Childcare-policy responses in the COVID-19 pandemic: unpacking cross-country variation

One of the many ways in which the current COVID-19 pandemic has been unique in the history of modern welfare states is the closure of early childhood education and care services (ECEC) and schools, which took place in most countries around the globe as from mid-March 2020. In some countries, ECEC/schools’ re-opening started in May or June, while others opted for a prolonged closure until the new school year. With care and home-schooling responsibility thus being ‘shifted fully onto parents’ (Yerkes et al. 2020: 4), these closures have raised concerns about the increased risk of social exclusion and growing social inequalities in children's educational opportunities (OECD 2020a), as well as about growing work-family conflicts and gender inequalities. German sociologist Allmendinger drew a looming picture in early May, warning that the Corona crisis might set back gender-equality progress by no less than 30 years.¹ In this respect, emerging evidence shows a nuanced view: working mothers are indeed at higher risk of being overburdened with the re-familiarised care work, shortening working hours, even quitting their job, or being dismissed (Czymara et al. 2020; Andrew et al. 2020). Yet women are also overrepresented amongst ‘key workers’, and fathers have increased their share of care- and housework during lockdowns, e.g. in the US (Carlson et al. 2020) and the Netherlands (Yerkes et al.2020). Without doubt, both working mothers and fathers are under increased pressure under COVID-19 vis-à-vis people without care obligations – with national ECEC/schools-closure ‘strategies’ playing a vital role in their ability to (continuously) work, especially as the greater illness risk for elderly is limiting other care alternatives such as grandparents.

Against this backdrop, there is a need to understand national crisis responses and their implications for various groups in society. The cross-country variation in immediate crisis responses is indeed striking (Capano et al. 2020; Yerkes et al.2020), as in the novel COVID-19 pandemic situation the knowledge base required to design interventions has been highly uncertain, while governments needed to react to an urgent, and severe crisis (t’Hart et al. 2001). Such conditions of uncertainty and urgency particularly applied to decisions regarding ‘lockdown’ – including ECEC/schools’ closures – due to lacking evidence-base on disease transmission, and its severity for children that is still highly debated (Mallapaty 2020). Countries’ responses have differed widely in type and timing, going far beyond the categorisation of keeping public ECEC and primary schools ‘opened’ (e.g. Sweden) or ‘closed’ (e.g. Italy) (UNESCO 2020; OECD 2020b): Rather, many countries exhibit hybrid approaches, varying not only from more lenient to strict, but also from universal to selective closures/re-openings, i.e. allowing certain groups to keep access or re-enter earlier. In such an uncertain situation, countries’ approaches may be affected by factors that go beyond public-health concerns (e.g. specific regime paths, government constellations) and have different implications for the groups targeted (or not), and therefore gender and social inequalities. Yet, until now, such patterns are far from evident and additional research is needed. Importantly, we first need to conceptualise the similarities and differences in immediate ECEC/schools’ closures to allow for future research into the factors that account for distinct policy choices, as well as their effects.

This contribution, therefore, aims to explore cross-country variations in the closures and re-openings of ECEC services and (lower grades of) primary schools (hereafter called: childcare-policy responses), which occurred in relation to the first phase of immediate crisis response. It presents a conceptual framework that allows to ‘unpack the design’ (Capano et al. 2020: 3) of immediate childcare-policy responses under COVID-19 and explore their cross-country variations. The focus is on education- and care-services up to age 12 approximately, as during that age the need for childcare and parents’ guided home-schooling is significantly increased, which is also reflected in out-of-school-hours care being typically provided up to age 12 (Plantenga and Remery 2013). We argue that specific childcare-policy responses are resulting from the country-specific combination of pandemic prevention strategy and childcare-policy concerns. These theoretical considerations are supported by a summative overview of modalities of ECEC/school closures and re-openings in European countries, with typical country-cases being empirically illustrated (see Appendix 1). The presented framework provides a basis for future research into the gender and social consequences of the COVID-19 pandemic.

It bears repeating that while there is ample analytical and theoretical ground to comparatively investigate childcare policies, the situation under COVID-19 is unique and novel. Existing conceptualisations – such as classifying childcare policies by their type of familialism (Leitner 2003), facilitation of parents’ capabilities (Yerkes and Javornik 2019), or inclusiveness in terms of eligibility (Dobrotić and Blum 2020) – may indeed contribute to unpacking cross-country variations in childcare-policy responses under COVID-19. However, the global closure of ECEC/schools – and therewith the restriction of fundamental social rights – is unprecedented in the history of modern welfare states. Indeed, childcare-policy responses in the pandemic situation transverse previous conceptualisations, because they became (primarily) driven by public-health-related goals, which are usually not in their core focus. As detailed below, countries chose a specific pandemic-prevention strategy which prompted the initial ‘shock’ response also in childcare/educational policies. After those initial responses, countries began to balance – in different pace and patterns – public-health with other, sometimes competing goals and concerns more specific for childcare- and education-systems (work-family reconciliation and employability, equal educational opportunities, etc.) (cf. Moss and Kamerman 2011; Scheiwe and Willekens 2009).

Against the new and uncertain risks of the Coronavirus, countries chose different prevention strategies. Generally, public-health research distinguishes between the high-risk (i.e. targeted) and the population approach to prevention (Rose 2001). While the first strategy targets high-risk individuals (e.g. ‘the elderly’ for COVID-19) and avoids generalised interventions, the second strategy addresses the whole population (e.g. through curfews and social-distancing rules for COVID-19). There can also be mixed approaches which combine both strategies. The high-risk approach may fulfil its goals of protecting targeted groups, but its prevention potential is lower, and it risks to be exclusionary (cf. Rose 2001). The population approach is generally associated with a stronger prevention potential, but also a higher risk of the ‘prevention paradox’, where ‘large overall health gains for whole populations […] might offer only small advantages to each individual’ (WHO 2002: 147).

These two approaches were also reflected in earlier school closures and other interventions aimed at controlling influenza and previous coronavirus outbreaks (e.g. SARS). Yet there were inconclusive results of epidemiological and public-health research on the effectiveness and prevention potential of different ‘school closure practices’, e.g. ‘national, regional, local, or reactive closure of individual schools’ or ‘less disruptive social distancing interventions’ (Viner et al.2020: 397). As Cronert (2020: 2) argued, the timing of school closures reflects government's choice of prevention policy, namely between

Regarding the latter, Scheiwe and Willekens (2009: 4) distinguish between two main motives behind the ECEC development in Europe – ‘a need of public education’ and ‘reconciliation of care work and paid work’ – connected to different notions on gender and social relations. The educational model builds on both the idea of young children's need for public education and a need to overcome social inequalities in children's educational opportunities; it thus tends towards universal coverage of children. The work-care reconciliation model was built on the need to protect children of working mothers (early nineteenth century) and, more recently, to enable parents (mainly mothers) to enter the labour market; it therefore often includes a gender-equality orientation. Bacchi (1999) also identifies a specific welfare approach of ECEC development, which targets families ‘in need’ (e.g. through ECEC subsidies for low-income families), and is closely related to the goal of mitigating social inequalities. The latter approaches tend to be more selective, i.e. primarily focused on those in need of care (e.g. children with both parents/single parent in employment). As usually with ideal types, these models are not mutually exclusive; they can, however, be used for heuristic purposes, i.e. as ‘ideal types from which particular sets of organisational and institutional principles can be derived and from which different kind of questions and problems follow’ (Scheiwe and Willekens 2009: 4). As discussed below, they can be indeed helpful for conceptualising different childcare-policy responses under COVID-19, and their variation over time.

Against that backdrop, we argue that the immediate ‘crisis’ childcare-policy responses were related to the chosen pandemic prevention strategy in a country, either high-risk (targeted) or population approach (Figure 1). Moreover, closure and even more so re-opening responses are reflective of how the public-health goals have been balanced with the partly conflicting ‘core’ childcare-policy concerns. Responses within this framework are thus theoretically-informed, yet – as discussed above – there is a lack of theoretical assumptions on pandemic, public-health driven childcare-policy responses. Therefore, the conceptual development also needs to be informed by the variety of empirical realities. We collected data on ECEC/school closures and re-openings modalities in the March-July 2020 period in 28 countries of the European Economic Area² (Appendix 1).

The distinguished ‘pandemic childcare-policy responses’ should be seen as ideal-types that may empirically exist in hybrid forms, and also change over time. Such a conceptual approach allows to unpack the patterns of responses, but also their pace. Namely, we can assume that pure high-risk- to hybrid-strategies to prevention opened more ‘space’ for specific childcare-policy goals than the hybrid- to whole-population-strategies – as children are not amongst the high-risk groups of COVID-19 (as they are for other diseases, which would make school closures more likely also under high-risk approaches). The realisation that the pandemic would be ‘here to stay’ for some time, and of the adverse effects of extended ECEC/school closures (e.g. child welfare, social isolation, nutritional problems, economic harms; Viner et al.2020), have slowly opened space for specific childcare-policy goals also in countries that initially opted for whole-population strategy and fully closed ECEC/schools. The result was a higher frequency of lenient responses in re-opening phases, but also a greater diversity of re-opening approaches. Next, we will detail and empirically illustrate the immediate policy responses displayed in Figure 1.

To unpack the defining lines of pandemic childcare-policy responses, it is not sufficient to ask whether ECEC/schools were ‘closed’ or remained ‘open’. Instead, there is a complex and partly interrelated set of characteristics that form different and more hybrid policies. First, the responses often show differences between ECEC services, primary schools, and secondary schools. Although such different regulations are relevant, it is their sequence that forms a specific childcare-policy response (e.g. first re-opening for younger vs older pupils vs ‘critical’ years in educational process). Second, childcare-policy responses have differed decisively between ECEC/school closures and their re-openings. Therefore, we can identify distinct responses of (1) closure, and (2) re-opening (Figure 1), which vary not only from a lenient (fully-open) to strict (fully-closed) approach, but also in their universal to selective nature. Namely, asking for whom ECEC/school was closed or (re-)opened can help to uncover childcare-related priorities and inequalities, as more often than not, certain groups kept access, or access was available to them earlier or more comprehensively.

In the first state of immediate crisis reaction, the influence of public-health concerns and different prevention strategies over childcare-policy responses was strong. Under the pure high-risk approach, we could find a targeting of high-risk groups in elderly-care homes, while ECEC/schools remain open. Yet we see an integration of high-risk prevention into childcare-policy responses through less ‘drastic’ interventions, e.g. stricter rules of attention (e.g. stay-at-home rules in case of mild symptoms such as blocked nose) or offering alternative distance-schooling for children at-risk. With Sweden (see also Pierre 2020) and Iceland there are two European countries which exhibit such a pure high-risk response; while few countries (e.g. Finland) closed primary schools but kept ECEC open (Appendix 1). The strategy is lenient, but carries exclusionary risks as e.g. children (in families) with pre-existing conditions cannot attend ECEC/school.

Vice versa, under the population approach, strict containment policies (see stringency index in Appendix 1) translate into ECEC/school closures. These closures, witnessed in many countries, have tended to be full (and in that sense, universal). Yet they typically uphold ‘emergency childcare’ for key workers (e.g. parents working in hospitals, food supply, police). This reflects the dominance of the public-health focus (‘keeping the system functioning’); however, there is also variation (e.g. in Germany, the federal states applied narrower or wider definition of ‘key workers’).

Then there are ‘hybrid’ forms that consist of partial (and targeted) closures, aiming to include other (childcare policy) goals. In a number of countries (e.g. Austria, Belgium, Latvia), we find indications of a work-family perspective as ECEC remained open for ‘parents with no other care option’ (Appendix 1). Still, with strict containment policies in place, where childcare remained (partially) opened, parents were typically encouraged to use it only if absolutely necessary. This was reflected in low participation rates (e.g. in Austria only 1–2% in March/April 2020; see Der Standard 2020). Full and partial ‘closures’ alike have been contentious, due to their potentially adverse effects on the economy, parents’ employment, children's education and health, or vulnerable grandparents providing childcare (OECD 2020a; Viner et al.2020).

Under the re-openings, the variety of childcare-policy responses increased, as these have developed into more long-term strategies. In pandemic prevention terms, countries have been moving from ‘the hammer’ (immediate closure to slow the infection) towards ‘the dance’ (Pueyo 2020) of lifting confinements or re-applying them in a more targeted way to keep the virus under control. In re-opening, childcare-policy responses have become increasingly reflective of differing priorities and perspectives. They could also be related to the Coronavirus incidence in a country (e.g. school openings being dependent on low infection numbers), yet immediate responses do not seem entirely determined by such ‘problem pressures’ (cf. France re-opening schools in May despite comparatively high infection rate vs Ireland opting for school closure until the new school year).

There are countries where immediate childcare-policy responses in the re-opening stage remained dominated by the population-approach and public-health concerns. Accordingly, those responses are all universal in style and rather strict. Strictest countries (e.g. Italy) have kept ECEC/schools closed at least until the end of the school year (Appendix 1). Others have first extended their emergency mode by applying a broader definition of ‘key workers’ to include more children (e.g. Germany), or allowed for individual face-to-face teaching activities (e.g. Estonia, Hungary).

Vice versa, re-opening could be lenient and consist of full re-opening to a regular mode for (in principle) all children. In some cases, this was the ‘immediate’ re-opening response (e.g. Denmark, Finland, Greece; Appendix 1). Such a ‘quick’ reopening may have happened because countries shifted towards a high-risk approach (e.g. in the face of increasing economic costs of ‘lockdown’), or infections decreased at rates that allowed them a lenient approach. Also, balancing with childcare-specific goals can be important here. For example, Denmark's early re-opening of ECEC/schools by 15 April was linked to the care needs of working parents (Tagesschau 2020).

However, many countries have not moved from some form of closure to ‘full re-opening’; rather they move there through (a sequence of) partial re-opening. This more hybrid, mixed type of ECEC/schools’ re-opening evokes further research interest, as different childcare-specific perspectives become very visible. The interesting difference between responses here is not on the lenient to strict end as those are all forms of partial re-opening, but on the selectivity dimension of who gains (earlier) access. First, there is a universal and more lenient response: Countries here have gradually re-opened ECEC/schools for all children, but in a partial (reduced) re-opening that is not primarily care-focused and aims at lowering contacts and enforcing ‘social’-distancing rules (e.g. through weekly/daily shifts with a maximum number of pupils in class, e.g. primary schools in Austria and Belgium, in Luxembourg also ECEC; Appendix 1; OECD 2020b). Second, there are partial (targeted) re-openings, where certain groups of children were prioritised (e.g. youngest children or dual-earners, children-at-risk). We identified four goals at the conceptual level, which go along with different targeted re-openings and childcare possibilities:

1. public-health focused, where only older pupils are allowed back to class, who can keep ‘social’-distancing rules, while ECEC services remain closed;

2. education-focused, where transition years are allowed back to class (e.g. older students taking school-leaving exams, younger children in last pre-school/primary school year);

3. social-inequality focused, where ‘disadvantaged’ children gain earlier access (e.g. from families social-assistance receivers, or asylum seekers);

4. work-family focused, where ECEC and (lower classes of) primary schools re-opened first (i.e. before secondary schools), or certain parent-based criteria are developed allowing earlier access (e.g. employed single parents, dual-earners).

While the latter (4) response may be the opposite of the (1) public-health focus; other foci can also be combined (e.g. Germany), but concrete indicators need to be developed to disentangle those approaches and classify (hybrid) cases based on empirical data.

Finally, to fully grasp different policy responses, we also need to consider their timing (e.g. starting date and length of closures), and sequence (e.g. re-opening starting with a certain response, then moving to another response). Moreover, in some countries (e.g. Germany, UK, Spain), at least parts of the closure and re-opening responses were made on a regional basis, so this variety must be taken into account. Not least, the latter phenomenon is of growing importance for future research, as pandemic policies, in general, have been moving from ‘general lockdowns’ to the targeted intervention into local outbreaks.

The presented conceptual framework can provide a basis for looking closer into the empirical reality of childcare-policy responses to COVID-19, and investigating their drivers and effects.

Our overview of responses identified lenient to strict, and universal to targeted approaches in European countries (Appendix 1). Yet the hybrid policy approaches –particularly reflective of both public-health and ‘core’ childcare-policy goals – will require further ‘unpacking’ through developing detailed indicators able to grasp their prevalence and sequencing across countries. Moreover, when moving from high-risk to population-strategies in pandemic prevention (and therewith: to fuller ECEC/school closures), the need for ‘pandemic parental leaves’ and benefits increases. While it went beyond this article, it will be crucial to consider both ‘main pillars’ of childcare policies, i.e. how ECEC/school closures were integrated into a coherent approach (or not) with accompanying (pandemic) leaves and/or benefits.

As regards countries’ combinations of pandemic prevention strategies with childcare-specific goals, the causal mechanisms behind those choices have yet to be uncovered. Therein, it will not only be important to investigate whether countries followed traditional childcare-regime paths (e.g. Leitner 2003) in crisis responses, but also whether in the long run, the COVID-19 crisis triggered the introduction of new (or even old) ideas in childcare policies. Not least, besides their potentially adverse effects (e.g. OECD 2020a), there is an indication that in some countries (e.g. Canada), the pandemic raised the importance of childcare for the economy and gender equality, putting it higher on the political agenda.

Finally, investigating unequal effects of pandemic childcare-policy responses will require asking whether they were also shaped by ‘blindspots towards the vulnerabilities of certain population segments’ (Capano et al. 2020: 1). Blindspots of childcare policies could result, for instance, from mentioned exclusionary tendencies within high-risk approaches (e.g. excluding children with pre-existing conditions from ECEC/schools), or from varying ‘inclusiveness’ (Dobrotic and Blum 2020) of pandemic childcare policy eligibility (e.g. children of non-working parents ineligible for ECEC; self-employed ineligible for pandemic leave benefits). Childcare policies were ill-prepared for COVID-19, and short-term responses often bridged the way towards developing longer-term solutions for integrating all children in the ‘new normal’ ECEC/school under pandemic conditions. Future ‘waves’ of COVID-19 (or other pandemics) will bring about new closure/re-opening sequences that are becoming increasingly managed by regions and municipalities according to local infection numbers. For future strategies, an understanding of the initial responses and their effects can contribute to policy learning. Furthermore, the immediate childcare-policy responses to COVID-19 may serve as a burning glass to see which priorities countries set during the closures/re-openings, and therewith how entrenched and enduring different ideas (e.g. work-family reconciliation) are in national childcare policies in times of severe crisis.

We'd like to thank the two anonymous reviewers, and the editors for putting together this highly topical Special Issue on fast-track. Further thanks go to Mario Špundak and to Johanna Kuhlmann for fast-track, critical reading of earlier drafts of this article and for providing – as always – most valuable comments. Ivana Dobrotić's research is part of the project that has received funding from the European Union's Horizon 2020 Research and Innovation programme under the grant agreement No786826.

No potential conflict of interest was reported by the author(s).

Ivana Dobrotić is Marie Curie Fellow at the University of Oxford, Department of Social Policy and Intervention, running the project ‘Social and Gender Inequalities in Care: Childcare-Related Policies and Parenting Practices in the Post-Yugoslav Countries and the Role of Policy Ideas’. She is affiliated as an Associate Professor of Comparative Social Policy to the University of Zagreb. Her research interests include comparative social policy, family policy, gender, and social inequalities. She is co-editor of the Annual International Review of Leave Policies & Related Research.

Sonja Blum is Research Associate and Lecturer at the FernUniversität in Hagen, and Fellow at the KU Leuven Public Governance Institute. Her specialisation is in public policy analysis, comparative social policy, and family policy. Recent publications include the co-edited Routledge Handbook of European Welfare Systems (2020), and she is co-editor of the Annual International Review of Leave Policies & Related Research.