Conservative parties opted for (predominantly male) foreign labour to expand the workforce, ... Recently, Devitt (2015) has looked at whether Irish migration ..... perspective, exemplified by the comparatively large share of women in Parliament.
Foreign Workers or Women? Immigration and Welfare Expansion as Alternative Labour Supply Strategies in Post-War Europe Alexandre Afonso, Leiden University Abstract This article explains why some European countries opted for migrant workers to increase the labour supply in the post-war period, while others opted for increasing female employment via the expansion of welfare programs such as childcare and paid maternity leave. In doing so, it looks at how immigration can act as a form of social policy by other means when it comes to labour supply. The paper argues that the existing setup of welfare regimes and power relationships was an important structuring factor in these choices: Bismarckian countries characterized by a clear male breadwinner model and a strong adherence to the principle of subsidiarity opted for foreign labour, while SocialDemocratic countries with a more egalitarian division of labour and a more favourable political context for public sector expansion opted for female employment. The paper uses a paired comparison of Sweden (Social-Democratic) and Switzerland (Bismarckian) as two contrasting examples of each strategy. While Switzerland based its labour force expansion strategy almost exclusively on foreign labour, in Sweden, guest-worker programs were quickly abandoned and substituted with a clear strategy to promote the labour force participation of women via the expansion of welfare services, public sector employment, paid parental leave and free childcare.
Word count: 8412
1. Introduction The increase in female and migrant employment have certainly been the two most important transformations in labour markets in advanced industrial countries in the last decades (Oesch, 2013). While the labour force in the immediate post-war period was mostly male, native, and in steady employment, women and migrants have come to constitute an increasing share of employment in rich countries. In the United Kingdom, the share of foreign-born people in total employment increased from 7.2% in 1993 to 16.7% in 2014. In the United States, this share increased from 10.8% in 1996 to 16.1% in 2012. Developments for female employment are even more far-reaching. In 1971, only 22.6% of Dutch women between 25 and 54 worked. In 2015, this percentage had reached 77%. Corresponding figures evolved from 47% to 70% in the United States for the same period, and increased to varying extents in all European countries, almost coming on par with male employment rates. However, the timing and extent of these developments has varied quite significantly across countries. While Scandinavian countries raced ahead in female employment in the 1970s and 1980s, coupled with the massive expansion of social services to enable it, many Continental
1
countries (e.g Germany, France, the Netherlands, Switzerland, Austria) relied extensively in the 1950s and 1960s on so-called “guest worker programmes” sourcing labour from poorer regions of the Mediterranean rim (Italy, Spain, Portugal, Morocco, Greece, Turkey), and female employment progressed much more slowly there, before catching up in the 1990s and 2000s. In this article, I seek to explain the mechanisms that underpin these different strategies, considering immigration and welfare state expansion as two alternative strategies to expand the labour supply. In line with the theme of this special issue, I look at “social policy by other means”, but in contrast to other contributions focusing on how policies or actors other than conventional welfare programmes may provide social protection, I look at how immigration may act as a substitute to welfare programmes as a labour supply policy . I argue that countries seeking to expand their labour supply in times of economic expansion could do so either by expanding the labour market participation of women, which entailed the expansion of social services (childcare) and tax revenues, or by bringing in migrant workers, which required minimal public investment and preserved traditional gender roles. The choice for one or the other strategy was shaped by underlying political coalitions, as well as institutional factors. While countries with a Bismarckian system dominated by ChristianConservative parties opted for (predominantly male) foreign labour to expand the workforce, Scandinavian countries opted for increasing female labour force participation via the expansion of childcare and family policies. Using a most-similar system comparison of Sweden and Switzerland, I show how institutions and power relationships structured policy choices for migrant labour or female employment. Conservative Switzerland relied almost exclusively on foreign labour to expand its workforce, while Social-Democratic Sweden drew on female employment. The main idea of this article is to present immigration policy as a form of “social policy by other means” (SPOM). It contributes to the growing literature connecting immigration, employment and welfare (Afonso & Devitt 2016), and links the issue of immigration with the gendered dimension of social policy (Sainsbury 1999, Lewis 1992). Esping-Andersen (1990, 144-161) already emphasised how ensuring labour market supply is an important goal of social policy, and welfare programs can have an important effect on the employment of specific groups (Iversen and Wren, 1998; Sapir, 2006; Häusermann and Palier, 2008). For instance, generous early retirement schemes likely lower participation among older workers, while childcare subsidies and care leave facilitate the labour market participation of mothers (EspingAndersen, 1990, 155.). While these “conventional” welfare programs can be used to activate specific domestic groups, immigration policy can be considered as an alternative strategy to activate workers brought in from outside. Hence, I look at one specific social policy goal
2
(ensuring the labour supply) which can be pursued either by conventional social policy (childcare) or SPOM (migration). Recently, Devitt (2015) has looked at whether Irish migration policies in the 2000s hindered the development of childcare services to facilitate female employment. Migration and childcare policies are considered as alternatives to increase the size of the labour force in a period of high demand. This paper draws on this idea but puts it in a comparative-historical context.
2. Female Employment or Guest Worker Programmes: Choices in Labour Supply Strategies The thirty years that followed the Second World War were arguably the longest period of growth in Western Europe since the Industrial Revolution (Eichengreen, 2008; Kindleberger, 1967). West European economies grew at an unprecedented pace, propelled by an increasing demand for consumer goods, infrastructure and services, and technological progress allowing for remarkable jumps in productivity. Accordingly, labour markets were close to full employment and many countries soon faced labour shortages which could no longer be met by the local (male) workforce. In Germany, for instance, official unemployment amongst men in September 1955 was at 1,8%, meaning that nearly all of the male workforce was de facto employed (Herbert 2001: 202). This tight labour market resulted in wage increases which could cause inflationary tendencies, causing growing concern for employers and public authorities. In this context, they were faced with different options: technological change; relying on the domestic labour market (women), or foreign labour (Kindleberger 1967). While the rationalization of production was a means to reduce labour demand through capital investments, and produce more goods with fewer workers (Eichengreen, 2008, 6.; Kindleberger, 1967, 132.), it faced a number of limits. Replacing people with machines was not possibly in every sector, especially in services and agriculture. It was also limited by the availability of technology and capital. Moreover, the durability of economic growth in the aftermath of World War II was unclear, and massive capital investments could not necessarily be scaled back easily should a new recession arise. While mechanisation was an important factor, I focus on choices determining which type of labour supply should be mobilised.
Expanding the Domestic Labour Supply: Female Employment With Welfare Expansion If an increase in demand cannot be met with technological change, one has to increase the labour supply (extensive growth strategy). Options range from increasing the working hours of
3
the core workforce to mobilising under-utilised sources of domestic labour, such as youngsters, older workers, and – most importantly - women. In 1950, the percentage of women participating in the labour force was 48.5% in Germany, 39.1% in Switzerland, 28.8% in the Netherlands, 42.9% in the United Kingdom, and 50% in Denmark (Kindleberger, 1967, 153.). An average of 14 OECD countries indicates a 45% female participation rate in 1965 (Brewster and Rindfuss, 2000). A number of social dynamics, have played out against the three first options (increasing working time, bringing in more young people, or more elderly people). First, given the tendency towards the reduction of working time and the increasing demand for leisure in the post-war period, increasing working time for the core workforce has been a difficult option for employers, and working hours have declined steadily across almost all advanced economies. Similarly, the mobilisation of older and young (male) workers has tended to clash with the development and expansion of (early) retirement schemes after the second world war, and of (higher) education, delaying entry into the labour market. As for female employment, “work for older women [above 29] in much of the Continent, especially the Netherlands and the Latin countries, is subject to social disapproval” (Kindleberger 1967, 154.). On much of the European Continent, female employment challenged the male breadwinner model championed by powerful Christian-Democratic parties, who were important forces in the building of social insurance programmes (Kersbergen 1995). An obvious problem of this strategy for women of birthing age was how to reconcile paid work with childcare. Enabling the labour market participation of mothers required the expansion of public facilities to take care of young children, and/or flexible arrangements allowing young female workers with children to take up employment, such as paid maternity and parental leave. If this was to be taken over by the state, it also required a considerable expansion of the welfare state, and an increase in tax revenues. In this context, the expansion of female employment could face potential opposition from Liberals because of its fiscal implications and the expansion of state intervention, and from Christian Conservatives because of its challenge to the traditional (male-centric) family model. While the promotion of fertility could also be an option to expand the labour supply in a way that would not necessarily challenge traditional family values – on the contrary – it could only be an option that could generate effects in the long term, and did not solve the issue of low participation rates. As I will show, Scandinavian countries were the only ones where political power relationships would prove favourable to the expansion of female employment rather than immigrants.
4
Expanding the External Labour Supply: Foreign Workers The second strategy is the import of foreign labour through temporary guest worker programs (Kindleberger, 1967, 196.; Martin and Miller, 1980). This strategy displayed a number of advantages. First, it seemed to be a less risky strategy to increase the labour supply in the context of uncertainty that characterized the post-war period because foreign labour possessed a higher degree of elasticity. Immigrants from poorer countries could be brought in to fill in shortterm labour needs, but could also – theoretically – be easily sent back to their country of origin should a new downturn arise. In this respect, it was a more secure strategy than capital-heavy technological change in the context of uncertainty of post-war economic growth. Implicitly, the use of foreign labour could also help preserve existing gender roles by making it possible to “keep mothers at home”, a view that that was seldom articulated explicitly (see, however Herbert, 2001, 204.). In fact, expanding female employment was often not even mentioned as an option. Arguably, importing foreign labour could generate tensions within the native workforce, especially if labour migration took place concomitantly with the persistence of residual unemployment. From the point of view of trade unions, it could also risk creating a secondary labour market with worse wage and working conditions because of the presumably lower bargaining power of migrant workers, and the difficulty for unions to organize transient workers (Piore, 1979). Hence, strong unions could be expected to oppose possibilities to differentiate employment terms, if not to block immigration altogether.
1. Explaining the Choice of Labour Supply Strategies How can we explain why certain countries opted for female employment, while others opted for the import of foreign workers? Each labour supply strategy entails a number of trade-offs in terms of state intervention and taxation, the preservation of traditional structures, and segmentation of the labour market. It is striking to note that the countries where guest worker programmes were the most widely used were the German-speaking countries characterised by a Conservative/Bismarckian welfare state (Germany, Austria, Switzerland) (von Rhein-Kress, 1993), while female employment increased much earlier and more quickly in Scandinavian countries characterised by Social-Democratic dominance. Guest-worker programmes were also established there, but they were progressively scaled down (Knocke, 2000). Drawing on this, it is possible to establish a number of causal links between welfare regimes and the choice of strategy.
5
I have argued that policies to enhance the labour market participation of women entail a challenge on traditional gender norms, and imply a considerable expansion of the public sector. These requirements tend to clash with the values of the conservative parties that dominated Continental welfare states, namely the male breadwinner model and the principle of subsidiarity. In these countries, the use of foreign labour appeared as a much more expedient solution given the values and constraints embodied in the Bismarckian model. First, the traditional Bismarckian model and the Christian Democratic parties that shaped it were strongly anchored in the idea that women should stay at home to take care of children. Various policies and tax incentives in these countries typically discouraged female employment (Morel, 2007, 620.; Esping-Andersen 1990). Second, a central organising logic of the Bismarckian welfare state was the principle of subsidiarity, that is, that the state should not take up tasks that can best be assumed by lower units, such as the family, social partners or subnational units. In Bismarckian systems, family policy would typically consist of cash transfers to families – essentially subsidising families to take care of children - and little in the way of public social services (Morel, 2007, 621.). This hindered the expansion of state intervention that would be required to free up the female workforce for paid employment via public childcare services. This principle of subsidiarity was also embodied in the central governance role of trade unions, whose anchorage in the blue-collar industrial sector made them weakly receptive to values of gender equality (Häusermann, 2009). In these contexts, foreign workers were a more attractive option in Bismarckian models. By using foreign labour, gender norms would be preserved, and the public investments required to welcome mostly young, able-bodied men were believed to be minimal, satisfying fiscal conservatives and the principle of subsidiarity. Besides, the segmented nature of social rights in Bismarckian welfare systems was fairly expedient to exclude them from certain benefits. Because social rights were tied either to occupation or sector, it was easier to exclude foreign workers from certain welfare schemes, such as occupational pension plans, making their employment less costly (Herbert, 2001, 209.; Sainsbury, 2012). The more limited coverage of collective bargaining in Conservative models had the same effect, allowing for a greater level of segmentation of the workforce. As a whole the segmented nature of social insurance schemes and labour markets in Bismarckian countries made foreign labour a relatively attractive, cheap and flexible option since migrants could be partly excluded from major social benefits, and sent back in case of recession. While guarantees had to be given to trade unions to prevent a race to the bottom in wage standards, their opposition could be overcome. In Germany, for instance, trade unions consented to guest worker programmes as they saw this was one of the only alternatives to allow for a reduction in working time (Herbert, 2001).
6
In contrast, employers and policymakers in Scandinavian welfare states dominated by Social Democratic parties faced a different set of incentives in the choice between foreign workers and women. Gender roles had been less rigid early on. The principle of subsidiarity was not such a hard constraint on the expansion of state intervention to provide public support for working women. While conservative and liberal parties in power elsewhere were reluctant to increase the size of the public sector, Social Democratic hegemony in Scandinavian countries could enable the expansion of service-heavy welfare states and large public sectors (Iversen and Wren, 1998, 538.). On the other hand, the incentives found in Bismarckian systems to use an “outsider” workforce at low cost were less present in Scandinavia because of the institutional setup of the labour market and the strength of unions. Because of the universality of eligibility in welfare programmes and encompassing collective bargaining coverage, it was difficult for employers to substantially differentiate employment terms for migrants, and exclude them from welfare benefits. The fairly heavy tax load necessary to fund a large welfare state also depressed demand for low-skill employment in which migrants tend to be concentrated. Since migrants needed to be entitled to the whole range of services offered by Nordic welfare states, they were also less attractive as an option for both government and employers.
Figure 1: Female employment rates and share of foreign nationals in European countries, 1980s Figure 1 about here Source: OECD international migration database, Brewster and Rindfuss, 2000, 271.
A cursory look at country groupings along these two variables (female labour force participation and immigration) seems to provide support for the argument. In Figure 1, I have plotted countries by labour force participation of women in 1980, and the share of foreign nationals in the 1980s (an average was calculated to overcome the unavailability of data for the same time points).1 Sweden, Denmark, Finland and Norway appear at the lower right corner of the plot, with both the highest levels of female labour force participation and lower shares of
1
I use data from the 1980s because one can assume that the choices made in the 1960s would have deployed their effects, the asylum migration waves of the 1990s which could conflate the impact of labour migration had not yet happened. It is also the decade with the largest divergence in female employment rates, Bismarckian countries catching up to Scandinavia in the 1990s and 2000s. In subsequent period, when mostly asylum-based immigration increased to Scandinavia and employment rates for women started to converge throughout Europe, my argument is loses traction.
7
immigrants in the population. France, Germany, Switzerland and Belgium have both lower levels of female participation, and higher share of immigrants. In the following sections, I assess this argument on two countries displaying a clear variation in their welfare state setup and underlying political coalitions, Switzerland and Sweden. These countries also clearly appear as contrasting cases in the plot.
3. Cases and Methods To assess my argument, I use a comparative-historical analysis of welfare and immigration policies in Sweden and Switzerland between 1945 and ca. 1990. Broadly speaking, the research design is a most-similar systems design where the main independent variable is the welfare regime, or rather the different components of the welfare regime outlined above (institutions and party power relationships). Sweden and Switzerland are appropriate cases to assess the argument insofar as they share a number of scope characteristics. They are wealthy West European countries where wage differentials were large enough to be attractive for Southern European immigrants. Following Katzenstein (1985), both belong to the group of small European countries that developed corporatist structures. Both countries had remained neutral during World War II, and were therefore in a similar position, with a relatively intact production apparatus faced with a large demand for exports. Finally, none of them had colonial ties which could – as in the case of France, Great Britain or the Netherlands – foster immigration without deliberate choices guided by labour market concerns. Yet, Sweden and Switzerland diverged with respect to central variables of interest for our purpose. Sweden is often presented as the most emblematic example of the socialdemocratic welfare state, with high levels of social spending, the prevalence of universal programs, and high unionization rates. In contrast, while Switzerland has been the subject of some disagreement in the welfare state literature (Obinger, 1998), it can generally be considered Bismarckian, with a more fragmented structure of occupational schemes and management, the primacy of the principle of subsidiarity in welfare arrangements, and entrenched conservative gender norms (Häusermann, 2010, 214.). This latter feature was particularly strong considering that women were only granted the right to vote at the national level in 1971, following a referendum. Regarding power relationships, these two countries have also varied starkly regarding the strength of social democratic parties. While the average share of parliamentary seats and cabinet posts held by Social-Democratic parties between 1960 and 1980 was 47.3% and 82% in Sweden respectively, it was 25.6% and 28% in Switzerland.
8
Comparative indicators on immigration and female employment provide support for the relationship expected. In 1973, the share of migrant workers in the labour force was 4.9 per cent in Sweden, while it was 27.9 per cent in Switzerland (von Rhein-Kress, 1993, 161.). Meanwhile, from an already higher threshold, the female participation rate increased more quickly in Sweden than in Switzerland. In 1973, it was 62.6% in Sweden and 54.1% in Switzerland. Ten years later, it had increased to 76.6% in Sweden, while it slightly declined to 53.3% in Switzerland (von Rhein-Kress, 1993, 161.). By 1983, women constituted nearly half (46%) of the Swedish labour force, while they only constituted a third (36%) of its Swiss equivalent. In order to obtain some insight into the causal mechanisms at work, for each country, I carry out a comparative case study analysis of policy developments in the areas of immigration policy and female-employment-friendly social policies. The analysis is essentially based on secondary literature, with a subset of primary sources and comparative statistical data.
4. Sweden: From Small Guest Worker Programs to the Promotion of Female Employment Sweden opted early on for the expansion of female employment, while immigration played a much smaller role than in Switzerland. This strategy cannot be considered exclusive, however, insofar as Sweden also experimented with guest worker programs in the immediate post-war period. Its immigration policy should also be considered in the context of free movement within the Nordic labour market. Figure 2: Share of immigrants in the population in Sweden and Switzerland, 1960-2000. Figure 2 about here Source: Swedish Statistical Office, Swiss Statistical Office (Switzerland: foreign nationals; Sweden: foreign-born).
Figure 3 Immigration and emigration flows in Sweden and Switzerland, 1950-1971 Figure 3 about here Source: SCB, Swiss Statistics Office. Swiss emigration data for the period not available.
In 1947, Sweden signed its first of a series of bilateral recruitment agreements with Italy and Hungary. The recruitment and organization of labour migration was carried out by the newly created National Labour Market Board (AMS), a tripartite body where employers and
9
trade unions where represented (Knocke, 2000, 158.). Local unions needed to be consulted for every work permit delivered (Knocke, 2000, 162.). The features of the universal Swedish welfare state regime and the strength of trade unions would have a significant impact on the way guest worker programs were administered, and by way of consequence, also made foreign labour a less attractive option for employers than in Switzerland. As shown in Figure 2, if the share of foreign-born people in Sweden increased since the 1960s, this happened at a much slower pace than in Switzerland. Data on immigration flows (Figure 3) also points to much higher levels of immigration in Switzerland in the post-war period. In many ways, incentives for employers to resort to foreign workers were weakened by the power of unions and the institutional setting. First, the influential position of trade unions in the policymaking process made sure that the opportunities for employers to differentiate employment terms and access to social benefits were limited, and rights were on par with those of native workers. Foreign workers were allowed to change jobs within the occupational sector for which they were recruited, which, as we will see, contrasted with Switzerland. Second, foreign workers were granted access to the whole range of welfare benefits available to Swedish workers. The guarantee of an equal footing between nationals and non-nationals was laid out in the 1954 Aliens act (Knocke, 2000, 159.). Migrants were also allowed to bring in their families. Third, the recruitment agreements signed between Sweden and partner countries contained an organization clause stipulating that recruited workers committed themselves to join the appropriate union for their profession, and to remain unionized during their entire stay in Sweden (Knocke, 2000, 159.). This clause was abolished in 1965, but replaced by an agreement between SAF (the main employer body) and LO (the main trade union) providing for employers to recommend joining the union. Union membership was informally understood as a precondition for obtaining a job, a method that remained effective throughout the (Knocke, 2000, 166.). In Sweden, immigration through guest-worker programs remained fairly limited, and other channels (Nordic migration in virtue of free movement) that were not subjected to all these conditions remained more important in terms of numbers. Progressively, however, immigration channels were liberalized, and it became easier for nationals of other countries to gain access to the Swedish labour market. Swedish unions were deeply concerned about the creation of an “industrial reserve army” of workers which could undermine the working conditions of the native workforce, and made sure that differentiation channels were contained. As a consequence, it was difficult for Swedish employers to substantially differentiate employment conditions between migrants and Swedes (in a context of wage compression through encompassing collective bargaining). Because of the high levels of taxation and the
10
inability to bypass regular employment terms, foreign labour remained fairly expensive, lowering its attractiveness for employers. Throughout this period, labour unions remained critical of labour migration, calling for a better use of untapped domestic labour reserves, such as married women, the elderly, and partially disabled persons (Johansson, 2008, 170-171.; Knocke, 2000, 160.; Borevi, 2014, 38). However, they had accepted it – providing it did not challenge wage and employment levels of Swedish workers – when faced with the fact that labour demand could only met with a quick expansion of the workforce enabled by immigration. This tolerant stance, however, changed in the end of the 1960s. Foreign labour had increased to higher levels, especially from Yugoslavia, Greece and Turkey (Johansson, 2014, 103.; Knocke, 2000, 163.). This was also a time when LO, the main trade union, started to envisage the existing liberal migration policy as a threat to an orderly functioning of the labour market (Knocke, 2000, 163.). First, the entry of low-skilled foreign workers could delay structural transformations that were otherwise promoted by other policies, and delay the labour market entry of other social groups such as married women (Johansson, 2014, 104.). The labour market participation of married women, or the wives of union members themselves, was also a way to increase household income for union members as well. At this time, the Social Democrats in power implemented a series of measures to limit immigration. In 1966, the labour market board AMS was given a monopoly on recruitment, and migrants coming to Sweden needed to have a work permit before entering the country. In 1967, immigrants coming on their own initiative – outside bilateral recruitment agreements - could no longer be given work permits (Johansson, 2014, 106.). In 1972, eventually, LO decided to put a stop to labour immigration. This did not happen through direct legislation but through an internal decision with the trade unions, which decided to systematically reject applications for work permits (Hammar, 1985; Knocke, 2000, 165.; Borevi, 2014, 39). As will be explained in the next section, these changes happened concomitantly with a number of policy reforms designed to expand female labour market participation, such as separate taxation and the massive expansion of public childcare in the 1960s. In Sweden, the labour market participation of women was already above the European average in the aftermath of the Second World War, and the conservative family model that prevailed elsewhere in Europe was also less dominant (Nyberg 2000, 7.) A network of foster homes was created for children of poor backgrounds at the turn of the 20th century, which can be understood as forerunners of childcare (Nyberg, 2000, 9.). State support for these structures started in 1944 (Gunnarsson et al., 1999, 21.).
11
While foreign labour recruitment was being pursued, women were already more involved in paid labour than elsewhere in Europe, which also made this path a more obvious policy option for unions and the government. In the 1960s, a debate started about long-term strategies to solve labour shortages given the limits of foreign labour recruitment. Propelled by the women’s wing of LO, who had started lobbying for public childcare in the 1950s already, policy proposals in the debate included the introduction of a care allowance for working mothers and the expansion of public childcare (Naumann, 2005, 54.). Towards the end of the decade, the balance would tip in favour of the expansion of childcare to promote female employment as a “lesser evil” than the continuation of guest-worker policies which could erode solidaristic wage bargaining (Naumann, 2005, 54.). From the 1960s onwards, the number of children in childcare centres increased dramatically (Nyberg, 2000, 9.), and the labour force participation rate of women as well, propping Sweden away from Continental countries. In 1965, the Swedish female participation rate was 5% higher than the European average. By 1985, it had become 18% higher (Brewster and Rindfuss, 2000). Figure 4: Share of men and women in the workforce in Sweden and Switzerland, 1960-2015 Figure 4 about here Source: OECD LFS, Swiss Statistical Office
The background strategy behind these developments was a report from the National Commission on Child Care commissioned by the government to devise a strategy that could reconcile education imperative with increased female employment (Gunnarsson et al., 1999, 22.). During the 1970s, the Swedish welfare state underwent a period of clear extension. For instance, the 1975 National Pre-School required all municipalities to offer all 6 year olds with at least 525 hours of free pre-schooling (Gunnarsson et al., 1999, 23.). In 1974, paid maternal leave was substituted with paid parental leave, allowing fathers to take subsidized leave to care for their children. Nowadays, Sweden is one of the top spenders in public childcare and early education, with 1.6% of GDP (OECD, 2014) (Figure 2) The connection between foreign labour and women is made explicitly in a number of analyses (Kyle, 1979, 214.; Naumann, 2005, 54.). It is for instance striking to note that in the same year, in 1972, Sweden simultaneously put an end to Non-Nordic labour migration and introduced separate taxation for spouses, making it more profitable for married women to work (Knocke, 2000, 164.). Accordingly, the share of labour migration in total migration flows in Sweden would become marginal, and the employment ratio of women would be boosted by the
12
expansion of welfare programs in health and social services. In many respects, this expansion can also be connected to the greater influence of women’s movements in comparative perspective, exemplified by the comparatively large share of women in Parliament. In that respect, the difference with Switzerland, female suffrage was only passed in 1971, is striking.
5. Switzerland: Importing Guest Workers, Keeping Women at Home While Sweden experimented with foreign labour on a modest scale to expand the labour supply, shifting afterwards to a clear strategy to promote the labour force participation of women, Switzerland provides a contrasting example of labour force expansion based almost exclusively on foreign labour at least until the 1980s (Kindleberger, 1967). As shown in Figure 3, the size of foreign migration flows in Switzerland in the post-war period, brought in proportion to size, totally outpaced what happened in Sweden and, for that matter, almost all other West European countries. The context and power relationships explain a great deal of the variation in policy choices. While Sweden exhibits the typical characteristics of Social Democracy, Switzerland can best be characterized as a Liberal-Conservative model whose structural characteristics would foster the mass expansion of immigration rather than female employment and welfare: weak and fragmented unions, the prevalence of conservative gender norms, the dominance of right-wing parties ultimately hostile to welfare expansion, and a set of political institutions (federalism, direct democracy) which prevented the emergence of a large public sector as in Sweden (Immergut, 1992; Obinger, 1998; Trampusch and Mach, 2011). Female labour force expansion clashed with core beliefs of both dominant political parties: the challenge to the male breadwinner model was opposed by Christian Democrats, and the expansion of taxation to fund the welfare state by Liberals and powerful employer organisations. In Switzerland, incentives for foreign labour recruitment were also stronger because of larger differentiation opportunities. First, in contrast to Sweden, the guest-worker programs put in place by Swiss authorities entailed a high degree of differentiation and restrictions in employment and welfare rights. This made foreign labour an attractive option for employers because foreign labour came to be an easy, expendable labour supply for which social risks did not need to be paid for. Because “guest” workers were not destined to stay, Swiss authorities encouraged a high turnover rate to discourage settlement, restricting rights in social and legal terms. For instance, a particularly popular type of work permit consisted in “seasonal” permits delivered for a period of 9 months for sectors such as construction, agriculture and tourism which did not entitle to family reunification, and did not grant access to unemployment insurance (the latter was not compulsory). Because there was no tied social entitlement, there
13
were also limited payroll contributions. The number of permits of this type delivered each year reached about 250’000 in the mid-1960s, or more than half of total yearly entries (Piguet, 2005, 41.). Second, unlike Sweden, work permits were tied to a particular employer, meaning that migrants could not change jobs without losing their right to stay in Switzerland. Their bargaining power on the labour market was also therefore constrained. Third, the Swiss welfare state at the time was much less developed than neighbouring countries. Social programmes had a lower degree of coverage, allowing for differentiation (Cattacin, 2006, 55.). The coverage of collective bargaining was similarly low (about 50%). As shown in Figure 5, public expenditure for parental leave and childcare was almost non-existent, which contrasts sharply with the levels observed in Sweden for the same period. In the direct aftermath of the war, Swiss authorities feared similar economic developments those that followed the First World War, with a massive economic crisis (Cerutti, 2000, 89.). In light of labour shortages in a number of economic sectors, Swiss authorities opted for seeking guest workers – first from Germany, and in light of the refusal by allied forces to consent to German emigration, to Italy - providing they would not settle. Trade unions were initially opposed to this measure, but the Government pursued this strategy nonetheless in light of strong employer demand (Cerutti, 2000, 89.). In 1948, the Swiss government signed a bilateral agreement with Italy, whose features were fairly different from those signed between Sweden and its own sending countries. There was no unionization clause – on the contrary, Swiss public authorities were very concerned that Italian workers could import socialist ideas with them (Cerutti, 2000, 106.) - and the duration necessary to access permanent permits was extended from 5 to 10 years, making it more difficult access residency status (Cerutti, 2000, 92.). The explicit guidelines of immigration control were to favour precarious permits, notably seasonal permits, so that immigration could be stopped in case of a new recession, and settlement of families was prevented (Cerutti, 2000, 93.). Somewhat paradoxically, this immigration policy designed to make settlement less attractive by granting few rights, combined with steady economic growth led to an explosion in immigration flows. The number of immigrants in Switzerland increased from 285’000 in 1950 to nearly 600’000 in 1960, and more than a million in 1970. In the beginning, the overwhelming majority of immigrants were from Italy, shifting then to Spain, Yugoslavia and Portugal (Piguet, 2005, 51.). Similar to Sweden, the 1960s witnessed growing concerns about the influence that immigration would have on Swiss society. However, both the nature of these concerns and the options presented to solve them were radically different. First, even if trade unions were a critical voice, unlike their Swedish counterparts they yielded little influence in immigration policymaking (Cattacin, 1987). The most important challengers to guest worker programs were
14
anti-immigration movements who threatened to use the tool of direct democracy to put an end to labour migration via a referendum; Überfremdung (“overforeignisation”) was becoming a growing cause for concern. A commission was appointed in 1964 to outline strategies to solve the problem of foreign labour recruitment and Überfremdung (Conseil Fédéral, 1965; OFIAMT, 1964). Compared to similar debates in Sweden at the time, the most striking difference is the absence of female labour force participation as a potential alternative to guest-worker programs. Among policy strategies to expand the labour supply, only the rationalization of the production apparatus and the increase in productivity to lower the demand for labour were considered as realistic options. The potential to increase labour force participation amongst the domestic workforce was considered exhausted in spite of the fact that the participation rate of women was below 40% at the time. Rationalization was favoured because the access to cheap labour had made it possible for many economic sectors to delay modernization and expand their labour force instead (Conseil Fédéral, 1965, 338.). The only policy proposals contained in the report which made explicit references to women were increasing opportunities for home-based work (so that married women could continue taking care of children in the home while doing paid employment) or part-time employment (OFIAMT, 1964, 191.), besides delaying male retirement. Strikingly, the expansion of public childcare to facilitate the employment of mothers was never mentioned. During the 1960s, the government put in place a system of quotas to limit the recruitment of foreign labour. However, immigration flows continued to grow and only slowed down in the mid-1970s, because of the oil crisis (Afonso, 2005). In the 1980s, labour migration resumed while it had essentially stopped in Sweden. One possible facilitating factor for the continuation of labour migration in Switzerland was precisely the absence of adequate welfare protection. Because most migrant workers were not covered by unemployment insurance, a large number of them who lost their jobs left the country, and unemployment did not increase in spite of a considerable contraction in jobs (Flückiger, 1992). Hence, the social costs of immigration policy were not made visible.
Figure 5: Public spending on family policies: family allowances, parental leave and childcare in Sweden and Switzerland, 1980-2010. Figure 5 about here
15
Source: Brewster and Rindfuss (2000, 276.). Socx database only provides data from 1980 onward. In Switzerland, policy choices in terms of labour supply clearly favoured migrant workers over working mothers. In contrast to Sweden, public spending on childcare remained marginal, and the rate of female employment, even if it was fairly close to the Swedish rate in the early 1960s, stagnated up until the 1990s while it increased considerably in Sweden. By any standard, Swiss family policies, and especially policies aimed at supporting working mothers, lagged behind most neighbouring countries (Kuebler, 2007, 217.). For most of the 20th century, state-sponsored family policies in Switzerland have been limited to family allowances introduced in all cantons during the 1950s and 1960s (Kuebler, 2007, 219.)(Figure 5). Until the 1980s, maternity insurance, maternal leave and care infrastructure were non-existent, and only began to catch up in the 1990s and 200s (Häusermann, 2010, 214.).
Figure 6: Share of Women in Parliament in Sweden and Switzerland 1945-2003 Figure 6 about here Source: (Paxton et al., 2008)
While family allowances could be reconciled with a traditional one-earner family model, attempts at welfare schemes which could facilitate a more balanced participation of women on the labour market repeatedly failed. While a constitutional article providing for a public paid maternity leave was accepted in 1945, it only materialized in actual legislation 60 years later (Häusermann, 2010). In 1964, the government rejected the introduction of maternity paid leave (Kuebler, 2007, 219.), and another attempt was voted down in a referendum in 1984. In line with the liberal-conservative model mentioned above, state intervention to promote the reconciliation of family and working life was not considered a policy priority as in Sweden, because it would challenge deeply entrenched traditional gender roles, but also entail a significant expansion of the state that dominant right-wing parties (particularly ChristianDemocratic and Liberal) were not ready to accept (Kuebler, 2007, 225.). A fairly obvious factor to explain this aversion to female employment, besides the male breadwinner model that prevailed in similar countries – Germany, Austria or the Netherlands – was the lack of female representation in parliament until 1971, when Swiss women were eventually granted suffrage. This can be considered a component of the gender norms mentioned above as an intrinsic component of the welfare model. Accordingly, the expansion and “catching up” of Swiss
16
welfare spending with respect to neighbouring countries only took place in a belated way, and female suffrage has often been considered an important driver behind this.
6. Conclusion In this article, I have shown the interactions between immigration policy and welfare as alternative labour supply policies in the post-war period, and how immigration policy could pursue social policy goals by other means. The comparison of Switzerland and Sweden has shown how the setup of welfare regimes and domestic political power relationships have structured policy choices in favour of one or the other strategy. In this context, guest worker programs emerged as an expedient complement to the Conservative Swiss welfare regime, as it allowed to expand the labour supply at a relatively low cost, without expanding welfare programs – and therefore taxation – and without challenging deeply entrenched gender norms. This took place in a context of dominance of Liberal and Conservative political forces, and the delayed extension of suffrage to women. In Sweden, guest worker programs were seen as a challenge to the Swedish social welfare model, and the expansion of female employment via employment-friendly welfare programs was fostered by a more favourable set of power relationships. Female employment had already been higher, strong trade unions were favourable to an expansion of state intervention, and the conservative one-earner family model was less prevalent. A number of the limits and options for further research can be outlined. First, it could be argued that female and (mostly male) migrant employment are not completely interchangeable, and that the available labour supply strategies could be dictated by different sectoral setups of the economy. Indeed, the largest receiver of foreign labour in Switzerland in the 1950s and 1960s were construction and metallurgy, two sectors with a very pronounced male gender bias. In contrast, the take-off in female employment in Sweden form the 1960s onwards was underpinned by the expansion of public health and social services, where the gender balance tips towards women. It remains an open question to determine whether this was a dependent or independent variable, however. The labour supply can also strongly shape the patterns of expansion of economic sectors (Wright and Dwyer, 2003). The second question is the external validity of the findings presented here based on two countries. If Switzerland may be the clearest case of labour expansion via guest worker programs, these also played an important role in other Bismarckian countries such as France (Weil, 1995), Germany (Herbert, 2001) or the Netherlands (Penninx and Roosblad, 2000). In contrast, other Scandinavian countries, particularly Denmark and Finland, have also witnessed a much smaller scale of labour
17
migration. The applicability of these findings to a wider set of countries should be explored, providing reliable indicators to control for the restrictiveness of immigration control can be used. In terms of temporal scope, I have focused on the period up to the late 1980s because migration was mostly driven by labour market concerns. In the 1990s and 2000s, the differences highlighted here tend to fade away as immigration and female employment increased almost everywhere. Drawing on the present analysis, an important future research agenda will consist in examining the different ways whereby immigration can pursue social policy goals by others means in both sending and receiving countries. While I have focussed on the labour supply function of social policy, in some countries migrant workers deliver private services (childcare; elderly care) that are not adequately provided by the state (Sciortino, 2004; van Hooren, 2012). In sending countries, remittances sent by emigrants can act as an insurance mechanism equivalent to social spending, potentially dampening demand for “conventional” social policies (Doyle, 2015). Hence, immigration constitutes a promising way to explore how social policy goals can be pursued by other means.
Bibliography Afonso, A. 2005. “When the Export of Social Problems is No Longer Possible: Immigration Policies and Unemployment in Switzerland”. Social Policy & Administration 39: 653668. Afonso, A, and C Devitt. 2016. “Comparative Political Economy and International Migration”. Socio-Economic Review 14: -613. Brewster, KL, and RR Rindfuss. 2000. “Fertility and Women’s Employment in Industrialized Nations”. Annual review of sociology 271-296. Borevi, K. 2014. "Sweden: The Flagship of Multiculturalism", in Immigration Policy and the Scandinavian Welfare State 1945-2010. Basingstoke: Palgrave: 25-97. Cattacin, M. 2006. “Retard, Rattrappage, Normalisation. L’etat Social Suisse Face Aux Défis de Transformation de la Sécurité Sociale.”. Studien und Quellen 31:49-77. Cattacin, Sandro. 1987. Neokorporatismus in Der Schweiz: Die Fremdarbeiterpolitik. Zürich: Forschungsstelle für Politische Wissenschaft. Cerutti, Mauro. 2000. “La Politique Migratoire de la Suisse 1945-1970.” In Histoire de la Politique de Migration, D’Asile et D’Intégration En Suisse Depuis 1948. Research Report for the Swiss National Fund for Scientific Research, ed. Hans Mahnig, and Sandro Cattacin. Neuchatel and Geneva: Swiss forum for the Study of migrations and population.
18
Conseil Fédéral. 1965. “Rapport Du Conseil Fédéral À la Commission Élargie Des Affaires Étrangères Du Conseil National sur la Limitation et la Réduction de L’Effectif de Travailleurs Étrangers”. Feuille Fédérale 1965:334-360. Devitt, C. 2015. “Mothers Or Migrants? Labour Supply Policies in Ireland 1997-2007”. Social Politics: International Studies in Gender, State & Society Doyle, D. 2015. "Remittances and Social Spending." American Political Science Review 109 (4):785–802. Eichengreen, B. 2008. The European Economy Since 1945. Princeton: Princeton University Press. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Flückiger, Yves. 1992. “La Politique Suisse En Matière D’Immigration.” In Main D’Oeuvre Étrangère: Une Analyse de L’Économie Suisse, ed. Beat Bürgenmeier. Paris: Economica. Gunnarsson, Lars, Barbara Martin Korpi, and Ulla Nordenstam. 1999. Early Childhood Education and Care Policy in Sweden: Background Report Prepared for the Oecd Thematic Review. Ministry of Education and Science in Sweden Hammar, Tomas, ed. 1985. European Immigration Policy: A Comparative Study Cambridge: Cambridge University Press. Häusermann, S. 2009. “Solidarity With Whom? Why Organised Labour is Losing Ground in Continental Pension Politics”. European Journal of Political Research 49:223-256. Häusermann, S. 2010. “Reform Opportunities in a Bismarckian Latecomer: Restructuring the Swiss Welfare State.” In A Long Good-Bye to Bismarck? The Politics of Welfare Reform in Continental Europe, ed. B Palier. Amsterdam: Amsterdam University Press. Häusermann, S, and B Palier. 2008. “The Politics of Employment-Friendly Welfare Reforms in Post-Industrial Economies”. Socio-Economic Review 6:559-586. Herbert, Ulrich. 2001. Geschichte Der Ausländerpolitik in Deutschland. Munich: C.H Beck. Immergut, Ellen M. 1992. Health Politics : Interests and Institutions in Western Europe. Cambridge: Cambridge University Press. Iversen, T, and A Wren. 1998. “Equality, Employment, and Budgetary Restraint: The Trilemma of the Service Economy”. World Politics 50:507-546. Johansson, Jesper. 2008. Så Gör Vi Inte Här I Sverige. Vi Brukar Göra Så Här: Retorik Och Praktik I Lo:s Invandrarpolitik 1945-1981. Växjö: Växjö University Press. Johansson, J. 2014. “Swedish Employers and Trade Unions, Labor Migration and the Welfare State—perspectives on Swedish Labor Migration Policy Debates During the 1960s and the 2000s”. Nordic Journal of Working Life Studies 4:97-118.
19
Katzenstein, Peter J. 1985. Small States in World Markets: Industrial Policy in Europe. Ithaca, NY: Cornell University Press. Kersbergen, Kees van. 1995. Social Capitalism : A Study of Christian Democracy and the Welfare State. London: Routledge. Kindleberger, Charles Poor. 1967. Europe’s Postwar Growth: The Role of Labor Supply. Cambridge, MA: Harvard University Press. Knocke, W. 2000. “Sweden: Insiders Outside the Trade Union Mainstream.” In Trade Unions, Immigration, and Immigrants in Europe, 1960-1993: A Comparative Study of the Attitudes and Actions of Trade Unions in Seven West European Countries, ed. R. Penninx, and J. Roosblad. New York: Berghahn books. Kuebler, D. 2007. “Understanding the Recent Expansion of Swiss Family Policy: An IdeaCentred Approach”. Journal of Social Policy 36:217-237. Kyle, G. 1979. Gästarbeterska I Mansamhället. Stockholm: Liberförlag. Lewis, J. 1992. “Gender and the Development of Welfare Regimes”. Journal of European social policy 2:159-173. Martin, PL, and MJ Miller. 1980. “Guestworkers: Lessons From Western Europe”. ILR Review 33:315-330. Morel, N. 2007. “From Subsidiarity to ‘free Choice’: Child and Elder Care Policy Reforms in France, Belgium, Germany and the Netherlands”. Social Policy & Administration 41:618637. Naumann, IK. 2005. “Child Care and Feminism in West Germany and Sweden in the 1960s and 1970s”. Journal of European Social Policy 15:47-63. Nyberg, A. 2000. “From Foster Mothers to Child Care Centers: A History of Working Mothers and Child Care in Sweden”. Feminist Economics 6:5-20. Obinger, H. 1998. “Federalism, Direct Democracy, and Welfare State Development in Switzerland”. Journal of Public Policy 18:241-263. OECD. 2014. Pf3.1: Public Spending on Childcare and Early Education. Paris: OECD. Oesch, Daniel. 2013. Occupational Change in Europe: How Technology and Education Transform the Job Structure. Oxford: Oxford University Press. OFIAMT. 1964. Le Problème de la Main D’Oeuvre Étrangère. Bern: OFIAMT. Paxton, P, J Green, and M Hughes. 2008. “Women in Parliament, 1945-2003: Cross-National Dataset”. Penninx, R., and J. Roosblad. 2000. “Introduction.” In Trade Unions, Immigration, and Immigrants in Europe, 1960-1993: A Comparative Study of the Attitudes and Actions of Trade Unions in Seven West European Countries, ed. R. Penninx, and J. Roosblad. New
20
York: Berghahn books. Piguet, E. 2005. “L’immigration En Suisse Depuis 1948 - Contexte et Conséquences Des Politiques D’immigration, D’intégration et D’asile.” In Histoire de la Politique de Migration, D’aile et D’intégration Depuis 1948, ed. Hans Mahnig. Zurich: Seismo. Piore, Michael. 1979. Birds of Passage: Migrant Labour and Industrial Societies. Cambridge: Cambridge University Press. Sainsbury, D. 1999. “Gender and Welfare State Regimes”. Oxford: Oxford University Press. Sainsbury, Diane. 2012. Welfare States and Immigrant Rights: The Politics of Inclusion and Exclusion. Oxford: Oxford University Press. Sapir, A. 2006. “Globalization and the Reform of European Social Models*”. JCMS: Journal of Common Market Studies 44:369-390. Sciortino, G. 2004. “Immigration in a Mediterranean Welfare State: The Italian Experience in Comparative Perspective”. Journal of comparative policy analysis: research and practice 6:111-129. Trampusch, Christine, and André Mach, eds. 2011. Switzerland in Europe: Continuity and Change in the Swiss Political Economy (Routledge Advances in European Politics) Routledge. Van Hooren, FJ. 2012. “Varieties of Migrant Care Work: Comparing Patterns of Migrant Labour in Social Care”. Journal of European Social Policy 22:133-147. von Rhein-Kress, Gaby. 1993. “Coping With Economic Crisis: Labour Supply as a Policy Instrument.” In Families of Nations, ed. FG Castles. Brookfield: Dartmouth. Weil, Patrick. 1995. La France et Ses Étrangers: L’Aventure D’Une Politique de L’Immigration de 1938 À Nos Jours. Paris: Gallimard. Wright, EO, and RE Dwyer. 2003. “The Patterns of Job Expansions in the USA: A Comparison of the 1960s and 1990s”. Socio-Economic Review 1:289-325.
21
Figures 1.
22
2.
23
3.
24
4.
25
5.
26
6.
27
