Following the landmark essay of T. H. Marshall, Citizenship and Social Class (1949), it has conventionally been assumed that the introduction and expansion of social rights in Europe happened as the final stage of a long process of democratization that included the granting of first civil and then political rights. Although Marshall acknowledged that the sequence of rights was neither linear nor teleological, existing work on citizens’ social rights either presumes the existence of encompassing civil and political rights—as the background for the democratic struggle over society’s material resources—or that expanding civil and political rights leads to better social rights because poorer citizens get a vote and therefore force the elite to introduce social programs.Footnote 1
We present a radically different perspective on the relationship between the extension of suffrage (under meaningful competition for government power) and social rights, that is, state-financed entitlements that make citizens’ livelihoods independent from the labor market in the instance of events such as unemployment or sickness (see Korpi Reference Korpi1989: 314). First, some countries institutionalized a poor relief system, organized and financed outside the church, much before mass democratization. In these countries, the primary effect of suffrage extension was to reduce public social spending, not expand it. Second, the way this retrenchment occurred was partly by creating a negative link between social rights, on the one hand, and civil and political rights, on the other. A pauper making use of his or her social rights had to give up several civil rights and, occasionally, also political rights—what we call “citizenship disqualification.” This created a wall of stigmatization around the poor relief system, which made claimants abstain from activating their social rights, thus lowering total social spending. These findings suggest that the effect of suffrage extension can be quite heterogeneous across national contexts.
We introduce a framework that contributes to explaining the origins of modern welfare states in the nineteenth century, honing in on countries with an existing poor relief system run by secular authorities rather than the church. These countries underwent a number of common developments in the late eighteenth and early nineteenth century that increased demand for social protection and raised the burden on taxpayers. We argue that citizenship disqualification was the most likely result of these developments. However, we also explain how it was more pronounced where land inequality was high and where there was a prolonged period with competitive elections and limited suffrage—what we term “competitive oligarchy.”
Land inequality implied a redistributive conflict between the landed elites as the most significant taxpayers, on the one hand, and the peasants, and later urban dwellers, as the most prominent beneficiaries of poor relief, on the other. In the context of rising demand for poor relief, the landed elite wanted to cut back the generosity of existing poor relief schemes and avoid further expansion. Yet the landed elite was not always in a position to decide policy. Instead, citizenship disqualification proliferated where absolutist monarchy was overthrown, and suffrage expanded gradually from the top down with only the comparative wealthy having the right to vote. The new, wealthy median voter thus stood to gain from reducing public spending on poor relief. The solution was innovative: Claimants of poor relief had to abandon some or all of their civil or political rights; their citizenship was “disqualified.” The associated marginalization and stigmatization was intended to make the needy refrain from claiming poor relief and thereby helped rein in costs. Well into the 1930s, welfare state politics was more often a fight over the extent of this citizenship disqualification than about how generous social rights per se should be.
We test our argument with case studies of nineteenth to early twentieth centuries England, Denmark, Norway, and Prussia, all of which are paradigmatic cases that represent the variation in welfare state types. Despite a number of socioeconomic and political differences, England and Denmark both saw high levels of land inequality and sustained periods of competitive oligarchy resulting in extensive citizenship disqualification. Norway and Prussia represent negative cases in their own distinct way. Land inequality was much lower in Norway resulting in much less salient and widespread redistributive conflict and, consequently, a pronounced absence of citizenship disqualification. In Prussia, the liberal revolutions of early 1848 replaced the absolute monarch with a freely elected parliament based on a three-class voting and suffrage system. However, the king (and later the chancellor) alongside the bureaucracy and Junker landholders were the de facto power holders. Consequently, the political system of Prussia and the German Reich from 1871 remained autocratic to the bones until the dramatic transition to mass democracy in 1918, meaning that citizenship disqualification was never more than a marginal phenomenon.
We begin by outlining the theoretical argument. We describe the major economic and political background conditions for welfare generosity that changed dramatically before and around 1800 and the logic of political action that followed. Hereafter, the four cases are analyzed one at a time beginning with the two cases of unambiguous citizenship disqualification, that is, England and Denmark. This is followed by an analysis of the different developments in Norway and Prussia. Our findings provide a historical corrective to the understanding of the interaction between different kinds of rights in nineteenth- and early-twentieth-century Europe but also suggest that the size and nature of the winning coalition, rather than simply the instance of democracy or autocracy, determines welfare state expansion.
The Argument
Our argument is bounded by specific socioeconomic conditions that characterized Western and Central Europe by the late eighteenth to the early nineteenth century and laid the foundation for dramatic social and political reforms.
-
1. The Agricultural Revolution led to population growth and higher land inequality (e.g., Allen Reference Allen2000; Kluge Reference Kluge1979: 33; Løgstrup Reference Løgstrup2015; Overton Reference Overton1996: 1–9, 63–132). Across Europe, the last decades of the eighteenth century witnessed the Agricultural Revolution: a dramatic reallocation of land granted property rights to a new class of farmers that created more compact and productive farms. This revolution increased food production dramatically, which in turn led to a rising population. Despite property rights, the land reallocation thus meant that farmland effectively became concentrated in fewer hands.
-
2. Destitution increased dramatically (e.g., Brundage Reference Brundage2002: 23–25; Engberg Reference Engberg2011: 176–296; Kluge Reference Kluge1979: 33). The rising population and the concentration of farmland in combination created a surplus labor force. However, this new class of landless laborers constituted a challenge for governments. They were almost entirely dependent on employers for an income, be it in the form of money or in-kind goods, and often they had no source of living if without a job. Previously, low-income farmers could rely partly on what they could produce themselves and partly on the local community of farmers in similar circumstances. With the concentration of farmland in fewer hands, that possibility disappeared for many. In addition, many of the landless laborers became urban poor as they moved to the cities in pursuit of jobs.
-
3. Existing poor relief arrangements were not geared to the new situation (e.g., Lindert Reference Lindert1998; Rimlinger Reference Rimlinger1971: 13–33). Prior to the Agricultural Revolution, the Reformation in Northwestern Europe had shifted responsibility for the destitute from the Church to the state. The state, in turn, typically delegated the job to local communities. While never perfect, the system worked reasonably well by its own logic exactly because the majority of people were nested in social networks such as the family, the village community, or guilds. The combined forces of the agricultural and industrial revolutions changed that, boosting demand for poor relief to a level not seen previously. The burden on taxpayers rose accordingly.
-
4. Poor relief was normally paid for by the better-off local elite (e.g., Brundage 1975; Reference Brundage2002: 9–10; Kluge Reference Kluge1979; Philip Reference Philip1947: 20–24, 36–40). The organization of poor relief varied greatly across time and between regions, not to say countries. Yet a commonality almost everywhere was the presence of a positive economic gradient in who paid. In societies where income was generally low and unequally distributed, comparatively few had the surplus to support the needy with food and shelter (the normal mode of relief) or money. Rules often emphasized property as the primary determinant of who was liable to pay for poor relief, which reinforced the division between the landless laborers and the property-owning elite.
Our argument answers how the decision makers of the day handled these new socioeconomic conditions as a function of the magnitude of land inequality, and thus the strength of the redistribution conflict, and the nature of the political regime. Regarding the latter, we distinguish between absolutism, competitive oligarchy (parliamentarism and free and fair elections with property/income restrictions on suffrage), and mass democracy (parliamentarism and free and fair elections without property/income restrictions on suffrage).Footnote 2 We assume that all actors—both decision makers and non–decision makers—in all three regimes were motivated to maximize their own current and future income. Yet the way in which decision makers could do so varied across the three regimes according to two parameters: (1) the position in the preredistribution income distribution, and (2) the ability to externalize monetary costs from redistribution (above all else taxes).
In absolutist regimes, a single monarch set policy without regard to other actors, except perhaps a small circle of supporters. Monarchs were typically located at the top of the preredistribution income distribution, normally being the wealthiest of all members of society. Given that they had little more to gain materially, the primary motive of the monarchs was to protect the status quo. After the French Revolution in 1789, elites perceived that the primary source of threat against the status quo came from an overturn of the monarchy and a reshuffle of economic wealth by the lower classes.Footnote 3
The best way for the monarch to face discontent was to provide social programs that protected against future income loss to those individuals that were perceived as most likely to revolt, that is, the poor. The fact that income loss for the poor in this historical period often meant starvation and ultimately could spell death only increased the likelihood of revolt. For instance, increasing bread prices and consequent massive food shortage contributed to the French Revolution (Baker Reference Baker1987: 1–3, 6; Hobsbawm Reference Hobsbawm1962: 60–61).
If monarchs in the era were motivated to deliver some form of income compensation, they were also able to externalize the costs of such poor relief schemes onto the better-off citizens through higher taxes. In other words, by promoting tax-paid poor relief, the monarch could protect himself against social unrest and have someone else pay for it.Footnote 4
Compared to monarchies, competitive oligarchies and mass democracies shared the central feature that the median voter rather than the monarch set policy. The two regimes were distinguished from each other in terms of the position of the median voter in the preredistribution income distribution and the ability of the median voter to externalize monetary costs from redistribution. In combination, this created powerful policy differences between the regimes. First, in competitive oligarchies the income of the median voter was above the mean income of society, while it was below in mass democracies. Second, in competitive oligarchies the median voter could not externalize the costs from redistribution because, in a nutshell, redistribution required that the well-off (i.e., those with an above-mean income) paid to the less well-off (i.e., those with a below-mean income).
The political implications of this difference flow from the median voter theory of Meltzer and Richards (Reference Meltzer and Richard1981). In this model, the median voter prefers redistribution until median and mean income are identical (minus a margin to allow for transaction costs related to such reallocation of material wealth). However, the Meltzer–Richards model only works under a mass democracy where the median voter sets policy and median voter income is lower than mean income in society. In this scenario, the median voter stands to gain from redistribution because taxes are paid disproportionately by the individuals with above-mean income.
Yet the median voter may not always have a below-mean income. In many European countries, suffrage for the better part of the nineteenth century only included the 5–20 percent of male adults that were wealthy enough to fulfil the stipulated requirements. Voting rights only started expanding in the last decades of the century until universal suffrage was achieved in the first third of the twentieth century (though some places, like France and Switzerland, only somewhat later). In these competitive oligarchies, the median voter was from the outset well-off and therefore had a clear-cut interest in maintaining the status quo. In this way, he resembled the monarch. Yet whereas the monarch could externalize the costs of providing poor relief, the median voter in competitive oligarchies needed to pay himself using taxes. All else equal, the median voter in competitive oligarchies should be less willing to provide poor relief relative to both the monarch and the median voter in mass democracies.
The question remains why the nineteenth-century median voter in competitive oligarchies set up such a politically complicated scheme as citizenship disqualification to rein in public spending on poor relief. The answer is that he faced a set of practical problems. First, because of the rising social needs caused by the Agricultural Revolution, spending would automatically rise without reforms. Doing nothing would often not do the trick. Second, the organization of poor relief made modern-style retrenchment of benefit generosity difficult. Although such retrenchment took place, the fact that poor relief was already very meagre set clear limits to this strategy. Moreover, poor relief often did not imply money transfers but rather access to food and shelter, which in the context of the nineteenth century were hard to regulate precisely. Even if benefit generosity was cut, poor relief would at any rate still be a fiscal cost. Much better, then, to stop the destitute from claiming benefits in the first place. Third, it was pivotal that the truly poor still had access to poor relief; otherwise, they could be forced to take desperate means such as revolution. This third concern counterbalanced the first two.
Given that the median voter wanted to maximize his own security and prosperity, there was no perfect solution to these problems. Many, however, perceived citizenship disqualification to be the best alternative. This particular welfare policy promised to reduce public spending while maintaining access to poor relief for the truly poor, thus allowing them to survive. The idea was to ensure cutbacks in social spending in more subtle ways by creating a negative link between claiming poor relief, that is, to make use of one’s social rights, and the claimant’s civil and political rights. When a citizen activated his or her social rights, the law stipulated that he or she would automatically lose a number of civil and/or political rights. Such citizenship disqualification could take a variety of forms. Best known is probably the workhouse; a prison-like institution meant to contain and exploit claimants. The workhouse legislation effectively took away claimants’ freedom of movement. However, the workhouse was merely one of several such instruments. Claimants could also lose their property rights, the right to enter into marriage, the right to have children, and, finally, the right to vote. The latter could scare off already enfranchised from claiming benefits or provide an incentive for disenfranchised to give up benefit claims. All these encroachments on the civil and political rights of claimants were not only unpleasant on their own but also created a firewall of stigmatization and marginalization around the poor relief system, making it highly unattractive. In consequence, the number of claimants and public spending would decrease.
We have so far assumed that the three political regimes existed in isolation from each other. That is clearly not the case because they tended to follow one another in chronological order from absolutism over competitive oligarchy to mass democracy—albeit at a varying tempo across countries. Where countries democratized in a one-step move from absolutism to mass democracy—without experiencing competitive oligarchy—the poor were immediately enfranchised, and thus citizenship disqualification was off the table. By contrast, where countries underwent a sustained period of competitive oligarchy, citizenship disqualification was extensive. Here, the new median voter did not just instigate redistribution; he also wanted to dismantle the rules regulating citizenship disqualification. Without such dismantling, expanding generosity made little sense because accepting even generous benefits was out of the question for many people if it meant losing key civil and political rights. Thus, in countries moving from competitive oligarchy to mass democracy the first wave of income security reforms from the 1890s right up until the 1930s also concerned the dismantling of citizenship disqualifications for needy groups.Footnote 5
Case Selection
Our four cases, that is, England, Denmark, Norway, and Prussia, represent the three welfare regime types popularized by Esping-Andersen and constitute a paradigmatic set as such. They all share a historical experience with a tax-financed poor relief system, yet they differ on the degree of land inequality and the existence of competitive oligarchy. EnglandFootnote 6 and Denmark both had sustained periods of competitive oligarchy. Although the English monarch had remained influential after the Glorious Revolution in 1688, domestic policy was unambiguously the prerogative of the Parliament in England by the nineteenth century. Strict property rules meant that only the well-to-do could vote in large parts of the century. Even after the Reform Act of 1867, only 16 percent of the adult population could register to vote (Johnston Reference Johnston2013: 4). It was not until the Reform Act of 1918 that property requirements for men were lifted. Up until 1849, Denmark was ruled by an absolutist monarch. However, already from the first decades of the nineteenth century, the king’s power dwindled substantially as a property-owning elite gradually gained strength, ultimately leading to a full-blown, competitive oligarchy with the 1849 constitution. Property requirements for men were lifted in 1915 (Elklit Reference Elklit and Büsch1980: 370–80, 388–89; Knudsen Reference Knudsen2001: 76–78, 84–87). Compared to England, Denmark allows us to track the process of how a monarch goes from being truly absolutist, over a period of de facto power sharing, to the introduction of competitive oligarchy.
Norway in many ways resembles Denmark. Indeed, up until 1814, it was part of the Danish kingdom. In 1814, it also got a free constitution although with hefty property requirements on voting until universal male suffrage was granted in 1898. However, compared to Denmark, land inequality was low in Norway with 70 percent of all farms owned by families (rather than nobility) in 1850 compared to only 35 percent in Denmark (Vanhanen Reference Vanhanen2003). This suggests that the median voter was much closer to the mean income than was the case in Denmark. Moreover, Norway witnessed the emergence of a powerful decentralization movement that enhanced the affinity between small and big farmers according to geographic location, nullifying, to some extent rural, distributive conflicts (Rokkan Reference Rokkan1987: 116, 244). In short, Norway allows us to explore how the relative absence of land inequality hindered citizenship disqualification.
Like the other three countries, Prussia had a secular poor relief system. Like England and Denmark, high levels of land inequality produced rural conflicts and pressures for reform. Yet Prussia, and later Imperial Germany, never became a competitive oligarchy. The revolution in 1848 marked the beginning of the unification of Germany but ultimately failed to democratize most of the princely states, most notably Prussia. Until full-blown, mass democracy was installed in 1918, the political system of the unified German Reich in 1871 combined systematically rigged elections and a formally unchecked executive (the monarch in conjunction with the chancellor) with a Reichstag based on universal suffrage that gradually increased its ability to affect policy making (Anderson Reference Anderson2000; Koch Reference Koch1984: 80–81). The pre-1918 regime thus saw separated elements of pure autocracy and mass democracy; the executive did not originate from a free but exclusive competition as in England and Denmark. In turn, those who ought to have been most keen on cutting social spending were relatively marginalized (see, e.g., Beck Reference Beck1995: 149; Frohman Reference Frohman2008: Ch. 2, 59, 66–68; Steinmetz Reference Steinmetz1993: 110). Therefore, by studying Prussia, we are able to investigate how the absence of competitive oligarchy hindered citizenship disqualification.
Table 1 summarizes the case selection logic and the expected outcomes in terms of citizenship disqualification. In the cases of sustained competitive oligarchy and high levels of land inequality, we find England (competitive oligarchy throughout the studied period) and Denmark (gradually morphing to a competitive oligarchy from the first decades of the nineteenth century). In both instances, we expect extensive citizenship disqualifications. Because Denmark and England in the current context are most different systems, we can rule out the stage of economic development and previous experience with quasidemocracy as the most important potential confounders. England advanced rapidly toward industrialization in the first part of the nineteenth century, whereas Denmark only began to industrialize in earnest from the 1870s and 1880s. England also had much more experience with quasidemocratic institutions than Denmark, which had boasted the most absolutist constitution in Europe since 1660. By analyzing Norway and Prussia, we test the assumed mechanism and expect no or only modest citizenship disqualification because of the lower levels of rural conflict in Norway and the absence of competitive oligarchy in Prussia.
Table 1. Degree of citizenship disqualification and country cases
Our case studies are organized and presented chronologically to make the reading easier. This also avoids us reading history backward. By focusing on how actors calculated based on their historical present and past, we avoid deducing their behavior from what we know happened later. Within this frame, the analyses first describe each country’s political regime traits and initial socioeconomic conditions, including the magnitude of rural inequality and strength of rural conflict. Next, they enlist the country-specific evidence of how, given the rural conflict dynamics and political regime developments, decision makers thought about social rights and spending, including citizenship disqualification rules, and finally how these welfare traits developed. The evidence is based on official statistics and the political and social historiographies of the four countries.
England
Like other places in Northwestern Europe, the Reformation disrupted the delivery of social protection in England. In the face of a perceived rise in the number of vagabonds and paupers, the Elizabethan poor laws were conceived. The main innovation of what became known as the Old Poor Laws of 1597 and 1601 was to place the administrative and fiscal responsibility of poor relief squarely with the parishes; that is, in effect, with the local property-owning elite (Brundage 1978: 1–4; Reference Brundage2002: 9–10; Fraser Reference Fraser2009: 38–42; Solar Reference Solar1995: 3–7).
The Agricultural Revolution had caused a slowly augmenting hike in farm productivity from around 1550, but it was only from the mid-eighteenth century that the socioeconomic consequences became evident as a substantial rise in the precariousness of many people. First, in the 50 years leading up to 1800, the population grew by 50 percent to 8.7 million and then doubled again to 16.7 million by 1851. Second, a key ingredient of the Agricultural Revolution had been the integration of communal areas into a more efficient farmland production system; the so-called enclosure movement. The downside of the enclosure movement was that the ability of smallholders to maintain a living off the land was severely constrained. Paid work became the predominant source of income for smallholders, making them vulnerable to even small fluctuations in the economy. Third, as industrialization caught on, textile cottage industry was crowded out. This closed an alternative income stream that had been important in many countryside regions (Brundage Reference Brundage2002: 23–25; Fraser Reference Fraser2009: 42–43; Overton Reference Overton1996: 63–132).
During the Napoleonic Wars, corn prices had been high and the economic situation generally tolerable, but a series of postwar economic slumps and failed harvests in 1817–19, 1825–26, and 1828–29 created widespread countryside unemployment (Block and Somers Reference Block and Somers2003: 293, 309–11; Brundage Reference Brundage2002: 42–45; Fraser Reference Fraser2009: 42–43; Harris Reference Harris, Harris and Bridgen2007; Overton Reference Overton1996: 64–67). Given the lack of other means of existence, this meant that demand for poor relief rose. Precise data is hard to come by, but by one estimate public spending on poor relief almost doubled both as a proportion of GDP and per capita from 1776 to 1820–21 (Lindert Reference Lindert1998: 114). The cost of this doubling of costs fell almost exclusively on the well-off landed elite who, to make things worse, often also suffered from the downturns, thus having both their income squeezed and taxes raised simultaneously.
A debate, which had already been on its way by the late eighteenth century, between members of the elite about the appropriate policy response now intensified. On the one side stood the Malthusian abolitionists who wanted to get rid of public poor relief altogether. This was the drastic solution to Malthus’s (and many others’) analysis that society’s poor bred much faster than food production increased in part because they knew that they would receive support from the public purse if they needed it. On the other side stood the Benthamite revisionists. This group, whose arguments eventually would be accepted by a big majority of the elite (as manifested in the formative Royal Commission report of 1834), argued that some form of poor relief was necessary to maintain social order. Without it, revolution akin to the French one, which was in fresh memory at the time, might sweep away the elite. That argument was reinforced by the so-called Captain Swing riot in 1830 where agricultural workers across large parts of the country protested violently against joblessness and depressed wages. The riot was rather easily put down by the authorities, but it emphasized the Benthamite point (even as it may simultaneously have reinforced the opinions of those believing that poor relief undermined social discipline) (Brundage 1978: 15–45; Reference Brundage2002: 29–36; Dunkley Reference Dunkley1981; Fraser Reference Fraser2009: 46–57).
However, the Benthamites also strongly believed that a fundamental shake-up of the poor laws was needed to rein in costs. More exactly, poor relief had to be so unattractive that no one able to feed himself or his family would ever want it. The solution was the workhouse test. The premise for the test was the construction of a vast net of workhouses whose primary function was to deprive the claimant of his or her freedom of movement (staying in the workhouse was mandatory), property rights (everything produced belonged to the workhouse), family life (families were systematically and permanently separated), and even afterlife (the bodies of the dead were sold off to anatomic and surgical experiments and training in medical schools; something that people of the day believed prevented resurrection after death—maybe the ultimate punishment).Footnote 7 In short, the workhouse was deliberately designed to stigmatize claimants by depriving them of the majority of rights ordinary citizens enjoyed (Brundage Reference Brundage2002: 77–81; Fraser Reference Fraser2009: 50–59). The desired result was that the poor simply refrained from asking for poor relief in the first place. In the words of a contemporary high-ranking official,
The pauper naturally enough concludes that the relief he receives in the workhouse is a very inadequate return for the surrender of his liberty—the full occupation of his time—the value of his labour—the humiliation he must endure in being associated with some of the depraved and abandoned members of the community and the painful consciousness that he has lost all self-reliance and self-respect. Who can wonder that the honest poor would make every effort to keep out of the workhouse? (as quoted in Fraser Reference Fraser2009: 331).
In other words, the workhouse test was designed to automatically distinguish between “true” and “false” poor because only the former—those that were absolutely and unquestionably poor with no other place to go—would submit to the conditions of the workhouse. This was the central logic of the New Poor Law of 1834. In the years after 1834, more than 500 workhouses were built around the country (Driver Reference Driver1993: 79–85). In most places, they were mainly meant as deterrents; a threat of what could happen if a person asked for poor relief. The reason is that the cost of poor relief inside a workhouse was around three times that of providing poor relief the old-fashioned way (Fraser Reference Fraser2009: 62). For the local elite, who remained firmly in control of the poor relief system also after 1834 (Brundage Reference Brundage1972, Reference Brundage1975), this balancing act was vital: The workhouse should scare paupers away but should not be used to any great extent because that would defeat the cost-saving purpose of the entire exercise (Fraser Reference Fraser2009: 62, 67). As such, they never became the dominant mode of provision. Still, the result was a very substantial drop in public spending of almost 50 percent over the duration of less than a decade from 1830–32 to 1840 both when measured as a proportion of GDP and per capita (Lindert Reference Lindert1998: 114). The proportion of claimants similarly fell significantly, perhaps with as much as 50 percent too (Fraser Reference Fraser2009: 58; see also Lindert Reference Lindert1998: 110).Footnote 8
The New Poor Law system operated throughout the nineteenth century and even saw an intensification in its implementation in the 1870s and 1880s, which included the so-called crusade against outdoor relief (Brundage Reference Brundage2002: 112–20; Driver Reference Driver1993: 76–85; Fraser Reference Fraser2009: 172–73; Harris Reference Harris2004: 53–58). Still, with the Reform Acts of 1867 and the Representation of the People Acts of 1884 and 1918, the electorate became increasingly nonelitist (Johnston Reference Johnston2013: 4–5), leading to a mounting pressure against the workhouse test. The poor relief system was operated at the local level by a large number of elected Boards of Guardians. These boards decided if a pauper would get any support and whether that support would be granted inside or outside the workhouse. The electoral system for the boards strongly favored the landed elite, who de facto controlled the poor relief system using the boards. However, by 1894, these electoral rules had become so blatantly out of sync with the electoral realities in the country, that a radical overhaul was inevitable. The new electoral system had no property requirements and eliminated plural votes, and the effect of the franchise extension quickly became visible in terms of rising spending on the poor and a laxer implementation of the workhouse test (Brundage Reference Brundage1972, Reference Brundage1975, Reference Brundage2002: 124–27).
However, it took several decades before the New Poor Law of 1834 was fully dismantled. Like in the Danish case relayed next, the dismantling of citizenship disqualification concerned one social group at a time, and as in Denmark the first groups to be lifted out of the Poor Law framework were the old (in 1908) and the sick (in 1911). The pauper disqualification clause was abandoned in 1918. The major protagonist of these reforms was the Liberal Party, which was competing with the rising Labour Party over the votes of the workers. With the economic upheaval after World War I, this interparty competition became less important for policy making. Therefore, it was not until 1930 when a Labour government was dependent on the support of the Liberal Party for its parliamentary majority that the workhouse test was finally abolished. It took four more years and the full blow of the Great Depression before the 1934 Unemployment Act lifted all abled-bodied males out of the poor relief system, which became a generalized relief agency for residual conditions not covered by the main social programs (Brundage Reference Brundage2002: 133–53; Fraser Reference Fraser2009: 175–233; Harris Reference Harris2004: 53–58).
Denmark
The Danish case resembles the English one in terms of the extent of citizenship disqualification placed on poor relief claimants. However, the path that led to this outcome was distinct. In Denmark, the policy trajectory was shaped by a gradually changing power balance in society. Around 1800, the king was still truly absolutist, but the Napoleonic Wars marked a watershed. Denmark had been on the losing side as an ally of France and suffered severely both politically (with the loss of Norway) and economically (with a state bankruptcy in 1813). Although the country formally remained an absolutist monarchy, power shifted toward the emerging class of freeholders (former serfs who had acquired property rights as farmers of, usually, small- or medium-sized lands) with large consequences for citizenship disqualifications.
Inspired by the Agricultural Revolution that was happening elsewhere in Europe, a series of land reforms that would radically transform the social and political landscape of the country were implemented from the 1770s. As a crucial innovation, the landlords were forced to sell off land to some of their old tenants who became a new class of freeholders. This led to a marked growth in farming efficiency over the course of the subsequent decades as industrious freeholders implemented new techniques and worked hard on their own land. A downside of the land reforms was the creation of a new and big class of smallholders (husmænd), that is, people who own or run a small piece of land with unstable means of self-support. The smallholders had to make a living from selling their labor to the freeholders or large estate holders, generally at a measly pay. As the farming became more productive, the population grew, but the available land was limited to fewer people. Several decades before the Industrial Revolution came to the country, the ranks of the smallholding class exploded and with it severe destitution (Engberg Reference Engberg2011: 176–259; Feldbæk Reference Feldbæk1993: 109–22, 164–70; Løgstrup Reference Løgstrup2015).
The political response to the rising number of poor came comparably swift with the poor law of 1799 for Copenhagen and in 1803 for the rest of the country. The law signaled the royal government’s intent to alleviate some of the hardships suffered by large groups in society and was, by the standards of the day, uncommonly generous and ambitious. The most important break with the past was the granting of taxing rights to the local parishes that were made responsible for improving the living conditions of claimants. Moreover, new county authorities were created to allocate means between parishes with few and many claimants. The Danish poor relief system from 1803 constituted the high tide in public generosity toward the poor between the Reformation and the next 100 years. Importantly, the increased generosity was paid by the new class of freeholders, not by the absolutist king deciding the reform. That would prove to be decisive in the coming century (Johansen and Kolstrup Reference Johansen, Kolstrup, Petersen, Petersen and Christiansen2010: 180–98; Jørgensen Reference Jørgensen1940: 26–41; Philip Reference Philip1947: 27).
The backlash against the poor law came quickly, but it was intensified by the economic downturn that followed the Napoleonic wars and left the Danish state bankrupt. Though still an absolutist monarchy, the representatives of the taxpaying freeholders voiced their opinion bluntly. One public debater observed—in a statement representative of the tone of the debate for several decades—that “most of those who become needy are lazy, horny and wasteful humans” (as quoted in Johansen and Kolstrup Reference Johansen, Kolstrup, Petersen, Petersen and Christiansen2010: 240; see also Philip Reference Philip1947: 41–42).
In subsequent years, the freeholders grew into a property-owning elite. They gained more and more political clout, which provided them with a strong leverage to affect public policy in their preferred direction. As a symbol of their importance, the first quasidemocratic public body was introduced in 1831 with the advisory Stænderforsamlingerne. The basic criterion of suffrage was property of a certain (high) value, which meant that only 2–3 percent of the population—or 11–12 percent of men above the age of 25—could vote (Elklit Reference Elklit and Büsch1980: 369, 388; Knudsen Reference Knudsen2001: 76–78). It is noteworthy that one of the main issues brought forward by essentially all the elected representatives of the freeholders in Stænderforsamlingerne was how to reduce spending on poor relief (Jørgensen Reference Jørgensen1940: 59–83; Kolstrup Reference Kolstrup, Petersen, Petersen and Christiansen2010: 210–12). In 1841, the freeholders got more formal power as democratically elected municipality councils were introduced. Again, there was a strict property requirement, effectively excluding anyone but the well-off freeholders and other citizens of high status (Christensen Reference Christensen1997: 93–97). Because the municipality councils were meant to handle day-to-day administration of the poor relief, this gave the freeholders considerable power over the scope and form of public assistance (ibid.: 156–59; Kolstrup Reference Kolstrup, Petersen, Petersen and Christiansen2010: 237–40). In 1849, the absolutist monarchy gave way for a constitutional monarchy with an elected parliament. However, this parliament could not formally determine government formation. As before, suffrage was limited and included only relatively well-off males. Nevertheless, property requirements were still sufficiently low to allow a large part of the freeholder class, especially those elected to the higher chamber of Landstinget, to vote and to exert substantial influence on government formation and policy (Elklit Reference Elklit and Büsch1980: 370–80, 388–89; Knudsen Reference Knudsen2001: 84–87).
The effect of the political pressure from the freeholders and their representatives was evident already during the absolutist monarchy as the embattled royal government struggled to appease this increasingly important class. Yet the pressure grew even stronger as the formal political powers of the freeholding class were strengthened. Loads of legislation passed with the explicit intention of making poor relief unattractive to potential claimants. Collectively it summed up to an unprecedented deprivation of citizens’ civil and political rights to cut social spending.
From 1808, and with additional legislation in 1810 and 1828, all property of the claimants effectively became public. If the claimant died, all belongings were sold off, and the revenue was collected by the local authorities; the family would get nothing. In 1820, the royal government allowed the local authorities to force able-bodied claimants to work in exchange for poor relief. In 1838, the government began granting loans to the local authorities to set up workhouses. Workhouses had existed previously in various forms, but now the usage increased rapidly and stayed at a high level until the 1880s. In the heyday of the workhouses, around 1,200 institutions were scattered around the country. The idea of the workhouse was to imprison claimants, thus both keeping them off the streets and making it feasible to exploit their labor force in quasiindustrial productions. From 1824, people who either currently received poor relief or who had not repaid previously received benefits could not marry without the consent of the local authorities. This intervention into the private life of claimants was also ruthlessly pursued inside the workhouses where families were split and intimate relationships prohibited. In 1849, as Denmark got its first democratic constitution, it was only natural to exclude all current claimants, as well as those that had not repaid previous benefits, from voting or standing for election. In the next half-century, these and several similar laws stood unchanged. In 1891, a new poor relief law that codified many old laws into a single, coherent piece of legislation was passed. Apart from a few minor issues, the 1891 law maintained all the key provisions that had been invented since 1808 (Jørgensen Reference Jørgensen1940: 42–111, 181–208; Kolstrup Reference Kolstrup, Petersen, Petersen and Christiansen2010: 208–44; Philip Reference Philip1947: 45–49, 72–73).
Good statistics are difficult to come by, so an assessment of how effective the rules of citizenship disqualification were at keeping people from claiming poor relief can only be partial. The best nationwide data comes from the censuses that were carried out with intervals of several decades. In 1787, before the Agricultural Revolution had boosted the number of poor smallholders, the proportion of the population receiving poor relief was 3.2 percent. In 1834, after the explosion of destitution and the generous 1799/1803 law, but before the backlash was fully in force, 3.6 percent claimed benefits. By 1860, that proportion had dropped to 2 percent and by 1890 to 1.8 percent, that is, a 50 percent reduction (Johansen and Holten Reference Johansen and Holten2015: 17). Based on this data, it seems fair to conclude that the rules of citizenship disqualification did exactly what they were supposed to do.
From the 1870s, a political mobilization of both rural and urban poor slowly began. At first, the mobilization primarily took place within different party factions of the rural Liberal Party. However, gradually from around 1900, the Social Democratic Party and the labor unions emerged as the major pro-poor forces in society. The dismantling of citizenship disqualification was a key point for these new actors. However, at first, the push for the dismantling was weak and by-and-large blocked by the powerful Conservative Party (Højre), often in alliance with those factions in the Liberal Party that represented the large freeholders (Jørgensen Reference Jørgensen1940: 129–65; Kolstrup Reference Kolstrup, Petersen, Petersen and Christiansen2010: 302–8). For the conservative elite and the freeholders, dismantling should only come about if it served other purposes.Footnote 9
In the 1910s and 1920s, dismantling of citizenship disqualification became an explicit and important goal for the Social Democratic Party. At the same time, the party grew in strength. In the 1929 general election, the party won 41.8 percent of the votes and formed a majority government with the Social Liberals, a party mobilizing the smallholders. Despite being a majority in the lower house (Folketinget), the new government lacked a similar majority in the upper house (Landstinget). This meant that the large-scale dismantling envisioned by the social democratic leadership at first came to nothing. Only a few years later, in the midst of the worst economic crisis of the century that had left thousands of freeholders on the brink of ruin, was it possible to forge a compromise between the Social Democratic–led government and the parliamentary representatives of the freeholding class. This compromise, passed in 1933, meant the abolishment of citizenship disqualifications for the vast majority of benefit claimants. From that moment, citizenship disqualification became the clear exception although it was only in 1961 that the very last legislative residue was eradicated (Kolstrup 2011: 196–210, 213–25; Reference Kolstrup, Petersen, Petersen and Christiansen2012: 175–88).
Norway
The Norwegian constitution of 1814 introduced a competitive oligarchic regime. The right to vote was tied to property and granted to the powerful civil service, the urban bourgeoisie, and the freeholding farmers (Rokkan Reference Rokkan1987: 113). Consequently, only about 6–7 percent of the population was eligible to vote (Statistisk sentral byrå 2010: 3). This gave the farmers a strong, formal position. However, in reality they would remain marginalized until the 1830s when they started to mobilize against the civil servants that dominated both the state apparatus and politics. The farmers eventually achieved local self-governance during the 1830s and led a series of regime reforms through the nineteenth century, including the introduction of parliamentarianism after 1880 and universal male suffrage later in 1898, which moved Norway into the group of mass democracies (see Pittaluga et al. Reference Pittaluga, Cama and Seghezza2015: 334).
However, in contrast to Denmark and England (and Prussia), small family farms dominated in Norway. By the 1850s, the proportion of family farms was around twice as big in Norway as in Denmark. Although farmers tended to advocate smaller budgets, the conflict between different landed groups never became as intense as in Denmark (Rokkan Reference Rokkan1987: 116, 244), and that social rights legislation consequently was much less strict (Seip Reference Seip1984).
The economic ills that plagued the rural economies of the three other countries were also prevalent in Norway. From 1814 to 1845, the population grew by about 50 percent without any increase in arable soil (Kluge Reference Kluge1979: 33). The resulting economic desperation created substantial migration to the cities and North America. The demand for poor relief increased accordingly, in the cities rather than in the countryside. In the whole period, the destitute migrated from rural to urban centers and increased pressure on urban relief centers (Sundt Reference Sundt1968 [1867]). Farmers therefore had a twofold interest; restricting relief at “home” and stopping transfers from rural to urban poor relief districts “abroad.” The latter meant that the Norwegian poor relief debate would be equally concerned with harshness as with decentralization of administration and costs. This in turn meant that the elite facing the poor relief bill would be split into an urban and a rural stratum with the rural being mostly interested in restricting any sort of expense compensation between urban and rural localities. Therefore, Norway lacked the unified elite interested in restricting poor expenses that existed in Denmark and England.
Norway’s poor laws of 1845 and 1863 stand out compared to the harsh poor laws of England and Denmark in the nineteenth century in that they included no citizenship disqualifications. Instead, the 1845 poor relief law reflected the ongoing power shift from civil servants to aspiring farmers. The latter used the 1845 law to circumscribe the role of the civil servants in the administration of benefits, formalizing a decentralized system of administration (Ehrhardt Reference Ehrhardt2016: 62). This meant that the farmers would be aptly placed to vote their members into the now decisive local poor law councils (Kluge Reference Kluge1979: 33). In addition, by introducing a split system between the rural and the urban areas, the farmers had effectively shielded themselves from contributing to the increasing costs of the urban poor relief (ibid.: 34). In effect, because there always tended to be more claimants in urban centers, farmers’ most important point was to stop any form of cost sharing between urban and rural strata (Ehrhardt Reference Ehrhardt2016: 62).
The 1845 law also codified several aspects that were meant to reduce poor relief costs. The distinction between “deserving” and “undeserving” claimants was made for the first time. The elderly and children belonged to the first category, and the unemployed and those with an “inclination to start families without consideration as to one’s financial situation” belonged to the second (Kluge Reference Kluge1979: 31). But even so, it is noteworthy that the Norwegian poor law granted the deserving poor the right to aid, and workhouses played only a marginal role in the new law (ibid.; Seip Reference Seip1984). Furthermore, the creation of local ombudsmen, who lived in the same municipality as the claimant, led to a less than harsh implementation (Ehrhardt Reference Ehrhardt2016: 62–63). For those wanting to cut spending on the poor, the 1845 law was disappointing, and the fact that the 1845 law neither reduced social spending nor the number of claimants motivated the harsher 1863 law (Ehrhardt Reference Ehrhardt2016: 62; Kluge Reference Kluge1979: 73). The representatives of the farmers were now adamant about curbing spending, and the 1863 law consequently removed any mentioning of deserving groups, except for orphans (Kluge Reference Kluge1979: 39). Even so, no restrictions on civil or political rights occurred (1863 Law on Poverty). In sum, the law might have been motivated by bringing down spending, but it lacked the essence of citizen disqualification because it did not link the reception of poor relief with the loss of civil or political rights. Neither did it reduce claimant numbers, as these continued to increase.
By 1898, the Liberal Party—representing freeholders, smallholders, and urban radicals—had won the parliamentary majority and achieved the internal coherence needed to push through universal male suffrage (Luebbert Reference Luebbert1991: 121). This removed the previous income restrictions, and for the first time all males older than 25 could vote. Yet while the working class was now eligible to vote, it lacked both parliamentary representation and the organizational resources to protect their newly won voting rights. As a consequence, adjacent with the extension of the vote, the constitution was revised, so that all poor relief recipients lost the right to vote (1898 Constitutional Change). However, this was more a question of removing the political influence of the most radical parts of the working movement than a strategy of curbing social spending. Evidence of this can be seen in the 1900 poor law that reintroduced the statutory right to claim benefits when in need, thus allowing the unemployed to receive relief (Seip Reference Seip1984: 161–63). In short, increased generosity and leniency came at the expense of political rights but without any major citizenship disqualifications. Therefore, claimant numbers remained stable after the new law was put into effect (Norges Offisielle Statistikk 1909: 6).
Even if part of the 1898 constitutional change is regarded as citizenship disqualification, it was short-lived and quickly pushed back. Constitutional counteramendments in 1902, 1908, 1911, and 1916 gradually dismantled the disqualification, culminating in its complete removal in 1919 (Kluge Reference Kluge1979: 62). The 1919 constitutional revision was the result of an increasingly strong worker mobilization both in parliament and labor market and the weakening of farmer interests. In 1900, 3 percent of the labor force had belonged to a union. By 1918, this share had grown to 16 percent, while the vote share of the Social Democratic Party had increased from 5.2 to 31.6 percent (Bull Reference Bull1955). This increasing strength, in turn, spurred a more worker-friendly stance of the Liberal Party, weakening the farmer group within the party (Luebbert Reference Luebbert1991: 124–25; Seip Reference Seip1984: 285–87). The competition over the votes of the workers in Norway’s mass democracy meant that a powerful coalition emerged in favor of eliminating the loss of voting rights.
Prussia
Given its relevant historical background, Prussia initially looked much like Denmark. At the end of the eighteenth century, the royal household of the Hohenzollern had a firm grip on political powers and supported the preservation of landlordism in which the Junker estates drew revenue from taxing smallholders and exploiting the large rural proletariat as workforce while being taxed themselves by the king (Eddie Reference Eddie2013: 90). Like in Denmark, serfdom produced widespread destitution. According to one estimate, the lower classes constituted around 70 percent of the rural population (Gray Reference Gray1986: 17–19). Ever since the Reformation, such poverty was managed by a state-led poor relief system. The poor had become “desanctified,” and Catholic monasteries had been dissolved, giving way to poor relief that was administered by urban magistrates and financed by landlords (Gorski Reference Gorski2003: 18–19). Access to the system of poor relief came through working and living on an estate (Eddie Reference Eddie2013: 90).
Also like in Denmark, the rural population grew rapidly in the decades around the mid-eighteenth century—around a 50 percent increase in the eastern provinces (Gray Reference Gray1986: 24). As this demographic change coincided with decades of decline in food crops, the landlords raised demands on their labor while prices on primary commodities rose. This initiated a social crisis of hitherto unseen proportions (Blackbourn Reference Blackbourn1997: 52; Eddie Reference Eddie2013: 138). However, from then, the political and social histories of Denmark and Prussia over the next century diverged. The royal response in Prussia was similar to that in Denmark and meant a widening and strengthening of social rights to combat social problems. However, the liberals in Prussia never became a powerful opposition that could succeed in demanding democratization. Instead, core liberal groups adopted the royal strategy of accommodating the poor to avoid social revolution, in turn making citizenship disqualification much less prominent.
During the eighteenth century, bureaucrats, primarily educated from universities, formed the emerging middle class. This “bureaucratic” middle class was empowered (although not enfranchised) by the strengthening of political and civil rights in the 1794 Allgemeine Landrecht and the Stein reforms implemented after the military defeat against Napoleon in 1806. From then, a bureaucratic absolutist regime crystalized as an alliance between a constantly expanding and increasingly autonomous bureaucracy, on the one side, and the monarch, including to some extent the Junker landlords, on the other. Their common interest was to save the old political order from increasing social pressures of urban and rural poor while trying to upgrade Prussia into a world power (Rosenberg Reference Rosenberg1958).
The alliance between nonenfranchised but politically powerful groups that basically backed the monarchy hindered the emergence of a strong freeholding class. In accordance with the logic of our argument, it also laid the foundation for the welfare state. In the eighteenth century, workhouses had been used, primarily as a means for the local landlords to keep track of paupers in their area and to deter claimants in the first place. Yet the laws from 1794 and 1806, respectively, removed the workhouse functions and took the first steps toward the abolishment of serfdom and installation of freedom of movement for paupers. These reforms were finalized in 1816 (Beck Reference Beck1995: 149; Frohman Reference Frohman2008: Ch. 2, 59, 66–68).
The king and the bureaucracy, who invented this so-called Elberfeld system of poor relief, were certainly worried about how to finance poor relief in a situation in which rural destitution was increasing and state revenues were decreasing. As a result, poor relief was only given after a means test and preconditioned on the worker actively applying for jobs. In this way, the system only addressed the needy and differentiated between the poor to attach stigma and combat welfare dependency (Crew Reference Crew and Geoff1996: 322; Steinmetz Reference Steinmetz1993: 42). However, the system was never primarily intended to keep taxpayer contributions at a minimum. Instead, the Prussian liberals adopted royal concerns for creating a mobile labor force and growing industries as well as paying the poor to abstain from rebellion, which made poor relief look more like social investment (Beck Reference Beck1995: 152, 157, 166; Crew Reference Crew and Geoff1996: 319). Thus, while the Danish Stænderforsamlinger pushed monarchs to set up workhouses and force paupers to work in exchange for poor relief, the Prussian executive set paupers free and never considered citizenship disqualification.
The European revolutions of 1848 also hit Prussia, but the monarch, the bureaucracy, and the Junker class managed to maintain decisive powers over government formation and policy. In 1849, Prussia was in fact the first European country to grant universal (male) suffrage in parliamentary elections (Pittaluga et al. Reference Pittaluga, Cama and Seghezza2015: 318). However, Prussia neither qualified as a mass democracy nor a competitive oligarchy. First of all, the rebellions proved too weak to install genuine parliamentarism as the autonomy of the new assemblies was quickly repressed by royal forces. Second, elections were systematically rigged by the landrats in favor of conservative supporters (Beck Reference Beck1995: 223). This caused low turnout rates among the poor and concentrated legislative powers on the few rich hands (Ziblatt Reference Ziblatt2008: 620–21). In turn, universal suffrage was more form than reality. The unified German Empire from 1871 to a great extent adopted the Prussian political system in which the ministers and the chancellor were de facto only responsible to the king but granted the Reichstag considerable influence through formal veto powers in social and economic legislation (Anderson Reference Anderson2000; Koch Reference Koch1984: 80, 111; Pittaluga et al. Reference Pittaluga, Cama and Seghezza2015, 328). In this sense, the continued rigging of elections to hold back liberal and socialist oppositions proved increasingly ineffective. Yet the point we make here is that social policy was either set by a nonelected executive or a universally enfranchised parliament, not an oligarchy.
In this peculiar, semiauthoritarian system, monarchic and mass interests reconciled. It should thus come as no surprise that the raison d’être of the Prussian Elberfeld system continued and remained the principle of the welfare state system that emerged in the latter half of the nineteenth century and early decades of the twentieth century. The Poor Law of 1855 did tighten the use of means tests by denying poor relief to those who had changed residence during the last year. But again, the main motive was to hinder a mobilization for revolution among the trade unions that organized the urban poor—receiving poor relief did not have consequences for civil or political rights (Blackbourn Reference Blackbourn1997: 131, 204, 280).
The king and Bismarck at first gradually had to strike a balance between advancing industrial growth by maintaining free movement of labor while deincentivizing factory-born social revolution, which they feared more than anything (Beck Reference Beck1995: 105; Steinmetz Reference Steinmetz1993: 110). Thus, the Elberfeld poor relief system was gradually substituted by social insurance schemes covering sickness, old age, and disability—policies that were supported by the Social Democratic Party (SPD) in the Reichstag (Frohman Reference Frohman2008: 100; Steinmetz Reference Steinmetz1993: 126, 129, 135–36).
The social insurances and remaining poor relief only came with very limited means of citizenship disqualification. At the annual meeting of the North German Confederation in 1869, both liberals and conservatives expressed concerns that the statutory poor relief system would lead to a false impression among the poor that they had a right to relief. However, despite these worries, a system of voluntary contributions to poor relief never passed, and statutory relief prevailed (Frohman Reference Frohman2008: 78–79). This reflected a more general development in the way the authoritarian elites and their liberal supporters viewed welfare. Apart from its instrumental value in attaching industrial workers to the state, welfare was increasingly seen as a way of shaping norms and social relations in society and creating a more competent and stronger labor force (Crew Reference Crew and Geoff1996: 323–24; Steinmetz Reference Steinmetz1993: 120). On the face of it, one case of citizenship disqualification was the 1869 suffrage law that stated that persons who had received public assistance in the preceding year were not allowed to vote in the upcoming parliamentary election. However, this was meant to exclude workers from enfranchisement more than making them abstain from receiving public assistance (Frohman Reference Frohman2008: 99–100). In turn, the law did not entail the mechanism of citizenship disqualification that we would expect. The threat of withdrawal of political rights was not a means to reduce social spending but rather a means to de facto exclude the poor from enjoying the franchise. Increased social spending thus reduced the number of enfranchised, not the other way around.
In the years between Bismarck’s withdrawal from power in 1890 and World War I, SPD yielded stronger influence on social policy. Alongside the liberal-monarchical, autocratic alliance, the social insurance model continued and social rights expanded. Laws like that from 1869, which coupled public assistance with withdrawal of suffrage, continued excluding public dependents from political participation while social spending kept increasing (Steinmetz Reference Steinmetz1991: 21).
In 1918, the Weimar Republic installed mass democracy in one stroke: The chancellor became responsible to the Reichstag (Koch Reference Koch1984: 80–81). The first elections immediately gave SPD the government power. Through the 1920s until Brüning’s emergency government in 1930, the SPD dominated social legislation accordingly, widening unemployment schemes and eligibility for being insured while strengthening health care and other welfare services. But the principle and purpose of granting social insurance by a means test to deter workers from extending their absence from the labor market remained the same (Crew Reference Crew and Geoff1996: 329; Steinmetz Reference Steinmetz1991: 23). Thus, the trend of rising social spending without citizenship disqualification continued because there was no intermezzo of competitive oligarchy in between the bureaucratic absolutist regime and mass democracy.
Conclusion
The analysis presented informs us about the intricate links between the expansion of civil and political rights, notably suffrage, and the development of the welfare state; arguably, two of the principal political events of the past two hundred years. Although Marshall (Reference Marshall1949) was highly sensitive to the historical context of the English case he explored, his argument has morphed into a rather rigid analytical schema, or common wisdom: Civil rights came first, then political rights, and, finally, social rights. What is more, Marshall only studies England, leaving aside the historical experiences of other European countries. This implies that it becomes difficult to understand the circumstances under which social rights developed as well as how these related to civil and political rights.
Our theory and empirical analysis suggests that in the nineteenth century this representation is too schematic; there are no necessary links by which social rights build on civil and political ones. Citizens in all our four countries—paradigmatic for research on welfare state formation—enjoyed certain social rights well before mass democracy came about, and, perversely, the expansion of voting rights to (comparatively well-off) people became a pervasive threat against these social rights. For most of the nineteenth century, those in need were typically better off living in an absolutist monarchy than in a quasidemocracy. We must therefore reinterpret citizenship rights and the origins of modern welfare state altogether.
Our findings, however, do not only provide a historical correction but also have implications for contemporary debates on the relationship between different political regimes and redistribution. Much large-n research indicates that democracies, despite the classic expectations of the Meltzer–Richards model, have no stronger propensity for redistribution than autocracies. However, this research also indicates that social rights granted in autocracies often function as club goods meant to secure regime survival, whereas the universality of social rights is stronger in democracies (e.g., Knutsen and Rasmussen Reference Knutsen and Rasmussen2018). Our findings speak to this literature by highlighting the need to consider carefully the mechanisms connecting different kinds of rights. Rights do not relate to one another in any constant, functionally specific way, but are molded by the size and nature of the winning coalition that defines the political regime in place. We may hypothesize that where the winning coalition has an above-mean income, welfare state expansion is more likely to come as club goods that exclude those outside the winning coalition.
There are few equivalents of the eighteenth- and nineteenth-century absolutist monarchies in today’s world. Yet today’s autocracies resemble nineteenth-century competitive oligarchies in key respects. Most contemporary autocratic regimes are “electoral authoritarian” in the sense that elections are held but biased against the opposition because the executive can rely on a small but powerful group of relatively well-off supporters—the winning coalition (Bueno de Mesquita et al. Reference Bueno de Mesquita, Smith, Siversen and Morrow2005). Much like in competitive oligarchies, this coalition limits the powers of the executive but is also tied to the executive by means of co-optation. In turn, the incentives for constraining social rights that existed in nineteenth-century competitive oligarchies constitute a striking parallel to today’s world of autocracies. On the one hand, this gives a micro-level foundation for extant critics of the Meltzer–Richards model’s prediction of linearity between levels of democracy and redistribution. On the other hand, it brings this classic model back in by reemphasizing how it is mass democratization, that is, the removal of income barriers to suffrage, that drives redistribution. Nevertheless, what we have shown is that mass democratization may under some circumstances restrain social spending.
