Poverty Vulnerability Measure And Its Influencing Factors

Poverty and vulnerability have become a forward-looking hot topic in development economics. However, there are some problems. Firstly, different researchers define and measure vulnerability in different methods. It does not have a unified approach in the definition and measurement of vulnerability until now. Secondly, a few researchers pay attention to the accuracy of the measuring. And they pay a little attention to analyzing the factors that influence on the vulnerability. Thirdly, as for vulnerability, the estimates of permanent income are very important, while most researchers only use multivariate statistics and a simple panel model to estimate the earnings function, without considering the problem of data heterogeneity. Therefore, the paper has two main purposes. First, we measure the vulnerability of the poverty, when we are the first people that use the statistical analysis of multi-level model that called growth model to predict the future development of permanent income, at the same time, we focus on how the choices of a different vulnerability line and poverty lines have influence the accuracy of measurement of vulnerability. Second, we apply the multi-level Logistic regression model to analysis vulnerability, focusing on the factor of geographical environment, human capital and production investment on the vulnerability of rural households and put forward policy recommendations.First, we analysis the situation of poverty of the border areas of Yunnan province, mainly paying attention on difference of the different regions, different geographical conditions and different ethnic on the poverty breadth, poverty depth and poverty intensity of rural households, and we also analysis the difference on the poverty breadth, poverty depth and poverty intensity of different ethnic households that on the same geographical environment. From these perspectives, we examine the poverty status of border areas in Yunnan province.When we measure the vulnerability, we use multi-level model---liner growth model to estimate the future permanent income of households, as well as focus on the difference of choice of the poverty line and different vulnerability line. We do the empirical research with the 2006-2008 panel data of Honghe County in Yunnan Province. Finally, we focus on analyzing of the geographical environment, human capital, physical capital on the vulnerability of rural households after calculating vulnerability.The main conclusions of the empirical analysis are as following:First, seen from the border areas of Yunnan current situation of poverty, Poor rural households are significant difference in different geographical conditions. Compared to other cities and counties in Honghe, the poverty of rural households in the southern six counties (which are in the border area in Honghe ) are significantly broader, deeper and the stronger. The poverty of farmers in mountainous and hilly areas are significantly broader, deeper and the stronger than plain poor farmers. The ethnic villages and the non-Ethnic village are not difference in poor breadth, depth, strength. Secondly, compared with the constant coefficients random effect model, growth model has obvious advantages, as for the growth model considering the heterogeneity of the longitudinal data. Besides, the statistic test result----- -2Loglikelihood of the growth model is smaller, which means that the growth model is better than the random effect model.Thirdly, the forecast accuracy of vulnerability is closely related to the choice of poverty line. The predicted results of vulnerability are the best than others, when the poverty line chosen as the international standard two U.S. dollars options. When comparing the domestic standard, predicted results of vulnerability are the better when selecting the low-income line than selecting the absolute poverty line. The vulnerability line at 50% and at the incidence of poverty has little effect on the accuracy of the vulnerability. Seen from the calculated results of the vulnerability, there are significant regional differences. The more natural regions with better economic conditions, the less they are vulnerability. The southern six counties (which are the border counties bordering Vietnam) significantly higher proportion of vulnerable households. The different locations have significantly effect on the vulnerability, and the different topography of farmers in the same location also are different greatly. Households in the plains and hilly areas are less vulnerable than the mountain areas'farmers.Finally, seen from the analyzing influence factors on the vulnerability, the factors---human capital, physical capital, the geographical environment have significant influence on the future chance of the poverty of farmers.