Business,labour and the costs of welfare state development

Abstract

This article presents a novel explanation for instances of business support for welfare state expansion. It emphasizes the importance of cost considerations in shaping business preferences and argues that their willingness to support and ability to oppose demands for increases in the generosity of social insurance programmes depends primarily on the extent to which labour unions accept that these increases are to be financed by workers themselves. Based on a comparative-historical analysis of postwar welfare state development in the Netherlands and the United Kingdom, the article shows that this willingness among others depended on the type of labour market risk, the margin for pay increases, and the extent to which social welfare initiatives were perceived as actual improvements to the social wage by labour unions.

Introduction

In recent years, the role of business in the postwar introduction and expansion of social programmes has developed into a popular research theme among welfare state scholars. This revival of scholarly interest has been accompanied by intense debate over the nature and extent of business influence over the postwar expansion of the welfare state. One side in this debate has criticized the traditional assumption of business-hostility to social protection by arguing that social policies not only impose costs and labour market rigidities on businesses but also provide benefits to them that have been sufficiently important to warrant active business support for their introduction and expansion (Mares, 2003; Martin and Swank, 2012; Swenson, 2002). Other scholars have been critical of these claims and hold that most instances of business support for social policy expansion can be explained by changes in political context that forced them to act as reluctant consenters to welfare initiatives (Gordon, 2014; Hacker and Pierson, 2002; Korpi, 2006; Paster, 2013).

This article provides an alternative explanation for instances of business support for welfare state expansion. Similar to both traditional labour-oriented accounts and recent critics of the business-interests scholarship, it holds that business groups mainly developed their position in strategic response to demands for increased protection against labour market risks from workers. Yet, it differs from those studies that contend that ‘the willingness of businesses’ to support, and ability to oppose, these demands depended primarily on the organizational strength of labour and its pro-welfare allies (Hacker and Pierson, 2002; Korpi, 2006). Instead, it argues that both depended foremost on the nature of these demands, in particular the extent to which labour unions were able to assuage business fears that welfare state expansion would be costly to businesses. And this ability, it shows, in turn depended on organized labour’s willingness to accept that this expansion was largely to be financed by workers themselves.

In addition, the article shows that there was a great deal of variation in organized labour’s willingness to do so. It identifies various factors that explain this variation and places particular importance on the extent to which labour unions viewed welfare initiatives as actual additions to the social wage. Whereas much of the literature assumes that labour unions generally supported the postwar expansion of the welfare state and were consequently willing to accept at least partial worker financing of this expansion (Eichengreen, 2007; Huber and Stephens, 2001; Siegel, 2005), this article shows that they by no means always did so. It consequently explains how this willingness affected both business attitudes and the political efficacy of business groups in matters relating to welfare state development. More specifically, it shows that business groups not only had less of an incentive to oppose worker demands for improved protection against labour market risks when union leaders accepted that the costs of this were largely to be borne by workers themselves, but they were also in a weaker position to resist labour’s demands for welfare state expansion under these circumstances.

The analysis illustrates these claims through a comparative-historical analysis of organized business’ response to postwar social insurance expansion in the Netherlands and the United Kingdom. As will be explained below, the British and Dutch labour union movements differed strongly from each other in their willingness to view the introduction and expansion of social insurance programmes as improvements to the social wage that needed to be financed by workers themselves. While several studies have noted the ‘the lack of consensus on the social wage’ in the United Kingdom (Crouch, 1999; Rhodes, 2000: 20), the extent of disagreement over who was to bear responsibility for the costs of public welfare initiatives and its consequences for the political feasibility of these initiatives has been far from fully recognized. Likewise, while various studies have observed instances in which Dutch labour unions ‘exchanged’ wage restraint for welfare state expansion (Eichengreen, 2007; Whiteside, 2006), the systematic nature of this exchange and its importance in facilitating the postwar expansion of the Dutch welfare state have been far from fully acknowledged.

In addition, the analysis shows that organized labour’s willingness to accept even partial worker financing of major welfare initiatives was also subject to strong variation over time in the two countries. Thus, while Dutch union leaders eventually came to accept the view that social costs constituted part of total wage costs, and therefore needed to be treated as such, they did not do so from the outset of the postwar period. And whereas many of Britain’s most powerful labour unions resisted even partial worker financing of most of the welfare in initiatives put forward by Labour in the 1960s, they did not resist worker co-financing of Labour’s early postwar welfare initiatives. The following section puts forward a theoretical justification for these differences. The empirical part of the analysis relies on both types of variation – cross-national and longitudinal – to illustrate how it affected the attitudes and influence of major business groups in the two countries.

In addition, the article draws on evidence from minutes of meetings and internal correspondence of the main business groups and labour unions in the two countries. In the United Kingdom, these were the British Employers’ Confederation (BEC), its successor the Confederation of British Industry (CBI) and the Trades Union Congress (TUC). In the Netherlands, these were the Christian Employers’ Federation (NCW) and its many Protestant and Catholic predecessors, the liberal Central Social Employers’ Federation (CSWV) and its successor the Federation of Dutch Industries (VNO), and their union counterparts the Christian Union Federation (CNV), the socialist Dutch Association of Trade Unions (NVV) and the Netherlands Catholic Trade Union Federation (NKV). These primary sources provide not only information on the attitudes and motives of these actors but also a clear overview of discussions over who was to pay for improved protection against labour market risks and, crucially, the repercussions this had for collective wage bargaining in the two countries.

Business interests and the costs of welfare state development

Most political observers would probably agree that businesses are generally reluctant to support the development of social insurance programmes, not only because they reduce workers’ dependence on the labour market but also, and perhaps primarily, because they threaten to impose a financial burden on firms. The importance of cost considerations in shaping business attitudes towards social insurance programmes has long been noted (Mares, 2003; Paster, 2013; Swenson, 2002). Yet, despite the prevalence of contributory modes of financing in most countries, recent studies on the role of business have paid little systematic attention to discussions between unions and employers over who was to bear responsibility for payroll contributions. As a result, they have failed to recognize the vast variation in the extent to which worker demands for increased public protection against labour market risks imposed a financial burden on businesses. Nor has there been much consideration for the possibility that their attitudes towards and influence over these demands may have depended on labour’s willingness to accept worker (co)financing of these programmes.

There are various possible reasons for this lack of attention, including the popular assumption that the fiscal incidence of a tax is the same whether its costs are carried by workers or employers (Riedl and Tyran, 2005). This article challenges this ‘invariance of incidence proposition’ by showing that business groups, labour unions and governments did care greatly as to whether the costs of social insurance expansion were to be carried by businesses or workers. ¹ It consequently outlines how discussions over who was to bear responsibility for payroll contributions affected both business’ willingness to support and its ability to oppose social insurance expansion. The latter can be explained as follows. Rapid increases in social insurance costs, unless offset by lower wage claims, have direct consequences for firm profitability, and thus for their ability to invest. As a result, businesses can be expected to find it easy to refer to this concern in order to exert pressure on governments. When labour unions, however, accept full worker financing of social insurance programmes, business will find it much more difficult to resist demands for welfare state expansion. To be sure, they may also refer to other concerns, such as the potential negative impact that improved provision against some labour market risks may have on labour supply (Esping-Andersen, 1990; Paster, 2013), but as these consequences are clearly more ambiguous (Howell et al., 2007), they can be expected to have much less of an impact on governments – especially during the area of welfare state expansion, when benefit dependency was not yet considered a major problem by policy-makers (Bean, 1994).

From this, we can draw the following expectations. First, business groups can be expected to take a more accommodating stance towards demands for improved protection against labour market risks when workers are willing to bear most of the costs of these improvements. Of course, they may still oppose particular welfare initiatives for other reasons than cost considerations (Paster, 2013). Yet, in such circumstances, they may find it more difficult to prevent these improvements from coming about. When labour unions are by contrast unwilling to accept even partial worker financing of social insurance programmes, employers will not only be much more likely to resist their development but, as there will be broad political concern over the consequences of improved welfare provision for the competitive position of business in this case, their resistance to their development is also much more likely to be successful. In sum, we can expect business’ willingness to support, and ability to oppose, welfare state expansion to have fluctuated significantly depending on labour unions’ willingness to accept worker responsibility for social insurance costs. And this is important, because labour’s willingness to do so varied significantly as well.

Business, labour and the notion of the social wage

To fully understand the importance of discussions over who was to bear responsibility for payroll contributions, it is important to emphasize that from the perspective of business, these discussions matter in two ways. First, it matters to them to what extent payroll contributions are carried by either workers or themselves. Second, it matters how these discussions affect subsequent wage negotiations. From the perspective of business, it makes little sense to impose worker (co)financing on labour unions when they respond to increased payroll contributions for workers by demanding higher wage increases than they otherwise would have done during subsequent wage negotiations. And vice versa, when labour unions agree to moderate their wage demands in order to facilitate welfare state expansion, this has the same material consequences as financing at least part of this expansion by increasing worker contributions.

In standard economic accounts of social insurance development, it is generally assumed that it does not matter who pays for social insurance costs, as any increases in worker payroll contributions tend to prompt increased wage demands. This article argues that this is not necessarily the case. It identifies three main factors that determine labour unions’ willingness to accept at least partial worker responsibility for social insurance costs and how they respond during subsequent wage negotiations. These include (1) the type of labour market risk. A major distinguishing feature of unemployment and work-related injuries and illnesses is that labour unions are far more likely to assign responsibility for their incidence to employers than they are with other major labour market risks such as old age and non-work-related injuries and illnesses (Alber, 1982). As the latter cannot in any way be attributed to employer actions (as is the case with unemployment) and do not occur while working under orders (as with work-related injuries and illnesses), labour union leaders generally found it easier to accept that responsibility for financing insurance against them primarily lay with workers themselves.

Of course, their willingness to do so does not only hinge on normative considerations. Economic conditions, such as (2) the scope or ‘margin’ for pay increases, matter as well. Under conditions that call for wage moderation, union leaders may find it much more difficult to accept even partial worker co-financing of social insurance expansion, and respond by demanding wage compensation for any loss of income that workers may experience as a result of higher payroll taxes. Of course, such a move would significantly increase the costs of social insurance expansion. This seemingly obvious point is not always recognized in the literature. Take for instance the popular tendency to explain the success of postwar European incomes policies by arguing that they rested on social policy expansion – and vice versa (Eichengreen, 2007; Huber and Stephens, 2001; Mares, 2006; Simoni, 2013). These ‘political exchanges’ were indeed common in Europe, but their role in persuading labour unions to participate in government incomes policies was quite minor, if only because such compensatory gestures conflicted with the very goals that were being pursued through these policies – which was to reduce the overall growth of labour costs. Wage settlements under which labour unions moderated their wage demands in exchange for social policy expansion were therefore much more common in periods of rapid wage growth than in periods that called for austerity and wage moderation. Yet even then, they were far from automatic. For starters, and as first emphasized by neo-corporatist studies from the 1980s, they were much easier to negotiate under centralized bargaining institutions (Golden, 1993; Schmitter, 1974) – although such institutions, as we will see, are by no means a necessary condition.

In addition, labour’s willingness to moderate wage demands in ‘exchange’ for welfare state expansion, to use or ‘reserve’ part of negotiated wage increases to finance improved benefits, or to accept increased worker contributions (which had the same material consequences in terms of labour costs) obviously depended on (3) the extent to which it viewed public welfare expansion as an actual improvement to the social wage. Contrary to what is often believed, this can by no means be taken for granted. Despite recent criticism of traditional class-oriented accounts of the welfare state, there has been little attention in the literature for the possibility and importance of consistent labour union disinterest in or opposition to public welfare expansion (Oude Nijhuis, 2013). As a result, there has also been little attention for the consequences of such opposition. For businesses, and as we will see, the most important consequence was that union leaders were much less likely to accept even partial worker responsibility for the costs of welfare initiatives when they felt that these did little to improve the security of their members. On the contrary, and as we will see, in such circumstances they were likely to attempt to make up for increased contribution levels by demanding higher than normal wage increases during subsequent wage negotiations. To businesses, this meant that they had either to accept further labour cost increases or face considerable industrial unrest.

The following analysis shows that this dilemma was particularly severe in countries, like the United Kingdom, with strong occupational labour unions, as these represented workers who were generally able to obtain generous private welfare provision, or simply had no need to secure themselves against certain labour market risks. Moreover, as organizations whose main organizational purpose was to defend the privileged position of their members, they also had a strong incentive to respond to the introduction of redistributive welfare initiatives by demanding higher than normal pay increases in order to make up for lost differentials during subsequent wage bargaining rounds. In this respect, they differed strongly from more inclusive types of labour unions, which were much more likely to view public and redistributive welfare initiatives as actual additions to the social wage. As a result of its tendency to view workers as a relatively homogeneous category (Korpi, 2006), the existing literature has paid little attention to the importance of this type of variation. As a result, it has also failed to notice how it affected the costs of redistributive welfare state expansion for businesses.

The following analysis of postwar welfare development in the Netherlands (where this type of labour unionism was almost completely absent) and the United Kingdom (where it was arguably more ingrained than in any other European country at the time) illustrates these consequences. The analysis shows that neither British nor Dutch unions were always willing to accept even partial worker financing of social insurance initiatives. Yet, it also shows that consistent union opposition to worker financing could only be found in the United Kingdom, and that British labour unions frequently threatened with wage repercussions in order to strengthen their opposition to unwanted social insurance initiatives. It contrasts this with the situation in the Netherlands, where the unions increasingly accepted that social insurance benefits constituted part of the social wage and, like other material benefits such as occupational provision, should be treated as deferred wages. The article consequently illustrates how this explains why British business groups were both more vocal and more successful in their resistance to postwar welfare initiatives than were their Dutch counterparts (Baldwin, 1990; Bridgen, 2000; Sluyterman, 2003; Therborn, 1989). To be sure, these differences cannot be attributed to variation in the political strength of business representatives in the two countries – at least not as measured by conventional indicators of their relative ‘power resources’ such as employer and union density (Lane, 1992; Visser and Hemerijck, 1997). Nor, as we will see, did they reflect different welfare preferences.

Cost considerations and the politics of welfare state expansion in the Netherlands and the United Kingdom

In both countries, business views of social welfare initiatives depended greatly on the extent to which these initiatives threatened to impose a financial burden on firms. To minimize these costs, the main business groups in the two countries all emphasized the need to opt for a contributory approach under which benefit entitlements were primarily to be financed by worker contributions, and opposed financing these entitlements out of general revenue (Baldwin, 1990; Cox, 1993). In its evidence to the Beveridge committee in 1943, the leading employer organization on labour affairs in the United Kingdom, the BEC, for instance, warned that social insurance programmes should be ‘based on the contributions of those who are to receive the benefits’ in order to prevent them from imposing ‘a heavy additional burden on industry’ or ‘find[ing] their way into prices’ (MRC, MSS.200.B/3/2/C216/5/50: Memorandum Beveridge report, 10 February 1943, 15; MRC, MSS.200/B/3/2/C216.5: Advanced copy Beveridge report, 11 January 1943). And during discussions over comprehensive welfare reform in the Netherlands some five years later, the largest employer federation there, the CSWV, likewise referred to the need to maintain economic competitiveness when it pointed out that ‘social costs are part of total wage costs and should therefore be treated as such’ (VNO, F15(3): CSWV inzake herziening sociale verzekering, 1 March 1948). To make sure that social insurance expansion did not undermine the position of Dutch industry, the CSWV therefore deemed it crucial to view social insurance entitlements as deferred wages whose expansion was to be financed out of the ‘margin for pay increases’ (in Dutch: loonruimte), meaning the extent to which businesses could afford to increase wages and non-wage benefits without undermining their competitiveness. There can be little doubt that the other British and Dutch employer organizations agreed with these statements. Nor can there be any doubt that the Dutch employer organizations would prove to be much more successful than their British counterparts in persuading union leaders to accept the principle of (partial) worker financing of social insurance expansion.

Business attitudes and the development of the social wage in the Netherlands

By the late 1940s, however, this was by no means already clear. In fact, when the CSWV made its statement on social costs being part of total wage costs, it did so in response to the unwillingness of labour unions to accept even partial worker financing of the public unemployment insurance programme that was about to be created. This unwillingness resulted from the conviction that responsibility for unemployment, and thus for financing provision against its financial consequences, lay solely with employers. The resulting stalemate was only resolved when employers succeeded in convincing parliament, including its labour delegates, to delay the programme’s introduction by three years out of concern over its impact on the competitive position of Dutch industry. This move eventually persuaded the unions to accept partial worker financing of the new scheme, after which they promptly demanded – and eventually received – full wage compensation (Ruijters, 1989).

By taking such an uncompromising stance on the question of how to finance the new unemployment insurance programme, the union federations had clearly undermined their aim of increasing the security of workers by as much as possible. And indeed, in order to limit the growth of labour costs, parliament seriously limited the scheme’s generosity (Oude Nijhuis, 2013). The unions’ stance was nevertheless quite understandable. In other European countries, including the United Kingdom, union leaders initially also took the principled stance that the costs of insurance against involuntary unemployment should be borne by those whom they deemed responsible for it (Alber, 1982). They likewise defended the view that employers should bear the costs of financing provision for workers who had incurred an injury or sickness while working under orders. It was for this reason that the British TUC had for instance strongly opposed even partial worker financing of the 1946 National Insurance (Industrial Injuries) Act (Lowe, 1998: 372). Such principled opposition to worker financing was mostly absent when it came to discussions over the introduction or expansion of public old age, and non-work-related disabilities and sickness insurance programmes. As these provided protection against risks that could not be attributed to employer actions, it was far easier for labour unions to accept that workers were in principle themselves responsible for financing them.

The Dutch union movement first came to acknowledge this principle during subsequent negotiations over the introduction of the 1956 General Old Age Act. Whereas the two confessional union federations had long ago come to support worker financing of old-age insurance because this fitted well with their emphasis on the need to preserve personal responsibility, the NVV, like the TUC in the United Kingdom, initially continued to insist on the introduction of a completely tax-financed pension because it feared that a contributory approach could not grant above-subsistence benefits to low-paid workers. The NVV eventually solved this dilemma by proposing to accept worker financing in exchange for a system that combined earnings-related contributions with flat-rate benefits. When the employer federations spoke out in favour of this proposal, mainly because they preferred it over the government’s mean-tested alternative which threatened the development of private pension schemes (IISG, NVV, HM4: Sociale Commissie, 30 August 1951), this not only paved the way for the introduction of the General Old Age Act but, as the unions agreed that the programme would be completely financed by worker contributions, it also formally introduced the principle of worker responsibility for the costs of welfare state expansion. The fact that union leaders once more demanded partial worker compensation did not undermine this principle, as they did so solely based on the argument that the statutory incomes policy had kept workers’ wages artificially down for many years (NA, CSWV, 2.19.103.06, 136: KvSO, 17 July 1956).

When the statutory incomes policy finally collapsed during the early 1960s, union leaders and employers continued with the practice of allocating which parts of negotiated wage increases were to be used to finance direct pay increases and improvements to public and private welfare programmes. As a result, the national union and employers’ federations continued to be heavily involved in wage bargaining, despite the transition towards industrial-level bargaining that took place in this period. By making sure that these improvements themselves did not add much to yearly increases in labour costs, this practice removed a major obstacle to obtaining broad employer support for them. While occasionally expressing concern over the fact that social welfare costs represented an ever-larger share of total labour costs, the employer federations seldom expressed worries over the consequences of social insurance expansion on total labour costs and national competitiveness (Oude Nijhuis, 2013). There is clear evidence of the effect this had on employer attitudes towards social insurance expansion. On various occasions, the employer federations, for instance, stated that increases in the generosity of public benefits, such as the expansion of a holiday pay to benefit recipients, were ‘acceptable … because it did not bring about additional costs’ (KDC, NKW, 641: CSV, 13 October 1965). On multiple occasions, they even professed their preference for improving private and public benefits over increasing net wages directly in order to ‘reduce inflationary pressures’ (NA, CSWV, 2.19.103.06, 134: KvSO, 7 December 1966; VU, FCWV, Memo verhoging AOW (FED 650-SV), 1966).

Under these circumstances, it may not be surprising that the employer federations were often quite willing to support increases in benefit generosity, especially when these did not have serious potential consequences for the supply of labour. They did often express reservations when confronted with union demands for major increases in the level of the old-age pension benefit (which, as they pointed out, left ‘less room’ for the development of occupational pensions (VNO, F119(4): Sociale Verzekering RCO, 29 October 1963)) and flat-out opposed most increases in the generosity of the unemployment insurance programme (the programme that was most likely to affect labour supply), although to little avail as we will see in the following section. Yet they, for instance, fully supported the coming about of exceptionally generous disability and sickness insurance schemes during the mid-1960s. In return for the unions’ willingness to accept that workers were to be responsible for the costs of improved provision against the risk of disability, the employer federations supported entitling all disabled and sick workers to a benefit that equalled 80 percent of their previous wages (or a percentage of that in the case of partial disability), regardless of the cause of injury, from the moment they started working and for as long as their disability or sickness lasted. During internal discussions on the matter in the CSWV social insurance committee, all members had agreed that a lower benefit level would be ‘insufficient’ for workers (NA, CSWV, 2.19.103.06, 132: KvSO, 30 January 1958). The only consistent source of disagreement with the unions was over the latter’s proposal to opt for contribution levels that were uniform among industries. The employer federations opposed this because they felt that ‘contribution levels need to reflect actual business risks’ (VNO, F5(6): Nota RvBA, 30 December 1963).

Business attitudes and the British rejection of the social wage

While not in any way limited to the Netherlands, there can be no doubt that the practice of viewing social insurance provision as part of the ‘social wage’ was particularly well developed there by the late 1950s. In the United Kingdom, on the contrary, it was ‘neither developed nor institutionalized’ in this period (Rhodes, 2000: 22). Both the extent and importance of this rejection have been far from fully recognized in the welfare state literature. Many studies for instance assume that while the tradition of voluntarism and the decentralized and fragmented nature of wage bargaining prevented the development of the notion of the social wage in the United Kingdom, the British labour union movement was generally quite willing to accept at least partial worker financing of new welfare initiatives (Hill, 1993: 50; Lowe, 1998; Mares, 2006; Whiteside, 2006). Such views neglect that both the voluntarist inclinations of this movement and the failure of postwar attempts to set up incomes policies in the United Kingdom can to a large extent be attributed to the presence of powerful occupationally organized labour unions that jealously guarded their autonomy in order to preserve the privileged status of their members (Pelling, 1971). In addition, they neglect the fact that these unions refused to bear even partial worker responsibility for the costs of major welfare initiatives when they felt that these contributed little to the security of their members or because they redistributed income and risk in favour of less fortunate groups of workers.

In other words, a substantial part of the British labour union movement rejected even partial worker financing of major welfare initiatives because they did not deem these initiatives to be in the interests of their members. To fully appreciate the importance of this rejection, it is first necessary to note that union leaders were generally quite willing to contribute to the costs of initiatives that were viewed as genuine improvements to the social wage by them. In the immediate postwar period, the Labour government had for instance found it quite easy to obtain TUC support for partial worker financing of the 1946 National Insurance Act, which granted a basic level of insurance against the inability to work as a result of old age, unemployment and non-work-related sicknesses and injuries. The reason for this support lay in the popularity of the Act, which did not affect the distribution of income and risk in a major way, among all grades of workers (Deacon, 1982). In addition, those unions that in subsequent years managed to complement the state benefit with ever more generous occupational entitlements generally did so through wage negotiations under which these benefits were viewed as part of the social wage. To be sure, these were generally unions that represented skilled, and thus higher paid and less risk-prone, workers. In subsequent years, it was these unions that refused to accept even partial worker financing of various major Labour initiatives aimed at improving the generosity of the national insurance benefit.

Their unwillingness to do so cannot be explained by arguing that as their members managed to obtain ever more generous private provision, Labour’s initiatives increasingly meant little to them. After all, the evenly rapid expansion of private pension provision in the Netherlands did not deter the labour unions there from pushing for more generous levels of public provision and, as we will see below, offer to either accept higher worker contributions or lower direct wage increases. Much more important was therefore that Labour’s welfare initiatives increasingly focused on improving matters for lower paid, more risk-prone workers who, incidentally, also tended to be the ones who were not able to complement the inadequate state benefit with private provision. The presence of a large occupationally organized labour union movement made sure that there was strong resistance to the redistributive consequences of this. Thus, whereas the Dutch union movement had, for instance, been united in its support for an old age pension insurance programme that benefited lower paid contributors, a large part of the British union movement opposed a similar proposal put forward by the Labour Party at roughly the same time (Baldwin, 1990: 233). In subsequent years, these unions continued to oppose Labour initiatives that ‘put too heavy a burden directly on those with the smallest risk’ and gave ‘advantage’ to lower paid workers (MRC.292B/116.2: TUC SCSUB, 14 February 1962; MRC.292B/116.1: TUC Minutes SIIWC, 11 January 1962). The most well known of these was national superannuation, which was condemned by unions representing skilled workers in a motion presented at the TUC Congress of 1969 for attempting to ‘improve the position of the worse-off section of the community at the expense of those whose earnings were in the middle salary range’ (MRC, MSS.292B/445.7/1: Occupational pensions, 9 March 1970).

The consistent resistance of powerful sections of the British labour union movement to redistributive welfare expansion not only made it impossible to view social insurance provision as part of the social wage but also threatened the principle of bipartite financing that had been established in the prewar period. After all, those unions that resisted these initiatives naturally also refused to accept even partial worker financing of their costs and frequently responded to them by threatening to increase their wage claims during subsequent wage negotiations in order to make up for any losses in income that were due to contributory increases. And this in turn had major consequences for business attitudes towards the postwar expansion of the British welfare state. It made this expansion more costly to them and meant that government welfare initiatives increased the potential for conflict with the union movement. The employer federations were quite aware of this and frequently complained that ‘increased payments to workers through social insurance schemes are inflationary’ as they would result in ‘additional wage claims based on the increased contribution liability of employees’ which meant that ‘in the long run [any] increases in employee’s contributions will be reflected in wage claims and so far as these are allowed, the whole increase will be met by employers’ (PRO, Lab 10/1824: BEC Bulletin 222, 22 January 1964; PRO, Lab 10/2024: Meeting with BEC, 3 February 1964).

As a result, British employers much more frequently expressed concerns over the costs of welfare state expansion than did their Dutch counterparts. After all, whereas both private and public provision were viewed as part of the social wage in the Netherlands, the BEC repeatedly pointed out that only private provision could be used as alternative ‘offers’ for direct wage increases (PRO, Lab 10/1932: BEC Bulletin, 222, 22 January 1964). In response to a Labour government initiative to improve the generosity of the state sickness and unemployment benefit by introducing earnings-related supplements during the mid-1960s, it for instance responded that ‘various proposals on unemployment and sickness benefits were items which the employers could use in bargaining with the unions about wages, but if the government imposed the schemes by legislation, employers would lose them as bargaining counters’ (PRO, Lab 10/2024: Meeting with BEC, 3 February 1964). As a result, the government’s initiative would not only ‘further push up wages’; it ‘might actually increase industrial conflict’ (MRC, MSS.22/C/4/61: British Industry, 164-1965; PRO, LAB 10/2024: BEC draft Bulletin, 19 September 1963). Its successor, the CBI, would later also frequently point out that public intervention ‘remove[d] from industry an important negotiating counter’ and would likely ‘result in strong [union] pressure to have the schemes financed by employers and the state only’ (PRO, LAB 130/602, Memorandum CPRS, 1972; MRC, MSS.200/C3/EMP/5/3: Minutes SSC, 3 March 1969).

Under these circumstances, it is not surprising that these business groups strongly resisted most of Labour’s postwar welfare proposals. They often invoked financial concerns to do so. During the late 1950s, the BEC and Association of British Chambers of Commerce (1959), for instance, condemned Labour’s first national superannuation proposal as a ‘costly’ initiative that would not only affect private pensions, but would also ‘jeopardize the value of sterling’ (MRC, MSS 200/B/3/2/C1221: A state-administered pension scheme, 4 December 1957). For the same reason, the CBI later asked the Labour government to ‘defer’ its plans to introduce earnings-related supplements to the unemployment and sickness benefits because of ‘repercussions on [industry’s] overseas position’. And when CBI representatives during the late 1960s rejected Labour’s initiative to present a ‘New Deal for the disabled’, they primarily did so because it ‘would raise costs’. In addition, they expressed their concern that the initiative might complicate subsequent wage negotiations as any increase in wage claims that followed as a result of union efforts to make up for increases in worker contributions ‘had to be checked’ (MRC, MSS.200/C/3/EMP/5/1: CBI SIC, 6 October 1965). Finally, there can be no doubt that cost considerations played an important role in the CBI’s rejection of Labour’s previously mentioned second national superannuation initiative. And if the scheme were to be enacted, the CBI argued that ‘the redistributive element of the scheme should be paid by the Exchequer’ in order to prevent it from having ‘repercussions’ in the area of collective bargaining (PRO, T 227/308, CBI memorandum 23 June 1969).

Cost considerations and business influence

Given the importance of cost considerations in shaping business attitudes towards the welfare state, it may not be surprising that the response of British and Dutch labour unions to demands for worker financing of social insurance expansion had strong consequences for business views towards this expansion in the two countries. Yet, and as noted earlier, cost considerations are not only of importance to businesses; because of their impact on firm profitability, they matter greatly to governments of all types of political persuasions as well. This meant that the response of British and Dutch labour unions to demands for worker financing also affected businesses’ ability to exert influence over policy choices in the two countries – at least in so far as this served to persuade policy-makers to abandon, postpone or reduce the generosity of public welfare initiatives. There are at least two reasons why this is worth exploring here. First, and as outlined in the second section of this article, doing so helps us to understand the limits of business power and the factors that affect its influence over the political process. Second, and by extension, it enables us to further explain instances of business support for welfare state expansion – at least in the Netherlands. After all, and as pointed out by a host of scholars over the years, business support for this expansion may also come about as a strategic response to the growing strength of pro-welfare demands (Hacker and Pierson, 2002; Paster, 2013).

As the salience of cost considerations is widely recognized in the welfare state literature, most scholars would probably regard it as self-evident that as a large section of the British labour union movement refused to accept even partial worker financing of major welfare initiatives, this not only provided an important reason for employers’ groups there to oppose these initiatives but also bolstered their opposition to them. In fact, despite its tendency to overestimate organized labour’s willingness to accept partial worker financing of social insurance initiatives in the United Kingdom, the existing literature on British welfare state expansion has always paid much attention to the importance of cost considerations (Baldwin, 1990; Bridgen, 2000; Lowe, 1998). That these considerations complicated Labour’s attempts to improve state provision in Britain during the 1950s and 1960s is therefore uncontested. Moreover, documentary evidence confirms that various Labour governments shared the concerns voiced by employers over the possible inflationary consequences of their proposals and the consequences they could have for wage bargaining, During meetings with the BEC over the possible introduction of a statutory redundancy payments scheme and earnings-related supplements to the sickness and unemployment benefits, government officials could, for instance, ‘only admit that they shared the BEC’s concerns on their costs, as many of them had already expressed doubts … for exactly this reason’ (PRO, Lab 10/1932: Meeting with BEC, 1963; PRO T311/106, Maude to Harding 19 March 1963). Later on, BEC officials gleefully noted that cost considerations had been an important reason for the government’s decision to keep these schemes ‘as modest and as simple as possible’ (MRC, MSS.200/C/3/EMP/5/1: CBI SIC, 6 October 1965). And in the period following on Labour’s announcement to introduce national superannuation during the late 1960s, there was so much discussion over its possible adverse repercussions for the area of collective bargaining that the government was forced to discuss this possibility in its own white paper on this (HMSO, 1969: 179).

There has, by contrast, been much less attention for the possibility that business groups were not only more likely to consent to, but also found it more difficult to resist, demands for increases in welfare generosity when labour unions were willing to accept that these increases were to be financed by workers themselves. An analysis of events in the Netherlands is ideally suited to illustrate these consequences. The penultimate section already explained that as the practice of financing public provision out of the margin for pay increases became more established in the Netherlands, the main employer federations continued to express strong reservations towards welfare initiatives that were expected to affect the supply of labour or threatened to thwart the development of private pension programmes in a major way. Yet, contrary to their counterparts in the United Kingdom, and as we have seen, they seldom invoked cost considerations to motivate their opposition to these initiatives. Instead, they generally espoused concerns over the consequences of increased benefit generosity for work incentives. And when confronted with union demands for increases in the level of the state old-age pension benefit, they consistently emphasized the importance of private pensions in generating investment (VNO, F119(4): RCO, 29 October 1963; VU, FCW, 17: Nota pensioenen, 6 June 1968; VU, FCW, 18: Pensioenvraagstuk, 17 October 1969).

The very fact that the employer federations were quite unsuccessful in their attempts to prevent the Dutch social insurance system from acquiring ever more generous features illustrates the lack of salience of these concerns among policy-makers at the time – including those of a liberal and confessional signature. Despite strong warnings from employers over its consequences for labour supply, a confessional–liberal government in the mid-1960s, for instance, introduced minimum benefits into the state sickness and unemployment insurance programmes, and greatly improved the latter’s overall generosity by increasing the duration and level of the benefit. The fact that the same government, despite strong employer protests, also introduced a generous unemployment provision that granted a benefit equal to 75 percent of the previous wage to all workers with a working history of at least 6 weeks, including the voluntary unemployed, further shows that policy-makers were not much concerned with the consequences of benefit generosity for worker behaviour in this period (Cox, 1993). As noted earlier, the employer federations had also been unsuccessful in persuading parliament that premium differentiation among industries was necessary to prevent the new state disability insurance programme from being used for redundancy purposes. Finally, and despite employer warnings that this would ‘undermine the function of [occupational] pensions for the savings rate’, governments of different political signatures raised the level of the state old-age benefit in two major steps during the 1960s, until it reached the level of the minimum wage by the end of the decade (VU, FCW, 18: Pensioenvraagstuk, 17 October 1969).

In other words, as concerns over the financial consequences of welfare state expansion were increasingly assuaged as a result of the emerging practice of financing this expansion out of the margin for pay increases, and concerns that were clearly less salient to policy-makers remained, it became much more difficult for Dutch business groups to resist demands for increases in benefit generosity. They seem to have realized this quite well. In the aftermath of the first step towards raising the level of the public pension benefit to that of the minimum wage, the representatives of one of the confessional employer federations, for instance, complained that ‘in matters concerning the public pension benefit … the government always sided with the unions’ (VU, VPCW, 172: Notities SER-stuk 640, 1964). Some years earlier, the largest employer federation at the time, the CSWV, had already come to the conclusion that increases in benefit generosity ‘could not be prevented. At the most, they could be delayed’ (NA, CSWV, 2.19.103.06, 139: CSV, 25 October 1963). This realization affected their behaviour in at least two ways. First, it meant that rather than responding with outright hostility to demands for increases in benefit generosity, it was often more fruitful to negotiate with union and government representatives in order to make these demands more palatable to businesses. Second, it occasionally motivated employers to come forward with welfare initiatives of their own in order to prevent the introduction of more problematic alternatives.

The employer federations, for instance, followed the latter strategy when they were confronted with a union proposal to introduce a secondary public pension layer on top of the existing state benefit during in the late 1960s. This proposal greatly disturbed employers, as it would effectively bring about an end to the existence of occupational pension funds. By offering to finance the proposal out of the margin for pay increases, the unions also made sure that the proposal was favourably received by the government (VNO, F2(2) DB, 9 September 1969 and 14 October 1969). To prevent it from being implemented, the employer federations responded by offering to increase the level of the existing state benefit to that of the minimum wage, introduce a generous holiday surcharge to the benefit, and oblige all workers to join occupational pension schemes. Together, the public and occupational pension benefit would come to provide at least 70 percent of the previous wage. Internal employer resistance to the scheme, which was partly based on principled opposition to compulsory membership of occupational pension schemes, was brushed aside by the leaders of the employer federations by arguing that they needed to find a position ‘that leads to a solution in the short term that is satisfactory to the unions’ (VU, FCWV, 18: Pensioenvraagstuk, 17 October 1969). When the union federations were informed of this counterproposal, they immediately endorsed it (VNO, F118(21): CSV, 6 November 1969). Within months, the government consequently adopted the proposal as well.

Conclusions

This article sought to show that the extent to which the postwar expansion of the welfare state imposed a financial burden on businesses varied considerably depending on organized labour’s willingness to accept that this expansion was largely to be financed by workers themselves. Based on an analysis of the major social welfare initiatives put forward in the Netherlands and the United Kingdom during the first decades of the postwar period, the analysis showed that there was considerable variation in organized labour’s willingness to do so. It emphasized that this willingness among others depended on the type of labour market risk against which workers were to be insured, the margin or scope for pay increases, and the extent to which social welfare initiatives were perceived as actual improvements to the social wage. Finally, it showed that this willingness had major consequences not only for business attitudes towards social welfare initiatives but also for their ability to influence these initiatives. In other words, when labour unions were willing to accept that improved provision against labour market risks was to be financed out of the margin for pay increases, business groups were not only more likely to support social welfare initiatives but also found it much more difficult to persuade policy-makers to abandon, postpone or reduce the generosity of these initiatives.

While this analysis focused on the Netherlands and the United Kingdom, its findings can be extended to other countries as well. In all advanced industrial societies, including those that initially leaned heavily towards tax financing, a large part of the welfare state is funded through contributory revenue. As a result, it is of crucial importance to investigate discussions over who was to pay for the costs of contributory social programmes. It has often been argued that social programmes based on contributory financing, because of their lower level of visibility to voters, were less contested and therefore easier to introduce and expand than programmes that are financed out of general revenue (Campbell and Morgan, 2005). This article shows that this is by no means always the case. Indeed, and as we will see, union opposition to intra-class redistribution can make it very difficult to improve the generosity of social insurance programmes in a contributory manner.

Finally, recent examinations of business involvement in welfare state expansion have almost exclusively focused on the claim that businesses may have supported the introduction and expansion of social insurance programmes because they expected to derive economic benefits from doing so. This claim remains contested, however. One major problem with this claim is that scholars have found it difficult to relate it to variation in business attitudes towards welfare state expansion between countries and over time. An approach that acknowledges that there has been strong variation in the extent to which welfare state expansion-imposed costs on businesses may be more successful in explaining this variation. By outlining various factors that explain this variation, this article thus provides several avenues for future research.