Dead Ends and Blind Spots in the European Semester: The Epistemological Foundation of the Crisis in Social Reproduction

Abstract

This article provides new perspectives on the persistent hierarchy between ‘social’ and ‘economic’ goals in European Union's (EU) economic governance. We operationalize insights from feminist economics and political economy to analyse the agenda‐setting documents of the European Semester – the Annual Growth Surveys (AGS) – showing how the much‐debated integration of social goals into the European Semester is fundamentally constrained by mainstream economic epistemologies. These epistemologies misrepresent interrelationships between the productive economy and the reproductive labour needed to maintain it. Using interpretive policy analysis, we show how multiple concepts and measurements used to conceptualize policy goals and impacts within the AGSs, coalesce to systematically misrepresent reproductive labour as a ‘social’ activity, an irrelevance, or a cost, rather than a macroeconomic input. This restricts the possibilities of enhancing the social dimension of the European Semester, in ways conspicuously ignored by the existing literature, which are of heightened salience in the wake of Covid‐19.

Introduction

Since its inception in 2011, the main framework of the EU's economic governance regime, the European Semester, has been a key site where long‐standing struggles over the EU's economic and social goals have been played out. Most analyses of these struggles have examined the content and framing of policy proposals and the power hierarchies between actors who compete to influence them. Some have argued that there has been a partial ‘socialization’ of the European Semester in terms of its policy content, processes and actors (Zeitlin and Vanhercke, 2018). Others, however, have stressed that social policy remains subjugated to economic goals and a market‐making ethos, and that the proponents of social policy continue to have less power than economic actors in policy discussions (Copeland, 2020; Copeland and Daly, 2018; Dawson, 2018).

Drawing on Mason's ‘facet methodology’ (Mason, 2011) we take a different approach that recasts these lines of enquiry. Mason points out that research fields are often created through processes where the same methodological and substantive foci are repeated over and over. Similar critiques have also been advanced in relation to EU studies, by authors who have pointed out how repetition of the same approaches and foci has not only co‐constructed shared understandings of the EU's contours, but also entrenched significant methodological blind spots. These critics call for the inclusion of dissident voices in the analysis of the EU to remedy these gaps (Guerrina et al., 2018; Manners and Whitman, 2016). Mason's facet methodology is not characterized by a particular set of tools or a process. Rather, it is an orientation that emphasizes the importance of critical reflection on two intertwined factors that are often left implicit: the ontological (what ‘is’/what are we studying?) and the epistemological (how do we gather and interpret meaningful evidence?) premises of the research process. Using the metaphor of grinding facets into a gemstone, Mason argues that the epistemology we use to understand a problem shapes our perspective of it, casting a particular light onto and through the object of study. Applying epistemological approaches that challenge existing background knowledge enables us to fashion a different facet, and shine fresh light onto the object of study. This will generate insights that are especially ‘telling in relation to existing background knowledge and theoretical debate’ (Mason, 2011, p. 80, our emphasis) so that existing lines of enquiry are recast, and prior concepts are ‘troubled’ (Mason, 2011, p. 83).

In our case, facet methodology has enabled us to recast perspectives on the relationship between economic and social goals within the European Semester. We suggest that the way the concepts of ‘economic’ and ‘social’ are specified, have become so firmly embedded in earlier literature, that they shape and limit insight into the problem at hand. This article problematizes these categories, and their associated economic ontology and epistemology. We argue that uncritical acceptance of definitions of ‘economic’ and ‘social’ policy has restricted our view on the issues at stake, drawing us into self‐limiting debates. To make these arguments, we draw together insights from two strands of scholarship: feminist economist and feminist political economist literature that criticizes the ontological and epistemological practices that dominate ‘mainstream’ economics and economic policy (Bakker, 2020; Çağlar, 2009; Cavaghan, 2021; Elomaki and Sandberg 2020; Grown, 2000; Heintz, 2019); and interpretive and discursive approaches that examine how ontological and epistemological assumptions are stabilized and with what implications (Callon and Latour, 1981; Cavaghan, 2017; Elomäki, 2020; LeGreco and Tracy, 2009).

Both feminist economists and feminist political economists have extensively critiqued the way categories of ‘economic’ and ‘social’ are deployed in mainstream economic thought ¹ (Bakker, 2020; Çağlar, 2009; Grown, 2000; Heintz, 2019; Pearson and Elson, 2015). Feminist scholars eschew these categories, focusing instead on the ‘reproductive’ and ‘productive’ sectors of the economy. They argue that mainstream economic epistemology labels reproductive labour and provisioning as ‘social’ issues, depicting them either as a cost (public reproductive provisioning) or as irrelevant (unpaid reproductive labour), rather than a foundational economic input. By definition, this constrains the scope for identifying problems, desirable goals, policy options and, importantly, measures of success. Extant research on social objectives and policies within the European Semester, and their relationship to economic goals conspicuously misses this fundamental insight.

In light of these critiques, we unpack what the European Semester assumes ‘the economy’ to be and how it draws boundaries of the ‘economic’ relative to the ‘social’. Following this, our research questions are: how is the interrelationship between the productive and the reproductive sectors of the economy constructed in the European Semester; what impacts do these constructions mean for actors and activities related to the reproductive economy; and what impact do they have on efforts to enhance the social dimension of the Semester? In unpacking these, we draw on research material comprised from the European Commission's Annual Growth Surveys (AGS) 2011–19, and the agenda‐setting documents of the European Semester, which provide a suitable site to analyse epistemologies and ontologies behind the European Semester's policy priorities.

Our analysis shows how the misrepresentation of the relationship between the reproductive and productive sectors of the economy typical of mainstream economic thinking and policy‐making, is woven into the European Semester and how this produces a persistent hierarchy between ‘social’ and ‘economic’ issues. Thus, we illuminate an important and under‐examined facet of the EU's economic governance. By exposing hitherto unidentified barriers to the integration of a meaningful social dimension to the European Semester we provide a dissident analysis in a core area of EU studies. Our findings suggest that even the recent efforts to enhance the social dimension of the European Semester by Ursula von der Leyen's Commission (von der Leyen, 2019), will prove impossible to achieve as long as the reproductive economy is mistakenly construed as a matter of ‘social’ policy and defined in opposition to the ‘economy’.

I. Towards a More ‘Social’ European Semester?

The relationship between social and economic goals and policies has been one of the key themes in scholarly debates about the European Semester. The Semester combines into one annual process different economic and social governance tools with distinctive legal bases and origins, such as the Stability and Growth Pact (SGP), the Fiscal Compact, the Macroeconomic Imbalance Procedure (MIP), the Europe 2020 Strategy and, since 2018, the European Pillar of Social Rights (EPSR). During the annual European Semester cycle, the European Commission, the Council and the European Council set key reform priorities for the EU (in the AGS), review national performance and policies (in Country Reports) and issue Country Specific Recommendations (CSRs) to member states that cover a wide range of themes from fiscal policy to wage determination, education and healthcare.

The European Semester has become the ‘only game in town’, into which social agendas have to be integrated in order to be pursued (Copeland, 2020, p. 138). However, whilst the tools of economic integration have been repeatedly enhanced, social policy tools have a consistently weaker legal basis than fiscal rules and the MIP, and are not monitored and reinforced to the same extent (de la Porte and Heins, 2015). This reflects the oft‐cited ‘constitutional asymmetry’ between the EU's economic and social policies, whereby formal economic integration has progressed whilst social policy choices have been seen as ‘secondary‐order issues’ and largely left to the national level (Copeland, 2020; Scharpf, 2010; Scheele, 2021). Some argue that this makes social policy goals subordinate to, or even more ‘dependent‐upon’, developments in economic integration than they were before the financial and euro crisis (Copeland and Daly, 2015, p. 156).

Nevertheless, scholars agree that despite the initial emphasis on fiscal consolidation, the priorities and recommendations in the European Semester have evolved in a more social direction. For instance, the proportion of social and employment recommendations in the CSR's have increased (Zeitlin and Vanhercke, 2018), alongside recommendations geared towards ‘social investment’ (Crespy and Vanheuverzwijn, 2019). Governance procedures have also changed in ways that give social actors within the European Commission and the Council a more prominent role, even if economic actors still ‘hold the pen’ (Copeland and Daly, 2018; Zeitlin and Vanhercke, 2018). The significance of these changes, however, remain contested.

Working from an actor‐centred perspective, Zeitlin and Vanhercke take an optimistic position, arguing that the Semester has undergone a ‘partial but progressive socialization’ (2018, p. 4), referring to greater emphasis and monitoring of social objectives and the enhanced access to relevant policy processes for social and employment actors. Other analyses of how social CSRs are framed, argue that economic goals and market‐logic continue to shape social policies. Copeland and Daly (2018, p. 1002) for example, argue that ‘EU social policy as enunciated through the CSRs is much more oriented to supporting market development than it is to correcting for market failures’. Going further still, others contend that social policy within the Semester has been reduced to ‘making individuals fit for the market’ (Dawson, 2018, p. 207) or that it serves to ‘commodify or partially commodify individuals and the labour market’ (Copeland, 2020, p. 125). On this basis, some suggest that social actors seem to become captured by the Semester's economic logic, obliging them to frame their proposals in economic terms (Copeland and Daly, 2018, p. 1012; Crespy and Vanheuverzwijn, 2019; Dawson, 2018, p. 207).

The critiques above share the salient inference that assessing the social dimension of the Semester requires looking beyond the mere presence of social goals, and examining instead the underlying and dominant assumptions reproduced in the European Semester. We share that sentiment. However, in our view, the focus needs to be shifted from the economic framing of social policies, toward the economic epistemologies underpinning the Semester. In doing so, we show that these epistemologies fundamentally subordinate ‘social’ goals to economic ones in ways hitherto unrecognized in extant literature.

II. Unpacking Economic Knowledge: Feminist Economics and Feminist Political Economy

The narrowness of mainstream economic epistemologies has been a persistent theme in critical political economy and heterodox economics literatures. Contributing to this debate, feminist economists and feminist political economists have long criticized the silence in mainstream economic thinking around the role of care or ‘social reproduction’ in the functioning of the economy (Bakker, 1994, 2020). In strong contrast to mainstream economists, feminist economists argue that human labour itself is a produced means of production, which is dependent on investment (Elson, 2016, p. 28; Heintz, 2019, p. 58). Some of this investment requires domestic and public spending – for example, buying food and paying for healthcare, or public spending on education, welfare services and benefits. However, much of this investment is not monetized. Domestic labour, such as cleaning, preparing meals or caring for family and community members, requires the investment of time. Feminist scholars describe these activities as the reproductive sector of the economy.

This reproductive sector includes ‘services directly concerned with the daily and intergenerational reproduction of people as human beings, especially through their care, socialization and education. It includes unpaid work in families and communities, organized unpaid volunteer work and paid work in public services like health and education that produce for use rather than for sale’ (Pearson and Elson, 2015, p. 10). Feminist economists, and increasing numbers of heterodox economists, therefore point out that human well‐being and the monetized productive economy are dependent upon the reproductive economy (Heintz, 2019; Pearson and Elson, 2015).

However, the reproductive economy is routinely undertheorized and undervalued, or even ignored, both in economic theory and economic policy (Bakker, 2020; Heintz, 2019). This pattern is embedded into the concepts and indicators used to discuss the economy, measure economic performance and legitimize economic policy to citizens (Bakker, 2020; Heintz, 2019; Scheele, 2021). Key economic indicators such as national accounts and Gross Domestic Product (GDP) exclude unpaid social reproductive inputs into the economy, and the amount and value of these inputs are not systematically imputed and calculated (Heintz, 2019). Similarly, despite policy rhetoric purporting to support social investment in many countries, definitions of public investment used in systems of national and regional accounting in EU member states exclude the monies spent to ensure the ongoing delivery of public services, such as the salaries of teachers, nurses and childcare workers (Elson, 2017, p. 21; Heintz, 2019 p. 33). Equally, standard economic concepts and measurements, including growth (Kabeer, 2016), competitiveness (Gillespie and Khan, 2016) and productivity (Elomäki, 2020), also reproduce this pattern. These kinds of economic measurements take a highly abstracted and simplified view, which elides the value of the reproductive sector of the economy. For example, commonplace assumptions that pay is an accurate measurement of productivity legitimize the systematic undervaluation of the labour‐intensity and the skills required for reproductive labour (Elomäki, 2020).

Feminist economic epistemologies, by contrast, begin from the assumption that the goal of economic policy should be maximum human welfare (Power, 2004), and are attentive to the interrelationships between the reproductive and productive sectors, human welfare, and ‘growth’. These analyses have shown how the productive economy is fundamentally dependent upon social reproductive labour, whilst also emphasizing positive social reproductive outcomes as an inherent good (Heintz, 2019). They also point out that social reproductive labour is overwhelmingly undertaken by women and that its burdens fall hardest on racialized or migrant women (Bakker, 2020; Grown et al., 2000). Feminist economists' perspectives then, enable an understanding of the ways in which public funding cuts for care lead to increased burdens of unpaid care work and a crisis in social reproduction (Bruff and Wöhl, 2016). Applications of these approaches include assessments of returns on investment that the care sector yields (de Henau et al., 2016), and the identification of tipping points where a community's or a person's reproductive capacities begin to deplete after austerity cuts, leading to a sustained decline in welfare and capacities (Rai et al., 2014).

Thus, the under‐theorization of the interrelationship between the productive and reproductive sectors of the economy is perpetuated by multiple, interlinked epistemic practices used in economics and economic policy‐making. These include concepts and measurements of growth or productivity and non‐measurement of reproductive work, its distribution and the depletion of collective or individual reproductive capacities. There are three overarching outcomes to this, each relevant to our analysis of the visibility and significance of social goals within the European Semester. First, unpaid reproductive labour is routinely (implicitly) misrepresented as non‐economic activity that is at worst irrelevant, or at best secondary, in economic policy discussions. Secondly, policies consistently reflect a failure to understand that growth and productivity are dependent upon a well‐functioning reproductive economy that meets collective human needs for care, nourishment and community. This leads to the third outcome. Mainstream macroeconomic thinking portrays spending on services such as health, or long‐term care, as a hindrance to growth or productivity, which can be tackled by cutting spending on these services. Macroeconomic parameters or goals that exclude consideration of the reproductive economy are set first; only after that do policy makers decide which ‘social’ policies are feasible (Heintz, 2019, p. 86).

This contrasts markedly with feminist economic approaches, which firmly portray social reproduction as an economic, specifically a macroeconomic, factor: ‘by acknowledging that human investments have macroeconomic consequences, an array of social policies are elevated to the status of macroeconomic policies including aspects of education, health and childcare policies. Unpaid work … becomes a critical factor of production … necessary to sustain our economies in the long run’ (Heintz, 2019, p. 86). Feminist economic approaches recognize that cutting spending does not ‘save’ resources, it simply reallocates the burdens of the reproductive economy back into the private sphere and renders the work non‐monetized–whilst the outputs of the reproductive economy continue to be collectively enjoyed. These insights are well established in feminist economics and feminist political economy. However, they remain marginal in EU studies, and to date they have also been ignored in examinations of the relationship between the social and economic goals within the European Semester.

III. Materials and Methods: From Increased Visibility of Social Goals to Epistemological Constraints

In the introduction we outlined the ontological and epistemological reflection process involved in Mason's (2011) facet methodology, which underpins our investigation. The methodological tools we use to operationalize it, draw on insights from discursive approaches (LeGreco and Tracy, 2009) and Interpretive Policy Analysis (IPA) (Çağlar, 2009; Callon and Latour, 1981; Cavaghan, 2017). IPA approaches have been used to examine the significance of epistemological practices by sorting out ‘relevant’ from ‘irrelevant’ information in organizations and policy processes. These approaches argue that the establishment of shared epistemological practices marginalizes dissenting voices, moving some issues into ‘black boxes … those things whose contents have become a matter of indifference’ (Callon and Latour, 1981, p. 285), and thus restricting the range of intelligible policy options to participants in a policy process (Cavaghan, 2017, p. 191). This scholarship is also particularly attentive to the different functions attributed to documents and the hierarchies between them (Cavaghan, 2017, p. 192; Le Greco and Tracy, 2009). Hence, agenda‐setting texts, such as EU Treaties or strategic policy documents, or in this case the AGSs, may establish goals and norms, but they are likely to be linked to implementation documents and evaluation criteria (Smith, 2005).

Repetitive cross‐referencing between documents and multiple actors' use of the same measures, plays an important role in policy processes. It establishes a coherent and durable epistemology, which becomes difficult to challenge because it is consistently reproduced in policy documents and discussions that refer to each other (Callon and Latour, 1981; Cavaghan, 2017). Analysing policy documents to reveal how problems are defined, what policy proposals they contain, and the consistencies in the evidence (that is, specific kinds of data, and the epistemology the data depends upon, to seem meaningful) they draw on, enables us to show how the scope of policymakers' attention, and plausible policy options, are systematically limited.

Our research material consists of the AGSs between 2011 and 2019; the agenda‐setting documents of the European Semester. Most research on the social/economic relationship within the European Semester has focused on the CSRs. In our view, the AGSs provide a better starting point for a qualitative, interpretative analysis of the Semester's epistemological underpinnings for the following reasons. The AGSs set out EU‐level priorities, by defining the issues, concepts and governance initiatives that are relevant and important within the Semester. For example, they insert new ideas and initiatives into the Semester, relating to social goals like ‘adequacy’ of pensions (see Copeland, 2020, p. 130). Unlike the concise CSRs, AGSs describe the factors that have influenced the EU's socio‐economic situation and the rationale for policy solutions and define key concepts. As such, they provide the most suitable material to analyse the dominant understandings of the economy and its functioning within one of the most important processes of EU policy‐making (European Commission, 2014, 2015, 2016, 2017, 2018). ²

AGSs commonly reflect the priorities of the sitting President of the European Commission, and the struggles between social and economic actors within and outwith the Commission (Copeland, 2020). They are structured along the Commission's policy priorities, which in the period covering our research, remained relatively stable for the Barroso Commission (2011–14) and the Juncker Commission (2015–19). Both Commissions devoted one section to fiscal policy, one or more sections to growth and investment, and one section to employment and social issues. Notably, these sections are linked to governance instruments with varying strength (SGP vs. Europe 2020 and EPSR) and are therefore organized in a hierarchical relationship with one another (see de la Porte and Heins, 2015). In addition, in keeping with the hierarchies within the European Semester, civil servants from Directorate General (DG) for Economic and Financial Affairs and the Secretariat‐General dominate the process of drafting the AGSs, even though DG Employment and Social Affairs increasingly participate (see Copeland, 2020; Copeland and Daly, 2018), and stakeholders, like EU social partners, attempt to influence the content of the AGSs (Elomäki and Kantola, 2020).

As outlined above, our analytical approach goes beyond examining how social policies are framed (see Copeland and Daly, 2018; Dawson, 2018) focusing instead on the relationship between the productive and the reproductive sectors of the economy. We analysed the AGSs using Atlas.ti in order to gain a systematic view. First, we coded all references to the reproductive economy. Drawing on feminist economic approaches, we defined the reproductive economy as unpaid care, alongside all monetized activities, public or private, that contribute to the maintenance of life (for example care services, welfare benefits, primary education), distinguishing between unpaid and monetized instances of it. Secondly, our coding interrogated these references to the reproductive economy in the broader context of the AGSs and the relationship between the ‘social’ and the ‘economic’ therein. For this purpose, we coded actors/sectors understood as valuable or as a cost to the economy; the causal assumptions, value hierarchies and contradictions connected to the relationship between productive and reproductive sectors and ‘social’ and the ‘economic’ goals more broadly (see supplemental material 1 for the coding framework).

Our analysis combined quantitative comparisons with an in‐depth interpretative reading of the texts. We paid particular attention to the concepts and measurements used in the AGS, which we regard as the embodiment of the European Semester's epistemology: abstractions of reality premised on disciplinary practices and assumptions, sorting out important information from that which was extraneous or irrelevant (see Callon and Latour, 1981; Cavaghan, 2017). In addition to tracing shifts over time, we identified differences and contradictions between the different parts of the AGSs. The analysis is also organized around three themes informed by feminist literature. First, we consider how the value and role of the reproductive economy within the functioning of the economy was measured and constructed. Secondly, we show how the ‘social’ impacts of economic policies were constructed and measured, with an emphasis on the causal relationships between the productive and the reproductive sectors of the economy. Finally, we are attentive to contradictory statements with regards to the reproductive economy.

IV. The Reproductive Economy: A Cost or an Invisible Economic Input?

Reflecting the silences within mainstream economics and economic policy‐making discussed above, the AGSs focused on the monetized economy, rarely mentioning the non‐monetized reproductive sector and its role in the economy. The ‘economy’ was quantified and measured through various indicators ranging from the GDP to general government debt and deficit, financial sector liabilities, house prices, employment and unemployment, and private sector debt. The increased usage of indicators and knowledge related to the social situation in the EU after the introduction of the EPSR, and the Social Scoreboard measuring its implementation, took some monetized aspects of the reproductive economy into account (for example unmet care needs) but neglected its non‐monetized forms. The tools which exist to measure the value of unpaid reproductive work and depletion of an economy's reproductive capacities (for example Hoskyns and Rai, 2007; Rai et al., 2014) were not used.

References to unpaid reproductive work were infrequent and often implicit. Four AGSs (European Commission, 2011a, 2012, 2013, 2018) included a single, implicit acknowledgement of unpaid reproductive work, framing it as a barrier to labour‐market participation: ‘tax benefit systems, flexible work arrangements and childcare facilities should be geared to facilitating the participation of second earners in the workforce’ (European Commission, 2011a, p. 6). Almost the exact same wording appeared in the AGS 2013, whilst the AGS 2014 argued that ‘access to affordable care services will help the participation of women in the labour market’ (European Commission, 2013, p. 11). AGSs for years 2016–18 discussed unpaid reproductive work under the theme of work‐life balance. Two of them explicitly referred either to ‘care obligations’ (European Commission, 2016, p. 7) or ‘caring responsibilities’ (European Commission, 2015, p. 11). Discussions about work‐life balance also focused on labour market participation. For example, the 2016 AGS argued that ‘Member States should have a comprehensive approach to improving the work‐life balance, including through care facilities, leave and flexible working time arrangements, as well as tax and benefit systems free of disincentives for second earners to work or to work more’ (European Commission, 2015, p. 11). We found one instance that weakly acknowledged the value of unpaid reproductive work. The AGS 2017 stated that the ‘pension system could credit care duties,’ followed by the caveat ‘taking account of fiscal positions and future expenditure implications’ (European Commission, 2016, p. 12, our emphasis). Overall, then, the value of unpaid reproductive work was largely elided. Unpaid care was constructed as something that should be transferred to the paid economy in the form of public service provision, combined more effectively with paid work or disincentivized through tax‐benefit reforms. In other words, it was constructed as a hindrance to the efficient functioning of the economy and the attainment of economic goals, rather than as something that keeps the economy in motion and has economic value in itself (see Heintz, 2019; Hoskyns and Rai, 2007)

Discussions of the monetized reproductive economy betrayed a similar under‐recognition of the reproductive sector's intrinsic value. For example, the location of mentions of public reproductive provisioning within the internal hierarchy of the AGSs sections (discussed above) revealed that the reproductive economy was commonly recognized as a macroeconomic issue in terms of a threat to public finances. Public reproductive provisioning was mainly discussed under employment and social priorities or in the sections focused on ‘fiscal consolidation’ and ‘fiscal responsibility.’ Sections on growth and investment either completely neglected public reproductive provisioning (2011–15), or, in later AGSs, depicted it as a marginal issue (2016–19).

‘Investment’, as opposed to an explicit and exclusive focus on austerity, has been a clear feature of the AGSs since 2015 (Crespy and Schmidt, 2017), and between 2015 and 19, the first priority was ‘re‐launching’, ‘boosting’ or ‘delivering’ investment. Using the concept ‘social investment’, the AGS 2016 was the first to explicitly argue that public services contributed to the functioning of the economy: ‘Social investment offers economic and social returns over time, notably in terms of employment prospects, labour incomes and productivity, prevention of poverty and strengthening of social cohesion’ (European Commission, 2015, p. 9). If we contextualize references to social investment within the AGSs' wider rhetoric of investment, however, we can see that the conventional understanding of public investment, one that excludes the on‐going delivery of public services crucial for the future productivity of the economy and for human well‐being (Elson, 2016; Heintz, 2019), remained unchallenged. The AGSs' investment priorities comprised environment and climate, education and human capital, traditional physical infrastructure, digital technologies and research, and development and innovation. The reproductive economy was not acknowledged as a site of meaningful investment.

In addition, discussions of social investment were strongly marked by traditional understandings of growth and the role of paid employment in it. This includes the pursuit of traditional macroeconomic metrics that measure the size of the paid economy (e.g., GDP) rather than the pursuit of maximum human welfare (see Power, 2004). Positive references to public care provisioning focused overwhelmingly on enhancing women's labour market participation (childcare and long‐term care), strengthening future labour market capacities (childcare) or creating jobs and keeping people able to work (healthcare). By contrast, healthcare, long‐term care and pensions were mainly constructed as threats to the sustainability of public finances, and unemployment and minimum income benefits as threats to labour supply. Clearly, the AGSs' commitment to social investment extends only to social reproductive services that provide returns in terms of labour supply and human capital. The needs of those requiring care and other forms of social provision, such as the elderly, the poor and the disabled, do not come under this umbrella. This understanding of the economy reflects traditionally conceived concepts of growth and associated anxieties over public spending – not an intention to use economic planning to maximize human welfare.

It is also notable, that although AGSs 2016–19 mentioned public care services in the investment section, only in 2017 were there explicit calls for more spending on care services: ‘Investments also need to focus on human capital and social infrastructure. The development of long‐term care services and affordable and flexible childcare facilities is particularly important to decrease care obligations towards the elderly and children, frequently affecting women.’ (European Commission, 2016, p. 7.) The investment section of the AGS 2018 mentioned care services without explicitly framing them as an investment, and the AGS 2019 only discussed investing in childcare as a means to improve educational outcomes. The increased rhetorical acknowledgment that social reproduction has the ability to yield a return was evidently not allied with calls for the levels of public spending which would enable these returns to be delivered (see on CSRs). Furthermore, when we read the AGSs in their entirety, taking account of the sections with the strongest enforcement regimes, the dominance of a cost‐perspective, implying further cuts to public services, was revealed. This lack of resourcing for the reproductive sector drives the intensification of crises of social reproduction (Bruff and Wöhl, 2016) and its depletion (Rai et al., 2014).

Even when ideas surfaced that challenged the ignorance of the economic contributions of the reproductive sector, our deeper analysis of AGSs shows that it did not lead to any form of consistent valuation evidenced by actions to support it. In any case, the idea of investing in ‘social infrastructure’ – a narrative used by some feminist economists to contest the narrow notions of investment that side‐line public provisioning (Elson, 2016) – used in AGS 2017, disappeared again in 2018 and 2019. This, we argue, shows the salience of the dominant economic epistemology around the reproductive economy. Even if social actors sometimes succeed in bringing new ideas about the reproductive economy on Semester's agenda (see Copeland, 2020), these ideas are crowded out by dominant economic premises and do not ‘stick’.

V. Social Consequences of Fiscal and Structural Policies

We now turn to examine how the AGSs discussed changes the reproductive economy in the wake of the economic crisis, and associated economic policy responses. The AGSs named these ‘social impacts’ and our analysis shows a moderate but inconsistent rise in their acknowledgement. AGS 2011 mentioned these once, whilst the AGS 2019 mentioned these seven times. The AGSs from 2012 to 2014 even named ‘tackling unemployment and the social consequences of the crisis’ as one of its five priorities. However, the way the social impacts were discussed, was nested in the same causal assumptions and epistemologies we found in the previous section. The discussions showed a very weak acknowledgement of the factors causing hardship combined with a mix of vague conceptualizations of social impacts, coupled with the use of traditional economic metrics that render the reproductive economy and human welfare almost unintelligible.

For example, in contrast to the detailed and repeated discussions of problems such as debt, deficit, growth and productivity, the descriptions of social impacts were infrequent and sometimes very broadly sketched and concise. The only reference to social impacts in the AGS 2011 was an acknowledgment that the crisis has taken ‘a heavy toll on Europe's societies’ – without any further comment (European Commission, 2011a, p. 2). The negative effects of austerity, and its role in causing hardships, were conspicuous by their absence. About half of all references to social impacts (26 of 50) included no identification of cause, and often ‘the crisis’, rather than policy responses, was identified as the cause of hardship. There were only five references to the social impacts of ‘reforms’ or ‘restructuring of our economies’ (two in AGS 2016 and 2017 respectively, and one in AGS 2019). Where austerity or ‘reforms’ were identified as the cause of hardships, it was accompanied with arguments that these processes were ‘necessary’. For example, ‘profound restructuring of our economies … is currently taking place. This process is disruptive … and socially difficult – but it is necessary to lay the foundations for future growth and competitiveness that will be smart, sustainable and inclusive’ (European Commission, 2012, p. 1). Four mentions of the need to ‘raise awareness of’ or ‘take account of’ the distributive impacts of the proposed economic policies appeared in the AGS 2016 and AGS 2017. These vague appeals to awareness‐raising were not observably linked to any concrete macroeconomic decisions.

In similar partiality, the acknowledgements of social impacts rarely identified who has been most affected by austerity. This despite the widespread evidence that identifying the disproportionate impact of austerity on people using, and working within, public services supplying social reproductive goods – i.e., women, especially those of migrant origin (see Cavaghan and O'Dwyer, 2018; Karamessini, 2014). In the nine AGSs analysed here, ‘youths’ were identified as being particularly affected, although unaccompanied with any explanation as to why. ‘Single parents’ were mentioned only once (European Commission, 2011b, p. 12). Other mentions included equally vague descriptions, such as ‘vulnerable groups’ (European Commission, 2011b), ‘vulnerable households’ (European Commission, 2012), ‘the disadvantaged’ (European Commission, 2013) and ‘the long term unemployed’ (European Commission, 2015). The AGSs showed a considerably less developed, or precise understanding of social impacts, compared to mainstream economic ones. Fuller comprehension of the reproductive economy, its function and its links to the productive economy would remedy this.

Where social impacts were discussed in a more precise way, employment was, once again, used as a proxy indicator: nearly half of all references to social impacts were conceptualized in terms of unemployment – that is, either defining the impact as unemployment or causally linking it to unemployment. The effect that cuts to public services may have had on access to employment, quality of life, or time use, were not acknowledged. Concepts such as depletion (Rai et al., 2014) that would enable the conceptualization of these effects – modelling the ways that cuts in public services may have depleted the economy's capacity to maintain and produce labour and human welfare – were lacking. The absence of this measure, and reliance instead on ‘unemployment’ as a proxy, once again depicts ignorance of how, and why, reducing the flow of resources into the reproductive sectors of the economy affects the productive economy. Unemployment is at heart an indicator of how many people are participating in the paid productive economy: it does not measure well‐being, wealth, health, security or resilience and it does not capture circular causal relationships between the reproductive and productive economy.

These kinds of measurements of ‘social impacts’ render them understandable within the mainstream economic paradigms discussed earlier. However, they do not consider the important redistributive impacts that feminist analyses of crises and austerity have emphasized: increased reproductive labour burdens (Bruff and Wöhl, 2016; Cavaghan and O'Dwyer, 2018; Karamessini, 2014). Thus, the increase in the ‘reproductive subsidy’ (Rai et al., 2014) extracted from households and individuals specifically due to the implementation of austerity is invisible in these documents.

VI. Contradictory Goals and Policies

Finally, we analyse examples of contradictions in the AGSs in relation to the reproductive economy, particularly with respect to public reproductive provisioning. Earlier research has drawn attention to contradictions between ‘economic’ and ‘social’ goals within the Semester. For instance, the CSRs simultaneously urge social retrenchment and social investment policies (Crespy and Vanheuverzwijn, 2019, p. 107). We argue that such contradictions are intrinsically linked to, and made intelligible by, the economic epistemology and ontological reasoning of the AGSs, that separates the productive economy from the reproductive economy, whilst also devaluing and under‐theorizing the latter.

First, our analysis showed contradictions between the text in different sections of the AGSs which are linked to different governance tools, and overseen by different actors in the Commission. This undermined calls for investment, social goals and policy solutions in other parts of the documents. However, our analysis also showed that goals related to the accessibility and quality of welfare services and benefits, even in the ‘social’ sections of the documents, were often presented alongside contradictory economic goals, such as fiscal sustainability and work incentives. For example, the AGS 2016 argued that ‘[m]ore effective social protection systems are needed to confront poverty and social exclusion, while preserving sustainable public finances and incentives to work’ (European Commission, 2015, p. 12). The AGSs 2018 on the other hand argued that ‘[r]eforms of health care and long‐term care systems need to be pursued to enhance their cost‐effectiveness, ensure their fiscal sustainability and ensure quality, affordable access’ (European Commission, 2017, p. 11). To understand the true implications of these competing goals, their relationship to other documents and policies needs to be considered. The ‘social’ goals expressed in the AGSs are vague and expressed in such broad terms that considerable room for interpretation is often left (for example ‘adequacy’). ‘Economic’ goals, however, have no such ambiguity; they are linked to precisely defined fiscal targets embedded in the SGP, an enforcement mechanism which enjoys legal clout. This produces incentives to resolve the contradictions in favour of fiscal rules (Crespy and Vanheuverzwijn, 2019, p. 107).

We argue that these contradictions illustrate the split between the productive and reproductive sectors of the economy systematically embedded in the economic ontology and epistemology that underpin these documents. When the productive and reproductive spheres are seen as separate, there is no need for the economic actors at the European Commission to really tackle how the priorities related to the productive sphere affect the reproductive sphere and vice versa (Grown et al., 2000). The resolution of these tensions is delegated elsewhere – for example, to those delivering care in member states. Feminist economic epistemologies, in contrast, allow the depiction of these relationships in greater detail, using measures such as ‘depletion’ of the social reproductive sphere (Rai et al., 2014) or assessments of human welfare (Power, 2004).

Conclusions

By questioning what is actually meant by ‘the economy’, and what this meaning marginalizes or obscures, we have provided a new perspective on the dynamics that underpin the continued hierarchy between social and economic goals in the EU's economic governance. Our analysis focused on the AGSs to enable a more rigorous examination of the common framework of goals for the EU member states, which the Commission aims to pursue. By examining the mainstream economic epistemology replicated in the AGSs, ³ we have uncovered important collective assumptions about the economy, how it works and how it should be discussed.

We have argued that the European Semester is underpinned by a false distinction between the productive and reproductive sectors of the economy, that is typical of mainstream economic thinking. In the AGSs, the productive economy is discussed and represented as though it is detached from, and independent of, the reproductive economy. We showed that depictions of the reproductive economy were vague and that its relationship to the productive economy were under‐theorised, and we maintain, that this leads to the misrepresentation of social reproductive labour as an irrelevance or a cost, rather than a foundational economic input. This systematic devaluation of the reproductive economy is, we argue, embedded in the key concepts and measurements of the AGSs, such as growth and investment. This limits what is collectively intelligible for actors participating in the European Semester process. It creates blind spots and dead ends. As long as the key measurements and concepts used in the AGSs replicate this ontology and epistemology, the integration of social goals and policies in the European Semester will remain, at best, incoherent and partial. This barrier has not been recognized in the extant literature on the social/economic relationship within the Semester. Indeed, we see the distinction between ‘the economic’ and ‘the social’ that is repeated in policy and existing literature, as a conceptual dead end.

When policy makers recognize that social reproduction is a form of investment that has macroeconomic consequences, ‘an array of social policies are elevated to the status of macroeconomic policies’ (Heintz, 2019, p. 86, our emphasis). In contrast, as long as the conceptual distinction between the economic and the social is used, ‘social’ goals will be misrepresented as mere expenses, which are optional and secondary to growth and employment, while fiscal targets will remain represented as ‘core’ or ‘real’ economic concerns. This distinction also means that the impacts of austerity policies that have been disproportionately felt by women, especially women of colour and/or migrant origin, and the causal processes behind them, will remain under‐theorized in economic policy, and that ameliorative policy prescriptions will remain vague and contradictory. In this respect, our analysis highlights the importance of including dissident voices in debates on core EU policies. Bringing the insights of feminist economics and feminist political economy to bear on the European Semester enabled us to reveal a facet of the Semester not yet considered in mainstream debates.

Shifting towards an economic model that views reproductive labour as an economic input, would entail drawing on an evidence‐base that includes nuanced conceptualizations of the reproductive economy, such as measurements of the reproductive subsidy or the depletion of social reproductive capacities (Rai et al., 2014). Whereas feminist economists' arguments about the links between the productive and reproductive economy have been ignored by mainstream economists and economic policymakers for decades, the Covid‐19 pandemic has made it much more obvious than before that the productive economy is dependent on the reproductive economy. We have experienced how under‐resourced healthcare services that fail to cope with a pandemic, make for a considerably less efficient economy that is vulnerable to a complete stand‐still. Subsequent shutdowns of schools and nurseries have in turn shown that participation in education or employment is dependent on freedom from care of dependents – it is not possible to study or work in paid employment and care simultaneously. Standard economic measurements have woefully misrepresented the relationship between productive and reproductive sectors for decades, and as we have demonstrated, these measurements are at the heart of the EU's economic governance.

Adopting an economic epistemology that views social reproductive labour as a macroeconomic input, opens possibilities for both a recognition of the collective economic benefit of social reproductive labour, and a consideration of its redistribution. Both paid and unpaid care work could be recognized in macroeconomic planning as infrastructure, while spending on it could be considered as investment. Inter‐relationships between public services, unpaid care, employment and economic growth would become more accurately portrayed (Heintz, 2019; O'Hagan, 2018, p. 25). This would also facilitate a movement away from individualized notions of ‘work‐life balance’ that depend upon the assumption that social reproduction, or ‘life’, is an awkward hindrance to economically meaningful and important work. Such an approach provides a counterweight to recently evolving EU narratives of ‘social rights’ which take a highly individualized approach and relate too closely to workforce participation capacity.

We note, however, that we cannot yet identify a nation state that successfully applies such a model (O'Hagan and Klatzer, 2018), and that doing so would represent a radical change. Despite the recent realities revealed by the Covid‐19 pandemic, early analyses of EU policy responses show the same entrenched assumptions shaping policy priorities (e.g., Klatzer and Rinaldi, 2020). This underscores how durable these assumptions are, and that we must acknowledge, study and tackle them. The real life consequence of continuing with the same mainstream economic ontology and epistemology is an inequitably distributed downward spiral of human welfare. Reproductive capacity is continually depleted by macroeconomic policies, whilst policymakers wait for increases in growth or productivity before tackling economic pressures in the reproductive sector that they misapprehend as ‘social’, rather than economic issues. This, we argue, is how crises in social reproduction are repeatedly driven.

Funding details

This research was funded by the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme, grant number: 771676.

Disclosure statement

No potential conflicts of interest.

Supporting information

Table S1: Frequency of references to different aspects of the social/economic relationship in the Annual Growth Surveys 2011–19, by year

Table S2: Frequency of acknowledgements of ‘social impacts’ in the Annual Growth Surveys 2011–2019, by year.

Data S1: Supporting Information

Cavaghan, R. , and Elomäki, A. (2022) Dead Ends and Blind Spots in the European Semester: The Epistemological Foundation of the Crisis in Social Reproduction. JCMS: Journal of Common Market Studies, 60: 885–902. 10.1111/jcms.13288.