Introduction

Since 1 January 2019, a Bulgarian working in Austria with a baby living in Bulgaria has been entitled to €51.30 Familienbeihilfe (family benefits), instead of €114 if the baby was residing in Austria. The reduced benefit is based on a calculation of lower living costs in Bulgaria. This so-called indexation of family benefits was at the centre of fierce controversies among and within EU member states in recent years: indexation was part of the British renegotiation deal preceding the Brexit referendum; it has divided EU member states within the Council and party groups of the European Parliament during the ongoing revision of the Social Coordination Regulation 883/2004; and it has entered national political debates in various EU member states, most prominently in Austria, against which the European Commission has initiated an infringement proceeding immediately after the entry into force of indexation. At the same time, indexation has not become a major political issue all over Europe, but mainly in a subgroup of (Western) EU member states.

In this text, we trace the political debate about the indexation of family benefits to make a more general argument about the free movement of workers in the EU. We ask why such a technical legal issue as indexation has become salient in several EU member states. As a first step towards answering this question, we discuss existing financial and political explanations of indexation, most importantly welfare chauvinism, and argue that important empirical puzzles remain. First, the financial argument is weak since indexation is unlikely to save significant amounts of national welfare expenditures as it also includes upward adjustments, and its application and enforcement create additional administrative costs. Secondly, the main political justification that indexation increases the fairness of family benefits does not stand up to close scrutiny as it rests on a very selective understanding of fairness. This selective interpretation of fairness lends support to a welfare chauvinist explanation of indexation. Yet, important empirical puzzles remain: Why does indexation target workers, i.e. welfare contributors? Why are member states willing to pay a high political price despite low financial savings? And why is the debate not just about welfare access but also about free movement more generally?

In a second step, therefore, we move beyond the welfare chauvinist explanation and argue that the debate on indexation cannot be fully understood without taking into account the broader dynamics of contestation and constitutionalization of the free movement of workers in the EU. On the one hand, free movement has become contested after Eastern enlargement with unprecedented levels of intra-EU mobility and wage competition. On the other hand, the legal framework of worker mobility and equal treatment in the EU is largely removed from political adjustments as it is constitutionalized in the Treaties and progressively interpreted by the Court of Justice (CJEU). At least symbolically, the indexation of family benefits allows EU member state governments to release this tension and serves as an “outlet” for contestation around free movement: it mainly affects EU migrants with low income, who may be perceived as a threat by domestic workers in terms of job and wage competition, and it is one of few EU legal options to qualify equal treatment without requiring Treaty amendment.

In the next section, we describe briefly how family benefits are regulated in the EU, how indexation works and why the main financial and political justifications cannot fully explain its salience. In section three, we develop our main argument on indexation as a symbol for the free movement of workers after EU Eastern enlargement. We illustrate the argument in section four with empirical evidence from the debates on indexation preceding the Brexit referendum and the Austrian reform of family benefits.

The policy issue: indexation of family benefits

We start by describing the EU legal framework for the export of family benefits (“Legal framework and financial implications” section) and the main justifications for their indexation (“Political justifications” section). From such a narrow perspective on the policy issue, the domestic salience of indexation is puzzling: it involves highly technical legal questions, its financial implications for national welfare budgets are modest, and domestic political debates about the “fairness” of indexation can only partly be understood from a welfare chauvinist perspective.

Legal framework and financial implications

In the EU, social security in cross-border family constellations is regulated by complex priority rules. The “country of employment principle” applies if only one parent is economically active. Accordingly, if the only active parent is working in state A and the child is living in state B, state A must pay the benefit. By contrast, the “country of residence principle” is decisive if the parents work in different countries. If one parent works in state A, the other one in state B and the child is living in state B, state B is responsible as the child’s country of residence. Besides this full responsibility for “primary” countries, “secondary” countries may have to grant a supplement: in case the benefits of the primary country are smaller than those of the secondary country, the latter has to compensate for the difference (Article 68 Regulation 883/2004).Footnote 1 As the CJEU clarified recently (Bogatu, C–322/17), parents may be entitled to child benefits for their children abroad even without being economically active. EU law thus clearly stipulates that member states have to export family benefits.

The indexation of family benefits concerns the question of which amount of family benefits has to be exported. Instead of paying the same amount for children living domestically and abroad, indexation means that member states adjust the benefit to the cost of living in the children’s country of residence. Accordingly, they pay a lower amount for children living in member states with lower costs of living and a higher amount for children living in member states with higher costs of living. These costs, for their part, are determined with reference to the price levels of final consumption by private households including indirect taxes of each member state.Footnote 2 For example, Austria as a primary competent state pays around €51 for children residing in Bulgaria or €151 for those residing in Denmark instead of €114 (standard benefit for children aged 0–2 living in Austria).

Indexation is unlikely to generate significant savings for EU member states. EU-wide, the current system entailed a total expenditure of around €942 million for the export of family benefits in 2013. Indexation would have led to a total expenditure of around €792 million, a decrease by 15.9% (Pacolet and De Wispelaere 2015: 31–38, data for 2013). In relative terms, this number appears even smaller. EU-wide, exported family benefits account for 3% of total expenditure on family benefits. In Austria, benefits going abroad (€273 Million) amounted to 6.2% of the total expenditure on family benefits (€4.4. billion) in 2016; the projected savings of €114 Million would thus have represented 2.5% of the total expenditure on family benefits.Footnote 3 Moreover, the Austrian government recently had to admit that actual savings in the first year of indexation only amounted to €62 rather than €114 Million—45% less than expected.Footnote 4

Upward adjustments are one reason why savings due to indexation are small in absolute and relative terms. In addition, indexation was known to entail higher administrative costs. According to the Brodolini Report Administrative Costs family benefits, which was prepared for the Commission in 2015, case workers need 2 h on average per case, whereas indexation requires an additional hour of work, i.e. an increase of workload by 50% (European Commission 2016: 27, 35). The adjustment factors need to be updated on a regular basis and, what is more, the determination and verification of the actual residence of children needs to be enhanced. Yet, the residence even of pupils is often not that obvious. Determining and controlling the residence of younger children was expected to create considerable additional burden on authorities already before the introduction of indexation.Footnote 5

To sum up, both in absolute and in relative terms, financial implications of indexation for national welfare budgets are modest. Therefore, it leads to a higher workload for administrations. The financial justification of indexation hence falls short of accounting for the remarkable political capital several governments have invested into the issue.

Political justifications

If the financial implications of indexation are modest at best, what else might account for the salience of the issue? Two of the most prominent political justifications of indexation do not provide (fully) convincing explanations either. If indexation was really about “fairness”, one would expect a broader debate about the distributive implications of family benefits. And while indexation is compatible with welfare chauvinism, important puzzles remain which call for additional explanatory effort.

To begin with, proponents of indexation often argue that the current export of family benefits without adjustment to local living costs contradicts their basic notion of fairness:

We do not consider it fair that family benefits are paid out in full in a country where the value of those benefits is worth twice or even three times as much as in the country which pays them. (letter of several EU member states to the Commissioner for Social affairs, 19 June 2018).Footnote 6

This statement certainly addresses an issue of distributive fairness, but it also raises the question why other distributive aspects of family benefits are not of significant political concern. First, as exemplified by the citation above, the public debate is exclusively framed in terms of downward adjustments. Arguably, however, paying “too little” is the bigger problem for the children affected and it should be at least equally concerning in terms of distributive fairness as paying “too much”. Proposals for indexation such as the Austrian reform mentioned above typically comprise the option of upward adjustments. Yet, these upward adjustments are only mentioned in public debates to assert compliance with EU legal requirements, but not as a matter of distributive fairness. Secondly, if indexation was really about a fair alignment of family benefits and living costs at the place of children’s residence, it would have to take into account other (in)equalities of living costs as well. Whereas living costs in border regions may converge to a greater extent than is reflected in cross-country comparisons, differences within countries are significant, most obviously between metropolitan and rural areas. And yet, these differences are hardly an issue in public debates about indexation and the fairness of family benefits. Thirdly, indexation is hardly discussed in terms of continuous inflation adjustments. If fairness was the main issue, however, it should be a matter of concern that the support for children born today is worth less than for those born in the past. Fourthly, the fairness argument is often accompanied by references to the specific needs of families. In the majority of EU member states, however, family benefits are not means-tested, i.e. they are paid independently of the income of the parents. For example, four of the five EU member states, which signed the letter to Commissioner Thyssen above, pay their family benefits without means test; from this group of countries, only Denmark reduces child benefits for families with high income.Footnote 7 Finally, it is striking that those EU member states advocating the indexation of family benefits have been among the most vocal supporters of “equal pay for equal work in the same place” during the revision of the posted workers directive.Footnote 8

The argument about fairness is weak, but such a selective interpretation of fairness may be compatible with a welfare chauvinist explanation. The label of welfare chauvinism (Hjorth 2016) is typically used for a right-wing populist agenda seeking to shield generous welfare benefits against access by immigrants (cf. Lefkofridi and Michel 2014). Moreover, welfare chauvinist arguments are shared beyond the “usual suspects”: while mainstream right parties sometimes adapt to the welfare chauvinism of populist right parties, no such effect has been found for the mainstream left (Schumacher and van Kersbergen 2016). In the past, universal family benefits attracted only limited attention by radical right parties (Ennser-Jedenastik 2018) and the picture has changed only recently. The indexation of family benefits has become a common policy demand among Western European right-wing populist parties, often voiced with welfare chauvinist undertones. For instance, the German Alternative für Deutschland demanded to introduce indexation in order to stop immigration into the German social system.Footnote 9 The Austrian Freedom Party highlighted that “Austria has to provide for its own families” and that “this social tourism has to be abolished”.Footnote 10 But indexation is not just a demand of classical welfare chauvinists. In those EU member states advocating an indexation of family benefits most strongly, this policy demand receives broad support across the political spectrum. For example, the letter to Commissioner Thyssen cited above was signed by three liberal-conservative ministers, one social liberal and one social democrat. One year earlier, a similar letter had been sent to the Commissioner by two liberal-conservatives and two social democrats.Footnote 11

Even if we interpret this development as evidence that welfare chauvinism has arrived in the middle of the political spectrum, puzzles remain. First, indexation mainly targets those EU citizens least likely to fit the stereo-typical “benefit tourist”—namely workers, who contribute to the welfare system responsible for the payment of family benefits. Hence, from a purely welfare chauvinist perspective, there should be much more pressing issues such as the exclusion of economically inactive EU citizens from non-contributory welfare benefits (Blauberger and Schmidt 2014). Second, why is a benefit which is not means-tested in a range of member states, and in the case of Austria even financed by contributions of employers, that controversial? Third, why are member states willing to pay such a high political price, i.e. taking a view which goes against the legal position of the Commission and which is perceived negatively by the governments and publics of sending countries, for such a minor problem? And fourth, why is the discussion on indexation not just about welfare access, but often linked to the free movement of workers or their access to the labour market? As will be explicated in detail in the next section, we argue that it is necessary to complement the welfare chauvinist explanation and to consider precisely the broader context of the free movement of workers, which is needed to understand the domestic salience of indexation.

The broader context: contestation and constitutionalization of free movement

The controversy about indexation of family benefits is tackling more than the technical and narrow question of indexation itself. Today, the free movement of workers is exercised under economic circumstances (“The contestation of intra-EU mobility after enlargement” section), which differ greatly from the EU’s foundational period, when the basic legal framework (“The constitutionalization of free movement law” section) was created. The domestic salience of indexation can only be understood by taking this broader EU level context into account. At least symbolically, indexation allows EU member state governments to respond to these changed economic circumstances without altering EU Treaty rules (“Indexation as a symbolic solution” section).

The contestation of intra-EU mobility after enlargement

While workers have been granted the right to move freely within the EU ever since the beginning of European integration, it was only within the past 15 years that this issue has become more contentious (Roos 2019). Albeit still being at a low level, intra-EU mobility has increased and the EU has become more heterogeneous, e.g. regarding wage levels, with several enlargements.

In particular since Eastern enlargement, intra-EU mobility has grown significantly. In 1987, around 5 million EU nationals lived in another member state and this number increased only slightly to around 6 million persons in 2004. Five years later, EU citizens living abroad in the EU totalled already almost 12 million persons and this number constantly increased since then, amounting to more than 16 million persons in 2016 (Recchi 2008: 202–204; European Commission 2011: 48f.; Eurostat 2018). Also in relative terms, intra-EU mobility has increased within the last years. Whereas 1.3% of the total EU population was mobile in 2003, the proportion amounted to 2.6% in 2012. Within EU-15 member states, EU nationals of other member states amounted to 1.6% of their citizens in 2003 and to 3.2% in 2012 (European Commission 2013: 20). Analogous to the growth of mobility in general, that of EU citizens of working age living abroad has increased from 2.5% of the total working age population in 2007 to 3.9% in 2016 (European Commission 2018; Alcidi and Gros 2019: 6).

The development within the past two decades has been driven by two events: the Eastern enlargements leading to migration flows from East to West and, to a lesser degree, the economic situation of Southern member states in the light of the Eurocrisis, inducing migration from South to North. Germany and the UK are key destination countries, hosting more than 50% of the EU citizens of working age who decided to leave their country. As concerns the countries of origin, around half of all EU citizens living in another member states are Romanians, Polish, Italians and Portuguese (European Commission 2018: 12f., 22, 31; Alcidi and Gros 2019: 3ff.). In particular, Austria has gained importance as a destination country. In 2015, it was the third highest net receiving country in the EU after Germany and the UK (European Commission 2018: 12f., 22, 31). The share of EU nationals of the overall population of working age living in Austria has grown from 3.7% in 2002 (185,000 persons) to 8.6% in 2016 (465,000 persons).Footnote 12

In terms of the employment situation of mobile EU citizens, several studies have shown that they are in general more likely to be economically active than nationals (for the UK Blanchflower and Lawton 2009; European Commission 2013: 18f.; European Commission 2018: 14). For instance in 2016, 83% of recent mobile EU citizens were in employment, compared to nationals with a share of 78%. Mobile EU citizens from CEE countries are disproportionately concentrated in elementary occupations with relatively low wage levels, which include, inter alia, cleaning, agricultural work or construction (European Commission 2018: 64–76). Whereas professionals with high income are very mobile throughout the Union, CEE nationals who leave their country also occupy a significant share of low and medium wage positions (Alcidi and Gros 2019: 10–12).

Closely related to greater intra-EU mobility is also the increased economic heterogeneity among EU member states. While the EU had become more heterogeneous already in the 1980s with the accession of Denmark, Ireland and the UK, e.g. regarding welfare regimes (Scharpf 2002: in particular 647), the differences between member states increased strongly with the Eastern enlargements in the 2000s. The welfare levels and institutions have become more diverse with different production and welfare regimes. Most importantly, labour costs vary to a high degree (Höpner and Schäfer 2012a, b: in particular 2, 3, 10f.). For instance in 2016, average hourly labour costs, i.e. wages and salaries and non-wage costs such as employers’ social contributions, ranged between €41.3 in Denmark with the highest level and €4.5 in Bulgaria with the lowest level. On average, hourly labour costs amounted to €30.7 in EU 15, to €11.9 in EU 10, to €4.9 in EU 2 and to €9.5 in Croatia.Footnote 13 The differences in wage, first, constitute a push and pull factor for migration (Alcidi and Gros 2019: 14), and, second, raise concerns about wage competition and labour standards, particular in Western European member states (Afonso 2012: 706).

With the Eastern enlargements, critical voices towards the free movement of persons became louder in Western EU member states (Schmidt Forthcoming). Studies show that the free movement of persons had a negative effect on public opinion towards the EU at the regional levels of Spain, France, the Netherlands and Ireland (Toshkov and Kortenska 2015). Concerns about immigration were a major driving factor for those UK citizens who voted to leave the EU in the Brexit referendum, who were typically less educated and part of the working class (Hobolt 2016). As Vasilopoulou and Talving find, the attitude of EU citizens in wealthy member states depends on their individual employment and economic situation. This confirms Hobolt’s finding for the UK that less educated persons in Western member states tend to be more sceptical towards the free movement of workers than higher educated persons. The former may perceive EU migrants rather as a threat to their job (Vasilopoulou and Talving 2018).

In short, the issue of free movement of workers has become more contentious within the past 15 years as EU 15 member states have experienced higher inflows of mobile EU workers. In addition, member states have become more heterogeneous, which, as Höpner and Schäfer already held in 2012, “fuels political conflict in the EU” (Höpner and Schäfer 2012a: 437).

The constitutionalization of free movement law

In stark contrast to the changed socioeconomic circumstances of today’s EU-28, the basic framework of EU free movement law still dates back to the founding Treaties of the EU-6. The free movement of workers was already included in the Treaty of Rome in 1957 and was explicitly linked to non-discrimination from the outset. By interpreting Article 45 TFEU (ex-Article 38, before 48) expansively, the CJEU has continuously broadened member states’ obligation to treat workers equally—often against explicit resistance by member state governments (Larsson and Naurin 2016: 396) and with little deference to secondary law (Martinsen and Falkner 2011). The term “worker” demonstrates this particularly well. Beginning with the Hoekstra decision in 1964 (Case 75/63), the Court has constantly interpreted the term worker—and, hence, also the rights associated with this particular status—broadly. According to the Court’s formula established in the Levin judgment in 1982 (C-53/81), to qualify as a worker, one’s employment has to be “genuine and effective”, but not “purely marginal and ancillary”. In subsequent judgements, the Court precluded any fixed threshold in terms of contract duration, working hours or minimum income used by member state administrations to establish the worker status in practice.

Whereas the free movement and equal treatment of workers have been enshrined in the Treaties from the beginnings of European integration, the provision of social security has remained national competence. European law, therefore, does not seek to harmonize, but only to coordinate national welfare systems in order to facilitate free movement. Accordingly, the Treaty provision on free movement was accompanied from the beginning of integration by secondary legislation coordinating the access to social protection of migrant workers. The early Regulations No. 3 and 4 of 1958 have evolved through multiple reforms into the current Regulations 883/2004 on the coordination of social security systems and 492/2011 on the free movement of workers.

Taken together, European rules on free movement of workers and social security coordination are characteristic for what Dieter Grimm has labelled the (over-)constitutionalization of European law (Grimm 2015). By establishing supremacy and direct effect, the CJEU has effectively constitutionalized European law (Weiler 1991), albeit with an important difference compared to national constitutions: the EU’s Treaties, i.e. its “constitution”, are not limited to defining the basic institutional framework, but they are very detailed on substantial policy, which is regulated by ordinary legislation at the national level. Even though social security coordination is subject to a fairly complicated regime of EU secondary legislation, its basic parameters such as the worker definition are directly derived from the Treaties.

The main beneficiaries [of constitutionalization, the authors] were the four economic freedoms …. These freedoms were transformed from objective principles for legislation into subjective rights of the market participants who could claim them against the Member States before the national courts. Their implementation thus became a matter of jurisdiction rather than legislation (Grimm 2015: 467).

The implications of constitutionalization become even clearer when contrasting the rights of mobile workers with EU citizens more generally. European citizenship was established in the Treaty of Maastricht of 1992 and extended free movement to all EU citizens, but it left much greater discretion to member state governments. In contrast to the free movement and equal treatment of workers, citizenship rights are “subject to the limitations and conditions laid down in the Treaties and by the measures adopted to give them effect” according to Article 21 TFEU. Whereas the CJEU has initially interpreted EU citizenship expansively, partly opening welfare access also to economically inactive EU migrants, it has confirmed significant member state “limitations and conditions” in its more recent case law (Blauberger et al. 2018). As one consequence, EU member states can still effectively exclude economically inactive EU citizens from social benefits (Kramer et al. 2018; Roos 2019; Martinsen et al. 2019). At the same time, this development has made the recognition as a worker even more crucial for claiming social rights in the EU, and workers’ rights now come to the fore.

Indexation as a symbolic solution

But what are the options for EU policy-makers if free movement is increasingly contested in the enlarged EU while being solidly enshrined in the Treaties? The last great EU Treaty revision was agreed in 2007, unanimity is still the decision-rule and free movement of workers is a core principle of European integration. Significant changes to the free movement provisions in EU Treaty law are not a realistic political option. And even without the unanimity requirement, changing EU secondary legislation poses high decision-making thresholds and often takes very long. As a consequence, member state governments may unilaterally explore legal grey areas and test the limits of what can be considered compatible with EU law (Blauberger 2012). The indexation of family benefits falls in such a legal grey area. Moreover, indexation is attractive also for mainstream parties, as it is unlikely to hinder free movement at large, but affects mainly low wage migrants.

Legally, indexation allows for a reinterpretation of the free movement of workers at the EU level or, as some argue, even at the member state level. At the European level, advocates of indexation see their position confirmed by the British renegotiation deal preceding the Brexit referendum (see “The British renegotiation deal” section): if it was possible to allow indexation in order to keep the UK inside the Union, it is just a matter of political will to allow indexation for other member states as well. At the domestic level, several voices were raised that indexation could even be introduced without changing EU law at all (see Austria’s reform, “The Austrian reform of family benefits” section): article 67 of Regulation 883/2004 may be interpreted as only demanding an equal treatment requirement, according to which a member state A would have to apply the same calculation method and treat children of nationals who live in another member state B like nationals of B (Thym 2018).

Economically, indexation is not only attractive for populist right, welfare chauvinist parties, but also for mainstream parties as a policy option. Scholars already demonstrated that mainstream right parties can be influenced by the discourse of the populist right and accommodate to their welfare chauvinist rhetoric at the same election (Schumacher and van Kersbergen 2016). And also social democrats, which are typically (blue-collar) workers’ parties, may have an incentive to position themselves in favour of indexation. While indexation formally concerns all workers with children abroad, they are de facto not equally affected. Workers who earn high wages do not depend on family benefits to make their living. In contrast, for persons in the low wage sector, family benefits are an important supplement as they make up a significant amount of the persons’ finances in relation to their wage: generous family benefits top up the income of such “working poor”. If such persons, who are already vulnerable due to their low wage, have children residing in member states with lower living conditions, they will hence feel the effects of an introduction of indexation the most; and the badly paid jobs in countries with generous benefits may consequently become less attractive. As was discussed above, migrants from CEE countries are overrepresented in the low wage sector. Domestic (blue-collar) workers for their part may perceive low wage migrants as a threat as they may have to compete with them for jobs (see “The contestation of intra-EU mobility after enlargement” section, cf. the findings of Hobolt 2016; Vasilopoulou and Talving 2018).Footnote 14 Social Democratic Parties may thus also favour indexation in order to address their important clientele, all the more if they are in a member state with a strong populist right party advocating the issue.

At least symbolically, the indexation of family benefits hence promises to address the economic and legal challenges presented in the preceding sections: it mainly affects EU migrants with low income, with whom nationals may fear to compete for jobs; and it is one of few EU legal areas to qualify equal treatment without requiring Treaty amendment. Indexation may hence serve as “outlet” for the contestation of free movement of workers and equal treatment.

Empirical evidence

In the remainder of this text, we illustrate our argument with the analysis of two cases in which the indexation of family benefits has been particularly salient. Indexation was one major element of the British renegotiation deal (“The British renegotiation deal” section), and it has influenced subsequent debates about indexation at the national and European levels. Frustrated by the exclusion of indexation from the revision of the EU social coordination regulation, Austria finally opted for unilateral action (“The Austrian reform of family benefits” section).

The British renegotiation deal

On 19 February 2016, the European Council concluded on a set of arrangements, which were supposed to facilitate the continued EU membership of the UK. Annex 5 of the conclusions contained a declaration of the European Commission on the indexation of child benefits.Footnote 15 As is well known, the European Council’s decision became invalid only 4 months later with the Brexit referendum, but the Commission’s declaration has resurfaced in all subsequent debates on indexation and its context of origin is exemplary for our general argument. Essentially, the British renegotiation deal was about limiting the free movement of workers while avoiding Treaty changes wherever possible.

Already in late 2013, then Prime Minister David Cameron set the tone for his re-election campaign, which eventually led to the renegotiation deal and the Brexit referendum. In an opinion piece for the Financial Times, he famously demanded that “Free movement within Europe needs to be less free”.Footnote 16 He justified this claim with the increased heterogeneity of the EU post-enlargement and called it a “monumental mistake” that the UK had opened its labour market without transitional arrangements in 2004. Moreover, Cameron explicitly linked the call for restricting the free movement of workers to ending the export of child benefits. After his re-election in 2015, Cameron specified the British negotiation position for a new settlement within the EU in a letter to Donald Tusk, then president of the European Council. One of four main issues in this letter concerned intra-EU migration and demanded, for example, measures against abuses of free movement and against expansive CJEU jurisprudence in this area. But the demands went beyond abuses of free movement and included the goal of an overall reduction of “the numbers coming here”.

As I have said previously, we can reduce the flow of people coming from within the EU by reducing the draw that our welfare system can exert across Europe. So we have proposed that … we should end the practice of sending child benefit overseas.Footnote 17

Eventually, the European Council’s conclusions did not propose to stop the export of child benefits, but only their indexation. And while indexation was not framed as a measure aimed at limiting free movement by the European Council or the Commission, it has to be understood precisely in this context. The indexation of child benefits was a compromise in response to the UK’s further reaching demand, and it was included in the renegotiation deal alongside more obvious restrictions such as the so-called “emergency brake” or “safeguard mechanism” in cases of exceptional inflow of workers to the UK.Footnote 18

Furthermore, in line with our argument, the determination of the British negotiating position as well as the actual deal was heavily shaped by the question of what could be agreed without changing the EU Treaties. Right after David Cameron’s electoral success in May 2015 and before the adoption of an official government position, the British think thank “Open Europe” evaluated the feasibility of potential renegotiation demands in terms of their compatibility with EU Treaty law.Footnote 19 Subsumed under the label “free movement reform”, restricting the export of child benefits was the only item which received the highest ranking (i.e. “treaty change is definitely not required”), whereas tighter restrictions on free movement were ranked 2.5 (i.e. “treaty change may be required”) or 0 (i.e. “Treaty change is required”). Even earlier, legal analysts such as Steve Peers had pointed at foreseeable problems once domestic electoral promises would have to be translated into EU legal changes.Footnote 20 Unsurprisingly from this perspective, the renegotiation deal concluded by the European Council did not contain any of the further reaching requests from Cameron’s re-election campaign. By contrast, the renegotiation deal started with an emphasis “that the following set of arrangements [is] fully compatible with the Treaties” and only referred to rather minor Treaty changes in other areas.Footnote 21

In sum, indexation was offered to the UK in response to calls aiming not just at cutting benefits for EU citizens, but at curbing the free movement of workers in the first place. And even though the British renegotiation deal was short-lived due to the Brexit referendum, it set the frame for continued debates about the indexation as a means to limit free movement without changing the EU Treaties. When the Commission published its legislative proposal for the revision of the EU Social Coordination Regulation 883/2004 in December 2016,Footnote 22 the deal was already invalidated by the Brexit referendum and no option for indexation was included. A group of Western member states continued to set the issue on the agenda in the Council and demanded a reform of secondary legislation, albeit without success: in a letter to the Council Presidency and to Commissioner Thyssen in July 2017, the competent ministers from Austria, Denmark, Germany and Ireland called for reconsidering the option of indexation during the legislative process.Footnote 23 The Netherlands joined in a repeated call in June 2018.Footnote 24 In both letters, the ministers reaffirmed their commitment to the freedom of movement in the EU, but they highlighted the need to “adjust rights” in light of “changing circumstances”. Thus, despite their explicit commitment to free movement, they challenged equal treatment in the enlarged single market—which is precisely why another group of governments, in particular those from Central and Eastern Europe, opposed and blocked indexation as unfair and incompatible with free movement in the EU.Footnote 25

The Austrian reform of family benefits

In Austria, the issue remained salient and indexation was introduced as of January 2019. Since then, family allowance (Familienbeihilfe) and tax credits for children (Kinderabsetzbetrag) have become subject to indexation whenever Austria is the primary competent state as well as differential payments (Differenzzahlung) in case of secondary competence. The Austrian case lends further support to our argument that with indexation member state governments test the limits of EU free movement rules more generally. Even though early calls for indexation originated from the right-wing welfare chauvinist spectrum, mainstream conservative and later even social-democrat politicians invested considerable political capital into the issue. While the British renegotiation deal continued to serve as an important argument for the compatibility of indexation with EU law, the Austrian government deliberately took the risk and introduced indexation unilaterally.

Austrian right-wing populists had called for indexation already for years, but the issue only became salient for the governing centre-right and centre-left parties in the context of the Brexit referendum in 2016 and in the run-up to Austrian national elections in 2017. When Reinhold Lopatka from the conservative Austrian People’s Party had forwarded the idea of indexation in May 2010, he was thwarted by his own party (John 2010). It was only in 2015 when Sebastian Kurz, at that time Minister for Europe, Integration and External Relations, re-started the debate in Austria by welcoming the British proposal for an indexation of family benefits.Footnote 26 Due to Kurz’s initiative, the issue of indexation also received increasing media attention since 2015 (XX 2018 [name eliminated for the purpose of anonymization]). Moreover, after initially opposing the idea, even the Social Democratic Party under then-Chancellor Christian Kern sided with Kurz’s calls for indexation in 2016.Footnote 27 As a consequence, the domestic debate about indexation was highly salient in the run-up to the 2017 national elections, but hardly polarized: indexation was not only advocated by the Freedom Party, but also by both governing, mainstream parties at the time.

When Austria introduced indexation unilaterally in 2018 under the chancellorship of Sebastian Kurz, the reform was mainly justified with reference to the rising amounts of money transferred for children abroad and potential cases of abuse.Footnote 28 Yet, these justifications have to be analysed in their broader political and economic context. In debates about indexation of child benefits during his electoral campaign, Kurz had continuously mentioned the risks of free movement and equal treatment of EU workers especially for Western European countries such as Austria. He highlighted that “free movement of workers does not mean that one can choose the best welfare system” and that the current system of export would lead to “massive distortion effects” in the receiving countries. Furthermore, he emphasized that “we have such a high immigration to Austria that we would need an economic growth of at least 3% in order to avoid rising unemployment”. Although Kurz tried to backpedal and insisted that free movement was an “important asset”,Footnote 29 his argument suggested that not only benefits for EU foreigners but also free movement at first place was contested and that indexation was seen as a means to curb it. The Social Democratic Party for its part called for a limited access of EU citizens to the Austrian labour market in its manifesto in 2017, also illustrating that free movement was contested within this centre-left party (SPÖ 2017: 29).

As regards the legal justification of indexation, the British renegotiation deal continued to play an important role far beyond the Brexit decision. Already briefly after the referendum, Kurz argued that he had always considered this deal as a package which would enter into force, independent from the UK’s decision to remain or to leave the EU.Footnote 30 In November 2016, together with two party colleagues, Sophie Karmasin (Minister for Families) and Hans Jörg Schelling (Minister of Finance), Kurz further pushed for a reform of Regulation 883/2004 at the EU level in a letter to the European Commission.Footnote 31 Given that no agreement on indexation could be reached at the EU level, however, Austrian politicians finally pushed for unilateral action.Footnote 32 The Federal Ministry of Finance asked the legal scholar Wolfgang Mazal to write a legal opinion on the indexation of child benefits and the possibility to introduce it at the national level. This legal opinion argued in favour of such a possibility highlighting the function of the Austrian benefit: to partially discharge the burden which results from the maintenance obligation, i.e. to reimburse a part of the expenses of the “basket of basic needs”. It had to be kept in mind that this basket varied from country to country and that one therefore had to consider the average living situation of the person who assumed the maintenance obligation and the purchasing power of the child’s country of residence when calculating the adequate amount of benefits. Otherwise, the system would constitute excessive support in countries with low purchasing power on the one hand, which would go beyond the demands of the four freedoms, and insufficient support in countries with high purchasing power on the other hand, which would hinder free movement. In order to comply with the principle of equal treatment, Austrian family benefits may thus be adjusted downwards as well as upwards. Importantly, this not only concerns other EU nationals working in Austria whose children reside abroad but also Austrian nationals themselves being employed in Austria but living with their children, e.g. in a border region.Footnote 33 In short, the current system would not correspond to the wording (“as if” the children were residing in the competent state) and aim of Article 67 Regulation 883/2004. As a consequence, Mazal reasoned Austria could—also unilaterally—introduce an indexation (Mazal 2017: 3f.).

Other legal scholars argued against this opinion (Marhold 2017; Leidenmühler 2018) and the Commission made clear that it would consider an infringement procedure if indexation was introduced.Footnote 34 Nevertheless, the Austrian government further pursued its plan and adopted a reform of relevant national legislation coming into force at the beginning of 2019. In 2020, the CJEU received the case even twice: a preliminary ruling procedure was introduced by the Austrian Fiscal CourtFootnote 35 and the CommissionFootnote 36 accuses Austria of an infringement of EU law, considering the indexation of family benefits as indirect discrimination. In sum, the Austrian government was willing to pay a considerable political price by introducing indexation as a rare occasion for testing the limits of EU free movement and equal treatment of workers.

Conclusion

Despite its technical legal character, the issue of indexation of child benefits has become salient in recent years. We argued that this development has to be understood in a broader context of economic and legal challenges of free movement and equal treatment of workers after EU Eastern enlargement. Indexation is paradigmatic for this broader context. At least symbolically, indexation of family benefits promises to tackle economic and legal challenges of the free movement of workers. While most aspects of free movement of workers are basically exempted from change because they are based upon the Treaties and their interpretation by the CJEU, indexation is a legal (grey) area which can be modified via legislation—and EU member state governments are likely to explore and to politicize this area. Moreover, it mostly affects EU citizens in the low wage sector, for whom generous family benefits constitute a wage subsidy and who may be perceived as a threat in terms of job competition.

We illustrated our argument with empirical evidence from the debates on indexation preceding the Brexit referendum, which has influenced all subsequent debates about indexation as it opened the window to discuss the issue, and the Austrian reform of family benefits. This evidence supports our argument that the debate, across cases, concerned the free movement of workers in general: the aim of indexation was not only to limit access to social benefits but to qualify free movement of (low wage) workers. Strikingly, such demands were raised not only by populist right parties, but also by mainstream parties. This suggests that welfare chauvinist and protectionist attitudes may have arrived in the centre of the political spectrum. While earlier research has demonstrated that the rights of economically inactive EU citizens and the definition of worker has increasingly been contested within the last few years (Heindlmaier and Blauberger 2017; Roos 2019; Kramer et al. 2018; Martinsen et al. 2019), we argue that member states even test the limits of the law when it comes to the rights of workers, meaning that even the EU’s core element of free movement of workers can no longer be taken for granted.