The concept of fiscal welfare refers to the use of fiscal instruments—such as tax reductions or credits—for social protection purposes. It involves social transfers that are not made through direct benefits or allowances, but through tax expenditures, which are less visible. This type of intervention is very common in liberal welfare states (such as the United States and the United Kingdom), but it has spread widely to other contexts. However, it remains under-analysed, both as a public policy instrument and in terms of its consequences for social protection.

What I show in this book is that this instrument has powerful effects in terms of wealth redistribution, the structuring of social protection provision, and the structuring of the labour market, while often escaping critical analysis because the fiscal instrument is perceived as neutral or technical. Yet, the use of this instrument reflects profound political choices: who deserves to be supported, under what conditions, and by what means.

I have chosen to compare France and Sweden because these two countries embody two very different models of social protection according to the typologies established in the literature⁽¹⁾: conservative (based on an insurance system and family-based principles) in France, and social democratic (a universalist and 'de-familialising' model in which the state plays a central role) in Sweden. The use of tax expenditures has been increasing since the early 1990s in France and since 2007 in Sweden. This comparison not only addresses the question of what happens to these two types of welfare states when an instrument from the public policy repertoire of liberal welfare states is introduced; it also highlights the specific properties of this instrument. Indeed, I find similar mechanisms and results in these two otherwise very contrasting welfare states. 

In France, the tax credit (initially introduced as a tax reduction) for personal services was introduced in 1991 through the "Aubry Law" ⁽²⁾. This initiative had three objectives: to combat high unemployment, reduce undeclared work, and meet the growing needs for childcare and assistance for the elderly. The scheme has been extended several times by raising the ceiling for eligible expenditure (which currently stands at €12,000, plus €1,500 per dependent child and/or per person over 65 in the household) and by broadening the type of services covered by the scheme.

In Sweden, although discussions began in the 1990s, the equivalent tax measure – the RUT reduction – which has a similar expenditure ceiling, was not introduced until 2007. The objectives were similar to those in France: to create jobs, combat undeclared work, meet the growing demand for services for the elderly, but also to improve the work-life balance of families and promote gender equality by lightening the domestic burden, which is often borne by women.

In France, the adoption of this tax benefit was originally based on political consensus. Both the Right and the Left saw it as a tool to stimulate employment and meet growing needs for childcare and home help. From the late 1990s onwards, it was the Right that championed this measure, with the aim of both getting people who were distant from the labour market back into work and supporting demand for services from households, particularly the wealthiest ones. Private personal services companies quickly became a powerful lobby contributing to the expansion of this policy. Despite some criticism from the Left, particularly in the late 1990s, this policy did not spark much debate.

In Sweden, on the other hand, the introduction of the tax credit in 2007 by a Right-wing government was much more ideologically divisive. It explicitly aimed to create a private sector in a country historically committed to the public and universal provision of social services. As in France, the arguments of free choice for the elderly and the activation of low-skilled women were put forward. However, this reform was strongly contested by the Left, trade unions and feminist circles. These opponents denounce the reproduction of an unequal class society, in which precarious women (often migrants) free up time for other more privileged women. The scheme is seen as reinforcing social, gender and racial inequalities. Thus, the schemes have crystallised debates on the role of the state, the universality of social services and new forms of social segmentation.

The introduction of these tax measures had a profound impact on eldercare in both France and Sweden. While these policies claimed to improve the response to the growing needs of an ageing population, their actual effect was more ambiguous and structurally unequal.

In France, by directing public support towards demand for services rather than supply, the scheme has contributed to transforming the sector: the rise of private for-profit providers, the increasing commodification of services, and the fragmentation of public management. Furthermore, the data show very unequal use of the tax credit: in 2022, more than 40% of the total tax expenditure benefited the richest 10%, and among the over-80s, only 15% of the least wealthy (those in the first income decile) used it, compared with 63% in the last decile. The scheme therefore primarily benefits those who do not necessarily have the greatest needs, but who have the means to purchase these services. At the same time, the coexistence of this tax credit with the Personalised Autonomy Allowance (Allocation personnalisée d’autonomie, APA), which is means-tested and depends on the level of dependency, has created a dual system. Those on the lowest incomes are covered by the APA, while those on higher incomes benefit from the tax credit. This distribution results in a U-shaped curve in the distribution of aid, leading to a phenomenon of non-take-up among people on middle incomes. These two schemes also differ in terms of their cost evolution, with the high cost of the tax credit growing faster (and uncontrollably) than that of the APA, while only marginally meeting eldercare needs. This questions the effectiveness of this tax measure in terms of its stated objective of meeting a social need, as well as its opportunity cost ("what could be done better for the same price?").

In Sweden, the disruption caused by the tax measure has been even more striking. Traditionally, care for dependent persons was part of a universal model based on municipal responsibility, with public services provided according to need. The tax credit has introduced a rationale based on the market and individual choice. This encouraged wealthier households to turn to the private sector, which may ultimately reduce their support for the public sector and its solidarity-based funding. The result is a form of partial privatisation of dependency, where wealthy older people find it easier to outsource care tasks to private providers, while the less well-off remain dependent on public services that are sometimes under pressure or even reduced due to budget cuts.

In both cases, public support through tax credits changes the terms of access and governance of care: it promotes a system based on solvency rather than solidarity. Finally, it obscures the precarious working conditions of home care workers, shifting the focus to consumer needs rather than the quality and continuity of care.

The introduction of tax credits for personal services has had a significant but mixed impact on the labour market in both France and Sweden. These measures have indeed contributed to job creation, but mainly in a low-skilled, precarious and highly gendered segment of the labour market.

In France, nearly 1.4 million people were employed in this sector in 2022, the majority of whom were women. However, most of these jobs are part-time, low-paid and offer no career prospects. Many workers have multiple employers or work in poor conditions. Tax measures have not led to any structural improvement in the quality of these jobs; rather, they have institutionalised a "low-end" segment of the labour market designed to meet the domestic needs of the upper classes.

In Sweden, a country historically committed to a model of public, skilled and stable employment in services for the elderly, the introduction of the tax credit marked a break with the past, characterised by the rapid creation of a new private sub-sector, often organised into very small businesses or self-employed entrepreneurs. The profile of employees in this sector is similar to that observed in France: women, often from immigrant backgrounds, in poorly protected jobs outside public collective agreements. The Swedish model has thus seen the emergence of a stratification of the labour market in personal services, between a regulated public sector and a deregulated private sector.

In both countries, the tax measure has therefore encouraged the growth of atypical and low-paid jobs. It has also reinforced social and gender inequalities in the workforce by promoting models of "domestic outsourcing" where women in precarious situations take on tasks that others, who are better off, delegate.

Numerous criticisms have been repeatedly made by the Court of Auditors in both countries, as well as by parliamentary committees and other public institutions responsible for evaluating public policy. The various reports highlight the high cost of these schemes in relation to their results: low effectiveness in terms of job creation, expenditure that mainly benefits the wealthiest, inadequacy in relation to real dependency needs, fragmentation of funding and public management of eldercare policies. 

Nevertheless, in both countries, this policy not only continues but has been extended several times. This is mainly due to the specific nature of this instrument: on the one hand, the cost to the public purse is less visible than that of direct expenditure (such as benefits or public employment). On the other hand, this tax advantage has contributed to structuring a private services market supported by economic actors and employers' confederations, which have become active advocates of the scheme, lobbying effectively with the public authorities.