Jobs Off the Farm: Wealth, Human Capital, and Social Group in Rural Eastern India

ABSTRACT

This study examines the effects of wealth, human capital, and social group on the occupational choice and income of farm households, to gain an understanding of the mechanism underlying the recently increasing income diversification in rural eastern India. The results show that wealth and human capital have a positive impact on the choice of high-return jobs, and low caste workers are more likely to work in casual, low-paying jobs, partly due to their dependence on personal networks in finding jobs. Notably, no clear evidence of association between social group and income is found.

Acknowledgements

We are grateful to the GRIPS Emerging State Project of the Japan Society for the Promotion of Science (JSPS KAKENHI Grant Number 25101002) for financial support. We would also like to thank Yoko Kijima of The National Graduate Institute for Policy Studies for her helpful and valuable comments.

The statistical code used in this study is available upon request; please email [email protected].

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental Material

Supplemental data for this article can be accessed here

Notes

1. The proportion of the population under the poverty line is 30.9 per cent in rural areas and 26.4 per cent in urban areas in 2011–2012 (Government of India, 2014).

2. ‘Off-farm’ refers to work other than self-employed farming and thus includes agricultural wage work; ‘non-farm’ refers to work outside the agricultural sector and thus excludes agricultural wage work.

3. The GDP shares of agriculture, industry, and service were 15 per cent, 19 per cent, and 66 per cent, respectively, in 2012 (GGDC, 2014).

4. Monthly income is wage for those engaged in wage work. In case of self-employment, it is the value of production minus paid-out cost per month. For example, the daily wage rate of daily wage earners is converted to monthly by multiplying it by 25, assuming that they work approximately 25 days a month. Therefore, monthly income is a proxy for monthly wage rate in the analysis.

5. The questions regarding income are asked on a recall basis for the past year; this is bound to involve measurement errors. However, the question asks the frequency of payment and the number of times actually paid as opposed to total income for the year; this could serve to reduce measurement error.

6. The eight categories are self-employed farming including household activities, agricultural wage work, local construction, local factory, local service and self-employment, migratory construction, migratory factory, and migratory service and self-employment.

7. The 68^th National Household Survey is the most recent (at the time of writing) nation-wide household survey on employment available to the public in India. However, the data does not provide information about migration and the means by which people found migration work. Furthermore, the survey was conducted in 2012 and the results are now too old to accurately describe current employment activities in rural areas.

8. For the history and evolution of the reservation policy in India, see Osborne ( 2001).

9. Unlike STs and SCs, OBCs are not given reservations in legislature and the quota is lower than their population share.

10. The four states are Bihar, Odisha, Eastern Uttar Pradesh, and West Bengal.

11. In the agricultural sector, Mano, Yamano, Suzuki, and Matsumoto (2011) found that workers in cut-flower farms in Ethiopia helped their friends and relatives obtain jobs at those cut-flower farms.

12. Occupations, classified into eight categories, are assigned depending on the occupation in which each individual earns the largest amount of income. The questions regarding off-farm jobs and income are listed in Table A1 of the Supplementary Materials.

13. The information about English skills is based on self-report, which equals one if the respondent reported having basic skill in English speaking, reading, and writing, and zero otherwise.

14. Since some workers receive their wages daily, we estimated the determinant of daily income with the same specification as monthly income and found no qualitative differences with the monthly income estimation results. The results are in Table A2 of the Supplementary Materials.

15. Since the definition of income is different for those in service and those in self-employment, a separate analysis was conducted, dividing the group into two categories, service and self-employment. The results were qualitatively the same as the reported results.

16. This methodology, developed by Lee (1983), is widely used in analysis of occupational choice and wage (for example, Kurosaki & Khan, 2006; Liu, Zhang, & Chong, 2004; Senaratna Sellamuttu, Aida, Kasahara, Sawada, & Wijerathna, 2014).

17. Considering the possible correlations between years of schooling and English skills and between land size and asset value, four separate analyses were conducted without including respective variables whose results are presented in Tables A6–Table A17 of the Supplementary Materials.

18. Qualitative results of the estimation using only male samples are consistent with results using all samples.

19. Land and asset distribution are largely unequal across social groups, as evident in Table 2. Therefore, additional analyses are conducted which include the interaction terms between social groups and land size, and social group and value of asset, whose results are reported in Tables A18–A20 of the Supplementary Materials. Both land size and asset value show differential impacts on the occupational choice by social groups. The impact of land size indicates that: (1) its negative effect on the choice of local self-employment/service is distinctively larger for STs and non-Hindus than others, (2) its negative effect on the choice of migrant construction is larger for upper castes than other social groups, (3) its negative effect on the choice of migrant self-employment/service is only present for upper castes and STs while its effect is positive for Non-Hindus. Asset value increases the probability of choice of off-farm jobs primarily for non-Hindus (agricultural wage work, local construction work, local factory work) and for OBCs (local factory and local self-employment/service).

20. India was ranked 11^th from the bottom in the world in terms of female labour participation rate (ILO, 2013).

Additional information

Funding

This work was supported by the Japan Society for the Promotion of Science [25101002];