Version classiqueVersion mobile

Essays on population and space in India

Christophe Z. Guilmoto
Alain Vaguet

Part II. Rural and urban poles

3. Diversification of Economic Activities in Rural India : Some Contrasting Trends

Hélène Guetat-Bernard

Texte intégral

1The new policy of economic liberalisation pursued by the Indian government since the beginning of the 1990s is the continuation of an economic reorientation that started ten years earlier. The concrete application of new measures at the local district level will however take long, and depend on political decisions taken by the various states. Improvements in some economic indicators over recent years do not mean that the economic difficulties of the preceding period have disappeared, while new uncertainties have made an appearance, particularly for small farmers. Reductions in subsidies on agricultural supplies and increasing prices of food and household necessities are affecting families of very modest means, while the division of inheritances and large dowry payments (connected with a still strong demographic growth, along with social pressures) continue to break up land-holdings (Landy, 1995). Already, “one quarter of the population engaged in agriculture, accounting for almost two thirds of the total working population, consists of landless peasants, condemned to work for wages or the hazards of tenant-farming.” As for the farmers officially classified as “marginal” —that is, those owning less than a hectare of land— they are far from marginal in demographic terms. The new measures intended to remedy imbalances and blockages in the national economy offer little hope to the small peasants. Increased prices of agricultural supplies will make it impossible for them to take advantage of improved methods of cultivation and soil-productivity. At the same time, the capacity of the agricultural sector to absorb an increasing rural labour-force is proving limited, since production methods connected with the green revolution are not providing enough jobs for all. Under these circumstances, imbalances affecting the future of the countryside, and questions about how to resolve them, remain persistent. For example, to what extent can accelerated diversification of non-agricultural economic activities help to retain in the countryside people of precarious economic resources, who have so far always managed to find ways to avoid a permanent move to the cities ? How can such diversification, taking advantage of local potentialities, contribute to rural development ?

The statistics are not easy to interpret

2No figures on occupational structures according to the residing place are available from before the Census of 1951, but dependence on the agricultural sector was more than 80 per cent in rural areas then, as was still the case in 1981 and 1991.

  • 1 The main sources of information about rural industries are the decennial population censuses, the r (...)

3Statistics do however indicate a significant trend towards diversification of rural occupations in the Indian countryside. The decennial censuses and national statistical surveys have registered a perceptible shift in the occupational structure of the rural labour-force towards non-agricultural activities since the early 1970s. For rural India as a whole, the censuses, but even more so the N.S.S.1 show that the rate of increase of the working population engaged in agricultural activities is less than that of the working population engaged in non-agricultural activities (Mahendra Dev, 1995). However, these figures need to be analysed with a lot of caution and they have been the topic of strongly contrasting commentaries.

  • 2 According to the census, the rural non-agricultural sector engaged 43.7 million individuals in 1991 (...)
  • 3 Strong fluctuations in the working population shown by successive surveys are evidence of under reg (...)

4On one hand, at the national level, only about one fifth of the rural working population is engaged in non-agricultural activity as a main occupation.2 This represents something of a failure in the development policies followed since Independence (whether as a result of attempts to decentralize or shift urban industrial activities to rural areas, of the “green revolution”, or of the integrated rural development project), for the proportion of the labour force engaged in non-agricultural activities is generally considered as an indicator of economic growth. Compared with other Asian countries, where one third to one half of rural workers are thus engaged, India ranks lowest according to this indicator. Nevertheless, it is obvious that international comparisons are not so simple, because of variations in the criteria utilized in census-taking, especially the distinction between main and secondary activities.3 At the national level in India as in other countries, marked contrasts exist between the various federal states, and even more so between different regions within the states, and at an even lower level between different rural areas and villages (Unni, 1991). On the other hand, the rapid increase in non-agricultural occupations in rural areas is often underestimated, because it is difficult to take into account both female labour and more generally all the complementary sources of income that contribute to the “total income” of individuals and rural families. Therefore, such proportions give a poor picture of the real scale of non-agricultural rural activities, since few families are dependent solely on agriculture for their total income. For example, most of the seasonal fluctuations in rural non-agricultural activities can be explained by shifts in occupational structure by temporary workers who move from an agricultural activity to a non-agricultural one and vice-versa.

5Lack of recognition of this sector is due to the inherent difficulties of an exhaustive census. In addition, the proportion of non-agricultural employment in rural areas would have shown a greater increase if certain areas previously classified as rural had not been reclassified as urban in the latest counts.

  • 4 A few figures allow us to see this issue more clearly : according to the Report of the Economic Cen (...)

6This proportion is also often difficult to estimate because it goes along with an increase in the proportion of paid jobs, often temporary and short-term, in undertakings classified as “informal” ;4 over the last 20 years, almost three quarters of the increase in non-agricultural rural employment relates to the informal sector (Vaidyanathan, 1994). Moreover, the share of the informal sector in the total number of manufacturing occupations has increased from 73 per cent to 82 per cent over the 1970s and 80s, with the greatest growth taking place in rural areas (Basant, 1994). Whether they belong to the informal sector or not, rural industries ought to attract the attention of planners, since their annual growth rate of 5.4 per cent (according to the economic census report of 1990) indicates an intense dynamism.

  • 5 Such jobs, which represented more than a quarter of the non-agricultural rural occupations for male (...)

7At the same time, the proportion of self-employed individuals is decreasing, particularly in the area of household manufactures, which have always constituted an important source of employment. The big drop since the 1960s is more marked amongst women than men, in both absolute and relative terms.5 On the other hand, the modern sector of household industries shows a tendency to rise.

8In this connection, the example of silk production is revealing (Guetat-Bernard, 1995) : from the first steps in transforming the silk threads, up to the weaving, this activity has traditionally created a lot of employment. The weaving sector in India has witnessed profound waves of structural change, punctuating the industrial history of the sub-continent. Here let us concentrate on the present revival being carried out by the dynamism of small entrepreneurs in the informal sector, who by modernizing their production techniques are creating a strong competition for bigger businesses. Amongst the small hand weavers, who are self-employed or work for a master weaver, only those who specialize in extremely sophisticated production techniques can survive, such as the silk sari weavers, whose skill has not yet been matched by any machine. The rest live in extreme poverty and will gradually be obliged to give up the traditional occupation of their caste.

Strongly contrasting local trends, and weak intersectoral linkages

9This change in the structure of rural employment in India is providing food for rich discussions and controversies, focusing especially on the existence of intersectoral linkages between agriculture and industry. John Mellor, whose thesis is still giving rise to discussion in India, presented a virtuous circle of growth whose internal dynamism results from an increasing demand for local products and services, sustained by the multiplier effect of incomes of farmers enriched by the green revolution (Mellor, 1976). In this sense, the effects generated by consumer demand would outweigh those created by increased agricultural production. This is said to be partially confirmed by the fact that, on the whole, the existence of an exploitable raw material in the locality is rarely the explanation for the location of a rural industry : more important factors are local tradition, skills and established networks. According to this model, part of the savings of wealthy farmers would moreover be invested, thus responding to the new demand.

  • 6 Harriss (1991). An examination of the account-books of small shops, showing that very low incomes a (...)

10In other words, is the present trend of diversification in rural occupations explained by the existence of growth-fostering linkages between the agricultural and non-agricultural sectors, to dynamic and productive activities (as J. Mellor suggests) ? Alternatively, does it rather reflect a changeover towards a “residual” non-agricultural sector, with lower productivity than the agricultural sector, whose growth is due to the unavailability of sufficient agricultural employment ? The outcome of this debate is important : we must evaluate whether the relative expansion of non-agricultural occupations reflects a positive development, in the sense that the jobs created are connected with a rising level of consumption and productivity, or on the contrary, if this expansion results from a survival solution for impoverished rural families who are trying to keep up their income level by engaging some of their members in ill-paid and unproductive non-agricultural activities.6

11Various studies on this topic in many different countries, particularly the important ones undertaken by the World Bank in the 1980s, give no clear answer : some could be interpreted positively, others negatively (Basu and Kashyap, 1992).

12A study relating to two rural systems, one based on irrigated agriculture, the other on rain-fed agriculture, in the south of Karnataka state in south India, has made it possible to modify this hypothesis of an almost automatic relation between agricultural growth and growth in the secondary and tertiary sectors (Guetat-Bernard, 1994). Above all it showed the complexity of the phenomenon, and suggested a typology of non-agricultural rural activities. This field study, carried out in the countryside around Mysore and Mandya, showed especially that problems of analysis result from an intermingling of attractive and propulsive factors in the same place, according to the types of activity in question.

  • 7 However, this relationship is not free from ambiguity, insofar as on one hand, the unemployment rat (...)
  • 8 Durand-Dastès (1995).

13On the scale of the country as a whole and the various Indian states, the analysis of the statistics shows the existence of a relation between numbers of non-agricultural occupations and rural unemployment rates.7 Villagers’ participation in non-agricultural work is however inversely proportional to the size of the land-holdings owned by the households, on account of the social priority that is given to working the land. But another somewhat contradictory correlation emerged, between greater inequality in land distribution and lower incidence of non-agricultural employment. This suggestion had already been put forward in the work of R. Islam, when he wrote that the degree of rural inequality tends to limit positive effects of agricultural growth upon the non-agricultural economy. It also echoes, at the local level. India’s inescapable problem : the low level of effective demand inevitably limits economic growth, even though the strategy being followed at present relies on the effect of increasing demand from the middle and upper classes alone, estimated at one quarter of the total population, that is no less that 220 million inhabitants.8 Jeemol Unni confirms that regions with the highest proportion of people living below the poverty line have low rates of male workers engaged in non-agricultural activities. So the hypothesis of a “residual” non-agricultural sector does not seem to be invariably confirmed : conditions of extreme poverty do not necessarily foster growth of activities in this sector, if there is insufficient demand. However, the same author, contradicting the findings of other researchers, shows that a concentration of landholdings in a region has beneficial effects on the percentage of non-agricultural jobs for males, particularly on activities related to commerce and transport.

  • 9 Basant (1987) ; Harriss (1992).
  • 10 The situation in the Punjab is unusual, not because of the rate of non-agricultural rural jobs, but (...)

14This apparent contradiction in the correlation of all these variables shows that there is no proof of the positive impact of new agricultural technologies upon non-agricultural employment. In other words, regions where the methods of the green revolution have been implemented are not necessarily those where the non-agricultural rural sector is most dynamic.9 This is proved by the case of Punjab,10 a bastion of the green revolution, one of the few Indian states to have registered a drop in the proportion of the female and male labour force engaged in non agricultural activities between 1961 and 1981. It is supposed that increasing agricultural production during the first decades had a more positive effect upon agricultural employment than upon non-agricultural jobs (Unni, 1991 : 121). In this case, the green revolution did not foster decentralized production creating many jobs as predicted by J. Mellor.

  • 11 Agricultural income is taxed only by the state of Karnataka, while other types of income are taxed (...)

15In fact, transfer of agricultural surpluses to non-agricultural activities also depends on the potential new entrepreneurs being able to foresee and anticipate the market. For example in the irrigated farming district of Mandya, although there are too many simple sugar mills for local productive capacities, they still earn enough to seem worthwhile to wealthier peasants. Income from agricultural activities is moreover taxed at a much lower rate than income earned from professional, industrial or commercial activities.11

16In the rural system of rain-fed agriculture at Nagamangala, not far from Mandya, peasants would like to diversify their activities, but they seek to multiply their sources of income by safe investments in “traditional” agricultural production or industry, such as sugar or rice mills, or in commerce connected with agriculture, which involve the least immediate risk. Moreover, these investment sectors also offer a possibility of development, as is shown by the increasing numbers of modernized rice-mills.

Growth of small rural towns plays a minor role

17Neither do strong agricultural growth nor consequent increasing consumer demands automatically lead to an increased demand for the goods and services provided by local producers. The opening up of village economies and increased human mobility favour urban markets more. Thus, many “traditional” rural services and activities are disappearing, because they no longer suit the tastes of their customers, who are now more able to compare them with manufactured products offered by the towns. This applies to some artisans such as potters, jewellers or basket-weavers, and service trades such as washermen and barbers. On the other hand it does not apply to carpenters and tailors, whose numbers are growing in villages. Moreover, the resources drawn from agriculture are mostly turned towards financial and commercial investment, money lending and the purchase of fixed assets, whether locally or in towns.

  • 12 Guetat-Bernard (1994) ; and Harriss (1987).

18Besides, towns have not always acted as nodes of growth. Their productive activities often have little connection with local economies, whether from the standpoint of production-links above or below existing channels, the financial resources utilized, or the employment of skilled labour. Thus beyond the small towns, rural areas have become closely linked to larger metropolitan centres. Under these circumstances, the terms of exchange have not favoured the agricultural sector, and money-flows have drained away from the countryside.12 In Mandya, income drawn from the surrounding countryside, although it is a prosperous one, is rarely invested in industry, whether in the town or its rural surroundings. Mandya seems to be a market and administrative centre, with little industry apart from a government sugar factory. In general, in Mandya or Mysore —only 50 kilometres away from each other— urban businesses have little effect on linkage as a whole or on sub-contracting links with other small or medium ones. At Mandya, there are only two industrial areas, which are far from being fully occupied. And yet the new demand from rural households for consumer goods such as soft drinks, plastic products, sweets, etc. is gradually being partly met by products from recently established small industries or home workshops in villages, rural industrial zones, or at Mandya itself. The first signs of this development have become visible only since the early 1990s.

19Certainly the greater or lesser proximity of an urban market may stimulate an increase in local supplies of goods and services, or conversely ensure a greater degree of investment of urban capital in the countryside with the idea of exploiting some local resource. This can be seen for example in the increasing numbers of chicken farms around the outskirts of towns, supplying meat or eggs, or the plantation of fruit orchards. It also happens when urban businessmen want to make use of the expertise of cheap rural labour (usually women and children under the supervision of the head of the household) in simple non-mechanized tasks such as the manufacture of matches, beedis or incense sticks. These activities, carried out in the home and often limited to specific job-pools, create a lot of low-productivity employment. Although not well paid, the incomes are appreciated by the family decision-makers : in the absence of other alternatives, they supplement income from agriculture. However, in terms of employment possibilities, adult males prefer to look for urban jobs, in the framework of shuttle migrations (not to mention more distant and longer-term migrations.) Within a radius of twenty-odd kilometres, depending on the urban activities, the size of the town and the transport available, and also according to the employment market in the village and existing networks (which implies differences between villages where agriculture is irrigated or rain-fed), the urban market exerts a very strong pull upon the rural work-force. As much as three quarters of the village workforce may be employed on building sites, for example ; and in any case, most commonly one quarter of the rural working population is employed in towns.


20Management of agricultural risk, the level of agricultural surplus resulting from intensified methods of cultivation, and the ease of capitalisation, as well as relationships between economic growth and agricultural dynamism, are all independent factors affecting the dynamism of rural industrialisation. Other equally important factors could also be mentioned, such as the developmental level of the human resources (especially level and quality of education), local infrastructures (especially those providing quick and easy access to urban markets for jobs, goods and services), and the nature of urban economic activities that are more or less favourable to sub-contracting connections with small and medium-sized rural enterprises. These positive elements are known, and more or less put forward according to the situations.

21On the other hand, one issue has not been examined in Indian studies of rural development. The people’s preference for country life, and even more for the village where their history lies, is well known (Racine 1997). Nevertheless, unlike some very recent work on rural areas in Europe, few studies have investigated the close connection between the value set on local identity and the potential for local development. How far can an attachment to a familiar environment and established social relationships become a growth factor ? In the future, we must not merely consider the results of official policies, but also investigate the real motivations of the people involved and interpret the connections between territory and development. We must also enquire into the nature of social bonds in such an unequal society, since these bonds may constitute a development resource.


1 The main sources of information about rural industries are the decennial population censuses, the regular surveys based on national samples (National Sample Surveys, N.S.S.), or economic surveys.

2 According to the census, the rural non-agricultural sector engaged 43.7 million individuals in 1991, which represents 19.7 per cent of the rural working population engaged as a main activity, according to Kanikar (1994). For comparison, these rates have increased, for men, from 1961 to 1981, from 16.3 per cent to 18.2 per cent. The apparent drop in the number of women engaged in the rural non-agricultural sector (from 5.7 million in 1961 to 5.6 million in 1981 according to the census, that is from 10.3 per cent in 1961 to 9.7 per cent in 1981) can be explained by a considerable underestimation, for in general, compared with male occupations, a greater proportion of women are engaged in non-agricultural rural activities.

3 Strong fluctuations in the working population shown by successive surveys are evidence of under registration (from one census to another) of unpaid family workers. and people with several occupations whose paid work is difficult to assess (such as for bonded labour). While the Censuses show large differences from one decade to another, the N.S.S. present more stable rates for the working population. It is generally accepted that the Census of 1961 gives a correct picture of these rates : its figures are comparable to those of the N.S.S., and its estimate of female workers is the most convincing. The Census of 1971, by introducing the concept of “main activity” for people who have worked more than 183 days per year, underestimates the size of the working population, especially in the case of women, since secondary activities are poorly represented. In the 1981 Census the concept of “marginal workers”— people who had worked no more than six months in the year preceding the census, was set against that of main activity. Despite this, figures from the 1981 Census are difficult to compare with those of 1961, since the estimates of male and female employment are lower. These difficulties of evaluation make comparisons of non-agricultural rural occupations, and their distribution according to occupational sectors at different times, a delicate task.

4 A few figures allow us to see this issue more clearly : according to the Report of the Economic Census of 1990, almost 13 million businesses exist in rural areas ; 83 per cent of these are non-agricultural. These employ an average of 2.2 persons and 80 per cent of them are active all year round. 71 per cent of them utilize manual energy alone (Kanikar, 1994).

5 Such jobs, which represented more than a quarter of the non-agricultural rural occupations for males in 1961, represent only an eighth, thirty years later ; for women, the proportions have dropped from more than half to a third.

6 Harriss (1991). An examination of the account-books of small shops, showing that very low incomes and earnings are very common amongst artisans and home-labourers, seems to support the second view sustained by Vaidyanathan.

7 However, this relationship is not free from ambiguity, insofar as on one hand, the unemployment rate recorded by the N.S.S. is significantly underestimated, and on the other, the highest levels of unemployment are found in the most dynamic areas. In fact, the rate is highest in areas where agriculture is prospering, partly because of labour migrations. In addition, these areas also show the highest proportions of households dependent on wage earners and temporary workers. Hence, it is easier for them to report their search for work (Unni 1990). Thus, the findings of Vaidyanathan are not in contradiction with the fact that there may be a positive correlation between numbers of non-agricultural jobs and a prospering agricultural sector.
The results of a regression analysis also enable Mahendra Dev (1990) to demonstrate the existence of a strong positive correlation between unemployment rate, soil productivity, and levels of non-agricultural occupations. The relationship between unemployment level and the rate of non-agricultural occupations is however not linear : in the Punjab, for example, a high proportion of non-agricultural occupations accompanies low unemployment, while in Kerala, both variables show high levels.

8 Durand-Dastès (1995).

9 Basant (1987) ; Harriss (1992).

10 The situation in the Punjab is unusual, not because of the rate of non-agricultural rural jobs, but because modern factories have been started in rural areas, as in the countryside around Coimbatore in Tamil Nadu.

11 Agricultural income is taxed only by the state of Karnataka, while other types of income are taxed by both central and state taxes.

12 Guetat-Bernard (1994) ; and Harriss (1987).

© Institut Français de Pondichéry, 2000

Licence OpenEdition Books

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search