit analysis predicting participation of RNF employment


In order to predict the probabilities of households participation in rural nonfarm activities in sampled villages, Probit analysis was done with the following
function:
Involvement of households in RNF activities = f (Male members in working
age, family size, operational area, leased-in land, leased-out land)
The results of the analysis are presented in Table 7.15. The results are tested
by value of Chi-Square at 1 per cent level of significance. This indicates that the
variables taken for the Probit model are significantly associated with the dependent
variable. It called for the goodness of the data to proceed for the analysis.
Table 7.15: Probit model for predicting participation in RNF employment in Punjab
Parameter
Estimate
t-value
Significance level
PROBITa
Male
0.035
3.353
0.01
Family Size
-0.027
-0.504
NS
Operational Area -0.182
-2.414
0.05
Leased in
0.027
0.272
NS
Leased Out
0.006
3.026
0.01
Intercept
-1.154
-3.611
0.01
a. PROBIT Model: PROBIT(p) = Intercept + bX
b. NS stands for Non Significant
Chi-Square Tests
Chi-Square
Pearson Goodness-of-Fit Test
18.223
Significance
0.01

162

On the basis of intercept term and different Probit coefficients in Probit


model, classification of households into RNF and agriculture sector was tested for
accuracy. The model predicted that 283 households involved in RNF activities
(against the actual number of 300 households). Thus 83.33 per cent outcomes were
correctly classified, while the remaining 16.67 per cent of the RNF households were
misclassified. Among agricultural households, 68.00 per cent were correctly classified
and the remaining 32.00 per cent were misclassified. The classification suggested that
there is a probable shift of 16.67 per cent households from RNF to agriculture and
32.00 per cent from agriculture to RNF. Hence, there is a probability of net shift of
15.33 per cent households towards RNF activities (Table 7.16).
Table 7.16: Number of households in farm and non-farm activities as per Probit Model
and Actual number in respective activities
Type of Household
Predicted
Actual
Probable shift
Yes
No.
Total
RNF
251
49
300
From Agri. to RNF = 32.00 per cent
(83.33) (16.33)
Farm
32
68
From RNF to Agri. = 16.67 per cent
100
(32.00) (68.00)
Net Shift to RNF = 15.33 per cent
Total
283
117
400
(70.75) (29.25)
Note: Figures in brackets indicate percentage shares.

The Chi-square test rejected the hypothesis that the pattern of prediction was
at random at well beyond 1 per cent level of significance. The significant calculated
value of chi-square implied that the predictive pattern was not random, but based on
the significant interaction of explanatory variables with the dependent variable.
The probit coefficients indicate that the increased number of male members in the
working age group increased the probability of involvement of rural households in the
non-farm activities. Similar was the role of leased-out agricultural land. The
operational area and leased-in area exerted inverse impact on probability of
engagement of rural households in non-farm activities. In the previously tabular
analysis, we had observed that larger the land holding which a household has smaller
was its participation in non-farm activities The Probit analysis also supports it.
Leasing-in land is an addition to the operational holding; hence its relationship also
came to be negative with the involvement of rural households in non-farm activities.
Leasing out land pushed the household towards non-farm activities, as with the

163
leasing out of the land the size of the operational holding with the household declines.
In a study from Punjab (Singh et al., 2007), it was seen that a large majority of
farmers who left farming and joined other activities in rural areas have leased out their
land. Out of the total 2 lakh farmers who left farming during 1991 to 2007, about 64
per cent farmers had leased out their land, while the remaining farmers had sold off
their land, totally or partially. This shows that farmers in Punjab generally participate
in RNF activities after leasing out their land.
The foregoing discussion has amply displayed that social and physical
infrastructures are the significant determinants for the development of RNF sector.
The participation of households in RNF activities, however, depends on the level of
education and skill. The land holding size also significantly impacts participation of
households in non-farm activities.

You might also like