Original The Study on Poverty Reduction Effects of Chinese Urban Minimum Living Standard Guarantee System—Empirical Analysis Based on CHIP 2002 and 2007

The Chinese urban minimum living-standard guarantee system, mainly functions to guarantee the poor people to have minimum living-standard life; at same time it can make some people to get rid of poverty by some poverty lines. But how much of the rates can be reduced? What differences among provinces, and what impact on all kind of families? The paper tries to answer these questions by using Chinese Household Income Project (CHIP) 2002 and 2007, taking international poverty line, minimum living standard line, Martin’s poverty line and relative poverty line respectively to measure the poverty reduction impact of urban minimum living standard guarantee. On the whole, the urban minimum living standard guarantee is effective to alleviate absolute poverty and even part of relative poverty. For example, taking international poverty line as standard, in 2002 the urban minimum living guarantee respectively decreased absolute poverty rate, poverty gap and squared poverty gap of national urban area by 4.58%, 11.41% and 16.32%, while in 2007 decreased respectively by 47.24%, 70.87% and 83.04%. Taking relative poverty line as standard, in 2002 respectively poverty rate, poverty gap and squared poverty gap of national urban area decreased by 1.92%, 4.65% and 8.37%, while in 2007 the 3 indices reached by 3.28%, 11.63% and


Introduction
After 35 years of reform and opening up, China has entered the middle-income countries from a low-income country with backward economic development and weak government financial resources.
However, China is still a developing country with a per capita GDP of just over $6,000. Therefore, except the poverty of rural residents need to be figure out, whether absolute or relative poverty of urban residents, both of them still have in different degrees. It has always been an important problem in the economic development of social, averting the poverty pass on the generation's gap and immobilization, and eliminating urban poverty by hardworking (Note 1). In the 1990s, in order to alleviate the social impact caused by the transformation of enterprises in urban areas (Note 2), and to guarantee the basic living conditions of laid-off workers, the urban minimum living standard guarantee system came into being, building the last survival "defense line" for urban residents, namely the livelihood income line (Note 3), or urban poverty line. After more than a decade of development, the system has become increasingly perfect, playing an important role in ensuring residents' right to subsistence and maintaining "bottom line equity".
However, since it was not until 1999, China formally established the minimum living standard guarantee system, the data which can be used to evaluate the policy effect is scarce at that time (Hong Dayong, 2004;Wang Weiping, 2007;Yang Lixiong, 2008;Zhang Haomiao, 2010). The good news is that the quantitative analysis of poverty reduction effect of minimum living standard guarantee by domestic and foreign scholars has been gradually carried out in recent years, and achieved some research results. For example, Chen, Ravallion and Wang (2006), used the China's Urban households for Short-term Survey (UHSS, 2003) data, to measure the aim of targeting and poverty reduction effect of the urban minimum living standard guarantee system in 35 large and medium-sized cities in China, found that urban poverty reduction effect is mainly reflected in reducing the intensity and depth of poverty, which may well prevent the leakage to the non-poor, but due to the low coverage affected its role in poverty reduction. Wang (2007) used the data of 14 small and medium-sized cities in the "China urban employment and Social Security Survey 2004" and five different poverty line, to measure the poverty reduction function of the urban minimum living standard guarantee system. The results showed that the urban minimum living standard guarantee reduced the absolute poverty rate by 11%-16% on average and the relative poverty rate by 2%-4% on average. Therefore, she thought that the poverty reduction effect of the urban minimum living standard guarantee could be measured and more effective.
Du Yang and Albert Park (2007) analyzed the poverty reduction effect of urban minimum living standard guarantee, laid-off subsidy and unemployment insurance based on the labor market surveys conducted by the Chinese academy of social sciences in Shanghai, Wuhan, Shenyang, Fuzhou and survey data, using Khan (2004) of high and low absolute poverty line for the standard, and estimates the poverty reduction effect of urban poor, the results show that in high poverty line, low-income residents and all samples of absolute poverty rate fell 16% and 5%, respectively, poverty gap fell by 29% and 12% respectively, the square poverty gap dropped by 38 percent and 20 percent, respectively. Gao et al. (2009) obtained similar results using the same data and methods (Note 4). Xia Qingjie et al. (2007) used the data of (CHIP) (1988, 1995, 1999 and 2002) to evaluate the impact of subsistence assistance, unemployment allowance and minimum living standard guarantee policies paid by work units on poverty rate with the international poverty line, and found that the effect of these three social welfare subsidy systems on reducing poverty rate was less than 1 percentage point. Therefore, the social welfare system still has a lot of space for improvement in urban anti-poverty.
In a word, by reviewing the empirical measurement and theoretical analysis of the poverty reduction effect of social assistance system in the above-mentioned literatures, we can clearly find that the effect of the current urban minimum living standard guarantee system in China is mainly to help the recipient families reduce the poverty depth and intensity, but not to lift them out of poverty. Some policy simulation results show that the poverty reduction effect will be further improved if the urban minimum living standard guarantee can be fully covered and the amount of minimum living standard guarantee is fully distributed. In addition, the above studies mainly focus on the early stage of the minimum living standard guarantee system, so it is difficult to carry out a comparative study on the effects of cross-period policies. At the same time, due to the short implementation time of the minimum living standard guarantee system, the inconsistency among different regions, and the relative lack of information, it has not been clearly answered how much the poverty reduction effect of the national urban minimum living standard guarantee is, and how much the poverty reduction effect is different among different regions and family types. This paper makes comprehensive analysis and conducts measurement of the poverty reduction effect of urban minimum living standard guarantee in China based on the two large-scale micro-survey data of "Chinese Household Income Project (CHIP)" (2002 and 2007), which can make up for the above deficiencies to some extent and provide the corresponding basis and reform ideas for the future development of China's minimum living standard guarantee system. www.scholink.org/ojs/index.php/elp Economics, Law and Policy Vol. 3, No. 2, 2020 4 Published by SCHOLINK INC.

Data Description
The data used in this paper are urban part of the There are two questionnaires in CHIP (2002 and, one is consistent with the questionnaire of National Bureau of statistics, the other is a more detailed questionnaire designed by CHIP research team. Both questionnaires ask questions related to urban minimum living standard guarantee income. In the questionnaire of the National Bureau of statistics, one item directly asks about the minimum living standard guarantee income received by family members, while the questionnaire in the appendix asks about the social assistance income received by family members. Using the former to calculate the rate of urban minimum living standard guarantee is 2.1%, which is far lower than the statistical data officially published by China's Ministry of Civil Affairs. It is also quite different from the measurement results of Chen, Ravallion and Wang (2006), Duyang and Albert Park (2007) using large-scale survey samples. This shows that the "minimum living standard guarantee income" item in the main questionnaire is seriously underreported, while according to the survey in the appendix, the rate of urban minimum living standard guarantee is 3.7%, which is closer to the official results.
Therefore, this paper refers to the practice of Gustafsson and Deng (2007) and Gao et al. (2009), and regards "social assistance income" in CHIP (2002 and as "minimum living standard guarantee income".

Selection of Poverty Indicators and Measurement Method
The selection of poverty indicators is related to the accuracy of poverty measurement. An appropriate poverty indicator will help to make a comprehensive evaluation of the poverty situation and the effect of poverty reduction policies. Poverty index usually needs to meet a series of normative requirements, such as monotonicity axiom, transferability axiom, transferability sensitivity axiom, subset monotonicity axiom, etc. These normative requirements are important criteria to judge the merits of poverty index. According to these standards, the comprehensive poverty index (FGT) proposed by Foster, Greer and Thorbecke (1984) was adopted as the poverty index in this paper, while the poverty reduction effect of the urban minimum living standard guarantee system was based on the difference value of the comprehensive poverty index (FGT) FE before and after the implementation of the minimum living standard guarantee measures.
The ratio of its difference to the former comprehensive poverty index RE: Where q, n, z and y respectively represent the number of poor population, the number of total population, the poverty line and income of the i-th poor population. While, "α" is the poverty aversion coefficient. The larger α is, the higher the society's aversion to poverty is.
The H index represents the incidence of poverty in the total population, reflects the extent of poverty; Here, FGT is the poverty gap index, which represents the relative gap between the income level of the poor population and the poverty line, and reflects the depth of poverty; When α=2, Here, FGT is the square poverty gap index, where C represents Gini coefficient of the income distribution of the poor population and reflects the income difference among poor population. As this index gives higher weight to the individual with more serious poverty, it is mainly used to reflect the intensity of poverty.
When α>0, FGT index satisfies the monotonicity axiom and the subset monotonicity axiom. When α>1, FGT index satisfies the transferability axiom. When α>2, FGT index satisfies the transfer sensitivity axiom. In addition, FGT index also satisfies the decomposability axiom, that is, the overall poverty index is equal to the weighted average of each group's poverty index, that is, Where p i represents the population weight of group i, and i i FGT y z  represents the comprehensive poverty index of group i. Since the FGT index has a good character and reflects the poverty degree comprehensively, this paper also uses the FGT index to measure the poverty degree before and after the implementation of the urban minimum living standard guarantee, so as to analyze the poverty reduction effect of urban minimum living standard guarantee.

The Choice of Poverty Line
In the above measurement formula, the determination of the poverty line is an important selection factor. The urban poverty line, as the basis for judging the urban poverty level, reflects the acceptable minimum living standard of a city under certain economic conditions. There is no unified national urban poverty line in China. In this study, the international poverty line, the line of minimum living standard guarantee, other poverty lines and the relative poverty line were used to measure the effect of urban poverty reduction policies on different occasions. Because of the sensitivity of the effect analysis of poverty reduction policy to the choice of poverty line (Duyang, Albert Park, 2007;Gustafsson & Deng, 2007;Gao et al., 2009;Xia Qingjie, 2007), it is a reasonable method to choose several poverty lines to illustrate the poverty reduction effect of policies from different perspectives. After careful comparison, this paper selects international poverty line, minimum living standard line, Martin's poverty line, relative poverty line and other related poverty lines to evaluate the poverty reduction effect of the policy.
According to purchasing power parity, the World Bank sets the standard of $1, $2 and $3 per capita per day as a representative poverty line standard for low-income countries (Ravallion, DATT and Van de walle, 1991). Although this kind of poverty standard ignores the differences of price levels in different parts of the country, it is a simple, effective and operable measure for uniform comparison of poverty in different regions and even different countries. This paper assumes that u$1 per person per day is the rural poverty line, and chooses $2 per day as one of the urban poverty lines (Note 5). It is found that the evaluation of purchasing power of RMB in 2005 (between 2002 and 2007) is 3.45 yuan/dollar (Deaton & Heston, 2008). Therefore, the international poverty line of urban residents is converted into RMB 2484 yuan in 360 days. By this single standard, the change of poverty of urban areas in China and other provinces can be measured. The results can also be compared internationally. This is a feature that none of the other criteria below have.
The minimum living standard guarantee line is a subsidy standard for the poor formulated by each city according to the local economic development level, residents' income and consumption and local financial resources. Therefore, the biggest deficiency of using the minimum living standard guarantee line as the poverty line is that the poverty line varies by the local financial resources, so that the poverty rate measured by the minimum living standard guarantee line cannot uniformly reflect its poverty situation in some areas. However, the benefit of measuring the poverty reduction effect using the minimum living standard guarantee line as the poverty line is that it can combine poverty reduction effects with assistance practices to examine the implementation effect of the policy. Some scholars have tried to measure the poverty reduction effect of the minimum living standard guarantee directly using the minimum living standard guarantee line as the poverty line (Note 6). In view of the unbalanced regional economic development in China, the urban minimum living standard guarantee   Table 1.
In addition, this paper also chooses the relative poverty line calculated by 50% of the median of per capita disposable income of households before receiving minimum living standard guarantee. In general, the relative poverty rate in urban areas is significantly higher than the absolute poverty rate (Note 10), which reflects the difference between the two poverty standards.

Measurement of Poverty Reduction Effect of Urban Minimum Living Standard Guarantee by Provinces
Based on the selected data, this paper takes the international standard of two dollars per day per capita, the urban minimum living standard and the poverty line calculated by Martin's poverty line as the absolute poverty line, and takes 50% of the median of per capita disposable income of households before receiving minimum living standard guarantee as relative poverty line, and uses the above comprehensive analysis formula to analyze the poverty reduction effect of implementing the urban minimum living standard guarantee in China.
The results of poverty reduction in urban areas measured by international poverty line of $2 per day are shown in Table 2. In 2002, the overall urban poverty rate dropped by 4.58%, the poverty rate in Guangdong, Chongqing, Shanxi and Liaoning all dropped more than the national average. Among them, the province with the largest decline was Guangdong, which reached 33.34%, followed by Chongqing, which declined by 11.11%; The average decline of other provinces is less than 5%, while Beijing, Jiangsu, Anhui, Hubei and Yunnan remained roughly unchanged.
The poverty gap index, which measures the depth of poverty, fell by 11.41% nationwide. In all provinces, except Beijing, the poverty gap of all provinces has declined to some extent. The most significant is Chongqing, Yunnan, Gansu and Anhui, which are located in the middle and west of China, The decline range is more than 20%, of which Chongqing reaches 32.51%, which is the largest, while Beijing, Jiangsu and Henan have the smallest decline range, which is less than 1/3 of the national level on average. The intensity of poverty, which reflects the extent of disparities within poor groups, has declined by 16.32% in the whole country. Chongqing, Yunnan and Gansu are 3.23 times, 2.46 times and 2.05 times of the national average, respectively, indicating that the minimum living standard guarantee plays an important role in alleviating poverty intensity in these areas. However, the average degrees of falling in the provinces of Beijing, Jiangsu and Henan was less than 5 percent.   (1) and RE FGT(2) in Table 2 respectively represent the percentage reduction of poverty rate, poverty gap and poverty intensity, that is, the decline rate of the same index before and after the implementation of the minimum living standard guarantee in the same year. It is calculated according to the FGT method introduced in the second part of this paper; 2) The "0.00" in Table 2 indicates that the implementation of urban minimum living standard guarantee has no effect on the poverty rate (as well as the poverty gap and poverty intensity) of the sample area; "100%" means that the urban minimum living standard guarantee reduces the poverty rate, poverty gap and poverty intensity of the sample areas by 100%, that is, poor households are lifted out of poverty; "-" means that there is no poverty in the sample area under the set poverty line.   The results of poverty reduction in urban areas measured by relative poverty line are shown in Table 5.
In 2002, the relative poverty rate of the whole country decreased by 1.92%, which was smaller than the previous two poverty rates measured by the absolute poverty line. The relative poverty rate of seven provinces and province-level municipality such as Beijing, Shanxi and Liaoning did not change.
Although the relative poverty rate of five provinces and province-level municipality such as Henan,    Table 6. The above findings indicate that there is a large gap in the poverty reduction effect measured by selecting different poverty lines. Compared with the poverty line in Table 1, the lower the poverty line is selected, the more significant the poverty reduction effect is measured, and vice versa. Because there are fewer people who can get rid of poverty through the minimum living standard guarantee according to the high poverty line, that is, the high (low) poverty line pulls down (high) the poverty reduction effect of the minimum living standard guarantee system. This is the sensitivity of poverty line selection.
Thirdly, compared with the decline rates of various poverty indicators, poverty intensity has decreased more significantly, followed by poverty gap and poverty rate. See the overall comparison of the three indicators in Table 6. This illustrates the sensitivity of measurement of poverty reduction effects to the selection of poverty indicators.
Fourth, the comparison of the poverty reduction effect between the two years shows that the poverty reduction effect in 2007 is more prominent, and the decline of some indicators in 2007 is more than 60 percentage points higher than that in 2002. See the comparison in the last row of Table 6. This shows the level of minimum living standard guarantee is an increasing function of the level of economic development.  Source: According to the relevant data in Tables 2-5.

Measurement of Poverty Reduction Effect of Urban Minimum Living Standard Guarantee System by Types of Urban Households
Next, this paper attempts to measure the poverty reduction effect of urban minimum living standard guarantee by household type, such as family size, gender, age, physical condition and employment status of household head. In order to simplify the analysis, only two poverty lines are selected: one is the international poverty line of $2 a day, which is roughly between the minimum living standard guarantee line and the Martin's poverty line; the other is the relative poverty line calculated by 50% of the median of per capita disposable income of households before receiving minimum living standard guarantee to measure the poverty reduction effect of urban minimum living standard guarantee on different types of families.
The poverty reduction effects of urban minimum living standard guarantee on different types of urban households measured by international poverty line of $2 per day are shown in Table 7. According to this poverty line standard, in 2002, the minimum living standard guarantee had no effect on single households; among two-family, three-family and multi-person households, the decline in poverty was greatest for multi-person households. Among the impacts on poverty gap and square poverty gap of all kinds of households, the decline range of three person households is the largest, and that of multi-person households is the smallest. This shows that the minimum living standard guarantee system can promote the family with large population to get rid of poverty, and also help the family of three to reduce the depth and intensity of poverty. For urban resident with different gender heads of households, poverty reduction effect of the female-headed households brought by the minimum living standard guarantee was more significant than that of male-headed households. For urban households with different age heads, the poverty rate and poverty gap of middle-aged households decreased more than that of elderly households and young households, but the square poverty gap decreased less than that of young households. For households with different physical conditions, the poverty rate of households with physical disability after receiving the minimum living standard guarantee decreased by 0.27 percentage points lower than that of households without physical disability, but the poverty gap and square poverty gap are 2.03% and 3.23% higher than the latter. For the urban households with different heads of employment, the poverty rate, poverty gap and square poverty gap of the unemployed households decreased by 4.55%, 14.14% and 18.74% respectively in 2002, among which the poverty rate of the unemployed households was 0.07 % lower than that of the employed households, while the other two poverty indexes were 4.97% and 5.48% higher than the latter. This shows that the urban minimum living standard guarantee system can promote the head of household employment families out of poverty, so that the \head of household unemployment family effectively reduce the poverty level.
In 2007, the ability of minimum living standard gurantee to reduce absolute poverty improved significantly. Among the four categories of households divided by population size, the effect of poverty reduction was the most significant for single-person households after receiving the minimum living standard guarantee. The absolute poverty rate, poverty gap and square poverty gap decrease by 80%, 95.85% and 99.14% respectively; However, the three poverty indicators of double-person households, three-person households and multi-person households decreased by more than 35%, 55% and 75% respectively, while the above poverty indicators of these three-person households in 2002 did not decline by more than 5%, 15% and 25%.In 2007. The decline of absolute poverty rate, poverty gap and square poverty gap of female headed households were 6.7 times, 4.9 times and 4.3 times of 2002 respectively, and the decline of three indicators of male headed households were 12.7 times, 6.4 times and 5 times of 2002 respectively; In terms of poverty rates alone, female-headed households declined less than male-headed households; In terms of poverty gap and square poverty gap, the decline of female headed households is significantly higher than that of male headed households. In 2007, the poverty rate of middle-aged households fell by the most after receiving minimum living standard guarantee, to 53.03%, nearly 8 times higher than in 2002. In terms of the poverty gap and the square poverty gap, the largest decline was in the elderly households, which reached 82.44% and 92.66% respectively, 9.1 times and 6.8 times higher than that in 2002, indicating that the elderly households' minimum living standard guarantee achieved remarkable poverty reduction effect in 2007. In 2007, the decline rate of the three poverty indicators of households with head disability reached 100%, which means that the absolute poverty of disabled households has been completely eliminated by the aid of minimum security; but all three indicators fell by less than 20% in 2002. In 2007, the three poverty indicators of unemployed households decreased by 52.80%, 73.86% and 84.21% respectively, 17.38, 14.25 and 9.10 percentage points higher than that of employed families. According to the employment status of household heads, the three indicators of unemployed household heads decreased by 1.34%, 4.73% and 8.53% respectively, which were 1.10, 2.48 and 4.14 percentage points higher than those of heads of employed families, indicating that the minimum living standard guarantee in 2002 played an important role in alleviating poverty of unemployed household heads.
In 2007, the relative poverty rate of single households was not affected by the minimum security assistance, while the poverty rate of double-person households, three-person households and multi-person households decreased by 1.62%, 0.65% and 2.79% respectively; In terms of the poverty gap and the square poverty gap of all kinds of households, the decline of single households was the largest, with 15.86% and 37.4% respectively, followed by 8.06% and 15.51% for multi-person households. According to the gender of the household head, the three indicators of female-headed households dropped by 1.95%, 6.80% and 15.91%, respectively, and were 0.82, 1.05 and 3.7 percentage points higher than that of male households. According to the age of the household head, the relative poverty rate of the middle-aged households decreased by 118%, which was the largest.
However, the poverty gap of elderly households decreased by 3.51% and 8.76% more than that of middle-aged households and young households, and the square poverty gap decreased by 9. 30% and www.scholink.org/ojs/index.php/elp Economics, Law and Policy Vol. 3, No. 2, 2020 19 Published by SCHOLINK INC.
20.52% more than that of middle-aged households and young households, respectively. According to the physical condition of the household head, the decline rate of the three indicators of the disabled family reached 37.5%, 52.63% and 71.94% respectively, which was the largest. According to the employment situation of the head of household, the decline of the three indicators of the head of household unemployed households reached 1. 97%, 12.50% and 25.58%, 1.13%, 9.15% and 18.61% higher than that of the head of household employed families respectively, and 0.47 times, 1.6 times and 2 times higher than that of 2002.

Brief Conclusions and Policy Recommendations
Based on the urban data of China's Household Income Survey (2002 and, this paper measures the poverty reduction effect of China's urban minimum living standard guarantee system and policies in a more comprehensive way than previous studies and draws the following brief conclusions: China's minimum living standard guarantee system are effective. The direction of the minimum living standard guarantee system is to help the majority of poor families to alleviate poverty. This goal has been basically achieved from all angles, and the result also makes a few families completely out of poverty. This paper measures the poverty reduction effect of the urban minimum living standard  2002; second, it shows that urban minimum living standard guarantee reduces absolute poverty more significantly than relative poverty; third, it also shows that urban minimum living standard guarantee reduces poverty intensity more significantly than poverty gap and poverty rate.
Comparing the poverty reduction effects of the urban minimum living standard guarantee system in different provinces and province-level municipality, it can be seen that the effect of most provinces and regions is significant, even all of them have been lifted out of poverty, but some of them are not of that year is low, so that all kinds of standards are similar, and the measurement results are almost.
In 2007 The comparison of poverty reduction effects of urban minimum living standard guarantee on different family types cannot be generalized, but must be treated differently. One is in addition to the individual circumstances all types of families after accepting threshold has the effect of different degrees, but various indicators, the poverty rate of decline is less than the poor from the peace party from declines, it shows that the threshold of poverty reduction effect on poverty reduction to the sensitivity of the index selection, namely not only single poverty reduction index for measure. First, except for some cases, all types of families have different effects after receiving the minimum security assistance, but various indicators are different, in which the decline of poverty rate is less than that of poverty gap and square poverty gap, which shows that the poverty reduction effect of the minimum living standard guarantee is sensitive to the selection of poverty reduction indicators, that is, it is not easy to use a single poverty reduction indicator as a measure. Second, the effect of different households is different in different years: for example, in 2002, according to the standard of absolute poverty, female headed households and three person households benefited the most; According to the standard of relative poverty, households with unemployed heads and disabled heads benefited the most. In 2007, according to the standard of absolute poverty, disabled households, single households and elderly households benefited the most; In terms of relative poverty, households with a disability, single households and households with an unemployed head benefited the most. This shows the sensitivity of household types to poverty reduction policies, that is, it cannot be assumed that the same poverty reduction criteria will benefit all households equally, which requires that poverty reduction policies should be more targeted and differentiated.
To sum up, the current urban minimum living standard guarantee system and policies in China can alleviate the absolute poverty and partial relative poverty of urban residents to a large extent, reduce the poverty level of most households, and the more poverty-stricken households, the more significant the effect of poverty reduction, and even make some people completely out of poverty, of course, In order to better cope with challenges of urban poverty and make all the residents including the poor, share the fruits of economic development, the urban minimum living standard guarantee system should be further improved. First, in the standard design of the urban minimum living standard guarantee system, we should break through the concept of subsistence and anti-absolute poverty, and emphasize the relative standard and enhance the self-development ability of the recipients. Second, in the process of institutional construction, we should gradually establish a normal growth mechanism and a dynamic subsidy mechanism in which the security standard is linked to the level of economic development, build a diversified financing mechanism with the government as the main body and multiple social participation, and ensure the supply of minimum living standard guarantee funds. The third is to gradually improve the level of management and the unity degree of urban minimum living standard guarantee system. Fourth, in view of the fact that the measurement of poverty reduction effect is quite sensitive to the selection of poverty line and poverty indicators, when evaluating the implementation effect of various social security policies, such as the minimum living standard guarantee, we must emphasize the comprehensive application of multiple poverty lines and indicators to prevent the possible biases caused by certain poverty lines and indicators.