2.1 Revenue-Generating Instruments

Revenue-generating instruments represent the financial mechanisms through which social welfare systems are funded, embodying the complex intersection of economic policy, social redistribution, and state–society relations. These instruments are not merely technical financial tools, but deeply political mechanisms that reflect and help reproduce social contracts, distributional preferences, and broader institutional arrangements. Revenue-generating instruments differ along at least six dimensions: (1) their public–private nature, (2) their tax base (what and who is being taxed), (3) their link to income (being progressive, regressive, or proportional), (4) their link to benefits (having a requited or unrequited nature), (5) their funding mechanism (being fully funded, partially funded, or pay-as-you-go (PAYGO)), and (6) their visibility.

The “Public–Private” Dimension – Social programmes can be publicly financed through various tax methods that are legally obligatory, such as the personal income tax (PIT), consumption tax (like VAT), and SIC.Footnote ¹ Yet, social programmes can also be financed through individualized forms of financing derived from private or occupational payments. While such payments can be voluntary, they can also be obligatory, as is the case when states require workers and or employers to contribute to individual private pension funds or when a private co-pay is required when utilizing a public programme. As the review in Section 2.1 demonstrates, the reliance on public or private forms of financing has far-reaching implications for welfare entitlements, programme coverage, characteristics, and the state’s redistributive capacity (Schmitt et al., Reference Schmitt, Lierse and Obinger2020).

The “Tax Base” Dimension – refers to the economic activities and assets that form the basis for applying the tax rate and determining the tax liability of individuals, businesses, or other entities. Different tax methods vary in their respective tax base. For example, in the case of PIT, the tax base typically comprises the overall income earned by an individual, encompassing wages, business profits, capital gains, and other income sources. In contrast, SIC applies exclusively to wages. Meanwhile, the tax base for VAT is determined by the total value of goods and services sold.

Two aspects of the tax base bear implications for the revenue generation capacity of welfare states: its broadness and its impact on economic performance. A broader tax base allows even small changes in rates to generate substantial revenue increases, unlike narrower tax bases. Among major tax instruments, VAT is potentially the most socially inclusive, as it covers not only the formal sector but also some input purchases by the informal sector (i.e., unregistered economic activities operating outside government regulation). In contrast, the tax base of PIT is limited to income earners in the formal economy (i.e., economic activities that are registered, regulated, and taxed by the government), entirely excluding the informal sector (Genschel and Seelkopf, Reference Genschel and Seelkopf2021; Wagner, Reference Wagner2012). The breadth of the tax base is also influenced by policies that create exemptions, loopholes, and thresholds, all of which can shape the overall scope of taxable activities.

The tax base also influences revenue generation potential through its impact on economic performance. The taxation of certain economic bases is argued to have fewer distorting effects, making them more conducive to economic growth and thus entailing a larger revenue potential (Prasad & Deng, Reference Prasad and Deng2009). Similarly, some activities are less mobile than others, rendering them less vulnerable, under conditions of globalization, to tax competition pressures (Genschel & Schwarz, Reference Genschel and Schwarz2011). Consumption taxes, argued to possess these two traits, feature a broad base that facilitates the development of large welfare states. Simultaneously, their reliance on consumption, a less mobile economic activity, insulates them from the pressures of globalization in contrast to, for instance, corporate taxes (Genschel, Reference Genschel2002; Morgan and Prasad, Reference Morgan and Prasad2009). To the contrary, several studies have argued that social policy funding by social security contributions pose an economic risk, as it increases labour costs and reduces firms’ competitiveness (Kemmerling, Reference Kemmerling2009).

The “Link to Income” Dimension – refers to how individual payments relate to income, manifesting in progressive, regressive, or proportional tax structures. In a progressive system, tax rates rise with income, reducing income inequality. Regressive taxes, however, place a disproportionate burden on lower-income earners by applying lower rates as incomes increase. Proportional taxes maintain a constant rate regardless of income, preserving existing income inequalities.

Different tax instruments create varying links to income and in turn impact redistribution. PIT is typically progressive, taxing higher incomes at higher rates, thereby decreasing income gaps after taxation. Conversely, VAT tends to be regressive, as it applies a uniform rate that burdens lower-income individuals more heavily relative to their earnings, thus widening income gaps. Social insurance contribution, usually seen as proportional, is neutral in distributive terms, because everyone contributes the same percentage of their income. However, when there’s a ceiling on insurable income, higher earners pay a lower effective rate, introducing regressive redistribution (Genschel, Reference Genschel2002).

While some tax instruments may be regressive, private financing mechanisms tend to be even more so (Koreh et al., Reference Koreh, Zemmour and Palier2024). For example, private health insurance premiums are lump sums based on coverage and individual risk, rather than income, causing higher-income individuals to pay a smaller share of their earnings compared to lower-income ones. However, it should be emphasized that the redistributive outcomes of tax systems are not determined by structure alone but depend significantly on tax levels – the total tax revenue as a percentage of GDP (Guillaud et al., Reference Guillaud, Olckers and Zemmour2020). Though progressivity enhances redistribution at any given revenue level, countries with lower overall tax levels, even if they have progressive systems, may only see limited redistributive outcomes. In fact, research shows a negative correlation between tax progressivity and tax levels, indicating that more progressive systems often generate less revenue, which in turn limits their capacity to reduce income inequality (Guillaud et al., Reference Guillaud, Olckers and Zemmour2020; Steinmo, Reference Steinmo1993). We further discuss this stylized fact in Section 4.

The “Link to Benefits” Dimension – examines the connection between payments and receipts in welfare provisions across different tax instruments. This link ranges from “requited” (where there is a direct relation between what you pay and what you receive) to “unrequited” (where there is no direct correlation). Private financing instruments, such as out-of-pocket payments and private insurance premiums, are highly requited in the sense that individual payments reflect service prices, coverage, and risk profiles. At the opposite end, general taxation instruments, such as PIT, or VAT, are unrequited and taxpayers cannot expect specific benefits in return (Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009).

Between these ends, hybrid forms exist. For example, SICs operate on a partially requited basis, but the connection between payments and receipt is weaker than in private financing and can vary in strength (Campbell and Morgan, Reference Morgan2005; Kemmerling, Reference Kemmerling, Genschel and Seelkopf2021; Wagner, Reference Wagner2012). In weaker forms, benefit eligibility of a SIC relies on a contribution record, but benefits themselves are uniform and unrelated to individual contributions. In contrast, under stronger linkages, contributions correlate with entitlement levels, underpinning the income-related nature of many social insurance programmes (Clasen, Reference Clasen2001).

These variations in the link between payments and benefits have significant consequences for social protection characteristics, impacting programme coverage and eligibility criteria (Cichon et al., Reference Cichon, Scholz and Van De Meerendonk2004; Manow, Reference Manow and Palier2010; Morel & Palme, Reference Morel, Palme and Greve2018). Contribution-financed programmes typically cover only wage earners in the formal economy who pay into the system. Where contribution levels are linked to entitlement levels, these programmes maintain a status-preserving function, allowing beneficiaries to sustain their standard of living during periods of unemployment or incapacity to work (Clasen, Reference Clasen2001; Esping-Andersen, Reference Esping-Andersen1990).

In contrast, tax-financed social protection, which is unrequited, can adopt different forms of coverage. It may provide universal benefits based on citizenship, as seen in many Nordic welfare models, where all citizens receive protection regardless of their tax contributions. Alternatively, it can be means-tested, targeting the very poor, as in Anglo-American welfare systems. In these cases, eligibility is generally conditioned on income levels, excluding those who contribute but exceed the income threshold.

These differences have also important redistributive implications affecting the state’s ability to reduce inequality (Schmitt et al., Reference Schmitt, Lierse and Obinger2020). The closer the link between payments and receipts, the less scope there is for redistribution. Out-of-pocket payments provide no redistribution, as individuals pay directly for services without any risk pooling. Private insurance, while offering some degree of risk pooling, still limits redistribution, because premiums are tied to individual risk profiles and coverage levels.

Social insurance contributions offer greater redistributive potential, particularly when contributions are linked to salary rather than set at a flat rate. This structure ensures that individuals contribute based on their ability to pay. In contrast, universal tax-financed programmes provide broad coverage and significant redistribution by spreading the cost across the population, including those who don’t directly benefit. At the highest end of redistributive potential sit means-tested, tax-financed programmes, which target resources specifically toward those with the lowest incomes while being funded by higher earners.

However, just like tax systems, redistributive outcomes in social protection depend not only on the structure of tax–benefit linkages but also on the size of the redistributive pool (i.e., both the taxes collected and the spending levels). Political factors, such as the legitimacy of the welfare state and the willingness to pay taxes, play a key role here. In their well-known work on the paradox of redistribution, Korpi and Palme (Reference Korpi and Palme1998) argue that overall redistribution tends to be higher in systems where contributors, through either social insurance or universal structures, can expect to benefit from the system themselves. In contrast, strictly targeted, means-tested systems, despite being the most progressive in structure, often result in limited overall redistribution due to a smaller resource pool, leading to higher poverty and inequality rates. This insight has been confirmed by more recent studies (Brady and Bostic, Reference Brady and Bostic2015; Jacques and Noël, Reference Jacques and Noël2018).

The “Funding Mechanism” Dimension – refers to the methods through which financial resources are accumulated and utilized to cover the costs of social programmes throughout their life cycle. This dimension reflects critical intertemporal choices regarding cost allocation between present and future, taking three primary forms: PAYGO, fully funded, or semi-funded approaches (Cichon et al., Reference Cichon, Scholz and Van De Meerendonk2004; on intertemporal choices see Jacobs, Reference Jacobs2011).

Private financing instruments, such as insurance premiums for private health insurance or contributions to private pension accounts, are typically fully funded. In a fully funded system, current contributions not only cover immediate expenses but also accumulate sufficient assets to fully cover future liabilities. For example, in a fully funded pension account, the fund’s total assets at any time are required to match its projected future obligations. In contrast, general taxation methods follow a PAYGO approach, where present tax revenues are used to cover current spending without creating a dedicated reserve. A tax-based pension scheme, for instance, uses current budget allocations to fund ongoing benefits.

Social insurance programmes typically employ a semi-funded method, which falls between these two approaches. In this model, a reserve is built up, but it does not cover all future liabilities. Instead, it serves two primary purposes: mitigating the frequency of contribution increases and providing a financial buffer during economic downturns.

The distinction between these financing methods is particularly significant in the debate over pension systems and their ability to cope with demographic aging. Advocates of fully funded pension systems argue that they are better equipped to handle aging populations because each generation accumulates the necessary funds to finance its own retirement. In this view, even if the ratio of workers to retirees declines, retirees will have pre-funded their benefits, preventing an increased burden on younger cohorts. This argument gained traction in the 1990s and 2000s, prompting pension reforms that shifted from PAYGO to funded models (Orenstein, Reference Orenstein, Kott and Droux2013).

Yet, this perspective remains contested. As Barr (Reference Barr2002) and others have argued, from a macroeconomic standpoint, funded systems do not inherently offer a clear advantage over PAYGO. The core issue is not whether pensions are funded or PAYGO but how demographic aging affects future output. Even if retirees accumulate financial assets, they still require goods and services produced by the working population. A shrinking workforce, therefore, constrains an economy’s ability to meet these demands, regardless of the financing model (Morel and Palme, Reference Morel, Palme and Greve2018). Thus, demographic aging presents fundamental challenges to pension sustainability across all funding structures.

At the same time, both models face distinct vulnerabilities. While fully funded systems are exposed to financial market risks – such as volatility, inflation, and economic downturns that can erode accumulated assets – PAYGO programmes rely on future governments’ political will, making them susceptible to shifting fiscal priorities.

This underscores why the fiscal sustainability of pension systems cannot be assessed solely based on their funding mechanism. Sustainability depends on broader factors, including economic growth, labour market dynamics, and political choices regarding benefit levels and contributions. While fully funded systems may alleviate immediate fiscal pressures on the working-age population, they are not immune to structural economic risks. Conversely, despite their reliance on intergenerational transfers, PAYGO systems offer built-in flexibility to adjust benefits and contributions in response to economic and demographic shifts.

Beyond pensions, funding mechanisms shape the financial capacity of social programmes more broadly. Fully and semi-funded approaches transform social programmes from mere expenditure mechanisms into potential capital generators. By accumulating and investing funds, these mechanisms generate additional resources through interest earnings and investment returns, thereby expanding the programmes’ spending capacity (Jacobs, Reference Jacobs2011).

The choice of funding methods also influences political dynamics. Pay-as-you-go systems typically focus political debates narrowly on benefit levels and contribution rates. In contrast, fully and semi-funded programmes introduce additional complexities, as the accumulation and investment of programme assets can serve additional societal objectives – such as industrial policy, government debt financing, or financial market development (Estevez-Abe, Reference Estevez-Abe, Ebbinghaus and Manow2001; Kangas, Reference Kangas, Hujo and McClanahan2009; Musgrave & Musgrave, Reference Musgrave and Musgrave1973; Park, Reference Park2011).

Moreover, in fully and semi-funded systems, the investment component reshapes political constituencies and fosters new interest groups. Stakeholders’ concerns extend beyond traditional welfare considerations to include the management, and strategic deployment of programme funds, generating new power struggles and policy trade-offs (Hassel et al., Reference Hassel, Naczyk and Wiß2019; Koreh & Béland, Reference Koreh and Béland2017). As further elaborated in Section 3, these broader political interests play a crucial role in shaping the adoption, design, and evolution of social policies.

The “Tax Visibility” Dimension – refers to how noticeable or transparent the tax imposition is to individuals. Visibility encompasses the clarity and awareness that taxpayers have regarding the taxes they are required to pay. Scholars argue that different tax instruments vary in their visibility and that direct taxes, like PIT, tend to be more visible than indirect taxes, such as VAT (Martin, Reference Martin2015; Steuerle, Reference Steuerle1992). This is because individuals directly see the amount deducted from their income when paying income taxes, making the tax obligation more apparent. In contrast, the VAT is included in the price of goods and services, making it less noticeable to consumers. Social insurance contributions are deducted from wages but are seen as less visible than income taxes because employers typically pay a portion of them. While scholars argue that both employers’ and employees’ contributions impact net income (Palme et al., Reference Palme, Nelson, Sjöberg and Minas2009), workers generally perceive only their direct deductions as costs. Even when employer contributions appear on pay checks, they are not always seen as affecting take-home pay, making SICs seem like a lower-cost way to fund benefits compared to more visible direct taxes.

Such visibility differences have influenced welfare state politics and state revenue capacity. Visible taxes are seen to provoke greater opposition, often leading to tax revolts and backlash, whereas invisible taxes are associated with greater tax consent and, consequently, a higher capacity for states to raise revenues and finance generous welfare states (Campbell & Morgan, Reference Morgan2005; Martin & Gabay, Reference Martin and Gabay2018; Steuerle, Reference Steuerle1992; Wilensky, Reference Wilensky2002). However, as discussed in Section 5, there are relatively few studies of public opinion of different taxes and the visibility argument is not always confirmed by empirical studies.

This brief review highlights the distinct characteristics of revenue-generating instruments and the trade-offs they entail. Value added tax, being broad based, regressive, and low visibility, is an efficient revenue-raising tool often dubbed a “money machine.” However, its regressive nature means it disproportionately burdens low-income earners. In contrast, PIT, with its progressive structure, often exempts lower incomes, alleviating their tax burden. Despite this, its high visibility and unrequited nature make it politically sensitive, often sparking resistance that constrains its use for mass revenue generation.

Social insurance contributions, characterized by their requited nature and lower visibility, provide broader coverage and political acceptability, though they often exclude non-workers and, in some cases, low-earning or part-time workers, leaving some individuals without social protection. Private financing mechanisms, like individualized pension contributions or private health insurance, are highly requited and regressive, and while they may exacerbate inequality, they remain crucial in countries with weaker state capacity for providing essential social services (Schmitt et al., Reference Schmitt, Lierse and Obinger2020). These varying features underscore not only the centrality of welfare state financing instruments in shaping social policy but also identifies the inherent political and fiscal challenges in designing comprehensive welfare state financing systems.

2.2 Tax Expenditures

Tax expenditures are policy instruments that intersect directly with welfare state financing, as they point to how the state can forgo tax revenues to advance social policy objectives. From this perspective, as Adrian Sinfield (Reference Sinfield, Lymer, May and Sinfield2023) explains, tax expenditures overlap with what Richard Titmuss (Reference Titmuss2019) called “fiscal welfare,” a form of social policy delivery distinct from “social welfare” (public social benefits and services) and “occupational welfare” (benefits tied to employment status). This perspective has long informed the comparative study of social policy, with a focus on the “changing boundaries of the welfare state” (Flora and Heidenheimer, Reference Flora and Heidenheimer1981), and it remains influential today (Morel, Reference Morel2025; Pavolini et al., Reference Pavolini, Jessoula, Natili, Greve, Moreira and van Gerven2024; Sinfield, Reference Sinfield2020). As for the concept of tax expenditures, which “mirrors the term public expenditures to show spending policies being run through the tax system” (Sinfield, Reference Sinfield, Lymer, May and Sinfield2023: 46), it was developed by economist Stanley S. Surrey (Reference Surrey1973), who published a major book on the topic focusing on the United States more than half a century ago.

Tax expenditure instruments encompass exemptions, deductions, allowances, or credits targeted at specific groups or “social” activities (Morel and Palme, Reference Morel, Palme and Greve2018). Tax expenditures subsidize and incentivize private welfare provisions such as pension savings, housing, healthcare, and childcare expenses. The OECD calls them “tax breaks with a social purpose” (Adema et al., Reference Adema, Fron and Ladaique2011). Despite not involving direct monetary transfers like cash benefits, tax expenditures constitute an indirect form of spending through foregone revenues (Howard, Reference Howard1997), making them a crucial element in welfare provision.

When we turn to the recent, in-depth, political and policy analysis of tax expenditures, much has been written on the topic in the United States (Bellafiore, Reference Bellafiore2018; Burman et al., Reference Burman, Geissler and Toder2008; Faricy and Ellis, Reference Faricy and Ellis2014; Hacker, Reference Hacker2002; Howard, Reference Howard1997; Mettler, Reference Mettler2011). For instance, in the widely cited book The Hidden Welfare State, Christopher Howard (Reference Howard1997) reminds us that tax expenditures in general “can take the form of tax deductions, tax credits, preferential tax rates, tax deferrals, or outright exclusion of income from taxation. They subsidize a broad range of activities, from oil exploration to the rehabilitation of historic buildings” (3). In his book, Howard (Reference Howard1997) documents the expansion of social policy-related tax expenditures in the United States.

The four policy areas studied systematically in the book point to the diversity of tax expenditures, which can belong to different areas of the welfare state such as housing, pension, and employment policies, including home mortgage interest deduction aimed at fostering home ownership, “favorable tax treatment of employer-provided retirement pensions” (Howard, Reference Howard1997: 48), the Earned Income Tax Credit for low-income workers, and the Targeted Jobs Tax Credit for members of vulnerable populations seeking employment. As Howard (Reference Howard1997: 24) shows, these four examples are only the tip of the tax expenditure iceberg in the United States, as the use of the tax system to promote social policy is both widespread and diverse in nature and scope.

Massive and fragmented, the “hidden welfare state” has a clearly regressive effect. This is the case partly because subsidizing employers to encourage them to offer health insurance and pension coverage to their workers means that “more affluent citizens are the main beneficiaries of the hidden welfare state. As a general rule, these benefits are most likely to be available to workers in larger companies, unionized industries, and better-paying occupations” (31). In this context, Howard (Reference Howard1997) claims that the “hidden welfare state” has redistributive patterns that are distinct from the “visible welfare state.”

Finally, partly because of this, the politics surrounding the “hidden welfare state” are distinct in nature, leading to the argument that “the differences between direct expenditures and tax expenditures as policy tools affect the nature of political support for each type of spending” (12). Yet, the politics surrounding the “hidden welfare state” varies greatly from one policy tool to the next, something that reflects the internal fragmentation of this welfare state (Howard, Reference Howard1997).

The findings from Howard’s book were widely influential, encouraging other scholars to directly consider tax expenditures in the analysis of social policy in the United States. For instance, Jacob Hacker (Hacker, Reference Hacker2002; Hacker, Reference Hacker2004) stresses the importance of tax expenditures in the development and politics of voluntary private health insurance and pension benefits in the United States. His 2002 book, The Divided Welfare State, draws on the work of Howard (Reference Howard1997) and points to the importance of tax expenditures aimed at stimulating the development of private health insurance and pensions in the United States: “the two largest expenditures in the federal code are the exclusions of employer-sponsored health insurance and employer pension contributions and plan earnings from taxable income – which together reduce annual federal revenues by nearly $200 billion, about what is spent on Medicare” (11). The broader analysis discusses the role of tax expenditures within the politics of private social policy in the United States. In that sense, his work is close to Jennifer Klein’s (Reference Klein2004) historical analysis of the rise of private health insurance and pension benefits in that country.

Even more than Hacker’s book, Suzanne Mettler’s Reference Mettler2011 The Submerged State: How Invisible Government Policies Undermine American Democracy stresses the importance of paying close attention to the “hidden welfare state” (Howard, Reference Howard1997) in the United States. A student of public opinion and public policy, Mettler (Reference Mettler2011) stresses the negative impact of “invisible” policies, such as tax expenditures on US democracy, as they make citizens largely unaware of key public actions that affect their lives profoundly. Turning to early fiscal and social policy reforms enacted towards the beginning of the Obama administration (2009–17), she argues that the lack of awareness of these policies on the part of citizens fostered political discontent and contributed to the rise of the Tea Party movement.

While not focusing exclusively on tax expenditures, Mettler’s book The Submerged State (Reference Mettler2011) does locate them in a broader conversation about democracy and social policy in the United States that explicitly builds on the work of Howard (Reference Howard1997) and his followers. For her, the “submerged state” and the tax expenditures featured in it are the product of an ongoing conservative restructuring of US social policy that favours narrow economic interests at the expense of the majority of citizens, who remain largely unaware of costly policies affecting them indirectly. In fact, the more a welfare state relies on “hidden” or “submerged” tax expenditures, the less importance voters put on social policy issues (Gingrich, Reference Gingrich2014).

This scholarship suggests that reliance on tax expenditures is not merely a technical debate between indirect over direct spending instruments but has instead distinct distributional and political outcomes. Tax expenditures typically favour middle- and upper-middle-income groups (Howard, Reference Howard1997) and encourage the development of private welfare markets, which also tend to benefit these groups (Morgan, Reference Morgan2005). Moreover, due to the oft-hidden nature of state provisions, dependence on tax expenditures affects welfare state politics and legitimacy (Hacker, Reference Hacker2002; Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009).

The public opinion implications of tax expenditures are explored systematically in The Other Side of the Coin: Public Opinion toward Social Tax Expenditures. In this recent book, Faricy and Ellis (Reference Faricy and Ellis2021) “find not only that social tax expenditures are generally popular, but that these programs are supported by subpopulations that normally are not inclined to favor public social spending” (13). They find tax expenditures are popular for three reasons: “a policy design’s appeal to citizens who favor limited government, its interference with attempts to politicize and racialize beneficiaries, and its signals to citizens about the deservingness of program beneficiaries” (9). In this context, there is broad public support for social tax expenditures in the United States. For Ellis and Faricy, this situation has important implications for the politics and future of both sides of the US welfare state:

Following Howard (Howard, Reference Howard1997; Howard, Reference Howard2021; Howard, Reference Howard2023), Hacker (Reference Hacker2002, Reference Hacker2004), and Mettler (Reference Mettler2011), Faricy and Ellis (Reference Faricy and Ellis2014, Reference Faricy and Ellis2021) emphasize the need for social policy scholars to take social tax expenditures seriously, a call consistent with the welfare state financing perspective featured in this Cambridge Element. This is the case because, at least in the United States, looking at how the state might forgo massive tax revenues to promote social policy objectives is a major piece of the welfare state financing puzzle.

Over the last three decades, the literature about the United States dominated the academic conversation on tax expenditures in social policy. This is the case because academic research about social tax expenditures has proved much more limited in scope in most other advanced industrial countries (for early exceptions see Castles Reference Castles1994 and Greve Reference Greve1994). For instance, less than a decade ago, Morel, Touzet and Zemmour (Reference Morel, Touzet and Zemmour2018) wrote that, with a few exceptions (Adema et al., Reference Adema, Fron and Ladaique2011), “in Europe the role of fiscal welfare has remained a blind spot in most social policy studies, national and comparative alike” (550). Yet, they show that, in Europe, tax expenditures and their reforms are important because they play an “instrumental role in the development of private health and retirement insurance plans across welfare states” (Morel, Touzet and Zemmour, Reference Morel, Touzet and Zemmour2018: 550). Linked to privatization, the expansion of tax expenditures in various European countries is described as a component of the broader trend towards pro-market social policy restructuring (Morel, Touzet and Zemmour, Reference Morel, Touzet and Zemmour2018).

Using a tax–benefit microsimulation model, Silvia Avram (Reference Avram2018) explores the scope and effect on income redistribution of tax expenditures in six European Union countries: the Czech Republic, Denmark, Germany, France, Italy, and Spain. She finds that tax credits and tax allowances “are by no means a policy instrument intended only/mainly for the rich” in these countries, but that “despite being widespread, their distributional consequences are generally not progressive” (284). This leads her to conclude that her research confirms “previous insights from the ‘hidden welfare state’ literature that suggested tax expenditures are more beneficial to middle and higher income groups” while stressing that “the distributional effects of tax allowances and tax credits are complex and often unanticipated” (284).

In a recent article about Europe, Salvador Barrios et al. (Reference Barrios, Figari and Gandullia2020b) explore the fiscal and equity effects of social tax expenditure in twenty-eight countries. Focusing on housing, education, and health, they document the widespread use of tax expenditures across Europe (Barrios et al., Reference Barrios, Figari and Gandullia2020b: 366). Their quantitative analysis “suggests that the impact of TEs on tax revenues and income inequalities is non-negligible. The redistributive impact of housing TEs is regressive in most countries. Education TEs tend to favour middle- to top-income working-age categories, while health TEs tend to favour the top-income class and the elderly” (Barrios et al., Reference Barrios, Figari and Gandullia2020b: 366). These findings are consistent with some of the scholarship on tax expenditures in the United States, which point to their regressive side (Howard, Reference Howard1997; Mettler, Reference Mettler2011).

In another quantitative analysis, Barrios et al. (Barrios et al., Reference Barrios, Coda Moscarola and Figari2020a) explore the fiscal and social effects of tax expenditures for pensions in twenty-eight European countries and find that “the revenue impact of pension-related tax expenditures can be sizeable.” More important, and in contrast with some of the US scholarship discussed above, they find that pension-related tax expenditures tend to be progressive by favouring lower-income elderly people and by draining resources from working age individuals at the top of the income distribution through partial or no deduction of pension contributions (Barrios et al., Reference Barrios, Coda Moscarola and Figari2020a). These findings are interesting as they point to the diversity of tax expenditures and of their divergent effects on income distribution.

Beyond this type of large-N quantitative comparative analysis and overview publications by international organizations such as the OECD (2010), single country case studies about the role of tax expenditures offer a closer look at the use of this type of fiscal welfare in specific national contexts (Branco and Costa, Reference Branco and Costa2019; Collins and Hughes, Reference Collins and Hughes2017; Morel et al., Reference Morel, Touzet and Zemmour2019; Provencher et al., Reference Provencher, Godbout and St‐Cerny2022; Sinfield, Reference Sinfield2020; Stebbing and Spies-Butcher, Reference Stebbing and Spies-Butcher2010). Simultaneously, small-N comparative scholarship on tax expenditures is also most helpful. For example, while much of the existing scholarship focuses on liberal welfare states, where the use of tax expenditures has long been extensive, in her recent book The Politics of Fiscal Welfare, Nathalie Morel (Reference Morel2025) examines the politics and effects of fiscal welfare in France and Sweden, two contrasted welfare states where the use of fiscal welfare is more recent. Focusing on the example of the introduction of a similar 50 per cent tax deduction on the purchase of domiciliary care and household services, she explores the politics behind the introduction of this tax benefit scheme and its effects on care provision, showing how this scheme has contributed to the privatization of care provision, as well as to inequalities in access to care linked to the highly regressive distributive profile of the benefit. She also highlights the policy feedback effects of this specific policy instrument, which has created powerful vested interests with a stake in defending this tax benefit. Morel’s findings for France and Sweden regarding the partisan politics of tax expenditures, the distributive profile of tax benefits, the role of tax expenditures in supporting the privatization of welfare, and the policy feedback effects of this policy instrument are largely consistent with the findings from the United States described above.

For social policy scholars at least, the “hidden welfare state” is no longer hidden, even if the politics of tax expenditures remains understudied in most parts of the globe, with the partial exception of the United States. Yet, even in this country, more research is needed to explore the politics of tax expenditures, including their ideological content and their potential feedback effects of the politics of welfare state reform (Russell, Reference Russell2018). Tax expenditure literature has primarily focused on its effects on inequality, social divisions, and welfare state politics, particularly regarding welfare state support. However, while tax expenditures by definition involve a loss in state capacity to generate revenues through taxes, their overall impact on states’ fiscal capacity has yet to be comprehensively addressed. There are compelling reasons to assume that extensive use of tax expenditures can hinder states’ fiscal capacity.

For instance, Mettler’s (Reference Mettler2018) finding that citizens are less aware of these types of benefits and that this fosters antagonism suggests that the hindering effect of tax expenditures goes beyond the straightforward loss of revenues generated by tax exemptions. One reason is that politics increasingly centres on attaining more tax exemptions instead of creating new or expanding existing programmes that provide cash transfers or public social services.

Furthermore, findings indicate that tax expenditures tend to reproduce and intensify divisions between citizens from different economic classes and particularly benefit high- and medium-income earners. This suggests that when countries go down a less redistributive path, politics will increasingly focus on expanding tax exemptions, further reducing states’ revenue capacities in a cycle that can exacerbate itself.

There is a need to integrate different scholarships from public opinion, interest groups, and welfare state politics to examine the political feedback effects from tax expenditures to fiscal development. This interdisciplinary approach could provide a more comprehensive understanding of the long-term implications of tax expenditure policies on state fiscal capacity and social cohesion.

2.3 The Governance and Administration of Financing Instruments

Tax administration refers to how taxes are assessed and collected. From a tax policy perspective, the efficiency of administration is a key concern, as it directly influences revenue collection. Efficiency is typically measured by the ratio between net tax revenue and the potential revenue given tax rates (Kiser and Karceski, Reference Kiser and Karceski2017). Well-organized systems and modern technologies can significantly enhance revenue outcomes. However, for social policy scholars, tax administration matters not only for efficiency but also because different tax instruments – such as social insurance contributions versus general taxation – are embedded within different governance structures. These structures shape not only revenue collection but also the politics, characteristics, and outcomes of social policy.

The central distinction lies in whether tax administration is centralized or decentralized. In systems where social policy is primarily tax-based and centrally administered, such as income tax and VAT, revenue is pooled into a single fund controlled by the Treasury. This centralization grants the Treasury significant power over social spending, as it controls resource allocation. Fiscal decisions regarding social programmes are integrated into the general budget, allowing the Treasury substantial leverage in determining how much funding welfare programmes receive (Koreh and Shalev, Reference Koreh, Shalev, Maron and Shalev2017; Truchlewski, Reference Truchlewski2020).

In contrast, SICs are often governed in a decentralized manner, typically through separate social security agencies (OECD, 2007). In such systems, contributions are managed independently from the general budget, often by semi-autonomous agencies within the state administration or even by social partners, such as trade unions and employer associations, as seen in many European countries (Palier, Reference Palier2010). This setup reduces the Treasury’s control and increases the influence of other actors, such as social partners. This decentralization limits both the capacity and the motivation of the state to intervene in social security funds, thereby shielding social programmes from immediate fiscal pressures (Koreh & Mandelkern, Reference Koreh and Mandelkern2023; Manow, Reference Manow and Palier2010).

These structural differences also impact the power dynamics among key actors. In decentralized systems, social partners, such as trade unions, hold significant influence, especially in managing social security funds. For example, French trade unions, despite having relatively low membership density compared to their British counterparts, have been more successful in resisting austerity measures. They have repeatedly managed to compel governments to address fiscal challenges by raising SICs rather than cutting social spending (Truchlewski, Reference Truchlewski2020).

Moreover, decentralized systems offer greater protection from fiscal pressures (Koreh & Mandelkern, Reference Koreh and Mandelkern2023; Koreh & Shalev, Reference Koreh, Shalev, Maron and Shalev2017). Social insurance funds, managed separately from general budgets, are less vulnerable to budget cuts or political interference, making it harder for governments, especially right-wing parties, to dismantle social rights. In contrast, tax-financed welfare states, where funding is subject to annual budget negotiations and heavily influenced by the Minister of Finance, make it more difficult for left-wing parties to prevent cuts to welfare programmes (Truchlewski, Reference Truchlewski2020). Hence, when governments impose fiscal consolidation measures to reduce budget deficits, social programmes funded by social insurance are less likely to be cut than programmes funded by general taxes (Jacques, Reference Jacques2024). Instead, governments have political incentives to raise contributions, especially as these increases are often automatic and social partners are able to block cutbacks (Manow, Reference Manow and Palier2010; Truchlewski, Reference Truchlewski2020).

Recent reform processes in countries with BismarckianFootnote ² welfare traditions further illustrate the centrality of these administrative structures. These countries – such as Germany, France, Austria, and Belgium – are characterized by social insurance systems primarily funded through employer and employee contributions rather than general taxation and administered by their representatives (social partners), rather than being directly managed by the state. As Palier (Reference Palier2010) explains in The Long Goodbye to Bismarck, reforms often targeted the fiscal autonomy of social insurance administrations and sought to centralize control. By increasing social insurance systems’ reliance on general tax revenues and shifting decision-making power to the state, reforms gradually eroded the influence of social partners. For example, a 1996 constitutional amendment in France gave parliament the authority to vote annually on the social insurance budget, thereby reducing the system’s autonomy. In Germany, the Hartz reforms of the early 2000s diminished the role of social partners in managing unemployment insurance, transferring more control to the state (Hinrichs, Reference Hinrichs and Palier2010; van Berkel, Reference van Berkel2010).

These institutional changes in Bismarckian systems were far from merely technical adjustments. By increasing fiscal centralization and enhancing social security agencies’ dependence on taxation, these administrative reforms served as prerequisites for later structural changes that cut benefit and altered eligibility criteria (Palier Reference Palier2010). Similar centralizing processes have emerged beyond Bismarckian systems. In recent decades, tax governance has become increasingly centralized, across various welfare regimes, with reforms unifying the collection and administration of SICs alongside general taxation under the growing control of ministries of finance (Barrand et al., Reference Barrand, Ross and Harrison2004; OECD, 2007). This development suggests that social policy practitioners and scholars should pay close attention to administrative-level changes. Understanding these dynamics is crucial, as seemingly technical shifts in governance can have profound implications for social policy outcomes and the protection of social rights.

2.4 Conclusion

Section 2 has explored the complex architecture of welfare state financing, highlighting the diverse array of instruments and structures that modern states employ to fund social programmes. It has shown that welfare state financing systems are far more varied than often assumed, with each instrument and structure having distinct implications for redistribution, programme coverage, and political sustainability. The choice between public and private financing, progressive or regressive tax structures, and centralized or decentralized administration all have far-reaching consequences for social policy outcomes. Moreover, the section underscores that financing is not merely about revenue collection. Rather, it fundamentally shapes the nature of welfare states, influencing everything from programme design to political support.

Given its significant causal impact on social policy, in the next section, we examine the drivers behind the evolution of tax and financing systems, focusing on how fiscal development and welfare state development are intertwined. This exploration will further illuminate the complex relationship between financing mechanisms and the broader landscape of social policy, providing insights into the historical and ongoing processes that shape social policies.

3.1 Contingent Events, Socio-economic Processes, and Fiscal Development

The transition from pre-modern financing methods to modern taxation marked a significant shift in fiscal capacity. In pre-modern times, states relied on narrow-based and non-tax revenues, such as plunder, tributes, tariffs, and various specific levies that yielded low tax revenues (Kiser & Karceski, Reference Kiser and Karceski2017; Morgan & Prasad, Reference Morgan and Prasad2009). However, starting in the late eighteenth century, the introduction of broad-based modern taxes, such as income taxes, corporate taxes, social insurance contributions, and VATs, marked a significant shift. These taxes, applied broadly and ad valorem, became dominant globally (Genschel & Seelkopf, Reference Genschel and Seelkopf2021). This transition, coupled with advancements in tax administration, led to a significant increase in revenues. For example, Scheve and Stasavage (Reference Scheve and Stasavage2016) note that tax revenues, which did not exceed 5 per cent of GDP in most countries by 1800, rose to around 9 per cent by 1900 and, in OECD countries, to around 20 per cent by the 1950s.

Social expenditures also increased dramatically as modern welfare states developed in Europe and in North America during the first part of the twentieth century and, especially, after World War II (Myles and Quadagno, Reference Myles and Quadagno2002). This is the case because welfare states expanded from minimal poor relief to comprehensive systems of social protection covering a wide range of socio-economic risks and needs. This parallel growth in fiscal capacity and social expenditures raises crucial questions about how these two phenomena are related and what explains their concurrent development.

The literature identifies four central drivers of fiscal development that help illuminate both the growth in state fiscal capacity and its key relationship with welfare state development: wars, recessions, modernization processes, and democratization. Some of the accounts suggest that enhanced fiscal capacity enabled welfare state growth, while others indicate that emerging welfare needs drove fiscal development.

Wars stand out as the first major driver of state fiscal capacity. Rooted in theories of state-building, Bellicist theories attribute the growth of fiscal capacity to the enormous financial demands of warfare (Levi, Reference Levi1989; Spencer, Reference Spencer1897; Tilly, Reference Tilly1992). Wars are costly, requiring states to expand their revenue-raising capacity (Obinger et al., Reference Obinger, Petersen, Starke, Obinger, Petersen and Starke2018). Furthermore, wars create a favourable environment for tax increases due to the “rally ‘round the flag” effect, which reduces public resistance to new and higher taxes (Feldman and Slemrod, Reference Feldman, Slemrod, Martin, Mehrotra and Prasad2009; Levi, Reference Levi1989; Scheve and Stasavage, Reference Scheve and Stasavage2016). In the twentieth century, both World War I and World War II fostered a massive expansion of state taxation power. In the United States, for example, World War II led to a drastic increase in tax rates, transforming the income tax from a “class tax” into a “mass tax” (Scheve and Stasavage, Reference Scheve and Stasavage2016).

Wars also spurred the development of tax administration. During World Wars I and II, innovations like third-party reporting and tax withholding were introduced in the United States, making tax collection more efficient (Kiser and Karceski, Reference Kiser and Karceski2017; Mehrotra, Reference Mehrotra2016). For example, during the world wars, the United States introduced two pivotal innovations: third-party reporting and withholding at the source. Third-party reporting required employers to inform tax authorities about workers’ incomes, while withholding at the source meant employers deducted taxes directly from pay cheques. These changes marked a significant shift from self-reporting and direct payment by individuals to an employer-mediated system. By enhancing efficiency and compliance, these innovations greatly improved revenue collection.

In Bellicist theories, the direction of causality goes from fiscal development to welfare state development and follows the observation that the enhanced taxation capacity created during wartime tends to persist after the war ends, enabling states to significantly increase welfare spending (Campbell, Reference Campbell1993; Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009; Morgan & Prasad, Reference Morgan and Prasad2009). This phenomenon termed the “displacement effect” by Peacock and Wiseman (Reference Peacock and Wiseman1961) refers to the maintenance of elevated tax levels after the war is over, allowing governments to use the “peace dividend” to finance expanded social and public policies.

Yet, the relationship between fiscal and welfare state development can also work in the opposite direction, when emerging social needs and extended demands for social protection pressure states to develop their fiscal capacity to meet these welfare needs. This is exemplified in the second major driver of fiscal development discussed in the literature: economic downturns (Katzenstein, Reference Katzenstein1985; Rodrik, Reference Rodrik1998; Truchlewski, Reference Truchlewski2020). Rising unemployment and business losses heighten societal demands for new or extended welfare state programmes, such as unemployment insurance, pressuring states to increase social spending. Simultaneously, economic contraction reduces revenues and access to capital, constraining state finances. As Papadia and Truchlewski (Reference Papadia, Truchlewski, Genschel and Seelkopf2021) show, this combination of increased welfare demands and reduced revenues compels governments to seek new revenue sources.

Modernization, the third major driver of fiscal development, offers another important explanation for the development of tax systems. Modernization theory has emphasized the roles of industrialization and urbanization for fiscal development, viewing the evolution of tax systems as driven by both emerging social needs and expanding economic opportunities. As societies industrialized, traditional family structures eroded and urban populations grew, giving rise to new social risks such as unemployment and old age poverty (Flora and Heidenheimer, Reference Flora and Heidenheimer1981; Montanari, Reference Montanari2001). These changes increased demand for social welfare and greater public expenditure, pushing states to enhance their revenue-generating capabilities (Wilensky, Reference Wilensky1974). Concurrently, economic modernization created more taxable wealth, allowing states to implement new and higher taxes without undermining the population’s basic needs (Hinrichs, Reference Hinrichs1966). Modernization processes affect fiscal capacities also by driving state administration capabilities. Wealthier societies are able to better staff and fund their tax administrations, thereby reduce the costs of tax compliance and enhance the efficiency of tax collection (Besley and Persson, Reference Besley, Persson, Auerbach, Chetty, Feldstein and Saez2013).

The fourth major driver of fiscal development identified in the existing scholarship is democratization. The connection between democracy and taxation is rooted in fiscal sociology literature and the fiscal contract approach, which we further describe in Section 3.3. According to this perspective, tax capacity increases when taxation is based on consent rather than just enforcement. This consent is achieved through fiscal contracts where, in exchange for tax revenues, rulers provide their citizens with public goods. Democracy affects tax capacity in two ways: first, democratic institutions lower the transaction costs of creating and enforcing fiscal contracts between taxpayers and rulers (Kiser & Karceski, Reference Kiser and Karceski2017; Levi, Reference Levi1989); second, they are argued to cultivate taxpayer consent to paying taxes (Einhorn, Reference Einhorn, Martin, Mehrotra and Prasad2009).

Interestingly, while fiscal sociology broadly views democracy as enhancing states’ fiscal capacities by providing a framework for more effective tax collection, it also sees democracy as a consequence of the state’s fiscal imperatives. Tilly (Reference Tilly2009), one of the most prominent figures of fiscal sociology, explains that democracy emerged in Europe when states shifted from directly controlling production to relying on taxes from private economic actors for revenue. These actors, in turn, demanded political rights to influence how their taxes were spent, initiating a bargaining process that eventually led to the establishment of representative institutions. Although in the nineteenth century political representation was initially limited to wealthy elites, this process ultimately paved the way for the expansion of the franchise and the attainment of universal suffrage (Tilly, Reference Tilly2009). In fact, relying on oil revenues rather than taxes to fund relatively generous expenditures allows governments of “Petro-states” to stay in power longer, whether they are autocracies or democracies, as they do not have to be accountable to the public regarding the levels of taxes necessary for the public services they receive (Morrison, Reference Morrison2015).

In brief, the relationship between fiscal capacity and welfare state development is complex, with causality flowing in both directions. While welfare needs can drive fiscal evolution, fiscal capacity may also develop independently. Regardless, increased revenue generation expands a government’s spending potential and its ability to influence the welfare state. This section has also highlighted the significant impact of contingent events and large-scale economic processes on the adoption of modern taxes and collection systems.

3.2 Structure, Politics, Institutions, Ideas, and Fiscal Development

Equally important are the mechanisms behind these developments, which can be understood through structural conditions, political struggles, institutions, and shifting ideational realms. In Section 3.3, we will critically review how these perspectives offer distinct insights into both the historical formation and current challenges of fiscal and welfare systems, highlighting their limits and the added value of integrating them for a more comprehensive understanding of fiscal development.

Structural explanations focus on the deep, enduring social and economic frameworks within a society that shape the conditions in which individuals and institutions operate. At the economic level, structures refer to how wealth is produced and distributed. At the social level, they encompass systems of social relations, such as class structures or racial and ethnic hierarchies. Demographic structures involve patterns of population growth, age distribution, migration, and urbanization, while geopolitical structures relate to the degree of economic internationalization and global integration.

Central to structural explanations is the idea that these frameworks – or the large-scale trends that change them – are crucial for understanding why societies develop in specific directions. Modernization theory, for example, is a classic structural explanation for the development of modern tax structures and welfare states in the late nineteenth and twentieth centuries. It is considered a structural explanation, because it focuses on how the transition from agriculture to industry, from rural to urban societies, and from feudal to wage-based economies created the foundational conditions for the emergence of modern taxes and the welfare state.

Changing structural conditions are key to understanding recent challenges to states’ fiscal and welfare systems. For example, aging populations, a common demographic trend in developed countries, pose significant challenges to welfare state financing by altering the dependency ratio – the number of beneficiaries compared to those funding the system (Kangas and Palme, Reference Kangas, Palme, Clasen and Siegel2007; Morel and Palme, Reference Morel, Palme and Greve2018). Similarly transformative are the effects of economic internationalization, which increases the mobility of people, business, and capital. This enhanced mobility constrains countries’ tax policy choices and their capacity to maintain revenue levels while pursuing redistributive goals (Genschel, Reference Genschel2002; Genschel and Schwarz, Reference Genschel and Schwarz2011; Steinmo, Reference Steinmo1994).

While structural explanations provide valuable insights into how large-scale changes in society, such as economic shifts or demographic trends, shape fiscal capacity and welfare systems, they can sometimes become overly functional and deterministic. They risk suggesting that specific outcomes are inevitable based solely on these underlying structural trends. This tendency is evident in both early modernization theories and more recent globalization theories (Montanari, Reference Montanari2001; Myles & Quadagno, Reference Myles and Quadagno2002).

This deterministic tendency was particularly pronounced in early modernization theories, which went beyond seeing industrialization as merely influencing fiscal and welfare development. Instead, it portrayed modern tax structures and expansive welfare states as the inevitable destiny of industrializing societies. According to this view, the transition from agricultural to industrial economies would automatically trigger a universal process of fiscal and welfare expansion, with all modernizing countries eventually converging on similar systems regardless of their distinct historical, cultural, or political contexts (Wilensky, Reference Wilensky1974).

Similarly, deterministic in its predictions, though pointing in the opposite direction, globalization theory of the 1990s foresaw an inevitable “race to the bottom.” As Genschel (Reference Genschel2002) shows, these theories predicted that increasing economic integration and capital mobility would inexorably force states to lower taxes to remain competitive, eroding their fiscal capacity and leading to welfare state retrenchment as governments could no longer sustain social protections (OECD, 1998; Sinn, Reference Sinn1992).

The stark contrast between these structural theories – with modernization predicting welfare state expansion and globalization anticipating welfare state decline – highlights the limitations of purely structural and functional explanations. Rather than converging towards the same outcomes, countries have responded to industrialization and globalization pressures in remarkably diverse ways. For example, in 2022, the average tax revenue and public social spending in OECD countries were estimated to be 34 per cent and 21 per cent of GDP, respectively. Yet, countries like France collected taxes at a ratio of 46 per cent of GDP and spent 31.6 per cent on public social expenditures, while the United States collected only 27.7 per cent of GDP in taxes and allocated 22.1 per cent to public social spending (OECD, 2023, 2024). In fact, while corporate tax rates are certainly influenced by tax competition, other forms of tax revenues, which represent much larger shares of state revenues, are not influenced in the same way (Genschel and Schwarz, Reference Genschel and Schwarz2011).

This significant divergence suggests that structural factors alone are insufficient to explain the evolution of welfare systems and fiscal capacity. Political and institutional factors play a crucial role in shaping fiscal regimes and welfare state development, making them central to understanding why nations follow different paths.

Political and institutional explanations offer a more nuanced understanding of the development of fiscal capacity and welfare state systems than purely structural accounts. Indeed, economic development, in itself, does not lead to a predetermined form of taxation or social protection. Instead, political conflicts, power dynamics, and institutional settings shape a variety of tax regimes and financing models in the modern world (Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009).

Political explanations emphasize the power struggles among labour, capital, and various societal groups, illustrating how their conflicting interests shape tax and social policies. Tax policies are particularly prone to political conflict due to the structure and composition of taxation, which can target distinct economic activities, actors, and classes. For instance, a progressive income tax imposes higher rates on wealthier individuals, facilitating income redistribution. In contrast, a consumption tax, such as a VAT, disproportionately burdens lower-income households, as they allocate a larger share of their income to taxable goods. Additionally, corporate taxes target profits, VAT taxes consumption, and PIT, along with social insurance contributions, tax labour. This complexity can lead to both class-based and sector-based struggles over taxation, reflecting broader redistributive conflicts (Campbell, Reference Campbell1993; Haffert, Reference Haffert2021a).

These diverse political dynamics are apparent in historical examples. In the United States, the push for progressive income taxation was rooted in the broader political currents of the late nineteenth and early twentieth centuries. A broad-based populist movement emerged in response to growing inequality and the concentration of wealth among a few monopolistic capitalists. Populists sought to regulate monopoly capital and address the widening wealth divide, viewing progressive income taxation as a key mechanism to achieve these goals. As Morgan and Prasad (Reference Morgan and Prasad2009) explain, the income tax was incorporated into this broader populist agenda as a tool to discipline capitalists. By shifting the tax burden from consumers paying tariffs to those with high incomes, the income tax became a method for redistributing wealth and addressing the power imbalances between labour and capital.

The introduction of VAT in Europe illustrates redistributive struggles between both sectors and social classes. As Haffert and Schultz (Reference Haffert and Schulz2020) note, the initial struggles over the adoption of VAT in the European Economic Community in the 1960s were largely sectoral rather than class based. Competitive export industries supported VAT as a standardized European consumption tax that would reduce trade barriers, whereas domestic industries opposed it, fearing increased exposure to foreign competition.

These political struggles, however, are embedded within the institutional context in which they take place, and this context shapes both the possibilities and constraints that actors face. Institutions, comprising formal rules, norms, and procedures, play a crucial role in shaping fiscal policies by structuring these political conflicts. They influence the incentives actors respond to and limit the range of policy options available to them. Political regimes, electoral systems, and systems of interest representation all impact the development of tax policies.

For example, democratic institutions, relying on majority rule, were long argued to be central to the development of progressive taxation. This argument was formalized in the well-known median voter model, popularized by Meltzer and Richard (Reference Meltzer and Richard1981), which suggests that democratic institutions empower those who benefit from redistribution. This is because there are generally more individuals on the lower end of the income distribution than on the upper end (i.e., more workers than employers, and more income earners than capital owners) creating an incentive to shift the tax burden onto capital and high-income earners. While the empirical evidence supporting this model for the latter half of the twentieth century is debatable (Iversen & Soskice, Reference Iversen and Soskice2006; Timmons, Reference Timmons2010), it does align with the situation prevailing during the early twentieth century, when democratization was positively correlated with the adoption of progressive taxes (Genschel and Seelkopf Reference Genschel and Seelkopf2021).

In recent decades, however, data suggest that differences in electoral systems – such as majoritarian versus proportional systems – and organized interest representation institutions like corporatism, discussed further in Section 4, play a more decisive role than democracy itself in shaping tax levels and structures (Iversen & Soskice, Reference Iversen and Soskice2006; Steinmo, Reference Steinmo1993). This institutional variation is particularly evident in how different countries responded to global tax competition. While structural accounts predicted that this competition would result in the defunding of mature welfare states, the actual outcomes varied significantly based on the electoral system and its associated institutional constrains (Hays, Reference Hays2009). For example, under the Conservative majority governments of Margaret Thatcher in Britain (1979–90), the majoritarian system required the approval of just one veto player: the governing party. This allowed Prime Minister Thatcher to implement much deeper cuts compared to Helmut Kohl’s Christian-Democratic Union government in Germany, where the proportional system with multiple veto players compelled the chancellor to compromise with coalition partners. This difference in institutional structure illustrates a broader pattern: similar structural pressures, when filtered through varying institutional contexts, can generate different political dynamics and lead to distinct fiscal outcomes (Hallerberg and Basinger, Reference Hallerberg and Basinger1998).

One of the central contributions of institutional explanations is the analytical toolbox they offer, particularly concepts like critical junctures and path dependence. These temporal concepts help explain why different countries embark on distinct trajectories in tax and welfare policy. A critical juncture refers to a period of significant change that sets countries on specific paths, which continue to shape their development long after the initial causes have disappeared (Collier and Collier, Reference Collier and Collier1991). Path dependence theory builds on this, emphasizing how institutional choices made at these critical junctures become entrenched over time, leading to the persistence of particular policy frameworks (Pierson, Reference Pierson2000).

The development of tax systems in the United States and France provides compelling examples of how critical junctures can lead to divergent fiscal paths (Morgan & Prasad Reference Morgan and Prasad2009). Both countries introduced progressive income taxes in the early twentieth century, following heated political debates in the 1890s. However, despite these similarities, their tax systems diverged significantly during and after World War I. In the United States, the income tax became a robust and flexible tool for revenue generation, shaping future fiscal policies and even withstanding conservative attempts to introduce national sales taxes in the 1920s. In contrast, France faced difficulties with income tax collection and consequently shifted towards a national sales tax, which eventually evolved into the VAT. This early divergence in tax policy has had lasting effects, with the United States relying more heavily on progressive taxation and France adopting a mix of regressive and indirect taxes partly to support its welfare state (Morgan & Prasad, Reference Morgan and Prasad2009).

These historical differences illustrate how institutional paths set during critical junctures continue to shape fiscal and welfare politics today. For instance, the central role of progressive income taxes in the United States has entrenched the country in ongoing political conflicts over revenue generation, while France’s VAT has become a significant source of funding for its welfare system. These distinct trajectories highlight the critical role of political institutions and decision-making processes at key moments in determining long-term public finance outcomes.

Political and institutional explanations offer valuable insights into the development of tax systems and welfare states, providing essential tools for understanding why different nations follow distinct fiscal trajectories. While Section 4 of this Cambridge Element explores how these dynamics have evolved into comparative differences in countries’ tax structures and trajectories, we must also consider the role of shifting ideas and policy paradigms. This ideational perspective, along with the reciprocal relationship of the fiscal/social contract examined in Section 3.3, provides the necessary foundation for understanding modern welfare financing.

While much of the literature on fiscal development focuses on material interests and institutional constraints, a growing body of research highlights the significance of ideational factors in shaping tax and welfare financing policies. The ideological divide between left and right – where the former generally supports higher and more progressive taxation, while the latter advocates for lower and more neutral taxes – remains a central assumption in studies of taxation. Despite a few exceptions (e.g., Béland and Waddan, Reference Béland and Waddan2015; Steinmo, Reference Steinmo2003), however, there is relatively little systematic analysis of fiscal developments through an ideational lens, especially as they relate to social policy. Still, existing evidence suggest that shifts in ideas – both about justice and fairness and about the economic consequences of different financing methods – can profoundly affect the trajectory of fiscal and welfare state development.

Policymakers’ view on taxation shifted significantly around the time of the Tax Reform Act of 1986, a cornerstone of Reagan-era fiscal policy, which fundamentally altered prevailing notions of what kinds of tax reforms were seen as possible and desirable (Steinmo, Reference Steinmo2003). Prior to this, taxation in the United States was widely seen as both a redistributive and economic policy tool, marked by steeply progressive rates – particularly on “unearned” income, such as capital gains and generous exemptions and deductions intended to steer investments into specific economic sectors. During the 1980s, however, elite views increasingly favoured lower marginal rates, a broader tax base with fewer loopholes, and preferential treatment of capital income (Steinmo, Reference Steinmo2003).

Elite perceptions of SICs also shifted significantly in the 1970s (Campbell and Morgan, Reference Morgan2005). Historically, SICs embodied a strong logic of justice and fairness grounded in equivalence and reciprocity (Hinrichs, Reference Hinrichs1997; Sachweh, Reference Sachweh, Sabbagh and Schmitt2016). Although public opinion polls consistently showed a stronger support for SICs than for other tax instruments, policymakers from both the left and the right increasingly questioned them – left-wing critics argued that SICs disproportionately burdened low-income earners, while right-wing critics blamed them for driving up welfare expenditures. These shifting perceptions contributed to a significant curtailment of SICs as a primary funding source.

Similarly, perceptions of the economic impact of SICs evolved over time. Throughout much of the twentieth century, SICs were viewed as economically significant not only for their role in social welfare but also for macroeconomic reasons. In the 1940s, Keynesian economists like Harvard’s Seymour Harris praised the unfunded character of the new Social Security programme for its ability to stimulate aggregate demand (Harris, Reference Harris1941). Additionally, as a broad-based tax levied from the first unit of income, SICs were seen as less distortionary than progressive income taxes (Berglass, Reference Berglas1969). These contributions were considered less likely to discourage work and investment compared to the more progressive income taxes. However, by the 1980s, critiques of SICs began to emerge, raising concerns about their economic impact on employment and competitiveness (ILO, 1984; Palier, Reference Palier2010). In Europe, these shifting perceptions played a key role in reshaping the tax mix of Bismarckian welfare states: while SICs expanded significantly until the 1990s, their relative importance has since declined (Manow, Reference Manow and Palier2010; Palier, Reference Palier2010).

These examples illustrate how evolving ideas about taxation and financing mechanisms – whether framed in terms of fairness or economic efficiency – can shape fiscal policies in ways that are not reducible to partisan politics or economic pressures alone. An ideational perspective on fiscal development thus offers a valuable complement to existing institutional and interest-based analyses, highlighting how changing perceptions and policy paradigms influence long-term welfare financing trajectories.

The four perspectives – structural, political, institutional, and ideational – discussed in Section 3.2 offer complementary yet distinct insights into fiscal development. Structural explanations provide essential context for understanding the background conditions that shape fiscal policy shifts, such as economic structures, demographic trends, and globalization. However, they often overlook the agency of actors and the role of politics. Political perspectives, in contrast, focus on power dynamics, conflicts, and alliances among different social groups, emphasizing how these struggles shape fiscal policies. Yet, political explanations can be limited by their failure to account for institutional constraints that structure these conflicts. Institutional theories highlight how formal rules, norms, and procedures shape policy outcomes by structuring political conflicts and limiting available choices. These insights are crucial, yet they can overlook dynamic processes that change over time, offering an overly static vision. Finally, ideational explanations emphasize the role of ideas and policy learning in shaping policy possibilities, pointing to how cognitive structures influence policy development. While each perspective contributes valuable insights, integrating them allows for a more nuanced understanding of fiscal policy evolution, illustrating how structural conditions, political power, institutional frameworks, and changing ideas intersect to shape fiscal systems in complex and historically contingent ways.

3.3 The Fiscal/Social Contract

In recent years, social policy scholars have increasingly engaged with the fiscal aspects of welfare states, recognizing their importance in understanding welfare state development, expansion, and current pressures for retrenchment and restructuring (Brownlee, Reference Brownlee2003; Pierson, Reference Pierson and Pierson2001; Steuerle, Reference Steuerle and Brownlee1996). This approach has evolved from a more functional perspective, which assumed that the forces driving social spending automatically adjusted financing levels, to a more nuanced view that considers contingency, structure, politics, and institutions as central to explaining countries’ fiscal capacities (Haffert, Reference Haffert2021a; Koreh & Béland, Reference Koreh and Béland2017).

However, as evident in the preceding review, much of this literature has framed the relationship between fiscal development and welfare state development primarily in terms of revenue availability and its effects on social spending. These accounts can be summarized as follows: when structural, political, and institutional factors, along with contingent events, create resource abundance, they facilitate social spending and welfare state expansion. Conversely, when these factors lead to resource scarcity, they drive expenditure constraints and welfare state restructuring.

Yet, the focus on revenue availability risks oversimplifying social policy by suggesting that other key elements, such as allocation logic or risk coverage, can be studied independently of fiscal policies and politics. This separation, often observed in the academic treatment of fiscal and social policy, overlooks the inherent interconnections between them (Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009; Timmons, Reference Timmons2005).

A more promising approach emerges when social policy is analysed through the lens of the fiscal contract, which frames taxation and social spending as intertwined processes (Koreh & Béland, Reference Koreh and Béland2017; Morel & Palme, Reference Morel, Palme and Greve2018; Schmitt et al., Reference Schmitt, Lierse and Obinger2020). This perspective, increasingly embraced in modern tax theories (Kiser & Karceski, Reference Kiser and Karceski2017; Martin et al., Reference Martin, Mehrotra, Prasad, Martin, Mehrotra and Prasad2009), has gained traction among social policy scholars. It offers new insights into both historical and contemporary social policy dynamics (Béland and Koreh, Reference Béland and Koreh2019; Koreh and Béland, Reference Koreh and Béland2017). Importantly, it emphasizes that key aspects of social policy, such as the creation of new programmes, allocation logics, redistributive structures and restructuring reforms, can only fully be understood when taxation, and spending politics are considered in tandem (Manow, Reference Manow2004; Martin and Prasad, Reference Martin and Prasad2014; Morel and Palme, Reference Morel, Palme and Greve2018; Steinmo, Reference Steinmo1993).

The fiscal contract approach views taxation not merely as a revenue extraction process but as a negotiation between states and citizens where tax compliance is exchanged for public goods (Levi, Reference Levi1989; Schumpeter, Reference Schumpeter and Swedberg1918/Reference Schumpeter and Swedberg1991; Tilly, Reference Tilly1992). While rulers historically relied on coercion to extract revenue, this method proved costly and risky, leading to conflict, tax evasion, and a loss of legitimacy (Brautigam et al., Reference Brautigam, Fjeldstad and Moore2008). Instead, states often engage in a form of quasi-consensual taxation (Levi, Reference Levi1989) where the provision of collective goods legitimizes their authority and enables increased access to public resources (Brautigam et al., Reference Brautigam, Fjeldstad and Moore2008; Levi, Reference Levi1989; Timmons, Reference Timmons2005).

The concept of taxation as a fiscal contract traces its roots to the founders of fiscal sociology, who portrayed taxation as a tangible social contract – the result of a historic bargain between rulers and the ruled forged in specific times and places (Schumpeter, Reference Schumpeter and Swedberg1918). These scholars linked the shift from coercion to quasi-consensual taxation to the rise of the tax state. As rulers’ personal resources became insufficient to sustain wars, they increasingly depended on taxation, pushing them to forge fiscal contracts with societal elites. Historical instances, such as medieval England’s Parliament forcing the monarchy to negotiate with taxpayers, resulted in increased revenues (Kiser and Barzel, Reference Kiser and Barzel1991). The Glorious Revolution of 1688 also demonstrates this dynamic, as the wealthy agreed to provide the monarchy with revenue in exchange for property and political rights (North and Weingast, Reference North and Weingast1989).

Fiscal sociology underscores the centrality of the tax state and fiscal contract to the development of the modern state. The need to legitimize revenue extraction prompted states to provide more public goods, leading to the creation of modern institutions, like tax collection bureaucracies and political representation systems, which improved revenue collection and enhanced the credibility of state commitments (Kiser and Karceski, Reference Kiser and Karceski2017; North, Reference North1993). More recently, scholars have extended this perspective to welfare states, suggesting that social rights and welfare benefits have become integral to the fiscal contract (Morel & Palme, Reference Morel, Palme and Greve2018; Schmitt et al., Reference Schmitt, Lierse and Obinger2020; Timmons, Reference Timmons2005).

However, as Koreh (Reference Koreh2017) points out, applying the fiscal contract approach to the study of social policy requires careful adaptation. A review of the literature using this perspective explicitly or implicitly indicates that three adjustments are necessary to fully harness its potential: (1) recognizing the distinct exchange opportunities provided by different financing methods, (2) broadening the notion of the contract to include exchanges between societal groups and the institutional context within they take place, and (3) empirically examining the motivations and dynamics underlying the contract and how it evolves over time.

First, it is essential to distinguish between the various types of exchanges that different financing instruments offer. Earlier studies primarily focused on taxation, often overlooking the significant differences between tax instruments (see, for example, North & Weingast, Reference North and Weingast1989). Yet, in social policy, whether programmes are financed through general taxation, social insurance contributions, or private financing greatly influences the nature of the exchanges that can take place.

For instance, as discussed in Section 2, SICs that account for an average of 44 per cent of total public social expenditure in OECD countries (Koreh et al., Reference Koreh, Zemmour and Palier2024, authors’ calculations based on OECD 2022 data) create a direct link between payments and benefits. This system generates specific interests and opportunities for exchange, distinct from general taxation. In social insurance-based systems, the principle of equivalence (“we are entitled because we pay”) prevails. It entails a direct relationship between contributions and entitlements, which means that only those who pay into the system, typically through employment, benefit from it. In contrast, general taxation’s unrequited nature implies no direct alignment between contributors and beneficiaries. This difference affects the scope of exchange opportunities and has significant implications for welfare coverage and allocation logic.

Moreover, tax and financing instruments also differ in the types of revenue opportunities they create. While tax-based social policies primarily impose spending demands on government budgets, semi- or fully funded social insurance programmes (as well as private funded programmes) can provide governments with access to substantial capital reserves that can address various fiscal and economic needs (Estevez-Abe, Reference Estevez-Abe, Ebbinghaus and Manow2001; Kangas, Reference Kangas, Hujo and McClanahan2009; Musgrave and Musgrave, Reference Musgrave and Musgrave1973). Koreh and Béland (Reference Koreh and Béland2017) demonstrate how this characteristic of social insurance played a crucial role in the adoption of key social insurance schemes, as these programmes provided governments with much-needed capital during times of resource scarcity.

Two prime examples of this dynamic are the social insurance programmes introduced by Otto von Bismarck in nineteenth-century Germany and the US federal old age insurance (Social Security) programme established during the presidency of Franklin D. Roosevelt in 1935. While both initiatives were designed to address social needs, such as income maintenance during sickness or old age, their adoption was also driven by fiscal imperatives, as they offered the state access to new revenue streams.

In the German case, Philip Manow (Reference Manow2004) emphasizes that, beyond the widely recognized motivation of Bismarck’s social insurance schemes as a countermeasure to the growing labour movement, there was a crucial fiscal and political objective to consolidate the central federal state vis-à-vis the German Länder (provinces) and establish autonomous sources of revenue. At the time, the German federal government had limited taxation authority because the 1871 constitution assigned tax collection powers and veto rights over central government tax increases to the Länder. Bismarck saw the new social insurance programmes as a tool not only to legitimize the creation of new revenue streams but also to remove them from the Länder’s control. This allowed the federal state to commit to social protection while gaining a degree of financial independence from the provinces.

Similarly, Leff (Reference Leff1983) identifies a comparable dynamic behind the adoption of the US federal old age insurance (Social Security) system in 1935. In response to the Great Depression, growing federal involvement in the economy during Roosevelt’s New Deal led to increased spending and mounting national debt. By implementing old age insurance on a fully funded basis and investing its surpluses in federal bonds, Roosevelt simultaneously addressed public demands for greater economic security and created a reliable, cost-effective source of funding for the federal debt (Jacobs, Reference Jacobs2011).

These examples highlight not only the unique exchange opportunities provided by social insurance but also how the state’s need for revenue can drive the creation of new social programmes. While this might seem exceptional, recent research suggests that this dynamic is more prevalent than typically recognized in social policy literature. Genschel and Seelkopf’s (Reference Genschel and Seelkopf2021) comprehensive study on the global spread of modern taxes supports this, revealing that during economic downturns, countries are less likely to introduce mass tax instruments, like income tax or VAT, but more likely to adopt SICs. This aligns with the fiscal contract logic that argues, when states urgently need revenue and taxpayers resist regular tax hikes, governments often offer tangible benefits, like social insurance, in exchange for greater revenues.

A second key adjustment to the fiscal contract framework concerns both the actors involved and the institutional context shaping these exchanges. Schumpeter (Reference Schumpeter and Swedberg1918) and Goldscheid (Reference Goldscheid, Musgrave and Peacock1917/Reference Goldscheid, Musgrave and Peacock1958), the founders of fiscal sociology, initially conceptualized the fiscal contract as a negotiation between rulers and elites in pre-modern, pre-bureaucratic societies. However, with the expansion of the franchise, the rise of democratic institutions, and the development of complex state bureaucracies, this framework must now account for exchanges, not only between the state and citizens but also among various social groups. Furthermore, both state and welfare state institutions must be seen as shaping and being shaped by the fiscal contract, a dynamic with profound implications for welfare state development.

This expanded view recognizes that modern democratic institutions like electoral systems and mechanisms of interest representation not only allow more actors (e.g., political parties, trade unions, and employers’ associations) to participate in the fiscal contract but also enable bargaining and exchanges among them. It also acknowledges that the structure of state institutions shapes these exchanges and affects the range of trade-offs possible.

For example, scholars have pointed out that countries with proportional electoral systems and/or corporatist decision-making frameworks, like Sweden or Germany, are more conducive to consensual policymaking and thus facilitate fiscal bargains between labour and capital. In these systems, labour often concedes to regressive taxation, like payroll and consumption taxes, in exchange for more robust social protections and broader welfare coverage. In contrast, majoritarian systems, such as those in the United States or the United Kingdom, have weaker institutions for consensual policymaking, leading to more intense class conflict. As a result, these countries tend to have more progressive tax systems, but significantly lower social spending (Beramendi & Rueda, Reference Beramendi and Rueda2007; Hertel-Fernandez & Martin, Reference Hertel-Fernandez, Martin, Huerlimann, Brownlee and Ide2018; Steinmo, Reference Steinmo1993).

While these scholars did not explicitly use the term “fiscal contract,” their focus on how tax and spending politics interact allowed them to explain redistributive patterns across countries that traditional theories struggled with. For instance, key class-based approaches, like the median voter model and power resource theory, suggest that strong labour mobilization should lead to both expanded welfare systems and more progressive taxation (Haffert & Mertens, Reference Haffert and Mertens2021; Meltzer & Richard, Reference Meltzer and Richard1981). However, the reality that the largest welfare states rely heavily on regressive payroll and consumption taxes, which disproportionately affect lower-income groups, while countries with smaller welfare states tend to use more progressive taxation, remains puzzling under this framework (Beramendi & Rueda, Reference Beramendi and Rueda2007; Haffert & Mertens, Reference Haffert and Mertens2021; I. Martin & Prasad, Reference Martin and Prasad2014). We come back to this relationship between tax progressivity and levels of expenditures in Section 4.

The fiscal contract perspective, by accounting for the interplay between taxation and social spending, offers a more nuanced understanding of redistributive outcomes. It also highlights that while general taxation, such as income or corporate tax, may not generate the same exchange opportunities as SICs, it still allows for cross-class and cross-party concessions that shape both tax and welfare policies.

While state institutions, such as electoral systems, shape welfare politics by facilitating certain types of exchanges, welfare politics and policies are also influenced by the welfare state institutions that governments establish to make fiscal and social bargains credible. As Koreh and Mandelkern (Reference Koreh and Mandelkern2023) explain, the fiscal/social contract has an inherent “time inconsistency problem,” where citizens are expected to pay taxes or SICs now, while social benefits, such as unemployment insurance or old age benefits, are realized later. As these authors show, key welfare state institutions like entitlements, indexation mechanisms, and independent social security bodies were originally designed to enhance credibility and trust in state commitments as means to ensure citizens consent to pay into this system. By ensuring the value of future benefits, limiting the discretion of elected officials and delegating decision-making power to social partners and social security bodies, governments signal their long-term commitment to safeguarding citizens’ social rights.

The design of these credibility-enhancing institutions mirrors historical fiscal contracts described by North (Reference North1993), where time inconsistency also plagued rulers. During wartime, rulers often borrowed money to finance conflicts but had little incentive to repay those loans afterward. This led to demands for political representation as a mechanism to ensure credibility. By granting political representation to those who lent money, rulers could signal their commitment to repay debts and secure future borrowing capacity.

However, like political representation institutions, the influence of social credibility institutions extends beyond merely ensuring trust and payment compliance. Entitlement programmes, for example, create constituencies with a vested interest in maintaining and expanding these programmes. Likewise, delegating decision-making authority to actors, such as social partners and semi-autonomous welfare agencies, generates veto points, creating multiple avenues for these groups to influence and shape social policies over time (Immergut, Reference Immergut1992; Pierson, Reference Pierson2000).

Equally important, these institutions also feed back into tax and financing policies. For example, in an aging society where the population eligible for benefits, like pensions, increases, entitlement-based systems – anchored in legislation with clear eligibility and benefit guidelines – trigger automatic spending increases. This puts pressure on governments to adjust fiscal policies to accommodate rising costs. In contrast, in systems without guaranteed entitlements, where budgets are predetermined, increased demand is generally addressed by reduced assistance levels rather than financing (King, Reference King2000; Koreh et al., Reference Koreh, Zemmour and Palier2024; White, Reference White2015).

While institutions, particularly credibility-enhancing institutions, are often associated with self-reinforcing feedback that supports continuity overtime, they can also generate pressures for change through self-undermining effects.Footnote ³ As the example of demographic aging suggests, institutional arrangements may compel governments and societal actors to renegotiate existing fiscal contracts. This observation points to a critical yet often overlooked aspect: the need to empirically examine the underlying motivations driving different actors to forge, maintain, or renegotiate these contracts. This brings us to the third key adjustment to the fiscal contract framework.

While classical fiscal sociology often frames the state’s primary motivation as revenue maximization and societal interests as focused on either receiving government services or securing political and property rights (Levi, Reference Levi1989; Schumpeter, Reference Schumpeter and Swedberg1918; Tilly, Reference Tilly1992), this simplistic attribution of motivations becomes insufficient when considering the variety of financing instruments available and the different types of exchanges they create. Viewing states solely as revenue maximizers highlights the importance of revenue for state survival but fails to explain why some states rely heavily on private funding or pursue policies like tax cuts and the privatization of social benefits. Similarly, assuming that societal interests are limited to what the state provides through spending overlooks broader motivations, such as how social funds are invested and circulated within the economy.

When applying the fiscal contract framework to social policy, it is therefore crucial to consider not only public but also private modes of finance and to empirically examine, rather than pre-attribute, the motivations of the parties involved. In fact, various scholars have shown that state involvement in establishing and regulating pension systems in the twentieth century was closely tied to state-building objectives (Béland & Koreh, Reference Béland and Koreh2019; Kangas, Reference Kangas, Hujo and McClanahan2009). Importantly, while some countries achieved these objectives through public financing, others relied on private financing methods, illustrating that states can pursue their goals through various funding approaches beyond tax revenue maximization.

Canada’s province of Quebec and Japan exemplify these contrasting approaches to achieving similar state-building objectives. In the mid-1960s, Quebec’s state- and nation-building efforts led it to establish a separate contributory public pension plan (the Quebec Pension Plan) alongside the programme covering people in the rest of Canada (the federal Canada Pension Plan) allowing the province, through a newly established public investment board (Caisse de dépôt et placement du Québec), to invest pension fund surpluses to drive provincial economic development (Béland and Koreh, Reference Koreh2017). In contrast, post–World War II Japan implemented a private pension system that was crucial for state-building by accumulating large reserves of patient capital. These highly regulated funds were directed towards public infrastructure and industrial development, helping Japan rise to global prominence (Estevez-Abe, Reference Estevez-Abe, Ebbinghaus and Manow2001; Park, Reference Park2011). These cases also suggest that business consent to co-finance workers’ pensions rested not only on labour relations but also on their potential benefits from investing these funds.

Recognizing that both states and key societal actors can have vested interests in private financing modes is particularly important to understand the evolution of the fiscal and social contract, especially in recent decades. From a revenue maximization perspective, the wave of pension reforms across OECD countries in the 1990s might seem puzzling, as these reforms generally involved reductions in public benefits alongside the expansion of private, pre-funded pension schemes (Ebbinghaus, Reference Ebbinghaus2011; Hassel et al., Reference Hassel, Naczyk and Wiß2019; Orenstein, Reference Orenstein, Kott and Droux2013). However, as a growing body of literature on financialization suggests, when states perceive financial markets as critical growth drivers or as sources for financing public debt, they may have vested interests in pension privatization as a way to create and strengthen financial markets (Fastenrath et al., Reference Fastenrath, Schwan and Trampusch2017; Maman, Reference Maman2021; Trampusch, Reference Trampusch2018). Furthermore, the growth of finance has empowered financial actors like insurance companies, whose interests in private pension provision have been argued to be a major driver of pension privatization reforms (Kemmerling and Neugart, Reference Kemmerling and Neugart2009; Naczyk and Palier, Reference Naczyk and Palier2014), altering the fiscal contract by increasingly tying pensions to financial markets and their performance.

3.4 Conclusion

To summarize, this section has explored the complex relationship between the development of fiscal systems and welfare states to examine the deep connections between taxation and social spending. Through an analysis of contingent events, socio-economic processes, structural conditions, and political-institutional factors, the section has demonstrated how the evolution of tax systems has both enabled and been shaped by welfare state development. It has also highlighted how different financing methods, from general taxation to SICs and private funding, create distinct exchange opportunities and shape social policy outcomes in various ways.

This section explains why the fiscal contract framework is a valuable analytical tool for social policy scholars. By focusing on the mutually causal relationships between taxation, finance, and welfare programmes, this perspective offers more comprehensive accounts of welfare state development and restructuring than traditional approaches. For scholars of social policy, the fiscal contract framework enhances our ability to understand multiple dimensions of welfare state politics. The framework helps explain the adoption timing of different programmes, illuminates why certain coverage and redistributive features are chosen over others, and provides insights into processes of both continuity and change in welfare states. By taking the tax and finance side seriously, this analytical approach broadens our understanding of welfare policies and politics, offering more intersecting explanations for empirical puzzles that have challenged conventional theoretical frameworks. As the section has shown, viewing taxation and social spending as fundamentally intertwined processes, with welfare benefits integral to the modern fiscal contract between states and citizens, provides a more nuanced and complete picture of how welfare states develop, adapt, and transform over time.

4.1 The Comparative Impact of Tax Systems on Welfare State Variation

Generous welfare states tend to exhibit significantly higher tax levels. Countries within Esping-Andersen’s Conservative–Continental and Social Democratic–Nordic clusters – what Jonas Pontusson (Reference Pontusson2005) calls “Social Europe” – have average tax-to-GDP ratios around 42.2 per cent, compared to 28.6 per cent in the Anglo-American (Liberal) cluster.Footnote ⁴ These differences in fiscal effort are reflected in public social expenditure levels: 26.5 per cent of GDP in Social Europe versus 18.9 per cent in Liberal America.

This variation seems to support a standard assumption: governments tax more because they choose to spend more. However, research increasingly challenges this view. It suggests that tax structures do not just mirror spending decisions – they also shape them. The instruments used to raise revenue significantly affect a state’s capacity to support expansive welfare systems.

Empirical comparisons support this. The European cluster relies more heavily on consumption taxes (e.g., VAT) and SICs. Consumption taxes account for 12.1 per cent of GDP in Europe, compared to 8.2 per cent in the Anglo-American cluster. Social insurance contributions are even more divergent: 12.6 per cent in Europe (13.9 per cent excluding DenmarkFootnote ⁵) compared to just 3.9 per cent in the Anglo-American group. They are even higher in countries that rely on Social Health Insurance systems (Austria, Belgium, France, and Germany), in which health expenditures are largely funded by social insurance contributions. Income taxes, by contrast, are similar across clusters as a share of GDP, but in Liberal regimes, they constitute a larger share of total taxation, indicating a narrower and more progressive base.

These differences matter. As discussed in Section 2, tax instruments vary in their revenue-generating capacity for both economic and political reasons. Economically, consumption taxes are less distortionary and more growth compatible than income taxes (Lindert, Reference Lindert2004). Politically, taxes differ in visibility and in their link to perceived benefits. Social insurance contributions, for example, are typically deducted automatically from wages and are matched by employers’ contributions, making them relatively invisible. More importantly, they are closely tied to social entitlements. As a result, many people perceive SICs not as taxes but as contributions to their own pension or health insurance plans (Campbell & Morgan, Reference Morgan2005; Morel & Palme, Reference Morel, Palme and Greve2018; Schmitt et al., Reference Schmitt, Lierse and Obinger2020).

4.2 Three Perspectives to Explain Tax Regimes

Having established the significance of taxation for shaping the structure and outcomes of welfare states, the next section turns to a deeper question: What explains the emergence and persistence of such different financing systems across countries? A conventional view holds that the development of tax policies mirrored the construction of the welfare state and was shaped by the same political and institutional forces. A traditional version of classical power resource theory, for instance, posits that left parties develop more generous welfare states funded by high and progressive taxes (Cameron, Reference Cameron1978). However, this theory struggles to explain a critical empirical puzzle: countries dominated by left parties often achieve higher levels of social spending, yet paradoxically rely more on regressive forms of taxation than countries in which the left never governs (Martin, Reference Martin2015; Andersson, Reference Andersson2022). To make sense of this divergence between tax regimes, the following section draws on three major explanatory perspectives from the comparative political economy literature: the functionalist, the distributional, and the sectoral perspectives (Haffert, Reference Haffert2021a).

The functionalist perspective offers a structural explanation for differences in countries’ tax systems, emphasizing that tax structures emerge as a response to two main imperatives: countries’ spending needs and the constrains imposed by the economic environment. In functional accounts, governments first decide on spending levels and then select tax instruments that are economically viable given broader structural conditions (Ganghof, Reference Ganghof2006; Lindert, Reference Lindert2004). A core assumption is that in order to sustain high levels of social expenditure, governments are functionally compelled to adopt tax instruments that minimize economic distortions and maximize revenue. In a situation of international tax competition, capital mobility and a structural dependence of the state on capital (Przeworski and Wallerstein, Reference Przeworski and Wallerstein1988), governments face a critical challenge: how to generate sufficient revenue without discouraging job creation or capital investment. This constraint leads them to prefer taxes that generate substantial revenue with minimal economic distortion. Hence, according to the functionalist account, the reliance of high-spending welfare states on regressive taxation such as VAT stems from its status as one of the most efficient revenue-generating machines, especially when compared to capital and corporate taxes, which generate relatively little revenues, while potentially creating significant economic distortions (Lindert, Reference Lindert2004).

The functionalist perspective also links tax regime variation to structural economic conditions: countries with open economies – more exposed to capital flight and global competition – are seen as more compelled to adopt business friendly taxes like VAT, creating a distinctive relationship between economic openness and tax structure (Seelkopf et al., Reference Seelkopf, Lierse and Schmitt2016).

While the functionalist perspective offers important insights into the structural constraints shaping tax regimes, it faces two key critiques. First, institutionalist accounts challenge its assumption of present-day optimization by emphasizing the long-term effects of early policy choices. These accounts argue that tax structures are not simply functional responses to current spending needs or economic pressures but are instead shaped by path-dependent processes. Kato’s (Reference Kato2003) analysis of tax regimes reflects this view. In contrast to functionalist accounts that see the adoption of VAT as a response to the need to generate higher revenues, Kato reverses the causal relationship. Drawing on a path-dependent institutional logic, she shows that countries that adopted VAT early on were in better position to fund social policy expansions at a later point in time. In contrast, late adopters did not have the necessary revenue base when international tax competition intensified in the 1980s and constrained their ability to rely on other forms of taxes (Kato, Reference Kato2003).

A second critique comes from the distributive perspective, which challenges the functionalist tendency to downplay political agency by treating structural pressures as uniform constraints. Instead, distributive accounts emphasize that both spending and taxation are shaped by political conflict and institutional variation across welfare state regimes.

An influential distributional perspective has emerged through the notion of tax bargain, which helps explain why European countries tend to reach a high tax and high spending equilibrium, whereas Anglo-American democracies do not (Andersson, Reference Andersson2022; Beramendi and Rueda, Reference Beramendi and Rueda2007; Jacques, Reference Jacques, Heaman and Tough2020; Martin, Reference Martin2015; Steinmo, Reference Steinmo1993; Timmons, Reference Timmons2010). This approach conceives taxation as a negotiated bargain between political actors, shaped by their competing objectives around expenditure and revenue extraction. It highlights how similar structural pressures, when mediated by different institutional contexts, can produce divergent fiscal trajectories.

Two key institutions take centre stage in this perspective: the electoral system and the system of interest group representation. Both proportional electoral systems and a specific system of interest group representation called a “coordinated market economy” (CME) (Hall and Soskice, Reference Hall and Soskice2001) help foster a high tax, high spending bargain between employers, unions, and political parties across the ideological spectrum. In contrast, institutional complementarities between majoritarian electoral systems and non-corporatist, liberal market economies (LMEs) tend to generate lower overall tax revenues, more modest welfare states, and generally more progressive tax structures. This divergence reflects not only the incentives created by electoral rules but also how the broader organization of the economy influences the feasibility and direction of tax and spending bargains.

Electoral systems influence three factors that are crucial for tax policy choices: political stability, parties’ responsiveness to the median voter, and political incentives regarding public spending. Majoritarian electoral systems are characterized by single-party governments that concentrate power and generate barriers to entry for smaller parties. Moreover, they are characterized by adversarial policymaking and policy volatility, because a minor change in votes can lead to a large difference of seats won. In contrast, proportional representation (PR) systems are characterized by multiparty systems, more consensual forms of policymaking in coalition governments, and greater political stability, as small shifts in the electorate generally do not alter the balance of power and opposition parties are still able to maintain an influence on policymaking (Lijphart, Reference Lijphart2012).

The stability in PR systems creates the conditions for reaching bargains between political actors in which the left achieves its egalitarian objectives with generous social spending rather than via progressive taxation, whereas the right ensures that capital holders do not disproportionally pay for higher spending. Furthermore, the risk associated with losing power is smaller, as opposition parties maintain an influence on policymaking because of coalition or minority governments and influential parliamentary committees (Andersson, Reference Andersson2022). Hence, left governments can pursue a strategy of funding expansions of social spending drawing on regressive taxes, even if the revenues they generate are not allocated exclusively to social programmes, as is the case of SICs. Knowing that they can block future right-wing governments that could be tempted to cut back social spending, left governments can credibly commit to tax the middle class and the poor with VATs and use this revenue source to fund more generous social spending (Andersson, Reference Andersson2022; Steinmo, Reference Steinmo1993; Timmons, Reference Timmons2010). Hence, PR systems are associated with an earlier adoption of VAT (Helgason, Reference Helgason2017).

In majoritarian systems in contrast, it is impossible for parties to credibly commit to a long-term tax and spending strategy. If a party is thrown out of office in a system of single-party government, it cannot maintain its influence on the next government. Hence, the left refrains from implementing regressive taxes to fund generous social spending because a future right-wing government could choose to reduce social spending to fund income tax cuts while maintaining regressive taxes. Because of the absence of credible commitment between parties in majoritarian systems, right-wing parties cut taxes, while left-wing parties increase taxes on the rich, but refuse to increase taxes on their own supporters (Andersson, Reference Andersson2022).

Electoral systems also influence how political parties respond to the preferences of the median voter. Research indicates that voters generally support two interconnected tax-related objectives: higher tax progressivity targeting the rich, and lower overall taxation rates, especially reduced consumption taxes that affect all citizens (Barnes, Reference Barnes2015; Barnes et al., Reference Barnes, de Romément and Lauderdale2024). Majoritarian and PR systems differ markedly in how they translate these voter preferences into political strategy. Two-party systems are more likely to converge towards the centre, because small shifts in votes have large effects in majoritarian systems, whereas parties are more insulated from having to rapidly react to voters’ changing moods in PR systems (Finnegan, Reference Finnegan2022). Hence, parties in majoritarian systems are more likely to adopt a short-term tax strategy pleasing as many voters as possible, preferably by using tax expenditures beneficial in districts that are key to their re-election, therefore limiting their efforts to broaden the tax base.

Electoral systems are also related to political incentives regarding tax and spending levels. Proportional representation systems are more likely to result in coalition governments. Because the multiple players involved in the budget process can block cutbacks that would displease their constituencies, it is more difficult to reduce spending in coalition governments (Hallerberg and Von Hagen, Reference Hallerberg and Von Hagen1997). Moreover, left-wing parties, which are associated with more spending and taxes than parties on the right, tend to have more electoral success in proportional systems, while right-wing parties are stronger in majoritarian systems (Döring and Manow, Reference Döring and Manow2017). Hence, revenues and expenditures tend to be higher in countries with PR systems than in countries with majoritarian systems (Iversen and Soskice, Reference Iversen and Soskice2006).

There are institutional complementarities between electoral systems and systems of interest group representation, as coordinated market economies have PR systems and liberal market economies have majoritarian systems (Beramendi et al., 2015). Within each of these two models, both the political and economic institutions create mutually reinforcing incentives regarding taxation. These typologies of systems of interest group representation are derived from the Varieties of Capitalism school that focuses on the preferences of employers to understand why they consent to large welfare states in some countries but not in others (Hall and Soskice, Reference Hall and Soskice2001). These institutions also contribute to reaching or failing to achieve the high (regressive) tax and high spending bargain. While theories based on the role of electoral systems focused on parties, those relying on Varieties of Capitalism focus on producer groups like unions, but especially employers.

Coordinated market economies are characterized by macro-corporatist institutions, where peak association representing unions and employers meet with government representatives to implement public policies related to wages, taxes, and spending. Another key distinguishing feature of CMEs is their unique industrial structure, which differs fundamentally from liberal market economies. These economies depend on workers with industry-specific skills that are not easily portable. This specialization creates a significant vulnerability for workers, who risk losing everything if their specific industry declines. To offset this ‘skill risk’ and encourage long-term specialization, employers consent to expansive social insurance systems. This provides the necessary security for workers to invest in specialized skills, forming the foundation of the high-tax, high-spending corporatist bargain (Hall and Soskice, Reference Hall and Soskice2001).

This industrial structure and the accompanying institutional arrangements fundamentally shape the tax and welfare regime strategies. The engagement of unions, employers, and the state in macro-corporatist institutions makes it possible to credibly commit to tax policies. Repeated contacts between unions and employers can raise awareness of both constituencies on the costs of high capital taxes on firms’ competitiveness and on the benefits of social insurance spending for employees’ retention (Martin, Reference Martin2015). In CMEs, employers and unions can consent to a tax and spending policy bargain that involves high social spending paid for by taxes on workers. Employers consent to high social insurance since they can tailor tax policies to their own needs, notably by making sure that the tax burden is not disproportionately theirs to shoulder. Unions consent to higher taxes on their own members, knowing that they can count on higher social spending (Jacques, Reference Jacques, Heaman and Tough2020; Martin, Reference Martin2015).

In contrast, in LMEs, industries rely on general skills, reducing the employer’s motivations to support public social insurance. Employers and unions are not represented by a peak association, rather they represent the particular interest of their own industry and hence cannot impose bargains on their members. Unions and employers aim for short-term objectives and prefer to shift the tax burden to the other group. The conflictual relationship between employers and unions makes the high tax/high social insurance bargain impossible to sustain in LMEs (Jacques, Reference Jacques, Heaman and Tough2020; Martin, Reference Martin2015).

The corporatist/non-corporatist character of these capitalism varieties fundamentally shapes tax regime differences. In CMEs, the corporatist structure constrains the left’s ability to rely on capital taxes to fund social spending as such actions would disrupt corporatist agreements between labour and capital. Hence, we also observe that CMEs are associated with an earlier adoption of VATs (Helgason, Reference Helgason2017). In contrast, non-corporatist settings allow the left more flexibility with capital taxation. At the same time, in non-corporatist settings, left-wing parties may hesitate to increase regressive taxes, fearing these won’t genuinely fund expanded social spending. Consequently, they might prefer higher income taxes on the wealthy.

These distributional accounts are based on stylized assumptions about political parties and interest groups preferences that are not necessarily tested by these theories. They assume that actors know the consequences of tax policies, make tax policy choices based on known distributional effects, and that each actor aim to maximize their own welfare, defined either in terms of income (for producer groups) or of votes (for parties). However, the relationship between parties’ tax policy preferences and those of voters is inherently more complex than what these rationalistic theories suggest, as we will discuss in Section 5. Moreover, the sectoral perspective reveals that producer groups may also have more nuanced preferences.

While distributional accounts emphasize class-based conflicts and compromises between labour and capital in tax regime formation, the more recent sectoral perspective shifts towards a cross-class view of tax policy choices, rooted in growth model theories (Haffert, Reference Haffert2021a; Haffert, Reference Haffert2021b; Haffert and Mertens, Reference Haffert and Mertens2021; Huo, Reference Huo2021). The growth model perspective builds on the Varieties of Capitalism school to explain how countries engage in different strategies to achieve durable economic growth (Baccaro and Pontusson, Reference Baccaro and Pontusson2016; Baccaro et al., Reference Baccaro, Blyth and Pontusson2022). It distinguishes between export-oriented growth models and consumption-oriented growth models, which resemble CMEs and LMEs, respectively, although some countries combine both growth models (i.e., Sweden) and others are unable to develop a coherent growth model (i.e., Italy). Growth models are sustained by cross-class social blocs uniting labour and capital in the economic sectors that benefit from the current growth model (Baccaro and Pontusson, Reference Baccaro and Pontusson2016; Baccaro et al., Reference Baccaro, Blyth and Pontusson2022).

The sectoral perspective on tax policy choices proposes that taxes can be used as industrial policies to encourage specific sectors and economic activities as they influence decisions related to savings, consumption, working, exporting, importing, and investing (Haffert Reference Haffert2021b; Steinmo Reference Steinmo1993). Haffert and Schultz (2020) note that the conflicts surrounding the introduction of VAT in 1967 in the European Economic Community were primarily reflecting sectoral rather than class-based interests. Because VAT is generally regarded as a regressive tax method that disproportionately impacts low- and middle-income classes, researchers often assume that the struggles over its implementation are class based. While this is accurate in terms of VAT’s expansion, Haffert and Schultz (Reference Haffert and Schulz2020) reveal that the initial political struggles over its adoption in the European Economic Community were sector based. Competitive export sectors advocated for its introduction, needing a standardized European consumption tax to reduce trade barriers, while domestic industries resisted the change because it exposed them to foreign competition.

According to this perspective, the size and structure of the tax state depend on the balance of power between different sectoral coalitions that are influencing tax policy. For instance, both unions and firms in sectors that depend on domestic consumption will favour lower consumption taxes (Haffert, Reference Haffert2021a). In contrast, Nordic countries rely heavily on service sector jobs and tax consumption so that citizens find it more attractive to rely on public services (Haffert, Reference Haffert2021a). Moreover, countries that rely on financial assets to finance debt-driven consumption will maintain a more favourable treatment on mortgage interest to encourage debt-financed home ownership (Fuller, Reference Fuller2015) and foster higher property value. In turn, higher property value can generate larger property taxes as a proportion of GDP, as seen in Section 4.1.2 in the case of Canada and the United Kingdom. Therefore, countries with consumption-driven and domestically oriented growth models maintain less consumption taxes and aim to tax production with direct taxes, whereas countries with export-driven growth models minimize taxes on producer groups and shift the tax burden towards consumption to curb it and to encourage exports (Haffert and Mertens, Reference Haffert and Mertens2021).

Still, apart from the pioneering work of Lukas Haffert, research on the link between growth models and tax regimes remain in its infancy and more needs be done to develop the sectoral perspective in order to explore the relationship between growth models and tax policy choices. More generally, future research should integrate the three perspectives together to explain the determinants of tax regimes. In fact, they are generally presented as separate theories by researchers, but the sectoral and distributional perspective are relying on similar variables and institutions to explain tax regimes (consumption-oriented growth models tend to have majoritarian electoral systems and a Liberal market economy for example), whereas both perspectives certainly agree that structural constraints highlighted by the functional perspective matter to explain tax policy choices.

The three perspectives share rationalist assumptions that political actors know the distributional effects of tax policies and choose them based on their own interests and on the institutional constraints that they face. Yet, future research could incorporate a constructivist perspective focusing on how political parties and producer groups perceive tax policies and their distributional effects, and whether these policies are deemed politically feasible or not (see, for example, Fastenrath et al., Reference Fastenrath, Marx and Truger2022).

4.3 Conclusion

The analysis of tax regimes reveals a complex interplay between economic structures, political institutions, and social compromises. While different theoretical perspectives – functionalist, distributional, and sectoral – offer unique insights, they collectively underscore that tax systems are not simply technical instruments but deeply political mechanisms reflecting societal negotiations and economic strategies.

Across different welfare state models, we observe a paradoxical pattern where seemingly progressive objectives are achieved through ostensibly regressive taxation mechanisms. The European cluster demonstrates how high social spending can be funded through consumption taxes and SICs, challenging simplistic notions of tax progressivity and highlighting the nuanced relationship between taxation, redistribution, and economic growth.

Future research should focus on integrating these perspectives and developing a more comprehensive understanding of welfare financing. This includes examining both public tax systems and private financing instruments to provide a more holistic picture of the comparative dimensions of the welfare finance equation, moving beyond their current compartmentalization to explore how different funding mechanisms interact and shape social policy and socio-economic outcomes.

5.1 Determinants of Tax Policy Preferences at the Individual Level

The study of public opinion on taxation has witnessed significant developments in recent years. From relatively vague survey questions about general support for taxation, the field has evolved towards using more complex online tools to survey the intricacies of the public’s preferences regarding taxation. These include direct trade-off questions comparing support for public spending with and without the costs associated to it (Busemeyer et al., Reference Busemeyer, Garritzmann and Neimanns2018), conjoint experiments measuring how specific tax policy changes influence opinion (Bremer and Bürgisser, Reference Bremer and Bürgisser2025), and online budget tools that allow respondents to simulate a public budget by choosing how to fund their preferred public programmes (Barnes et al., Reference Barnes, Blumenau and Lauderdale2022; Tuxhorn et al., Reference Tuxhorn, D’Attoma and Steinmo2021). The literature has highlighted three main individual-level determinants of preferences for taxation: self-interest related to one’s socio-economic status (income and education), ideology, and fairness considerations.

5.2 How Tax Instruments and External Events Feedback into Preferences

Section 5.1 has analysed the determinants of individual preferences on taxation and their consequences for the politics of tax reforms. In this subsection, we turn the tables and analyse how tax policies feedback into public opinion since tax policy design influences how the public feels about various taxes (Campbell, Reference Campbell2018). We discuss how three main dimensions of revenue-generating instruments discussed in Section 2, influence public opinion: the progressivity of taxes and of social policies, the degree of public and private funding, and the governance structure of the funding mechanism. Then, we analyse how external events like wars and economic downturns influence tax policy preferences.

An important strand of this literature has analysed whether the size, progressivity, and degree of redistribution of the spending and taxing dimensions of the welfare state influence public opinion. On the one hand, the reciprocity argument suggests that citizens are more willing to pay taxes if they expect to receive something in exchange for the taxes they pay (Levi, Reference Levi1988). Therefore, if social policies are universal, in the sense that the whole population is allowed to receive benefits or services, citizens are more likely to be willing to pay taxes and, as a consequence, the social budget should be larger and therefore more redistributive (Jacques and Noël, Reference Jacques and Noël2018; Korpi and Palme, Reference Korpi and Palme1998).

In line with this argument, Berens and Gelepithis (Reference Berens and Gelepithis2019) have found that support for tax progressivity is higher when the middle class benefits from the welfare state, and lower when social spending is targeted at the poor. Similarly, tax progressivity, but more importantly social transfers’ progressivity, increases the effect of income on preferences for redistribution (Beramendi and Rehm, Reference Beramendi and Rehm2016). Hence, less progressive social transfers would dampen class conflict about taxation. This argument is contested, however, as another study has found that both more redistribution and more progressivity in social transfers generates bigger class cleavages concerning support for redistribution (Fernández and Jaime-Castillo, Reference Fernández and Jaime-Castillo2018). In fact, the relationship between social policy universalism and willingness to pay suggested by Korpi and Palme (Reference Korpi and Palme1998) deserves more direct empirical tests.

Section 2 highlighted the importance of the funding mechanism and the governance structure of taxation. We can assume that SICs are perceived differently by the public than income taxes. Some argue that SICs are perceived as deferred wages rather than taxes (Palier, Reference Palier2010), and that it is very easy for the public to see where the money goes, which would increase their popularity (Campbell, Reference Campbell2018). The fact that SICs are tied to the programmes they are funding could help to build credible policies and enhance citizens’ trust (Koreh and Mandelkern, Reference Koreh and Mandelkern2023). In fact, some have found that support for social insurance programmes like Social Security in the United States is very high because the payment is directly earmarked to a popular social programme whose beneficiaries are deemed deserving of the benefits they received because they contributed to the programme (Campbell, Reference Campbell2018; Williamson, Reference Williamson2017). However, more research needs to be done to analyse how citizens perceive social security contributions relative to other forms of taxation. Are citizens more willing to pay social contributions than taxes? One study has shown that SICs are the least contested type of tax across twenty democracies (Martin and Gabay, Reference Martin and Gabay2018), but another study has found that in the United Kingdom, SICs are not particularly popular (Barnes et al., Reference Barnes, de Romément and Lauderdale2024). Moreover, we do not know if the nature of the programme funded by SICs influences public opinion: Are respondents more supportive of paying pension contributions that are more likely to be seen as deferred wages than healthcare contributions common in Social Health Insurance systems? Are contributions to social insurance covering life cycle risks (pensions and healthcare) inherently more popular than contributions for social insurance coverage of a more targeted labour market risk like unemployment insurance? Hence, more research should be done to confirm if funding by SICs is politically easier to sustain.

Similarly, public opinion on private financing methods is also understudied. As discussed in Section 2.1, private funding of social policies entails significantly less redistribution than funding by income taxes or social security contributions. It is expected that the beneficiaries of private financing methods would oppose an expansion of public financing that would crowd out private insurance. Moreover, several studies have found that private financing shapes public opinion as it reduces support for the state’s involvement and influences expectations about the provision of social services by the market (Busemeyer and Iversen, Reference Busemeyer and Iversen2020; Lindh, Reference Lindh2015; Zhu and Lipsmeyer, Reference Zhu and Lipsmeyer2015). Hence, private financing methods would create a vicious circle against the public funding of social policies by increasing support for the private system among its beneficiary and potentially reducing citizens’ willingness to pay taxes for the public system.

However, most of the previous research on the influence of privatization of funding on public opinion rely on time series cross-sectional designs that cannot establish causality, mostly because of the lack of long panel series of public opinion that include the relevant questions on support for specific public or private programmes. Future research must identify a semi-exogenous programmatic change in financing methods towards private or public financing and rely on a panel survey of public opinion to see if the programme changes influence the public in the expected direction.

5.3 Are Governments Reactive to Public Opinion on Taxation?

Political parties are reactive to public opinion, at least under certain conditions (Busemeyer et al., Reference Busemeyer, Garritzmann and Neimanns2020). However, the previous review suggests that policy elites may have different preferences than the general public. For example, there seems to be strong public support for funding social policy by SICs, but policy elites are often wary of SICs’ effects on employment, as it increases labour costs and reduces firms’ competitiveness (Kemmerling, Reference Kemmerling2009). Moreover, experts on climate change praise the merits of carbon taxes but sustaining high enough levels of carbon pricing to influence consumers and producers’ decisions is often politically difficult to achieve because of the public’s resistance (Carattini et al., Reference Carattini, Carvalho and Fankhauser2018). While policy elites may prefer broad tax bases that are economically efficient and hard to avoid, such as VATs, citizens often prefer narrow tax bases on rich citizens and corporations (Steinmo, Reference Steinmo2003; Barnes et al., Reference Barnes, de Romément and Lauderdale2024).

The policies resulting from the disagreements between the public and policy elites on taxation is a fascinating research topic. This is illustrated in the following debate over government’s responsiveness to taxes on the rich. Sometimes, tax policies do not follow public preferences. This is the case regarding taxes on the rich. As shown in this section, there is an overwhelming support for higher taxes on the rich. In fact, increasing taxes on the rich would even increase overall support for tax reforms (Bremer and Bürgisser, Reference Bremer and Bürgisser2025). At the same time, the national income share allocated to the top 1 per cent has risen in most countries over the last decades and there are significant revenue needs across OECD countries. However, governments have generally refrained from increasing taxes on the rich, such as by raising top marginal tax rates, taxes on capital gains, inheritance taxes, or wealth taxes. Can this phenomenon be explained because governments are more responsive to the preferences of the rich, or is it because support for taxes on the rich is not as pervasive as generally assumed?

Since the publication of Gilens and Page (Reference Gilens and Page2014) classical study on the issue, it has become common to argue that the policy process in most democracies is biased in favour of the preferences of high-income citizens and of the interest groups representing them, while being unresponsive to those of median voters. Some argue that the lack of reforms towards increasing the tax burden on the rich represents a case of unequal responsiveness, because tax cuts on the rich are opposed by a majority of voters (Mathisen, 2024). Hence, parties would not be responding to the public’s demand for higher taxes on the rich.

However, rather than blaming unequal responsiveness, some contend that the public’s preferences on taxes on the rich are elusive. Indeed, some taxes that target mostly the rich, such as inheritance taxes, are very unpopular (Emmenegger and Marx, Reference Emmenegger and Marx2019). Moreover, even if they support higher taxes on the rich, citizens tend to be indifferent to how much higher taxes on the rich should be raised, which gives leeway to governments that do not wish to implement significant reforms (Ballard-Rosa et al., Reference Ballard-Rosa, Martin and Scheve2017). Also, many people concerned about inequality do not believe that progressive taxes are the main solution to the problem as they hold strong beliefs about the importance of the equal treatment of individuals by the state, which means, to them, that everyone should pay the same rate of taxes (Scheve and Stasavage, Reference Scheve and Stasavage2023).

Finally, there is a large difference between what people say in surveys and the extent to which they demand taxes on the rich in the real world. In focus groups, people display strong internalized pro-business tax preferences and meritocratic justifications of wealth. As such, politicians often perceive survey responses about taxes on the rich to be meaningless to actual politics and believe that there is no appetite among voters for taxing the rich (Fastenrath and Marx, Reference Fastenrath and Marx2023). For policies to change, the party system must supply electoral pledges about taxes on the rich (Elsasser et al., Reference Elsässer, Fastenrath and Rehm2023). Yet, left-wing parties have generally converged with the right regarding wealth-related taxes in the 1990s (Lierse, Reference Lierse2022). Moreover, the logic of collective action suggests that tax reform targeting a few rich voters or corporations are likely to be resisted by the groups concerned, whereas few interest groups form to defend citizens’ interest to improve the revenues sources of the general budget to have better funded public services. This debate suggests that there is room for development of more qualitative tools of public opinion research that can provide different insights into the political dynamics of taxation.

5.4 Conclusion

In summary, this section has critically reviewed studies on the micro-foundations of tax policies by focusing on public opinion dynamics. It revealed that preferences for taxation and for public spending are not driven by the same factors, as income, income growth, and education have somewhat counter-intuitive associations with tax preferences. We have also reviewed studies revealing strong contextual effects on tax preferences as citizens react to the existing tax and spending system differently when forming their preferences. Finally, we have argued that more studies need to be done on public opinion of specific taxes and on causal impact of tax changes on government’s popularity, as assumptions about the unpopularity of tax increases may not be as straightforward as commonly expected.