Global social policy ideas in the COVID-19 crisis: Ideational change and continuity in the ILO, the OECD, the WHO, and the World Bank

Abstract

Global crises constitute challenges for social policy. While social policy is predominantly a national concern, international organisations (IOs) contribute frames of reference for state decisions. In this article, we explore whether the COVID-19 pandemic led to changes in IOs’ social policy ideas and recommendations in health care, labour market, and social protection policies due to how IOs perceived the crisis’ specific nature, severity, and global scope. We focus on four IOs regarded as key actors in global social policy, namely the ILO, OECD, WHO, and the World Bank. Theoretically, we employ a framework of ideational policy change combining different levels (recommendations – including parameters and instruments – and paradigmatic ideas) with different types of change (layering, conversion, dismantlement, and displacement). We find that IOs have not fundamentally reimagined their pre-pandemic stances during the pandemic. The IOs’ perceptions of the crisis do not undermine IOs’ ideas and recommendations but highlight their appropriateness.

Introduction

The COVID-19 pandemic has not only put an extraordinary strain on health sectors worldwide but – through the economic crisis that it brought in its wake – also challenged labour markets and social protection systems alike. While most research on social policy responses has focused on country cases (see e.g. Béland et al., 2021; Cantillon et al., 2021; Devereux, 2021), less attention has been paid to how international organisations (IOs) have dealt with the crisis and its implications (but see Debre and Dijkstra, 2021; Leisering, 2021; Van Hecke et al., 2021). However, on the one hand, IOs routinely develop policy proposals for crisis-induced risks and threats. Their ideas frequently become blueprints for national crisis management, either directly translated into policies (policy transfer) or indirectly becoming frames of reference (Olsson and Verbeek, 2013; Kaasch, 2016). On the other hand, whether crises affect policymaking depends on how they are perceived (Hay, 2001; Rosenhek, 2013). We ask: did the COVID-19 pandemic affect IOs’ policy recommendations and broader paradigmatic ideas, or did these remain consistent compared to pre-pandemic times?

Based on existing research about political action in crisis contexts, change would be expected. After all, crises constitute potential turning points as they call into question existing ways of operating (Kuisma, 2013; Van Hooren et al., 2014; Bali et al., 2022). However, there are gaps in scholarship on policy shifts due to crises: Most scholarship discusses changes at the institutional – and not ideational – level, and views crises frequently as objective triggers of change rather than focusing on crisis perceptions. To identify ideational change or consistency, we apply a theoretical framework adapted from historical institutionalism (HI), combining different levels of ideas (recommendations – parameters and instruments – and paradigmatic ideas) with different types of change (layering, conversion, dismantlement, and displacement). Based on this framework, we investigate whether there is path dependency on the level of IO ideas and recommendations despite the crisis, where IOs develop policy proposals in line with their existing ideological positions, or whether IOs depart from their traditional approaches. Using qualitative content analysis, we analyse IO ideas and recommendations in the fields of healthcare, labour market, and social protection during the COVID-19 crisis, in addition to IOs’ crisis perceptions. We focus on four major IOs in global social policy: the International Labour Organization (ILO), the Organisation for Economic Co-operation and Development (OECD), the World Health Organization (WHO), and the World Bank – key players in global social policy and the subject of extensive study (Deacon et al., 1997; Martens et al., 2021).

While the academic literature on COVID-19 policymaking is slowly evolving, work on IOs’ responses usually refers to the first wave of the pandemic (Debre and Dijkstra, 2021; Leisering, 2021; Van Hecke et al., 2021). By analysing IO documents from the early phases of the pandemic up to late 2022, we are able to observe almost the entirety of the COVID-19 crisis, taking May 2023 as an endpoint, as proclaimed by the WHO (UN, 2023).

The article begins by presenting how we apply theoretical perspectives on different levels of ideas and types of change to IOs. After a methodological section in which we describe our approach, data, and case selection, we present the empirical results on the ILO, OECD, WHO, and the World Bank. We conclude with a discussion, comparing IOs’ responses of ideational change or continuity in crisis settings. We find that IOs are persistent in their ideas and recommendations on social policy despite the most severe pandemic circumstances in recent history, a finding we explain by reference to their framing of the crisis, which allowed for ‘business as usual.’

Theoretical framework: IOs and ideational policy change

Despite the urgency and uncertainty of the COVID-19 pandemic, IOs were confronted with high expectations to provide guidance commensurate with their competencies. As in previous global crises (Kaasch, 2016; Leisering, 2021), their proposals have the potential to guide states to respond to situations perceived as deeply threatening, uncertain, and urgent (Olsson and Verbeek, 2013; Boin et al., 2016). They are expected to coordinate policies between states, serving as intermediary actors in networked relationships (Haas, 1992; Abbott et al., 2015). Crises may thus require an IO to update its policy proposals in light of these expectations, and even broader ideological positions may be in jeopardy. Scholars treat such ideational change in crisis conditions as common, initiated, e.g. through policy entrepreneurs (Kingdon, 1984) such as IOs. In institutionalist research, ideas are mostly treated as variables to explain institutional shifts, which are otherwise characterised by much more continuity and path dependency (Pierson, 2000; Carstensen, 2011; Béland and Powell, 2016). Cases of ideational change in IOs indicate that crises may lead to far-ranging departures (Allan and Meckling, 2021), however, not necessarily abruptly but in an incremental way (Moschella, 2015).

The article reflects on this issue by exploring whether IOs’ perceptions of the COVID-19 pandemic have indeed led to actual changes of IOs’ ideas, or whether a similar process of path-dependent development can be observed in the generation and modification of ideas, even during an extreme case – a crisis. To be able to discern between changes at different orders of magnitude, we differentiate between different levels of ideas – policy recommendations and paradigmatic ideas that inform them – as well as between different variants of change, inspired by HI scholarship, namely layering, conversion, dismantlement, and displacement.

Ideas exist at different levels of abstraction – from the various elements of policy proposals to broader “world views” or “normative frameworks” (Campbell, 2002, pp. 21, 23). Some scholars focus on the latter, more abstract paradigmatic ideas of IOs (Goldstein and Keohane, 1993; Béland and Orenstein, 2013), as they are often already embodied in IOs’ mission statements or founding documents and incorporated in subsequent documents. Changes at this level apply if an IO adapts its underlying broader and more abstract ideas in a particular policy field, indicated by, for instance, a change in the conceptual framework that an IO uses to frame its policies, a change in the problem construction that orients its policies’ goals, or even a change of the fundamental normative backdrop (see also Hall, 1993, on paradigm shifts).

In addition to paradigmatic ideas, we also analyse more concrete instances in the form of policy recommendations. In the literature, these are also called “programmatic ideas” (Campbell, 2002, p. 28). Policy recommendations are usually in line with IOs’ broader ideas but are characterised by much more technical language and necessitate the translation of abstract ideas into concrete solutions (Kaasch, 2013; von Gliszczynski and Leisering, 2016). They come in the form of policy parameters or policy instruments (Hall, 1993). Changes in parameters refer to select policy elements – this includes coverage (who is to be covered) or generosity (how much is appropriate). Changes in instruments would amount to more far-ranging modifications to recommendations, e.g. switching to promoting another policy model to combat a particular problem.

Adapting insights from HI on how institutions change to our analysis of ideational transformations (Béland and Powell, 2016), we can differentiate between variants of change. For each IO and each policy field observed, these indicate the direction of change and whether changes take place at the level of recommendations or paradigmatic ideas. Layering describes a process in which new recommendations (parameters or instruments) are superimposed to an IO’s existing recommendations, which does not involve changes at the broader level of ideas. Displacement means the adoption of potentially rivalling recommendations to an existing set of recommendations when there is a perceived need for quick adaptation. Conversion denotes an adjustment of recommendations in their content, leading to changes in objectives of policies and thus indicating a need for adaptations at the broader level of guiding ideas. Dismantlement describes a wholesale termination of recommendations, which, however, not only means that parameters and instruments change but also the underlying ideas, since broader guiding ideas also need to be adapted (oriented on Vetterlein and Moschella, 2014, pp. 148–149). Whereas layering and displacement are less far-ranging types of change, since they only add elements to an existing set of IOs’ policy recommendations, conversion and dismantlement can be considered more fundamental transformations by necessitating adaptations of the broader set of ideas that inform recommendations.

To explain our findings of ideational change or continuity in IOs, we refer to the IOs’ understandings of the crisis and its implications. The existing literature on policymaking under crisis conditions understands crises as exceptional events that allow actors to overthrow traditional practices and implement new ideas because existing approaches are challenged – but it predominantly approaches crisis as an objective external trigger of change. We argue that ideational change depends on the IOs’ perceptions of a crisis (see Hay, 2001; Rosenhek, 2013). Ideational changes can be time-consuming and subject to resistance due to ideational path dependencies. When proposing concrete recommendations and outlining broader ideas, IOs conceptualise the need for these policies and relate them to their crisis perception. By observing not only which policy recommendations and ideas the different IOs promote – and comparing these to pre-pandemic times – but also which reasoning underlies their arguments, we show the conditions and constraints of ideational change in crisis contexts.

Methodology and case selection

In social protection, labour, and health policy, several actors cooperate and compete in diverse and plural networks. We selected four of these IOs – the ILO, the OECD, the WHO, and the World Bank. As a global health authority, the WHO has been crucial from the beginning of the pandemic, but the other IOs have also contributed important policy recommendations to counter the pandemic’s social and economic effects (Kaasch, 2021). The WHO and the ILO share an orientation towards universalist policies, underpinned by progressive or social democratic ideas, while the OECD and the World Bank have often been characterised as neoliberal. Despite being an organisation with a smaller membership base of mainly developed countries, the OECD’s policy advice has wider repercussions also beyond its members (Ougaard, 2010; Mahon and McBride, 2011). Our cases thus intentionally cover a variety of policy perspectives, ideological orientations, and membership bases. While the IOs’ remit, influence, and impact differ, we can therefore explore whether we observe generalised trends or IO-specific responses.

The empirical analysis of IOs’ recommendations and paradigmatic ideas was guided by principles of qualitative content analysis (Mayring, 2014). First, relevant documents were collected through institutional repositories and websites. In the case of the ILO and the OECD, we could draw on pre-selections; the ILO offered a directory of COVID-19 related documents, and the OECD published a distinct working paper series on COVID-19. In the case of the WHO and the World Bank, we needed to identify relevant documents through a more open keyword search. Based on these criteria of relevance, 33 ILO documents, 61 OECD documents, 55 WHO documents, and 74 World Bank documents were analysed in detail. While the purpose, remit, and length of these documents differ, our methods allowed us to identify patterns that are indicative of both broader paradigmatic ideas and recurring (not singular) recommendations. Once no new insights were generated through the empirical analysis, we stopped incorporating new documents.

Coding began with a series of broad deductive concepts. This includes the differentiation of paradigmatic ideas and more specific recommendations, as well as the types of benefits recommended, inclusiveness and generosity criteria, exclusion and retrenchment, long-term or short-term duration, and binding or non-binding character. Further codes, particular to the IOs, were inductively added, such as their crisis frames or problem constructions. As a point of comparison for our identification of ideational changes, we drew on key policy documents for each IO that describe the respective IO’s policy ideas before the pandemic. To increase comparability, the selected documents mostly cover the period from 2017 to 2019 (ILO, 2017; OECD, 2017a; OECD, 2021a; WHO, 2019; World Bank, 2019).

The presentation of the empirical results only includes references to pertinent examples.

Empirical analysis: IO policy recommendations and ideas in times of crisis

International Labour Organization

For the ILO, the pandemic illustrated the indispensability of public social policy. The organisation thus used the COVID-19 crisis as a window of opportunity for promoting the expansion of social protection, health, and labour market interventions (ILO, 2021a, p. 18), hoping that “mindsets have shifted” (ILO, 2021a, p. 23). It blamed austerity for leaving poor and vulnerable people exposed to shocks and called for an end to fiscal consolidation (ILO, 2021a, p. 23) by using “deficit-based spending” (ILO, 2021b, p. 1). A key concern of the ILO was equality: preventing the entrenchment of existing disparities across gender, migration, income, or employment lines, as well as between the Global North and the Global South (ILO, 2021c, pp. 14–15). Due to the shared pandemic challenges, the ILO called for international support, particularly to countries with low financial means, cooperation between agencies, and better coordination between policy fields (ILO, 2021a, pp. 23–24).

Simultaneously, the ILO questioned whether the COVID-19 crisis was fundamentally different from past crises, apart from its severity, which is mirrored in the broad continuation of key ideas. Instead, the organisation argued that the crisis has accelerated existing trends (ILO, 2021d, p. 10), exposed existing problems (ILO, 2021a, p. 18), worsened conditions for vulnerable groups, and cemented existing inequalities (ILO, 2020a, p. 3; see also e.g. ILO, 2020c), demonstrating the merit of its pre-pandemic policy goals. While calling for new approaches (ILO, 2021d, pp. 10, 16–18, 22), its policy proposals were largely in line with existing recommendations, underpinned by its current set of labour standards (ILO, 2020a, p. 3), its conviction of the value of formalising informal employment, and its foundational paradigmatic ideas, among them universality, decent work, equality, and sustainability (ILO, 2021d, p. 24; see also Berten and Kaasch, 2025). A cross-cutting policy concern continued to be social dialogue (ILO, 2021c, p. 15), i.e. the strong involvement of the social partners in the development and implementation of policies. However, despite the continuation of key paradigmatic ideas, there are two cases of conversion: the ILO framed its policies also as a preparation for future crises, thus pointing to ‘resilience’ as a new key guiding concept; furthermore, the organisation pointed to the interconnectedness of crises, especially of climate change and COVID-19, and suggested strengthening sustainability and greening economic growth (ILO, 2020b; see also Leisering, 2021).

In terms of healthcare, the ILO promoted “collective financing, broad risk-pooling, and rights-based entitlements” (ILO, 2021a, p. 22), which is the same set of recommendations on Universal Health Coverage (UHC) that the organisation has promoted before the pandemic. A case of layering is the organisation’s promotion of universal vaccine access and distribution, pushing for an “equitable” solution for countries without the necessary financial resources (ILO, 2021a, p. 23). This added recommendation is in line with the existing paradigmatic idea of universalism (see also Kaltenborn, 2020) while responding to a pandemic-specific challenge.

Concerning labour, the ILO promoted strengthening income compensation and activation measures to deal with the widespread loss of working hours and jobs (ILO, 2020d, p. 1). Active labour market policies should be integrated with social protection measures to overcome the economic crisis, cushion its impact on people, and ensure lasting and decent working conditions (ILO, 2020d, p. 1). To preserve jobs, the ILO promoted work compensation measures (ILO, 2020d, pp. 12–13). At the intersection of health and labour, the ILO focused on occupational safety and health (ILO, 2020a, p. 4). Apart from strengthening security and preparedness of strongly affected groups of workers due to the special pandemic circumstances (see e.g. ILO, 2020e), which is a case of layering, the ILO did not stray from its existing policy recommendations (see also Silva, 2022), drawing on evidence from past crises.

The ILO claimed that social protection policies have been effective for people and economic systems alike, functioning as an economic stabiliser (ILO, 2021a, pp. 1, 18). However, interventions should function as “building blocks for establishing rights-based national social protection floors” (ILO, 2021b, p. 1) and thus be made permanent. Policy proposals largely mirrored existing emphases (ILO, 2021a, p. 23), focusing on achieving broad inclusiveness first, followed by increasing generosity, until coverage was “adequate and comprehensive” (ILO, 2021a, p. 23). Non-contributory measures should complement contributory ones (ILO, 2021e, p. 37), especially for workers in informal, temporary, or non-standard employment. While promoting universal expansion of social protection, the organisation highlighted coverage gaps of vulnerable groups, such as children, the elderly, women, migrants, and workers in informal or non-standard employment (ILO, 2021b, p. 1), also in line with pre-pandemic discourses (see also Deacon, 2013). Indicating a case of layering, the ILO discovered new vulnerable groups “who seemed to be getting by relatively well but were not adequately protected from the socio-economic shock waves” (ILO, 2021a, p. 18), including the self-employed (ILO, 2020a, p. 3). New data was published on their crucial economic role immediately before the pandemic (ILO, 2019), but the self-employed have not been a core focus of the ILO before.

Organisation for Economic Co-operation and Development

For the OECD, the COVID-19 crisis constituted an unprecedented challenge, with an intense impact on every aspect of society, including health systems, job markets, and societal well-being (OECD, 2020a). According to the OECD, the pandemic played the role of “a major accelerator of several trends and transformations,” which exacerbated pre-existing socioeconomic concerns (OECD, 2020b, p. 1). Given that diagnosis, the OECD’s reaction mainly leveraged its existing prescriptions by extending them to the context of the pandemic. Moreover, despite its devastating impacts, the OECD emphasised that the crisis was an opportunity for its member states to “build back better” (OECD, 2020b, p. 2) and learn lessons for “increasing governments’ resilience in the long run” (OECD, 2022, p. 3).

In health policy, the OECD reinterpreted health systems performance by prioritising resilience to recover from the pandemic. Upon the outbreak of the pandemic, the OECD highlighted the capacity of health systems to mobilise healthcare staff, secure essential medical supplies, and scale up spaces to diagnose and treat patients to respond to the short-term surge demands (OECD, 2020c), which later evolved into a comprehensive framework to enhance the resilience of health systems against future shocks (OECD, 2023). However, although emphasising resilience was novel, the recommendations for enhancing resilience predominantly built on previous policies prescribed to improve the efficiency and equity of health systems, such as strengthening primary care to prevent the spread of diseases (OECD, 2021b) and providing UHC to enable open access to diagnostic tests and treatment (OECD, 2020c; see also Kaasch, 2015). Moreover, as an organisation oriented towards fostering economic growth, the OECD focused on balancing the proposed containment and mitigation policies with their economic costs (OECD, 2020c, 2020d). This suggests that the fundamental ideational orientation has remained consistent, even when it comes to short-term recommendations. Given the rearrangement of existing policies to the new idea of resilience, the OECD’s health policy response can be characterised as a case of conversion, despite large continuities at the level of recommendations.

Regarding labour market policy, the OECD extended its existing recommendations for improving job quantity, job quality, and inclusiveness of the labour market to the pandemic situation. The OECD’s Jobs Strategy, updated multiple times since its launch in 1994, balances economic growth and protection of workers in the face of rapid transformations of society (OECD, 2018; see also Schmelzer, 2016). Policies proposed include facilitating flexible adjustments in working conditions, offering opportunities and incentives for workers to acquire skills that match labour market demands, and ensuring appropriate labour market regulation in protecting workers (OECD, 2018). Applying these policies to the short-term pandemic response, the OECD suggested extending the duration and coverage of paid sick leave, activating short-time work schemes, and providing access to unemployment benefits to informal workers (OECD, 2020e, 2020f, 2020g). Some of these recommendations leverage the pandemic as an opportunity to promote the OECD’s existing ideas: The OECD called for structural paid sick leave reforms to provide all groups of workers with access to paid sick leave and improve the adaptability of paid sick leave systems in case of future pandemics (OECD, 2020f); it also underlined the potential economic advantages of remote working after the pandemic (OECD, 2020e). Taken together, these additions of both pandemic-specific and long-term instruments to existing labour recommendations constitute a case of layering.

In social protection policy, one of the OECD’s main concerns was the well-being and inclusivity of vulnerable groups, particularly children and women, which had long been an important component of the OECD’s strategic agenda. For vulnerable children, the OECD urged governments to adopt emergency measures to provide children with access to food, protection against abuse and neglect, digital learning resources, and physical and mental health services (OECD, 2020h, 2021c). It additionally highlighted a long-term perspective on children’s well-being in the post-COVID-19 decade for an inclusive and resilient recovery, proposing a framework consisting of five pillars of action, including the mobilisation of financial resources and promotion of political commitment (OECD, 2021d). The OECD suggested to align both short-term and long-term measures for children with its existing call for early investment to improve their well-being and inclusive growth (OECD, 2009). Similarly, the OECD’s proposals on gender equality leveraged its previous agenda on eradicating violence against women and closing the gender wage and caregiving gap (OECD, 2017b; see also Mahon, 2015). Drawing on this agenda, the OECD showed the pandemic’s disproportionate effects on women, diagnosing that women faced increased risks of job and income loss, violence during quarantine, and an unequal burden of caregiving responsibilities (OECD, 2020i, 2021d). In response to its gender-specific impacts, the OECD argued that “all policy responses to the crisis must embed a gender lens and account for women’s unique needs, responsibilities and perspectives” (OECD, 2020i, p. 1). As a long-term response, the OECD also called for a gender-inclusive recovery by further utilising governmental tools such as gender budgeting and data disaggregated by gender (OECD, 2021e). Thus, changes in social protection proposals, both in terms of immediate and long-term action, constitute a case of layering pandemic-specific aspects to the OECD’s existing agenda.

World Health Organization

WHO warned COVID-19 will certainly not be the last health emergency and placed it as one of many anticipated future shocks (WHO, 2020a). In line with its mandate within the UN system, WHO took a health-focused approach to COVID-19. For WHO, COVID-19 demonstrated a lack of learning from previous pandemics and added urgency to a need to build resilient health systems. At the level of paradigmatic social policy ideas, the WHO did not seem to deviate from previously held positions. Instead, COVID-19 was understood to threaten progress towards existing goals, such as extending financial protection for health expenditures, while also serving to underscore their importance (WHO and World Bank, 2021). The COVID-19 pandemic revealed that “critical health systems gaps and vulnerabilities exist in countries from all income groups” (WHO, 2021a, p. 2). While underscoring that the organisation faced recurring funding issues (Reddy et al., 2018), the current crisis was in some cases presented as an opportunity, e.g. “for a ‘reset’ in countries with weak health financing systems to progress towards universal health coverage” (WHO, 2020b, p. 53; see also Kaltenborn, 2020).

WHO continued to promote UHC – finding this overarching goal more salient than ever in light of the COVID-19 pandemic and anticipated future emergencies. There was continuity in the promotion of a strong foundation of primary health care (PHC) as the best way to achieve UHC and contribute to (global) health security (see also Kaasch, 2015). The addition of resilience to ongoing and future shocks as a new goal of health systems strengthening constitutes a case of conversion.

In terms of the labour market, WHO focused on recommendations for the health sector and a strengthened public health workforce (WHO, 2022b). WHO called for protecting and supporting health workers, as they were simultaneously a vulnerable group and a key resource for maintaining essential health services during COVID-19 and beyond. If possible, countries were to “absorb qualified, yet unemployed, health workers into the health labour market to cope with rising demand, while setting a basis for greater long-term human resources for health capacity and resilience” (WHO, 2020d, p. 22). This emphasis on the vulnerability of health workers can be characterised as a case of layering, since it leaves both broader ideas and recommendations otherwise intact.

A focus on “serving the vulnerable” was highlighted in WHO’s General Programme of Work (WHO, 2019). In response to the pandemic, WHO recommended an “intensified focus on vulnerable and marginalized populations in all countries” (WHO, 2021a, p. 7). Layering can be observed in recommendations for vulnerable groups, e.g. in recommendations to make health systems more inclusive for migrants and refugees (WHO, 2022a) and people living with disabilities (WHO, 2020c), as well as in the identification of certain groups such as informal workers (WHO, 2021b, p. 20) and persons deprived of their liberty (OHCHR and WHO, 2020; WHO, 2021c, p. 4). Emphasising vulnerable and marginalised groups as those hit hardest by the adverse impacts of the pandemic, WHO called for integrating a social determinants approach “into pandemic prevention, preparedness, response and recovery – to manage COVID-19, to build back fairer and to prepare for future outbreaks” (WHO, 2021c, p. v).”

World Bank

Perhaps unsurprisingly due to the remit of the World Bank, the COVID-19 crisis was approached as a potential threat to existing economic norms and agreements, resulting in a focus on potential economic repercussions and fiscal stability (Evenett et al., 2020). Moreover, COVID-19 was understood as one of many crises – past, present, and future – including the environmental crisis, economic crises, the 2008 food crisis, and the war in Ukraine, re-emphasising the need for resilient and adaptive social security systems for the most vulnerable. “COVID-19 has also highlighted the need for more effective crisis response and greater household resilience to future shocks” (World Bank, 2021, p. 50). As such, COVID-19 was approached as an opportunity to ‘build back better’ and become more prepared for future crises by prioritising green and resilient infrastructure and economies (World Bank, 2021, pp. 70–74). This signified no paradigmatic ideational shift but allowed for existing instruments to be built upon and re-purposed for this goal.

Approaching COVID-19 through an economic lens, the immediate recommendations of the World Bank called on states to prioritise keeping cash flows going for especially small and medium enterprises and households (World Bank, 2021). Yet, some structural reforms of social protection systems were necessary so that they could be “shock-responsive and adaptive” to future crises; this could include a “shift toward financing models based on general taxation, instead of contributory systems that are largely limited to formal sector workers” (World Bank, 2021, p. 63). The key instrument that could successfully target the most vulnerable was cash transfers – a key World Bank policy recommendation before COVID-19 (see also Leisering, 2019) – now referred to as cash transfers plus, where the ‘plus’ includes complementing cash with additional inputs, such as “a combination of cash or in-kind transfers, skills training, coaching, access to finance, and links to market support” (Andrews et al., 2021, p. 2). This would be achieved by scaling up existing cash transfer programs in terms of both generosity and coverage while placing continuous emphasis on groups deemed most vulnerable, such as women, migrants, persons with disabilities, the elderly, sexual and gender minorities, indigenous people, and racial and ethnic minorities (World Bank, 2021, pp. 10, 63–66; 2022, p. 22). By doing so, these programs would increase economic inclusion and resilience of vulnerable households (Berten, 2024).

Beyond the household, there were several more structural economy-wide recommendations made, which targeted the fiscal situation of states going forward. This included tax reforms such as cutting business taxes or debt relief, and in higher-income countries, a focus was on employment support payments (Andrews et al., 2021). The IO recommended that the recovery phase should focus on lowering market entry requirements and simplifying regulations (World Bank, 2020), alongside the need for ‘fiscal space’ by strengthening financial sector resilience through deepening capital markets and leveraging private capital (World Bank, 2021, p. 75); recommendations in line with traditional emphases (see also Stone and Wright, 2007).

In focusing on the economic repercussions of COVID-19, the World Bank did not offer any paradigmatic ideational nor programmatic shifts; however, we do see some layering and conversion of existing mechanisms. The crisis reaffirmed their position that key to resilient economies and households were programs focusing on pro-poor economic inclusion and cash transfers at the household level and regulatory reforms that would open up national fiscal space and prioritise investment conditions. The shift from cash transfers to cash transfers plus can be understood as a layering of existing mechanisms as the instrument was expanded to include new degrees of coverage and generosity. Yet there was also a degree of conversion as these instruments were reframed as necessary responses to, and foundations for, future interlinking crises and shocks; they are now understood as necessary mechanisms for resilient and adaptive economies and the promotion of a digital economy for green growth, “helping countries emerge from this crisis and move toward a green, resilient, and inclusive recovery” (World Bank, 2021, p. 12; see also Leisering, 2021).

Discussion and Conclusion: Have IO ideas and recommendations changed or remained stable during the crisis?

Within IO discourse, social policy has received high priority during the COVID-19 crisis. Unsurprisingly, health policy was discussed by all organisations analysed, focusing on the immediate needs due to the pandemic. Labour market policy was also a central concern, since the crisis responses undertaken worldwide had grave consequences on employment that needed to be addressed. Social protection policy has been discussed as an economic stabiliser, a job preservation measure, and a safety net, particularly for the most vulnerable groups in society, which are often not fully covered by preexisting schemes. We find that there are clear parallels between the global financial crisis and COVID-19: whereas in early stages of the pandemic, IOs concentrated on short-term proposals of effective crisis management, they later focused on outlining long-term and future-oriented ideas for recovery.

However, we clearly observe path dependency in paradigmatic ideas and policy recommendations alike, which contradicts the expectation that actors perceive the crisis as a “window of opportunity” (Kingdon, 1984) for more fundamental path departures. Rather, IOs have been largely reactive, resorting to existing concepts that are quickly applied to new frames, maintaining recommendations in line with their existing paradigmatic ideas. We find that the perceived urgency of the crisis has been more dominant than its uncertainty; while its uncertainty could lead IOs to develop innovative responses that diverge from their usual approaches, its urgency led IOs to quickly refer to existing solutions. Based on our empirical observations of how policy recommendations and paradigmatic ideas are related to the crisis, we posit that the type and understanding of crises make an important difference in the actual realisation of path-breaking change. While there are crises that (are seen to) necessitate responses that conflict with ideological preferences and traditional emphases, others allow for ‘business as usual.’ After all, neither IOs’ recommendations nor their paradigmatic ideas were cast into doubt by the COVID-19 crisis in fundamental ways, and the benefit of social policies remained undisputed. We infer that the way in which the crisis and its outcomes were understood and communicated was thus a primary factor in this ideational continuity.

As Table 1 illustrates, we observe the pattern that, predominantly, IOs have added new recommendations in line with their existing broader ideas (layering). Recommendations were not abandoned (dismantlement), and there were also no contradicting recommendations introduced (displacement). Less often than layering, in some policy fields, recommendations were supplied with a different rationale or policy goals (conversion), introducing the future-related ideas of resilience and preparedness.

Table 1. Comparing ILO, OECD, WHO, and the World Bank.

Source: The authors.

Despite broad similarities, there are differences observable between the IOs. In the case of the ILO, there have been changes in parameters of instruments proposed rather than any paradigmatic ideational changes. For instance, the ILO acknowledged the struggles of the self-employed during the pandemic and proposed an expansion of existing social protection schemes to this group. These policy proposals were in line with the ILO’s existing ideas, such as universalism in coverage and an emphasis on workers’ struggles. The OECD tends to rely heavily on antecedent policy experiences to suggest solutions to given problems. Its outputs are closer to a collection of ‘effective’ policies rather than innovative solutions. Policy briefs are contextualised based on the OECD countries’ concerns and needs. Thus, the relatively narrow remit of policy briefs reflects limits of their global applicability. In the case of the WHO, advice has been technical and functionally limited to the WHO’s main mandate: healthcare. While acknowledging a need for multisectoral action, the WHO response centres on drawing renewed attention and commitment to health. In addition, the organisation entered the pandemic with, and continues to face, funding issues. While a lack of predictable and flexible financing may have limited its social policy response, the framing of the crisis as highlighting the relevance of pre-existing ideas and recommendations is consistent with continuity and the dominance of layering and conversion. The World Bank frequently connects COVID-19 response measures with broader adaptation goals, e.g. climate change, to strengthen the resilience of social policy institutions and individuals, but otherwise keeps its recommendations largely intact compared with the pre-pandemic phase.

There are also differences observable in the three policy fields. On the one hand, for health and labour, challenges involved a variety of ‘unknowns,’ such as developing a vaccine for a novel disease in record time and dealing with the effects of widespread shutdowns of the economy, which have both been unprecedented in recorded history. Thus, IOs had to depart somewhat from their conventional policy wisdom to meet these new challenges; they have done so, though, in line with their preexisting positions. On the other hand, social protection faced generally ‘known’ challenges, since the pandemic exacerbated preexisting inequalities and weaknesses, hitting already vulnerable groups the hardest. The main solution promoted by IOs was to ensure the provision of income- and health-based social protection to these groups and include them in labour market interventions – which did not constitute a new measure but rather the continuation and expansion of an existing one. This partly explains why we find cases of conversion concentrated in the field of health policy, where issues of future-oriented resilience were common in IO policy discourse.

It was outside our scope of analysis to investigate whether the changes observed continued after the official end of the pandemic. Due to the pandemic-related nature of a number of these instances, we assume that most changes are indeed confined to the pandemic period. We found that there are strong ideational path dependencies, which contradict the common assumption in the literature on the role of crises in policymaking that crises frequently lead to ideational changes. We posit that there is a need to do more research on the independent influence of crisis perceptions on ideas, complementing existing research on changes in institutions, where ideas figure mostly as an independent variable.

The existing literature on crises and ideational change points to the possibility of paradigmatic transformations in light of overwhelming uncertainties. But the article’s empirical results strengthen arguments for a general path dependency on the level of not only the paradigmatic ideas of IOs but also their specific policy recommendations. We find that ideational change occurs mostly incrementally, in line with what the institutionalist literature posits regarding institutional change. Opportunities as well as pressures for change depend on crisis perceptions.