Abstract
China has achieved high economic growth for over 20 years but still has a sizeable number of people living in poverty. Using large household survey data, this paper attempts to present a comprehensive picture and identify the main determinants of poverty in both the urban and rural sectors. It concludes that despite rising urban poverty in recent years, rural poverty is still predominant in China. The slow progress in reducing poverty has been caused by rising inequality,particularly the urban–rural divide and interregional inequality. In sharp contrast to the government’s optimism in poverty reduction, this paper shows that poverty in China is still serious and has become increasingly more difficult to reduce. Although China may easily meet the international development target of reducing poverty by half from 1990 to 2015, the battle against poverty is far from over, even by 2015.
Keywords: Inequality; Poverty; China
1. Introduction
Economic growth in China has been impressive by international standards. Although it is the most populous country in the world, China has outperformed the majority ofcountries in the past two decades. Remarkable economic growth helps improve people’sliving standards and reduce poverty.
Based on official statistics, the total number of rural poor declined from 270 million in1978 to 26 million by 2000 (People’s Daily, 2000), but the official figures are ques-tionable. This paper estimates that there are up to 5.5–15 million urban poor and 103–187million rural poor in 1998. The battle against poverty is far from over in China, and newgroups of poor population are emerging, making poverty reduction more difficult in thefuture than before.
Poverty has been considered to be primarily a rural phenomenon due to the huge per capita income disparity between the rural and urban populations. For historical reasons,Chinese peasants have been disadvantaged by the urban-biased policy, which severely restricts urban–rural migration and provides subsidies to the urban population at the expense of its rural counterpart.
However, recent research shows that urban poverty exists in a significant scale due to massive unemployment and urban–rural migration. Meng (2001) shows that 17.5 million state-owned enterprise (SOE) workers were made redundant by the end of 1998. Many of these unemployed workers may be trapped in poverty due to lack of income. Millions of rural residents working in the cities may also live in poverty as they either engage in lowpay jobs or face unemployment.
One of the most difficult problems faced by researchers on poverty study in China is how to compare the poverty situations of two highly heterogeneous populations (rural and urban). Apart from the huge disparity of per capita urban–rural incomes, comparable data are usually difficult to obtain, not because they are not available, but because the statistical authority is unwilling to provide them to independent researchers.
This paper is the first attempt in the literature to study urban and rural poverty at the same time using household level data. The urban data cover all the provinces in 1998. The rural data cover 19 provinces in 1995 and 1998. Such complete data sets have never been possible in any other studies.
It is concluded that despite rising urban poverty, rural poverty is still a predominant problem. The slow progress in poverty reduction has been due to rising inequality, particularly the urban–rural divide and interregional inequality.
The rest of this paper is organised as follows. Section 2 estimates the incidence of urban poverty and studies the effects of income and inequality on poverty. Section 3 estimates the incidence of rural poverty and studies the effects of income and inequality on poverty. Section 4 analyses and compares urban–rural inequality. Section 5 has some simulations on poverty reduction in rural China. Section 6 concludes and makes policy recommendations.
2. Incidence of urban poverty and its response to income and inequality
This section uses the urban household survey data in 1998 for all the provinces. The data set contains 17,000 households in 31 provinces. To estimate the incidence of poverty, the first step is to define an urban poverty line. Because there are significant differences across regions in prices, consumption patterns, and tastes, it is useful to estimate a poverty line for each individual region. However, we cannot have a separate poverty line for each individual city because there are not enough observations in each city to have a robust estimate of a poverty line. Instead, we opt to estimate poverty lines for individual provinces.
Following Ravallion (1994), the estimation of poverty line starts from the definition of a food poverty line. It then evaluates the basic nonfood component. Depending on how the basic nonfood component is evaluated, two poverty lines, i.e., low and high, are estimated for each region.
A basic need basket of food items is defined to evaluate the cost of food consumption.In this paper, the poorest 20% of households ranked by per capita income are considered to be the relevant group whose consumption pattern is close to that of the households who may live in poverty. As a result, all the food items with complete price and quantity information from this group of households are listed. The food quantities are converted into calorie equivalents using a nutritional conversion table provided by the National Nutritional Society. The unit cost of calories consumed is then computed. This unit cost multiplied by 2100 (the minimum requirement of calorie per capita per day) and 365 (the number of days in a year) yields the value of a food poverty line. At the national level, the food poverty line in 1998 is 1390 yuan per capita per year, which is 23.63% of the national average per capita income.
The values of food poverty lines vary substantially across provinces, ranging from 941 yuan in Qinghai to 2361 yuan in Shanghai. This implies that the same amount of food energy costs 2.5 times as much in Shanghai as in Qinghai. The differences in the values of food poverty lines reflect the extent to which food prices differ spatially.
Given a food poverty line, denoted by ZF for any particular region, the low and high poverty lines for that region can be derived from running the following regression.
(1)
where the subscript j denotes a household, S is the ratio of food expenditures over income,x is food expenditures, and n is the number of household members.
Once the parameters in Eq. (1) are estimated, the low and high poverty lines can be evaluated by the following formula.
The low poverty line, denoted by :
The high poverty line, denoted by :
where ln(n) is the logarithm of average household size. The same regression is repeated for all provinces. The national poverty lines are the weighted averages of the provincial lines by provincial populations instead of the survey sample populations.
In 1998, the total urban population was 312.4 million people. The average per capita disposable income was 5633 yuan per year. The low and high poverty lines are, respectively, 1869 and 2316 yuan (Table 1). These are 32% and 39% of per capita income, respectively. Like the food poverty lines, the values of low and high poverty lines vary significantly across provinces. The evaluations of food poverty lines reflect consumers’ tastes and consumption patterns as suggested by Ravallion (1994) and Deaton(1997).
With the provincial poverty lines, the incidence of urban poverty, measured as the proportion of urban people living at or below the poverty lines, can be derived for each individual province and for the whole country. The poverty headcount rates by province derived from the 1998 urban household survey data are presented in Table 1.
The incidence of poverty varies widely across provinces, ranging from less than 1% in Beijing and Guangdong to over 11% in Tibet, Shaanxi, and Ningxia. In general, the poverty incidence in the central and west regions is significantly
上一页 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] ... 下一页 >>