With the entry of Greece (1981) and of Portugal and Spain (1986), the southern European welfare state made its debut in the political arena of the European Union, although (southern) Italy already represented this group to some extent in the original six Member States. Divergence was facilitated partly because the Member States were able to implement guidelines in very different ways and because the EC could exert little control over whether directives were upheld. In the more developed economies of north-western Europe this led not only to fear of possible objections to decisions regarding new initiatives, but also to a fear of social dumping through concession wage bargaining and cross-border tax competition. The perception was that unequal levels of social protection and the very different traditions among the Member States could make social rights detrimental to competitiveness. Countries with higher standards of social protection would be at a great financial disadvantage compared to those with lower standards. Generous welfare states unable to respond adequately to the extremes of policy and tax competition would have to pay a high price in terms of economic stagnation and unemployment. The conjecture of social dumping is reminiscent of the double bind of social Europe, in the sense that market expansion unleashed competition between different countries with different levels and structures of social protection. The expected social dumping never really materialized. In the first place, social costs turned out to be only one of many factors considered by enterprises making investment decisions, in addition to productivity, education and training, innovative potential, infrastructure, business climate, and labour relations stability. Secondly, the proclivity of southern European Member States to engage in a 'race to the bottom' was mitigated by EU cohesion policy as a kind of quid pro quo. In the Maastricht Treaty of 1992 a doubling of resources for the Structural Funds was agreed to, whereby they became a prominent instrument for regional economic development and social cohesion in a more competitive, enlarged EU market (Marks, Hooghe, and Blank, 1996: 354). The relatively easy expansion of the cohesion policy, compared with the small steps made in the areas of labour market regulation, labour relations, and social security, can be partly explained by the fact that the cohesion fund, although costly for those making net payments, did not threaten the policy autonomy of the generous welfare states.
In hindsight, it is interesting to observe how southern enlargement was double-edged. The southern socialist parties strengthened the Delors camp, but at the same time this greatly increased diversity in welfare institutions and interests. Hence, the compromise to add a social dimension through cohesion policy. Cohesion policy also usefully enlarged the coalition beyond socialist or centre-left parties to include conservatives in recipient countries. The principle of partnership in cohesion policy, in addition, opened up national models to enable territorial subnational governments and later also social partners to engage directly with the European Commission. It is interesting that social democrats in the donor north remained generally supportive of the fourfold increase (and then doubling) of the cohesion funds until the late 1990s to early 2000s. One reason was that the dispersal of funds to needy deindustrialized areas, also included economic regions in northern countries, particularly Britain. Notwithstanding, the expansion of the social funds surely encouraged southern countries to catch up on welfare state modernization—albeit with strong insider/outsider dynamics—which also muted potential social dumping (see Hooghe and Marks, 1999, 2009; Barnard, 2000; Ferrera, 2005a).
Throughout the 1980s, the ECJ began to challenge nationally defined social assistance benefits in a number of important cases (Piscitello 1983; Giletti et al. 1987: see Ferrera, 2005a: 134-5). In spite of the fierce opposition of Member States, minimum benefits became exportable from the country of payment to the country of (new) residence. In 1992, agreement was reached over a rather weak commitment on the part of the Member States to seek convergence on social protection objectives (Recommendation 92/442/EEC), in an attempt to bolster the legitimacy of social policy ambitions: from the goal of harmonization to that of convergence. Proposals for solutions to the problem of coordin- ation—the 'thirteenth' or 'sixteenth state' and the European social snake—on the other hand, were never enacted due to lack of political support. In the same year, Regulation no. 1247 was adopted, which inserted a specific coordination mechanism for non-contributive 'mixed' transfer benefits into Regulation no. 1408/71. The two main novelties were: (a) the principle that such benefits, though regarded as social security benefits, shall be granted exclusively in the territory of the Member State in which the beneficiary resides; and (b) the inclusion of a positive list (amendable) of benefits for each country as a prerequisite for imposing residence requirements. On the portability of social assistance, under the revised umbrella of the 1992 Regulation, means-tested benefits can be consumed only within national boundaries, however, under certain residence and income conditions.