International trade-unionism and migration: European integration and the post-war ‘Free’ movement of labour

ABSTRACT The aim of this article is to study the significance of international trade-union cooperation during the first two post-Second-World-War decades as regards the development of a more liberal labour-migration regime in Western Europe. The article illuminates the mutual interplay between the national and international level of the trade-union movement. Mainly, it focuses on discussions within the International Metalworkers’ Federation, but also the International Confederation of Free Trade Unions. Thereby it creates an analytical bridge between the two different structures of the socialist/social democratic international trade-union movement. The article shows that trade unionists generally articulated positive attitudes to migration, at least as long as they were possible to control. Trade unionists connected migration with population and reconstruction issues, and tied movements of labour to a broader policy that aimed at a more effective use of all the means of production available. At the same time, they construed migration as a short-term solution to counter imbalances in Europe between the supply and demand for labour. The goal in the longer run was national labour markets in full employment, where all workers had similar conditions, which would decrease migration incentives.


Introduction
Scholarship on European trade unionism and international migration often adopts a state-centred approach or a cross-country perspective (e.g. Merino et al., 2017;Penninx & Roosblad, 2000). This is not surprising, considering the importance of nation-state borders since the First World War, and the fact that workers have been unionised in local and national trade unions. Yet there is a risk that this approach overlooks significant transnational contacts between trade unionists. When trade unionists from different countries met within international organisations and talked about migration -among many other issues -they learned from each other's experiences and normalised their practices, at the same time as they developed a better understanding of both coinciding and colliding national interests.
After a very restrictive period, stretching from the outbreak of the First World War to the end of the Second World War, many European politicians and industrialists favoured a more liberal migration regime (on employers, e.g. Caviedes, 2010;Jansson, 2018Jansson, , 2020. Following Italian pressure, the agreement of the European Coal and Steel Community (ECSC), signed in 1951, included a paragraph stipulating free movement for (skilled) coal and steel workers. The Treaties of Rome, signed in 1957, followed this path of liberalisation, and paved the way for a European labour market open for all citizens within the community (e.g. Goedings, 2005). However, neither labour historians nor minutes of the smaller Advisory Board (which became the Executive Committee in the middle of the 1950s), where migration issues were sometimes discussed as well.
In addition, I have analysed sources from the conferences I mentioned above, organised by the ICFTU and the IMF respectively. The minutes from the European-integration conference in 1962 are detailed, but unfortunately, I have not found any equivalent discussion protocol from the ICFTUmigration-conference, only preparatory documentation and the statement that the participants formulated by the end of the conference. Yet the ICFTU Secretariat wrote the preparatory documentation based on a previous questionnaire where the national centres had the opportunity to answer questions about migration and state their opinions, and these answers are preserved in the archive. The final statement from the conference is also important, considering the dynamics of international trade-unionism. Generally, trade-union discussions on the international level ended up in official resolutions, which the organisations published in their journals. The IMF, for instance, spread its resolutions on different issues through the Bulletin of the International Metalworkers' Federation. These resolutions would supposedly function as guidelines for the affiliates. Simultaneously, they were tools for the international trade-union organisations to communicate with a single voice outwards, with the surrounding society. In other words, the resolutions/statements were international trade-union agency.

Migration and trade-union dilemmas
International migration affects the relationship between the supply and demand for labour on local and national labour markets. In theory, immigration undermines the relative strength of the trade unions, considering their negotiations with employers about wages and working conditions; when the supply of labour increases, employers' incentives to pay higher wages or improve the working environment decrease. Accordingly, since the breakthrough of industrial and agrarian capitalism in the middle of the nineteenth century, migration has been an important aspect of the relationship between labour and capital (e.g. Guerin-Gonzales & Strikwerda, 1998).
Because employers usually initiate labour migrations -often in cooperation with the statetrade unions are in this respect secondary organisations. Trade unions have to react if they want to defend their labour-market interests. The migration scholars Penninx and Roosblad (2000, pp. 4-12), discern three basic dilemmas which confront trade unionists in such situations. First, 'should trade unions cooperate with employers and authorities in the employment of foreign workers ' (p. 4), or should they use their power resources and try to resist? Accordingly, they have to balance migration as a potential labour-market threat against potential welfare gains for the native working class (if the state can turn growth into welfare). Possibly, cooperation could also find strength in an internationalist ideology stressing solidarity with unemployed fellow class-mates in other countries.
Penninx and Roosblad's second dilemma emerges when workers arrive in another country: 'should trade unions include them fully in their ranks or exclude them as a special category? ' (2000, p. 4) Exclusion would mean a weaker position in the relationship with the employers, but inclusion could be perceived as a 'threat to nationally defined labour market or trade union interests' (p. 9). In their pioneering study, Castles and Kosack (1973) pointed out that it could seem logical for trade unionists to oppose immigration for economic reasons, but once they were in place it became very important to unionise all migrant workers. For the trade unionists in the receiving society there was thus a risk of a contradiction in the policies, 'between trade union policies towards immigration on the one hand', and policies towards migrant workers 'once they are in the country, on the other' (p. 128). If trade unionists oppose immigration, they may find that immigrants do not trust them and are unwilling to join. Where this happens, the unions have the worst of two worlds. Not strong enough to prevent immigration, their attempt to do so only serve to alienate the new workers from them. The result is a weakening of the unions and deepening split in the working class. (Castles & Kosack, 1973, p. 128.) If trade unionists in the receiving country choose inclusion, a third dilemma arises, according to Penninx and Roosblad: 'should they advocate and implement special measures for these immigrants or should they insist on general, equal treatment for all workers? ' (2000, p. 4) Trade unionists might use a supposed lack of knowledge about the organisational structure, the culture within the specific national movement, or language difficulties, to justify educational efforts, but as the migration scholars point out, preferential treatment of migrant workers might lead to dissatisfaction among native-born workers.
However, these basic dilemmas assume a rather static view of the labour market. The 'free' movement of labour and other features of the European labour market were not something fixed and finished, just waiting for trade-union reactions. The international labour market in Western and Northern Europe rather developed in a tug of war between different organised interests. I argue in this article that the international trade-union movement tried to gain influence and act -not only react -in relation to the development of the liberal labour-migration regime. Therefore, it is important to study the first post-war decades, when this migration regime started to evolve. Furthermore, I argue that the trade-union interests were not static over time, but continuously renegotiated internally at different levels of the trade-union movement, with respect to changing circumstances in the political and economic context.

An important issue in the margin of the agenda
Migration was not one of the questions that the IMF dealt with most often during the post-war period. As regards the early process of European integration, the trade unionists more often talked about power relations in the Ruhr and the supposed danger of communism. In general, the developing Cold War permeated all post-war discussions. 1 Nevertheless, migration was an issue considered now and then during the IMF's meetings, regardless of whether it was on the actual agenda or not. On such occasions, trade unionists usually connected migration with other economic or political matters that actually were on the agenda. The IMF's integration-positive 'Resolution on the European Coal and Steel Community' from 1955, for example, foresaw and accepted increased migrations; it described 'the need' for 'transferability of workers'. 2 Already when the IMF Central Committee discussed the Marshall Plan, in 1948, the American trade unionist -and hard-core anti-communist -Irving Brown took the opportunity to argue in favour of a positive attitude to labour mobility. Such an attitude was necessary, he pointed out, because only through international migrations could the contemporary problem of 'too much' and 'too little' manpower in different European countries be solved: for the Marshall Plan to work to the full extent, migrations were required. 3 Thereby, migration became a part of Brown's anti-communist campaign in Europe: labour emigration, seen in this perspective, could potentially undermine communism's growth possibilities in perceived over-populated areas (read Italy).
So, even if migration was not on the official IMF agenda, individual delegates could raise the issue anyhow. A telling example is the Italian Arturo Chiari's plea during the Central Committee meeting in Stockholm in 1953, under the agenda point 'any other business', to get his fellow trade unionists to pay more attention to Italian workers' migration interest. Chiari tried to convince his colleagues to accept an official resolution which would pledge the IMF to do all it could 'to remove obstacles in the way of the free transferability of manpower'. He pointed out that Italians were 'in a very difficult position on account of the serious unemployment'. According to Chiari, Italian metalworkers were particularly exposed: 120,000 Italian metalworkers were unemployed, and he foresaw 'that a further 27,000 workers may be thrown out of employment by coming into force of the Schuman Plan'. His understanding of the Plan, hence, included a structural transformation of the European trade and industry, which meant that Italian industries could be outcompeted. Therefore, he finally emphasised, it was the 'duty of the International to promote free transferability of manpower', which had been 'the normal state of affairs prevailing in the past'. 4 Chiari thus mustered strong historical arguments for his plea, with a rather explicit reference to international solidarity and a more implicit reference to the liberal migration regime of the late 1800s and early 1900s. In fact, migration had been an important issue when European socialists created their very first international forums for cooperation during the second half of the nineteenth century. As the proletarianised European population increased during the development of agrarian and industrial capitalism, and when powerbrokers deregulated labour markets, both inter-European and trans-Atlantic migrations increased in scope. An ambition of the First International (International Working Men's Association), founded in 1864, was to gain control over the international labour market and if possible prevent employers from recruiting strike breakers in other countries. Likewise trade unionists' formation of several different ITS by the turn of the century was partly a reaction to ongoing migrations. From the contemporary socialists' point of view, cooperation at international level was necessary if they wanted to influence or respond to an international phenomenon like migration (Olsson, 2001, pp. 8195-6;Van Der Linden, 2003, pp. 14, 16, 19).
During that earlier period in time, not only Italians but also millions of other European workers had had the opportunity to leave unemployment or other dire social situations behind, and search for better opportunities abroad and overseas. Several of the other men -they were all men -in the IMF Central Committee were from countries from where large numbers of workers had made international journeys in the past; the delegates represented workers in countries where emigration had eased previous pressures of over-population and processes of proletarianisation. That, and the fact that the international trade-union movement had a tradition of dealing with migrations, was the content of Chiari's narrative and history lesson.
In the end, Chiari's plea won approbation in the Central Committee, even though other delegates voiced both warnings and more restrictive attitudes to migration. Harry Douglas and Jack Tanner from Great Britain emphasised, in turn, 'that some countries' which had recruited Italian workers 'are bothered with the problem of manpower being infiltrated into their territory for the purpose of Communistic indoctrination', and that 'all European countries' therefore want to make sure 'that none of that dangerous type of worker is to be employed'. 5 Konrad Ilg -Secretary and a leading figure in the IMF from the early 1920s until his death in 1954 -appealed to the present trade unionists' 'conscience as internationalists', although he urged at least some caution too. From his Swiss point of view, he argued that migrant workers were difficult to unionise: if labour migrations were to be liberalised in the manner suggested by Chiari, the trade unions needed to improve their strategies for unionising all migrant workers, otherwise they would lose relative strength vis-à-vis the employers. Ilg pointed to Sweden as a good example. According to the IMF Secretary, Swedish trade unions had 'reached an agreement with the employers that no Italian should be permitted to enter employment without joining the organisation'. 6 Probably, he had in mind the bilateral agreement between Sweden and Italy from 1947. That actually stipulated mandatory trade-union affiliation for all migrant workers whom Swedish employers recruited in accordance with that settlement (Järtelius, 1987). Even though Ilg did not elaborate on all the details about the Swedish-Italian agreement before the Central Committee, he did use the international trade-union arena to spread knowledge about supposed successful unionising strategies.
Before the Central Committee could approve the resolution, it needed to clarify one question. Some language problems hampered the Italian initiative. Obviously, the English translation went further in its demands than intended by the Italians. According to the English translation, the IMF should work to 'abolish' all regulations which restricted the free international transferability of workers. Tanner probably spoke for many in the meeting room when he underlined that they did 'not want one word in Italian which has not the same meaning in the English, German or French text'. It was easier for them all to agree about a policy that only spoke about modifying the migration regime. The final resolution stated that the IMF had 'examined the state of employment in various countries', which besides causing economic hardship for the workers, is leading to serious difficulties in the way of strengthening and development of the democratic trade-union movement. The Central Committee, with a view to contributing to the mitigation of such special difficulties [. . .] calls on all national organisations to exert pressure on their respective Governments to modify regulations which restrict the free transferability of manpower from one country to another. 7 The Cold War, and the conflict between social democrats and communists within the labour movement, certainly affected IMF's discussions about migration. On the one hand, the trade unionists concerned warned against opening up for free migration, because communists could infiltrate industries in other countries. Among the IMF delegates from supposed destination countries, there was seemingly a consensus that governments should keep some regulatory measures and perhaps even control migrants' political sympathies. On the other hand -and in line with the overall positive IMF stance towards European integration -the trade unionists perceived the high unemployment rate in Italy as a fertile ground for the communist movement to grow. From that point of view, it was better to decrease the pressure on the Italian labour market, and allow Italians to seek employment abroad, than to let the Italian working class sink deeper into social despair and thereby be radicalised.
It is important to note that the IMF resolution was wider than only considering movements between 'the inner six'. Besides the initiating Italian, several of the trade unionists who took part in this discussion were from countries outside the ECSC. During the 1950s and 1960s, national laws and bilateral agreements regulated the lion's share of all European labour migrations. That the issue of migration still stood relatively free from the process of European integration was also obvious when the IMF talked about the more technical question of international transfer of trade-union membership rights.

International membership transfer
In September 1951, Algemene Nederlandse Metaal Bedrijfsbond reminded Konrad Ilg about the great migrations from the Netherlands after the war, mainly to Canada, Australia and New Zealand, and to some extent to South Africa. It pleaded that the Secretary should pay attention to those problems which often arose when Dutch migrants tried to convert their trade-union membership in the country of emigration to an affiliation in the immigration country. It was notable, the letter argued, that the migrants were treated as if they were joining a trade union for the very first time. In Canada, migrants had to pay admission fees although they had been members of the Dutch Metal Workers' Union; considering that both the Canadian and Dutch trade unions concerned were affiliated with the IMF it should not be so difficult to transfer individual memberships between the countries. As regards Australia, New Zealand and South Africa, the membership issue was more problematic, the author of the letter admitted, because these countries did not have any IMF affiliates. But at least organisations within the IMF should cooperate more, the Dutch plea concluded. 8 Similarly, the Swedish Metal Workers' Union argued that the IMF should regulate free membership transfers on international level, and that migrant workers should be able to credit their already earned membership rights in their new countries. 9 In the early 1950s, the IMF was not the only ITS that discussed migration and free international membership transfers. Seemingly as a source of inspiration for the IMF, the International Graphical Federation (IGF) emphasised in January 1951 that 'a most liberal application of the principle of freedom to move from country to county' was in the interest of individual trade-union members. The IGF underscored that it was an expression of 'trade-union solidarity when the associations and their organs are in every way helpful to members of other associations who express the wish to take up positions abroad'. 10 For the IGF, this principle rested upon a long tradition of inter-European journeymen travels, not least among typographers (see e.g. Edgren, 1993;Edgren & Olsson, 1993;Olsson, 1993).
However, in developing welfare states the issue of transferring individual membership rights was more complicated than during the journeymen migrations of the nineteenth century. Rights and benefits, such as unemployment support and pensions, were often tied to membership of a trade union, but in different ways in different countries. Trade unions often also co-managed socialinsurance arrangements together with state authorities. Therefore, the trade unions could not singlehandedly change the provisions. In that way, different national circumstances, and the tradeunion movement's increased integration with different welfare states, restrained the practice of trade-union internationalism.
With such hindrances in mind, the Dutch argument and the Swedish initiative stopped at a recommendation from the IMF to the national trade unions to 'grant members of other national organisations affiliated with the IMF free admission'. The IMF also encouraged its affiliates to note in the individual membership books how long the individual worker had been a member. That would make it easier for the shop stewards to see which of the migrant workers had been loyal to the trade unions in the emigration country. Instead of making free membership transfers something collectively organised by the IMF, the international organisation advocated bilateral agreements between trade unions in different countries. When the affiliates signed such bilateral agreements, they were also encouraged to notify the IMF Secretariat, so that it could inform all other affiliates through the organisational journal. 11 During the 1950s, IMF member organisations signed several such bilateral or multilateral agreements. For example, in 1954 the three affiliates from the Benelux countries agreed on mutually accepted membership for all unionised metal workers in the region, in a deal that carefully described which social insurances were applicable in the three countries (e.g. in the event of labour-market conflicts or work accidents). 12 In 1956, the Swedish and the West German Metal Workers' Unions signed a similar agreement, stipulating free transfers of membership rights. Besides a mutual Scandinavian agreement, by the middle of the 1960s the Swedish trade union had made similar contracts with sibling organisations in Switzerland and Austria. The Dutch IMF affiliate by that point in time had signed agreements with trade unions in West Germany and Austria, and so on. 13 At first glance, it is noteworthy that none of the IMF affiliates from the labour-recruiting countries in Northern and Western Europe signed agreements with either of the two Italian member organisations. Probably this was partly because trade unionists in the labour-recruiting countries supposed that the differences between those forms of social insurance that were tied to the trade-union membership were too great. Most likely, this was partly also due to the relatively low level of unionised workers in the two non-Communist Metal Workers' Unions in Italy. Furthermore, the trade unionists in Northern and Western Europe generally regarded the northbound migrations as a part of an agrarian-urban movement. The discussion in the IMF about free international transfer of membership rights only concerned already unionised metal and engineering workers.
This dialogue developed into a lengthy one, continuing well into the 1960s. Still, trade unionists from the Netherlands put the question on the IMF agenda, but by this point in time as representatives of an organisation in a country on the verge of transforming into an 'immigration country'. Alfred Dannenberg, Assistant General Secretary, told the Central Committee that the Dutch were keenly interested in finding a common solution, 'as members of the Common Market', and because of increased labour mobility both within and to Europe. In 1964, almost thirteen years after the Dutch trade union first put the issue on the IMF's agenda, the international organisation appointed a fourdelegate committee to investigate both possibilities and problems. 14 Another two years later -the trade-union machinery could work rather slowly -the IMF discussed the suggestions from the committee. As had been pointed out already during the 1950s, rapporteur Dannenberg was sceptical considering the national differences in sickness allowances, unemployment relief and in all other welfare benefits that were tied to the individual trade-union membership. Because the benefits often depended on members' prior payments to collective funds, the national trade unions could not take full responsibility for arriving migrant workers. Nevertheless, Dannenberg underlined that some 'exceptions dictated by tactics' were desirable: in case of labourmarket conflicts, it would be wise to pay all workers participating in the conflicts the same amount of support. 15 In that way, similar treatment of all workers was a vested interest. International membership rules might be considered as a strategy aiming at a restraint of the free market; the trade unions should not allow employers to exploit migrant workers only because they omitted to pay the newly arrived workers any relief. Yet the fact that the discussion ended up only in a recommendation of free transfer of membership rights between IMF affiliates illuminates the limitations of trade-union internationalism. 16 If the IMF gave directives to the national trade unions about strike benefits for migrant workers, it would have forced the international organisation to ensure that such funds were available. The international level of the trade-union movement developed guidelines, not directives.
Likewise, the national trade-union centres paid attention to the post-war labour mobility in Europe. From the founding congress in 1949, trade unionists discussed population issues in general and migration in particular within the ICFTU. Under the ICFTU umbrella, the international tradeunion movement could also connect its two different organisational structures, as when the confederation invited the ITS to participate in a special migration conference in 1956. That the ICFTU chose to gather a special conference on this theme is in itself a sign of the importance that trade unionists assigned to migration. 17

Migration on the ICFTU agenda
At its first congress in London, in 1949, the ICFTU stipulated an overall manifesto for the reconstruction of Europe. The slogan 'Bread, Freedom and Peace' summarised the confederation's social, economic and political aims. The ICFTU prioritised reconstruction in Western Europe, while downplaying trade-union activities in the rest of the world. Anti-communist Europeans and North Americans founded the organisation, it was a Cold-War product, and during the late 1940s the conflict between the two belligerent power blocs was centred on the European scene . The ICFTU supported the Marshall Plan with the same anti-communist motive as the IMF: for Western Europe to be able to rise again out of the devastation of war, and in the longer run to be able to keep up the defence against the eastern threat, there was no place for large numbers of idle workers. That was the starting point for many ICFTU affiliates when the trade unionists started to talk about migration. At the London congress, the ICFTU declared that migration was a potential strategy to deal with the sustentation problems of 'over-populated' countries: We recognise that in some few countries the needs of the population may be so great in relation to available resources that no reasonable amount of national or international effort can provide them, where they are, with an adequate standard of living. In such cases the voluntary migration of workers should be encouraged and facilitated, with adequate safeguards both for themselves and for the workers of the countries to which they migrate. In all cases immigrants should be employed on conditions of work and at wages which are not less favourable than in the country of adaptation. (Cited after ICFTU, 1954, p. 10.) At the following ICFTU congresses, in Milan 1951, Stockholm 1953and Vienna 1955, the organisation repeated the need for emigration from (European) countries with high levels of unemployment. Accordingly, the official rhetoric was affirmative. The ICFTU encouraged labour mobility, at least as long as it did not threaten to increase the level of unemployment in the receiving countries or restrain wage increases. 18 Just as Arturo Chiari put migration on the IMF agenda, the Italian centre, Confederazione Italiana Sindicati Lavoratori (CISL), made sure that the ICFTU treated labour mobility more carefully than just as something mentioned in general statements. In 1952, the ICFTU accepted a proposal from the CISL, that the international trade-union movement should investigate migrants' work situations abroad. The Italian organisation demanded that the ICFTU work more actively to ensure that employers paid migrant workers the same wages and offered them the same working conditions and social benefits as the native-born. 19 The Italian representatives were obviously not pleased with how their fellow countrymen were treated abroad.
Within the ICFTU, it was the responsibility of the General Secretary, J.H. Oldenbroek, to conduct the investigation demanded by the CISL. He submitted a questionnaire to member organisations in both countries characterised by immigration and countries with labour emigration, but apparently he was mostly interested in experiences from trade unionists in labour-recruiting countries. The questions were generally asked in a receiving perspective, for example, the important one about the selection of migrants: 'Do you think that a stricter and more careful selection of immigrants should be made?' What was seemingly forgotten when the question was asked in that perspective was that the selection of migrants probably was just as important for the future of the Italian labour market as for the labour markets in northern and western Europe. Moreover, the questions dealt with migrants' willingness to join the trade unions, and asked which role language problems or cultural differences played during the unionising process. Otherwise, Oldenbroek asked those questions that interested the CISL, such as the level of migrants' wages, the migrants' working conditions and access to social benefits. 20 Affiliates from eleven countries answered the questions. Eight of them were organisations in countries that recruited foreign-born workers at this point in time (Belgium, Canada, France, Great Britain, New Zealand, Sweden, Switzerland and the USA), while three of them belonged to countries characterised more by emigration (Austria, Italy and the Netherlands). The results from the investigation were published in the ICFTU's journal, Free Labour World (1953, September, pp. 6-8), as well as in a seventy-page booklet with the title Problems of International Migration (ICFTU, 1954).
The centres from labour-recruiting countries stressed that employers did not discriminate against migrant workers as regards wages or working conditions (ICFTU, 1954, p. 34). The Schweizerischer Gewerkschaftsbund (SGB) claimed that Swiss trade unions were careful to ensure that employers never gave foreign-born workers inferior conditions in comparison with nationals: 'This is primarily a form of self-defence, to avoid competition from foreign workers', the answering trade unionist pointed out honestly. Landsorganisationen in Sweden (LO) similarly maintained that 'obviously it is of great importance that immigration can occur in an orderly way', and that 'chiefly workers with appropriate education are admitted into a country'. 21 At the same time, the SGB mentioned that an ethnic division of work had been established on the Swiss labour market. It was mainly employers in deficit sectors of the economy that recruited foreigners, sectors that Swiss workers 'now only seldom or hardly ever enter'. Assignments in agriculture, at hotels, as domestic servants, or as unskilled workers in the industries, meant working conditions that 'in themselves are slightly less good than in other trades and branches'. The LO normalised the ethnic divide too, stressing that it was 'quite obvious that migrants' in most cases 'have to begin with assignments which the native workforce consider less good'. 22 The CISL viewed the ethnic division of work in another perspective. The Italian organisation underlined that even if employers paid migrant workers equivalent wages when foreigners and natives performed similar tasks -which the trade unions demanded -the division of work on plant level nevertheless meant that Italians received lower wages. Probably with the piecework pay in mind, the CISL pointed out that employers often allotted migrant workers to assignments or machines with low wages. It furthermore underscored that employers discriminated against Italians regarding promotions (ICFTU, 1954, p. 35). Explicitly, the critique aimed at employers who made a distinction between foreign-born and native-born workers, but implicitly the Italian trade unionists also criticised trade unions in immigration countries for their lack of ability/will to defend the interests of migrants.
Several of the answering centres from labour-recruiting countries emphasised the lower level of unionisation among migrants. According to the LO, migrants' lack of interest in the trade unions was 'the main reason for the dissatisfaction with foreign comrades' among Swedish workers. The SGB, in turn, claimed that many migrant workers perceived the trade-union dues as unnecessary costs, because collective agreements or laws stipulated wages and work conditions in any case. 23 The CISL viewed the organisational issue in another light as well. The Italian organisation maintained that the relatively low level of unionised migrant workers was not only due to unwilling foreigners, but also to prejudices among trade-union organisers (ICFTU, 1954, pp. 38-39). Nevertheless, it is possible that these two tendencies reinforced each other. If migrant workers' lower degree of unionisation irritated native-born workers, and thereby strengthened the social dividing line between 'us' and 'them', the migrants might have interpreted this indignation as another expression of xenophobia among their new workmates.
Through this questionnaire and the published results, the ICFTU had pinpointed perceived migration-related problems. The confederation took the next step towards a more coherent migration policy in 1956, when it gathered and invited the ITS to a special migration conference at Nervi in Italy.

The Nervi conference
ICFTU-affiliates from 18 countries participated during the three-day migration conference, together with representatives from the IMF, the IGF, the International Federation of Industrial Organisations and General Workers' Unions and the International Landworkers' Federation. Charles Levinson represented the IMF, otherwise employed at its Secretariat. Present were also observers from the International Labour Organisation, the Organisation for European Economic Cooperation, the United Nations High Commissioner for Refugees and the Intergovernmental Committee for European Migration. 24 The agenda included broad issues such as the relationship between international migration and the standard of living for workers in both immigration and emigration countries, cooperation between trade unions and authorities dealing with migration, and the familiar questions about international transfer of trade-union membership and migrants' access to social benefits. 25 The written documents that the ICFTU Secretariat prepared before the conference expressed the same positive attitude to migration as the declarations of principle at the previous congresses. Supposedly, immigration as well as emigration countries gained from labour mobility. As soon as migrants started to work in another country, they become sources of wealth, both through the product of their work and through their consumption capacity, and this in turn benefits the economy of the immigration country. Furthermore, by relieving some of the pressure of population on the scarce resources of the emigration countries, international migration also contributes to the economic well-being of the workers living there. 26 Such general statements did not single out individual countries. Anyhow, it must have been obvious for the conference participants that the author meant Italy while writing about population pressure in a non-defined emigration country. Italian trade unionists had been active in putting migration on the international trade-union agenda, the conference was arranged in Italy with an Italian chairman, and the Italian centres had the largest delegations on site. That Italy was overpopulated was a widespread idea, not only among trade unionists.
Nonetheless, the conference problematised the idea of overpopulation. Already in one of the preparatory documents, the participants could read that over-as well as underpopulation were relative notions. When the ICFTU used the term overpopulation, it meant that available resources and/or capital were too scarce to reach full employment, or that powerbrokers and industrialists used available resources/capital far too irrationally to decrease the unemployment levels. If investors supplied more capital to the national market concerned, and/or powerbrokers and industrialists used the available resources more rationally, at least parts of the excess population could be absorbed on site. 27 When the ICFTU, with this faith in development, equated overpopulation with economic underdevelopment, it thus found the solution in investments in combination with a more rational exploitation of available resources. Thereby, new work opportunities could be created within the framework of the nation state, which would render migrations from Southern to Western and Northern Europe unnecessary. Accordingly, the ICFTU showed some faith in capitalism and its ability to correct structural problems, even though the capitalist system needed state or international steering to work to the working class's advantage. However, the economic gap between the richer and the poorer countries of Europe would presumably take a long time to bridge: In certain cases economic development would have to be on a very large scale and sustained at a very rapid pace over a considerable period in order [. . .] to absorb the excess of manpower by the creation of new opportunities of employment. It may therefore justly be held that a more rapid solution is needed, at the early stages of development in particular, to alleviate immediately the poverty and misery of unemployment and underemployment from which many workers suffer. This alleviation could come through emigration. 28 For that reason, the ICFTU advocated migration as a short-term solution to Europe's population and unemployment problems. To reach full employment in the whole area west of the Iron Curtain, workers needed to move from places with a relatively low level of demand for labour to areas with a perceived shortage.
The ICFTU in any case expressed the connection between migration on the one hand, and growth and development on the other, most clearly as regards potential immigration countries: There are still countries where the available manpower is not sufficient to exploit the available resources to the full. It is clearly in the interest of these countries to foster immigration of people needed to exploit the resources which would otherwise not be fully used and in this way to contribute to the raising of living standards of the people. 29 But the ICFTU did not promote completely free movement of labour. Instead, a more restrictive approach imbued the official statement, which the trade unionists authored during the third and final day of the conference. Even though the ICFTU retained the positive attitude to migration in this externally targeted announcement, the confederation simultaneously underlined that migration needed regulation and that migrants had to be steered carefully for the growth potential to bear fruit. Only through regulation could the trade-union movement gain influence over migration processes. 30 During the conference in Nervi, the participants also discussed the danger of employers who used migrations and poor living conditions in emigration areas to depress the overall development of wages and working conditions in the immigration countries -i.e. migration as a labourmarket threat. Migrants sometimes lacked knowledge about the (higher) wages in immigration countries, the author of one of the preparatory documents reflected. That could lead to their acceptance of conditions which, although better than the conditions offered at home, were worse than the common standards in the receiving area: 'This obviously endangers the standards enjoyed by the nationals.' The author, however, then emphasised that migrants and natives had a common class interest in countering such underbidding competition. Here the national and local trade unions were assigned great responsibility, first to unionise all arriving workers and then to inform them about existing standards, collective agreements and labour laws. Furthermore, the ICFTU encouraged the trade unions to create social spaces for meetings between workers from different countries, to arrange language courses and lectures about trade-union history. 31 Accordingly, this discussion pinpoints the main background to the importance of a simplified system for international transfer of trade-union membership and benefits. Although not all migrant workers had been unionised in the areas of departure, the trade unionists believed that the level of unionisation could be increased at least a bit if only all those workers who actually had a trade-union past started to pay the dues soon upon arrival.
The Nervi conference also clarified the division of work between the two different organisational structures of the 'free' international trade-union movement. The preparatory documents as well as the official statement of the conference urged the ITS to facilitate labour migrations with simplified rules about international membership transfer, for example, by ruling out new entry fees in the statutes. 32 In addition, the ICFTU encouraged the affiliates of the ITS to do much of the actual work to integrate the migrants into the trade unions concerned. National and local trade unions should include the migrants by directed efforts that to a certain degree meant preferential treatment; only then could the trade unions build and uphold a coherent front in their relationships with the employers.
Afterwards, Walter Reuther from the United Auto Workers and Walter Freitag from IG Metall, who were prominent actors within the IMF, approved the official conference statement. 33 Reuther stressed that the trade unions should see the question of migration in relation to the great internationalist undertaking, 'of rebuilding a world which has fallen apart and in which nations have become fearful of one another'; from this point of view, it was a task for the trade unions to simplify mobility over nation-state borders. Nevertheless, Reuther too agreed that great migrations were not a solution in a long-term perspective for countries with a surplus of idle workers: However, large-scale migration cannot be viewed as 'the solution' for economic underdevelopment. Surplus labor in many areas can and should be overcome by speeding up economic growth so that workers can be drawn into local industries. This, it seems to me, is the basic approach, which the ICFTU should constantly emphasize. 34 In that way, the trade-union discussions during the 1950s reproduced ethnic dividing lines between European workers. Trade unionists perceived migrations as something necessary during a transformational phase -although perhaps a long phase of development -before the economic gap between richer and poorer countries could be closed and full employment reached in all of the countries concerned. Once the hardships of unemployment were defeated, workers would not have to move; instead, they could stay put and be productive within the national community where they had been born and raised.

Picking (migrant) raisins out of the cake
When the IMF grew out of its European roots and developed into a global organisation during the 1960s, many European trade unionists obviously felt a simultaneous need to create new platforms for discussing specific European questions. In 1971, European trade unionists founded a new organisation, the European Metalworkers' Federation (see Henning, 2013). Earlier, they also organised special meetings with only European questions on the agenda.
In November 1962, the IMF convened a special conference to discuss the process of European integration. The IMF Secretariat prepared documentation, but mentioned the free movement of workers only in passing. Probably, the IMF leadership did not intend to give this issue much space. However, either it misjudged the importance of migration-related questions among the trade unionists, or it suspected that they were controversial and therefore rather kept them off the agenda. Yet participants wanted to talk about free labour mobility; again, some trade unionists managed to include migration among the discussed topics, although the issue was not on the formal IMF agenda.
While it mainly was Italians who put migration on the international trade-union agenda during the 1950s, it was trade unionists who calculated with increased immigration following the Treaties of Rome that wanted to talk about labour mobility during the 1962 conference. 35 Accordingly, when the free mobility of workers in the western part of Europe, which the Italian trade unionists had propagated, seemed to become reality, other trade unionists became more inclined to discuss migration. Hans Rasmussen pointed out that Denmark only recently had achieved full employment and that Danish trade unionists would never jeopardise this status with increased migrations; if Denmark accepted the Treaties of Rome in a near future and if that meant that 'workers from abroad streamed across our frontiers, not subject to any control', that would certainly 'provoke some misgivings and various social problems'. 36 The British trade unionists treated the free movement in Europe with respect to rights within the Commonwealth: if Great Britain were to join the European Economic Community, and the British labour market thereby opened up for Italians, there was a risk that inter-European migrants would out-compete Commonwealth migrants. Joseph O'Hagan emphasised that people belonging to the Commonwealth have had the right of entry into Britain, from the very inception of the Commonwealth. That right could be denied to them -or not actually denied, but it could become of very little value to them if there was a free-for-all entry into Britain, by other nationals than Commonwealth members. 37 Ernst Wüthrich, from Switzerland, in turn, clarified that the social problems that had been mentioned thus far were not only grounded in worry about unemployment in the event of a recession, but also in a fear of decreased trade-union power in the wake of immigration. About 770,000 foreign-born workers were employed in Switzerland, he pointed out, which corresponded to 14 per cent of the total population: The Swiss trade unions believe that this is enough, for these foreign workmates are raising almost insoluble problems for them. In Switzerland, for instance, 450,000 Italians, of whom 25% at the best are unionised, have jobs. Before there is talk of extreme mobility of labour, the trade unions should be able to solve the problems of trade union membership. For these very reasons, in the interest of trade-union stability, we oppose unrestricted labour mobility, which brings us into conflict with the corresponding terms in the EEC Treaty. 38 Probably in an implicit defence against potential accusations of xenophobia, one of Wüthrich's fellow-countrymen reminded the audience that Switzerland and Sweden -two countries expected to stay outside the EEC for reasons of neutrality -had received hundreds of thousands of migrant workers from the Common Market since the Second World War, even 'without demanding some return in trade policy, as would match the Rome Treaty'. 39 Also the Austrian delegation maintained that its country had accepted hundreds of thousands of migrants, in this case mostly Hungarian refugees, and Heinz Kienzl emphasised that this fact was important to remember because 'on several occasions [it] has been heard in the discussion that neutrality is somewhat suspicious, a pretext for picking the raisins out of the cake'. 40 Although the Italian trade unionists had been more passive in this discussion, in comparison with their active role during the 1950s, Bruno Corti, from the social-democratic Unione Italiana Lavoratori Metallurgici, quickly replied to these statements. Apparently, he was very annoyed with his colleagues whom he thought had a reprehensible attitude from a Marxist point of view. Corti could not understand how trained trade unionists could speak about labour recruitments as a favour from immigration to emigration countries. In a theoretical sermon with socialistic pathos, the Italian underscored that indeed it was the immigration countries in general and the employers in particular that gained most from the migrations; the contemporary European labour market was characterised by a great and unfulfilled demand for workers. The structural emigration pressure had been palpable during the post-war years in Italy, but Corti stressed that this was no longer the case: Brothers who are anxious about seeing their factories invaded by foreign manpower, which would lead to tradeunion competition under present conditions, should realise that from now on in Europe and also in Italy there are no longer proletarian masses disposed to go to other countries. 41 In parallel with a growing Italian economy, Italians' desire to migrate generally decreased. The migrations from Italy reached a peak in 1960, when almost 400,000 Italians left the country, but by the time of the IMF's European integration conference, the number of emigrants had already started to dwindle (Goedings, 2005, pp. 29, 33-34).
Corti admitted that the unionisation problem, which mainly the Swiss delegation underlined, was a real dilemma. However, he emphasised at the same time that both the relatively weak trade unions in Italy and the stronger trade unions in the labour-recruiting countries did too little to prepare and receive the migrants properly. Therefore, Corti rounded off his speech by emphasising that the ITS must take a greater responsibility concerning the ongoing migrations. 42

Concluding discussion
In this article I have studied which migration-related issues different national trade-union organisations put on the international trade-union agenda during the first post-war decades, as a more liberal labour-migration regime evolved in Western Europe. I have also analysed the occurrence of coinciding as well as opposing interests among and between national affiliates of international tradeunion organisations.
Italians often took the initiative. Italian trade unionists talked about both a structural emigration pressure and Italian workers' precarious situation abroad. Thereby, they used their international networks and available platforms to stress a national migration interest, and to try to defend their fellow-countrymen's working conditions abroad. Italian representatives managed to open up discussions about migration in the international trade-union arena, although such issues were not on the prepared agenda.
On the one hand, the trade unionists concerned were generally in agreement and expressed positive attitudes as regards workers' possibilities to move over nation-state borders and work in different countries, at least if the migrations were regulated and possible to control. A selfconsciousness of historical and ideological international working-class solidarity could support that starting point.
Trade unionists within both the IMF and the ICFTU connected migration with more overall questions of European reconstruction and over/underpopulation. They linked labour mobility with a wider economic context, and with a policy that aimed at a more effective use of all available means of production. In general, they showed a rather great faith in economic planning on both international and national level. If policies exhorted people to move from areas with relative population surpluses (i.e. mainly Italy) to industrially more developed areas with great demand for workers, migrants supposedly could raise their own standard of living, at the same time as they contributed to prosperity in immigration countries and also decreased the pressure on the labour market in the emigration country. The trade-union understanding of migration was, accordingly, well in line with the functionalistic theories that still dominated the academic discourse after the Second World War, which presupposed that rational choice drove migrants and that they moved from low-wage to high-wage areas. Like the trade unionists, the neo-classical theory of that time predicted that migrations would result in a new equilibrium between supply and demand for labour in national labour markets (cp. De Haas, 2014, p. 7;Castles et al., 2014, pp. 27-29).
In addition to the economic and demographic aspects, leading IMF trade unionists incorporated the issue of migration into a Cold-War narrative that claimed that social distress was fertile soil for the Communist organisations to grow. Hence, letting unemployed Italians migrate northwards had an anti-communist motive. From the IMF leadership's point of view, liberal migration policies could serve the same reformist and anti-Communist purpose as the more overall propaganda and organisational work that the ITS supported in Italy (see Svanberg, 2019).
On the other hand, trade unionists from receiving countries were careful to stress that they did not allow that migrations resulted in a surplus of labour, thereby undermining trade-union power. Migrations should be planned and directed towards certain sectors and migrants carefully selected, chiefly from the immigration country's point of view. Seemingly, the trade unionists included migration in a pattern of thought where the nation state constituted 'the natural order of things' (Malkki, 1995). In that case, their ideal was a Europe where no country was so underdeveloped that it needed emigration to reach a balance between the supply and demand for labour. The looming goal was national labour markets in full employment, and national collective agreements/labour laws, which were so similar that they decreased migration incentives. Moreover, even if it were representatives from emigration countries that usually raised migration issues in the international tradeunion arena, the following discussions showed little concern for potential problems in the areas of departure (e.g. recruitment of already employed workers). The immigration perspective dominated the discourse.
All trade unionists framed their argumentation within a discourse of problems: migrations meant problems, either for the migrants or for the trade unions. While Italian trade unionists implied xenophobia among workers in receiving countries, stressed social borders as difficult to penetrate, and pointed out employers who discriminated migrant workers, trade unionists from labour-recruiting countries talked about migrants as difficult to unionise and about employers using them to undermine the power of organised labour. However, the latter did not explicitly criticise the ethnic division of work; as long as the migrants had jobs that the natives anyhow rejected, they were less of a labour-market threat.
The restrictive line of argument was noticeable in the discussions about international membership transfers between IMF affiliates. The goal was that the trade unions should enrol all migrant workers soon after arrival. At the same time, however, the trade unionists confined the recommendations of equal treatment to migrants who had paid their trade-union dues to a metalworkers' union before departure. Therefore, the IMF could overlook many southern European migrants with a past in the agricultural sector. In that perspective, the national trade unions were interest organisations for their respective members even on the international level. In turn, the organisational borders meant that the solidarity dimension of trade-union internationalism, and its positive attitude towards liberal migration policies, was circumscribed rather narrowly to include chiefly male workers who had proved themselves loyal to trade unions even before their departure. Judging by the primary sources analysed in this article, the (male) trade unionists regarded the migrant worker as a male person as a given.
Resolutions from the international trade-union movement could be pompous, although they in practice lacked sting. The international organisations only pointed out guidelines, and formulated, for example, an official approach to the relationship between migration and development. Then it was up to the national organisations to develop real policy at the intersection between cooperation and opposition, based on their specific conditions and circumstances. Likewise, the IMF leadership concluded that it was too difficult to formulate detailed international directives to the national trade unions about how exactly they should include the migrant workers or how they should ensure equal treatment. The IMF affiliates were too integrated in national welfare systems for the organisation to develop international coordination. Instead, the IMF encouraged the affiliates to follow the general pattern in the post-war European migration regime, where bilateral government agreements to a high degree regulated the migrations, by signing their own bilateral trade-union agreements. In that way, the IMF functioned as a forum for exchange of information and experiences at the backdrop of the formation of the post-war European labour market. The trade unionists mutually normalised their practices through the discussions at international level.