Introduction
From the mid-1980s, social policy in East Asia has attracted growing attention from academic researchers, with an emerging literature identifying basic features of East Asian social policy, or suggesting a distinct model for the welfare states in the region (for example, Midgley, Reference Midgley1986; Jones, Reference Jones and Jones1993; Kwon, Reference Kwon1997, Reference Kwon and Kwon2005; White and Goodman, Reference White, Goodman, Goodman, Gordon and Kwon1998; Holliday, Reference Holliday2000, Reference Holliday2005; Tang, Reference Tang2000; Holliday and Wilding, Reference Holliday, Wilding, Holliday and Wilding2003; Ramesh, Reference Ramesh2004; Aspalter, Reference Aspalter2006, Reference Aspalter2011). Some commentators have argued that Esping-Andersen's (Reference Esping-Andersen1990, Reference Esping-Andersen1997) threefold framework of welfare regimes is unhelpful for explaining the basic features of welfare states in East Asia (Goodman and Peng, Reference Goodman, Peng and Esping-Andersen1996; Kwon, Reference Kwon1997; White and Goodman, Reference White, Goodman, Goodman, Gordon and Kwon1998; Holliday, Reference Holliday2000). Consequently, several new models for explaining the basic features of East Asian welfare regimes have emerged in the literature since the 1990s (for example, Jones, Reference Jones and Jones1993; Holliday, Reference Holliday2000, Reference Holliday2005; Holliday and Wilding, Reference Holliday, Wilding, Holliday and Wilding2003; Gough, Reference Gough, Gough and Wood2004; Aspalter, Reference Aspalter2006, Reference Aspalter2011).
The major weakness of approaches such as Jones’ and Aspalter's is that they try to explain all of the commonly examined East Asian welfare systems (Hong Kong, Singapore, Korea and Taiwan) with a single model. Yet it is evident that these four states have taken different paths to social policy development over the past two decades (Holliday, Reference Holliday2005; Wilding, Reference Wilding2008). By contrast, Ramesh (Reference Ramesh2004) constructs a twofold framework of East Asian welfare regimes that comprises the liberal and conservative models, while Kwon's (Reference Kwon and Kwon2005) developmental welfare state model suggests that there are inclusive and selective forms of welfare developmentalism in the East Asian contexts. They similarly classify Korea and Taiwan into one sub-group and Hong Kong and Singapore into another. Both Ramesh's and Kwon's frameworks appear to be more useful for describing the basic features of East Asian welfare regimes, as their approaches identify the heterogeneity of social policy development in the region. Although Korea and Taiwan have been confirmed as belonging to the same regime type (Lee and Ku, Reference Lee and Ku2007), there is little empirical support for the homogeneity of Hong Kong and Singapore in terms of basic features of social policy (Chau and Yu, Reference Chau and Yu2013). In the case of Hong Kong, Ramesh (Reference Ramesh2004) regards the city-state as an example of the liberal welfare model, while Kwon (Reference Kwon and Kwon2005) suggests it as a case of the selective form of developmental welfare state.
Despite the fact that Macao is located in East Asia, it has been neglected in comparative studies on East Asian welfare models (for example, Jones, Reference Jones and Jones1993; White and Goodman, Reference White, Goodman, Goodman, Gordon and Kwon1998; Holliday, Reference Holliday2000; Holliday and Wilding, Reference Holliday, Wilding, Holliday and Wilding2003; Ramesh, Reference Ramesh2004; Kwon, Reference Kwon and Kwon2005; Aspalter, Reference Aspalter2006, Reference Aspalter2011). The exception is Lai's (Reference Lai2012) study, which concluded that Macao does not appear to fit with the inclusive developmental welfare state. Rather, he suggests that the city-state is a prototype of a ‘regulatory welfare regime’, which is characterised by a weak modification impact on the capitalist social structure and a predominant role for social policy in maintaining the workings of the capitalist system (Lai, Reference Lai2012).
In looking to advance debates about welfare regimes in the region, there are good reasons for comparing Hong Kong and Macao: they are neighbouring city-states situated at China's southern coast; an overwhelming majority of their populations are of Chinese nationality; both were under the colonial administration of European countries for over a hundred of years; they are the Special Administrative Regions of the People's Republic of China; and have semi-democratic political systems at present. Despite these similarities, there have been no such comparative studies to date, and the main objective of this article is therefore to compare key features of the welfare regimes of the two city-states and to discuss their current welfare models. Before turning to this task, however, we first provide a brief sketch of the key social policy provisions in Hong Kong and Macao.
An overview of social welfare in Hong Kong and Macao
Social security programmes
There are four main pillars of Hong Kong's social security system. The Comprehensive Social Security Assistance Scheme provides a means-tested payment for the poorest, covering approximately 6 per cent of the population (Hong Kong Social Welfare Department, 2013). The Social Security Allowance Scheme comprises four universal payments for older people and disabled people, while the Old Age Living Allowance is provided for older people who have limited economic resources (Hong Kong Social Welfare Department, 2013). Finally, there is a Mandatory Provident Fund Scheme that is a contributory and privately managed saving scheme for retirement protection. Macao's social security system similarly consists of four main components although with differing composition and some differing mechanisms. The Social Security Fund, the largest element, is a social insurance scheme financed through the government's funding and monthly contributions from employers and employees, with coverage extending to over 80 per cent of older people in Macao at present. The Financial Assistance scheme provides means-tested support to the most deprived, and the government also provides low-rate social allowances for the elderly and the disabled. Finally, there is a wage subsidy scheme that supports low-paid employees (Lai, Reference Lai2012).
Key human services
With regard to education, provision in Hong Kong and Macao has many similarities. The Hong Kong government provides twelve years of free public school education, or education at ‘aided schools’ that are heavily subsidised by public funds. Since 2007-8, the government has implemented a Pre-primary Education Voucher Scheme, which offers a subsidy for parents to exercise their consumer choice in selecting kindergartens for their children. In Macao, meanwhile, the government provides a fifteen-year free education policy, covering kindergarten to higher secondary schooling provided by both public and heavily subsidised private schools. In addition, the government gives a school fee subsidy to eligible students in self-financed private schools.
Hong Kong's healthcare system is characterised by a large-scale provision of public healthcare that consists of a range of heavily subsidised and accessible primary healthcare and hospital services. The public hospitals cover the vast majority of in-patients in Hong Kong. In addition, the government has recently begun to provide healthcare vouchers that help older people to purchase primary health care in private clinics. As far as healthcare provisions in Macao are concerned, the scale is modest in comparison, partly due to its smaller population: it has one public hospital and several district health centres. Although the health centres provide free universal primary healthcare, the public hospital offers free services only for certain social groups, such as children, older people, poor people and cancer patients. In addition, the government subsidises several kinds of medical services in a private hospital and charitable clinics, and, similar to Hong Kong, in recent years the government has implemented a healthcare voucher scheme to subsidise medical services in private clinics (Lai, Reference Lai2012)
While the Hong Kong government plays a major role in financing personal welfare services, non-government organisations (NGOs) bear the main responsibility for the delivery of these services with the support of public funding (Chui et al., Reference Chui, Tsang and Mok2010). The government also provides financial support for older people to purchase personal care in private residential homes. However, government departments are still responsible for providing a small part of social services provision, particularly statutory services such as those relating to domestic violence and probation casework. In Macao, non-governmental organisations are the major providers of personal social services too, running more than more 80 per cent of services, although the NGOs are heavily subsidised by public funds. In addition to financing and monitoring the services provided by NGOs – again with similarities to Hong Kong – government departments provide a range of services related to certain statutory duties, including medical social services, services relating to domestic violence and services for offenders and prisoners (Lai, Reference Lai2012).
Public housing
The Hong Kong government is the world's second largest provider of public housing, with only Singapore offering a more extensive programme. There are currently two kinds of public housing: rental housing units and government-built for-sale flats. The former is targeted at low-income families and the latter is a kind of means-tested provision with higher limits of income and assets than the rental public housing (Chui, Reference Chui, Aspalter, Aldosary, Dashkina and Singh2008). Nearly half of Hong Kong's population resides in these two types of subsidised housing. After a suspension of building for-sale flats lasting for several years, the fourth-term Chief Executive, who took office in July 2012, promised that the project of for-sale flats would be re-launched. The scale of intervention in Macao is modest by comparison with Hong Kong, but by no means insignificant. About 13 per cent of Macao's households reside in subsidised housing (Macao Statistics and Census Service, 2012), and there are four major types of housing intervention programme: a social housing scheme for low-income households, a home ownership scheme for low- to middle-income families, a temporary rental allowance for households that are on a waiting list for social housing and a mortgage interest subsidy for households that have purchased housing flats in the property market.
Analysing welfare regimes
As the discussion above makes clear, the welfare mix in Hong Kong and Macao is complex, and the state's role is smaller in many policy sectors than is the case for the welfare regimes explored by Esping-Andersen (Reference Esping-Andersen1990). The selection of indicators for the exploration of the welfare regime in Hong Kong and Macao requires careful consideration therefore. Here we examine two major dimensions of welfare systems in order to allow us to compare Hong Kong and Macao with key cases from Esping-Andersen's welfare regimes. The first relates to the pattern of the welfare mix, and the second is concerned with the impact of social policy on the capitalist social structure. While the second dimension is an important consequence of social policy, the first defines the institutional framework within which the impact of social policy is produced. These two dimensions together provide us with a comprehensive view of the characteristics of a welfare regime. We examine the basis of each in turn, before including both in a broad comparison of welfare types that includes Hong Kong, Macao and key cases representing Esping-Andersen's three regime types.
The welfare mix in Hong Kong and Macao
According to Powell and Barrientos (Reference Powell and Barrientos2004: 86–7), welfare mix refers to the ‘articulation of the markets, the state and the family in welfare production’, and it ‘constitutes the centre of gravity of welfare regimes’. Besides these three welfare providers, NGOs have played a very significant role in the delivery of social services in Hong Kong and Macao. The NGO sector in Hong Kong, as well as that in Singapore, has been characterised as a ‘state-corporatist regime’ (Lee and Haque, Reference Lee and Haque2008) in which the government has incorporated NGOs as partners in social service provision. Likewise, the government and NGOs in Macao have long built a partnership in the delivery of social services. In light of this, the welfare role of the NGOs is also included in the comparison. These four welfare providers are inter-related, and, in particular, the state's policy plays a key role in shaping the scale of provision of the market, the family and the NGOs. When welfare production is dominated by the state, the welfare roles of the other providers are relatively less significant. When the state's welfare production is limited, the market, the family and the NGOs play a more significant role in addressing social needs
Table 1 presents a comparison of the welfare mixes between Hong Kong and Macao. In general, Hong Kong's and Macao's scales of public welfare provision, in terms of the level of public social expenditure, are relatively moderate compared to Western welfare states. For example, the public social expenditure of Sweden, Germany and the United States accounted for 30.3 per cent, 29.7 per cent and 20.8 per cent of their GDP respectively (see Table 3).Footnote 1 The Hong Kong government's public social expenditure accounted for a total of 8.2 per cent of GDP, while the spending of the Macao government constituted 5.0 per cent of GDP. It appears, therefore, that intervention in addressing social needs is of a larger scale by the Hong Kong than by the Macao government.
Hong Kong's and Macao's welfare mixes
Source: Data for Hong Kong was extracted from Hong Kong Census and Statistics Department (2012a); Hong Kong Census and Statistics Department, Hong Kong Annual Digest of Statistics (various editions); Hong Kong Food and Health Bureau (2004); Hong Kong Mandatory Provident Fund Authority, Mandatory Provident Fund Schemes Statistics Digest (various issues). Part of the data was provided directly by Hong Kong Census and Statistics Department and Hong Kong Social Welfare Department. Data for Macao were extracted from Health Bureau, Yearbook of Statistics (various years); Lai (Reference Lai2012); Macao Statistics and Census Service (2004, 2009, 2013); Macao Statistics and Census Service, Yearbook of Statistics (various years); Macao Social Welfare Bureau, Departmental Report (various years); Monetary Authority of Macao, Private Pension Fund Annual Report (various years). Some of the data were provided directly by Kiang Wu Hospital and Macao Social Welfare Bureau.
The government's relatively moderate role in welfare provision in both Hong Kong and Macao has reinforced the emergence and expansion of market welfare, which plays a major role in addressing certain social needs. It is interesting to note that the significance of market provision in the two city-states varies across policy areas. Table 1 indicates that, in the case of old-age financial security, the coverage of private pensions in Hong Kong (86.1 per cent) was much larger than in Macao (25.5 per cent). The difference is largely explained by the fact that Hong Kong's major pillar of retirement protection is the Mandatory Provident Fund Scheme. Similarly, Hong Kong's private residential homes accounted for a greater share of the provision of institutional care for elderly people (69.8 per cent Hong Kong versus 38.3 per cent Macao) and private nurseries had a larger proportion of the provision of childcare services compared with Macao (69.3 per cent Hong Kong versus 11.8 per cent Macao). Conversely, in Macao there were more in-patients treated in private hospitals (41.6 per cent Macao versus 17.7 per cent Hong Kong) and a larger share of total households living in private housing than in Hong Kong (79.7 per cent Macao against 52.1 per cent Hong Kong).
Families in Hong Kong and Macao generally serve as the primary caregivers, despite a variation across groups of care recipients. Table 1 shows that 92.8 per cent of older people in Macao did not receive any kind of care services, compared with 88.9 per cent in Hong Kong. For this group of older people, family is expected to be the primary source of the care they receive. As regards the provision of childcare, children under the age of three not placed in day nurseries accounted for a larger proportion in Hong Kong (88.4 per cent) than in Macao (76.0 per cent).
The provision of personal welfare services in both Hong Kong and Macao is marked by a dominant role for NGOs in the delivery of the services. In particular, the role of NGOs in Macao is noticeably larger than in Hong Kong. As Table 1 indicates, Macao's NGOs had larger proportions of the provision of residential care for older people (61.7 per cent Macao versus 30.2 per cent Hong Kong), day child care (88.2 per cent Macao versus 30.7 per cent Hong Kong), and non-institutional rehabilitation services, compared with their counterparts in Hong Kong (100.0 per cent Macao versus 98.3 per cent Hong Kong). Nevertheless, it is important to note that the services provided by NGOs are heavily subsidised by public funding. While the NGOs in Hong Kong and Macao dominate the delivery of personal social services, the governments actually play a vital role in financing the services.
The modification impact of social policy
The second dimension of the comparison relates to the modification impact on the social structure of a capitalist society. Since social policy is a form of state intervention that modifies the play of market forces or the operating principles of capitalism (Marshall, Reference Marshall1981; Briggs, Reference Briggs, Pierson and Castles2006), what we call the modification impact of social policy reflects the very basic nature of a welfare regime. The modification impact is mainly concerned with two salient features of modern capitalism: the dominance of the market economy and the structure of class divisions. Market dominance inevitably results in a process of the commodification of people and poses a threat to the personal autonomy of labourers (Lai, Reference Lai2012). Social policy carries two kinds of effects that challenge the domination and coercive forces of the market. The first is decommodification, which refers to the extent to which social policy makes individuals and families capable of obtaining ‘a socially acceptable standard of living independently of market participation’ (Esping-Andersen, Reference Esping-Andersen1990: 37). The second is the promotion of personal autonomy by offering a choice between working in paid jobs and being supported by welfare payments (Goodin et al., Reference Goodin, Headey, Muffels and Henk-Jan1999). With regard to social structure, it is widely accepted that the abatement of class divisions is partly achieved by the development of social policy (Marshall, Reference Marshall1950; Briggs, Reference Briggs, Pierson and Castles2006). There are two major ways through which social policy combats class divisions: redistributing income by means of taxation and social policy provisions, and promoting equality of status through the universal provision of a range of social services.
In summary, the concepts of decommodification, autonomy promotion and class abatement serve to capture the degree to which the basic structure of (laissez-faire) capitalism is modified or reformed through a range of welfare provisions and social legislations. These concepts are useful criteria that are used for comparing basic features of welfare regimes. The effects of decommodification, autonomy promotion and class abatement are measured by nine indicators, which are presented in Table 2. Since it is acknowledged that ‘welfare production must be examined as a whole’, the nine indicators cover the areas of both cash transfers and services: social security, healthcare and education (Powell and Barrientos, Reference Powell and Barrientos2004). It should be borne in mind that these three effects are distinguished only at the conceptual level. In reality, one social provision can carry more than one effect. For example, universal free education carries the effects of decommodification and class abatement; and the poverty rate reflects the degree of income disparity (class inequality) and also the extent to which personal autonomy of the people is constrained.
Indicators of the modification impact of social policy
Notes: a decommodification, b autonomy promotion, c class abatement
d Macao's figure for Indicator 4 includes public in-patient services and publicly funded services in the private hospital.
e Originally the poverty rate is the indicator of the effects of class abatement and autonomy promotion, but their relationships are indeed negatively correlated. Thus, ‘the non-poverty rate’, which is equal to 100 minus the poverty rate, is used instead for the comparison and the calculation of the overall modification impact. The same method is also applied to ‘Degree of equality of income distribution’.
We measured the overall modification impact of social policy with an overall modification score, calculated by averaging the scores of all the nine indicators in Table 2. These indicators were selected with reference to the methodologies of Esping-Andersen (Reference Esping-Andersen1990), Bambra (Reference Bambra2005) and Goodin et al. (Reference Goodin, Headey, Muffels and Henk-Jan1999). A higher overall modification score meant a stronger modification impact. Instead of simply examining numerical values for all the indicators, we used their z scores to gain an accurate understanding of the difference in each value relative to the others. The z scores were derived from the means and standard deviations of the eighteen welfare states studied in Esping-Andersen's (Reference Esping-Andersen1990) original work.Footnote 2 This guaranteed that the comparative analysis would be based on the standards of advanced capitalist economies. In other words, Hong Kong's and Macao's social policy effects were evaluated by the standards of the most developed countries in the world. Finally, we noted the difference in the peculiarity of welfare regimes between Hong Kong and Macao when comparing their overall modification scores. Using the same method of calculation, we expressed the effects of the decommodification of people, autonomy promotion and class abatement in a decommodification score, an autonomy promotion score and a class abatement score, respectively; these were equal to the mean values of the z scores of their respective indicators presented in Table 2. Moreover, the comparison involved four more welfare states, Sweden, Germany, the United States and Korea, which are the representative cases of the social democratic, the conservative, the liberal welfare regimes and the developmental welfare state, in order to examine their relevance to explaining basic features of Hong Kong's and Macao's social policies.
Two major limitations exist with respect to this methodology. First, the selection of indicators is partly data-driven, rather than completely theory-driven. Certain indicators that might be more relevant to measure a particular effect are not used due to the unavailability of their raw data in Macao and/or Hong Kong. For example, statistics on the ‘change of Gini coefficient after post-tax post-social transfer’, which is a relevant indicator of the class abatement effect, is not available in Macao. Second, the assessment of the modification effect of social policy in terms of the overall modification score is based on the assumption that all the indicators carry equal weight with respect to the effect. This assumption is inevitably subject to the criticism that it lacks adequate empirical justification.
Results of the comparison
Table 3 reports the modification impact by social policy of the six states. It shows that Hong Kong, with an overall modification score of –2.40, achieved the poorest performance in modifying the capitalist social structure through its social policy. Hong Kong was followed by Macao, which scored –2.20, but their overall modification scores have only a very small difference. Hong Kong's and Macao's scores were much lower than those of Sweden (0.70) and Germany (0.27) and Korea (–1.02), but their difference from the United States’ score (–1.79) was relatively small, albeit still noticeable. These findings suggest that the modification impact of Hong Kong and Macao is not comparable with those of the social democratic regime, the conservative regime, the developmental welfare state or, roughly speaking, the liberal regime. It is noteworthy that the United States’ score was much lower than Korea's, implying that the developmental welfare state was more powerful than the liberal regime with respect to reforming the social structure of a capitalist system.
The modification effects of social policy on the capitalist social structure
Notes: a overall modification, b decommodification, c autonomy promotion, d class abatement.
e The means and standard deviations are derived from the 18 welfare states in Esping–Andersen's (1990) study.
f For most indicators, the figures are the mean values of their actual percentages over the 2000s with the exception of Indicator 4 (only 2007 and 2009).
Source: Data for Hong Kong was extracted from Hong Kong Census and Statistics Department (2012b); Hong Kong Census and Statistics Department, Hong Kong Annual Digest of Statistics (various editions), Quarterly Report on General Household Survey (various editions); Hong Kong Food and Health Bureau (2004); The Hong Kong Council of Social Service (2012). Some of the data were provided directly by Hong Kong Social Welfare Department. Data for Macao and the other welfare states was extracted from Lai (Reference Lai2012).
As explained earlier, the modification impact can be divided into three broad categories: the decommodification of people, autonomy promotion and class abatement. Table 3 indicates that Hong Kong's social policy, with a decommodification score of –1.72, had a stronger effect on decommodifying people than Macao's (–1.93). Conversely, Macao's autonomy promotion score (–1.41) was higher than Hong Kong's (–2.37). This implies that Macao's social policy possessed a stronger capacity to secure personal autonomy than its counterpart in Hong Kong. Interestingly, Hong Kong's and Macao's social policies had about the same effect on combating class divisions with the same class abatement score (–2.73). In other words, the two states’ social policies had about the same achievement in promoting class equality.
Despite Hong Kong's and Macao's overall modification scores being similar, some indicators related to the modification impact have a remarkable difference. While Macao had a higher level of public healthcare expenditure (Indicator 3), a smaller proportion of poor people (Indicator 6), and a lower degree in income disparity (Indicator 9), Hong Kong had a higher level of public social spending (Indicator 8) and a larger scale of public in-patient services (Indicator 4).
Conclusions and discussion
Three main conclusions can be drawn from this study. First, Hong Kong's and Macao's social policies are marked by a relatively low modifying effect on the capitalist social structure.
Second, on the whole, these two city-states’ patterns of welfare mix have much in common: their scales of public welfare provision in terms of the level of public social expenditure were relatively moderate; their markets and families played a significant role in addressing certain domains of social needs; and their NGOs are responsible for the delivery of a major proportion of personal social services which were heavily subsidised by the governments.
Third, Hong Kong and Macao do not fit with the social democratic, the conservative, the liberal welfare regimes or the (inclusive) developmental welfare state as their modification impacts are noticeably lower than those of the prototypes of the models. Given that Macao is identified as a prototype of the regulatory welfare regime, it is empirically justified to classify Hong Kong into the same regime type as the levels of their modification effect on the capitalist social structure are similar. As a result, it is suggested that both Hong Kong and Macao are grouped in the regulatory welfare regime, which is characterised by its low degree of the modification impact on the capitalist social structure compared with the dominant welfare models. To put it another way, the two city-states retain the central features of capitalism to a significant extent, although their social policies have developed for several decades. This is directly related to the comparatively moderate scale of government intervention in social affairs and the market system.
This article makes two major contributions to the research on social policy in East Asia. First, this study is a pioneering attempt to compare the welfare regimes of Hong Kong and Macao. The conclusions add to the understanding of East Asian social policy or welfare regimes: the idea of a regulatory welfare regime appears to be a more accurate label for the peculiarity of the two city-states’ social policy than the other, dominant, models. Second, the article suggests a useful framework for comparative study in both the West and the East. The framework is concerned with the overall modification impact of social policy on a capitalist social structure, which is a relatively comprehensive concept for comparative analysis and, we suggest, more informative than any single criterion used in a comparative study such as ‘decommodification’ (Esping-Andersen, Reference Esping-Andersen1990; Bambra, Reference Bambra2005) and ‘benefit equality’ (Castles and Mitchell, Reference Castles, Mitchell and Castles1993).
