China’s Multi-Dimensional Poverty and Trade

Elisabetta Croci Angelini and Yang Liu


Since the late 1970s, China has achieved remarkable economic growth. According to the 2010 poverty standard, in 1978, 770.39 million people in China lived below the poverty line of RMB 366 per capita, comprising 97.5% of China’s population.1 In 2015, under the 2010 poverty standard, the poverty line was RMB 2,855 per capita, and 55.75 million people were considered to be living in poverty, corresponding to 5.7% of the population. In less than 40 years, China had lifted 714.64 million people out of poverty, with a reduction in the rate of poverty of more than 90 percentage points. Over the same period, according to data from the National Bureau of Statistics (NBS) of China, the total value of China’s imports and exports rose from $20,640 million (RMB 35,500 million) in 1978 to a peak of $4,301,527 million (RMB 26,424,177 million) in 2014.

Poverty is a complex concept, and evaluating it in monetary terms neglects many aspects that are relevant to policies aimed at fighting it. Widespread agreement on the advantages of a multi-dimensional assessment of poverty has existed for many years. Multi-dimensional poverty represents a state of deprivation and has no direct impact on imports and exports. Yet the degree of deprivation expressed by each variable in an assessment of multidimensional poverty has a direct impact on production and consumption and affects international trade indirectly.

Evolution in China’s Reduction of Poverty

The dramatic changes in China’s international trade increased income and contributed to the fight against poverty. These changes can also be seen in different poverty standards. Based on China’s 1978 poverty standard, between 1978 (when the poverty line was RMB 100 per capita and counted 250 million poor) and 2005 (when it was RMB 625 per capita and comprised 23.65 million), 226.35 million Chinese were lifted out of poverty, and the poverty rate dropped from 30.7% of the population to 2.5%. If this is measured in terms of the 2010 poverty standard, China achieved further progress by lifting 23.09 million people out of poverty between 2005 (when

China’s population living in poverty based on the standards in 1978, 2008 and 2010. Source

Figure 6.1 China’s population living in poverty based on the standards in 1978, 2008 and 2010. Source: National Bureau of Statistics of China, Poverty Monitoring Report of Rural China 2016 compiled by the Department of Household Surveys, Table 1-3-1, p. 6, and Table 8-1-1, p. 182.

the poverty line was set at RMB 1,742 per capita) and 2015 (when it was RMB 2,855 per capita), as the poverty rate dropped from 30.2% of the population to 5.7%.

As China’s economy developed, the poverty line, which initially targeted the rural poor, rose, too. Consequently, the poverty line raised after every adjustment, and at the same time the calculation methodology changed. The 1978 poverty standard was set up with the problem of hunger in mind, the 2008 poverty standard was meant to address the problems of hunger and clothing, and the 2010 poverty standard employed a multi-dimensional welfare approach.

Figure 6.1 shows the number of the poor (in 10,000 people on the vertical axis) calculated according to the three above-mentioned poverty lines. Some overlap exists because the NBS does not provide information about the three poverty standards for every year between 1978 and 2015.

Evolution in China’s International Trade

The evolution in China’s international trade can be split into three periods bounded by the economic reforms in 1978 and membership in the World Trade Organization (WTO) in 2001.

Before 1978

There is little doubt that 1978 was a turning point in China’s trade history. Following the economic reforms initiated in 1978, China began to engage in trade and liberalisation on a huge scale, almost unparalleled in economic history. After Deng Xiaoping rose to power in China, he quickly abandoned most of the ideological measures and policies of the Mao era in economics and ended the country’s isolation from the rest of the world. The series of reforms, labelled Reform and Opening Up, enabled the Chinese to benefit tremendously from integration into the global trade system. From this point onwards, China engaged in trade with as many countries it could and opened up its foreign trade system as much as possible. The value of imports and exports, which in 1950 was RMB 2,130 million and RMB 2,020 million, respectively, increased in 1978 to RMB 18,740 million and RMB 16,760 million, respectively; this represented great progress, as did the annual growth rate of 7.8%. Despite this high growth rate, these values were low compared with what followed.

Between 1978 and 2001

Before 1978, China’s foreign trade system was fully centralised, dominated by about a dozen specialised foreign trade companies organised along product lines and based in Beijing. Since 1978, China has established special economic zones (SEZs) and successively opened coastal cities as part of opening up. State-owned enterprises and joint ventures were permitted to manage their own materials and products. In later years, some joint ventures could engage in international business as well. After 1978, trade rose slightly every year, and between 1984 and 2001 it increased sharply. The value of imports and exports increased from RMB 18,740 million and RMB 16,760 million, respectively, in 1978 to RMB 2,015,920 million and RMB 2,202,440 million in 2001. In this period, the annual growth rate was about 25%.

Between 2001 and 2016

Since China joined the WTO in 2001, as it became part of the world market economy and the multilateral trading system, it has fully engaged in bilateral trade agreements and acted as a driving force for foreign trade. China’s export markets expanded by increasing supply in other countries’ markets. A strong expansion in investment in China followed, and then it began its own outward foreign investment. Annual economic growth from 2001 to 2014 was 15.5%, a decline from the previous period. Nevertheless, in 2014 the value of imports and exports rose to RMB 12,035,803 and RMB 14,388,375 million, respectively, making China the world’s leading exporter and second-largest importer. In 2008, because of the global financial crisis, the value of imports and exports dropped but recovered in 2010 and thereafter continued the previous growth momentum. In 2015 and 2016, the value of imports and exports fell slightly but did not change China’s global trade ranking. Figure 6.2 traces the growth in the value of China’s imports and exports. The figure shows China became integrated into international

Evolution in the value of China’s exports and imports, 1950-2019

Figure 6.2 Evolution in the value of China’s exports and imports, 1950-2019.

Source: National Bureau of Statistics of China.

trade (including suffering from and reacting to two global crises), rising from a rank of 32nd in international trade in 1978 to the sixth largest importer-exporter in the world in 2001, and then to the largest exporter and second-largest importer in the world in 2009.

Multi-Dimensional Poverty

The first assessment of poverty in terms of food and non-food goods, subdivided into these two parts of poverty in a monetary variable, dates back to the book Poverty: A Study of Town Life (Rowntree, 1901), which identifies the budget required to afford a ‘shopping basket’ of ‘the minimum needed to maintain physical strength’. But the actual conditions of poverty cannot be entirely captured by income alone. Then, in 1987, Hagenaars introduced leisure into the study of poverty and constructed the first multi-dimensional poverty index using the dimensions of income and leisure.

In 1990, the World Bank (1990) introduced an international poverty line of $1 a day, which was raised to $1.90 per day in 2015 (Ferreira et al., 2015).

Later, Sen (2000, 2004a, 2004b) studied poverty at the level of personal welfare, then at the level of social development. He pointed out that poverty encompasses many dimensions, some of which are objective (e.g. having clean drinking water, roads and sanitation), and others are subjective feelings of poverty.

Based on Sen’s capability approach, Alkire and Foster (2011) constructed a multi-dimensional poverty index to fully reflect the multi-dimensional deprivation experienced by the poor. The index includes important indicators (e.g. environmental poverty and asset poverty) in addition to income. In 2010, based on Alkire’s measurement, a multi-dimensional poverty index (MPI) was added to the Human Development Report by the United Nations Development Programme (UNDP), replacing the Human Poverty Index (HPI) and enhancing the measurement of poverty.

Multi-Dimensional Poverty in China

Few studies have been conducted on multi-dimensional poverty in China. This new concept mainly focusses on measuring the differences between urban and rural areas and in some poor areas.

Wang and Alkire (2009) measured multi-dimensional poverty in China in 2006, based on Alkire and Foster’s method. They measured poverty in nine provinces along eight dimensions (housing, clean drinking water, sanitation, electricity, assets, land, education and health insurance) and concluded that multi-dimensional poverty is more serious in rural areas than urban areas, and deprivation in sanitation, electricity, education and health insurance contributes most to poverty'.

Gao (2012) calculated the MPI in Chinese urban and rural areas with the education, health and living standard based on the China Health and Nutrition Survey (CHNS) dataset in 2000 and 2009. His main conclusions are that multi-dimensional poverty is more serious in rural areas than urban areas, and health insurance, sanitation, cooking fuel and housing are important dimensions of poverty reduction. Moreover, the government should pay more attention to the provinces in central and western China.

Guo and Wu (2012) measured the MPI in Shanxi Province with county- level data. They concluded that the lack of education and health care are the main reasons for multi-dimensional poverty in Shanxi, so the government should focus its efforts on improvements in those two domains. Shen and Alkire (2017) estimated MPI for China using China Family Panel Studies (CFPS). They found that deprivation in nutrition, education, safe drinking water and cooking fuel contribute most to the overall non-monetary poverty in China.

Empirical Evidence on Chinese Household Income Project (CHIP) 2013

Whereas the MPI combines health, education and living standards, we separate income and living conditions to divide information about the source of income from the employment level within a household. Therefore, multidimensional poverty is assessed through four dimensions (health, education, income and living standards) employing the CHIP 2013 dataset, which has the most complete information but does not cover all Chinese provinces. The four dimensions are evaluated based on three data levels of analysis: the individual level, the household level and the provincial level. People live in a household, where the consumption of several goods is usually collective, and some income redistribution exists between earners and dependent members of the household. So, like many other studies about distributive issues, we use the household as one level of analysis. We also investigate multi-dimensional poverty at the provincial level because of its relevance to international trade.

Data and Methodology

Multi-dimensional poverty has also been investigated by Cerioli and Zani (1990), Cheli and Lemmi (1995), Betti and Verma (2008), Betti and Lemmi (2013), Betti (2017) and others employing a fuzzy-set methodology,2 but we did not find any studies that apply this method to examine China. Here, we employ a fuzzy-set methodology, in which each level of analysis, whether a household, an individual, or a province, is evaluated based on the relevant variables related to the dimensions mentioned earlier. All variables are subsequently aggregated to assess the level of multi-dimensional poverty in a province. The fuzzy-set method introduced by Zadeh (1965) is based on the generalisation of the classical concepts of sets and the membership function.3

Whereas poverty' at the individual and household level is evaluated with micro-data in the CHIP 2013, the provincial data prepared for the fuzzy- set method to measure multi-dimensional poverty are macro-data from the NBS.

At the individual level, the CHIP 2013 dataset offers information about the 61,162 individuals in the sample. Six variables are used to explore the four dimensions: two variables for education: the highest level of education completed and the years of formal education; one variable for health status; one for living standards in terms of employment status; and two variables for income: main job and other jobs.

At the household level, the 17,891 samples were assessed with 23 variables. The number of variables at the household level is higher because the dataset considers each household separately, and observation of each household member as an additional variable. Therefore, household income is the outcome of the following variables: household total disposable income for three years (2011, 2012 and 2013), total living expenditures by the household in 2013, the balance of financial assets (the total amount), spot cash, demand deposits, time deposits and the estimated net present value of fixed productive assets. Household education, in addition to the respondent’s and spouse’s education level, considers the education level of the household head’s father, mother, spouse’s father, spouse’s mother and the siblings of the respondent. Similarly, household employment considers the employment status of the respondent, the household head’s father, mother, spouse’s father, spouse’s mother and the employment status of siblings of the respondent. As for the standard of living, we rely upon subjective questions related to the living level: the variables refer to the amount of money needed to maintain the minimum living standard for the family, the evaluation of the living standards compared to the average, and whether the living standard is comfortable and evaluate economic conditions according to economic shocks.

At the provincial level, multi-dimensional poverty is evaluated by including 15 macro variables as follows: four variables for health: life expectancy, local government expenditure on medical and health care per capita, the number of medical personnel per capita, and the number of beds in healthcare institutions per capita; three variables for education: the ration of people with no schooling (people aged 6 and over with no schooling as a percentage of people aged 6 and over), the rate of illiteracy (illiterate population aged 15 and over as a percentage of the population aged 15 and over) and local government expenditure on education per capita; three variables for income: gross domestic product (GDP) per capita, national disposable income per capita and national expenditure per capita; and five variables for the standard of living: total national investment in fixed assets per capita, the rate of urban population with access to cooking gas, the rate of urban population with access to running water, the number of public toilets per 10,000 people and the number of public transport vehicles per 10,000 people.

Table 6.1 summarises how the four dimensions are evaluated at the three data levels with the relevant indicators.

For each of the k variables (with k = 6 for i = 61,162 individuals; k = 23 for i = 17,891 households; and k = 15 for i = 31 provinces) used as indicators (X,, X2 ... Xk) of the four dimensions, we applied the fuzzy-set method to calculate the deprivation measure g(xi;) of level i for the indicator j and a system of weights wv tv-, ... wk as follows:

the membership function is therefore:

Thus, we evaluate poverty at three data levels (individual, household and provincial), employing the same methodology and obtaining different results to convey the complexity of deprivation in a more detailed way.

Table 6.1 Variables for each dimension (code or measurement in parentheses)



6 at the individual level

23 at the household level

IS at the provincial level


Highest level of education completed (a 13_1)

Years of formal education (a 13_2)

Household education level (highest for up to six people in the family: a 13 1, h03 1, h03 2, h03 3, h03 4, i05_l)

Rate of no schooling at six years old (%)

Illiteracy rate at 15 years old (%)

Public expenditure on education (RMB)


Health conditions (a 16_1)

Household subjective self- evaluation of four items (p02, p06 1, p07 1, p07_2)

Life expectancv (age in 2010)

Government health-care expenditure (RMB) Medical personnel (number) Beds in health-care institutions (number)


Income from main job (C05_l) Income from other jobs (cl 1_4)

Household income (the indicator is composed of nine variables: fOl 1, fOl 2, fOl 3, f02 1, f03, f03 1, f03 2, f03 3, f07_l)

Per capita GDP (RMB) National disposable income per capita (RMB)

Per capita expenditure (RMB)



Employment status (c03_l)


employment (for up to six family members: сОЗ 1, h04 1, h04 2, h04 3, h04 4, Ю6 1)

Total national investment in fixed assets per capita (RMB)

Rate of urban population with access to cooking gas (%)

Rate of urban population with access to running water (%)

Number of public toilets per 10,000 people (%)

Number of public transport vehicles per 10,000 people (%)


The variables are used to measure individual poverty and household poverty in terms of the dimensions of education, health, living standards and income and to aggregate them at the provincial level.

We defined someone as poor in education if the person has ‘no schooling’ or ‘no more than two years of formal education’; someone as poor in health who is in a ‘very poor health condition’ based on self-assessment; someone as poor in employment who is a ‘family worker’; and someone as poor in income who has ‘less than 75% of the average income’. Then we calculated the weight of every sample of poverty according to the Neperian logarithm and average value. In the end, we determined a total multi-dimensional value for every person, and the final value was the fuzzy value that summed up all the samples for each province. A higher value meant the province had higher poverty. We processed provincial data with the same method.

Based on the individual, household and macro-data, we calculated multi-dimensional poverty (MP) values for each province to represent provincial multi-dimensional poverty. The results are presented in Table 6.2. Specifically, an MP rank of 1 means the province is the richest in multidimensional terms, whereas a rank of 14 means the poorest. To facilitate understanding, we ranked the provinces by their import rank in reverse order, with the top import provinces at the top of the table.

Table 6.2 shows the multi-dimensional poverty values calculated as explained above at three data levels as well as the value of provincial imports and exports; the rankings are portrayed in Figures 6.3 - 6.5. They show, with few exceptions, clustering of the highest rankings and of the lowest. The poorest provinces, in multi-dimensional terms (e.g. Gansu and Yunnan), are also the provinces with the lowest imports and exports and the richest provinces have the highest (e.g. Guangdong, Jiangsu, Shandong and Beijing).

Figure 6.4 shows that Chinese provinces’ imports and exports tend to have a similar rank.

Figure 6.5 shows the individual, household and provincial ranking of multi-dimensional poverty using a fuzzy methodology.

The Effect of Multi-Dimensional Poverty on Imports and Exports

Poverty not only has a direct impact on production and consumption but also has a profound impact on trade. At the individual and the household level, poor people lack production materials and the production activities that can be carried out tend to depend on their physical ability. They engage in low-tech activities or unskilled manual labour, which are unlikely to be in a trade-oriented industry, which contributes to exports. The poor tend to have low consumption levels, mainly local low-quality products to meet subsistence needs.

Countries in poverty tend to export mostly primary goods in exchange for importing industrial manufactured products. This trade pattern creates a series of dilemmas in the global market (Williamson, 2011), including deindustrialisation, the Dutch disease, commodity price volatility and rising

Table 6.2 Value and the MP rank by individual data (AC), household data (FP), macro-data (MC), imports and exports











































































































































































Source: Authors’ calculations based on CHIP 2013 for multi-dimensional poverty and NBS for trade.

The MP rank and imports and exports

Figure 6.3 The MP rank and imports and exports.

The rank of imports and exports

Figure 6.4 The rank of imports and exports.

The rank of MP

Figure 6.5 The rank of MP.

inequality, thus creating a poverty trap. The impact of living standards on production and consumption is unclear. Most research on living standards shows a significant degree of deprivation.

If we relax our hypothesis, regardless of the standard of living, we can understand that the various sub-variables of multi-dimensional poverty have an impact on production and consumption, so exports, as production surplus, and imports, as import demand, may also be affected by them. This represents only a kind of transmission, but it shows no strong link. Therefore, we cannot directly conclude whether the volume of import and export trade in a province with multi-dimensional poverty will increase or decrease. So we relax the assumption of the impact of these indicators on imports and exports and assume that if more production is produced, more surplus production is produced, and more exports are transported. The stronger domestic demand is, the greater import demand it is, and the more imports are required. In this way, we can provide a perspective with a cross-sectional data in discussing the volume of import and export trade in provinces with different levels of multi-dimensional poverty.

Concluding Remarks

We selected four dimensions based on MPI, employed CHIP data for 2013, and then measured multi-dimensional poverty using fuzzy-set theory. The results show that the top five exporters and importers in China, except Guangdong, are multi-dimensionally rich provinces. If we ignore the potential mutual causation, we could simply conclude that multi-dimensionally rich provinces have access to importing and exporting and get benefit on the top seat, even they do not rank first, and have the highest volume of imports and exports. Provinces at the mid-level in imports and exports have mid-level values of multi-dimensional poverty. Yunnan and Gansu show high levels of multi-dimensional poverty, and they are much less involved in imports and exports.

Based on the results, we do not conclude that multi-dimensional poverty reduces participation in imports and exports, and neither of them has a positive effect on multi-dimensional poverty. To some extent, they have a correlation, and the relationship should be tested with a regression model, even with panel data.

Multi-dimensional poverty was calculated based on a fuzzy-set at the provincial level. Although in recent years China has made a tremendous and successful effort to reduce poverty, it has not been eradicated, especially in rural areas. Because different observations affect production and consumption in different ways, in this chapter we assessed provincial-level poverty with three data levels of analysis, at the individual level, the household level and the provincial level. The micro-data, both individual and household data, were aggregated into a measurement of provincial multi-dimensional poverty.

We found some correlations between multi-dimensional poverty and importing/exporting, although we compared them only for 2013, which is the most recent CHIP data available. The multi-dimensionally poor provinces export and import less, while multi-dimensionally rich provinces export and import more. That means that the poor contribute less in production and consumption; therefore, they import and export less. Because the micro-data on which multi-dimensional poverty is based are cross-sectional, they do not survey all the provinces, and the CHIP data are not annual, so we could not determine whether a causal relationship exists between multidimensional poverty and importing/exporting.


  • 1 In China three versions of poverty line were subsequently devised: (1) 1978 Poverty Standard (with 85% meant to be enough to buy 2,100 calories), (2) 2008 Poverty Standard (with 60% for 2,100 calories and 40% for some items of daily use) and (3) 2010 Poverty Standard (with 2,100 calories including 60 g protein, nine years of compulsory education, essential health care, basic housing). The World Bank’s international poverty line was SI a day in 1990, $1.08 in 1993, $1.25 in 2005 and SI.90 in 2015. The exchange rate per dollar was RMB 1.744 in 1990, RMB 2.271 in 1993, RMB 3.737 in 2005 and RMB 4.319 in 2015.
  • 2 For an updated review of the method and its applications, see Betti, D’Agostino, Gagliardi and Giusti (2020).
  • 3 In classical set theory, the membership of elements in a set is assessed in binary terms according to a crispy dichotomic condition: an element either belongs or does not belong to the set. By contrast, fuzzy-set theory permits the gradual assessment of the membership of elements in a set; this is described with the aid of a membership function valued in the real unit interval [0,1]. Fuzzy sets generalise classical sets, since the indicator functions of classical sets are special cases of the membership functions of fuzzy sets, if the latter take only values of 0 or 1.


Alkire, $., Foster, J.E. (2011). Counting and multidimensional poverty measures.

Journal of Public Economics, 95(7-8), 476-487.

Alkire, S., Shen, Y. (2017). Exploring multidimensional poverty in China: 2010 to 2014. Research on Economic Inequality, 25, 161-228.

Betti, G. (2017). What impact has the economic crisis had on quality of life in Europe? A multidimensional and fuzzy approach. Quality & Quantity, 51(1), 351-364.

Betti, G., D’Agostino, A., Gagliardi, F., Giusti, C. (2020). The integrated fuzzy and relative index for poverty analysis: A review of applications in the social sciences. Estudios de Economia Aplicada, 38(1), 1-14.

Betti, G., Lemmi, A. eds. (2013). Poverty and Social Exclusion: New Methods of Analysis. London: Routledge.

Betti, G., Verma, V. (2008). Fuzzy measures of the incidence of relative poverty and deprivation: A multi-dimensional perspective. Statistical Methods and Applications, 12(2), 225-250.

Cerioli, S., Zani, S. (1990). A fuzzy approach to the measurement of poverty. In Dagum, C. Zenga, M. (eds.), Income and Wealth Distribution, Inequality and Poverty Studies in Contemporary Economics, 272-284. Berlin: Springer.

Cheli, B., Lemmi, A. (1995). Totally fuzzy and relative approach to the multidimensional analysis of poverty. Economic Notes, 24, 115-134.

Ferreira, F., Jolliffe, D.M., Prydz, E.B. (2015). The international poverty line was raised to $1.90 a day, but global poverty is basically unchanged. How is that possible? line-has-just-been-raised-190-day-global-poverty-basically-unchanged-how-ev en/.

Gao, Y.Y. (2012). The multidimensional poverty in urban and rural China: Measurement and comparison. Statistical Research, 29(11), 61-66.

Guo, J.Y., Wu, G.B. (2012). Jiyu butong zibiao ji quanzhong xuanzede duowei pinkun celiang: yi Shanxisheng pinkunxian weili [Multidimensional poverty measurement based on different indicators and weight selection: A case study of impoverished counties in Shanxi Province]. Chinese Rural Economy, 2, 12-20.

Hagenaars, A. (1987). A class of poverty indices. International Economic Review, 28(3), 583-607.

National Bureau of Statistics (2016). Poverty Monitoring Report of Rural China. Beijing: China Statistics Press.

Rowntree, B.S. (1901). Poverty, A Study of Town Life. London: Macmillan.

Sen, A. (2000). A decade of human development. Journal of Human Development and Capabilities, 1(1), 17-23.

Sen, A. (2004a). Elements of a theory of human rights. Philosophy and Public Affairs, 32(4), 315-356.

Sen, A. (2004b). Dialogue capabilities, lists, and public reason: Continuing the conversation. Feminist Economics, 10(3), 77-80.

Wang, X.L., Alkire, S. (2009). Zhongguo duowei pinkun zeliang: guji he zhengce hanyi [The measurement of multidimensional poverty in China: Estimation and policy implications], Chinese Rural Economy, 12, 4-10.

Williamson, J. G. (2011). Trade and Poverty: When the Third World Fell Behind. Cambridge: MIT Press.

World Bank. (1990). World Development Report 1990: Poverty. New York: Oxford University Press.

Zadeh, L.A. (1965). Fuzzy sets. Information and Control, 8(3), 338-353.

7 JRR Variance Estimates for

< Prev   CONTENTS   Source   Next >