Introduction

There has been a substantial growth in the literature on global social policy (GSP) in recent years. Most authors acknowledge that the forms that GSP takes are distinct from those of national welfare states, since GSP depends on the activities of an array of international organisations and transnational actors (Reference Yeates and YeatesYeates, 2014). Deacon (Reference Deacon and Utting2006) has defined GSP as the mechanisms, policies and procedures used by intergovernmental and international organisations, working with other actors, to both influence and guide national social policy and to provide for a supranational or global social policy. In this second sense, GSP therefore constitutes global social redistribution, global social regulation and global social rights. Examples given by Deacon (Reference Deacon and Utting2006) for each of these respectively that are already being developed include emerging mechanisms for global social transfers such as the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM); core labour standards and the United Nations (UN) global compact on transnational corporations; and the advancement up the UN agenda of social rights and their monitoring and enforcement through ‘soft law’. Similarly, George and Wilding (Reference George and Wilding2002, p. 192) have argued that, ‘global social policy will be multi-dimensional – a mix of regulation, redistribution, provision of services and guaranteeing of basic rights.’

A number of authors have argued for the development of more comprehensive and social democratic forms of GSP (Reference DeaconDeacon, Hulse, & Stubbs, 1997; Reference DeaconDeacon, 2007; Reference MishraMishra, 1999; Reference Townsend, Townsend and GordonTownsend, 2002; Reference Townsend and DonkorTownsend & Donkor, 1996). Yet most authors have acknowledged the currently fragmented and piecemeal structure of global social governance, dispersed as it is across a wide array of actors and institutions (Reference Kaasch, Martens, Kaasch and MartensKaasch & Martens, 2015). Furthermore, in a period of neoliberal hegemony, when global economic integration and its related governance structures are designed in such a way as to facilitate ‘free’ markets and constrain social policy, it is particularly difficult to realise an expansive vision of global social policy.

This article seeks to explicate in a systematic way why GSP has taken the forms that it has. Although these forms are different to the forms taken by national welfare states, and in part for that reason, the article contends that we can gain insight into why it takes the forms it does by applying theories of the development of national welfare states to the development of GSP. There are various explanations for welfare state development, but three broad approaches have dominated the literature (Reference Huber, Ragin and StephensHuber, Ragin, & Stephens, 1993): those focused on processes of economic development, industrialisation and urbanisation; those focused on class struggle and political mobilisation; and those focused on the effects of differing political institutions. In the following sections, each of these three theoretical approaches is applied to GSP, in order to illuminate the nature of GSP, its likely future development, and the constraints upon such development. While the article draws on empirical evidence, the primary contribution is theoretical.

Economic development and global capitalism

Early welfare state theory focusing on economic development and industrialisation was largely functionalist in its approach, arguing that the welfare state is essentially the outcome of the ‘needs’ of industrial society (Reference WilenskyWilensky, 1975, Reference Wilensky1976). As Flora and Alber put it (Reference Flora, Alber, Flora and Heidenheimer1984, p. 38): ‘In the tradition of Durkheim, structural-functional differentiation is the fundamental process characterising modernisation.’ The key factors are therefore the increasing specialisation and division of labour associated with economic growth and industrialisation. Closely related to this are processes of urbanisation and increasing labour mobility as labour markets expand, which undermine ‘ascriptive bonds’ and therefore the security functions of the family. Human relationships are increasingly based upon exchange rather than close informal ties. These processes are seen as leading to a number of social problems, which can only effectively be tackled by the state. The development of factories leads to a change in working conditions that is much more likely to result in industrial accidents; labour contracts may be unrestrained, involving long working hours or child labour; the question of income security for those not employed is raised, including disabled people, children and their carers, older people, and the unemployed. In order to respond to these pressures, the state must control, supplement or substitute for the market (Reference Flora, Alber, Flora and HeidenheimerFlora & Alber, 1984, p. 41). This does not mean the state is against the market; rather, the state plays a key role in facilitating the market as its own institutional capacity develops alongside the extension of market relations.

On first consideration, the industrialisation thesis has little to tell us about GSP, since it suggests that welfare states will develop only where there is industrialisation. Thus, we may examine newly industrialised countries in order to see how social policy has developed in such countries and to compare this to the earlier processes of welfare state development in Western Europe, for example, but there is little in the industrialisation thesis per se that would lead us to expect the development of GSP, particularly since industrialised countries remain a minority of the world's states. In the existing industrialised (or ‘post-industrial’) welfare states, and in the context of economic globalisation, the literature has analysed how the state plays a role in both protecting citizens from, and in enhancing, global competition (Reference CernyCerny, 1990; Reference Esping-AndersenEsping-Andersen, 1999, Reference Esping-Andersen1996). Bonoli (Reference Bonoli2005) has also examined how the welfare state might adapt to incorporate the needs of new social risk bearers, in the context of more flexible labour markets and other developments. However, these are all national-level adaptations of existing welfare states.

Nevertheless, on the basis of industrialisation theory, it is possible to identify circumstances where continuing economic or technological development has tended to lead certain industrial economies to become more closely integrated in a way that would have possible implications for the further development of welfare provision. In particular, greater labour mobility between nations rather than within them has lead to international cooperation through measures such as reciprocal agreements on social security entitlements and the portability of health insurance or treatment arrangements. This tends to take place through agreements between sovereign nation states, although there are also supranational arrangements which supersede this.

Where this happens it is often on a regional basis, where similar and geographically adjacent countries become more economically interdependent. The best example of this is the European Union (EU), where there is an explicit project to create a single European market. Common European trade policies, and particularly the creation of the single currency, are indicators of how far this process has gone. As Hirst and Thompson (Reference Hirst and Thompson1999) observe, the current period has seen the national state cede power both ‘upwards’ and ‘downwards’, and in the EU this has led to a system of ‘multi-level governance’ (Reference Hooghe and MarksHooghe & Marks, 2001). Yet, while there have been some ‘positive’ forms of social policy integration in the EU, aimed at creating common standards or policies, most EU social policy integration has taken a ‘negative’ form; that is, it has been driven by efforts to remove barriers to competition and the single market, particularly in labour markets and social policies that affect them (Reference AndersonAnderson, 2015).

While the limited form of ‘positive’ regional social policy that can be observed in the EU might be argued to be consistent with the industrialisation thesis, this thesis has difficulty explaining global social transfers from industrialised to industrialising or mainly agrarian countries. Given its reductionist nature, the industrialisation thesis is not in any case a sufficient explanation of the development of national welfare states, let alone GSP. We therefore need to examine social and political relationships, rather than simply the level of economic and technological development. The next section discusses class struggle and political mobilisation (including, briefly, the potential for global alliances between industrial workers and farmers). Prior to that, however, those aspects of neo-Marxist welfare state theory concerned with the level of economic development are discussed.

Some approaches in the Marxist tradition have also contained a functionalist element, pointing to the role of the welfare state in maintaining a healthy and educated workforce capable of meeting the needs of advanced capitalism. However, Marxists have tended to emphasise the contradictory nature of the welfare state, in both attempting to meet the accumulation needs of capitalism while also legitimising the system by meeting certain needs of the working class within it (Reference GoughGough, 1979; Reference OffeOffe, 1984; O'Connor, Reference O'Connor1973). These arguments culminated in an extensive literature concerning the apparent crisis of the welfare state in the 1970s and 80s. From a Marxist perspective, this crisis was often seen as a crisis of capitalism itself, prompting a debate about the role of the welfare state in a capitalist economy where growth rates significantly below those of the post-war period were coming to be seen as the norm. These analyses were written in a period largely prefiguring the phase of economic globalisation that began in the 1980s. Barriers to international trade and investment have been substantially reduced during this period, in a process that might be conceived as an attempt to resolve earlier crises of national capitalism, but which in 2007–8 resulted in a global economic crisis. We might therefore ask whether there is something about capitalism in its globalising phase that suggests a need for some form of GSP.

By the nature of such theories, there is little in either the industrialisation thesis or in the functionalist aspects of its Marxist equivalent that would lead us to expect capitalism to provide more than that which is functionally necessary for its continued existence and smooth working. However, it is possible to conclude that the existing low levels of GSP are not functional for capitalism. This argument relates particularly to the work of Polanyi and Marxist writers such as Muller and Neususs. Polanyi (Reference Polanyi2001) argued that the market operated on the basis of a ‘fiction’ that labour was like every other commodity, in that it was produced solely for the purpose of exchange. This fiction meant that, without state intervention to provide social security for those who could not find a buyer for their labour, capitalism was unable to secure its own reproduction. Similarly, and following Marx's analysis in Capital (Reference Marx1976, pp. 389–416), Muller and Neususs (Reference Muller, Neususs, Holloway and Picciotto1978) argued that the English Factory Acts were necessary to ensure the reproduction of the labour force, since the unrestrained competitive struggle by the employing capitalists threatened to destroy it through long hours, dangerous working conditions and child labour. In other words, capitalism could destroy itself without state intervention.

Current processes of globalisation require internationalising firms and their governments to pay greater heed to the political and social conditions of other countries, particularly developing ones. Processes of economic globalisation tend to widen inequalities both within and between countries (Reference Holden and YeatesHolden, 2014a), and these may become dysfunctional if not moderated. If the transnational corporations (TNCs) of the developed world want to invest in developing countries, thus making use of the cheaper labour there, they must have stable conditions for investment and a fit and healthy workforce, and one with basic educational skills. It has been argued (by Cameron, Reference Cameron1978, for example) that welfare states expanded in order to respond to the social risks posed by more open economies. Thus, it can be argued that a more competitive global environment actually provides an imperative for welfare state expansion (Reference Holden and YeatesHolden, 2014b) and, potentially at least, this might be the case at the international level as well as the national level. Townsend (Reference Townsend, Townsend and Gordon2002, p. 19), for example, has noted that growing international social polarisation may bring with it ‘social self-destruction’. From a Marxian perspective, Cammack (Reference Cammack2004, p. 192) has argued that the World Bank's anti-poverty policies are linked to the needs of capitalism, aiming at the creation of ‘a global proletariat, on a wage of two dollars a day, with a reserve army of labour acting as a disciplinary force’.

The expansion of the global market without the complementary globalisation of governance arrangements may also undermine state capacity at the national level, even where it advantages TNCs. The increased opportunities for tax avoidance and evasion evident in the contemporary global political economy provide one pertinent example. States and international organisations are struggling to effectively contain such processes, particularly via the Organisation for Economic Cooperation and Development's (OECD) ‘base erosion and profit shifting’ programme (OECD, 2017). Wilensky (Reference Wilensky1975, Reference Wilensky1976) argued that as capitalism evolved so too did institutional capacity, and logically this may be the case at the global level too. Yet current international tax initiatives, such as those of the OECD, seem to fall short of what would be necessary to fully restore state capacity, and are not aimed at building global tax systems.

Theories of economic development therefore provide some explanation for the emergence of forms of GSP aimed at ensuring stability and the incorporation of developing countries into the world market. They also point to aspects of the global political economy that may be dysfunctional, both for individual states and for the system as a whole, such as opportunities for tax evasion. However, such theories can do little more than tell us whether the preconditions exist for the further development of GSP. Whether such developments take place and, crucially, the form they take, will be determined by political factors. The next section looks at theories of class struggle and political mobilisation, while the following section examines questions relating to political institutions.

Class struggle and political mobilisation

Most theories of welfare state development accord some importance to political mobilisation. For modernisation theory, the development of mass democracy was a key factor in the development of welfare states because it allowed working class demands to be heard (Reference Flora, Alber, Flora and HeidenheimerFlora & Alber, 1984) (although the development of the Bismarckian welfare state demonstrates that democracy is not a precondition for governments to make concessions to the working class). For T.H. Marshall (Reference Marshall1950), the emergence of ‘social rights’ was the culmination of the development of citizenship. Marxist theories argue that welfare states developed partly to head off revolution or instability caused by mass revolt. Other theories, such as Korpi's ‘power resources’ model (Reference Korpi1983), emphasise the importance of political parties to the development of welfare states.

Korpi's model takes a left social democratic approach, emphasising the importance of left parties and organised workers’ movements in the development of the welfare state, particularly where left parties are able to gain and hold office over a long period of time. As Pierson (Reference Pierson2006, p. 31) puts it, in the power resources model: ‘The more successful the forces of the organised working class, the more entrenched and institutionalised will the welfare state become and the more marginalised will be the principle of allocation through the market.’ Esping-Andersen (Reference Esping-Andersen1990, p. 111) built on this approach in explaining the development of different welfare regimes. As Esping-Andersen (Reference Esping-Andersen1990) points out, it is not simply the overall level of welfare expenditure which is influenced by the relative power of left and other parties, but the type of welfare state that is created.

This article does not attempt to measure in any precise way the current or likely future influence of left parties, but rather is concerned with the politics and ability to organise transnationally of those parties and movements that may be considered as part of the ‘left’, and which therefore, in the past at least, would have favoured an expansion of the welfare state. Although it is not only left parties that have been associated with welfare state expansion, in a period when neoliberalism has been the dominant political ideology, it is a reasonable assumption that any radical welfare state expansion, including the further development of GSP, would be more likely to be associated with the left than with any of the other mainstream political currents.

Globalisation has often been considered to have led to welfare state retrenchment at the national level, as national governments have had to take account of the increased mobility of businesses by creating more ‘investment friendly’ environments (see, for example, Reference MishraMishra, 1999). A more sophisticated approach, however, sees national governments as both shaping and responding to growing economic interdependence, and in the process reconfiguring the welfare state, rather than simply cutting it back. It has been argued that such reconfiguration has led to the development of what may be called the ‘competition state’. Phil Cerny (Reference Cerny1990, p. 179) has conceptualised the competition state as the shift away from the maximisation of welfare within the nation to ‘the promotion of enterprise, innovation and profitability in both private and public sectors’. This entails not the withdrawal of the state from social and economic life, but rather a pivotal role for it in creating opportunities for the operation of the market. The ‘competition state’ model of social policy therefore seems to run counter to a social democratic variant of GSP, by seeking to align the national welfare state with the needs of a competitive economy, rather than to moderate the outcomes of global competition as social democratic aspirations for GSP would do.

The competition state may take many forms, but mainstream left parties increasingly pursued a version of this from the 1990s onwards. While there are differences between social democratic parties across Europe reflecting differences between national political economies, a shift towards a more pro-business / pro-market stance took place from the 1990s, involving restrictions on the growth of public expenditure, and supply-side rather than demand-side approaches to unemployment (Reference Driver, Martell and SchmidtkeDriver & Martell, 2002; Reference Hall and SchmidtkeHall, 2002). It has been argued that such developments have led to social democracy suffering from what Finlayson (Reference Finlayson1999, p. 274) has called the ‘absence of a motivational ethical core’, or what Schmidtke (Reference Schmidtke and Schmidtke2002, p. 16) calls the ‘lack of a mobilising normative goal’. The fight against global poverty and inequality has the potential to provide such a mobilising goal. Furthermore, as Mishra argues (Reference Mishra1999, p. 114), acting alone even powerful developed countries may not be able to resist pressures for deregulation and the scaling down of social standards which affect their own welfare systems. There is therefore a clear rationale for social democrats to take a more global approach to questions of social justice. However, despite the elaboration of a global social democratic vision by some academics (Reference HeldHeld, 2004; Reference Holden, Hudson, Needham and HeinsHolden, 2017; Reference PatomakiPatomaki, 2000), such platforms have not generally been adopted by left parties or mass movements.

There has been a relative swing to the left in many countries following the 2007–8 financial crisis, manifested in the growth of new left parties such as Syriza in Greece and Podemos in Spain, as well as in a shift to the left in some established social democratic parties like the British Labour Party. Yet the experience of Greece has demonstrated the difficulty of forging a common movement across countries, even within the EU (Reference Papadopoulos, Roumpakis, Romano and PunzianoPapadopoulos & Roumpakis, 2015). No global movement with the coherence of the international socialist and communist movements that challenged capitalism prior to and during the creation of the first welfare states exists today. This has particularly been a feature of the world political situation following the collapse of the Soviet Union and the weakening of traditional forms of social democracy in the 1980s and 1990s, with Cerny (Reference Cerny2010, pp. 128–156) arguing that neoliberalism has become a ‘hegemonic paradigm’. There is evidence, nevertheless, that this neoliberal hegemony is beginning to weaken, an issue discussed further in the next section.

There are examples of trades unions organising effectively on a transnational basis (O'Brien, Reference O'Brien and Yeates2014, p. 135; Reference YeatesYeates, 2002, p. 78), and Evans (Reference Evans2010) outlines the considerable potential that exists for labour organisations to build on this. However, while there have been significant moves towards transnational organisation by both labour and business interests, as Farnsworth shows (Reference Farnsworth2004), globalisation has increased the power of capital by increasing its mobility. At the global level in particular, the power and the demands of labour have so far generally been subordinated to those of business (Reference FarnsworthFarnsworth, 2005). While most transnational labour organisations appear to have accommodated their demands to those seen as acceptable within the current pro-market consensus, transnational business has been more successful in having its preferences adopted by international organisations and governments alike. These preferences are not for the complete absence of social policy, but rather for social policy measures (for the most part at the national level) which enhance competitiveness and the skills of the workforce, while providing minimum levels of support for workers who genuinely cannot find work (Reference FarnsworthFarnsworth, 2005).

As Pierson (Reference Pierson2006, pp. 39–40) points out, there is substantial evidence that classes and social groups other than the traditional working class have played a major role in the development of welfare states, as have parties other than social democratic ones. In this context,

In particular, a number of writers have pointed out the importance of agrarian interests in influencing welfare settlements (Reference CareyCarey, 2005; Reference Esping-AndersenEsping-Andersen, 1990, pp. 30–31; Reference FaheyFahey, 2002). The nature of alliances between agrarian and other interests is crucial in this respect, with alliances between agrarian interests and those of workers or social democratic movements likely to give rise to more progressive outcomes than agrarian alliances based on nationalism or with more middle-class interests. This would seem to be particularly important to the development of GSP, given the largely agrarian nature of many developing economies. Yet the forging of meaningful alliances between developed country workers and farmers on the one hand, and those in developing countries on the other, seems especially difficult. In particular, farmers in the developed countries (who often receive large subsidies from their governments) may see their interests as being diametrically opposed to those in developing countries. There may be some scope for transnational alliances of farmers, possibly together with workers, at the world regional level, given regional trade integration projects, such as that of the EU. However, even in the EU, such alliances may be difficult to form, when farmers often see their interests lying in maintaining subsidies rather than allying with worker interests, and agricultural and other workers from one EU member state may regard those from others as competitors for jobs rather than as allies.

One feature of the current period is the growth of civil society organisations and social movements. Smith and Wiest (Reference Smith and Wiest2012, p. 46) estimate that the number of active international non-governmental organisations (INGOs) grew from less than 1000 in the 1950s to nearly 20,000 by the 2000s. Of this overall population of INGOs, the number committed to social changeFootnote ¹ increased from 127 to over 1000 during the same period, while the number of transnationally organised labour unions increased at a slower rate, from 39 to 87 (Reference Smith and WiestSmith & Wiest, 2012, p. 59). Using the same data set, Kruse and Martens (Reference Kruse, Martens, Kaasch and Martens2015) show that the numbers and resources of INGOs working on global social governance issues has also grown substantially during the post-war period, and that increasingly they do this within a broad global justice perspective.

The creation of the World Social Forum (WSF) process in 2001 provided a means for such civil society groups to collaborate and constitutes a regular counterpoint to the meetings of the business elite's World Economic Forum (Reference Navarro and SilvaNavarro & Silva, 2006). By the 2005 forum, the number of participants had increased to 155,000 (Reference Navarro and SilvaNavarro & Silva, 2006) from 135 countries (Reference Smith, Karides, Becker, Brunelle, Chase Dunn, della Porta and VazquezSmith et al., 2016, p. 52), and regional, national and local forums have been held around the world. However, the WSF's purpose is to provide spaces within which ideas and tactics can be exchanged and developed, rather than to form a single movement with a common programme (Reference Smith, Karides, Becker, Brunelle, Chase Dunn, della Porta and VazquezSmith et al., 2016). Although understanding the connections between global and local processes, and seeing themselves as part of a global movement, WSF participants often see the strengthening of local communities as more important than national or global reforms (Reference Smith, Karides, Becker, Brunelle, Chase Dunn, della Porta and VazquezSmith et al., 2016, p. 133). Furthermore, despite the participation of labour organisations and oppressed groups, students, professionals and those with higher levels of formal education are disproportionately represented at the forums, reflecting the structural barriers to participation by more marginalised people (Reference Smith, Karides, Becker, Brunelle, Chase Dunn, della Porta and VazquezSmith et al., 2016, pp. 58–61).

Conceived at its broadest, the global movement against neoliberal capitalism contains within it a spectrum of groups and opinions, from protectionist unions and small farmers in the developed countries to guerrilla groups and landless labourers in the developing countries. It is therefore most able to construct coherent coalitions when conducting focused single-issue campaigns (Reference YeatesYeates, 2002, p. 82). As Patomaki and Teivainen have noted (Reference Patomaki and Teivainen2004, p. 114), ‘the movement’ has seldom gone beyond the tendency of some participants to ‘name the enemy in order to unify the heterogeneous global civil society and create a basis for efficacious actions’, usually against the international institutions. ‘The movement’ thus influences the course of GSP primarily by campaigning against what it finds unacceptable in the agenda of the international institutions and by challenging their legitimacy, rather than by promoting a common global vision. While important ideas, and specific platforms such as the G19 Declaration and the Bamako Appeal,Footnote ² have emerged from the WSF process, in line with the way the forums work, these statements are not endorsed by all participants within them and are not the ‘official’ views of the forums.

However, despite their largely undemocratic nature (discussed in the next section), international institutions have to some extent acted as a conduit for reform demands, much as national parliaments often did in the period of initial welfare state development. The attention of NGOs and social movements has shone a ‘spotlight’ on international institutions, with their activities now more closely monitored and visible to a wider public, a development that may exercise a brake upon some of the worst excesses of neoliberal policy. Furthermore, some civil society actors are increasingly engaging with international institutions in a more sophisticated way. In contrast to the increasingly challenged legitimacy of the intergovernmental institutions, international NGOs often seem to carry particular legitimacy in the current period. The intergovernmental institutions, arguably therefore, do need to respond to this challenge to their legitimacy. Therefore, in order to understand more fully the forms of GSP that are emerging, and the scope for further development, we need to turn to an analysis of political institutions.

Political institutions and international organisation

Some writers have placed particular emphasis on the role of political institutions in the development of the welfare state (Reference Bonoli and PiersonBonoli, 2001, Reference Bonoli2005; Reference Flora, Alber, Flora and HeidenheimerFlora & Alber, 1984; Reference Huber, Ragin and StephensHuber et al., 1993; Reference ImmergutImmergut, 1992). Institutionalists have pointed to the importance of both temporal aspects of institutional development, such as sequencing and path dependency (Reference PiersonPierson, 2004), and to the effects of different institutional configurations on policy outcomes. In regard to the latter, Bonoli outlines (Reference Bonoli and Pierson2001, p. 239) how political constitutions which concentrate power with the executive branch of government have generally been found to be associated with big welfare states, whereas more fragmented political systems, which include a greater number of veto points within them, have tended to produce smaller welfare states. Fragmented systems, such as that of the USA, have provided opportunities for anti-welfarist groups to prevent the adoption of social policy programmes or water them down. However, the development of GSP is at least as dependent on international political processes as it is on those occurring in individual states.

Kaasch (Reference Kaasch, Kaasch and Martens2015) points out that GSP actors are not necessarily institutionalised in any narrow, constitutional, sense. Instead, a multiplicity of intergovernmental and non-governmental actors are engaged in changing and often horizontal relationships with each other. As Kaasch notes (Reference Kaasch, Kaasch and Martens2015), to the extent that there is some kind of structured, hierarchical, system, this is provided by the United Nations (UN) system, which encompasses a wide range of intergovernmental institutions. While these institutions have a degree of autonomy within their specific areas of responsibility, veto points (and other means of influencing outcomes) therefore usually lie with national governments, and the capacity to utilise these is likely to reflect the differences in wealth and power between governments. A range of such international institutions may have a role in the current and future delivery of GSP, including the International Monetary Fund (IMF), the World Bank, the World Trade Organisation (WTO), the World Health Organisation (WHO) and the International Labour Organisation (ILO) (Reference DeaconDeacon, 2007). The current UN-based system of global social governance therefore involves a range of overlapping and sometimes competing agencies. This fragmentation may exert a brake on the further development of GSP, in a manner that is consistent with the thesis outlined by Bonoli and others.

At its creation in 1945, the UN was conceived as being at the apex of the international system, and its Economic and Social Council (ECOSOC) was to have coordinated the work of other agencies, including the IMF and the World Bank. However, the latter organisations are legally separate and in practice have operated independently from the UN. Patomaki and Teivainen (Reference Patomaki and Teivainen2004, pp. 18–19) show how from the 1980s the USA has often both marginalised and subordinated UN agencies, partly through its dominance of funding. The role of the WHO in global health policy (Reference LeeLee, 2009) and of other UN agencies in global poverty reduction (Reference HulmeHulme, 2015), for example, has tended to be challenged by that of the World Bank and IMF, primarily because of the extra resources the latter organisations can command. It is the World Bank and the IMF that provide the main existing mechanisms of global redistribution, and these transfers usually come with strings attached. The two organisations are fundamentally neoliberal in their approach. Together with the WTO, O'Brien, Goetz, Scholte, and Williams (Reference O'Brien, Goetz, Scholte and Williams2000, p. 233) have referred to them as ‘the iron triangle of liberalism’, and note that their ‘rule-creating and rule-supervisory decisions’ give them a degree of influence that contrasts with that of institutions such as the ILO ‘which must rely on moral suasion and argument’ (O'Brien et al., Reference O'Brien, Goetz, Scholte and Williams2000, p. 11).

The IMF provides a useful example of the importance of both a focus on path dependency in the historical development of institutions and of veto power, both of which privilege the USA. During its creation in 1944, the IMF was designed so that the decisions of its Executive Board were made based on weighted voting shares that reflect a member state's quota, which is equivalent to the amount of capital contributed to the Fund's liquidity. The IMF claims that it follows the principles of non-discrimination in setting members’ quotas and their associated voting shares. However, since the USA has the largest economy in the world, it has the largest quota and therefore the largest share of votes. Even after reforms agreed in 2010 to redistribute quotas and voting shares are complete, the USA will have 16.5% of total voting shares (IMF, 2016a). The gap in voting power between the USA and other member states is illustrated by the fact that 46 Sub-Saharan African states are represented by two Executive Directors with a combined voting share of 4.5% (IMF, 2016b; see also the volume edited by Carin & Wood, Reference Carin and Wood2005).

The IMF argues that decisions on loan conditionality are apolitical because agreements are drawn up by its staff on the basis of technical criteria. However, Momani (Reference Momani2004) has shown that the USA has intervened to secure its own specific foreign policy goals. IMF staff in any case work within a neoliberal ideology and culture which continues to subordinate social policy goals to the IMF's view of what constitutes macroeconomic stability (Reference Farnsworth and IrvingFarnsworth & Irving, 2017; Reference Kentikelenis, Stubbs and KingKentikelenis, Stubbs, & King, 2016). Furthermore, many important IMF decisions require the support of 85% of votes, giving the USA an effective veto. Decisions over which the USA can exercise its veto include measures to increase the number of Executive Directors and to adjust quotas. The USA therefore has a veto over precisely those measures that would be necessary to give developing countries more influence in the IMF and transform it in a more egalitarian direction. Even in those cases where only a simple majority vote would be necessary or where the Board makes decisions on the basis of an apparent consensus, the size of the US vote has proven essential in building winning coalitions. Meanwhile, the World Bank's organisational model is, as Patomaki and Teivainen put it (Reference Patomaki and Teivainen2004, p. 62), ‘essentially a replica of the IMF system’.

On the face of it the WTO is more democratic than the IMF and World Bank, in that trade negotiations are conducted on the basis of bargaining and agreement between nominally equal national governments. However, powerful states have utilised a range of measures to exercise influence over weaker ones, from greater access to expert advice during highly technical negotiations to direct threats and manipulation on a bilateral basis (Reference Jawara, Kwa and SharmaJawara, Kwa, & Sharma, 2004). Furthermore, WTO dispute panels are composed of trade experts who reach decisions on the basis of a narrow interpretation of WTO rules, with little regard to social issues. The negative actual or potential impact of WTO agreements such as the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) and the General Agreement on Trade in Services (GATS) on health and other areas of social policy have been well documented (see, for example, Reference CorreaCorrea, 2002; Reference Holden and YeatesHolden, 2014b; Reference Price and PollockPrice & Pollock, 2003).

The emergence of important middle-income countries such as the ‘BRICS’ (Brazil, Russia, India, China and South Africa) has allowed some developing countries to take a more effective role in global social governance (Reference Surender, Urbina-Ferretjans, Kaasch and MartensSurender & Urbina-Ferretjans, 2015). This has included the movement of the G20 to the centre of global governance following the 2007–8 financial crisis (Reference Kirton, Kulik, Bracht, Kaasch and MartensKirton, Kulik, & Bracht, 2015) and Brazil's role in the formation of the Sustainable Development Goals (Reference HulmeHulme, 2015, p. 149). Brazil similarly played a key role in the ‘G20’ alliance on agriculture in WTO negotiations (not to be confused with the G20 previously referred to). Although such coalitions have tended to be transitory and subject to fragmentation in the past, the WTO ‘G20’ was able to sustain an unprecedented level of cohesion in the Doha round of trade negotations (Reference Narlikar and TussieNarlikar & Tussie, 2004), although this has led to deadlock. However, while the growth of the BRICS’ economies may in time threaten the dominance of the USA in institutions such as the IMF, there is little evidence currently that these countries are likely to favour a more comprehensive approach to GSP. Similarly, while the creation of new development banks by these countries is threatening the hegemony of the IMF and World Bank as institutions (Reference HulmeHulme, 2015, p. 202), this process is at a relatively early stage. To the extent that it undermines the power of the IMF and World Bank, it will add to rather than decrease the fragmentation in international institutions. Furthermore, the stalemate in the Doha round led the USA and the EU in particular to seek to partially bypass WTO negotations, pursuing instead a number of bilateral and plurilateral negotiations, including the Trans-Atlantic Trade and Investment Partnership (TTIP), the Trans-Pacific Partnership Agreement (TPP) and the Trade in Services Agreement (TISA). While putting some of these new agreements in doubt, the election of Donald Trump as US president is likely to further fragment the governance of global trade, as the US turns towards even greater bilateralism. The inclusion of investor-state dispute settlement (ISDS) mechanisms in agreements creates a series of new veto points and gives rights to corporations to directly initiate disputes with governments (Reference Hawkins and HoldenHawkins & Holden, 2016).

However, despite the neoliberal bias of the ‘iron triangle’, recent developments offer some evidence in support of those who have highlighted the importance of ideas to international institutions and their changing practices over time (Reference Boas and McNeilBoas & McNeil, 2004). The World Bank's social policies have moved towards a more comprehensive anti-poverty and social protection strategy (Reference DeaconDeacon, 2007, pp. 24–45; Reference HulmeHulme, 2015, pp. 178–182), including in its 2012–2022 Social Protection and Labour Strategy (World Bank, 2012). Similarly, since the financial crisis of 2007–8, the IMF has begun a re-examinition of some of its core ideas, which has included a greater willingness to challenge the now extreme levels of inequality in many countries (Reference Ostry, Berg and TsangaridesOstry, Berg, & Tsangarides, 2014), the acceptance of the benefits of public investment (Reference Abiad, Furceri and TopalovaAbiad, Furceri, & Topalova, 2015), and even a questioning of some of the key tenets of neoliberalism (Reference Ostry, Loungani and FurceriOstry, Loungani, & Furceri, 2016). Despite this, Nunn and White (Reference Nunn and White2016) show that there has been relatively little change in IMF practice, at least in relation to inequality, while Farnsworth and Irving (Reference Farnsworth and Irving2017) demonstrate that even ideational change has been limited. The IMF, the World Bank and the WTO have also increasingly engaged with social movements and civil society actors, leading to a greater degree of transparency and consultation, although this too has been limited (Reference Hannah, Scott and WilkinsonHannah, Scott, & Wilkinson, 2017; O'Brien et al., Reference O'Brien, Goetz, Scholte and Williams2000; Reference ScholteScholte, 2012; Reference Woods and NarlikarWoods & Narlikar, 2001). ILO policy has shifted towards a more resident-based, rather than simply worker-based, model, especially through the development of the Social Protection Floors Recommendation (Reference DeaconDeacon, 2013).

This tentative shift in ideas appears to have gained momentum since the UK's referendum decision to leave the EU and the election of Trump to the US presidency in 2016. The IMF, the World Bank and the WTO have, for example, published a joint paper for the G20 arguing for greater attention to the social policies necessary to mitigate trade adjustment (IMF, World Bank & WTO, 2017). Yet, despite an apparently genuine search for a new consensus on social policy, these ideas remain subject to the institutional constraints discussed above. While their emergence indicates the importance of ideas, they also arguably demonstrate the continuing importance of broader economic forces and of political mobilisation, in that the extreme growth of inequality in a context of economic crisis and stagnation is now seen as dysfunctional for the global economy, and may itself be seen as a factor underlying the political discontent manifested in the votes for Brexit and Trump (Reference Holden, Hudson, Needham and HeinsHolden, 2017). The political shifts represented by Brexit and Trump, while reflecting a shift to the right rather than the left, have been seen by defenders of the neoliberal global order as an unwelcome challenge, and may well be the cause of a partial rethinking of the neoliberal project in an attempt to head off discontent (see for example, O'Sullivan, Reference O'Sullivan2016).

The important role that ideas can play in international institutions is linked also to the sometimes disproportionate role that individuals can play within them. Deacon (Reference Deacon2013), for example, notes that this was an important aspect of the development of the ILO's Social Protection Floors Recommendation. However, despite significant elements of multi-level governance (Reference CernyCerny, 2010), and the partially autonomous capacity to act of international bureacrats, the enduring importance of the nation state within the world polity makes substantial transfer from the richer parts of the world to the poorer ones difficult, although not impossible. The development of the very idea of GSP challenges the normative bases of social policy and of welfare states as they have developed, which have primarily been about redistribution within states rather than between them. This has been reflected historically in the academic subject of Social Policy, which until relatively recently had largely ignored questions of international development and global social justice. It is also reflected in the fact that while domestic social expenditure constitutes the largest element of public expenditure in the advanced capitalist countries, very few such countries have reached the UN target of 0.7% of Gross National Income (GNI) for international development assistance. While mean public social expenditure in OECD countries was 21.6% of Gross Domestic Product in 2014 (OECD, 2016), net official development assistance by members of the OECD's Development Assistance Committee was 0.29% of GNI in the same year (OECD, 2015).

Given the prevalent discourse within economically advanced countries about the affordability of welfare states and the need for fiscal consolidation following the 2007–8 financial crisis, a substantial rise in international redistribution is unlikely in the absence of a strong political movement in favour of it. However, global coalitions in favour of GSP are difficult to form. As argued above, to be more effective the global movement against neoliberalism would need to develop a clearer agenda. Moreover, those that stand to gain most from the further development of GSP are the poorest citizens of low-income countries, that is, of those states with the least power in the international arena. While there are also some positive signs within the developed world of a desire to effect greater global redistribution, and that emerging middle-income countries may be able to play a progressive role, in the current political climate progress is likely to be slow.

Conclusion

By its very nature, GSP takes different forms to those of national welfare states, and often must therefore be analysed using different concepts and approaches than those applied to national welfare states. Nevertheless, this article argues that by applying the broad theoretical approaches of the literature on national welfare-state development, we can illuminate the nature of GSP and gain insight into the reasons why it takes the forms that it does. The analysis suggests that the forms taken by GSP will continue to be piecemeal, minimalist and essentially neoliberal for as long as an effective global political movement in favour of a more extensive GSP is absent. A basic form of GSP exists based upon the institutions set up at the end of World War Two, and the increasing integration of the world economy provides a material basis for its further development, in that as capitalism develops at the global level it increasingly ‘needs’ mechanisms of regulation and governance which operate at this level. However, responsibility for social policy at the global level is currently fragmented across a number of institutions, and those that control most resources are currently structured so as to limit their use for more progressive and extensive GSP goals. Rather, their structuring reflects the existing balance of power within the world political economy, in which, despite important changes as a result of globalising processes, the nation state remains the primary political form. The inequalities of power that exist between states mean that the policies pursued by the most powerful states may not even be those that seem to be functional for the system in the long-term, but those that are in the perceived short-term interests of the most powerful. Moreover, given the attempts in advanced capitalist states to contain the expansion of the (national level) welfare state, there will be stiff resistance to proposals for the funding and maintenance of more extensive structures of global social redistribution, regulation and rights.

Ultimately, the building of a more progressive and extensive GSP is a political project. Deacon has pointed to the work of Gill (Reference Gill1993, Reference Gill2003) and Cox (Reference Cox1995) concerning the necessity and possibility of building a new counter-hegemonic bloc in the pursuit of a post-neoliberal world order, that may be constituted from a global alliance between progressive forces in the developed and developing world, and that may involve both governmental and civil society forces (Reference Deacon and UttingDeacon, 2006). For those in favour of a more progressive and extensive GSP, the task articulated in the following quote by Cox, highlighted by Deacon (Reference Deacon and Utting2006, p. 144), is surely accurate: