Climbing the development ladder: Economic development and the evolution of occupations in rural China

Abstract

We study how occupations evolve across space and time during the development of an economy. Using a data set on more than 200 villages from 8 provinces in China, we examine the main occupations that have characterised China's labour markets since the economic reforms. Our findings reveal a systematic evolutionary pattern of occupational emergence: the evolution of occupations proceeds from traditional and fairly simple forms of subsistence agriculture to modern, more complex manufacturing and service firms. Our findings suggest that rural development in China is being built by a process that can be described by the climbing of a development ladder with each step up the ladder denoting the economy's transition into a more complex occupational regime.

Acknowledgements

The authors would like to thank two anonymous referees for their comments and James Eaves, Arizona State University, for his encouragement and helpful suggestions.

Notes

1. Although there are several reasons why wealth is important to the emergence of different occupations, the one most pertinent to developing countries is that in poor areas innovators of new productive activities face restricted access to capital (Eswaran and Kotwal, 1986). Low income regions are characterised by poorly developed financial markets and credit is invariably rationed according to the ability to offer collateral (Pische et al., 1983; Rudra, 1983). High factor costs such as that for capital which arise in poor areas due to the small size of markets, are an important determinant of the of occupations that emerge in the economy (Chenery, 1960; Kuznets 1966; Chenery and Taylor, 1968). Wealthier regions with better access to capital and credit, will therefore, experience faster expansion of their off-farm occupations than poorer regions; especially in the occupations such as large scale entrepreneurship, that are more capital intensive and bound by working capital and credit constraints

2. Areas that are closer to the urban centers have a larger demand for goods and services and better access to information, modern technology, markets, infrastructure and lower transaction costs. Therefore, occupations such as large-scale industries that require access to information and technology in order to operate efficiently, are more likely to develop more rapidly in rural areas that are closer to urban areas than in more isolated regions.

3. Micro-enterprises are believed to develop by taking advantage of market niches and isolated locations in which they do not have to face competition from larger firms but where they have sufficient access to small venture capital (Sherer, 1970; Pryor, 1972; Fafchamps, 1994). The literature on the location of large scale manufacturing emphasises the importance of attributes that are linked to urbanisation and affluence of a region; the location decision seeks to minimise costs of marketing and distribution, procuring material inputs, transportation and infrastructure and access to capital in the region.

4. While it is somewhat unusual for village leaders to be able to provide employment data for their villages, in China it is possible and we believe that the data are fairly accurate. To make the estimates, what is needed is an estimate of the size of the labour force (the denominator) and a count of the number of individuals that are participating in each type of occupation (the numerator). Information on the size of the village labour force is relatively straightforward, since it is a figure that villages are required to report on year-end statistical forms to township officials. In other words the denominator is a figure that is updated by the village accountant each year as a standard part of his job. To get the counts of individuals in each village that are in each occupation, during the survey, the village leader and accountant used a comprehensive list of all of the village's households. They then went down the list household-by-household and created running tallies of the number of individuals that they believed were involved in each occupation category. We have used this procedure many times and because of the intense interaction between village leaders and households (leaders are responsible for many village activities: they distribute land to farmers; collect taxes; implement family planning; oversee irrigation; co-ordinate health care; and many other activities. When going through the list of households it is clear from their interactions with each other that local leaders know each household and have a reasonably good idea of what members of each are doing.

5. Local rural industry workers include in-migrants, commuters and local wage earners. Among these, in-migrants are those who come from other villages in search of wage earning jobs. Commuters include those workers who come into the village from a nearby area for wage jobs and those, from within the village, who commute daily to work in a factory in a nearby area. Local wage earners, as explained before, comprise of people from within the village working in a local factory.

6. We would like to thank an anonymous referee for suggesting the decomposition in occupation shares.

7. Relative to Figure 2 which depicts the village means of occupational shares, in Figure A2 we present a vertical collection of dots for each income category, with each dot representing a village in that category. Note that the mean level of employment shares in each income group are indicated by a horizontal line of pluses, and solid lines indicate the mean plus or minus the standard deviation of each group. This provides a sense of the extent of dispersion in occupational shares within each of the income categories.

8. It is important to note that we are dealing with village-level not individual-level data. Therefore the transition parameters refer to the aggregate shift of communities across occupational categories which characterises the observed occupational evolution and not to the probability of any one individual worker transitioning from one occupation to another.

9. We treat land as a fixed factor in production due to the administrative restrictions on land use in rural China. The absence of hired labour in the agricultural production function also reflects another feature of China's rural economy; thin on-farm labour markets (see Brandt et al., 2001).

10. We assume that migration earnings are realised within the production period, at the end of which the agricultural and enterprise output is realised and factor costs are paid.

11. We assume for simplicity that the household uses its own labour and capital purchased from the market as inputs into the enterprise.

12. The underlying model of occupational choices in a village is discrete but the observed dependent variable is a proportion (of people making a particular discrete choice).

13. In the high income villages the share of micro enterprises increases by 10 per cent while the share of large enterprises increases by almost 20 per cent.

14. We treat land as a fixed factor in production due to the administrative restrictions on land use in rural China. The absence of hired labour in the agricultural production function also reflects another feature of China's rural economy; thin on-farm labour markets (see Brandt et al., 2001).

15. We assume that migration earnings are realised within the production period, at the end of which the agricultural and enterprise output is realised and factor costs are paid.

16. We assume for simplicity that the household uses its own labour and capital purchased from the market as inputs into the enterprise.