Regulating the Poor through Internal Borders: The EU in Historical and International Perspectives

Tensions surrounding internal migrants’ access to welfare and the associated politicisations about who should shoulder the ‘fiscal burden’ are not unique to the European Union (EU). Based on a Most Different Systems Design and following an institutionalist approach, this article analyses the developments associated with freedom of movement and access to poor relief/social assistance in four economically and politically diverse jurisdictions. It also considers the implications of these developments for the EU. The four cases analysed are industrialising England, contemporary China, Germany, and the United States. Although economic integration was a necessary, it was not a sufficient condition for the abolishment of residence requirements for internal migrants in all four jurisdictions. Moreover, it took political power, various coalitions, or the leadership of actors to overcome the barriers and hurdles on the path to social citizenship in the wider territorial jurisdictions. Solidarity as a precondition did not play a significant role.


Introduction
Increased migration and mobility within the European Union (EU) have led to a politicisation of freedom of movement and highly contested debates about 'unreasonable burdens' caused by 'welfare tourism' (Roos, 2019;Verschueren, 2014). EU citizens have the right to freedom of movement and to take up residence in any Member State; but this right to residence does not automatically entail a right to access welfare. According to the EU Citizenship Directive (2004/38), only those mobile EU citizens who have worker status are entitled to all social benefits and services in a Member State of destination from day one of their residence. Those with sufficient means and health insurance have the right to residence but have to prove a minimum five-year duration of residence before they have the same social rights as nationals. During the past decade, various rulings by the Court of Justice of the European Union (CJEU) have confirmed the possibility of excluding 'economically inactive' EU migrant citizens 1 from access to social assistance by Member States during their first five years of residence (Jacqueson, 2018;Verschueren, 2017). This policy leads to an exclusion of a significant number of EU migrant citizens from accessing social protection minima and exposes them to labour market exploitation in destination countries. Once needy EU migrant citizens apply for social assistance, a number of destination Member States revoke their residence status, request them to leave the country, and/or 'support' them to travel 'home' 2 (cf. Kramer, 2020;Lafleur and Elsa, 2018;Schrauwen, 2021;Valcke, 2020).
During the past years, a rich literature on tensions between freedom of movement and access to social benefits has emerged. For instance, political scientists have focused on normative questions of the duties and conditions required of citizens (cf. Ferrera, 2019), the judicialisation of citizenship (Schmidt, 2019), and the politicisation of intra-EU migration (Roos, 2019). Economists and political scientists have analysed the cost-benefit relationship of internal EU migration (e.g. Dustmann and Frattini, 2014;Martinsen and Werner, 2019). Sociologists and social policy analysts have focused on the relevance of institutional welfare arrangements (Ruhs and Palme, 2018) and the differences between Member States of origin and destination with respect to social rights . They have also analysed the role of local and street-level bureaucrats Ratzmann, 2021) and the lived experiences of EU migrant citizens themselves (Ehata and Seeleib-Kaiser, 2017). Meanwhile, socio-legal scholars have highlighted the importance of 'residence' and its meaning thereof as a precondition for benefits (Stendahl, 2016). 3 To be able to systematically assess the current debates, we must ask whether and to what extent the tensions associated with internal migration and accessing welfare within the EU are sui generis or a new phenomenon from historical and international-comparative perspectives. 4 Hence, I have used a diachronic and international comparative approach that follows a Most Different Systems Design. Moreover, my analysis builds on comparative historical institutionalism (Thelen, 1999). Primarily based on secondary literature, I compare the institutional arrangements restricting poor internal migrants' access to social protection minima and analyse the conditions and drivers that have contributed to abolishing these restrictions in four jurisdictions: industrialising England, current-day China, and the two federal states of Germany and the United States. Despite stark differences in the political and economic systems of these four cases, all have historically restricted 1. I define EU migrant citizens as EU citizens who are, or intend to become, habitually resident in a Member State of which they do not hold nationality. 2. E.g., the German government 'supports' poor EU migrant citizens with a loan to travel 'home' (cf. § 23, Para. 3a of the Social Code Book XII). 3. Cf. the 2016 Special Issue of the European Journal of Social Security on the issue of residence. 4. Although internal migration and resulting tensions around accessing social rights are neither historically nor internationally unique, very few researchers have placed the current EU debate into a broader historical and/or international comparison (but see Bruzelius and Seeleib-Kaiser, 2021;Kovacheva et al., 2012;Maas, 2021;Prak et al., 2018;Schönberger, 2005;van der Mei, 2003 access to minimum social protection at the local and/or regional levels to those persons who hold settled status, normally based on birth right. Whilst England was the first country to industrialisetriggering great internal migration (Pooley and Turnball, 1998) and subsequent tensions between jurisdictions with the duty to support the poor (Lees, 2006;Snell, 2006)current-day People's Republic of China (PRC) has had to deal with similar problems of reconciling the internal migration of workers and the provision of welfare for the poor at the local level since the 1980s (Zhang, 2018: 3). In addition to tensions between local jurisdictions, federations such as Germany and the United States of America have been confronted with tensions between interstate freedom of movement and the provision of welfare at the state levela challenge that is also currently faced by the EU (cf. Bruzelius and Seeleib-Kaiser, 2021). This article will demonstrate that economic and fiscal concerns and considerations have initially been important factors determining the restrictive access to social minima for internal migrants within territorial jurisdictions, very similar to concerns voiced within the EU. The comparative approach will expound that, irrespective of the economic and political system, economic integration has been a necessary condition to overcome restrictions for internal migrants in accessing social assistance. However, economic integration has not been a sufficient condition, but has needed to be augmented by political or judicial action. Solidarity or broad public support, often mentioned as a precondition for providing unconditional access to social assistance for internal migrants in the EU, was not an important factor in any of the cases analysed. It will be shown that specific actors and political coalitions, which were contingent on the respective political institutions, as well as measures taken to overcome institutional barriers or veto points were of key importance to the easing and the eventual abolishment of conditional residence requirements for internal migrants to access minimum social protection.

Territoriality and Social Citizenship
Historically, it has not been the nation state, itself largely a creation of the 18 th and 19 th centuries, that defined the provision of welfare; moreover, local parishes and municipalities (as well as the landed gentry) were responsible for defining and providing welfare to 'their' needy residents in much of Europe and North America. Initially, citizen meant the member of a city. Prak et al. (2018) highlight that, throughout modern history, (internal) migration has regularly challenged the definition of community membership as well as the boundaries of citizenship. Increasing mobility and migration thus led to tensions between territorial jurisdictions, as the domains of economic activity and residence of an individual no longer automatically coincided with the duty of a territory to support individuals in need. To ensure that newcomers would not automatically and unconditionally enjoy membership or citizenship duty-free, territorial jurisdictions have routinely used various means to restrict access to welfare or make it conditional. Hence, membership and citizenship often needed to be earned through contributions, or were related to occupational status or wealth. 5 Tensions and frictions between local jurisdictions regarding responsibility for the mobile poor are said to have been overcome in many affluent European countries with the establishment of the nation-state, which over time developed into a welfare state based on 'national closure' (Ferrera, 2005): that is, 'internally inclusive' and 'externally exclusive' (Brubaker, 1992: 72). The modern nation-state and its responsibility for welfare in many countries of Europe is largely 5. Similar to the conditions of earned social citizenship within the EU (Kramer, 2020). a development of the post-World War II era. For many observers, the nation-state continues to be the ideal locus or, in some respect, the endpoint in the development of social rights within a European Social Union, conceptualised as a holding environment for national welfare states (Vandenbroucke et al., 2017).
Within the EU, only workers or persons with 'sufficient resources and health insurance' enjoy the full benefits of freedom of movement and non-discrimination in Member States of destination based on EU citizenship within the initial period of residence. Member States can exclude 'economically inactive' EU citizens from access to social assistance for up to five years . As a consequence, EU migrant citizens without employment (and without worker status) have no recourse to social support and constitute a large proportion of the homeless population in many Member States of destination (Striano, n.d.). At the political level, we have been witness to a politicisation of freedom of movement. For example, concepts such as 'welfare magnetism' and 'benefit tourism' have been instrumentalised in the debate leading up to the Brexit referendum. Destination Member States have become anxious that immediate benefit access by newcomers might lead to an 'unreasonable burden' and have subsequently (further) restricted access to social protection minima (Roos, 2019).
The restricted approach towards granting access to social protection minima for EU migrant citizens is often justified by the concept of bounded solidarity within a nation-state, where rights and duties are clearly defined, and the principle of reciprocity (Bellamy, 2019;Offe, 2000;Sangiovanni, 2013). Bellamy andLacey (2018: 1410) argue: Although there is evidence of an emergent European identity, it is doubtful that this would one day replace national identity, while any such post-national identity may in and of itself be undesirable. … it is unlikely that the conditions that brought about the nation-state and established bonds of solidarity among citizens could be replicated at the supranational level.
However, as Börner (2015) highlights, solidarity should not be understood as a static concept and a precondition for social policy beyond the nation-state, but should instead be approached from a dynamic perspective. Moreover, new institutional arrangements and territorial configurations may shape new solidarities. Furthermore, observers often seem to overlook the complexity of citizenship and the internal contestations between domestic actors; jurisdictions are frequently characterised by multiple levels of citizenship at the local, regional, national, and supra-national levels (Bauböck, 2014). According to Maas (2013: 2), 'Unitary citizenship is the historical exception; more common are varieties of multilevel citizenship. ' Faist (2001) has tried to capture this complexity with the concept of 'nested citizenship'.
But: what does it mean to be a member or citizen of a community? Arendt (1951) argued that at the core of citizenship was the 'right to have rights'. Building on his triad of civil, political, and social rights, Marshall (1950) argued that social rights are crucial for the ability of all citizens, irrespective of class, to fully participate in the community and to enjoy their political and civil rights. For Dahrendorf (1985: 94), social rights are: 'the final stone in the arch which holds up the roof of citizenship.' Substantively, social citizenship can encompass a range of meanings, 'from the right to a modicum of economic welfare and security to the right to live the life of a civilised being according to the standards prevailing in the society' (Marshall, 1963: 74). In the social policy literature, the concept of social citizenship was often not well-elaborated and used in very different ways (for an excellent critical discussion, see Powell, 2002). In practice, social citizenship has developed quite differently across time and countries, with very dissimilar meanings (cf. Alber and Gilbert, 2009; Kaelble and Schmid, 2004).
In this article, I will conceptualise social citizenship as encompassing both the duty to provide members of a community with a minimum provision of subsistence, either in the form of poor relief or social assistance, as well as the right of those members of the community to access such minimum provision. A social minimum is fundamental because it constitutes a recognition of human dignity and is designed as a floor under which nobody should fall as a member of a community; thus, it constitutes the foundation of (a basic concept of social) citizenship or membership in a community (Bahle et al., 2011;Leisering, 2018: 88-93;Seeleib-Kaiser, 1995).

Internal Migration and Regulating the Poor in Industrialising England and Current-Day China
Rapid economic development has provided cities in industrialising England and current-day China with much-needed migrant workers from rural areas. Nevertheless, cities preserved a hierarchy of rights by excluding the poor or even migrant workers and their dependents from social support. When migrants became dependent on relief, cities returned them to their municipality of origin, which was defined to be their 'home'. However, both jurisdictions eased and eventually abolished these restrictions.
In industrialising England, the local parish was the key unit for administration in general, and regulating poor relief and the poor in particular. Based on the Poor Law, the parish of 'settlement' had the duty to provide relief. Boyer (2018: 14) even speaks of a 'well-defined legal right to relief'; hence, the location of 'settlement' was of utmost importance in determining which parish was responsible to provide relief for a pauper (Harris, 2004: 47). 'The poor laws more than any other legal mechanism defined the boundaries of community within England and Wales, linking rich and poor in a tight chain of dependency and responsibility' (Lees, 2006: 352). As parishes were more than reluctant to provide for residents of other locations, the determination of boundaries between parishes was often a matter of legal dispute (Snell, 2006). According to Snell (2006), the 'foreigner' was often considered to be someone from another parish or municipality as late as the 1940s. This conceptualisation of the 'other' or the 'foreign' was embedded in a culture of 'local xenophobia'.
'Settlement' was mainly based on place of birth but could also be acquired by marriage, property ownership, and some other conditions. Local authorities developed strategies to deter poor internal migrants from settling and would regularly remove paupers and vagrants to their last respective parish of settlement to prevent them from becoming a burden on the parish of destination (Feldman, 2003: 84-5). Although in absolute terms, local authorities made only limited use of removals, 6 the threat of removal was real and would deter migrants from applying for relief and incentivise them to move on. 'Migrants could stay if they were needed for work and were not on relief' (Lees, 2006: 51).
As time went by, the existing system of settlement was increasingly perceived to be a burden for economic development (Harris, 2004: 43). By the beginning of the 19 th century, many parishes provided non-resident poor relief but would charge the parish of settlement for supporting non-settled 6. For instance, between March 1827 and March 1828 43,677 individuals were removed from parishes in England and Wales (Feldman, 2003: 91).
residents (Feldman, 2003: 88). The New Poor Law of 1834 marked the onset of a shift in responsibility away from the local parish and towards the state, as it created Poor Law Unions who were responsible for the poor at a regional level, facilitating the migration of able-bodied rural poor to manufacturing districts (Anderson, 2013: 25), and minimising conflicts between parishes (Hurren, 2013: 133). Irrespective of the substantial international migration at the time, internal migration reached significant levels in England during the mid-19 th century and was seen as a way to escape dire economic conditions in depressed areas; consequently, long-term stability of residence became less dominant (Lees, 2006: 215). In industrial areas in the North of England, more than a quarter of the resident population were internal migrants; in cities like Birmingham, Liverpool, London, and Manchester, this figure was almost 60% (Pooley and Turnbull, 1998: 3). This high level of internal migration once again placed under stress the concept of settlement, which was perceived as limiting labour mobility and thereby the competitiveness of English industry (Hurren, 2013: 132). Subsequently, the principles of residence and 'irremovability' were introduced for people in England in 1841. Residents became irremovable when they had lived in a parish for a specified number of years, initially five years and since 1865 one year (Feldman, 2003: 93). 7 Costs for relief provision would no longer be borne by the Parish, but by the Poor Law Union (Boyer, 2018: 84). Nevertheless, pushing out the poor to neighbouring jurisdictions continued to be a core element of the 'crusade' to eradicate outdoor relief in the last two decades of the 19 th century (Hurren, 2013: 142-43).
At the beginning of the 20 th century, social insurance became an alternative to poor relief through the successive introduction of categorical social insurance schemes as part of the Liberal Reforms (Boyer, 2018: chapter 7). In addition, the 631 Poor Law Unions were abolished by the Local Government Act of 1929 and responsibility to support the poor was transferred to 145 counties and county boroughs, thereby greatly enlarging the tax base (Boyer, 2018: 240). While more than 20% of the population received some form of public support by the 1930s, 'only' 3-6% of the population remained on assistance. This contrasts with a century earlier, when 8-17% had been in receipt of poor relief (Boyer, 2018: 17-18;Lees, 2006: 180). The Poor Law and the system of 'local settlement' were finally abolished with the introduction of the status of 'national settlement' through the National Assistance Act of 1948. 8 Key to determining eligibility was now residence in Britain, as it was the purpose of the act 'to assist persons in Great Britain who are without resources to meet their requirements' (National Assistance Act 1948, Para 4). 9 The financing of assistance was fully centralised and funded by the national government (Boyer, 2018: 279). 10 To sum-up: "Nationalising welfare provision made the boundaries and scale of welfare entitlements symmetrical with those of the economy and the state, homogenising practices that had differed substantially throughout the British Isles" (Lees, 2006: 351).
In the People's Republic of China (PRC), local citizenship, which is largely based on birth right, was crucial for accessing social minima until the beginning of 21 st century. The so-called hukou 7. However, this irremovability did not apply to the Irish, who had the right to freedom of movement in England but could still be removed if they applied for poor relief until 1900 (Lees, 2006: 219). 8. National assistance continued to be an essential part of the post-World War II social security arrangement, as the value of national insurance benefits failed to rise in line with Beveridge's recommendations (Harris, 2004: 301). 9. The first restrictions for non-citizens were put in place in the 1960s, when the number of immigrants began to rise (Anderson, 2013: 38-41). 10. However, local residence continued to play a minor role in various communities even after 1948, especially with regard to housing (Feldman, 2003: 98).
system of household registration regulates movement within China and determines whether a person has access to social rights in the place of residence. It has been argued that citizens are only citizens when they stay in their 'native' places, and instead become 'alien nationals' or secondclass citizens once they migrate (Jieh-Min, 2010: 65). Full 'state-endorsed urban citizens' are only those that have a 'form of valid, official membership or affiliation in a city' (Solinger, 1999: 7). Whilst internal migrants have been welcome to contribute as workers to urban economies ever since the economic reforms of the 1980s, they have been regularly sent back to their localities of origin once they became an economic burden to relieve urban centres from unemployment and the costs of welfare support (Zhang, 2018: 10). However, more affluent migrants and highly educated people have had the opportunity to 'buy' the local hukou registration or citizenship in many cities of China (Wang, 2010: 343). The extent of second-class citizenship was quite significant at the turn of the century: migrant workers and their familiesoften called 'floating people'accounted for an estimated 20% of China's population (Zheng et al., 2020: 2). The level of secondclass citizenship becomes even more apparent when focusing on the population of specific large cities. For example, whilst the city of Shenzhen in the province of Guangdong had a de jure population of 1.25 million in the year 2000, the census counted a city population of more than 7 million inhabitants including migrant workers and their families (Chan, 2018: 33). A number of initiatives were undertaken at both the provincial and central government levels to reform the hukou system during the first two decades of the 21 st century. It is the declared goal of the central government to make access to social benefits and services easier for those who are locally resident but do not hold the local hukou registration. In general, larger cities have stricter limitations on accessing welfare benefits and educational opportunities, whereas medium-sized cities might relax hukou restrictions to encourage migrants to settle. In smaller cities with less than 500,000 inhabitants, acquisition of the local hukou should be open to all migrants who are habitually resident (Zhang, 2018).
According to the 'Provisional Regulations on Social Assistance' of 2014, social assistance is meant to provide a means of last resort and offer support in urgent situations and hardships. A local hukou was no longer mentioned as a precondition for eligibility; moreover, the responsible Minister stated that the temporary assistance programme should also cover needy urban residents without local hukou if they have lived in the city for a certain period of time (Zhai and Gao, 2019: 100). However, these policies have not (yet) been fully implemented across the country and cities continue to use the hukou system as a technique to render social citizenship a reward to 'deserving citizens' (Zhang, 2018: 15). At the beginning of the 21 st century, homeless and unemployed poor migrants in urban centres of the PRC without the proper hukou registration could still be ('voluntarily') repatriated to their (rural) places of origin (Economist, 2019), even though the respective law has been formally abolished in 2003 (Zhang, 2018: 10).

Internal Migration and Support for the Poor in Federations: Germany and the United States
Barriers to accessing poor relief and social assistance in federations are theoretically not only limited by the boundaries of the respective cities or the nation-state, as boundaries between subnational territorial jurisdictions can constitute additional barriers. The unification of a relatively large number of small German states to a so-called nation-state was only accomplished in the second half of the 19 th century. Historically, poor relief was provided by the municipality based on the principle of 'settlement' (Heimatrecht) within small regional jurisdictions (Reidegeld, 1999). Disputes between provinces and towns over the question of who was responsible for supporting a poor person were quite frequent (Frohman, 2008: 80-82), especially once migration from rural areas towards cities significantly increased during the 19 th century (Reidegeld, 1999: 215). In 1850, about 25% of low-skilled workers were no longer living in 'their' municipality of 'settlement'; this increased to almost 50% by the turn of the 20 th century (Sachße and Tennstedt, 1998: 195;207), underlining the inadequacy of a poor law based on Heimat at the local and regional levels (Sachße and Tennstedt, 2001: 216).
The North German Confederation (NGC), established in 1867, constituted an important step towards the economic integration of the fragmented German-speaking territorial entities, which subsequently led to the establishment of the German Empire only three years later. One of the first legislative acts of the NGC was to establish freedom of movement among the subjects and citizens of the Member States of the Confederation. Freedom of movement was considered by liberal politicians as a vital catalyst for industrialisation and economic development (Reidegeld, 1999: 250). However, the pre-existing poor relief regulations of the Member States initially remained in force, which hindered the mobility of workers. Otto von Bismarck, Governor of Prussia and Chancellor of the NGC, understood that a fragmented system of poor relief was a barrier to freedom of movement and economic development (Sachße and Tennstedt, 2001: 208;216). In 1869, he proposed the Relief Residence Law (Unterstützungswohnsitzgesetz) based on a law enacted in Prussia in 1842 (Sachße and Tennstedt, 1998: 199), which stipulated that every needy 'Northern German' be provided with poor relief at the place of residence irrespective of whether the person had already acquired the status of local residence. If a person had not been resident in the respective municipality for a minimum of two years or was younger than 26 years of age, the municipality of origin would have to reimburse the cost (Steinmetz, 1993: 115). Irrespective of these steps, Member States continued to have a high degree of flexibility in determining need and the level of support. Hence, it is not surprising that municipal officials also continued to try to move poor migrants out of their city limits. At the end of the 1870s, an estimated 400,000 poor people were homeless and living on the road without adequate support (Frohman, 2008: 162), 11 and it was largely up to voluntary associations to provide relief to poor migrant workers (Frie, 1997;Frohman, 2008: 160-172).
Although Bismarck's proposal was politically highly contentious, it was enacted by a two-third majority within the upper house, the Bundesrat. 12 During the deliberations of the law, the State of Saxony, for instance, emphasised that the change in law would result in it having to shoulder higher costs, as the net migration of the state was very high (N.N., 1869: 108, 113). Meanwhile, smaller states argued that the legislation would create an 'unbearable burden' for them (Pergler von Perglas, 1870). Even after the legislation was enacted, representatives from rural areas continued to complain about the cost they had to shoulder and demanded a reform of the law. In a reform of 1908, the age and the number of years required to acquire formal residence were reduced to 16 and one year respectively. Furthermore, the duration for which a municipality of destination had 11. This compares to approximately 1.2 million poor people who received poor relief at the end of the century (Frohman, 2008: 106). 12. The reasons for this high level of support remain unclear; Member States might have voted for the legislation due to the powerful position of Prussia or because they did not want to jeopardise the process of further economic and political integration.
to provide relief to needy sick persons without local resident status was extended as a further measure to relieve the fiscal burden of the rural regions (Steinmetz, 1993: 116-120). Article 7 of the Weimar Constitution of 1918 provided the Federal Government with the legislative competence to regulate the domain of poor relief and the support of needy migrants. Although most states favoured some reform of the poor relief system, they continued to insist that states retain most authority. According to Württemberg's Interior Ministry: 'Welfare is primarily the responsibility of the states and must remain so' (cited by Hong, 1998: 119). It took the hyperinflation of 1923 and the legislation of an emergency law, which provided the Federal Government with the authority to take all necessary measures by decree to minimise the hardship of the people, to overcome potential veto points. Using its newly acquired competence, the government issued the 'Regulation of the Duty to Poor Relief' (Reichsverordnung über die Fürsorgepflicht) in 1924. Accordingly, cities and districts would now be responsible to provide poor relief to all habitual residents in times of need, thereby abolishing all residence requirements (although some level of reimbursement still applied for those who were not habitually resident) (Hong, 1998: 120;Sachße and Tennstedt, 1988: 142-152).
One year later, the Federal Government issued guidance regarding the need standard as well as the overall benefit level (Richtsätze), whilst the cities and districts had to fund 90% of the support payments (Sachße and Tennstedt, 1988: 173-184). After World War II, this arrangement remained more or less in force until the enactment of the 'Federal Social Assistance Law' (Bundessozialhilfegesetz) in 1961. This new law introduced a legal entitlement to social assistance and provided for (nearly) uniform eligibility standards and benefit levels throughout the federal state. Whilst the local and state levels had to bear the highly varied costs, welfare bureaucrats involved in drafting the law were of the opinion that reimbursement of costs should, in principle, be something of the past (Föcking, 2007).
The right to interstate travel and freedom of movement (for white people) has been a core element of the US legal system from the beginning of the Republic. However, aside from immigration and naturalisation policies, the 'rights and obligations of American citizenship were defined at the state level' (Mettler, 2002: 231); moreover, states have held ''the police power, which included the responsibility to deal with public health, safety, and good order' of the community' (Mettler, 2002: 237). Based on these rights, many states restricted the freedom of movement of poor people well into the 20 th century (Katz, 1996). For example, according to California state law, it was unlawful to transport a non-resident 'indigent person' into the state. Eventually, this legal provision was challenged in the US Supreme Court in 1941, which ruled it unconstitutional, as the law violated the established 'interstate commerce' clause (Edwards v California, 314 U.S. 173). This was a big step towards removing barriers to freedom of movement for all citizens.
Historically, support for the poor in the United States was rooted in the English poor law tradition, whereby welfare was regulated at the state level and provided at the local level (Katz, 1996). The US Federal Government had no established constitutional competence to intervene in social policy matters (Kelly et al., 1991). This interpretation changed in the 1930s with the enactment of President Franklin D. Roosevelt's New Deal. The Social Security Act of 1935 introduced categorical social assistance programmes for needy single parents with children, the elderly, and disabled people. The Social Security Act provided states with a high degree of freedom in determining eligibility criteria and the right to impose minimum residence requirements. Within the programme for single parents, residence requirements were limited by the Social Security Act to a maximum of one year. This required 30 states to reduce the length of their existing residence requirements (Schwartz, 1985: 321). In subsequent years, technocrats and civil servants, such as the Commissioner of Social Security Arthur J. Altmeyer, continued to recommend the complete abolishment of residence requirements, as they were considered an anachronism in times of relatively high internal migration (Altmeyer, 1946). More than 60 congressional bills were introduced to modify or abolish residence requirements between the introduction of the law in 1935 and 1968. However, none of these proposals made it out of the Committee stage (Schwartz, 1985: 323), demonstrating the power of Southern conservative committee chairmen in the US Congress (Patterson, 1967), trying to preserve the political economy of the South (Gordon, 1994: 275).
By 1960, approximately 25% of all people in the United States had been born in a state other than the one in which they were currently residing (U.S. Census Bureau, 1963, Tab. 9). Nevertheless, residence requirements continued to be enforced: in 1960, approximately 100,000 needy persons were denied access to welfare based on residence requirements (Piven andCloward, 1971 [1993]: 391). In many states, poor people continued to be removed if they applied for welfare and did not fulfil residence requirements (New York Times, 1956). States feared that instant access would lead to increased immigration regardless of the migrant's willingness or ability to contribute to the economy. Furthermore, such a policy was argued to have detrimental effects on state budgets (cf. arguments in Shapiro v. Thompson;394 U.S. 618 (1968)). The Welfare Rights Movement of the 1960s challenged the residence requirements not only politically, but also in the courts, including all the way to the Supreme Court (Davis, 1993). Eventually, in 1969, the Supreme Court ruled that residence requirements violate the Equal Protection Clause of the 14 th Amendment of the U.S. Constitution: 'Neither deterrence of indigents from migrating to the State nor limitation of welfare benefits to those regarded as contributing to the State is a constitutionally permissible state objective' (Shapiro v. Thompson;394 U.S. 618 (1968): 633).
Subsequently, President Richard Nixon proposed a federal assistance programme with common standards. It was believed that such a federal programme would: a) be applauded by Northern taxpayers based on the assumption that national standards would minimise the perception among Northerners of unskilled poor Southerners as migrating into the North's welfare programmes; and b) benefit the South fiscally through a greater allocation of federal funds (Burke and Burke, 1974: 44). Yet, despite the benefits of more federal funding, Democrats in the South opposed the proposal in fear of changes to the regional political economy, as the proposed benefit level would significantly increase the reservation wage (Steensland, 2007: chap. 4). As a consequence of the opposition to Nixon's proposal of creating one federal welfare programme and guaranteeing equal standards for all in need, only the categorical social assistance programme for the elderly and disabled was fully nationalised in 1972. These target groups were considered as 'deserving' by all relevant actors, and the limited federal welfare programme was not considered as undermining local economies (Berkowitz and DeWitt, 2013;Burke and Burke, 1974).

Discussion
The four case studies have shown clear similarities and differences. In all cases, social citizenship was initially bounded and restricted to non-migrant permanent residents within local or regional jurisdictions. Internal migration has functioned and was perceived as a catalyst for economic development. People from poorer areas were migrating to developing and economically more prosperous regions, thereby contributing to economic growth primarily in the areas of destination. Clear parallels to freedom of movement and the concept of worker status within the EU may be drawn. However, once internal migrants became unable to support themselves and needed support, they were 'asked' to leave or their place of origin was requested to reimburse the cost for the support offered by the providing authority. Time and again, these arrangements led to tensions between jurisdictions. Needy newcomers without settlement status were often constructed as constituting a burden in the jurisdiction of destination, very similar to current debates in a number of EU Member States. Over time, conditions were relaxed in all jurisdictions, making it easier to earn social citizenship at the place of destination. Measures included introducing (lengthy) minimum residence requirements, reducing the length of said residence requirements and eventually abolishing them, and making larger territorial jurisdictions responsible for supporting the poor, thereby allowing for local/regional migration as well as broadening the tax base.
What were the drivers of change? Based on the logic of a Most Different Systems Design, it can be argued that the main 'variable' shared by the four politically and economically diverse jurisdictions is increased economic integration, of which increased internal migration was the crucial element. Thus, economic integration of the fragmented territorial jurisdictions can be interpreted as a key factor that is necessary for the abolishment of residence requirements.
In England, reforms introduced by the central government initially created larger territorial jurisdictions that were responsible for the poor and subsequently introduced the concept of 'irremoveability' (similar to the status achieved in the EU after five years of lawful residence). The introduction of various social insurance schemes led to a withering away of the importance of the Poor Law, and with it the concept of local settlement, as many needy people received support through these other categorical programmes. However, it was the critical juncture of establishing a national welfare state after the end of World War II that finally created a national social citizenship (including rights for migrants from the Empire). In the People's Republic of China, numerous steps taken in the early 21 st century by central government have made it easier for local authorities to deviate from the strict hukou registration system. Whilst some cities have taken the opportunity to make it easier for internal migrants to access social support, others remain strict in providing citizenship rights only to people with the appropriate registration.
A stepwise approach also characterised developments in the two federations analysed. In Germany, the creation of national social citizenship was achieved in a number of steps. Firstly, the introduction of freedom of movement and the enactment of the Relief Residence Law in the second half of the 19 th century were key in the process of nation building, enacted against the vehement opposition of a number of states and rural interests. Secondly, a number of reforms were achieved through legislation at the federal level as a result of political pressure by rural regions to shift the financial burden of providing poor relief towards the growing, more urban regions. Finally, the Weimar Constitution gave the federal level the competence to legislate within the domain of poor relief; nevertheless, it required a special enabling emergency law in 1923 to overcome political opposition to further reform and abolish all residence requirements for social assistance via decree. In Germany, we witnessed a dominance of the political realm to abolish residence requirements, whereas in the United States, political actors-primarily from the South-blocked a political solution with the exception of welfare for needy disabled and senior citizens. The US case demonstrates most clearly how particular economic and fiscal interests of certain states dominated the political discourse and constituted a veto point. Moreover, it took the welfare rights movement and their legal approach for the Supreme Court to establish full freedom of movement and abolish residence requirements. The Court based its rulings on the 'interstate commerce clause' of the US Constitution and the 'equal protection clause' of the 14 th Amendment.
In all four cases, residence requirements were (formally) abolished after significant steps towards economic integration and the stock of internal migrants had reached more than 20% of the population. In the EU, the percentage of persons living in a Member State of which they do not hold citizenship stands at 4.3% in 2021 (EU Commission, 2021: 20). Arguably, with the exception of England, residence requirements were abolished through hierarchical, non-democratic decisions. Popular public support or some form of national solidarity did not seem to play a significant role. 13

Conclusion
Barriers to accessing social support by internal migrants and highly charged political debates about who should shoulder the fiscal burden are in no way unique to the EU. The historical and international comparisons show that social citizenship had to be earned by internal migrants in jurisdictions that differed economically and politically, including state capitalist, liberal, and coordinated market economies, and authoritarian and democratic political systems as well as federal and unitary ones. In these very different systems, residence requirements initially fulfilled an economic logic and function for the local and/or regional economies. However, economic integrationwhich, to a significant degree, was driven by internal migration-led to tensions among jurisdictions about who should shoulder the burden of providing relief to migrant citizens. The above analysed cases demonstrate that these tensions were not resolved by a discursive recourse to some form of solidarity based on shared values. Moreover, as Durkheim has so eloquently argued, 'mechanical solidarity', based on a common collective consciousness and shared values, is becoming ever weaker in modern societies. According to him, it is the division of labour that is increasingly fulfilling the role of 'common consciousness' in ever more differentiated modern societies, based on 'organic solidarity' (Durkheim, 1997(Durkheim, [1893). Yet, whilst economic integration and an increasing division of labour were necessary conditions in all four cases, they were not sufficient conditions for the abolishment of residence requirements and the creation of social citizenship within the wider territory. Moreover, it took political power, various coalitions, or the leadership of political or judicial actors to overcome the barriers and hurdles on the path towards social citizenship in the wider territorial jurisdictions.
This does not mean that the arrangements of social citizenship within territorial jurisdictions, once achieved, remained uncontested. In Britain, for instance, the Localism Act that came into force in 2012 14 gave powers to local authorities in England to set their own rules about who is eligible to go on a housing register or waiting list. In 2013, the government issued further guidance recommending that councils in England require people applying for housing to have lived there for a minimum of two years. 15 In the United States, legislation reforming welfare for single parents during the 1980s and 1990s once again allowed states to introduce residence requirements. For instance, California introduced a rule whereby people who had been resident in the state for less than a year would only be entitled to the equivalent benefit level of the state of origin. 16 A key 13. Unfortunately, we do not have public opinion polls for England, the North German Confederation, or the PRC that demonstrate public support for residence requirements; nevertheless, given local and regional governments' opposition to granting 'foreigners' access to poor relief or social assistance, it can be inferred that the abolishment of residence requirements across these jurisdictions was not based on broad solidaristic support. Moreover, various public opinion polls provide clear evidence that a majority of Americans continued to support residence requirements up to the 1990s (Asen, 2001;LA Times, 1991). 14. http://www.legislation.gov.uk/ukpga/2011/20/contents/enacted. 15. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/5918/2171391.pdf. 16. This is similar to an approach included in the pre-Brexit referendum deal with the UK government and an approach chosen by Austria with regards to family benefits for the children of EU migrant citizens living in the Member State of origin. justification by California for the residence requirement was to protect the state budget (cf. Allard, 1998). Once again, however, the Supreme Court ruled the clause unconstitutional, arguing that, 'Citizens of the United States, whether rich or poor, have the right to choose to be citizens 'of the State wherein they reside'. The States, however, do not have any right to select their citizens' (Saenz v. Roe;526 U.S. 489 1999: 510-11). These two latter examples show how important decisions of political and juridical actors can be in undoing but also guaranteeing social citizenship.
What are the implications of this comparative analysis for the EU? Although history does not repeat itself, the institutional tensions associated with accessing social assistance in the analysed four multi-layered jurisdictions are similar in kind and have many obvious parallels with the tensions, developments, challenges, and discourses in the EU. 17 These include the concept of worker status, having sufficient means, becoming a burden, and the introduction of residence requirements to minimise the 'burden' on the jurisdiction of destination. The implication of the analysed cases is that, over time and assuming further economic integration, the EU may eventually introduce social citizenship for every EU citizen and abolish minimum residence requirements in the future, depending on political and judicial leadership (perhaps once again by the CJEU 18 ) and/or the formation of effective political coalitions. Economic integration is necessary, solidarity is helpful, but institutions and power really matter when it comes to ending practices that discriminate against migrants' access to minimum social protection!