Bringing Household Services Out of the Shadows

4. International approaches to household service formalisation

With the goal of increasing work-life balance for households – particularly for women in employment – and reducing the wide‑ranging practice of undeclared employment, a number of countries across the OECD have implemented policies to formalise and boost the non-care household service sector. By subsidising formal non-care household services (reducing both consumer prices and increasing take‑home pay) and easing administrative procedures, policy interventions make the use of formal non-care household services more attractive to households. Facilitating the outsourcing of housework intends to increase the labour force participation of women, who carry out most unpaid housework, but objectives also include creating new job opportunities for low-skilled workers and the (long-term) unemployed.

This section reviews different approaches to formalising household services in the OECD by introducing some more or less comprehensive household service policy systems in Belgium, Finland, France, Germany and Sweden. The design of policies is considered as well as their implications for the size of the undeclared sector, employment in the non-care service sector, and the quality of work. Other European countries also have relevant policies, but these are usually less comprehensive and not discussed in detail (a brief non-exhaustive overview over some of their policy frameworks is available in Annex A). Non-European OECD countries generally have less developed policy frameworks that mainly centre around working permit programmes for live‑in care workers, such as in Korea and Canada (with the exception of the province of Québec, see Annex A). In the United States there is no specific policy towards formalisation of the (non-care) household service market, and here undocumented migrants often carry out work on an informal basis (Jokela, 2017[58]).

Table 4.1 presents the main policy instruments as operational in Belgium, Finland, France, Germany and Sweden. The fiscal aspects of the different instruments are calculated assuming a two‑adult household with two dependent children that pools the available deductions and social vouchers as shown in national currency as well as 2019 USD Purchasing Power Parities (PPPs). The bottom part of the table pools the instruments together and highlights the total available subsidy, which is the sum of all available household service tax deductions at the maximum threshold and the maximum social vouchers volume. The cost reduction at maximum subsidy is the share of total expenses subsidised by the government when reaching the tax deduction ceiling or buying the maximum amount of social vouchers available per household.

Table 4.1. Overview of main policy instruments in five countries

		Belgium2	Finland3	France4	Germany5	Sweden6
Main policy instrument		Social voucher, tax credit	Tax credit	Social voucher, tax credit	Tax credit	Tax credit
Name		Titres services / Dienstencheque	Kotitalousvähennys	CESU, Crédit d’impôt pour l’emploi à domicile	Steuerermäßigung für haushaltsnahe Dienstleistungen	RUT-Avdrag
Operating agencies		Regional governments, Sodexo, Belgian Tax Administration	Finnish Tax Administration	URSSAF (+ private contractors), French Tax Administration	German Tax Agency	Swedish Tax Agency
Covered services		Non-care	Non-care, care & renovation	Non-care & care	Non-care, care & renovation	Non-care & care
Work arrangements		Tripartite	Tri- & bipartite	Tri- & bipartite	Tri- & bipartite	Tripartite
Provider types		Provider organisations (for- & non-profit) Micro‑enterprises	Provider organisations (for- & non-profit, public) Direct employment	Direct employment Provider organisations (for- & non-profit) Micro‑enterprises	Provider organisations (for- & non-profit) Direct employment	Provider organisations (for-profit) Micro‑enterprises
Public Spending (GDP)l	Total	0.48% (2017)	0.21% (2019)	0.31% (2014)	0.08% (2019)	0.11% (2017)
	Non-care	0.48% (2017)	~0.06% (2019)	0.18% (2014)	~0.02% (2019)	~0.10% (2017)
Workers Total	Total	150 000 (2019)	4 900 (2012)	1 300 000	~370 000 (2018)	21 686 (2019)
	Non-care	150 000 (2019)	/	/	/	/
Tax credit	Applicable to	Voucher costs	Labour costs3	Labour costs	Labour costs	Labour costs
	Rate	10‑20% (varying by region)	40% (service provider) 15% (direct emp.)3	50%	20%	50%
	Tax claims	Non-refundable	Non-refundable	Refundable	Non-refundable	Non-refundable
	Reporting	Tax return (annual)	Tax return (annual) Tax card (immediate)	Tax return (annual)	Tax return (annual)	Through provider (immediate)
	Max. credit	EUR 440 PPP 581	EUR 4 500 PPP 5 332	EUR 7 500 PPP 10 267	EUR 4 0 005 PPP 5 3865	SEK 75 000 PPP 8 448
Social voucher	Price & value	EUR 9 / 10 (800 / 200 vouchers) exchangeable for 1h of work (EUR 23.50 / PPP 31)	/	Pre‑specified by issuer	/	/
	Paid by	Households	/	Employers Local authorities Social organisations	/	/
	Maximum net value	EUR 14 300 PPP 18 869	/	EUR 3 660 PPP 5 010	/	/
Total subsidies1	Maximum annual subsidy	EUR 14 740 PPP 19 450	EUR 4 500 PPP 5 331	EUR 11 160 (or 14 801)4 PPP 15 276 (or 20 261) 4	EUR 4 000 PPP 5 386	SEK 75 000 PPP 8 448
	Cost reduction at maximum subsidy	62.72%2	39.65%	59.81% (or 79.32%)4	20,00%	50.00%

Notes: Entries indicated with “ ~” mark rough estimations based on calculations and assumptions laid out in the following sub-sections, entries marked with” /” indicate statistics that could not be identified by the authors or that specific policy instruments are absent.

1. For all countries calculations are based on maximum available subsidy per year for a two‑earner household with two dependent children. That is, the maximum subsidy pools all available tax credits and voucher subsidies at the respective tax credit ceiling or maximum voucher volume. The effective cost reduction at the maximum subsidy calculates the share of total labour costs for household service work reduced through the policies at the maximum subsidy.

2. For Belgium, calculations assume a household in the Brussels region (tax credit of 15% on first 163 vouchers; value of voucher EUR 23.50 / PPP 31). The effective cost reduction higher for lower voucher volumes (1‑300: 68%; 301‑800: 64%). The number of workers refer to employees in the sector in the fourth quarter of 2019.

3. For Finland, the effective cost reduction is lower for direct employment (14.9%). For direct employment, the tax credit applies to the wage costs, while employer contributions are 100% covered.

4. For France, the housework share is solely based on services provided through tripartite modes. The declarative CESU voucher eases administrative access and additional social security exemptions of 2 EUR (PPP 2.73) per hour of work are available for direct employment. Therefore, the maximum annual subsidy for France depends on the nature of the service provision. In the case of direct domestic employment, the maximum subsidy increases by PPP 4 9 185 (calculated assuming a minimum wage). This higher value is presented in parenthesis.

5. For Germany, marginal employees (earning less than EUR 450, or PPP 606, per month or work less than 3 months or 70 days per year) are exempt from income taxation. A household voucher eases their employment. The German tax credit has a lower deduction ceiling for renovation work, i.e. EUR 1 200 (PPP 1 615)

6. For Sweden, the total subsidy together with the ROT-reduction cannot exceed SEK 75 000 (PPP 18 448). The RUT-deduction can also be applied to direct domestic employment, but in this case, it only applies to the employer contributions and not to the wages paid. As such, RUT-subsidised direct employment is relatively rare and thus excluded from this overview.

4.1. Policy objectives

While most household service policy systems aim at formalising the sector, the underlying mix of objectives may differ, but generally also includes improving job-opportunities for the low skilled and enhancing their social protection coverage as well as easing work/life balance issues within families. As one of the first comprehensive policy frameworks, the French Government introduced in the early 1990s tax credits for domestic employment, the Crédit d’impôt pour l’emploi à domicile, and two social vouchers, Chèque Emploi Service and Titre Emploi-Service, which could be used to pay personal and household services. The main motivation behind formalising the sector was to increase employment levels among the lower qualified as well as providing support for the elderly (Guiraudon and Ledoux, 2015[59]). Both vouchers were replaced by the universal social Chèque emploi service universel (CESU) system in 2006, which comprises the pre‑financed CESU, a social voucher, and the declarative CESU, a simplified declaration and remuneration system for directly employed domestic workers.

After running initial regional pilot trials across the country from 1997, Finland introduced the Domestic Help Service Tax Credit (DHSTC), called Kotitalousvähennys, in 2001. Besides increased formalisation of the household service sector, the main objective of this framework was to increase the work-life balance within families and to boost employment among lower-qualified population segments (Aalto, 2015[60]).

Since 2003, the German Government grants a tax credit on household services, called Steuerermäßigung für haushaltsnahe Dienstleistungen. The objective of this system is an expansion of employment in the household service sector, which traditionally faced labour shortages. Along with the tax credit, and overall large‑scale labour market reforms in the same year, the government also introduced the concept of Minijobs (marginal employment) and Haushaltsschecks (household cheques), aimed at greatly simplifying small-scale (domestic) employment. With an extension of the tax credit in 2005, the focus of the policy framework was shifted to encompass support for mothers and dual-earner families (Shire, 2015[61]).

In 2004, Belgium implemented a comprehensive system of social vouchers on the non-care household service market, called Titres services or Dienstencheque. With a primary objective of incentivising the consumption and provision of declared services, the system was also designed to offer new employment opportunities with better social protection coverage to lower qualified workers. Since then, the motivation for the social vouchers has shifted to some degree, so that they are currently seen more as a middle‑class tax break that enables better work-life balance and more employment chances for higher qualified women (Marx and Vandelannoote, 2015[62]; Raz-Yurovich and Marx, 2018[63]).

The Swedish Government introduced the RUT-Avdrag in 2007, in order to tackle the high tax wedge for domestic household services and to boost formal employment in this sector, along with increasing female employment more generally by reducing time‑stress associated with routine household tasks. Over the years and with a large influx of asylum seekers, the RUT-deduction recently has been regarded as a tool for the labour market integration of refugees (Rickne, 2019[64]).

4.2. Policy instruments

Governments can use different tools to formalise the non-care household service sector. Usually governments choose demand-side measures, such as tax credits and social vouchers, but favourable tax treatment can also include reduced VAT-rates and reduced employer social security contributions. Each of these instruments may be used in isolation, but often the non-care household service sector is supported by a combination of instruments. As such, some countries have complex systems that involve multiple actors and instruments.

Most of the policy instruments work by reducing the effective prices for formal household service work. With lower prices, the consumption of undeclared services is supposed to decrease as the incentives for this form of employment weaken. However, price levels are not the only consideration for either household or service providers. Policies can also aim to ease access to the services and simplify employment formalities.

4.2.1. Tax credits

Tax credits serve as a tool to reduce net service prices, by granting a deduction on the annual tax liability of households that use non-care household services, relative to the volume of service work consumed and up to a certain threshold – a tax deduction ceiling. National tax authorities operate these instruments.

Tax credits may be the preferred mode for delivering social support as they are often seen as tax cuts, and, for example in the United States, tax cuts are likely to be more palatable to policy makers and the general public rather than increased government spending on welfare programmes (Forman, 2010[65]; Ashok and Huber, 2020[66]).

Of the countries presented in Table 4.1, each one has some form of tax credit towards the cost of buying household services, but the design of tax credits varies considerably across countries in terms of value and ceilings of tax deductions. The largest tax credits can be found in Sweden and France, where governments grant a deduction of 50% on the labour costs for household service work, up to a deduction ceiling of SEK 75 000 (PPP 8 448) and EUR 7 500 (PPP 10 267) per household, respectively. In Finland, the tax incentives differ between work arrangements and only apply to costs above EUR 100 (PPP 118) per year: the tax credit is 40% on the labour costs for services purchased through provider organisations, but only 15% on the wages for directly employed domestic workers, along with 100% on all employer contributions. In both cases, the maximum reimbursement amounts to EUR 4 500 (PPP 5 332) per household. Germany grants a tax credit of 20%, up to a deduction ceiling of EUR 4 000 (PPP 5 386) for household services or EUR 1 200 (PPP 1 615) for renovation work. As laid out in Annex A, similar tax credits exists in, for example, Denmark (BoligJobordningen) and Luxembourg (Abattement pour charges extraordinaires). The Belgian tax credit differs from those in other countries as it only applies to the cost of the service vouchers themselves. The level of the deduction and the reimbursement ceiling varies across the Belgian regions (see below).

Refundability of tax credits

Tax credits generally only benefit those households with sufficiently large incomes and tax liabilities – i.e. the tax credit is “wasteable” or non-refundable. Low-income households may not earn enough to pay tax and thus do not benefit from the tax credit, or their tax liability is so small that they cannot benefit from the tax credit in full. To overcome this issue and bolster fairness in programme design, tax credits can be made “non-wasteable” or refundable: low-income households receive a cash payment of equivalent value to the tax credit made to richer households who offset the tax-credit to their tax liabilities. Most countries only have non-refundable tax credits for household services Table 4.1, France is the only one that currently grants refundable tax credits (see Box 4.1).

Box 4.1. The French household service tax credit and social voucher system

Since 1991, the French Government grants a refundable tax credit of 50% for household service work, up to an expense ceiling of EUR 15 000 (PPP 20 534), depending on the number of dependent children in the household.1 Thus, tax-paying households can be reimbursed up to EUR 7 500 (PPP 10 267). Initially, the tax credit was introduced as non-refundable, though between 2007 and 2016 the French tax code was amended, such that households in employment would receive a tax refund, even if their tax liability was smaller than the applicable reimbursement. In 2017, the refundability of the tax credit was generalised to all households with a further amendment of the tax code.

The additional French Universal Service Voucher (CESU), available either as a declarative system or a social voucher,, is applicable to a wide range of activities inside and outside of the consumer’s household.2 With its creation in 2006, along with precursors from the early 1990s, it aims to formalise the (household) service sector and to increase the work-life balance of employees and their families. As such, the two different types of CESU follow somewhat different objectives and mechanisms and are managed by different private contractors:3

• The declarative CESU aims to simplify the direct employment of service providers in a household by enabling households to declare their workers through the CESU. This declaration automatically generates payslips when households indicate the number of working hours, wages and bonuses to be paid. For the employee, it guarantees rights to health insurance, unemployment benefits and pension. Since 2020, the newly introduced CESU+ also allows for direct handling of salaries with automatic transfers from bank accounts declared on the CESU+ platform.

• The prefinanced CESU is a social voucher of a predefined value, paid for and distributed by employers to their employees, who can use them to pay for various service work (fully or in part). Each employee can receive vouchers up to a value of EUR 1 830 (PPP 2 505) per year, while employers can claim a tax reduction of 25% up to a ceiling of EUR 500 000 (PPP 684 450) against the costs of the vouchers. In addition, employers are also exempt from social security contributions on the value spent on the vouchers for their employees. In contrast to the declarative CESU, the prefinanced voucher can also be used pay for services through provider organisations. Both local authorities and insurances can also distribute prefinanced vouchers to distribute social allowances (social CESU).

In addition to tax credits and social vouchers, French households additionally benefit from an exemption of EUR 2 (PPP 2.74) per hour worked on employment contributions for their domestic employees if they are directly employed by the household and not through an intermediary.4

In terms of demand-side measures, non-profit service provider organisations benefit from full or partial exemptions from employer contributions as well as reduced VAT rates. Instead of the usual 20%, their household services are taxed at 10% while services to the frail and dependent are taxed at 5.5%.

1. The baseline expense ceiling is EUR 12 000 (PPP 16 427) per household, but EUR 1 500 (PPP 2053) are added for each dependent child in the household. It can be increased to EUR 20 000 (PPP 27 378) in case of a disabled person in the household

2. The household services include, among others, cleaning, ironing, gardening, childcare, elderly assistance, maintenance work, tutoring, IT support.

3. The declarative CESU is issued by URSSAF, who then handle all administrative tasks of the domestic employment, such as payroll records and social security contributions. The prefinanced CESU is issued by multiple private entities (Domiserve, Edenred, Groupe UP, Natixis Intertitres and Sodexo).

4. The exemption on social security contributions was abolished in 2011, but reintroduced in 2015.

Source: Ad-PHS (2020[67]), Williams (2018[68])

With the final generalised refundability of the French tax credit, the public expenditure on the household service sector increased by 15.4% (DARES, 2019[69]). Since the support was made available to more low-income households and the elderly, the number of tax credit claims increased by about 6% compared to 2016, after the number of claims had been relatively stable for some years (DGFiP, 2021[70]). However, the overall number of hours worked in the sector did not change significantly over the same year – if anything they fell by about 2 million (DARES, 2020[71]). Rather than drawing new consumers to the household service market, this suggests that the main effect of reform was to allow for new claims among low-income households and the elderly who already used household services in the absence of tax benefits or those who could only claim a small amount. Making tax credits towards household services refundable may not be particularly effective in reaching entirely new consumers and boosting the overall size of the sector (see e.g. Carbonnier (2015[72]) or Marbot and Roy (2014[73])).

Operational issues

Tax administrations were set up to collect money rather than distribute cash transfers. Embedding “refundability of tax credits” in the operations of tax administrations can introduce a complicated administrative challenge prone to errors (Forman, 2010[65]). In the United States, for example, between 22 and 26% of tax refunds related to the Earned Income Tax Credit are paid out in error, either resulting from errors in the reporting, owing to the complexity of the system, or intentional misreporting (Tax Policy Center, 2020[74]). One issue around income‑tested refundable tax credit is to assess entitlements in a timely manner, which “[…] involves a series of decisions on what to include as income, whose income in a household is included, over what time period income is measured, how long awards should last, and how responsive the system should be to changes in income and circumstances during the period of the award [.]” (Millar and Whiteford, 2020[75]). Given the different “timing elements” involved, it is easily imagined that payments made can be incorrect, requiring repayment or, instead, supplementary payments at a later stage, all at considerable administrative cost. Furthermore, in an environment, which is strongly focused on “tackling fraud”, this can lead to legitimate claims for support to be denied while excessive penalties on alleged wrongdoing can be applied. For example, an investigation by the Dutch Parliamentary commission on the implementation of the childcare supplement by the Dutch tax authorities found that many parents were wrongly branded as deliberate fraudsters (Parlementaire Ondervragingscommissie Kinderopvangtoeslag, 2020[76]).

The timing of awarding tax credits is another key issue. Tax credits, like tax liabilities may accumulate over the year and only be granted upon reporting with the annual tax return. This means that users of non-care household services pay for the service upfront and wait for the final annual tax settlement to obtain the relevant amount, as for example in Finland and Germany.1 While this system is common, it may exclude certain households, as paying upfront may be too expensive for low-income earners. In addition, tax payers are often not fully aware of tax incentives – for example through psychological frictions, such as programme confusion and informational complexity – leading to incomplete take‑up of tax credits (Bhargava and Manoli, 2015[77]; Chetty and Saez, 2013[78]).

Instead, tax credits can be “granted at source” and thus immediately reduce the service prices for consumers. In Sweden, the service fee already accounts for the tax credit, so consumers do not need to cover the subsidy upfront (see Box 4.2). France, which currently operates the household service tax credit through annual tax returns, will grant the tax benefits at source from 2022 (Ministère de l’Action et des Comptes publics, 2021[79]). Such granting at source makes the fee reductions more visible and reduces confusion, which can increase consumptions particularly among low-income households who would find it harder to pay the full service price upfront. Indeed, after the Swedish tax credit was changed to granting of the benefits at source in 2009, the number of consumers almost doubled (Skatteverket, 2011[80]). In addition, consumers tend to under-react to taxes that are not immediately visible and thus, conversely, may have stronger demand responses to granting tax credits at source (Chetty, Looney and Kroft, 2009[81]). However, awarding the tax credit at source requires service providers to undertake the necessary administrative actions. This does not pose insurmountable cost issues for service providers of a certain size, but would be difficult to stomach for small companies or those who work directly for the service user.

Box 4.2. The Swedish RUT-Deduction

The RUT-deduction (RUT-avdrag)*, which came into force in 2007, is a non-refundable tax credit for consumers of domestic household services. The tax credit applies to cleaning, maintenance, and laundry services in the household, as well as some other personal care services. The tax credit is generally available for services provided by registered companies or self-employed micro‑entrepreneurs.** Households that apply for the RUT-deduction receive a 50% reduction of the service price charged by service providers. The providers then request the difference between the actual payment received from the consumer and total labour costs, including VAT, from the Swedish Tax Authority. As such, the scheme is simple and straightforward for users, though it involves some additional administrative costs for service providers.

The tax credit reduces labour costs by half, up to a yearly maximum deduction ceiling of initially SEK 50 000 (PPP 5 632) for each person, while spouses can pool the deduction to increase the ceiling. In 2016, the ceiling was reduced to SEK 25 000 (PPP 2 834) for individuals below the age of 65, yet was increased to SEK 50 000 (EUR 5 632) again in 2019. Since 2021, the ceiling is set at SEK 75 000 (PPP 8 448).*** It is open to individual customers who are liable for tax in the given year and who hire a registered business, either self-employed or companies, to provide domestic services. These domestic household services need to be carried out in the customer’s household and cannot be undertaken by a relative.

Alongside RUT, Sweden grants the 30% ROT tax credit for repair and renovation services, up to a yearly deduction ceiling of SEK 50 000 (PP 5 632). Together with the RUT tax credit, the total deduction for both cannot exceed SEK 75 000 (PPP 8 448) per year.

Note: RUT is an acronym for Renhållning, Underhåll, Tvätt, which translates to cleaning, maintenance, laundry. ** The RUT-deduction can also be applied to direct domestic employment, but in this case it only applies to the employer contributions and not to the wages paid. As such, RUT-subsidised direct employment is relatively rare. ***

Source: AD-PHS (2020[82]); Skatteverket.

Indeed, when comparing the Swedish and Finish tax credits, which operate in relatively similar national environments, it appears that the RUT-deduction reaches more low-income households (see page Section 4.6.1). For example, while only 0.4% of Finish households with incomes below EUR 10 000 (PPP 11 849) received tax credits through the DHSTC, about 8% of Swedish households with comparable incomes below SEK 100 000 (PPP 11 264) received a RUT-reduction. As both countries have only small differences in their tax credit rates, and while deduction ceilings are likely not important at these income levels, this difference may well be related to the Swedish RUT-reduction being more easily accessible as it is granted at source.2

When designing tax credit schemes, governments also need to be clear in communicating the rules and regulations. In particular, frequent adjustments of tax credit rates and deduction ceilings may confuse households, which can negatively affect the efficacy of the instrument. For example, the Finish Domestic Help Service Tax Credit (DHSTC) has seen frequent adjustments since its implementation in 2001 (see Box 4.3). As a result, Harju et al. (2021[83]) finds that many Finns are unware of the relevant replacement rates and DHSTC ceilings as well as unable to calculate effective net prices accounting for the tax credit. Many recipients therefore frequently also claim less support than what they are entitled to. In particular, there is significant bunching of claims around the maximum credit ceiling, many claims refer to servcies bought up to EUR 4 500 (PPP 5 332), while in reality that amount is the maximum yearly reimbursement. When using a service provider an amount EUR 11 350 (PPP 13 450) could be claimed instead.

Box 4.3. The Finish Domestic Help Service Tax Credit

The Domestic Help Service Tax Credit (Kotitalousvähennys) introduced in 2001 is a non-refundable tax credit for domestic household services, including both care and non-care work as well as renovation activities.

The tax credit is available to all individuals with taxable income who do not already receive some form of service voucher or care allowances. It deducts 40% of the labour-costs for household and/or renovation services provided by private agencies, VAT-registered as well as non-profit organisations, or 15% of the wages and all employer social security contributions for directly employed workers. As such, it is available in both tripartite and traditional domestic employment settings, but service organisations are the predominant provider. In both cases, the tax credit only applies to costs above EUR 100 (PPP 118) and below a ceiling of EUR 2 250 (PPP 2 666) per person. Spouses can pool the tax credit, increasing the maximum deduction per household to EUR 4 500 (PPP 5 332). Individuals can claim the tax credit for services performed in their homes, including holiday homes. Work in the homes of their or their spouse’s parents or grandparents are also eligible, as long as the claimants themselves pay the costs.

Over recent years, the ceilings and deduction rates have seen much adjustment. For example, between 2009 and 2011 the credit for service consumed through provider organisations was as high as 60% up to a EUR 3 000 (PPP 3 341) ceiling per person.

Note: In addition, the Finish Government introduced a system of Social and Health Care Vouchers for vulnerable citizens, such as the elderly, in 2004. These vouchers are mainly applicable to health and care services using a more market-based mechanism, but can cover non-care household services, as municipalities have discretion over whom they wish to grant the vouchers to, and for what they can be used. Overall, the social voucher system is a very small part of the wider household service policy framework.

Source: Harju et al. (2021[83]); AD-PHS (2020[84]); Hiilamo (2015[85])

4.2.2. Social vouchers

Social vouchers, defined as social benefits entitling recipients to specific goods or services to improve their work life balance (see Ad-PHS (2020[86])), are another tool to promote the use of non-care household services. Such vouchers are usually transparent and easy to use, making them attractive to many households in need for support. In Belgium, for example, the voucher takes the form of a cheque that entitles the buyer to exchange it for non-care household service work (Box 4.4). Households can buy the vouchers themselves. In France, employers regularly issue pre‑financed CESU vouchers that entitle their employees to exchange them against household service work.

Social vouchers also make it easier to target support to target support at population groups with specific needs for household support. For example, the pre‑financed CESU voucher mentioned above can also be distributed by French local authorities and insurances to distribute social allowances to households in need (social CESU). In Belgium (and Italy, see Annex A), voucher prices are fixed, and each of the vouchers can be exchanged for one hour of household service work. This makes the costs involved transparent for the user and the underlying benefits are not dependent on claiming through complicated tax returns, which is likely to affect the overall use of household services. With the pre‑financed CESU in France, the value of the voucher is set by the issuing employer or local authority and can then be used to pay for services at the market rate. If the price of the services is higher, households have to pay for the difference themselves.

Box 4.4. The Belgian Service Voucher Scheme

The Service Voucher Scheme (Titres services or Dienstencheque) is a comprehensive voucher system for non-care household services, including cleaning, laundry and ironing, cooking and grocery shopping. Implemented in 2004, the vouchers are only available in a tripartite system for work provided by service provider organisations.

The system is organised in an interplay between multiple actors. Per year, each resident is eligible to purchase up to 500 service vouchers from a private contractor – Sodexo – who issues the vouchers at two prices fixed by the public authorities. The first 400 vouchers are sold for EUR 9 (PPP 11.88), while the remaining 100 vouchers can be bought for EUR 10 (PPP 13.20). The costs of part of the vouchers can be claimed against a tax deduction which varies across the three Belgian regions. It amounts to 15% on the costs of the first 163 vouchers in Brussels, 20% on the first 169 vouchers in Flanders, and 10% on the first 150 vouchers in Wallonia.1 Together, the tax deduction and two‑tier pricing system create a progressive support system in which households with lower service volume receive a higher average support per hour of service work consumed.

Households can pool the vouchers, raising the number of available vouchers to 1 000 per family. With the use of each voucher, consumers can pay for-profit and non-profit service provider companies for one hour of non-care service work in their household. Service provider companies – who directly employ household service workers eligible to provide these services – get paid a fixed rate, which marginally varies across the Belgian regions. This reimbursement amounts to EUR 23.50 (PPP 31.00) in Brussels, EUR 23.48 (PPP 30.98) in Flanders, and EUR 23.86 (PPP 31.48) in Wallonia.

1. The effective prices on the initial vouchers thus differ across regions: EUR 7.65 (PPP 10.09) in Brussels, EUR 7.20 (PPP 9.50) in Flanders, and EUR 8.10 (PPP 10.69) in Wallonia.

Source: AD-PHS (2020[87]); European Commission (2018[88]); Regional Governments of Brussels, Flanders and Wallonia

4.2.3. Declarative systems

An alternative approach involve declarative systems that aim to simplify administrative processes regarding direct domestic employment. Such systems establish a contract between service provider and user, and simplify the handling of social contributions and payroll records as these administrative tasks can be outsourced to external system managers. In France, for example, URSSAF, a private contractor, handles all administration regarding direct domestic employment after households fill out a declarative CESU form. A similar approach exists in Austria (Dienstleistungsscheck), Germany (Haushaltschecks) and Italy (Libretto di famiglia), where households can use declarations to buy services provided by marginal employees (for Austria and Italy; see Annex A). While such systems ease the process of hiring domestic employees, they do not absolve the household from entering a legal employment relationship that comes with certain obligations in terms of employment conditions and labour protections.

Like social vouchers, such systems also facilitate targeting of support at population groups with specific needs for household support. For example, the Chèque Emploi-Service in the Canadian province Québec is available to individuals who have experienced a loss of autonomy, such as the elderly or individuals with disabilities (see Annex A). The system, which functions very similar to the French declarative CESU, can be used to simplify administrative processes in hiring personal care or none‑care housekeeping activities, in conjunction with direct allowances granted to these groups (Boivin, 2017[89]).

4.2.4. Other instruments

There are a number of additional instruments (not covered in Table 4.1) that are used in the household service market. For example, countries can choose to fully or partially exempt actors in the household service market from employer contributions or other levies. In France, households benefit from an exemption of EUR 2 (PPP 2.74) per hour worked on employment contributions for their domestic employees, if they are directly employed by the household and not through an intermediary. A total exemption of these employer contributions is granted to households with members aged 70 or older as well as those with acute care needs. In the Netherlands, households are exempt from all taxes and employer contributions for directly employed care and (non-care) household service workers who work for four days or less per week (see Annex A). In Spain, direct employers of non-care household service workers can benefit from an exemption of 20% or 45% on their employer contributions to social security, depending on the size of the household (see Annex A).

While most measures discussed so far have been demand side measures, either reducing net consumer prices for services or easing the access to services, countries can also choose supply-side measures that reduce the tax burden on service providers. In addition to the comprehensive French policy framework, for example, service provider organisations are granted reduced VAT-rates on a range of their services.

Another approach is to reduce the high tax-burden on household service employees directly, by reducing or completely waiving taxation of their income. Along with potential, though often partial, social insurance coverage (see section 4.5), this incentivises formal employment as the difference between the take‑home pay in formal and informal work arrangements decreases. In Germany, for example, the so-called Mini-jobbers (marginal employees) with limited earnings are exempt from income taxation (Box 4.5).

Box 4.5. The German marginal employment scheme and household service tax credit

In 2003, the German Government introduced Mini-jobs: workers who earn less than EUR 450 (PPP 606) per month on a regular basis, or who work for less than 3 months or 70 days per year are exempt from income taxation. The minimum wage for these workers is EUR 9.50 (PPP 12.80) per hour. Under this scheme, the mini-jobbers are covered by accident insurance through employer contributions, but are exempt from health- and unemployment insurance. Personal contributions towards pension funds (3.7%) are voluntary on an opt-out basis.

Since the introduction of household cheques (Haushaltsschecks), hiring marginal employees under this scheme is made relatively easy for households, which functions relatively similar to the French declarative CESU (see above). After providing information on the employing household and the employee, the cheques are submitted to the Minijob-Zentrale, who handles the administrative details, including calculation accident insurance contributions. In addition, households can claim a tax reduction of 20% on their expenses on mini-jobbers up to a maximum refund of EUR 510 (PPP 687) per year. For other domestic employment relationships that do not qualify as marginal employment, a 20% tax credit with a ceiling of EUR 4 000 (PPP 5 286) applies for care and non-care household services, as well as EUR 1 200 (PPP 1 616) regarding renovation work.

Source: Ad-PHS (2020[90]).

4.3. Services covered

Within their policy frameworks, countries can cover a range of different household services. Variation in the household services covered by tax credits and/or vouchers can result from different policy objectives and /or target groups and differences in national contexts (Table 4.2). In general, household services can be split into object-centric non-care services, such as cleaning, laundry, cooking, gardening and smaller home maintenance and repair, or person-centric care activities, such as child-minding and support for the elderly and those with disabilities (Baga et al., 2020[3]).

Non-care household services such as cleaning and laundry services are covered in all five countries. When it comes to cooking, the Swedish RUT-deduction is the only instrument that explicitly excludes it from the list of eligible services and the Belgian service vouchers are the only instrument that excludes gardening as well as smaller maintenance and repair work in the household. The Belgian system is also the only one that currently fully excludes care services, while child-minding and care for the elderly and disabled is covered in all other four countries. The French system of household service policies also covers tutoring services, while the Swedish programme covers small-scale help with school homework.

In some cases, other services that are not easily grouped as non-care or care services are covered too: for example, IT services in the household are covered everywhere but Belgium, or transport services of either furniture or vulnerable people, which are only excluded in Finland. Renovation work may be covered by a separate scheme, such as the Swedish ROT-deduction. In contrast to all other schemes, the Finnish DHSTC and the German tax credit encompass renovation work besides household and care services with the household service framework. Given different national needs, Sweden and Finland also cover snow-shovelling services, while Germany allows for this in the context of road cleaning on private property.

Table 4.2. Services eligible within national household service policies

	Belgium	Finland	France	Germany	Sweden
Non-care Cleaning Cooking Laundry & ironing Gardening & small DIY Maintenance & repair Grocery shopping	Yes	Yes	Yes	Yes	Yes
	Yes	Yes	Yes	Yes	No
	Yes	Yes	Yes	Yes	Yes
	No	Yes	Yes	Yes	Yes
	No	Yes	Yes	Yes	Yes
	Yes	No	Yes	Yes	No
Care Child-minding Tutoring Care (other than child)	No	Yes	Yes	Yes	Yes
	No	No	Yes	No	Yes3
	No	Yes	Yes	Yes	Yes
Other Renovation work IT services Transport Snow shovelling & road cleaning	No	Yes	No	Yes	No
	No	Yes	Yes	Yes	Yes
	Yes1	No	Yes1	Yes2	Yes2
	No	Yes	No	Yes	Yes

Notes: The table only considers the household service policies as presented in Table 1. Each country may have other policy frameworks that cover other services, such as the ROT-reduction for renovation work in Sweden.

1. Transport is covered for the elderly and non-independent persons.

2. Transport is covered for moving purposes.

3. Tutoring is only covered for small-scale support with homework of children.

Given the focus on non-care household services, it is important to identify the importance of care‑service support within national policies. In the case of Belgium, this is straightforward as the service vouchers only cover non-care service work. In Sweden, approximately 91% of the RUT work are done in cleaning, gardening, and cooking services (Riksrevisionen, 2020[91]), while about 56% of all household service in German households work falls onto non-care tasks (Juncke, Krämer and Weinelt, 2019[92]).The breadth of the services covered through the DHSTC in Finland makes it complicated to compare it to other countries. According to Harju et al. (2021[83]), the majority of the tax credit volume, about two‑thirds of all claims, are applied to renovation work, while only about 28% of the total volume is applied to cleaning and housekeeping. In France, DARES (2020[71]) suggests that about one‑third of all work in the household service sector concerns non-care services.

4.4. Size of the programmes

Based on the differences in scope and generosity of the policies, as well as national contexts, the total size of the programmes in terms of employment and public spending also differs across countries. For example, systems with higher tax credit rates may attract more households and larger claim volumes than systems with a low tax credit rate, leading to more employment in the sector but also higher public spending. Likewise, a strong subsidy for social vouchers or other policy instruments can weigh heavily on the public budget while strongly boosting the size of the market. A further factor is the scope of services covered within each policy framework, as discussed in the previous sub-section. A narrow set of services eligible for either tax credits or social vouchers may lead to limited employment and relatively low public spending in the sector. Given the differences in scope, however, it is not straightforward to compare national policies and this sub-section highlights indicators on the size of the programmes supporting the use of non-care household services.

4.4.1. Public spending

Since the administration of the service vouchers was delegated to the regions in 2014, there has been no central tracking of the public expenditure on these instruments. Based on separate sources for 2017 however, the Belgian service vouchers scheme is the largest among all countries considered, costing roughly 0.48% of the GDP (see Figure 4.1). Due to the restricted set of eligible services in Belgium, all of this expenditure was on non-care work. Likewise, France, for which the available data is also fairly dated, also has considerably large public spending on fiscal incentives in the household service market (0.31% of GDP in 2014), with non-care services making up about two‑thirds of this. The public spending in Finland, Germany and Sweden is lower, with public expenditure on non-care services falling below 0.10% of GDP in each of these countries.

Figure 4.1. Public spending on household service policy instruments

Total public spending on major household service policy instruments as a share of GDP, non-care and other services, latest year available

Notes:

1. For Belgium, public spending sums the costs of the service voucher subsidy and the tax credits for service voucher costs.

2. For France, public spending includes fiscal incentives in the household service market, i.e. tax credits, VAT-reductions and exemptions on employer contributions. It excludes about EUR 5 billion of direct aids paid departmental councils and the pension and family allowance funds. About 90% of these expenses are going towards care activities.

3. For Sweden, non-care spending is based on the total expenditure on the RUT-reduction and a back-of-the‑envelope calculation using an average price for cleaning services of about SEK 200 per hour (after tax credit) and the fact that 91% of RUT work are performed in non-care household services.

4. For Finland, non-care spending is based on the total DHSTC volume and a back-of-the‑envelope calculation extrapolating from survey results in Harju et al. (2021[83]), according to which non-care household services make up 28% of all DHSTC work.

5. For Germany, non-care spending is based on the total expenditure on the German tax credit for household services and a back-of-the‑envelope calculation because non-care represents about 56% of all non-renovation household service work provided.

Source: Lebrun and Fourna (2016[93]); OECD calculations based on Vlaamse DWSE (2020[94]), IDEA Consult (2020[95]), Cour des comptes de Belgique (2018[96]), Parlement de Wallonie (2019[97]), Riksrevisionen (2020[91]), Verohallinto (2021[98]) and Bundesminsterium der Finanzen (2019[99]).

4.4.2. Employment

In contrast to public spending, it is harder to get a sense of the number of workers in non-care services, especially as the same worker might provide a range of different services in the household. Nevertheless, it is possible to get an idea of workers associated with the specific household service policy instruments that may cover non-care, care and/or renovation work (the policies effects on employment over time are discussed in Section 4.6.1). In France, for example, the breadth of household service policies has led to a sizeable sector. For 2018, DARES (2020[71]) reported that about 1.3 million workers provided approximately 850 million hours of service work, two‑thirds of which were in direct employment. Likewise, official German statistics do not differentiate between different care and non-care service types. Overall there were about 50 000 employees in service provider organisations in 2018, along with 304 000 marginal employees directly employed in households and 20 000 micro‑entrepreneurs in 2016 (Ad-PHS, 2020[90]).

In Belgium and Sweden, all or almost all work is provided in the non-care sector. According to the Belgian National Social Security Office, there were about 150 000 providers employed as non-care household service workers in the fourth quarter of 2019, providing about 128 million hours of work in Belgium (IDEA Consult, 2018[100]; ONSS/RSZ, 2020[101]). Despite only small in differences in population size to Belgium, the Swedish RUT-sector only employed 21 868 workers that provided 22.9 million hours of services in the sector in 2019 (Skatteverket, 2021[102]).

The most recent employment figures for Finland indicate that about 4 900 service workers were providing approximately 10 000 hours of service work in 2012, which indicate particularly small employment levels (Ad-PHS, 2020[84]). However, this is substantially smaller than estimations based on domestic cleaners and helpers in the Finish Labour Force Survey (approx. 20 000, see Figure 2.1) and employment levels in private provider organisations, who predominately offer non-care service work and employed about 30 000 workers in 2013 (European Commission, 2015[103]).

4.5. Work arrangements and working conditions

4.5.1. Tripartite set-ups

The tripartite model is the most common work arrangement in many countries. This model includes third parties, usually service provider organisations, which mediate the non-care service provision in households. The service provider organisations can take a variety of forms, for example, for-profit businesses or non-profit associations, but public authorities can also provide services. In each of these settings, households purchase the services from the provider organisations, who then send their employees to provide services in the consumers’ homes (Baga et al., 2020[3]).

Both the Swedish and the Belgian non-care household service policy frameworks are embedded in a tripartite system that generally does not allow for significant subsidies for direct domestic employment.3 As such, only services bought through provider organisations are generally eligible for tax credits or social vouchers. In Sweden, these providers need to be registered as paying corporate tax and are usually for-profit, half of which are micro‑entrepreneurs (Ad-PHS, 2020[82]). In Belgium, both for- and non-profit service providers are active in the market, along with some micro‑entrepreneurs (Ad-PHS, 2020[87]). In Finland, 90% of household services are provided through intermediaries, both for- and non-profit. While both private and public agencies provide household services, it is predominately private providers that operate in the non-care sector (Ad-PHS, 2020[84]).

In general, tripartite arrangements are linked to relatively good working conditions. In countries where tripartite systems are the norm, wages are typically set by collective agreements (wages for other countries, where bipartite employment is more common, are discussed in the sub-section below). For Sweden, these specify a minimum salary of SEK 21 347 (PPP 2 405) per month at 40 hours per week in 2019, while receiving the same access to social security and employment protection as any other employee the country (Bowman and Cole, 2014[104]). The few workers not covered under collective agreements directly negotiate wages with their employers. Domestic cleaners in Finland earn a minimum of EUR 1 790 (PPP 2 121) per month at 37.5 hours per week in 2019 (Eurofound, 2020[105]; Ad-PHS, 2020[84]; Ad-PHS, 2020[82]). For all workers in the Belgian social voucher system, wages are set by collective agreement determined in a sectoral-level “Joint Committee” between worker and employer representatives. About 90% of these workers are registered under Joint Committee 322.01 for accredited social voucher agencies, which fixed gross wages at EUR 11.35 (PPP 15.05) to EUR 12.06 (PPP 15.99) per hour in 2020 (depending on seniority in the sector). This is substantially larger than the minimum wage of EUR 9.65 (PPP 12.73), and the small share of workers covered by other collective agreements can receive somewhat higher wages. For all workers, full social security rights are guaranteed and more than two‑thirds of all contracts are indefinite, additionally guaranteeing at least 10 hours of work per week. The organisations that operate on the service voucher market are at all times requires to guarantee their employees well-being at work. As a result, the Belgian service voucher sector can be considered as “[…] the most regulated and protected […]” domestic service sector in the world (European Commission, 2018[7]; Jokela, 2017[58]; Ramos Martin, 2020[106]; Lens et al., 2021[107]).

As such, working conditions in the sector are comparatively favourable, even though there are risks of exploitation for vulnerable workers, especially migrants (Ollus, 2016[108]; Michielsen, 2018[109]). In addition, substantial increases in numbers of employees on sick leave highlight increased workloads over the years and the relative arduousness of the tasks (Ramos Martin, 2020[106]). In Belgium, for example, entering the service voucher system increases the likelihood of claiming disability benefits in the short and long term (Leduc and Tojerow, 2020[110]). Even if sufficient health and safety regulations are present, there is also a general lack in monitoring and enforcement of existing regulations regarding occupational health and safety (Mousaid et al., 2017[111]; van Gerven, 2020[112]).

The household service policy instruments in France and Germany operate in mixed systems, where both bipartite and tripartite work arrangements eligible for tax credits or social vouchers and exist side by side. In Germany, households directly employ about 81% of workers as marginal employees under the Minijob-scheme, while about 19% work in for- and non-profit tripartite arrangements (Ad-PHS, 2020[90]). In France, about half of the households employ service workers directly, while the other half buy services from for- and non-profit provider organisations or other intermediaries in a tripartite system. The latter can take various forms, such as associations, businesses, or public organisations. There are two different models among them, the mode prestataire, which represents normal tripartite providers, or mode mandataire, where an organisation handles the administrative tasks of hiring domestic workers for private households while not employing service workers themselves (Ad-PHS, 2020[67]).4 Overall, care work is predominantly provided by non-profit and public organisations, while for-profit organisation most often provide household service work (European Commission, 2018[7]).

Direct employment by households and its drawbacks

The bipartite model, also referred to as direct domestic employment, is another common formal work arrangement in which households hire service workers without an intermediary. In this case, the household acts as the legal employer and is therefore generally responsible for employer contributions (if the workers are covered). This work arrangement has been the most common model in numerous countries for many years. While still prominent in some countries, it has often been (partially) replaced by tripartite arrangements (Baga et al., 2020[3]).

The wages in bipartite employment are comparatively low. For example, directly employed French workers in the non-care household service sector earn on average about EUR 10 (PPP 11.84) net in 2012 (DARES, 2015[113]). Most of the German household service workers are entitled to the federal minimum wage of EUR 9.35 (PPP 12.60) per hour in 2020.5 In addition, there is a range of different collective agreements that apply in specific regional contexts or service categories (Jaehrling and Weinkopf, 2020[114]).

More so than on the tripartite side of the market, bipartite workers remain vulnerable in terms of job security and legal protection. This differs somewhat from bipartite employment relationships in other sectors. For example, for outsourced business support services, workers in tripartite arrangements are usually paid less than directly employed workers and may be subject to worse health, safety and labour standards compliance by their employers (see OECD (2021[115])). In the context of household services however, many countries exclude domestic workers employed by households from health and safety regulations typically offered to other groups of workers. While this is grounded in difficulties in the application of the regulations in the domestic context, it often leaves household service workers unprotected and exposed to occupational hazards (Martin Ramos and Ruiz, 2020[4]).

In France, for example, labour inspectors need to obtain the occupants permission before entering the household, which is a strong limiting factor in terms of guaranteeing safety and quality (Ledoux and Krupka, 2020[116]).6 In Germany, the Minijob-scheme exempts marginal workers from health- and unemployment insurance. At the same time, contributions towards pension insurance are relatively costly for employees in private households and often avoided due to the possibility of opting out (Jaehrling and Weinkopf, 2020[114]).

Related concerns apply in the United States, where many labour protections available to other groups of workers do not apply to domestic workers in bipartite work arrangements, such as the Occupational Safety and Health Act (Wolfe et al., 2020[117]). Even when workers are covered by state‑level labour regulations, there can be substantial enforcement barriers based on fears around immigration-based retaliation, as more than half of all non-care household service workers are, often undocumented, non-citizen immigrants (Wolfe et al., 2020[117]; Theodore, Gutelius and Burnham, 2019[118]). In addition, only a small number of domestic workers are covered by social insurance. In addition, only 7% of housecleaners receive employer-provided health insurance coverage and only 2% have employer-provided retirement plans (Wolfe et al., 2020[117]).

Direct employment is also linked to worse job quality. For example, even though most bipartite work contracts in France are of a permanent nature, employing households can nevertheless terminate them easily. Overall, Devetter and Lefebvre (2015[119]) note that high degrees of part-time work, frequently changing private employers, limited employment protection, and relatively low net wages make bipartite household services in France a precarious sector.7 However, while the French Labour Code explicitly excludes bipartite workers from many of its general provisions, most labour protections are provided through the Collective Agreement for Private Domestic Employees (Ledoux and Krupka, 2020[116]).8

Just as with in terms of working conditions, bipartite work arrangements may also entail worse service quality than in tripartite providers. As such, buying services from provider organisations ensures continuity of the service provision. For example, provider organisations can, for example, in case of illness send another worker, which is not necessarily possible for bipartite workers. Household purchasing services also avoid necessary administration tasks when employing a service worker (Ramos Martin and Ruiz, 2020[13]). Given that many countries have expanded the services eligible for tax credits and social vouchers over the years, tripartite systems may also be able to better react to this by diversifying their offered service portfolio, as they can easily hire new workers.

Other work arrangements

While bi- and tripartite work arrangements can be clearly classified, other arrangements fall somewhere between these two categories. For example, service workers may also register as self-employed micro‑entrepreneurs. In this case, the service provider organisation consists of a single service worker, who, in contrast to the bipartite employment, benefits from the same regulations as for service providers. Similarly, micro‑entrepreneurs sell services on the market, rather than being legally employed by the household. However, in terms of labour and social protection, the service provision through self-employment can resemble domestic employment. As such, this employment model shares aspects of both direct employment and service provider arrangements (Baga et al., 2020[3]; European Commission, 2015[12]).

In France, some businesses are micro‑enterprises, though they provide just about 1% of all hours worked in the sector (Ad-PHS, 2020[67]). In Germany, about 40% of tripartite employees are in fact self-employed, which excludes them from minimum wage entitlements. In addition, online service platforms have become an important player on the German household service market in recent years (Ad-PHS, 2020[90]). As laid out in Section 2.3, the exact employment form in this arrangements if often unclear, which can undermine job quality and labour protections for these workers.

Against the backdrop of widespread precarious conditions in the household service market – predominately for bipartite arrangements, but also for workers in provider organisations – the ILO Convention 189 on Domestic Work from 2011 stipulates a set of minimum rights for domestic workers (see Box 4.6). To this date, the convention has been ratified by 32 countries, including a number of OECD countries, prompting new and strengthened employment protection and social security provision. When Finland ratified the convention in 2015, household service workers were granted the same labour rights as other workers in the country. As such, regular working time was limited to 8 hours, in contrast to earlier legislations prescribing 9 hours as a standard. Workers were also covered by social security, much like the rest of the country (van Gerven, 2020[112]). In other cases, for example in Belgium, most of the prescribed labour protections were already present in the national legal framework for domestic work before ratification in 2015 (European Commission, 2018[7]; Ramos Martin, 2020[106]).9 However, even if ILO Convention 189 is ratified, some countries still have considerable gaps in the implementation and enforcement of guarantees as stipulated under the Convention, both regarding informal and formal workers (ILO, 2021[120]). For example, despite the general guarantee of equal treatment for marginal employees in Germany, deviations from these stipulations are widespread in practice (e.g. in terms of sick pay and paid holidays) (Jaehrling and Weinkopf, 2020[114]).

Box 4.6. ILO Convention 189 on Domestic Work

The ILO Convention 189 on Domestic Work was signed in 2011 with a strong majority in favour among delegates, after a 2‑year negotiation with substantial involvement of a broad coalition of workers themselves. The convention aims at ensuring decent work conditions for domestic workers and requires ratifying member countries to implement a range of legal protections. Among others, it stipulates the following guarantees:

• Protection from abuse, harassment, and violence (Article 5)

• Fair terms of employment and working conditions (Article 6)

• Sufficient information on employment terms (Article 7)

• Protections for migrant workers (Article 8)

• Equal treatment in terms of employment conditions (Article 10)

• Minimum wage entitlement, where applicable (Article 11)

• Safe and healthy working environments (Article 13)

• Equal treatment in terms of social security protection (Article 14)

• Labour inspections and penalties (Article 17).

The convention thus postulates that all domestic workers are entitled to protective labour legislations and social security provisions. As of today, it has been ratified by 32 ILO member countries and therefore demands often significant improvement in working conditions for domestic employees.1

1. Of the countries primarily discussed in this report, only France did not ratify ILO convention 189 yet.

Source: ILO Domestic Workers Convention – C189, Pape (2016[121]), ILO (2021[120]).

4.6. Professionalization and training

A critical aspect in ensuring high service quality and professionalization in the household service sector is the enforcement of qualification standards and provision of training. In most of the considered countries, there is no direct qualification enforcement for the non-care household service sector, however. As such, service quality regulations are much looser than in the care sector, which is often rather strict in qualification requirements. The only exception is Belgium, where service vouchers are limited to licensed service provider companies. While there are no strict qualification requirements for workers themselves, regional and sectoral training funds provide regular training. As such, 12 hours of training per year are compulsory for workers under the service voucher scheme. Training is subject to financial incentives and provider organisations can receive partial reimbursement on the training costs as well as extra subsidies from regional training funds for new workers coming out of unemployment or integration benefits. In addition, the sectoral training funds finance up to 18 hours of training for new workers employed under collective agreement 322.01 for accredited service voucher agencies (Ramos Martin, 2020[106]; Ad-PHS, 2020[87]).

Other countries provide voluntary certification or qualification systems for service workers and provider organisations. For example, a voluntary, yet not well recognised, quality charters for intermediary organisations exist in the Swedish RUT sector. Workers themselves can get certified by the Service Industry Professional Board (Servicebranschens Yrkesnämnd) after completing theoretical, oral and practical tests on cleaning and household services (Ad-PHS, 2020[82]).10 In Finland, the government established a 3‑year basic vocational qualification in Household and Cleaning Services in 2000 to improve professionalism in the sector, which was integrated with the Cleaning Service Qualification in 2010 (European Commission, 2015[103]).

It is also tripartite arrangements that offer better opportunities for training and upskilling of the service workers. In Sweden, many provider companies, especially larger ones, have their own training system that ensures quality provision of household services (Ad-PHS, 2020[82]). In addition, French service workers in provider organisations more often benefit from training through their employers (Devetter and Lefebvre, 2015[119]). In Germany, a conceptual qualification framework for new employees of service provider organisation has been developed by the German Society for Housekeeping. Within this framework, new workers can receive up to 12 weeks of structured training aimed at equipping them with the necessary skills for the job (Ad-PHS, 2020[90]; Jaehrling and Weinkopf, 2020[114]).

Resulting from widespread employment in bipartite arrangements, most of the directly employed household service workers in Germany and France have little or no professional training in domestic work however. In these arrangements, sufficient qualifications and standards can become secondary to price levels, which can negatively affect the professionalization of the providers. However, there are some mechanisms in place that aim to improve quality in the sector. For example, through collective agreements in France, private employers pay a direct contribution to the vocational training of their employees, who can thus access trainings in order to improve service quality. In the absence of formal qualifications, workers with a minimum of one year full-time work experience can have their skills recognised through a vocational certification (Validation des acquis de l’expérience), conditional on passing general and professional tests (Devetter and Lefebvre, 2015[119]; Baga et al., 2020[3]; European Commission, 2018[7]; Ad-PHS, 2020[90]; Jaehrling and Weinkopf, 2020[114]).

4.7. Effects of the policies

4.7.1. Consumption patterns

Non-care household service policies have undoubtedly created financial incentives to use declared over undeclared services. As such, the overall formal household service sector has generally seen growth in terms of consumption (see Figure 4.2). For example, for the Belgian service voucher, which is the most widely used incentive scheme considered in this report, the number of consumers steadily increased from 4% at its inception to about 23% of all Belgians by the end of 2016. This overall success may be due to a comparatively high elasticity of demand with respect to the price of household services (Desiere and Goesaert, 2019[122]).

Both Finland and Sweden have also seen increased household service consumption over the years, even though both differ in terms of how much of this can be attributed to the tax credits. With slow, but almost steady growth in recipients, the Finish DHSTC reached about 17% of households in 2019. Only when the generosity of the tax credit was reduced between 2011 and 2012 did consumption drop about 2 percentage points. Despite these trends, the growth in the number of households using the tax credits was weaker compared to neighbouring Sweden, where most users state they would not consume household services if it were not for the price reduction of the tax credit (Ad-PHS, 2020[82]). The RUT-sector itself has seen stronger growth over the years, reaching 19% of Swedish households in 2019. This may therefore reveal a larger demand elasticity with respect to the net price of the services than in Finland (Harju et al., 2021[83]).

Figure 4.2. Consumption patterns

Share of private households consuming household services with specific policy incentives, all eligible services

Note: The presented data divides the total reported data by the total number of private households as reported by Eurostat. Therefore, the share of actual household consumers might be different. For example, some households may have two or more members consuming tax credit or vouchers. Data for all countries, except for Belgium, also contains care related services, as well as renovation work for Finland. For Germany, the tax credit data concerns only household service work and not renovation services. The minijob data concerns only marginal employment in private households. For France, the tax credit data concerns tax filings for the deduction, but it is not necessarily the case that each claimant was granted a credit. For 2016 and 2017, there is no public data on the number of households receiving prefinanced CESU vouchers.

Source: OECD calculations based on Eurostat Household Data and Skatteverket (Sweden), Bundesrechnungshof (Germany), Minijob-Zentrale (Germany), IDEA Consult (Belgium), Suomen virallinen tilasto (Finland), Verohallinto (Finland), Direction Générale des Finances Publique (France), ACOSS (France), Direction Générale des Entreprises (France), Association Professionnelle des Emetteurs de CESU (France).

With the exception of two amendments of the tax code to allow for refundability, the French household service tax credit has not been subject to major changes. As such, its consumption has remained largely stable at around 14% of all French households. Both CESU instruments (declarative and pre‑financed) have been equally stable, but used by no more than 5% of all households.

There is limited public data on the number of households consuming the German tax credits over the years. However, the Federal Audit Office shows that the number of households receiving tax credits for household services increased by roughly 50% between 2008 and 2012 (excluding renovation services). About a third of these households would have used formal household services even in the absence of the tax credit (Bundesrechnungshof, 2016[123]). The share of private households hiring marginal employees for household service work has been about 0.8% between 2016 and 2019.

4.7.2. Consumption groups

Most household service policy frameworks aim – at least to some degree – to ease the housework burden for families, especially those with two earners and/or young dependent children. Given that these households are in most cases also earn more than an average household, one might expect that the fiscal incentives are more likely to reach better-off households. These patterns have been criticised by a number of researchers as there are concerns that the current systems predominately cater to high-income earners, effectively subsidising their leisure time (e.g. Nyberg (2015[124]) or Carbonnier (2015[72])). At the same time, one must note that there is an overall lack of criticism for tax subsidies that target work traditionally associated to men, such as tax credits for renovation work (e.g. the ROT-deduction in Sweden, which has comparable consumption patterns to the RUT-deduction). That the criticism is geared toward female‑dominated work may reveal the general undervaluation of these tasks relative to male‑dominated work.

It is indeed in the nature of tax credits that more often households of higher incomes claim tax credits and their deduction volume is in most cases larger than at lower incomes (Figure 4.3). For example, more than one‑third of the highest earners benefits from the RUT-deduction in Sweden, while also receiving service work of the highest volume (Figure 4.3, Panel A.). At the same time, a significant number of customers with medium-low incomes benefit from the tax reliefs that are of only slightly lower volume than credits for moderate income earners. In Finland, the average credit volumes are relatively stable and similar across the income distribution, while there is a much stronger gradient across income groups in terms of usage (Figure 4.3, Panel B.). Many of the households with lower incomes may receive social vouchers or care allowances that disqualify households from the tax credit, but they might also not be able to claim tax credits due to insufficient levels of taxable income. The higher consumption among lower-income households in Sweden may therefore be a result of granting the tax credit at source, which relieves households from paying the full service price upfront.

While there is no unified national data on consumption across income groups for the German tax credits and Belgian service vouchers, regional data, for example on Flanders (Belgium) and Rhineland-Palatinate (Germany), show similar patterns of predominant consumption by better earning households (Vlaamse DWSE, 2020[94]; Statistisches Landesamt Rheinland-Pfalz, 2018[125]). In Flanders, these numbers mirror earlier findings that highlighted the inability to reach a substantial share of less-affluent households and predominately cater to two‑earner household in full-time work (Marx and Vandelannoote, 2015[62]). For France, there is no data that breaks down consumption by income groups since the generalisation of refundability (for earlier data see DARES (2014[126])).

In terms of other household characteristics, those of retirement age are also more likely to use household services. In Sweden, for example, the retired were the biggest population group using the RUT deduction, reaching 20% of those above the age of 75 in 2018. Among all age groups, it is women who make the most use of the RUT deduction, especially between ages 35 and 59 (Skatteverket, 2020[127]; Ekonomifakta, 2020[128]). For Finland, Aalto (2015[60]) shows that those with particular need for household help, namely families with children and the elderly, make higher use of the DHSTC than the average household.

Figure 4.3. Tax credit consumption by income group

Source: Statistiska Centralbyrån (2020[129]), Verohallinto (2021[98]).

4.7.3. Household service employment

With increasing users, household service policy instruments have also led to sizeable employment in the sector (Figure 4.4). These employees might come from both previously declared and undeclared positions, or from entirely other sectors following increased attractively of household service employment (see e.g. Marx and Vandelannoote (2015[62])). However, the growth in the number of employees differs strongly across countries, which can be linked to differences in scope and, particularly, the generosity of the policies.

The French household service policies have, in some shape or form, been in place for about three decades. As such, employment in the overall sector has not seen major increases over recent years. In fact, by 2018 it fell to approximately 4.9% of all employment, down from 5.7% in 2010. This decrease is predominately due to a reduction of workers in bipartite arrangements, while employment in provider organisations is slowly but continuously increasing (DARES, 2020[71]). In contrast, employment in the Belgian service voucher sector increased (almost) steadily, despite only applying to no n-care services. These are making up 3.1% of the Belgian labour force in 2019, many of which work part-time while caring for their own household (ONSS/RSZ, 2020[101]). However, the dynamics of employment in the Belgian sector may not be perfectly comparable over time, as measurement has changed, for example by integrating those employed by provincial and local government services from 2017 onward. More complete data may be found in (Lens et al., 2021[107]).

Recent evidence shows that the Belgian service voucher system is drawing almost half of its workers from individuals with previously weak attachment to the labour market, such as the long-term unemployed or inactive as well as those with highly erratic employment. The positive employment effects are permanent, as workers entering the system are significantly more likely to be employed in the sector over the short and long term than comparable workers on the labour market (Lens et al., 2021[107]; Leduc and Tojerow, 2020[110]). However, a third of all workers are coming out of regular employment in other sectors, potentially entrapping qualified women, often immigrants, in jobs with low career prospects. As such, these hidden effects highlight that the service voucher system may be somewhat too generous (Lens et al., 2021[107]; Raz-Yurovich and Marx, 2018[63]; Marx and Vandelannoote, 2015[62]). A further problem for lower-qualified Belgians in finding employment is that labour migration may have crowded them out of the sector. Especially in Brussels, more than three‑quarters of all service voucher workers are foreigners, many of which are coming from Eastern Europe (Marx and Vandelannoote, 2015[62]; IDEA Consult, 2020[95]).

Figure 4.4. Employment related to household service policy instruments

Share of employees associated with specific household service policy instruments among total employed work force

Note: These figures here generally report employment related to one or more of the household service policy instruments. They thus differ from data reported through Labour Force Surveys, as displayed in Figure 2.1. As explained above, the latter are no perfect measure of employment in the overall sector and only include domestic cleaners and helpers. Data for Belgium refers to workers associated with the service voucher sector in the fourth quarter of each year. Data collection methods have changed after 2013 and 2017, thus the data may not be fully comparable over time. Data for Germany only concern only marginal employment in private households, which can be either care or non-care work. Data for France contain all household service employment, which covers both care and non-care work. This might be financed by tax credits, CESU vouchers as well as VAT or employer contribution exemptions. The data for Sweden only concern work in the RUT sector, which covers both care and non-care work. There is no sufficient data for Finland.

Source: OECD calculations based on Skatteverket (2021[102]), ONSS/RSZ (2020[101]; 2016[130]), DARES (2020[71]), Minijob-Zentrale (2020[131]), Enste (2017[132]).

Employment in the Swedish RUT-sector is on a substantially lower level, despite continuous positive trends. Over the last 10 years, employment in the sector has almost doubled, reaching 0.4% of the Swedish labour force in 2020. In total, estimates suggest that the RUT-deduction has directly led to the creation of at least 8 500 jobs, almost half of all employment in the RUT-sector (Tillväxtanalys, 2019[133]). While a primary motivation of the tax credit was to integrate female refugees into the Swedish labour market, recent evidence highlights that the majority of the foreign-born workers in the sector are temporary labour migrants from other European countries (Rickne, 2019[64]).

Unfortunately, it is only possible to measure the development of marginal employment in private households over time for Germany. Despite remaining relatively low, marginal employment roughly doubled since 2005, reaching 0.7% of the labour force in 2019 (Minijob-Zentrale, 2020[131]; Enste, 2017[132]).11 Likewise, it is complicated to obtain clear statistics on the number of Finnish household service employees over time. However, Harju et al. (2021[83]) suggests that the tax credit likely did not substantially increase employment in the Finish household service sector, much in line with similar evidence for VAT-reduction experiments in other labour-intensive service sectors (see e.g. Benzarti et al. (2020[134]) and Kosonen (2015[135])).

4.7.4. Undeclared work

With the before‑mentioned financial incentives for formal household service consumption, the incentives to consume undeclared services are likewise reduced. As such, there have been noticeable effects on the undeclared household service sector across the selected countries.

For example, a survey-based study of the Swedish Tax Authority shows the share of households consuming undeclared household services decreased from 6% in 2005 to 1% in 2019. As a result, RUT and the related ROT-reduction led to an approximate decrease in the consumption of undeclared services from SEK 11 billion (PPP 1.16 billion) to SEK 3 billion (PPP 0.34 billion) over the same time (Skatteverket, 2020[127]). Likewise, the French household service policies had a substantial effect on undeclared work in the sector as well. While half of all household service work was provided informally in 1996, it fell to 20% by 2015 (CRÉDOC, 2017[15]; DARES, 2014[126]).12 The reduction in undeclared employment in the sector was particularly strong over the first 10 years the initial Chèque Emploi Service and Titre Emploi-Service systems were active, but was further enhanced with introduction of the CESU (DARES, 2014[126]).

The effects of household service policies on the incidence of undeclared work is related to its capacity to set a competitive price on the formal market. In Sweden where substantial tax credits prevail, “after tax credit” prices of non-household services are relatively competitive and have led to sizeable reductions in informality. In contrast, undeclared household service work is widespread in Germany where the tax credits are relatively low and and have limited impact of the price of the services. As such, about 89% of all household service jobs in Germany were provided in an informal setting in 2017. This is only slightly lower than 94% in 2005, shortly after the introduction of tax credits and marginal employment schemes (Enste, 2019[8]). As a result, the Federal Audit Office strongly doubts the effectiveness of the existing German household service policy framework in reducing undeclared employment (Bundesrechnungshof, 2016[123]).

Despite previously estimating a reduction of undeclared work in the sector from 60% to 25% by 2010, the Finnish Ministries of Employment and Economy see methodological problems in measuring the deadweight effect of the schemes. Therefore there have been no recent evaluations of the DHSTC, especially regarding the extent of undeclared work in the sector (European Commission, 2015[103]). As the only recent study on these effects, Harju et al. (2021[83]) use the introduction of comparable household service tax credits at different times in Finland and Sweden, two countries with relatively similar economic development and demographics. The findings suggest that the tax credit likely had very little impact on decreasing the incidence of undeclared work and tax evasion.

In some instances, the reduction in the relative share of undeclared work may also be linked disproportionate increases in new consumers of declared services, rather than actual reductions in informal service provision. For example, survey data for Belgium suggests only a modest shift from undeclared to declared work (Marx and Vandelannoote, 2015[62]). Therefore, most of the current users of declared services are new consumers rather than previous users of undeclared services, as evident in Flemish survey data (Vlaamse DWSE, 2020[94]).

4.7.5. Female employment

A main motivation for household service polices is to relieve households from the burden of unpaid housework. Given that women predominantly bear this burden (see page 19) and the majority of the opportunities for economic value creation lies with them, most of these effects are to be expected among them.

There is some evidence on policy effects on increased female labour force participation. For example, Halldén and Stenberg (2018[136]) found that 40‑70 hours of RUT-deducted household service work per year proportionally increase the income and employment of married women in Sweden, who use 60% of their freed time in the labour market. In addition, positive employment and income effects of the tax credits seem to be concentrated among dual earner families with children, rather than for single parents and people above retirement age (Riksrevisionen, 2020[91]). For Belgium, the service voucher system appears to have had positive short- and long-term effects on women’s employment for both high- and low-skilled women as well as female unemployed (Desiere and Goesaert, 2019[122]; Raz-Yurovich and Marx, 2018[63]). However, estimations of employment effects of the social voucher scheme are not straightforward and and because of insufficient data have to rely on assumptions.

Evaluations on the effect of the French household service policies on female labour supply are rare. However, women in households that used domestic services were 9.4 percentage points more likely to be in employment than in the average household in 2005, increasing further with the number of dependent children in the household. While this may be tainted by reverse causation, it nevertheless exhibits a clear link between employment and household service consumption (Marbot, 2009[137]). At the same time, the consumption of household services gradually increases with educational qualification and the number of hours worked (Devetter, 2017[138]). This suggests that household services are an important substitute for unpaid housework in families were both adults engage in full-time and career-oriented work.

In both Finland and Germany, there is a lack of evaluations on household service policies and the labour force participation of women. However, about one‑third of German household service consumers report that they do so in order to have more time to engage on the labour market. At the same time, 22% of those without domestic help report that they would like to do so in order to increase their engagement on the labour market (Juncke, Krämer and Weinelt, 2019[92]).

4.7.6. Earn-back effects

At face value, household service policies can appear as expensive fiscal measures. However, a naïve focus on the cost of these instruments curtails potential earn-back effects that may substantially lower the net costs and, in some cases, could even lead to net fiscal benefits. These earn-back effects are of a direct or indirect nature. Reduced expenditure on unemployment benefits and increased income tax revenues and social contributions are directly earned back. Indirect effects relate to increased business activity as well as consumption and labour market substitution effects. Against this backdrop, it is important to account for such earn-back effects, even though their estimation is complicated and usually based on a number assumptions (EFSI, 2013[139]). While presenting the overall costs for household service policies, if possible narrowed down to non-care services in isolation, this sub-section thus also provides figures on earn-back effects wherever possible.

In practice, estimates often point toward budget neutrality of household service policies, such as in France and Belgium. In the French non-care household service sector, the earn-back effects were estimated to be EUR 70 million (PPP 87 million) larger than the public expenses in the sector in 2014 (Lebrun and Fourna, 2016[93]). After accounting for direct and indirect earn-back effects, the net cost of the Belgian social voucher in 2016 were estimated at EUR 1 200 (PPP 1 536) per full-job and thus remarkably small compared to the overall size of the policy intervention (IDEA Consult, 2018[100]).13 However, these calculations assume a particularly small role of deadweight, substitution and displacement effects of the social vouchers, leading to high estimated net job creation (e.g. 90% in Flanders, see IDEA Consult (2019[140])).14 This, in turn, results in an estimation of stark reductions in expenditure on unemployment benefits. In light of the strong labour migration into the service voucher system, particularly in Brussels, as well as the entry of many workers from non-subsidized jobs (see Section 0), these assumptions may prove to be too optimistic.

There is considerable doubt on whether the Swedish RUT-system is self-financing. Summing up direct and indirect earn-back effects, a deficit of approximately SEK 1.1 billion (PPP 0.12 billion), or 0.03% of GDP, remains. The real wedge between costs and earn-back effects may even be larger as this calculation excludes consumption effects. These can reduce the earn-back, as households shift their expenses toward household services and decrease revenue elsewhere (Riksrevisionen, 2020[91]). However, given that the Swedish household service policies appear to have reduced undeclared work and increased female labour force participation, despite missing the target of integrating refugees into the labour market, these moderate net costs may have been warranted.

There is no recent evaluation of earn-back effects of household service policies in Finland and Germany. However, given the limited effects mentioned above, Harju et al. (2021[83]) note that the costs of the Finish tax credits are likely higher than their benefits. Early estimates concluded a net benefit of almost EUR 150 million (PPP 154 million) for the overall tax credit in 2004 (Niilola and Valtakari, 2006[141]). A German pilot-project, was operational in 2 districts in the state of Baden-Württemberg between 2017 and 2019. The project introduced a social voucher for non-care household services, based on the Belgian system. The subsequent cost-benefit evaluation highlights potential earn-back effects, particularly in terms of tax revenue and social security contributions in the long term that are larger than the public expenditure on the social vouchers (Meier-Gräwe, 2019[142]). For potential increases in public subsidies, model simulations based on different scenarios on the overall German non-care household service market predict direct earn-back effects that are lower than the aggregated costs. However, increased returns through indirect effects may substantially increase the earn-back and make future implementations of substantial non-care household service policies viable in the mid to long term (Juncke, Krämer and Weinelt, 2019[92]).15

Notes

← 1.  The Finish tax credit can also be claimed through the tax card, right after service consumption. However, consumers still need to pay the full service price upfront.

← 2.  Another difference is that the Finish DHSTC is granted for renovation work as well. However, also considering use of the Swedish ROT-reduction that concerns renovation work does not fundamentally change the outcome: the Swedish tax credit seems more effective in reaching low-income households.

← 3.  Direct domestic employment instead benefits from RUT-deduction on employer contributions, which is a substantially smaller incentive than the tax credit on the work of service providers and generally not very popular (see Box 4.2).

← 4.  The mode mandatories only makes up about 10% of all tripartite work arrangements and has been decreasing over the years (DARES, 2020[71]).

← 5.  Workers in service provider organisation usually earn more, though this often makes their services too costly for many households.

← 6.  When services are purchased through provider organisation instead, labour inspectorates can be better able to monitor the workplace, even in the household, and thus ensure better protection for the workers (Ramos Martin and Ruiz, 2020[13]).

← 7.  The hourly prices in the non-care sector were on average EUR 21 in 2012, while the exact rates vary across provider type and service task.

← 8.  A notable exception is the provision of a 40 hour work week, in contrast to the 35 hour week specified in the General Labour Code.

← 9.  Germany ratified the ILO Convention 189 on Domestic Work, in 2013. Sweden did so in 2019 and it came into force in April 2020. France is the only country focussed on in this report which did not ratify the convention by 2021.

← 10.  Among others, these certification tests cover basics of cleaning work, organisation and finances, service quality, as well as health and safety (including ergonomics and chemical product safety).

← 11.  An evaluation of the tax credit for renovation and artisan services (also under the general German policy framework) showed only small employment effects in the relevant occupations (Walter Eucken Institut and Ernst & Young, 2013[162]).

← 12.  Earlier estimates for 1995, 2005 and 2011 are based on a DARES survey, indicating a stepwise decrease from 50% to 25%, while estimates for 2015 are based on a CRÉDOC survey. It is thus not entirely clear whether the undeclared household service sector decreased further between 2011 and 2015, or whether this reflects differences in survey methodologies.

← 13.  The net costs per full-time job vary between the Belgian regions, primarily as tax credits vary between them. As such, the net costs per full-time job were negative in Brussels and Wallonia, while they were positive in the most populous region Flanders.

← 14.  The implicit subsidies in the social voucher system may go to workers who were already employed in the system (deadweight effect) and may lead to the loss of jobs in sectors that compete with the service cheque sector (displacement effect) or for groups not directly targeted by the subsidy (substitution effect).

← 15.  Earlier simulations estimated that subsidies in the non-care household service market would lead to earn-back effects that are substantially larger than the cost (German Federal Ministry of Family Affairs, Senior Citizens, Women and Youth, 2011[163]).