Jump to ContentJump to Main Navigation
Social Security in Developing Countries$

Ehtisham Ahmad, Jean Drèze, John Hills, and Amartya Sen

Print publication date: 1991

Print ISBN-13: 9780198233008

Published to Oxford Scholarship Online: September 2011

DOI: 10.1093/acprof:oso/9780198233008.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 28 October 2020

Social Security in China: A Historical Perspective

Social Security in China: A Historical Perspective

(p.246) (p.247) 6 Social Security in China: A Historical Perspective*
Social Security in Developing Countries

Ehtisham Ahmad

Athar Hussain

Oxford University Press

Abstract and Keywords

This chapter provides a schematic historical background, relevant for the present purposes, and presents an overview of social security in China. It discusses the details of the social-security programmes and the relevant features of the economy with reference to rural areas and urban areas. The chapter notes that when analysing the Chinese social-security system, it is important to bear in mind that the Chinese economy has undergone a series of revolutionary transformations since 1949. It emphasizes that, in China, the pervasive State control of industry and fundamental changes in the land tenure have made it possible for measures for preventing and alleviating deprivation to be built into the fabric of the economy to a degree that simply would not be feasible in an evolutionary framework.

Keywords:   historical background, social security, China, rural areas, urban areas, revolutionary transformations, industry, land tenure, deprivation, evolutionary framework

1. Overview

We take social security to cover various social mechanisms for preventing contingencies such as interruption of earnings, sickness, and crop failures, or alleviating their effects, and for protecting vulnerable social groups such as the elderly and the chronically poor. These social mechanisms need not take the form of special programmes, but instead may be embedded in the economic organization or social relations. This is of special importance in the case of developing economies, because unlike developed economies they often lack comprehensive social-security programmes, and may not be able to afford them in the foreseeable future. Thus the effects of the organization of the economy on the incidence and alleviation of deprivation assumes a central importance in the discussion of social security in developing economies.

When analysing the Chinese social-security system, it is important to bear in mind that the Chinese economy has undergone a series of revolutionary transformations since 1949. The development of social-security programmes has been intertwined with changes in the pattern of ownership and State control of the economy. In China, the pervasive State control of industry and fundamental changes in the land tenure have made it possible for measures for preventing and alleviating deprivation to be built into the fabric of the economy to a degree that simply would not be feasible in an evolutionary framework. Thus, apart from discussing social-security programmes to cater for contingencies and particular categories of persons, we devote considerable space to the features of the rural and the urban economy which have a direct bearing on social security.

We begin in Section 1.1 with a schematic historical background relevant for the present purposes, and then go on to present an overview of social security. The details of the social-security programmes and the relevant features of the (p.248) economy are discussed in Section 2 with reference to rural areas, and in Section 3 with reference to urban areas. We end in Section 4 with a conclusion.

1.1. Historical Background

Schematically, Chinese economic history since 1949 may be divided into three phases marked by the following events:


Main features


Consolidation of social and political control; formation of a planned economy; land reforms; industrialization on the Soviet model.


1958–60: the Great Leap Forward, involving rural collectivization, infrastructural works and rural industrialization;

1958–62: the great famine;

1966–76: the Cultural Revolutionary Period with its emphasis on equality and the collective economy.


Reversion to family farming; removal of restrictions on house-hold enterprise; granting of financial and operational autonomy to industrial enterprises; promotion of mixed ownership; shift away from egalitarianism; erosion of some collective welfare arrangements.

The economic reforms since 1978 have spanned the rural and urban economies, and the most far-reaching of the transformations have been the shift to family farming and the disbanding of collective agriculture that had been the central feature of the Chinese economy since 1958. Agricultural land remains under collective ownership but has been leased out to households under what is termed the ‘Responsibility System’. In contrast to agricultural reforms, the industrial or urban reforms started later and have been comparatively modest as yet. As in East European economies, their central feature has been the granting of financial and operational autonomy to State-owned enterprises. But unlike them, China has gone further towards adopting the principle of diversified ownership, allowing State, collective (co-operative), private, and foreign-owned enterprises to proliferate side by side. Freed of many of the stifling restrictions, household economic activities have flourished. A notable feature of industrial development in recent years is the exceptionally fast growth of rural industry, owned collectively by rural townships and villages. Thus in many areas the collective economy still remains strong, with the difference that it is now anchored in industry rather than in agriculture. (p.249)

The economic reforms have had a dramatic effect on incomes and consumption. During the nine years of reforms (1978–87), per capita income more than doubled (ZTN 1988: 799). The increase in per capita consumption expenditure during this period, which is a more accurate index of improvement in the living standard, has been even more striking: it exceeded the total increase over the previous twenty-five years from 1952 to 1977 (ZTN 1988: 801). However, the increase in incomes and consumption has been unevenly spread among different regions and occupational groups. The rural areas along the densely populated coastal rim in the east and in the vicinity of large cities have experienced a higher increase than the rest. And in urban areas, there have been significant changes in relative incomes. The incomes of those engaged in the production of goods and services for sale and of the self-employed have risen faster than the incomes of groups engaged in providing non-marketed goods and services such as education, health, and public administration. Lately with the acceleration in the rate of inflation from 1987, a substantial proportion of the urban population has not merely fallen behind in the race for personal betterment, but suffered a decrease in real incomes. With the unfolding of the economic reforms, regional disparities in rural areas, which had always been substantial, and an increase in income inequality in urban areas, which used to be very low, have come to the fore as major social issues in the Chinese economy.

Turning to a brief history of social security in China, in the 1950s the government formally guaranteed to its citizens provisions for basic needs. And it has reiterated the commitment in successive constitutions. More important, preventing deprivation has been a permanent concern of policy, notwithstanding the changes in leadership and abrupt shifts in the organization of the economy. Generally speaking, this concern has taken two forms: first, building protective measures in land tenure and employment practices, and, second, instituting social-security programmes to cater for specific contingencies and for the protection of vulnerable groups. Preventing rural households from falling into destitution has been a continuous strand running through the transformations in the rural economy from the land reforms, which began in the 1940s, to the shift to household farming from 1979. Similarly, the prevention of mass unemployment in urban areas has remained a permanent preoccupation of the government. The assumption of State control of industry in the 1950s went together with the introduction of an ambitious system of labour insurance, covering health care and old-age pensions, for the employees of the government and State-owned enterprises. Since then the coverage of labour insurance has widened with the increase in employment in the State-owned sector and the government (on the history of labour insurance, see Hussain and Liu 1989: sec. 6).

With the nationalization of trade in 1953, the government assumed an obligation to provide rations of staple items at low prices in urban areas, and to (p.250) redistribute grain from surplus to deficit areas in the countryside. Both these still hold true. The 1950s saw the enunciation of the health charter, which gave primacy to preventive over curative health care, and the development of government-financed or supported hospitals and health centres. In the field of primary health care, the most important event was the spread of health insurance to over three-quarters of the rural population following Mao’s instruction to put stress on rural areas (Mao 1974: 232–3). Equally significant has been an almost total dismantling of the collective rural health insurance systems set up during the 1960s and the 1970s in the wake of the decollectiv-ization of the rural economy. However, a large part of the urban labour-force and their dependents still benefit from partly or fully paid health care through labour insurance.

In comparison with other developing economies, China has been notably successful in reducing the threat of starvation and destitution to its citizens (World Bank 1985:16–19). This success, however, is not unblemished. As the Chinese authorities themselves admit, the natural calamities of 1959–61, combined with the upheavals of the Great Leap Forward, condemned tens of millions to an early death. The estimates of excess deaths vary from ten to thirty million (see Coale 1984; Ashton et al 1984; Peng 1987; see also the population figures in the Statistical Year Books). Liu Shaoqi, who was an object of popular wrath during the Cultural Revolution, remarked that the Great Leap claimed more lives than the building of the Great Wall. Pockets of transitory or permanent poverty have always existed in China, particularly in rural areas, and do so even now despite a massive increase in farm output and rural incomes since 1978. Only a few years ago Deng Xiaoping candidly admitted that around 100 million people or about 10 per cent of the population still did not have enough to eat and wear (Beijing Review, 5, Jan. 1985).

China is a low-income economy, yet in terms of welfare indicators it is akin to low or high ‘middle-income’ economies. The per capita intakes of principal nutrients in China markedly exceed those common in low-income economies, attaining the levels common in middle-income economies. The crude death rate, except for a sharp peak during the famine years of 1959–61, declined steadily from 20 per 1,000 in 1949 to6.3in 1978. Ithas since then risen due mainly to an increase in the percentage of the elderly in the population (see Hussain and Stern 1988). Life expectancy rose from 40 years in 1953 to 68 years in 1981. Adjusting for the underrecording of deaths, the figure for the 1980s is around 65 (SYB 1986: 91; Banister 1987: 116). Among low-income countries only Sri Lanka matches the Chinese level (see WDR 1989). According to a UN estimate, China’s achievement in increasing life expectancy at birth since 1949 is matched by a few small countries such as Taiwan, Chile, and Sri Lanka (see UN 1984: 126–45; Banister and Preston 1981; and World Bank 1985: 174–5). This achievement is all the more impressive because a sustained gain, in life expectancy is more difficult to achieve in a large heterogeneous country such as (p.251) China than in a small country such as Sri Lanka. The high life expectancy is corroborated by the cause-pattern of mortality: diseases related to old age rank high and infectious and parasitic diseases, which are the hallmarks of poverty, rank low—again a pattern which is similar to that of middle-income economies (World Bank 1984:13). Thus in its income class, China is an ‘outlier’ in terms, of the average level of nutrition, life expectancy at birth, and the cause pattern of mortality. This raises the central question: why has China been more successful than many other developing economies?

The steady gain in life expectancy coupled with low birth rates since the early 1970s has hastened China’s transition to the ‘tertiary demographic’ phase of low birth and life expectancy. This has brought in its train a rising percentage of the elderly in the population—a trend expected to gather pace in the coming decades. In 1986, there were around 57.5 million individuals over 64 (5.43 per cent of the population), and their numbers are expected to rise by 64 per cent to 94.5 million (around 8 per cent of the population) in the year 2000, soaring to over 200 million by 2030 (World Bank 1985: 138–9). The ageing of the population has been quicker in China than in European countries or even in Japan (see Hussain and Liu 1989: sec. 4.7). As a result, China faces a ‘premature’ problem of supporting the elderly similar in magnitude to that in middle-income countries. As we shall see, how to provide for a relatively high percentage of the elderly, when per capita income is still low, is a central problem of social security in China.

1.2. The Social-Security System

The important point about the Chinese social-security system is that it operates in an economy which has built into it the following two features:

  1. (1) all rural households have had access to agricultural land, either individually or collectively;

  2. (2) recruitment and employment practices have been geared to maintaining a high rate of employment in urban areas.

Since the extension of the land reforms to the whole of mainland China in the early 1950s, all rural families have had guaranteed access to agricultural land either individually or collectively, enabling them to earn a living by deploying their labour. Details are discussed in Section 2 on rural areas. Turning to urban areas, most of the employed labour-force in China has been in government departments, and in State-owned and collective (co-operative) industry. Labour recruitment, especially in government departments and State-owned enterprises has been geared to providing employment to new entrants to the permanent urban labour-force. Life-long employment, which still remains a prevalent norm in China, has helped to keep unemployment in check by ruling out spells in between jobs. Broadly speaking, the urban population has (p.252) benefited from, as it were, a comprehensive employment guarantee. This guarantee has in recent years begun to erode at the edges. Formally, the government no longer promises jobs to all new entrants to the labour force with urban registration (hukou). More important, the massive influx of rural, inhabitants has swelled the ranks of urban residents without urban registration, who are outside the umbrella of employment guarantee altogether. Nevertheless, as we elaborate later in Section 3, a large majority of the urban labour-force still benefits from some form of guaranteed employment.

The guaranteed access to land in rural areas and the employment guarantee in urban areas have together provided the Chinese population with a degree of economic security which is rare in developing economies. Although they are not components of the social-security system, they have had a strong bearing on it. They limit the numbers in need of assistance and reduce the contingencies for which social-security programmes have to cater. They have had two striking effects that mark out China from other developing economies. The first is the absence of landless poor, though not of poor with land. The second is the absence of large visible unemployment, though not of ‘disguised unemployment’ in China’s urban areas. However, this has begun to change rapidly with the increase in immigrant workers from the countryside, who are exposed to a much greater risk of unemployment than permanent urban residents and are not covered by urban social-security programmes. Notwithstanding this change, the map of poverty for China still looks very different from those for other developing economies. Urban destitution and squalor is much less visible, and extreme poverty is still a predominantly rural phenomenon, confined mostly to remote and sparsely populated areas. Arguably, there may be less poverty in China, but whatever is there is often not immediately visible.

In China, greater economic security goes in tandem with a web of controls on personal conduct. The guaranteed access to land in rural areas and the employment guarantee have rested on stringent restrictions on population migration. The control on migration from the countryside to cities has served to keep in check the urban labour-force and thus the numbers benefiting from the employment guarantee. The guaranteed access to land has not been unqualified, but has taken the form of the permanent residents of a village having access to the village land. Normally, ‘outsiders’ have had no right of access, but until recently their numbers have been very small because of the low rate of population migration within the countryside. The point which needs emphasizing is that the two guarantees have been premissed upon a very limited migration of the population. Their coverage has begun to erode, as the loosening of controls on movement swell the numbers of ‘temporary’ residents who are outside the umbrella of the two guarantees.

In China changes in the place of residence have been rationed through a system of household registration (hukou), which we discuss further in Section (p.253) 3. In particular, rural inhabitants are not allowed to settle permanently in cities without permission, which for a big city may be especially hard to obtain. Since the 1950s, successive Chinese leaderships have sought to encourage the development of industry in rural areas, and thereby accomplish the usual historical process of the shifting of labour from farming to other activities without a large-scale migration from urban to rural areas. ‘Leave the land but not the countryside’ (Li tu bu li xiang), as the slogan goes. Until the early 1980s, household registration in combination with restrictions on travel and a diverse array of administrative controls did succeed in keeping the proportion of urban population remarkably low. Thus, as compared to other economies, China has exhibited an anomalous combination of a high share of industrial output in national income and a low proportion of the urban population in the total (for a discussion see Perkins 1986: pt. 2). With the unfolding of the economic reforms, whilst household registration remains as an instrument for controlling migration, many of the other economic impediments to migration have withered away. And the government has come to accept the inevitability of an increase in urbanization with economic development, and has considerably relaxed the issue of urban registration for small or medium-size towns, but not for large cities such as Beijing, Shanghai, and Guangzhou. Increasingly, rural inhabitants are leaving not only the land but also the countryside for urban areas, including large cities—li tuye li xiang. The result is a massive increase in that proportion of the population of large cities who are not recognized as permanent urban residents, presenting the government with huge problems of social control and deprivation.

The Chinese population is marked by two forms of division which bear upon the social-security system. The first is the distinction between rural and urban residents, which has existed since the 1950s. And the second, applying to the urban population, is the distinction between permanent residents—namely, those with registration for the locality—and temporary residents. This distinction is assuming greater importance with the increase in population migration. Urban registration confers a privileged economic status, providing the holder with possibilities for a comparatively higher living standard. Rural inhabitants have guaranteed access to land, but, as economic opportunities vary widely with the locality, this does not rule out poverty. In some areas it may mean no more than a precarious existence between poverty and distress. Many of China’s large cities, in particular Guangzhou, are flooded with large numbers of itinerants who have no permanent abode and no regular source of income. Recent years have also seen a significant increase in migration within the countryside. Migration poses a special problem for a social-security system which still remains premissed on the assumption that the population will be immobile and has no special provisions for migrants either in cities or in the countryside.

China has a complex collection of disparate programmes that may be (p.254) grouped under the general rubric of social-security system. Social-security cover provided to citizens varies according to place of residence and employment status, which are closely related. Most of these social-security programmes date from the 1950s, but have undergone transformations with the twists and turns of government policies. And they are surrounded by a formidable array of social and economic imperatives for self-sustenance. Principal among these have been participation in labour up to physical capacity, and obligation to provide assistance to kin in need, especially to parents in old age. Means-tested public assistance has been seen and projected as the safety net of last resort. Notwithstanding a permanent concern with basic needs on the part of successive Chinese leaderships, the discussion of social security from a comprehensive perspective is relatively recent, dating only from the period of economic reforms since 1978. We provide an analytical account of the system in terms of the agencies responsible for social security and then turn to the social-security programmes.

1.2.1. Main Agencies

Generally speaking, the responsibility for social-security programmes is diffuse, divided between five agencies: the family, the government, work units (danwei), collective (or communal) organizations in rural areas, and ‘mass organizations’. Collective rural organizations date back to the collectivization of agriculture in the 1950s and still survive following the shift from collective to household farming, albeit radically transformed in status and functioning. They perform governmental functions, yet are not regarded as part of the formal governmental apparatus. Mass organizations include inter alia the ‘All China Federation of Trades Unions’, the ‘Women’s Association’, and neighbourhood organizations, which are concerned with the welfare of their respective constituencies. The last, specifically Chinese institutions, are urban organizations covering residential neighbourhoods. They perform a variety of social-welfare and policing functions, and are supervised by the Ministry of Civil Affairs.

Three features characterize the institutional structure of the Chinese social-security system. First, the public social-security system, which excludes family support, is tiered. The first tier consists of rural collective organizations, work units (danwei), and neighbourhood organizations. These grass-roots bodies are the proximate sources of funds and organize and disburse assistance. The second tier consists of the government agencies which supervise the organizations in the first tier, and provide them with funds when they are unable to meet their obligations or cope with certain contingencies. The second feature of the structure is that it is ‘cellular’. The organization and the financing of the various social-security programmes are divided among the large number of units in the first tier—in the main, among work units (danwei) in urban areas and among collective organizations in rural areas. Thus relative to other developing economies China has a surfeit of agencies concerned with (p.255) social welfare. The cellular structure goes together with a wide variation in the levels and methods of provision. This is especially pronounced in the case of rural social-security programmes, but also holds for urban programmes as well. Within the government itself, responsibility is split between the provincial governments and various ministries of the central government. The third significant feature of the Chinese social-security system is the important role played by collective institutions and mass organizations in the provision of social security, especially in rural areas. We would suggest that China’s relative success in developing rural social-security programmes is due in large measure not to the government but to collective organizations.

As in other developing economies, family and kin support remains a significant source of protection against destitution for the population at large. All the same, what is special about China is that the government has sought to incorporate family and kinship support as a basic component of social-security arrangements. The obligation of children to support their old parents is written into the constitution, which the 1980 Family Law has prudently extended to include grandparents as well. Authorities take pains to make sure that discretionary public support does not displace but merely supplements family support. Local government and collective institutions, especially in rural areas, police the discharge of filial obligations. Reliance on social relief, especially what is termed wubao, is projected as a shaming last resort, branding both the recipients and their immediate and extended families with social stigma (see Wolf 1985: 196, 199, 248). The family is, and is likely to remain for the foreseeable future, the mainstay of old-age support in rural areas. Generally speaking, urban residents get old-age pensions and rural residents rely on sons for old-age support.

The family may not, however, continue to be as important a source of old-age support as it has been, especially in rural areas. Notwithstanding the long tradition of filial piety in China, a variety of factors point to a gradual diminution in the role of the family. The support of the elderly by their children, although not conditional on, has rested on multi-generational households. An increasing number of the young prefer nuclear households, and the shortage of housing in urban areas and the household registration system limit the formation of multi-generational households. Added to these, the demographic changes in the offing will increase the burden of old-age support on families. With an increase in life expectancy, at a mature age in particular, each succeeding generation of the elderly will have to be supported for a longer period than the preceding one. Further, the one-child policy promoted by the government does not fit in with heavy reliance on support of the elderly by their families, which the government also encourages. As most rural couples do not have old-age pension but only social relief—or wubao—to look forward to, they have an incentive to produce more than one child in order to make sure that they have at least one male offspring. Conversely, if the one-child policy (p.256) does succeed, which it has in urban areas, then the ensuing change may simply make it impossible for many families to support their elderly (see Hussain and Liu 1989: introduction).

1.2.2. Social-Security Programmes

In China there are four programmes which fit in with the conventional definition of a social-security system:

  1. (1) labour insurance (laodong baoxian);

  2. (2) occupational and communal provisions—often termed in China social welfare (shehui full), a term which we use subsequently;

  3. (3) social relief (shehui jiuji) and disaster relief (ziran zaihai jiuji);

  4. (4) public provisions for health care.

We provide an outline of these programmes in this section, leaving details for the later sections. Labour insurance extends only to the employed labour-force, excluding the peasants altogether, because they are regarded as self-employed. Moreover, it covers only a percentage of the formally employed labour-force, consisting of government employees and employees of State-owned enterprises and larger collective enterprises. The employees of smaller collective enterprises in urban areas and of rural township and village enterprises are outside the umbrella of labour insurance. In recent years, the percentage of the labour-force covered by labour insurance has remained fairly steady around 23 per cent (ZSTZ 1987: 111). Promising disability and old-age pensions, maternity and sickħess benefits, and subsidized health care, labour insurance is akin to schemes in Welfare States, except that unemployment benefit has not been its main component. For unemployment has been regarded as a problem to be solved by job creation. However, the government has finally come to recognize that job creation may lead to disguised unemployment with a heavy cost to the economy. It now accepts that at least some unemployment is inevitable in a socialist market economy, and has created an unemployment insurance scheme. However, the scheme is, as we shall see in Section 3, still severely restricted in scope. As almost all organizations under the umbrella of labour insurance are in urban areas, and a large percentage of the urban labour-force is employed in such organizations, labour insurance is broadly synonymous with urban-cum-industrial social security. This is, however, subject to the important qualification that the percentage of the urban labour-force outside the umbrella of labour insurance has been increasing in recent years. This includes a proportion of the urban labour-force and the growing numbers of rural immigrant workers in cities. The rapidly increasing labour-force of rural industry is excluded from labour insurance altogether. Unless checked, this trend would turn labour insurance into a selective scheme, applying only to a proportion of the employed labour-force. Details of labour insurance are discussed in Section 3.

What is descriptively termed here ‘occupational and communal welfare (p.257) provisions’—social welfare (shehui fuli) in China—covers a wide range, including inflation, hardship and bereavement allowances, social facilities such as kindergartens, educational and recreational facilities, and institutions for orphans, the mentally ill, and the disabled. In sum, it is a miscellaneous category covering diverse income-maintenance provisions and social facilities at the work unit or in residential areas, villages included. In urban areas, it supplements labour insurance, with which it is often grouped together.

Social relief (shehui jiuji) covers for the most part assistance to the rural poor, including to the elderly without relatives to depend on, and is termed ‘5 Guarantees’ (wubao). Broadly, social relief is synonymous with ruralcum-agricultural social security. Unlike labour insurance, it does not cover contingencies as a matter of right, it is discretionary. The eligibility for social relief is highly restrictive: not only should the recipient be non-self-subsistent, but also have no relatives to depend on. In organization, it varies between private charity and public philanthropy. Unlike labour insurance, it lacks an underlay of laws and regulations specifying entitlements. Further details of social relief are discussed in Section 2 on rural areas. Disaster relief (ziran zaihai jiuji) is of special importance for rural areas, as around 70 per cent of the population lives in rural areas, and over 60 per cent of the labour-force is still employed in agriculture. Vast and varied as China is, natural disasters are recurrent contingencies in rural areas.

Public provision for health care is not unified, it consists of two main components. The first is government provision in the form of a network of hospitals and preventive health services. And the second is the coverage of health care costs through health insurance, predominantly through labour insurance. Until the shift to the Responsibility System, there used to be an extensive rural health insurance system, but that has all but disappeared. Health care costs of those who are neither covered by health insurance nor able to pay for themselves are covered as part of social welfare (shehui fuli) and social relief (shehui jiuji). Because health care is a large topic in need of a separate treatment, we confine the discussion to the selective features of health care in rural areas (see Section 2).

Besides these four arrangements, which would normally be grouped under the social-security system, we would also include the pricing and trading policies of the government. These have an important bearing on personal real incomes in both rural and urban areas. Price controls and the rationing of necessities play an important part in income maintenance in many developing economies, but their importance is all the greater in the Chinese economy because of the pervasive State control of prices and trading. Notwithstanding the reduction in government control of the economy since 1978, the control of prices and trading by national and local government is still more extensive in China than in most developing economies. Government agencies remain the largest purchasers of agricultural commodities, and a ration of grain and (p.258) cooking oil at low prices is provided to all permanent, but not to temporary urban residents. Added to that, municipal governments in urban areas often provide low-price rations of ‘non-staples’ such as vegetables and meat. The cash income of farmers depends crucially on the purchase prices paid by the government, and the real incomes of urban consumers depend heavily on government-determined price and on the rationing of consumer goods. We discuss the expenditure on price subsidies further in this section.

1.2.3. Social-Security Expenditure (SSE)

The economic importance of social security is usually expressed in terms of the total expenditure on social-security programmes and its ratio to national income. The time series on the total expenditure are available only for the period since 1979 (Table 6.1), and do not include government expenditure on health, because in the government accounts it is grouped together with culture, education, and science in one category.

Table 6.1. Social security expenditure (SSE): 1979–1985


SSE as % of national income

Labour insurance as % of SSE

Social relief as % of SSE

Social welfare as % of SSE



45 —

12 —

43 —



52 (32)a

9 (−20)

39 (4)



55 (15)

9 (9)

36 (2)



57 (16)

8 (1)

35 (7)



58 (18)

7 (6)

35 (18)



59 (18)

7 (10)

34 (12)



60 (29)

6 (11)

34 (27)

(a) Figures in parentheses represent percentage growth rate at current prices.

Source: ZSTZ (1987: 123).

As a percentage of national income, the expenditure on social security is comparatively high in China, and has risen since 1979, with the start of the economic reforms. The increase of 0.7 percentage points over 6 years to 1985, as shown by Table 6.1, is due to the dramatic rise in expenditure on labour insurance at 21 per cent a year, 60 per cent higher than the exceptionally rapid rate of growth of 13 per cent per year in national income at current prices—9 per cent per annum at constant prices. As may be seen from Table 6.2, the other two components of social-security expenditure, social welfare and social relief, grew more slowly than national income. The steep rise in labour insurance expenditure is not due to an increase in the proportion of the labour-force covered by labour insurance, which in recent years has remained constant at around 23 per cent. In fact, as we point out in Section 3, the proportion of the employed labour-force covered by labour insurance has in (p.259) fact declined. Rather the increase in expenditure is largely the result of an increase in the number of the retired entitled to pensions. The wide variation in annual growth rates of labour insurance expenditure reflects the fact that the number of pensioners increased by highly uneven steps. In particular, the massive growth rate of 32 per cent in 1980 was caused by the bunched retirement of large numbers who should have retired earlier but did not do so because of the Cultural Revolution.

Table 6.2. Average annual rates of growth: 1979–1985 (%)a

National income

Total social-security expenditure

Labour insurance

Social relief

Social welfare

Price level







(a) All rates of growth refer to magnitude in current prices.

Source: ZSTZ (1987: 123) the inflation figure is from ZTN (1988: 778).

In contrast to labour insurance, expenditure on social relief in nominal terms increased relatively little over the period. As may be seen from Table 6.1, the expenditure fell sharply in 1980 with the start of the rural reforms. The decrease came immediately after a massive rise in the procurement prices of agriculture commodities in 1979, leading to large increase in rural incomes (TPS 1987: 115). Moreover, the annual expenditure on social relief during the three years between 1980 and 1983 remained lower than it had been in 1979, and began to rise only from 1984. In fact, taking into account that the rate of inflation averaged 4 per cent per annum, expenditure on social relief fell in real terms. The main reason for the low rate of increase of 2 per cent in social-relief expenditure in nominal terms over the six years to 1985 is simply that social relief is subject to a stringent means test, and is only meant to alleviate extreme poverty. Prima facie, the sharp increase in rural incomes after 1979 may have reduced the number qualifying for relief.

In sum, social-security expenditure has risen sharply since 1979, with an increasing proportion of it going to labour insurance. The change is adverse from the point of view of equity, as labour insurance covers barely a quarter of the labour-force. Further, the fact that the percentage of the labour-force with labour insurance cover has remained constant over the six years to 1985 despite a steep increase in expenditure makes it doubtful that the percentage would rise in the near future. We discuss this in Section 3. The general implication is that for financial reasons, at least, the Chinese social-security system is likely to remain segmented, with a comprehensive labour insurance for a small proportion of the labour-force, almost all in urban areas, and a stringently means-tested social relief for the rest.

(p.260) 1.2.4. Financing of Social Security and Price Subsidies

The rapid growth in social-security expenditure, especially labour insurance expenditure, raises the question of how it is financed. Generally speaking, all but a small percentage of social-security expenditure is financed from current revenues of work units and rural collective organizations. As we shall see in Section 3, this is now a major source of problems facing the labour insurance system. As yet personal contributions do not play an important part in the financing of social security. A proportion of the employees of collective enterprises and the employees of State-owned enterprises recruited since 1986, termed ‘contract employees’, contribute towards their labour insurance, but they still constitute a small proportion of the employed labour-force. Thus an overwhelming proportion of social security benefits are non-contributory. Given the mounting cost of labour insurance and the increase in personal incomes, the argument for personal contributions towards labour insurance is strong. This would not only help to diversify the sources of finance, but may make it feasible to extend labour insurance cover to a larger percentage of the formally employed labour-force. Apart from a limited introduction of personal contributions on the part of contract employees, there is as yet no plan to tap personal contributions as a source of finance for labour insurance. Rather the policy seems to be to encourage the growth of private or contributory group insurance for employees outside the umbrella labour insurance, which adds to the segmentation of the social-security system.

Schematically, the three main social-security programmes are financed as follows:

Labour insurance Work units

(danwei) and the government;

Social welfare

Work units, collective organizations and the government;

Social relief

Collective organizations and the government.

To get an idea of the relative importance of different sources of social-security expenditure we look at the government’s share in total expenditure over the period 1979–85 (Table 6.3). Two features stand out in this table. First, looking at the composition of government expenditure as a whole, the share going to social security is remarkably small. In fact, since 1952 the share has usually been below 2 per cent (see ZTN 1988: 758). Second, the government share in total expenditure on social security, which was relatively low to start with, has fallen rapidly to just over half of its original proportion. Over the period from 1979–1985, whilst total social-security expenditure in nominal terms rose by an average of almost 16 per cent per year, government expenditure on social security, also in nominal terms, increased at a comparatively low average rate of 6 per cent per year, although government expenditure in general rose by an annual average of 9 per cent. Thus not only is most of social-security expenditure (p.261) in China financed by work units and rural collectives; but also their share in the total has been increasing.

Table 6.3. Government expenditure on social securitya








Government expenditure as % of total spent on social security








Government expenditure on social security as % of total government Expenditure







(a) All figures refer to nominal magnitudes.

Source: ZSTZ (1987: 123) ZTN (1988: 747, 761).

At first sight both these features seem unusual, because a high and increasing expenditure on social security relative to national income usually goes hand in hand with a high and increasing proportion of government expenditure devoted to social security. The main reason for the anomaly is that what would elsewhere be financed by general or social-security taxes is in China financed largely from the current revenue of a vast number of work units. Prior to the reforms, the difference between government and non-government financing was purely nominal, as the budgets of enterprises were not separated from the government budget. But now that they are, the difference matters, and the requirement that each enterprise is responsible for financing the labour insurance of its employees on an individual basis has important implications for maintaining labour insurance cover. For the labour insurance liabilities of an enterprise bear no particular relation to its financial capacity to meet those liabilities. This is a source of problems for labour insurance, which we discuss in Section 3. The current system of financing social security in China is highly decentralized, even in comparison to systems functioning under a federal government. Thus the stylized fact that decentralized financing of social security goes in tandem with variations in the level of provisions holds with even greater force in the Chinese context. We discuss specific examples of variations in Sections 2 and 3 with respect to rural and urban social-security programmes.

If government expenditure on social-security programmes is low, price subsidies from the government budget are exceptionally high, as shown by Table 6.4. These figures underestimate price subsidies because they do not include municipal government subsidies on ‘non-staples’ such as meat, vegetables, sugar, and eggs. The fact that a high proportion of total subsidies goes (p.262) to ‘consumer goods’ is due to the procurement prices of agricultural goods being typically higher than the sale price. Chief among these are grain and cooking oil which are supplied at subsidized prices to all permanent residents of urban areas and some rural residents. Therefore a substantial part of subsidies on consumer goods is tied to the rationing of consumer goods in urban areas. Most of the subsidy on agricultural inputs goes on fertilizers, although this has been reduced along with the increase in the procurement prices of agricultural goods. The subsidy on imported goods arises out of the fact that domestic prices of grain, fertilizers, sugar, and agricultural chemicals have been lower than border prices at the official rate of exchange. Generally speaking, most price subsidies are directed towards urban inhabitants. The main problem with these subsidies is, however, not their urban bias, but that such a large proportion of government expenditure is devoted to a non-targeted programme.

Table 6.4. Price subsidies by central government

Composition of subsidies (%)


Subsidies as % of government expenditure

Consumer goods

Agricultural inputs

Imported goods














































Source: ZTN (1988: 763).

In the next two sections we discuss social security in rural areas and urban areas respectively. In China the distinction between rural and urban is administrative, and in many instances at variance with an economic classification. Some of the areas classified as rural are better regarded as urban from the point of view of population density and of the relative importance of non-agricultural income. This is especially true of the regions where rural industry has grown rapidly in recent years. The arbitrariness of the category may be seen from the fact that following an administrative reclassification, the percentage of the rural population as a proportion of the total decreased from 77 per cent in 1983 to 68 per cent in 1984 (ZTN 1988: 97). Nevertheless, the (p.263) distinction remains important in any discussion of social security in China because residential status plays a central role in the organization of the social-security system.

2. Social Security in Rural Areas

Earlier we pointed out that the rural social-security programmes operate in an environment where all rural families have guaranteed access to land. So we shall begin by looking briefly at the institutional transformations in rural areas since 1949, and then turn to other issues in the following order:

  1. (1) agricultural pricing and trading policies of the government;

  2. (2) social relief (shehui jiuji) and disaster relief (ziran zaihai jiuji);

  3. (3) health provisions.

The rural social-security system consists of the programmes under headings 2 and 3. Whilst social relief includes various income maintenance programmes, disaster relief and health provisions cover specific contingencies. Agricultural pricing and trading are included because of the important role they have played in determining real personal incomes and alleviating nutritional contingencies. Although China has been comparatively successful in preventing and alleviating deprivation in rural areas, the rural social-security system is, as we shall see, strikingly sparse. Social relief is contingent on stringent means-testing, which is often concerned more with preventing the non-deserving from claiming relief than with ensuring a comprehensive coverage of all those in need. Although social-security programmes in rural areas are by no means significant, on their own they do not seem weighty enough to account for China’s success in preventing deprivation in rural areas. Thus we would suggest that the success is due in large measure to the organization of the rural economy.

2.1. Transformations in the Rural Economy

Since 1949, the rural economy has undergone three fundamental transformations: first, land reform, which began before 1949; second, formation of rural cooperatives and then rural communes from 1956; and third reversion to family farming from 1979. For our purpose, the land reform has been the most fundamental of the transformations in that it laid the foundations of the income-generating mechanism in farming and rural areas generally. The shift back to family farming since 1979 has reversed collective farming but not the essential feature of the land reform: guaranteeing all rural households access to land. In fact, land is now more ‘equally’ distributed than it was on the eve of collectivization in the mid-1950s. We take a brief look at the three transformations in turn. (p.264)

Given that 89.4 per cent of the Chinese population in 1949 was rural, the land reform left a strong imprint on income distribution (SYB 1983: 104). A fundamental aim of the reform was to provide each rural household with at least enough land and the wherewithal to meet basic needs; and its most important feature was the redistribution of land to the poor peasants. To assess the exact impact of the land reform one has to turn to the specific form it took. First, given that the reform was in effect a war waged by poor peasants against landlords and rich peasants, the possibilities of evading the transfer of land were non-existent. Second, the range of beneficiaries was wide, encompassing ‘middle’ as well as poor peasants. Third, the reform concerned not only land but also agricultural implements. Yet there were severe limits to what could be achieved by these measures. Improvements in the standard of living of the poor were limited by land-to-labour ratios and the level of agricultural technology. As in other East Asian economies, the ratio of cultivable land to rural population in China was, at the time of the land reforms in the early 1950s, exceptionally low by international standards. Since then the ratio has steadily decreased due to the increase in the agricultural labour-force. But the rate of growth of agricultural productivity in China seems to have been higher than in most other developing economies (see Hussain 1989). Besides, as elsewhere, land reform in China took the form of local redistribution of land, usually within the boundaries of the village (for a discussion of the geographical scope of land redistribution see Schurman 1971: chap. 7). As it did not involve a redistribution of rural population, the land-to-population ratio at village level set the upper limit on minimum landholding. Although the land reform did lead to a more equitable distribution of land within a village, it could not address the problem of poverty on any broader scale. In particular, it left untouched the important problem of regional poverty, which has always existed in rural China, but has become more striking because of the rapid but uneven increase in rural incomes since 1978.

Land reform had a profound and still persisting effect on the characteristic attributes of needy households in rural China. It eliminated landlessness as a cause of destitution. Herein lies a crucial difference between the attributes of the rural poor in China and those in most other developing economies, where rural poverty and landlessness often go together. There have been, since the land reform, landless households in rural China, and their numbers have grown rapidly in recent years with the economic reforms. But ‘specialized households’, as they are termed now, are not poor; on the contrary, they are landless by choice and are more likely to be among the more prosperous households.

Broadly speaking, poverty in rural China is associated with two groups of factors. The first consists of natural circumstances such as poor soil, inhospitable climate, and geographical isolation. These are the main attributes of poor regions, or causes of a low average income over a whole area. The second arises (p.265) from lack of sufficient labour power due to age, death, and a large number of children, and it is causes such as these which determine whether any one household will be poorer than the rest in the area.

The land reform dovetailed into the formation of co-operatives and shortly after into full-scale collectivization of rural economic activities. In retrospect, the co-operatives were no more than a preliminary stage of the rural commune. The history of collectivization is turbulent, and the flux of institutional changes is too complex for a short account. We confine ourselves to a minimal sketch of the rural commune and its effects relevant for the present purpose.

The rural commune combined local government and the organization of economic activities encompassing not only farming but also infrastructural investment and rural industry. Collectivization of land went in tandem with restrictions on household economic activities and on migration, tethering rural inhabitants to their place of birth. The leadership saw collectivization as an organizational device by which to mobilize surplus rural labour for investment in the infrastructure and to transfer labour from farming to non-farming activities without geographical migration. The low mobility of the population, which has characterized the Chinese economy until recently and on which the Chinese social-security system is still premissed, could in large measure be attributed to rural collectivization. In geographical terms rural collectivization covered a much wider area than land redistribution in that it bound together a large number of villages—‘big in size and collective in operation’ (yida ergong), as described by a common slogan. To convey an idea of its scale, after 1962 the rural commune, for the most part, consisted of three tiers: the production team (around 30 households), the production brigade (around 250 households), and the commune (around 3,000 households) (see SYB 1981: 134). Treating the production brigade as equivalent to a natural village, the rural commune encompassed around twelve villages, on average. This is no more than a very rough estimate because the size of villages varies widely.

Broadly speaking, collectivization was marked by two features: first, the centralized allocation of labour and, second, the collective disposal of income. Households were left with little discretion over the deployment of their own labour. In effect, the collectivization of land also implied a collectivization of rural labour. The distinguishing feature of centralized deployment of labour was the heavy allocation of labour for infrastructural investment. Aside from a relatively small income from heavily constrained private activities, all income accrued initially to the collective. This gave the collective power to decide upon the disposal of income in cash and kind, which was done according partly to need and partly to work (see Parish and Whyte 1978: chap. 5; Croll 1982; Perkins and Yusuf 1984: chap. 5; Endicott 1988: 123–6). Remuneration for labour and income maintenance were intertwined; the relative weight assigned to each varied across units. Distribution according to needs reduced the risk of loss of income to a household from the diminution of its labour force. To use an (p.266) adage common in China, rural collectivization provided each rural family with an iron rice bowl—the bowl symbolizes provision and iron the fact that the level of provision remains unaffected by everyday knocks and bumps, or, in other words, variations in labour performed by a household.

The adverse effect of collectivization on incentives has been emphasized in the literature, but it should also be borne in mind that the iron rice bowl provided a guaranteed income not only to shirkers but also to households lacking sufficient labour due to sickness, old age, and deaths. Under collectivization, the financing of social security, rural health insurance, and infra-structural work posed no special problem. Local cadres decided upon the uses of income and the deployment of resources with or without the consent of households. However, coercion was not the sole prop of the system. Collective cultivation did engender a spirit of solidarity conducive to social security through local co-operation, albeit not to the same degree everywhere. Besides, the fact that income accrued to the collective in the first instance had a special significance in terms of insurance: that is, any diminution of income or damage to assets due to natural factors was automatically borne by all and not just a few households. By its very nature collectivization involved a pooling of risks; and as with most insurance systems, it went together with adverse effects on incentives.

A major aim behind the formation of the rural commune was to mobilize labour for infrastructural work during slack seasons. During the period of collective agriculture, there was a massive investment of resources and labour in infrastructure. Heroism rather than economic calculations guided much of this, and in terms of consumption forgone and return on investment, the economic propriety of many of the schemes undertaken is open to question. We would, however, single out the massive investment in hydraulic construction as having a special significance for income maintenance in rural areas. At the end of the era of collective farming 45 per cent of the cultivable area in China was irrigated—an exceptionally high figure for a large and diverse country (see World Bank 1983: ii. 23–4). Much of this hydraulic construction was undertaken during collectivization, involving a massive mobilization of labour, and bears the hallmarks of bursts of heroic but ill-planned effort. Notwithstanding this, the high proportion of irrigated area prima facie did help both to increase and to stabilize yields.

Collective agriculture is now heavily discredited in China, and the speed with which it was displaced by family farming suggests that many of its features were strongly resented by the rural population. Assessed from the vantage point of what has happened in the rural economy since decollectivization, there is little doubt that the collective rural economy stunted incentives, misallocated resources, and held back personal incomes by putting household economic activities in a strait-jacket of collectivism. Yet it is also true that the shift to family farming has adversely affected communal social-security arrangements (p.267) and the investment in and maintenance of the infrastructure, particularly irrigation systems. We turn now to the reversion to family farming.

Since 1979, the collectively owned land and productive assets of the production team have been parcelled out to households under the Responsibility System, and most of the restrictions on non-farming activities by rural households have been lifted. The pattern of distribution of land varies geographically, and so too do the terms on which the land has been parcelled out. Predominantly, the land has been assigned to households in proportion to their size. Notwithstanding regional variations, implicit in the distribution of cultivable land there remains a concern to provide each household with the means to make a living. The living which households are able to make depends crucially on the locality. Since the land is still dejure collectively owned, there exists a possibility of changing the distribution of land to take account of any movement of population from agricultural to non-agricultural activities, and of demographic changes. (For examples of readjustments in the distribution of land, see AYB 1986: 40–1.) Households in receipt of collectively owned land have to pay land tax, local accumulation and welfare tax, and are obliged to sell to the government a portion of the produce. Taxes and the procurement quota obligation are divided pro rata to the land area. In effect, households are subject to a lump-sum tax. However, things are neither as orderly nor as uniform as this sparse description might suggest. The shift to family farming has cut the ground from underneath the local government in rural areas, which was intertwined with collective organization of economic activities. Generally speaking, the reorganization of local government has lagged behind economic changes. As a result, there are problems in collecting accumulation and welfare taxes and in organizing public activities, at least in some rural areas. Under the Responsibility System diversity in the countryside is much greater than it was under collectivism.

Turning to its effects on social security, the Responsibility System has undone many of the features associated with collectivization. Granting each household the power to dispose of any income generated from the plot allocated to it does not necessarily accord with notions of distribution according to needs, which is now limited by the size of the local welfare tax. This is intentional, since the aim has been to ‘smash the iron rice bowl’. Ipso facto, the household has to bear the risk of any loss in income or assets due to a diminution of its labour-force, and to damage from natural factors. The general point is that the Responsibility System has increased incentives, but it has also increased the risks to be borne individually by rural households. The financing of social-security programmes can no longer be decided by an administrative fiat; it depends on the willingness of households to contribute, and on the welfare tax the rural administration manages finally to collect. The parcelling out of land and restoring to households the power to allocate their own labour have unravelled the arrangements built around the collective (p.268) deployment of income and resources. In particular, the rural health insurance system has, in most areas, been replaced with ‘payment according to treatment’. This means that in case of illness a household not only faces the risk of loss of income but also of larger expenditure. The Responsibility System has also decreased the ability of rural collectives to mobilize labour for infrastructual works. In fact, the government has explicitly forbidden the use of labour and resources without adequate compensation (for a case-study see Zweig 1989: chap. 7). The irrigated area declined steadily from 1978 to 1985, though since then the government has succeeded in partially reversing the trend (see ZTN 1988: 233). From the point of view of social security, the displacement of the collective economy by the household economy has to be coupled with a growth in either tax-financed social insurance or private insurance to cover the extra risks created by this displacement, but that has yet to happen.

When assessing the impact of the Responsibility System on rural social security, one should bear in mind that rural incomes have risen at a record pace since the introduction of the system. Between 1978 and 1987, rural income per capita at current prices grew by 13 per cent annually (ZTN 1988: 799). Given that the rate of inflation over the period has averaged around 6 per cent per annum, the rate of growth of real incomes still comes to a high 7 per cent. By doing away with restrictions on private activities, the economic reforms have provided rural households with wider and more varied opportunities for earning an income. Such opportunities, however, vary greatly with the locality. Added to that, the economic reforms have given a fillip to rural industry. Employment in rural industry (township and village industry) occasionally brings with it occupational benefits such as old-age pensions and subsidized medical care, though not on the same scale as in urban areas. And, since most of that industry is collectively owned, a portion of its profits is used for subsidizing agriculture and financing rural social welfare. Rural industry’s contribution to rural welfare in nominal terms has been growing at 10 per cent per year over the ten years 1978–87 (see YRSES 1986: 168; ZTN 1988: 287). However, rural industry is very unevenly distributed. For example, in 1987 around 60 per cent of the total national profits from rural industry accrued to only six out of twenty-nine provinces, containing 28 per cent of the total rural population (ZTN 1988: 289; SYB 1987: 69). All these provinces are along the eastern seaboard of the country.

2.2. Pricing and Trading Policies of the Government

Although land tenure is the most important determinant of personal incomes in rural areas, trading arrangements have had a significant bearing on rural incomes. A part of the agricultural produce is sold to the government, voluntarily or not. In addition, farmers purchase agricultural inputs such as fertilizers, manufactured consumer goods, and also, in the case of a shortfall, (p.269) staples such as grain. As government agencies have been by far the most important buyers of agricultural produce and the sellers of inputs to farmers, the pricing and trading policies of the government have had an important effect on the cash income of farmers. These policies have frequently been used for income maintenance in rural areas. The welfare impact of the pricing and trading policies of the government on rural areas is a large subject. Here we concentrate on two issues: first, the role of the State marketing network in alleviating nutritional contingencies in rural areas, and, second, the implications of the pricing policy for agricultural commodities as far as personal incomes are concerned.

We analyse the first issue in terms of how much grain is purchased from rural areas and the geographical pattern of sale of the purchased grain. On average, the government has purchased around a quarter of grain output, and resold around a third of its purchases back to rural areas—termed ‘return sales’ (fan xiao). Both these proportions have gone up since 1978, reflecting an increase in specialization in cropping patterns (see TPS 1987: 57). Until 1978, under the regime of ‘taking grain as the key link’ each rural locality, except a few permitted to specialize in cash crops, was expected to devote enough of the cultivated area to grain to ensure self-sufficiency. With the relaxation, but not the disappearance, of the regime, there has been an increase in specialization, hence the increase in the proportion of return sales in recent years. The time series of return sales make it clear that the State marketing network has not been used merely to siphon grain from rural to urban areas. It has also performed the important function of redistributing grain within rural areas. During the period of ‘taking grain as the key link’, prima facie, an overwhelming proportion of grain resold in rural areas was designed to alleviate nutritional contingency. Although the grain-rationing system does not apply to rural areas, there has been in operation a system of guaranteed grain. All rural inhabitants are supposed to have at least a certain amount of grain for their own consumption. The minimum level, which has varied historically and regionally, is taken into account in setting procurement quotas. Villages (or production teams or brigades) falling short of the minimum level are supplied with grain from the State marketing network to make up the shortfall. The terms of supply are determined by a means test. Poor areas or areas struck with natural disasters are supplied with grain either free or on loan (for details see Hussain and Feuchtwang 1988: 64–70). The provision of grain is the most important component of social relief (shehui jiuji) in rural areas. In general outline, the system has not changed during the period of reforms, except that with the increase in personal incomes in rural areas, the government is less willing to supply grain on concessional terms.

Turning to government pricing policy, since 1978 there has been a substantial increase in procurement prices for agricultural produce. Specifically, in the nine years between 1978 and 1986, the index of purchase prices of agricultural (p.270) commodities rose by 77 per cent, which exceeds the total increase in the twenty-six years between 1952 and 1977 (TPS 1987: 115). In addition, the government has allowed farmers to sell any output above the procurement quota in private markets, where prices are higher than the procurement prices. The emergence of free markets means that the government cannot set both the purchase price and the procurement target, as it did earlier. The increase in procurement prices has been an important source of the increase in personal income in rural areas since 1978. However, the effect of procurement prices on the cash incomes of the farmers is complicated by changes in the prices farmers pay for their input. In recent years the government has been cutting subsidies on agricultural inputs (see Table 6.3 above; for a discussion see Sicular 1988). The net effect of government pricing policy on the distribution of incomes in rural areas remains an issue for research.

2.3. Rural Social Security

For the purposes of discussion here we divide social-security arrangements in the countryside into two functional categories:

  1. (1) measures to maintain income and promote the development of poor regions;

  2. (2) public provisions for health, including co-operative rural insurance.

2.3.1. Income Maintenance and Income Promotion

This category covers three programmes: first, social relief (shehui jiuji), including wubao; second, disaster relief (ziran zaihai jiuji); and third, development assistance to poor regions. For reasons of space, we restrict ourselves to the first two, whose main purpose is to provide a safety net to stop the rural population from slipping into the depths of deprivation. Social relief is means-tested, and not meant to eliminate poverty in the sense of bringing everyone up to an adequate level. Thus, as we shall see, many who are officially recognized to be poor may not receive any assistance. Social relief is for the most part financed by the government. Rural collectives usually finance only that part of social relief which is provided to wubao households (explained below); and some also have special programmes to help poor households in their locality. Such programmes vary greatly between localities, and depend crucially on the initiative of the local leadership. It is these local programmes which have been affected by the recent changes in the rural economy—adversely in some cases, and favourably in others. However, social relief provided by the government seems to be better organized in the post-reform than in the pre-reform period. Before going into the details of the scheme, we turn first to disaster relief.

In contrast to social relief, which is targeted towards the very poor, disaster relief is a much larger programme providing cover to the whole of the rural (p.271) population against severe damage from natural causes. To give an idea of the relative sizes of the two programmes, in the four years 1983–7 between 70 and 80 per cent of the total expenditure on social relief and disaster relief has been devoted to disaster relief (see NTN 1986: 296; NTN 1987: 274). Around a half of disaster relief is provided to households in the form of living allowances consisting mainly of grain. Thus a significant part of the resale of grain to the countryside (fan xiao) consists of grain provided to households in disaster-affected areas. The eligibility for disaster relief has traditionally been assessed in terms of the shortfall in grain per capita from a minimum level. The usual practice has been to provide relief to all households in eligible areas regardless of means. In recent years, the government has tried to shift the emphasis of the programme from assistance after the incidence of disaster towards the prevention of damage from disaster, and there have been increasing criticisms of the ‘grain test’ to assess the extent of damage and eligibility for relief. The rural reforms have rendered the ‘grain standard’ increasingly obsolete. The cropping pattern has shifted away from grain, and an increasing proportion of income is derived from non-farming activities which are less affected by natural disasters. There is a gradual, but still not a complete, switch from the ‘grain standard’ to the ‘monetary standard’ based on money income from all activities (for further discussion see Hussain and Liu 1989: introduction and sec. 3).

What is the extent of protection provided to the rural population against natural disasters, which are among the main risks it faces? When assessing the extent of protection one needs to take into account not merely disaster relief provided by the government, which covers only severe damage, but also self-protection by rural communities. For rural inhabitants are expected to bear at least a part of the damage from natural disasters. As pointed out earlier, collective cultivation involved a pooling of risks from crop loss. With the parcelling out of land under the Responsibility System each household has to bear the risk of damage to crops on its land plot individually, which increases the risk of loss of income facing households. Thus from the point of view of social security against natural hazards, the growth of the household economy should be coupled with the growth of insurance, either private or social, to cover the extra risk, but this has yet to happen on a significant scale.

Returning to social relief, this covers assistance to two types of households, as distinguished in the Chinese literature: 5-Guarantees households (wubao hu) and poor households (pinkun hu or kunnan hu). Wubao, which is a term dating from the 1950s, refers to the guarantee of food, health care, shelter, clothing, and funeral costs for all citizens, hence the term 5 Guarantees. However, over time the term has come to designate a special category of households or persons unable to earn a living and lacking relatives to depend on. Thus, as we shall see, the recipients of wubao relief are predominantly the elderly without family support. Although narrowly focused on people without support and available to a relatively small number of people, wubao has a special significance in that it (p.272) is linked to the tradition of looking after parents in old age. Wubao is seen as an obligation of rural collectives. In recent years, this obligation has been cited in birth control propaganda to demonstrate that couples without sons need not fear for support in old age, as the collective will be there ready to look after them. In contrast, poor households (pinkun hu) constitute a wider category, and although distributed over the whole countryside, they are much more prevalent in underdeveloped areas, often with poor soil and an inhospitable climate.

The details of social relief for recent years are given in Table 6.5. The figures in column 1 refer to the recipients of relief, predominantly in kind; they do not include the recipients of assistance in the form of employment or production subsidies. For a low-income country, the percentage of the rural population receiving assistance seems to be strikingly small, and, more important, so too is the percentage of the poor in the rural population. As shown by column 3, only a fraction of those officially recognized as poor receive social relief, and this fraction has been falling since 1983 for reasons explained below. Yet the percentage of the recipients of social relief in the rural population has been rising due to the decrease in the rural population. The percentage of the poor, including both the recipients and the non-recipients, in the rural population, as shown by column 4, is an indirect estimate derived from the percentage of the rural poor receiving social relief. Considering the rapid and widespread rise in personal incomes over this period, there is no plausible explanation for this apparent rise in poverty other than a better identification of the poor due to a greater sensitivity to rural poverty on the part of the government. This brings (p.273) home the point that the existing figures for the rural poor in China have to be treated with great caution because to our knowledge there is as yet no systematic attempt at estimation.

Table 6.5. The recipients of social relief


No. of recipientsa

% of rural population

% of poor assistedb

% of the poor in rural populationc




































(a) The figures for the recipients are in million, those for 1978–84 are from YRSES (1986: 259, and 264); the rest are from ZTN (1988).

(b) As provided in NTN (1986: 294), the figure for 1986 is from NTN (1987: 272).

(c) As implied by the percentage of the poor who are assisted.

The decrease in the percentage of the recognized poor receiving social relief shown by column 3, we would suggest, does not represent an erosion of social security, but a steady shift from consumption subsidies to production subsidies and public works programmes. The recipients of the former are counted under social relief but the latter are not. In recent years, there has been a proliferation of employment provision schemes in rural areas. The numbers of such schemes increased from 20,000 in 1980 to 50,000 in 1986 (NTN 1986: 294; NTN 1987: 252). Such schemes have been commonly used in developing economies such as India to alleviate deprivation in rural areas. But until recently China has not used public works schemes for this purpose. This may seem surprising because China has long been famous for massive mobilizations of rural labour for infrastructural works. The reason for this apprent paradox lies in the method of mobilization of labour and resources for capital projects during the era of collectivism.

During the period of collective cultivation, the labour and resources used for infrastructural projects were frequently not remunerated at all. This resembled the traditional Chinese practice of requiring rural communities to supply labour and resources for the construction and maintenance of monuments and economic infrastructure. On occasions when labour and resources were paid for, more often than not the remuneration came from collective rather than government funds. Infrastructural works were seen and commended by the leadership as attempts by rural communities to lift themselves from poverty through their own efforts. They were not used by the government for making income transfers to poor areas. For raising current consumption in deprived areas, the government tended to use the State marketing system to supply grain and agricultural inputs on concessional terms. There has been an important change in the government policy towards infrastructural works in rural areas. One of the first acts of the reform leadership in China was to ban unrewarded exactions of labour and resources by rural collectives (for a case-study see Zweig 1989: chap. 7). This, together with the change in the ideological climate, has cut the ground from underneath revolutionary mobilizations of labour for collective ends. A result of this is the neglect of irrigation works to which we referred earlier. Now that the deployment of labour has been devolved from collectives to households, relieving rural poverty by providing employment on government-funded public works schemes is gaining in favour.

Broadly speaking, there has been an important difference between the methods for relieving rural poverty in China and in India. The Chinese government with its tight control of trade in agricultural produce and inputs, and with its relatively underdeveloped public finance system has tended to relym (p.274) more on assistance through commerce than on assistance from the government budget. In contrast, the Indian government with its more developed public finance system and its looser hold over trade in agricultural produce and inputs has tended to rely more on budgetary assistance than on assistance through commerce. This contrast still holds but not as starkly as it did in the pre-reform period. As the system of public finance still remains comparatively underdeveloped, the Chinese government is naturally disposed towards using trade as a policy instrument for a wide variety of purposes.

Turning to the details of wubao relief, Table 6.6 presents the available figures. As indicated earlier, the recipients of wubao relief are predominantly the elderly without relatives to depend on. For example, in 1985–6, they constituted on average 83 per cent of the recipients; the rest were orphans and the disabled (NTN 1987: 266). In effect, wubao is a support scheme for indigent elderly in rural areas. What proportion of the elderly in rural areas receive wubao? Taking the 1982 population census figure for over-65s in the countryside and assuming that 83 per cent of the recipients of wubao relief are elderly people, the estimate for 1982 comes to just over 3 per cent. The proportion for recent years would be lower because, whilst the numbers of over-65s has increased, the number of recipients has since 1982 remained nearly constant. This percentage may seem derisory by the norms of a Welfare State, but it is exceptional for a developing economy. In most developing economies, public provisions for the indigent elderly in rural areas simply do not exist. Turning to column 2, there has been a sharp decrease in the percentage of wubao elderly supported by collective institutions. According to circumstantial evidence, this represents a centralization of wubao relief away from villages, which are collective institutions, to townships, the lowest tier of the government in rural areas (for case-studies see Hussain and Liu 1989: sec. 3).

Table 6.6. Recipients of Wubao relief


Persons in millions

% supported by collectives

























Sources: STZ (1987: 120) NTN (1987: 266) the figures are rounded.

(p.275) The essential feature of the wubao system is that it is local, founded upon solidarity in rural communities. The levels and methods of provision have not been uniform. They have varied greatly both in time and space, as do the criteria of eligibility for assistance. Wubao relief, although grounded in the constitution, lacks a detailed legal and regulatory framework specifying the sources of funds, the range of coverage, the rights of recipients, and the obligations of collectives. In terms of organization, it ranges between charitable assistance, which carries a social stigma, and a formal social-security programme with codified details of provision. Wubao relief is predominantly nonresidential: the recipients live on their own or with relatives. However, an increasing proportion of them are accommodated in ‘Houses of Respect’ (jinglao yuan), and they receive a higher level of support than those living at home. Some of the richer areas have gone a step further and have introduced old-age pensions for farmers. As yet such schemes are restricted to a few prosperous rural areas (for case-studies see Hussain and Liu 1989: introduction and sec. l).

In recent years wubao relief has come under strain from two sources. Prior to the reforms, it was organized at the level of the village (production team) and was intertwined with collective cultivation. There is evidence that, following the introduction of the Responsibility System, it has been neglected or in some rural areas has broken down completely (for examples see World Bank 1985: 30, 92–3, and 165; Bernstein 1985; Hussain and Liu 1989: sec. 3). There is a trend towards the township government taking over the responsibility for wubao relief, which in future may become a norm for the whole of the countryside and provide the basis for a formalization of the system.

The second source of strain arises from the changing age structure of the population and the birth control policy. As pointed out earlier, the percentage of the elderly in the population has been increasing. Added to that, the percentage of the elderly is higher in the rural than in the urban areas (Hussain and Liu 1989: sec. 4.2). The discrepancy between rural and urban areas is likely to increase with additional migration from the countryside, which occurs mostly among the young. This raises a doubt whether the wubao system, informally organized as it is, will be able to cope with the mounting burden implied by demographic trends. Wubao support is used to persuade couples to forgo children. Yet, as presently organized, it does not have much power of persuasion. On the contrary, it provides an incentive to breach the one-child policy. For it is strictly restricted to the elderly without children or near relatives to depend on. In most areas, wubao is designed more to prevent the ‘non-deserving’ elderly from claiming relief than to ensure that all the ‘deserving’ elderly in fact get relief. For example, the property of wubao recipients is inherited by the collective not by their relatives. Rural authorities take measures to prevent families from transferring the burden of supporting their elderly to the collective. And wubao is intended as no more than a safety net (p.276) against destitution. For couples of child-bearing age, wubao is an inferior alternative to having a son who would support them in old age. The general point is that a scheme which is expressly designed for the elderly without relatives cannot appeal to couples who have the option of depending on their children in old age. This is increasingly realized in China, and concern to decrease rural birth rates is the main source of pressure for an improvement in public support for the elderly in rural areas.

2.3.2. Public Provision for Health in Rural Areas

Aside from the land reform, which guaranteed access to land to the whole of the rural population, the extension of health care to a vast majority of the rural population has been a major achievement of the Chinese economy in the field of social security. Here we confine ourselves to two aspects, which have been distinguishing marks of health provision in rural China. The first is the emphasis on public hygiene and preventive care and the second is the provision of primary medical care at the grass-roots level in the 1960s and 1970s.

Historically, the promotion of public hygiene and preventive health care has relied heavily on mass campaigns. The biggest of these was the ‘Great Patriotic Health Campaign’ of the 1950s which was aimed against four pests: flies, mosquitoes, rats, and (mistakenly) sparrows. Since then there have been many other campaigns directed against opium addiction, a legacy of colonialism, and specific diseases such as smallpox, venereal diseases, typhus, plague, and schistosomiasis, a severe debilitating disease spread by snails which was once endemic in the Yangtze basin (for details of campaigns see a personal account in Horn 1971). Such health campaigns involved promoting hygiene, eliminating disease ‘vectors’, and mass screening to identify and treat the afflicted. They relied on elementary medical precepts and owed their effectiveness to the mobilization of the population on an epic scale. Thanks to them, within two decades (1950–70), China managed to reduce drastically, if not to eliminate, the incidence of common parasitic and infectious diseases. A testimony of their success is provided by the massive national mortality survey of 1973–5, which revealed that the cause pattern of mortality in China was closer to that of middle-income than of low-income economies, where parasitic and infectious diseases rank high among the causes of death (on the details of the mortality survey, see Banister 1987: 96–8). In form, health campaigns were similar to campaigns in other fields such as those for investment in infrastructure or politics, all involving a massive mobilization of the population through a mixture of propaganda and coercion, which the collective rural economy facilitated. From an economic point of view, their main feature was their very low financial cost, because in most cases labour was not remunerated at all, or, on the few occasions when it was, rates were very low.

The decollectivization of the rural economy has, as it were, weakened the social immune system (for an account see Endicott 1988: chap. 12). For (p.277) example, schistosomiasis, which had been eradicated by the 1970s, has now reappeared, and the percentage of inoculated infants decreased in the 1980s. Health campaigns have continued, but they have neither the same appeal nor the same effect as they did in the pre-reform period. The economics of campaigns has changed radically. With the shift to household economy, the opportunity cost of labour to rural households has increased; neither are they willing nor can they be easily coerced into performing free or low-paid labour for wider social ends, as they did before. The basic problem is that campaigns can no longer be sustained by propaganda alone, they require an adequate remuneration of labour for their success. The maintenance of preventive health care and public hygiene in rural areas requires a system of public finance which has been slow to develop. The government has taken a few steps to arrest the decline in preventive health care. For example, alarmed by the evidence of rise in infant mortality (for a discussion see Hussain and Stern 1988), the government in conjunction with UNICEF has now embarked on a national plan to achieve inoculation of at least 90 per cent of infants by 1991.

A major achievement of the period of the Cultural Revolution (1966–76) was the extension of primary medical care to villages. This involved dispatching medical personnel from cities and county towns to villages, the establishment of village health stations, training a huge number of paramedics—‘barefoot doctors’ and part-time medical personnel—and developing co-operative health insurance in villages. The policy was a success because it addressed the two issues central to the provision of primary medical care in rural areas of developing economies. The first is how to train medical personnel in sufficient numbers and, more important, to deploy them in the countryside. The second is how to finance medical care. The barefoot doctors, who became famous throughout the world, were the main medical personnel at village level. They were rural residents who had received an elementary medical training which was short, focused on a few tasks, and did not presuppose a high level of education. They worked part or full time aided by medical workers whom they trained. They dispensed basic primary health care and referred patients with severe illnesses to commune and county hospitals. Notwithstanding their name they did wear shoes, especially when performing medical tasks. The use of paramedics for primary health care is not peculiar to China; earlier the Soviet Union too relied on paramedics (known as feldshers) for the same purpose. China’s main achievement was to train a very large number within a very short period of time. For example, in 1979 there were around two barefoot doctors per 1,000 inhabitants. This had its negative side too, in that their level of training was highly variable, and many of them were entrusted with tasks well beyond their competence (see Chen 1989: pt. 2). Further, barefoot doctors could not be the permanent mainstay of primary health care in rural areas. For their training, elementary and narrowly focused as it was, ran the (p.278) risk of being made obsolete by successes in improving public hygiene and eliminating parasitic and infectious diseases.

Starting from 1965 there was a concerted attempt to develop co-operative health insurance in the countryside. On the eve of decollectivization in 1979, around 80 per cent of production brigades (the middle tier of the commune, now termed villages), or about 85 per cent of the rural population had some form of health insurance system—an unparalleled achievement for a large developing economy (see World Bank 1984). Brigade members paid an annual fee in return for the reimbursement of out-patients and hospital costs, including medicines. The extent of coverage varied widely. Contributions from brigade members only covered a part of the costs, villages (production teams and production brigades) paid the rest from collective funds. The co-operative health insurance system was founded on collective agriculture, and together with barefoot doctors it came to be closely identified with the discredited period of the Cultural Revolution. Thus it seems that the leadership made no attempt to retain the system as land was parcelled out to households. As a result, in 1985 the rural health insurance system survived in a mere 5 per cent of brigades or villages as compared to 80 per cent in 1979 (Shao 1988). The rural health insurance system did suffer from some major defects, but its almost total disappearance is due in large measure to wilful neglect by the leadership.

The Responsibility System has had a debilitating effect on the rural health care system. Not only has health insurance been replaced with payment according to treatment, including the inoculation of children, but there has also been a massive decrease in medical personnel in rural areas (for details see Hussain and Stern 1988). The government has decontrolled drug production without developing an effective system to monitor quality (see Chen 1989: pt. 2). Many village health stations are left unattended because of lack of personnel. The main reason for the decline in the number of personnel is financial. Subsidies for health care from collective funds have decreased, and the government tried to keep medical fees under a tight control, while rural incomes were rising. As a result, many barefoot doctors and other part-time medical personnel have found it more profitable to engage in non-medical pursuits.

Nevertheless, there have been some attempts to develop an alternative health care system for rural areas. Following Deng Xiaoping’s advice that barefoot doctors should start wearing straw sandals, if not leather shoes, a proportion of them have acquired further medical training and been upgraded to the status of ‘rural doctors’ who work full time. In fact, the term ‘barefoot doctors’ is no longer used. Further, the Ministry of Public Health has introduced pilot rural insurance schemes in a number of villages (for details, see Shao 1988). However, the cost per head of these schemes is much higher than that of previous schemes. While such schemes may be applicable to richer (p.279) rural areas, their introduction in poorer areas would depend crucially on subsidies from the government. Given the pressure on the government to keep its expenditure on a tight rein, comprehensive health insurance in poorer areas is a distant possibility. But one may well witness a rapid spread of health insurance in richer areas.

3. Social Security in Urban Areas

The urban counterpart of access to land for rural households is an employment guarantee for residents with urban registration. As in the previous section, we first outline the environment in which urban social-security programmes operate. For the present purposes, the two most important components of the environment are:

  1. (1) entitlements, employment and earnings;

  2. (2) household registration and rationing of consumer goods.

The urban social-security system consists largely of what is usually termed ‘occupational welfare’. That is, most urban residents receive their social-security benefits and a wide range of social provisions through their work units (ıdanwei), and not from a social-security agency. In fact, work units have a pervasive influence on every aspect of their employees’ lives. There is also a residual social-security system built around the residential neighbourhood to cater for those cases which are not covered by occupational welfare. As we shall see, unlike in rural areas, the urban social-security system is fairly elaborate.

3.1. Entitlements, Employment and Earnings

The formally employed labour force, most of which is in urban areas, is heavily segmented by differing entitlements to social security. These vary with employee status and the status of the work unit (danwei), which makes the distribution of social-security provisions in urban areas heavily dependent on the employment pattern. As we shall see, the reforms have had a profound effect on this pattern. By status, employees may be divided into permanent (guding), contract (laodong hetongzhi) and temporary (linshi) workers, each with a different entitlement to labour insurance. The permanent and contráct employees together form the regular labour-force of work units. Such or similar divisions have existed since the 1950s but have assumed greater significance in the wake of the reforms. Beside employees, there are also the self-employed, who existed in large numbers up to 1955 and became almost extinct during the Cultural Revolution period. They have proliferated since 1978 and accounted for 4 per cent of the non-agricultural labour force in 1987 (ZTN 1988: 153). Their growth, which is expected to continue apace, is of (p.280) special significance for urban social security, as the self-employed are outside the ambit of occupational welfare altogether.

Work units (danwei) are officially classified into ‘owned by the people’—the State sector—collective enterprises, and the rest—private and foreign enterprises. The first two constitute what may be termed ‘the socialized sector’, which has since the early 1950s accounted for almost all of the employed labour-force. Despite the encouragement given to private and foreign enterprises in recent years, they accounted for only 0.5 per cent of the employed labour-force in 1987 (ZTN 1988, p. 153). The term ‘socialized sector’ denotes no more than the absence of private ownership; its employees do not all have the same entitlement to labour insurance. Whilst all employees of the State sector are covered by labour insurance, either partially or fully, only two-thirds of the employees of collective enterprises are. Furthermore, the employees of collective enterprises often contribute towards their labour insurance, but a vast majority of employees of the State sector still do not.

As compared to the employees of collective enterprises, State employees have more secure jobs and higher pay on average, 28 per cent higher in 1987 (ZTN 1988: 194, 198). The higher-tier collective enterprises tend to emulate State-owned enterprises. Collective enterprises are graded into two categories according to the tier of the subprovincial government supervising them. The first consists of those supervised by the county or supra-county tiers of the government, and the second of the rest. The lower-tier includes the collective enterprises run by urban neighbourhoods and rural townships and villages. Generally speaking, the State sector is heavily regulated and is comparatively uniform in matters of pay and labour insurance. In contrast, the collective sector is highly diverse.

To avoid going into lengthy detail, we focus mainly on the State-sector employees, who at the end of 1988 constituted around 73 per cent of the formally employed labour-force (ZTYB Dec. 1988: 5). In 1978 this figure was around 79 per cent (ZLGTZ 1987: 15). Such employees are nearly all in urban areas, as rural industry is collectively rather than State-owned. For the present purposes, the State-sector employees may be divided into the following two broad categories, relating to the extent of labour insurance coverage: (p.281) Contract employees are recently introduced replacements for permanent employees, and the two together constitute the regular labour-force of work units.

% of State-sector labour-force

Fully Covered

Permanent Employees


Contract Employees


Not Fully Covered

Temporary workers


Permanent employees, who are all in the State sector, also constitute a majority of the formally employed labour-force, around 57 per cent at the end of 1988 (ZTYB 1988: 5). Until recently, they had the right to a job in their work unit for life, which frequently was taken to mean the right to occupy the same job for life. They amounted, therefore, to a fixed factor of production, as it was rare for permanent employees to change their work unit. Permanent employment for the regular labour-force has had a special significance for social security in that it restricted the incidence of unemployment to what has usually been a small minority of the urban labour force, consisting of temporary employees and new entrants, and now contract workers as well. The labour market reforms of recent years have sought to reduce the security enjoyed by permanent employees, and to encourage greater labour mobility between jobs and work units. Permanent employees may now be dismissed for persistent indiscipline and be made redundant in the event of bankruptcy, which is not possible but still rare. They no longer have the right to remain in the same job and can be redeployed in a variety of ways. They can be laid off, but only with pay. They still have a privileged status as compared to other categories of employees in that they have the right to income, if no longer to a job. In as yet infrequent cases of dismissal or redundancy due to bankruptcy, they are entitled to unemployment benefit, which, aside from contract workers, is not available to the rest of the urban labour-force. This includes temporary employees of the State sector and almost all of the collective-sector employees. And, unlike other categories of employees, they do not contribute towards their labour insurance. In principle, this is taken into account in determining their pay. The 1986 Labour Regulation put an almost total end to permanent employment by providing for all new recruits to be on fixed-term but renewable contracts. As intended, this will lead to a steady decrease in the numbers of permanent employees, as those reaching retirement will not be replaced. But it will also increase the ratio of retired to still working permanent employees. This, as we shall see, is likely to pose a special problem for the financing of pensions for permanent employees in the future.

Contract employees, or ‘labour contract system employees’ (laodong hetongzhi gong) as they are termed, have increased rapidly in numbers since the 1986 Labour Regulation. The government’s intention is that in future the regular labour-force of enterprises will consist entirely on contract employees. Employment on contract in various forms has existed since the 1950s, but until 1986 this covered only the residual category of temporary workers (for details, see Walder 1986: chap. 2; White 1987). Unlike permanent employees, contract employees contribute towards their labour insurance, in particular old-age pensions. The change-over from permanent to contract employment is meant (p.282) to promote efficiency by allowing workers greater freedom to switch jobs on the one hand, and by enabling enterprises to adjust their regular labour force on the other. However, the change-over is also likely to create new problems for social security. First of all, it introduces the possibility of recurrent large-scale unemployment arising out of spells between successive jobs, which in the past applied only to temporary workers. Second, since, as elsewhere, unemployment spells in between jobs are likely to vary with the age, sex, and qualifications of the unemployed, contract employment may lead to long-term unemployment among certain categories of urban workers such as the middle-aged, the unskilled, and women. The full consequences of contract employment will take some time to surface, because contract employees still constitute a small percentage of the regular labour-force, and the system has not been in operation long enough for many of even the initial set of contracts to expire.

Temporary employees form a residual category, including both workers with urban registration, and also growing numbers of rural immigrants with only temporary urban registration. Although temporary employment is a transitory state for both, they do not face the same range of future possibilities. For workers with urban registration, temporary employment may lead on to a regular job, preferably contract employment in the State sector. But for immigrant workers, temporary employment leads either to another temporary employment, perhaps only after a spell of unemployment, or forced return to the countryside. They have a limited prospect of graduating to regular employment, because they stand a very small chance of obtaining urban registration. Almost all immigrant workers are in unskilled and manual jobs; many are employed on building sites. Being temporary urban residents, they are not entitled to low-price rations of grain and cooking oil. They may be housed by their work units, often in makeshift single accommodation on building sites. Temporary workers in the State sector are usually better protected than temporary workers in the collective sector, in that they benefit from some of the labour insurance provisions. But they are not entitled to old-age and disability pensions.

As well as rural immigrants in temporary jobs, Chinese cities also attract from the countryside a large number of itinerant casual labourers, petty traders, female domestic servants, and people in search of a better life. They constitute a marginal population, completely outside the umbrella of urban social security. They are covered neither by occupational social welfare nor by programmes organized around urban neighbourhoods, which are restricted to permanent residents. The so-called floating population, including all types of rural immigrants, has grown rapidly in recent years and is already sizeable. For example, it is estimated that in 1985 as much as 12 per cent of Beijing’s population consisted of residents without urban registration (JPRS 1986: 48- 52). More recent figures for Beijing would be higher, and higher again for Guangzhou. A rough guess is that the nation-wide floating population may be (p.283) as much as 50 million, which is around 9 per cent of the total urban population. The rapid growth in their numbers is creating a social and economic rift in China’s urban areas which is far deeper and more extensive than any existing in the pre-reform period. This rift is not temporary, because it is founded upon the system of household registration, which is likely to stay in force for the foreseeable future. And the economic factors responsible for immigration into urban areas are likely to remain strong.

China has had a complex system of administrative assignment to jobs, which has been more extensive than those in East European economies. In the 1950s when it was introduced, the system covered only the State sector, but was extended to the collective sector during the period of the Cultural Revolution. The job assignment system does much more than bring together prospective employees and work units. It is as much concerned with employment creation as it is with matchmaking. Built into the system there has been an obligation on the part of the government (labour bureaux) to provide jobs to university graduates and new entrants to the permanent urban labour-force. The reforms have cut down on this obligation, but as yet only marginally. In the past, job assignment by labour bureaux imposed considerable restrictions on both prospective employees and work units. Employees were not free to choose their occupation or work unit. And work units had to accept candidates channelled to them by labour bureaux, even if they were unqualified for the job and redundant to their requirements. Work units were not allowed by bypass labour bureaux by recruiting employees directly, except in special cases and within limits. In effect, it was the labour bureaux rather than the work units which made recruitment decisions. Administrative assignment to jobs also helped to keep in check unlicensed immigrants from rural areas, as they were not entitled to job assignment and had very limited possibilities of being hired directly by work units.

As part of the economic reforms, the government has been seeking to transform recruitment and employment practices. One aim is to relieve the pressure on the State sector to create jobs by encouraging the growth of employment in collective enterprises, and the proliferation of private businesses. Although formally the government no longer guarantees employment to new entrants to the labour-force with urban registration and to university graduates, they are still covered de facto by an employment guarantee. What has changed, however, is that they now stand a lower chance of getting a regular job in the State sector than they did before the reforms. However, the creation of jobs so as to keep in check unemployment among the urban labour-force still remains a central aim. Recent years have witnessed the proliferation of so-called ‘labour service companies’. These are established by a diverse variety of work units, including labour bureaux, State-owned enterprises, and government organizations. They are, in effect, job creation schemes. But the jobs they provide are usually not on a par with regular jobs: the pay is lower and there is little or (p.284) no labour insurance cover. Thus employment creation through labour service companies is a cheaper alternative to creating regular employment in work units, because it saves on both pay and labour insurance benefits. Added to that, investment per extra worker is low in such companies.

A long-term aim of the labour reforms is completely to replace administrative assignment with direct recruitment by employers. Work units are now allowed a much wider leeway in direct recruitment, which has weakened the previous controls on hiring rural immigrants. Direct recruitment does give both job-seekers and work units a greater freedom of choice, but it also removes the employment guarantee which goes with administrative assignment. As channels of direct recruitment are still very underdeveloped, administrative assignment to jobs is likely to remain important for a long time to come (for a discussion of labour reforms see Feng and Zhao 1984).

Backed by stringent restrictions on immigration from rural areas, employment and recruitment policies have succeeded in keeping urban unemployment rates exceptionally low by international standards. For example in the decade 1978–87 (inclusive), the unemployment rate among the labour-force with urban registration averaged around 3 per cent; and the average over the last five years of the decade was even lower at 2 per cent (ZTN 1988: 175). However, China has gone through periods of comparatively serious urban unemployment, for example in the early 1950s, in the late 1950s to the early 1960s following the collapse of the Great Leap, and, recently, in the years from 1979 to the early 1980s when youths sent to the countryside during the Cultural Revolution returned. Recently recorded levels of urban unemployment would be much higher if urban residents without registration were included in the figures. Notwithstanding these qualifications, an exceptionally low rate of urban unemployment by international standards still remains a distinguishing feature of the Chinese economy. This may change in the future. The unemployment rate would rise steeply if the government were to press ahead with labour reforms designed to reduce feather-bedding in State-owned enterprises. And, as we pointed out earlier, contract employment may in time lead to a higher unemployment rate.

The absence of sizeable urban unemployment is, however, not without serious negative features. The low rate of manifest unemployment goes together with a high rate of disguised unemployment, or in-house unemployment’ (neibu diaye) as it is aptly termed in China. According to the Minister of Labour, there may be as many as 20 million surplus workers in State-owned enterprises, which is equal to a fifth of their labour-force in 1987 (Beijing Review, 19–25 Dec. 1988, p. 18). The figure is likely to be no more than a rough guess, but it does convey the magnitude of the problem. Feather-bedding has had an adverse effect not merely on labour productivity but also, we would argue, on total factor productivity. For the existence of surplus labour goes together with an organization of production in which an efficient (p.285) utilization of factors does not figure as a central consideration. How to reduce feather-bedding without causing a social upheaval is one of the most difficult problems facing the economic reforms in China. The problem cannot be solved quickly without a significant increase in at least temporary unemployment. The urban social-security system is not equipped, either financially or organizationally, to cope with a sizeable number of unemployed. In the past, it never had to face such a problem, and, as we discuss later, the newly introduced unemployment insurance scheme is designed to cater for only a limited category of the unemployed.

Historically, the government has sought to keep the task of providing jobs to urban residents within manageable limits by imposing stringent restrictions on the growth of the labour-force with urban registration, though, interestingly, not by reducing labour participation. Prior to the reforms, rural-urban migration was exceptionally low, and urban registration was strictly rationed. The former no longer holds, but the latter still does. And the government periodically pared the urban labour-force by dispatching immigrants back to the countryside (fang xia), as in the late 1950s and the early 1960s, or by sending the educated youth to the countryside, as during the period of the Cultural Revolution. The latter move, however, was dictated more by ideology than by a manpower policy. Reducing urban unemployment by repatriating rural immigrants depends crucially on the ability of the government to force rural communities to absorb the rusticated labour-force. This was relatively easy during the period of collective agriculture, but now it is more difficult. Added to that, the mechanisms for ensuring that repatriated people do not drift back to cities are now much weaker.

Yet the government has not attempted to contain the problem of urban unemployment by reducing labour participation. The reason is that work has not been regarded simply as a source of income, but also as a source of virtue. Much of the ideological education and social control of adults in China is based around the work unit. A host of measures has served to keep labour participation high. For example, facilitated by maternity benefits under labour insurance and the widespread provision of child care facilities by work units, female participation is exceptionally high in urban areas. However, attitudes towards female participation have begun to change. For the reasons discussed later, many enterprises now exert pressure on women workers to withdraw from the labour market. Moreover, there are increasing reports of discrimination against women in labour recruitment. The leadership, however, still remains committed to maintaining a high rate of female participation.

Turning to labour compensation, it may be argued that the urban labour-force has had implicitly to trade high employment for low pay and administrative assignment to a job, which has left little room for occupational choice. Low wage rates have also served to sustain a high rate of female participation by making it difficult to subsist on one income alone. In the past, wage rates were (p.286) governed by the national wage scale, which varied little across work units, and was independent of the financial performance of enterprises. Income inequality in urban areas was exceptionally low. As the economic reforms have unfolded, there has been an increase in bonuses and other forms of earnings linked to the financial performance of enterprises. Higher profits mean higher bonuses. Conversely, as we point out later, losses may reduce pay and erode labour insurance cover. Added to that, the reforms have also driven a wide wedge between the earnings of different occupational groups; in particular, the earnings of workers producing marketed goods and services have risen faster than those of people producing non-marketed goods and services such as medical care and education. A common refrain has been that a barber earns more than a surgeon, and a waiter more than a university professor (Liu 1989). The proliferation of labour service companies adds yet another dimension to differences in income by occupation.

3.2. Household Registration and Rationing of Consumer Goods

Household registration (hukou)), which is compulsory for all households, is not merely a source of data but also a means for controlling citizens. Whilst rural registration is simply a matter of record, urban registration is a privilege which is either inherited or acquired under certain restrictive conditions. Given this asymmetry, household registration can be used for rewarding and punishing citizens, indeed it has been and still is. Acquiring hukou in a small town is relatively easy, but very difficult in cities such as Beijing, Shanghai, or Tianjin. ‘Temporary residents’ are required to have a temporary household registration, but the requirement is increasingly violated. Household registration booklets (hukou bu) contain full details of the family and the history of births, deaths, and marriage. In effect, they contain all the information needed to target social-welfare payments according to family size and income. However, household registration has been used less for social welfare and more to control population migration. The way in which it is implemented has an important influence on household formation. For a spouse is not automatically entitled to reside in the same place as the other, nor parents in the same place as grown-up children.

In the pre-reform period, household registration, in conjunction with other restrictions, did succeed in keeping migration exceptionally low. But in recent years, the host of restraints which in the past kept the numbers of temporary visitors to cities low have either disappeared altogether or lost their rigour. Travel is no longer restricted. With the shift to the Responsibility System, leaving the land has become a matter for individual choice rather than for official discretion. In addition, the economic attraction of migrating to cities has increased. The urban construction boom has increased the demand for manual labourers. The removal of controls on private trading, self-employment, (p.287) and direct recruitment by enterprises has provided rural immigrants with much wider opportunities for earning a living. With the opening of private markets in grain and other necessities, household registration is no longer needed for their purchase. In sum, population movement in China is increasingly governed by the usual economic forces. Since urban registration is still strictly rationed, the result has been a massive increase in the numbers of people who live in urban areas but are not recognized as permanent urban residents. Thus household registration, as well as restricting migration to large cities, is also creating a marginal population with uncertain residence status and a precarious economic position.

Apart from opening doors to jobs with higher pay and in most cases with labour insurance cover, urban registration also confers entitlement to obtain rations of selected necessities at low prices. A central feature of the rationing of consumer goods in China is that it is targeted according to residential status and age, but not according to income or other economic characteristics. The rationing of consumer goods in China has lasted longer and has been more comprehensive than in any other economy in the world. Since the early 1950s when it began, the range of rationed goods has varied in time and place. The two decades spanning the Great Leap and the Cultural Revolution (1958–78) saw the extension of rationing to all but a few goods. Since 1978, most commodities have been de-rationed, leaving only the three main staples of grain, cooking oil, and cotton cloth, which are rationed nationally. In addition to these, there may be local rationing of such goods as meat and eggs. Further, the government has allowed the emergence of free markets in rationed commodities (for details of rationing of consumer goods see Tian 1989).

With the reforms, the financial burden on the government of maintaining rationing has mounted, as we pointed out in Section 1. This has put pressure on the government to raise the sale price of rationed goods on the one hand, and to restrict the issue of urban registration on the other. In smaller towns, the government has increasingly resorted to granting urban registration with the condition that the holder should not be entitled to low-price rations, thus introducing yet another status distinction in the urban population (see Fei Xiaotong et al. 1989: 199–200). In fact, were it not for the probability of resistance from households with urban registration, the government would like to do away with rationing altogether. Initially at the start of the reforms, consumers welcomed the removal of rationing because it meant the lifting of quantitative restrictions on purchases. But with the emergence of parallel markets in rationed commodities, consumers are free to buy more than the ration at higher market prices. The ration coupons simply entitle the holder to receive price subsidies, determined by the size of the ration and the price difference. Further, ration coupons have a cash value because they can be sold, a practice which, although illegal, seems to be widespread. As a result, (p.288) households with urban registration have a vested interest in the continuation of rationing.

How valuable has the rationing of consumer goods been as an income maintenance measure? Historically speaking, the introduction of rationing was motivated by the desire to ensure a just distribution of daily necessities. Theoretically, rationing of consumer goods is justified when the government lacks the means to control household incomes or transfer income, and when households differ widely (Weitzman 1977). Neither of these two conditions applied strictly to the pre-reform economy. Salaries and wages were controlled by the government, the degree of inequality in urban areas was low, and the household registration system provided possibilities for targeting assistance. Thus the continuation of rationing beyond the periods of emergency such as the early 1950s and the period of the Great Famine in 1959–62, lacked a coherent rationale in terms of social welfare. Rationing of consumer goods was a natural extension of government monopoly control of trade and prices, and motivated by a profound distrust of markets. Paradoxically, some of the arguments for rationing have gained in validity since the reforms. The government has much less control over incomes than before, and income inequalities in urban areas have considerably widened. But the traditional practice of targeting rations by residential status and age rather than by economic characteristics is becoming increasingly anomalous. The association between residential status and economic characteristics has become much looser. The increase in the numbers of urban residents without urban registration means that a substantial section of the low-income group in urban areas is excluded from low-price rations. In fact, household registration which is used to target rations by residential status can also be used to target rations by the economic characteristics of households.

3.3. Urban Social Security

The urban social-security system is largely anchored in the work units (danwei). Modelled on the army, danwei do not just perform a particular activity, but also take care of all aspects of their employees’ lives. For example, a large majority of urban residents live in housing provided by their work units, and the rest in municipal housing. Private housing is still extremely rare in urban areas. Scarce consumer goods are often distributed through work units. Retired workers receive their pensions and health care from their former work units, which, in most cases, continue to house them. Work units often provide employment for children of their employees, though this is not as important as it used to be (on the tutelary role of work units see Walder 1986; Henderson and Cohen 1984). Generally speaking, Chinese work units perform social-welfare functions which in other economies would be performed by the government. Preliminary attempts at enterprise reform have highlighted the (p.289) need to change urban social security, making it increasingly apparent that any radical reform at enterprise level is not possible without an overhaul of the occupational-welfare system, covering labour insurance and diverse range of benefits provided by work units.

The main benefits provided by labour insurance are, for the most part, grounded in the laws and regulations of the 1950s, when the system was introduced. Schematically, they are (for details see Wong and Macquarrie 1986: sec. 2.4)

  • health care and paid sick leave;
  • disability and retirement pensions;
  • maternity leave and benefit;
  • unemployment insurance (introduced only recently).

In addition to labour insurance, occupational benefits also include inter alia a diverse variety of cash allowances, such as those to compensate for inflation and for transport. In addition, work units provide various benefits in kind, such as child care facilities and subsidized meals (for a case-study see Wong 1989). By far the most important benefit in kind is heavily subsidized housing, which does not figure in social-welfare expenditure. Thus the total income of urban workers, including benefits in kind, far exceeds their salaries or wages. Moreover, as occupational benefits vary widely with employee status and work unit, total income inequality is prima facie much higher than wage and salary inequality.

For a developing economy, labour insurance in China provides generous benefits and covers a substantial percentage of the labour-force. For example, between 1978 and 1985, on average around 23 per cent of the total labour-force and around 91 per cent of the employed labour-force were covered by labour insurance (calculated from ZSTZ 1987: 111; and ZTN 1988: 153). After the 1950s, the percentage of the insured labour force increased with the expansion of employment in government departments and State-owned enterprises. However, the percentage has remained almost constant since 1978, because of the policy to shift the employment patterns away from the State sector. Although labour insurance regulations apply only to the State sector, the higher-tier collective enterprises in urban areas have tended to follow State-owned enterprises in adopting such measures, fairly closely in the pre-reform period but less so in recent years. For example, the percentage of the labour-force in collective enterprises covered by labour insurance fell slightly from 70 per cent in 1978 to 68 per cent in 1985 (calculated from ZSTZ 1987: 111; and ZTN 1988: 153).

By the norms of developing economies, the coverage of labour insurance in China is indeed impressive, but the question remains whether it is likely to extend to the whole of the employed labour-force, including that in rural industry, in the foreseeable future. This question is of some significance (p.290) because the percentage of the employed labour-force covered by labour insurance fell from 94 per cent in 1978 to 88 per cent in 1985. The reason for this reduction is the change in the distribution of the labour-force away from the State sector, where the coverage is total, to collective enterprises, where the coverage is less total and has fallen slightly. The shift in employment away from the State sector seems set to continue. This by itself would lead to a decrease in the coverage. Besides, future enterprise reforms are likely to transform the status of a large number of State-owned enterprises. In fact, this has already happened with the leasing out of small State-owned enterprises to managers. As we point out later, such enterprises have cut down on benefits. In sum, the developments in the offing seem to indicate a steady decrease rather than an increase in the percentage of the employed labour force covered by labour insurance. This impression is reinforced by the problems besetting labour insurance.

The fundamental problem with labour insurance is that its method of financing is ad hoc, and not commensurate with the long-term liabilities entailed by the labour-force presently covered by labour insurance. During the period from 1978 to 1985, whilst expenditure on labour insurance rose steeply (see Table 6.2 above), the percentage of the employed labour-force covered by labour insurance fell by 6 percentage points. Added to that, the ratio of expenditure on occupational welfare (labour insurance and other occupational provisions) as a proportion of the wage bill has risen from 14 per cent in 1978 to 27 per cent in 1987 (ZTN 1988: 203). At the time when the system was introduced, labour insurance was financed by joint contributions from employees and employers, and provision was made for future pension liabilities. With the beginning of the Cultural Revolution in 1966, employee contributions were abolished, the funding of pensions ceased, and so too, it seems, did any forward planning of labour insurance expenditure. Since then, all labour insurance expenditure, including pensions, has been financed on an ad hoc basis out of the current revenue of work units. This did not pose any problem until 1977, as the proportion of the total wage bill spent on occupational welfare remained almost constant between 1966 and 1977 (ZLGTZ 1987:189). But ad hoc financing has not been able to cope well with the sharp and unforeseen increase in expenditure on occupational welfare witnessed since 1978.

This sharp rise is due partly to the massive increase in expenditure on old-age pensions, which alone accounted for nearly half of the total increase between 1978 and 1985. And the proportion of total expenditure devoted to pensions rose from 24 per cent to 42 per cent (ZLGTZ 1987:191). The reasons for the increase lie partly in the legacy of the Cultural Revolution and partly in the demographic trend towards a higher percentage of elderly people in the population as a whole. During the period of the Cultural Revolution retirement due to old age was suspended. Employees kept on working past retirement age (p.291) unless physically incapacitated. The post-Mao leadership inherited a large backlog of employees past retirement age. It not only reinstated retirement, but also provided generous incentives for early retirement. This was done do make place for the returning youths who had been sent to the countryside during the Cultural Revolution, and also to clear the ranks of party cadres and government officials. It is this move that accounts for the sharp jump in the ratio of the retired to the employed between 1978 and 1983, as shown in Table 6.7.

Table 6.7. Numbers of retired people as % of employed



















Source: ZSTZ (1987: 114).

However, the increase in the percentage of the retired is not due to the legacy of the Cultural Revolution alone. It is also a result of increasing numbers of elderly people in the population, a trend which is expected to continue apace and even accelerate. For a low-income economy, China has an unusually large proportion of the aged in the population. The process of ageing has been faster in China than in most other economies because of the rapid decrease in fertility from the beginning of the 1970s and the rise in life expectancy. The rise in the percentage of the elderly in the population has not been evenly spread but highly skewed towards recent years. The rising trend seems to date from the latter half of the 1970s, which has presented the social-security system in the 1980s with a problem which it was ill equipped to handle after the upheavals of the Cultural Revolution. The financing of pensions, which is a topical subject in China, concerns two distinct issues. The first is a highly uneven distribution of the percentage of the retired to the employed across enterprises and provinces, and the second is the increasing percentage of elderly in the population. The first poses a problem only because the financing of pensions is, as we shall see, highly decentralized.

Turning now to problems posed by the economic reforms, the industrial reforms have sought to grant financial and operational autonomy to State enterprises, and have re-established the principle of the autonomy of collective (p.292) enterprises. Many small and medium State-owned enterprises have been leased out to managers. China has around half a million enterprises, most of them dating from the pre-reform period. By and large, they were not created on the basis of financial viability, immediate or distant. Granting them financial autonomy immediately raises the issue of survival for many of these enterprises. This poses a problem for labour insurance, which each enterprise is supposed to finance for its employees. Until the onset of the industrial reforms, the decentralized financing of social security was of nominal significance because enterprise budgets were not completely separated from the government budget. The separation of enterprise from government budgets, as sought by industrial reforms, converts nominal into effective liabilities. This brings in its train the possibility that some enterprises may not be able to maintain labour insurance cover for their employees. In fact, the rapid increase in the number of retired from 1979 onwards coincided with the onset of industrial reforms, and soon made it apparent that many of the now financially independent enterprises could not meet their pension liabilities.

Following the industrial reforms, the labour insurance system has to take account of the consequences of the risk that some enterprises may be insolvent. The risk of insolvency is all the greater in the Chinese economy because its enterprise population has not been ‘naturally selected’ for financial viability. The basic dilemma of enterprise reforms in China is that strict adherence to financial autonomy erodes the coverage of labour insurance, which, in principle, is meant to be independent of the place of employment. Conversely, efforts to maintain labour insurance cover through grants from the government seriously compromise the main aim of reforms. The government has tried to deal with this dilemma through a series of compromise principles. Loss-making enterprises, except in extreme cases, are not allowed to go bankrupt, though there is now a bankruptcy law.

What is needed is a disengagement of the coverage of labour insurance from the financial performance of particular enterprises. In the main this would involve a centralized funding of labour insurance, covering a group of enterprises. Such funding would insulate the coverage of labour insurance from sporadic financial insolvency. The extent of insulation would depend upon how heterogeneous the financial prospects of enterprises in the group are. The Chinese government has already taken some steps towards centralized funding by encouraging the piecemeal pooling of pension liabilities. The pooling schemes are local and predominantly organized by city governments. Under these schemes, participating enterprises contribute a percentage of their payroll to a pool which pays the pension of their retired employees. The percentage of the payroll paid into the pool and the proportion of pensions the pool pays out vary between schemes but not within schemes. In some cases, the pool pays out only a percentage of retirement pensions. Whilst in most areas the pooling schemes cover only the State-owned enterprises, some cities have (p.293) also introduced separate schemes for collective enterprises. Although 2,000 cities and counties have introduced pooling schemes, there is as yet no national framework for them. Neither is there any arrangement for pooling other items of labour insurance such as health care costs and maternity benefits.

The pooling schemes do no more than redistribute the current financial burden of pensions, and even that only among a relatively small number of enterprises usually in one city. As the ratio of the retired to the labour-force in the State sector varies widely across provinces, prima facie the burden is very unevenly distributed among the schemes (see ZSTZ 1987: 115). More important, the schemes do not address the problem of redistributing the financial burden of pensions over time raised by the increasing percentage of the elderly in the population, which is expected to continue well into the next century. The government has taken a small step in that direction by reintroducing the funding of pensions by joint employee and employer contributions. But this covers contract employees alone, the new recruits in the State sector from 1986. It does not address the pressing problem of redistributing the rising burden of pensions in the immediate future. The reason is that for several decades to come pensioners will be drawn from the ranks not of contract but of permanent employees, whose pensions have not been funded.

By opting for a separate funding of the pensions of contract workers, the government has forgone the promising possibility of financing the pensions of permanent employees from the contributions of the rising numbers of contract employees. Given that no new permanent employees are to be recruited, the ratio of the retired to still working permanent employees will keep on rising. Further, as contract employment applies only to new recruits, for the next twenty to thirty years almost all those reaching the age of retirement will be permanent employees, whose pensions have not been funded. As a result, funding the pensions of contract workers would do nothing to solve the proximate problem of the increasing burden of old-age pensions over the next 20 to 30 years. This would be possible if the pensions of both permanent and contract workers were financed from the same fund. The general point is that pooling the pension liabilities and contributions of various segments of the labour-force in such a way as to include all age-groups makes it possible to redistribute the immediate burden of pensions, which separate financing of pensions for each segment fails to do. Extending this point, it may be argued that as the formally employed labour-force is rapidly rising, and most of the new entrants are relatively young, China could easily solve the problem of financing old-age pensions in the immediate future by instituting a contributory pension scheme covering the whole of the employed labour-force. But the Chinese government seems more disposed towards maintaining labour market segmentation and instituting a separate scheme for each segment.

The problem with labour insurance does not consist merely of financing what the pre-reform laws provided for. As in agriculture, reforms in industry (p.294) have created new contingencies that labour insurance needs to take into account. The most important of such contingencies are, first, unemployment, and, second, a decline in income due either to poor performance by particular enterprises or to inflation. Unemployment, which was once assumed to be a problem particular to capitalist economies, is being recognized as a recurrent contingency. The economic reforms have created four potential sources of large-scale urban unemployment; as yet only one of them has led to significant unemployment. This is the system of employing large numbers of rural immigrants on short-term contracts or as casual labourers. The curb on investment imposed in the autumn of 1988 triggered large-scale redundancies among rural immigrants. Contrary to government expectations, those who had been laid off did not return immediately to the countryside, creating a massive social and economic problem in large cities, particularly Guangzhou. On the contrary, even while rural immigrants were being laid off, Guangzhou continued to attract large-scale migration from the country. The problem was accentuated by the fact that, being regarded as temporary residents, the migrants were not covered by urban social-security programmes. This episode brings home the general point that the massive influx of rural immigrants creates problems of unemployment and destitution in urban areas. It is no longer feasible for the government to make these problems disappear by dispatching the unemployed and the destitute back to the countryside, as it did in the pasŧ.

Among the remaining three potential sources of unemployment, the first is contract employment, which has built into it the possibility of spells of unemployment in between jobs. This may in future lead to significant unemployment, as the percentage of contract workers in the urban labour-force becomes large. The second is the possibility of loss-making enterprises, including the State-owned ones, going bankrupt. As pointed out earlier, bankruptcies among loss-making enterprises are still rare. The third is disguised unemployment in enterprises, which the government wants to reduce in order to further technical efficiency. But, as we point out below, until now enterprises have been expected to tackle the problem of disguised unemployment without large-scale redundancies. In 1986 the government set up a limited unemployment insurance scheme financed by a payroll levy on enterprises (see Hussain and Liu 1989: sec. 5.1). The scheme is designed specifically to reduce resistance to the replacement of permanent employment with contract employment and to bankruptcies. It is restricted to the employees of State enterprises only and promises relatively generous compensation, including inter alia a cash income related to basic wage in employment, health care costs, an allowance for dependents on death, and funeral expenses.

Until now, the scheme seems to have played little role in reducing feather-bedding. Enterprises use a variety of methods to redeploy their surplus labour. They may retrain workers in order to redeploy them internally, or pay part (p.295) wages to their redundant workers while they look for a job outside the enterprise. Another common method is the establishment by individual enterprises of auxiliary or ‘labour service’ companies to create jobs for their surplus labourers. These auxiliary companies may engage in activities far removed from the parent enterprise’s main line of activity. Thus, many enterprises have their own separate unemployment insurance schemes and labour placement and employment creation agencies—hence the common term ‘in-house unemployment’ (neibu daiye). The widespread reliance on internal redeployment is due mainly to the fact that the unemployment insurance scheme set up by the government does not cover redundant labourers in viable enterprises. As a scheme to reduce feather-bedding, internal redeployment is bound to be uneven in its effectiveness, because the opportunities available to an enterprise to redeploy its labour internally may not be commensurate with the amount of surplus labour it has. The government wants to encourage the reduction of surplus labour through external redeployment across enterprises; and it recognizes that this would involve redundancies and dismissals at a rate much higher than at present. The main question is, if that were to happen, would the unemployment insurance fund still be capable of providing the same generous coverage to all the unemployed as is provided for in the law? As indicated earlier, a rough estimate is that up to 20 per cent of the labour-force in State-owned enterprises may be redundant. And the unemployment scheme is financed by only a 1 per cent levy on the payroll.

Apart from unemployment, the other contingency which has arisen in the wake of the reforms is a decline in income due to poor financial performance by particular enterprises and to inflation. As we have already mentioned, by way of incentive the reforms have introduced a link between financial performance and industrial earnings. Such an incentive scheme has implications for social welfare because it opens up the possibility of earnings falling below the sustenance level. Prima facie, this is likely to be the case for a significant section of the industrial labour-force, because there are, and will continue to be for some time, large numbers of loss-making enterprises which are regarded as not yet ripe for bankruptcy. There is some evidence that employees of such enterprises suffer a substantial decrease in wages and occupational benefits (Hussain and Liu 1989: sec. 2.17). This points to an anomaly: whilst the employees of bankrupt enterprises are provided with adequate income and benefits on a par with labour insurance, there is no mechanism to ensure the same for the employees of loss-making enterprises. The general point is that, prior to the reforms, the indepedence of wages and occupational benefits from the financial performance of their work units amounted to an income guarantee, which has been partially invalidated by the industrial reforms. For social security, this raises the important issue of the extent to which employees ought to bear the risk of a loss of income and occupational benefits which is due to poor financial performance on the part of their enterprises.

(p.296) Urban residents are increasingly exposed to the risk of a decline in real income due to inflation, which did not exist in the pre-reform period. During the reform decade (1978–88), the inflation rate averaged 6 per cent per year and accelerated sharply to over 20 per cent in 1988. As a result, as much as 35 per cent of the urban population experienced a reduction in real income in 1988 (Beijing Review, 6–12 Mar. 1989, Documents). The Chinese economy still lacks regular procedures for reducing the adverse impact of inflation on social welfare. The official price indices are generally regarded as unreliable. Incomes, allowances, and pensions are periodically adjusted to take account of inflation, but they are not tied to any particular index of inflation. The group most at risk from the effects of inflation is pensioners. A survey conducted in 1987 revealed that as many as 20 per cent of retired workers in large cities had incomes below subsistence level (Hussain and Liu 1989: sec. 2.12). China has yet to introduce a regular uprating of pensions in line with inflation.

We turn to two particular cases to illustrate the problems arising out of the situation in which the industrial reforms have been introduced without corresponding changes in the labour insurance system. The first concerns leased-out enterprises, and the second maternity benefits in enterprises with a large percentage of female workers. A large number of small and medium State-owned enterprises have been leased out to management. Although formally such enterprises are expected to maintain labour insurance cover for their employees, many of them do not, skimping on paid sick and maternity leave, and injury and disability compensation. Similarly, enterprises with a large percentage of women employees increasingly default on maternity leave and on the benefits normally provided in labour insurance (for details, see Hussain and Liu 1989: sec. 2). Both these raise general issues of incentives and monitoring in respect to labour insurance. Prior to the reforms, enterprises, being financially dependent on the government, had no economic incentive to reduce labour insurance cover. But after the reforms, enterprises, concerned as they are meant to be with the pursuit of profit, have a strong financial incentive to cut down on those items of labour insurance which they finance individually. These include all but old-age pensions, for example, sickness and maternity leave and benefit. The decentralized financing of maternity benefits creates widespread anomalies because the percentage of women workers varies widely with enterprises and industries, and a reduction in the percentage of women workers may therefore appear attractive to certain enterprises. This may explain the increasing reports of discrimination against women referred to earlier.

The problem, however, is not that all enterprises have cut down upon the coverage of labour insurance. The industrial reforms have led to a divergence in the goals pursued by individual enterprises. Whilst some have switched over to the pursuit of profit, as intended by the reforms, many still remain occupied with enhancing the income and welfare of their employees. Profit-seekers try (p.297) to shed their welfare obligations, expecting the government to take on the burden—termed in China ‘shifting the burden to a maid’ (baomu diti). In welfare-oriented enterprises, managers and workers form a coalition, and use financial and operational autonomy to further their own parochial interest. This happened under the traditional system too, but it was heavily constrained. Both these pose problems. Profit-seeking reduces the Coverage of labour insurance; and whilst welfare-oriented enterprises do not undermine the labour insurance system, they nullify the central aim of enterprise reform.

A reform of the labour insurance system, which is generally considered to be urgently needed, has to keep in view the origins of the system and the general implications of the economic reforms. In the main the present labour insurance system dates from the 1950s and was designed for State enterprises and government agencies. This is evident from the fact that most of the laws and regulations governing labour insurance apply only to the State sector. At the time of the inception of the system the assumption was that in time all enterprises would evolve into State-owned enterprises, thereby extending the coverage of labour insurance. The process of evolution was accelerated during the two decades from 1958 to 1978. The government took over a large proportion of non-State enterprises, and those which had not been taken over behaved like State-owned enterprises. In the wake of these reforms, the perspective on the evolution of enterprises has undergone a fundamental change. It is now generally accepted that the Chinese economy will retain diverse forms of ownership; State, collective, foreign, and private, indefinitely into the future. Further, State ownership itself is likely to undergo a fundamental change.

The close link between the labour insurance system and State ownership is becoming increasingly anachronistic. This would suggest that the labour insurance system should cover the whole of the employed labour-force. Such a move would be desirable on the grounds not only of equity but also of efficiency, because the segmentation of the labour-force by differing entitlement to labour insurance is a massive hinderance to labour mobility between different categories of enterprises. But this cannot be done without a fundamental reassessment of the levels of benefits provided by labour insurance. The benefits are generous for a developing economy, especially for a low-income economy such as China. This is due to the fact that benefits stipulated by laws and regulations have been governed more by ideals than by financial feasibility. The problem of financing is not particular to old-age pensions. Although less pressing, similar problems exist with respect to maternity benefits and subsidized health care.

(p.298) 4. Conclusion

We return to the question raised in Section 1 together with a supplementary question: why has China been comparatively successful in preventing deprivation, and will it be able to sustain that success in future? It should be noted that China’s comparative success is neither unblemished nor unqualified. Apart from the famine of 1959–61, there have been local shortages of grain occasionally verging on famine. The threat of natural disasters leading to local starvation is still not completely ruled out. Pockets of poverty have existed in China and still do. The post-Mao leadership has been candid in admitting that there is still a sizeable problem of poverty in rural areas. Nevertheless, we still lack a systematic account of the patterns of poverty. There is need for further research, particularly into the patterns of rural poverty and into rural immigrants in cities. The usual assumption has been that poor households are largely confined to poor areas. This may be broadly correct, but there has been little work on the incidence of poverty in relatively prosperous areas. At various points in this chapter we have emphasized the problems posed by the growing number of rural immigrants in urban areas. Apart from impressions, there is very little material on either the size of the immigrant population or their economic and social conditions.

Taking a broad historical view, we would attribute China’s comparative success to the joint effect of the following three sets of factors:

  1. (1) capacity to mobilize the population for collective ends;

  2. (2) providing rural households with access to land, and maintaining a high rate of employment in urban areas;

  3. (3) social security programmes.

They are, however, not all on a par. The second set of factors has been far more important than the other two for reasons indicated below. More important, some of these factors no longer apply or, are not as effective as before.

We have included campaigns involving the mobilization of the population because China’s achievements in certain areas such as improvements in public hygiene and preventive health care cannot be explained without these. As instruments of change, campaigns have some severe limitations. By their nature they are transitory, they may help to launch initiatives but cannot sustain them. Whatever may be their pros and cons, the point is that the ability of the leadership to mobilize the population for collective ends has diminished, and is likely to diminish even further.

Providing all rural households with access to land either collectively or individually, and guaranteeing some form of employment to all urban residents with urban registration—the two guarantees as we termed them—have provided the Chinese population with a degree of economic security which is rare in (p.299) developing countries. The two guarantees, by eliminating landlessness and keeping down unemployment, have heavily reduced the population in need of assistance, leaving only those people who are unable to work and lack family ties to be cared for through social-security programmes. Broadly speaking because of the two guarantees, the Chinese social-security system has had a relatively light burden to carry. However, the effectiveness of the two guarantees in reducing the incidence of deprivation has depended crucially on keeping the numbers of urban residents without urban registration to a minimum, and on virtually banishing unemployment in urban areas. Before the reforms both these conditions held true, but increasingly less so after the reforms.

As we have pointed out, there has in recent years been a massive influx of rural immigrants into cities. Although resident and working in urban areas, they are officially classified as rural residents. A vast majority of them are in temporary employment which may be terminated at short notice. The rest are either self-employed, in casual employment, or unemployed. They do not benefit from any employment guarantee, as they are expected to return to the countryside. In the pre-reform period, the government simply repatriated rural immigrants when they were no longer needed. But repatriation is no longer a straightforward option. Rural immigrants are much larger in numbers, around 50 million according to a common guess. And there is now a problem of enforcing repatriation orders.

Unemployment is no longer a purely urban phenomenon, but now constitutes a problem in rural areas as well. In the wake of the reforms, there has been a massive shift of the labour-force within rural areas from farming to other activities. In particular, recent years have witnessed an explosive growth in rural industry—the township and village enterprises. This shift has a particular significance for the incidence of unemployment in rural areas, because whilst self-employment is almost a universal norm in farming, formal employment is prevalent in the non-farming sector. The sectoral shift in the rural labour-force creates a potential problem of unemployment, in that the employed can be laid off. In fact, this has already happened. The curb on investment in township and village industries imposed since autumn 1988 has led to redundancies on a large scale.

The general implication is that an employment guarantee extended only to the labour-force with urban registration is increasingly insufficient to maintain a low rate of unemployment in the economy as a whole. And the provision of land to the agricultural labour-force is no longer adequate to ensure that the means of earning a living are given to the whole of the rural labour-force, which also includes the workers in the non-farm sector. In fact, households which opt out of farming altogether, termed ‘specialized households’, surrender their claim to agricultural land. It is not easy for them to revert to farming when they cannot make a living out of their non-farming activities. China needs extensive employment schemes in rural areas and also special provision for rural (p.300) immigrants in cities if it is to maintain the level of economic security its citizens have enjoyed in the past.

In addition, the traditional method of keeping unemployment in check by creating jobs in State or collective enterprises has paid a heavy price in terms of efficiency, as is illustrated by the estimate that as much as 20 per cent of the labour-force in the State şęctor may be unemployed in disguise. The drive for efficiency and the traditional method of creating employment are increasingly in conflict. The implications of the economic reforms for unemployment have been hidden by spectacular rates of growth in national income, ranging between 10 and 14 per cent over the last five years (ZTN 1988: 53). The acceleration in the rate of inflation since 1988, which has continued into 1989, suggests that such rates of growth are unsustainable. A slow-down, which the government is aiming for, would lead to a massive increase in unemployment. Even though there is an unemployment insurance scheme, the Chinese social-security system įş neither organizationally nor financially capable of handling unemployment pn a large scale.

Turning to şpcial-security programmes, we have seen that these are fairly sparse in rural areas, but elaborate in urban areas, particularly for those who are employed in the State sector. Rural social-security programmes were to a considerable degree intertwined with the collective economy; and the end of collective farming has cut the ground from underneath many of them. Rural sociaļ-seçurity programmes require a new financial foundation, which takes account of the fact that increasingly income from economic activities accrues to households and not to collectives. There is also a need for a legal framework to meet social-welfare obligations, which can no longer be left entirely to benevolence and the spirit of solidarity in villages.

The comparatively elaborate social-security system in urban areas has come under strain due to enterprise reform and also to the increasing numbers of the retired. The system suffers from two main problems: first, it largely takes the form of ‘occupational welfare’ based on work units; and, second, labour insurance, which is the main component of the system, does not cover the whole of the employed labour-force. Not only is the proportion of the employed labour-force with no labour insurance cover increasing in urban areas, but also labour insurance does not apply at all to the rapidly growing labour-force in township and village enterprises. The rural labourers are regarded as self-employed, as it were, even when employed. Unless the labour insurance system is overhauled, it may in time be reduced to a selective scheme applying only to a part of the urban labour-force.

A pervasive feature of the Chinese economy is the segmentation of the labour-force by differing entitlements to social security. In the main, the segmentation runs along two lines which overlap partially. The first is the distinction between the rural and the urban labour-forces, which is based on household registration and not on the current place of residence. The second, (p.301) which applies to the urban labour-force, consists in the distinction between regular employees, including permanent and contract workers who are all in the State sector, and non-regular employees, including temporary workers and workers in the collective sector. These distinctions have become increasingly anomalous with the economic reforms. We have seen that there are now large numbers of rural immigrants who live and work in urban areas. They are excluded to a great extent from urban social-security programmes, since for administrative purposes they are regarded as rural residents. In principle, labour insurance does not extend to rural areas because rural inhabitants are regarded as self-employed. Yet there is now in rural areas a large employed labour-force. In urban areas, labour insurance laws and regulations, strictly speaking, apply only to the State-sector employees, even though the policy of the government is to shift employment away from the State sector. By default, if not by design, the government is seeking to reduce the percentage of the employed labour-force which is covered by labour insurance.

The Chinese social-security system still remains tied to pre-reform distinctions even though the changes brought about by the reforms are rendering them obsolete. In particular, the assumption that ‘rural’ corresponds to ‘self-employed’ and ‘urban’ to ‘employed’, which underlies labour insurance, is increasingly at variance with the actual pattern. The general tenor of reforms is to reduce security of employment or income where it conflicts with efficiency. This is illustrated by the attempt to replace permanent employment with ‘terminable’ employment. Further, the reforms have given rise to phenomena which increase the risk of deprivation, such as immigration from rural areas on a large scale. Apart from the problems posed by the economic reforms, the Chinese social-security system also needs to take account of the growing percentage of the elderly in the population. The implication is that the system has to shoulder a heavier burden of alleviating deprivation than it did in the past, if the population is to enjoy a reasonable degree of economic security. But, as we have pointed out, this would require a fairly major overhaul of the present system.

(p.302) References

Bibliography references:

Ashton, B., Hill, K., Piazza, A., and Zeitz, R. (1984), ‘Famine in China: 1958–6’, Population and Development Review, 4.

AYB 1986 (China Agricultural Yearbook 1986) (1987), Beijing: Agricultural Publishing House.

Banister, J. (1987), China’s Changing Population. Stanford: Stanford University Press.

Banister, J. and Preston, S. H. (1981), ‘Mortality in China’, Population and Development Review, 1.

Bernstein, T. (1985), ‘Reforming Chinese Agriculture’, China Business Review, Mar./ Apr.

Chen, C. C. (1989), Medicine in Rural China. Berkeley: University of California Press.

Coale, A. J. (1984), Rapid Population Change in China: 1952–82, Report No 27, National Research Council Committee on Population and Demography, Washington DC: National Academy Press.

Croll, E. (1982), The Family Rice Bowl. London: Zed Press.

Dixon, J. (1981), The Chinese Welfare System. New York: Praeger.

Endicott, S. (1988), Red Earth: Revolution in a Sichuan Village. London: I. B. Tauris.

Fei X (HS) (Fei and Others) (1989), Small Towns in China. Beijing: New World Press.

Feng L. and Zhao L. (1984), ‘Urban Employment and Wages’, in Yu Guangyuan (ed.), China’s Socialist Modernization. Beijing: Foreign Languages Press.

Feuchtwang, S. (1987), ‘Changes in the System of Basic Social Security in the Countryside Since 1979’, in A. Saith (ed.), The Re-Emergence of the Chinese Peasantry: Aspects of Rural Decollectivization. London: Croom Helm.

Harding, H. (1987), China’s Second Revolution. Washington DC: The Brookings Institution.

Henderson, G., and Cohen, M. S. (1984), The Chinese Hospital: A Socialist Work Unit. New Haven, Conn.: Yale University Press.

Hinton, W. (1966), Fanshen: A Documentary of Revolution in a Chinese Village. New York: Monthly Review Press.

Horn, J. S. (1971), Away With All Pests: An English Surgeon in People’s China. New York: Monthly Review Press.

Hussain, A. (1989), ‘Science and Technology in the Chinese Countryside’, in M. Goldman and F. Simon (eds.), Science and Technology in Post-Mao China. Cambridge, Mass.: Harvard University Press.

Hussain, A. and Feuchtwang, S. (1988), ‘The People’s Livelihood and the Incidence of Poverty’,in S. Feuchtwang et al. (eds.), Transforming China’s Economy in the Eighties, i. London: Zed Press.

Hussain, A. and Stern, N. (1988), ‘On the Recent Increase in Death Rates in China’, China Programme, Research Working Papers, No. 12, STICERD, London School of Economics.

Hussain, A. and Liu H. (1989), ‘Compendium of Literature on the Chinese Social Security System’, China Programme, Research Working Papers, No 3, STICERD, London School of Economics. (This work contains an annotated bibliography of the Chinese literature on social security.)

(p.303) JPRS (Joint Publication Research Service) (1986). ‘Beijing’s Floating Population During Economic Reforms’, CEA-86–107, 6 Oct., Washington DC.

Liu G. (1989). ‘A Sweet and Sour Decade’, Beijing Review, 2–8 Jan.

Mao Z. (1974), ‘Directive on Public Health’, in S. Schram (ed.), Mao Tse-Tung Unrehearsed: Talks and Letters. London: Penguin Books.

NTN (Nongcun Tongji Nianjian) (Agricultural Year Book), (1986, 1987), Beijing: Zhongguo Tongji Chubanshe.

Parish, W. L., and Whyte, M. K. (1978), Village and Family in Contemporary China. Chicago: University of Chicago Press.

Peng X. (1987), ‘Demographic Consequence of the Great Leap Forward in China’s Provinces’, Population and Development Review, 4.

Peng X. (1989), ‘Major Determinants of China’s Fertility Transition’, China Quarterly, 117.

Perkins, D. (1986), China: Asia’s Next Economic Giant. Seattle: University of Washington Press.

Perkins, D. and Yusuf, S. (1984), Rural Development in China. Baltimore: Johns Hopkins University Press.

Perkins, D. (1988), ‘Reforming China’s Economic Structure’, Journal of Economic Literature, 2.

Perkins, D. and Yusuf, S. (1984), Rural Development in China. Baltimore: Johns Hopkins University Press.

Schurman, F. (1971), Ideology and Organization in Communist China. Berkeley: University of California Press.

Shao Y. (1988), Health Care in China. London: Office of Health Economics.

Sicular, T. (1988), ‘Plan and Market in China’s Agricultural Commerce’, Journal of Political Economy, 2.

SYB (Statistical Year Book of China) (1981, 1983, 1986, 1987), Beijing: State Statistical Bureau.

Tian J. (1989), ‘Consumer Goods Rationing in China’, mimeo, China Programme Research Reports, STICERD, London School of Economics.

TPS (China Trade and Price Statistics in 1987) (1987), Beijing: State Statistical Bureau.

UN (1984), Mortality and Health Policy. New York: UN.

Vermeer, E. B. (1979), ‘Social Welfare Provisions and the Limits of Inequality in Contemporary China’, Asian Survey, 9.

Vermeer, E. B. (1982), ‘Income Differentials in Rural China’, China Quarterly, 89.

Walder, A. G. (1986), Communist Neo-Traditionalism: Work and Authority in Chinese Industry. Berkeley: University of California Press.

WDR (World Development Report) (1989), Oxford: Oxford University Press.

Weitzman, M. L. (1977), ‘Is the Price System or Rationing More Effective in Getting a Commodity to Those Who Need It Most’, Bell Journal of Economics, 8.

White, G. (1987), ‘The Politics of Economic Reform: Introduction of the Labour Contract System’, China Quarterly, 111.

Whyte, M. K., and Parish, W. L. (1984), Urban Life in Contemporary China. Chicago: University of Chicago Press.

Wolf, M. (1985), Revolution Postponed: Women in Contemporary China. Stanford: Stanford University Press.

(p.304) Wong, L., and Macquarrie, L. (1986), China’s Welfare System: A View From Guangzhou. Hongkong: Hongkong Polytechnic.

Wong, T. T. (1989), ‘The Salary Structure, Allowances and Benefits of Shanghai Electronics Factory’, China Quarterly, 117.

World Bank (1983), Socialist Economic Development, i and ii. Washington DC: World Bank.

World Bank (1984), China: The Health Sector. Washington DC: World Bank.

World Bank (1985), China: Long-Term Development Issues and Options. Washington DC: World Bank.

Yang, C. K. (1959), The Chinese Family in the Communist Revolution. Cambridge, Mass.: MIT Press.

YRSES 1986 (Yearbook of Rural Social and Economic Statistics 1986), Bei jing: State Statistical Bureau.

ZLTGTZ 1987 (Zhongguo Laodong Gongzi Tongji Ziliao: 1949–1985), Beijing: Zhongguo Tongji Chubanshe.

ZSTZ (Zhongguo Shehui Tongji Ziliao) (1987), Beijing: Zhongguo Tongji Chubanshe.

ZTN (Zhongguo Tongji Nianjian 1988) (1988), Beijing: Zhongguo Tongji Chubanshe.

ZTYB (Zhongguo Tongji Yuebao, 12, 1988) (Dec. 1988), Beijing: Guojia Tongji Ju.

Zweig, D. (1989), Agrarian Radicalism in China: 1968–1981. Cambridge, Mass.: Harvard University Press.


(*) The paper on which this chápter is based arose out of the programme of research funded by the Overseas Development Administration of the UK government, the Ford Foundation, and the Bŗadley Foundation. We are grateful to Anthony Atkinson, Meghnad Desai, Jean Drèze, John Hills, and Nicholas Stern for comments and suggestions, and owe special thanks to Hong Liu for research assistance. Athar Hussain would also like to thank the British Academy for a personal research grant during the initial stages of his work on the Chinese social-security system.