Week 5 (03 - 09th Mar 2014)

• Work Finished
  

Visiting Nanmazhuang Coop and Huzhai Coop in Lankao County of Henan Prov, Hongfeng Coop and Lantian Coop in Shangqiu City of Henan Prov, and Minlong Coop in Shouguang County of Shandong Prov.

• Work to Do

Visiting Jihetang Coop in Taian City and 2-3 coops in Dezhou City of Shandong Prov.

• Questions

1. This issue of migrant workers is not simply just related to the social capital. I think it also concerns the farmer’s characteristic of multi behavioural goals. Farmers do not just seek to earn money. When having sufficient money to live a good life in the village, some of them tend to remain in the village and live with their parents and children. Working as a village leader might earn less than being employed in the city, but they can take care of their parents and support children’s education conveniently, and also enjoy more respect. Moreover, this characteristic differentiates between the young and the old. General speaking, those who age above 35 (or 40 in some regions) tend to stay in the village if this does not reduce their income too much. While those age below 35 never think about taking farming as their vacations and spending the rest of their lives in the village. It also differentiates between the regions. Farmers in the southern areas (especially the coastal areas) seem never stop earning money even through they are much richer than farmers in other regions of China. They will definitely not remain in the village if they can earn more in the city. However, the inland farmers will probably stop working hard and to “enjoy” their lives once they think they can.

In some regions, both the young and old labours, male or female, migrant to work in the city. In some regions, the young women do not work in the cities. They take the responsibility of supporting the children’s education, working on their farm land and looking after the old after marriage.  While in some other regions, no labour above 40 move into the city. There are probably some good farming industries (usually done by agri companies, agri coops, etc) in the village. The money they earn from the industries can satisfy them. Thus, the percentage of migrant workers in a village is different in different regions. This phenomenon might relate to the per capita farmland, the agri product varieties they farm, the distance to the nearest city and so on. However, there is no record of the data of migrant workers in villages, which makes it difficult to test these guesses.

2. There is no unified mechanism of forming the cooperation among the smallholders, even through in this study we focus on only one of the forms of cooperation, the so called quasi-coop.  Therefore we cannot take any coop as a representative one, build a mathematical model to describe the mechanism and use the statistical data of random selected samples to evaluate the parameters. A sensible approach is to do case study and computer simulation. We should first delve into a few typical cases and make clear under what conditions they form, what affect the formation significantly and how the formation proceeds. Then we probably can extract the rules that different types of concerned agents behave and the key variables that affect the formation of the cooperative order. The behavioural rules can be simulated by an agent-based model, in which different circumstances of forming the order can be reflected by different settings of parameters of the key variables. Meanwhile, there are some necessary conditions for generating the order.  The conditions, existing in all circumstances, can be reflected by the mathematical model. We then can use the statistical data to test the hypotheses and evaluate the parameters.

To sum up, we will use the methods of case study, agent-based simulation and statistical modelling in this study. They will be used to deduce the mechanism from the reality, induce and test the mechanism in simulation experiments, and test the common quantitative relations among variables embedded in the mechanism respectively.

3. The relations that we would test are like

(1) Village level: the higher the social capital or entrepreneurship of a village, the higher the likelihood that it has coops.

(2) Household level: the higher the social capital or entrepreneurship of a smallholder (largeholder), the higher the likelihood that it joins (leads) a coop.

Can these relations be tested in the function like

Coop = α + β*SC + γ*Ep + δ*Ctrl +  ε

This function can be used at both village level and household level.

Village level:

• Coop can be a dummy variable, i.e. it equals to 1 if there is a coop in the village, 0 otherwise. It can be the number of coops that the village has. It can also be the percentage of households that join a coop in the village.

• SC is the social capital of this village 5 or 10 years ago.

• Ep is the entrepreneurship of this village 5 or 10 years ago.

• Ctrl is a set of control variables, i.e. other variables that affect the formation of coops in a village (e.g. per capita income, per household land area, market condition) 5 or 10 years ago.

Household level:

• Coop can be a categorical variable, i.e. it equals to 0 if the household does not join a coop, 1 if it is a regular member of a coop (smallholder), 2 if it is a core member of a coop (largeholder), 3 if it is the leader of a coop.    

• SC is the social capital of this farmer 5 or 10 years ago.

• Ep is the entrepreneurship of this farmer 5 or 10 years ago.

• Ctrl is a set of control variables, i.e. other variables that affect a household’s participation of a coop (e.g. per capita income, per household land area) 5 or 10 years ago.