Content analysis is a valuable tool for analysing policy discourse, but annotation by humans is costly and time consuming. ChatGPT is a potentially valuable tool to partially automate content analysis for policy debates, largely replacing human annotators. We evaluate ChatGPT’s ability to classify documents using pre-defined argument descriptions, comparing its performance with human annotators for two policy debates: the Universal Basic Income debate on Dutch Twitter (2014–2016) and the pension reforms debate in German newspapers (1993–2001). We use the API (GPT-4 Turbo) and user interface version (GPT-4) and evaluate multiple performance metrics (accuracy, precision and recall). ChatGPT is highly reliable and accurate in classifying pre-defined arguments across datasets. However, precision and recall are much lower, and vary strongly between arguments. These results hold for both datasets, despite differences in language and media type. Moreover, the cut-off method proposed in this paper may aid researchers in navigating the trade-off between detection and noise. Overall, we do not (yet) recommend a blind application of ChatGPT to classify arguments in policy debates. Those interested in adopting this tool should manually validate bot classifications before using them in further analyses. At least for now, human annotators are here to stay.

By 2025, over eight million UK households will be receiving Universal Credit (UC). Introduced in 2013 to simplify the benefit system and improve work incentives for working age adults, UC has been criticised for causing hardship and exacerbating inequalities. There is limited research on children and young people’s (CYP) views of UC, as well as its health and social impacts. In this pilot qualitative study, creative methods were used to understand the views of UC among CYP (n = 40) aged 12–16 years in North East England. Findings showed diverse and nuanced understanding of UC as well as contested views about conditionality, sanctions, lower UC rates for under-25s and the two-child limit alongside recognition of the stigma and shame associated with benefits. While CYP value paid employment, they stressed the importance of minimum income standards and tailored employment support for UC claimants, taking account of their personal, health and family circumstances. Findings suggest CYP are aware when parents and carers are struggling financially and may try to ease pressures on parents. Debates about principles of equality, fairness, social justice and deservingness were present in young people’s accounts. We conclude by exploring future directions for a CYP-centred approach to social policy.

We are living through cruel and frightening times. How should a progressive policy studies respond? Critique undoubtedly plays a role: the task of exposing the structural conditions, political interests and power asymmetries that lie beneath the ‘prosaic surface’ of policy is an urgent one. But are these primarily deconstructive efforts enough? Can they lead us out of this quagmire, alone? In this article, we argue that something additional – something more generative and hopeful – is also required. In response, we introduce ‘critical utopian policy analysis’ (CUPA) a methodological elaboration of critical policy analysis (CPA) designed to support its use in both deconstructive and reconstructive policy efforts. This approach builds on the theoretical offerings of critical policy analysis, utopianism and prefiguration, to posit a methodological embrace of critique, imagination, enactment and play. It seeks to mobilise a complex nexus of affect – including heartbreak and hope – to motivate and support a range of intellectual undertakings and emancipatory politics.

The past decade has been marked by cuts in public funding of adult social care alongside an increased policy focus within the UK on extending working lives through ‘50 PLUS Choices’. This study uses the UK Household Longitudinal Study (2009/10–2018/19) to examine the relationship between informal care provision and labour market participation. The analysis focusses on mid-life, a period of life course characterised by both the uptake of informal care provision responsibilities and withdrawal from the labour market. Across the observation period, employment increased amongst both mid-life carers and non-carers, but the gap widened – with carers being much less likely to be employed. Discrete-time survival models assess the effect of caregiving on the likelihood of changing from full-time to part-time work or leaving work altogether. A range of indicators of caregiving, including care intensity, type of care provided and relationship to the person cared for, are all associated with reduced employment. The analysis supports the argument that policies promoting higher labour force participation amongst older workers are incompatible with cuts in funding for adult social care; to realise ‘50 PLUS Choices’, older working carers need to be better supported in juggling the competing demands of care and work.

Welfare state attitudes make up an interactive feedback loop of defining popular legitimacy and future policy trajectories. Understanding attitudinal drivers is thus essential political knowledge. However, as existing research is mainly based on the work-nexus of welfare, this article expands the literature to the welfare state’s care-nexus, examining drivers of family policy attitudes. We argue that conventional attitude predictors of self-interest and ideology are insufficient to explain the attitudinal cleavage in family policy. Instead, justice perceptions in the division of physical and cognitive household labour represent an important normative battleground. We test this with Norwegian survey data (N = 3500), using a unique vignette experiment to operationalise justice perceptions. Findings show that individuals who do not perceive a disproportional household labour division as unfair prefer optional familialism within family policy. Individuals who do perceive unfairness in a disproportional household labour division prefer de-familialism, which facilitates gender equality in public and private spheres. This is consistently found for the physical division of labour, while the cognitive dimension seems less politicised. We conclude that the battleground for different family policy approaches is fundamentally normative and linked to justice considerations on gender roles.

The aim of this study was to understand how and why relational welfare works to support young people who are not in employment, education or training (NEET). It builds on research discussing the limitations of work-first and human capital strategies in social policy while responding to calls for theory-driven insights into initiatives that move beyond employability and rapid employment. The material for this realist evaluation includes programme documents, fieldnotes and 75 interviews with practitioners and participants in community-based multicomponent initiatives delivered by Swedish municipalities. These data were scrutinised against programme theories while integrating literature on relational welfare as underpinned by co-creation and capability approaches. The results illustrate how flexible, challenging and coordinated programming strengthen beings and doings of young people in NEET situations while improving their wellbeing by overcoming isolation and forming a future orientation. The study provides guidance for supporting NEET-situated young people through a relational approach to welfare. It also offers a model against which local initiatives provided to a youth group high on the policy agenda can be mapped.

Among the factors identified to account for non-take-up of social benefits, there has been limited research on ‘process costs’, particularly regarding the impact of geographic access. Using Israeli data on field office openings from 1993 to 2021, this paper investigates the impact of geographic access on the take-up of the five largest social security programs in Israel. Based on staggered openings and closings of social security field offices, we find that geographic access has no significant impact on the take-up of either automatic enrollment programs, such as child allowances, or non-automatic programs, such as disability benefits. These findings suggest that the effect of geographic access on the take-up of social benefits may have been overstated in previous studies. We propose the following hypothesis to explain the surprising findings: If enhanced geographic access is driven by political favoritism, opening of new service points may lead to the misallocation of resources and, in effect, increase administrative burdens, thereby undermining rather than improving the take-up of social benefits.

This article addresses whether responses to COVID-19 created opportunities for future policy change. We explore this matter by presenting a framework rooted in political economy and the literature on pandemics. We argue that the opportunities created by emergency responses are context-specific and that narratives, policy tools, and pro-equity state actors are variables that mediate emergency responses and future opportunities. We ground our analytical contribution on the emergency cash transfers deployed during 2020 following the COVID-19 outbreak in two contrasting Central American countries, Costa Rica and Guatemala. The paper promotes further policy discussion on the opportunities for progressive change in unequal contexts.

Work-related conditionality policy in the UK is built around the problematic assumption that people should commit to ‘full-time’ work and job search efforts as a condition of receiving benefits. This is potentially in conflict with the idea that what is required of people should be tailored to their circumstances in some way – ‘personalised conditionality’ – and implies a failure to recognise that conditionality is being applied to a diverse group of people and in a context where the paid work that is available is often temporary and insecure. Drawing on thirty-three qualitative interviews with people subject to intensive work-related conditionality whilst receiving Universal Credit or Jobseeker’s Allowance in Manchester, the paper explores the work-related time demands that people were facing and argues that these provide a lens for examining the rigidities and contradictions of conditionality policy. The findings indicate that expectations are often set in relation to an ideal of full-time hours and in a highly asymmetric context that is far from conducive to being able to negotiate a reasonable set of work-related expectations. Work search requirements affect people differently depending on their personal circumstances and demand-side factors, and can act to weaken the position of people entering, or already in, work.

Over recent decades it has consistently been shown that disabled adults in the UK fare worse in the labour market and have lower levels of wellbeing than non-disabled adults. However, this is in part due to the selection into dis-ability of those with existing socio-economic disadvantages. In this article, we use panel data from the combined British Household Panel Survey and Understanding Society, covering the 27 years from 1991 to 2018, to distinguish between the effect of selection, the effect of dis-ability onset and the effect of dis-ability duration on a range of labour market and wellbeing outcomes. We show that there is important selection both into dis-ability and into longer experience of dis-ability on the basis of observable characteristics. We also show the importance of controlling for time-invariant unobservable individual characteristics that similarly affect selection into dis-ability and duration of dis-ability. Even after controlling for both forms of selection, we find significant negative effects of dis-ability onset and duration, and offer policy solutions to address them.

As climate change progresses, natural hazards are projected to continue to increase in frequency and intensity, posing a new form of social risk, implicating both the welfare and environmental state and raising the salience of ecosocial policy as a mechanism to attend to the distributional effects of climate change mitigation and adaptation. This study posits a novel conceptual framework for ecosocial policy and offers the US ecosocial safety net as a case analysis. While we conceptualise disaster relief policy as a mode of the environmental state, it includes unique ecosocial policies that constitute the backbone of the US ecosocial safety net. This study describes and compares the developmental and functional synergies between the US welfare and environmental state manifested in the form of an ecosocial safety net by explicating the Individual Assistance Program and the National Flood Insurance Program. Our findings reveal synergies between US disaster relief and welfare, including parallel developmental trends, philosophies of deserving/undeserving, functions of racial capitalism and relationships with economic growth. This study and its conceptual framework of ecosocial policy offer a groundwork for the study of ecosocial policy in other contexts.

Social policies convey normative assumptions about how households should make ends meet and organise care, but how do these ideals withstand crises such as the coronavirus disease 2019 (COVID-19) pandemic? Previous research shows continuity of welfare state models in the crisis, but mostly looked at single policy fields and produced mixed findings regarding the role of pre-crisis reform trajectories. This paper contributes a detailed analysis of assumptions about the ‘standard productive household’ in terms of three dimensions: labour market participation, coverage of economic needs and coverage of care needs. Drawing on original policy documents enacted in 2020 in Germany – which had dismantled many of its institutional strongholds for the male-breadwinner model before the crisis – we provide two novel insights. First, social policy responses to the pandemic were relatively coherent regarding assumptions about labour market participation, but expectations towards households’ abilities to make ends meet and parents’ care involvement were less coherent. In addition to relaxing conditions on stable employment and income, policy responses normalised patchwork incomes and relied on parents to compress paid and unpaid work. Second, we propose that crises may slow down reform processes that are already underway by reverting to ideas that were dominant in the past.

The United Kingdom’s National Referral Mechanism (NRM) is a framework for identifying potential victims of modern slavery (slavery, servitude, forced labour or human trafficking) and ensuring that they receive adequate care. This research explores differences in referrals and outcomes of potential modern slavery victims within the NRM on the basis of individual attributes, geography and first responder. Findings are based on exploratory analysis of data on 55,000 cases released by the Home Office in spring 2022 plus data from four Freedom of Information requests. Findings confirm that there are significant differences in rates of positive outcomes between native and immigrant groups, with native populations more likely to receive positive conclusive grounds (CG) decisions. Our key contribution is in identification of the role of the first responder in negatively influencing outcomes for victims of particular forms of exploitation. We suggest the differences in outcomes may be explained by the dual role played by first responders within the immigration system in identifying victims and implementing immigration control measures. We situate this finding within a broader critical migration literature on polymorphous borders pointing to the NRM as one mechanism through which bodies are differentially excluded from territorial access and associated rights or benefits.

The social protection of platform workers is considered one of the most precarious features and political challenges of this new form of employment. Still, there have only been a few empirical investigations on this issue to date. This article presents an explorative empirical analysis of the social protection of platform workers in Germany – a conservative welfare regime with a strong link between standard employment and institutionalised social protection. On the basis of an online survey amongst 719 self-employed platform workers, we examine how different employment patterns correspond to institutionalised protection against sickness and old age. We empirically explore different protection types and analyse how they differ regarding working conditions in platform work and individual social policy preferences. Findings reveal that conditions of platform work and social protection as well as demands and regulatory preferences vary notably across different clusters of platform workers. Still, the vast majority votes against obligatory social insurances for platform workers and favours self-employment over dependent employment. Against this background, we discuss challenges for future attempts aiming at improving social protection for platform workers. This study adds to the literature by empirically exploring platform workers’ social protection and social policy preferences, which have been overlooked to date.

Crises create opportunities for policy change, yet the extent to which they encourage redistribution is under-researched. We adopt a narrative approach to study how crisis frames are mobilised to support or oppose redistribution, and whether that redistribution is progressive or regressive. A typology of crisis narratives with different redistributive implications is presented: retrenchment narratives promote deregulation and cuts to welfare; Robin Hood narratives advocate progressive redistribution with expanded rights; and restoration narratives favour bringing back the status quo ex ante. We apply the Narrative Policy Framework to examine how Australian parliamentarians used the language of ‘housing crisis’ during and after COVID-19. Despite existing research suggesting crisis narratives mostly support retrenchment, Australia’s pandemic housing debates were dominated by Robin Hood and restoration narratives. We show that party ideology matters for the redistributive content of crisis narratives, but the effect of ideology is mediated by incumbency status. We conclude that shifts in the parliamentary balance of power lead to changes in political parties’ rhetorical support for redistribution.

Governments around the world have increasingly adopted behavioural public policies, in which behavioural insights units (often designated as ‘Nudge units’) play an increasingly important role. Such units are typically assigned with implementing behavioural insights in various social policy fields. However, research on the contextual specificities of Nudge units is still scarce. This article addresses that gap by providing an in-depth qualitative study of MineduLAB, a Nudge unit active within the Peruvian Ministry of Education. Informed by an interpretive approach, our analysis shows that the behavioural insights that MineduLAB makes use of and develops, are enacted by a unique local interplay of three different dimensions: a political-financial, an epistemic community, and a sector-specific (here: educational) dimension. Moreover, our analysis of a Nudge unit in the Global South showcases both similarities (e.g. focus on behavioural evidence and experimentation) as well as differences (e.g. the challenge of corruption; the deployment of other vocabulary with regards to behavioural insights) with the predominantly Western Nudge units that are more commonly researched. Consequently, we argue that for a profound academic evaluation of behavioural public policy and the role of Nudge units herein, it is crucial to pay close attention to their contextuality and geographical diversity.

A substantial international body of evidence links housing to health outcomes. In 2021, the World Health Organisation (WHO) evaluated a small selection of policies from its six geographic regions and found that, in Australia as in the rest of the world, existing healthy housing measures fall short of the systemic response required to address health impacts and inequities. This paper takes the novel step of applying Bacchi’s (2009) ‘What is the Problem Represented to Be?’ approach to a wide-ranging thematic analysis of over 300 Australian policies across the domains of health and housing and related policy areas. In so doing, it offers an overview of existing healthy housing policy as well as illuminating the conceptual understandings and priorities of policy makers, shedding light on the policy paradigms that see housing under-utilised as a preventive health and health equity measure.

The UK government has called for employers to make work adaptations in response to changes in health individuals may experience as they age. However, government assumptions place too much emphasis on the voluntary actions of employers and managers, without placing the management of health in a wider context. Drawing on insights from Thompson’s disconnected capitalism thesis, we explore whether financial/competitive pressures facing many private and public sector organisations today, alongside other factors, contribute to organisations not considering or implementing work adaptations. In this context, it is suggested that older workers may also hide health issues because of anxiety, or ‘ontological precarity’, regarding working longer. Qualitative case studies compare the delivery of work adaptations in three organisations: ‘Local Government’, ‘Hospitality’, and ‘Trains’. Work adaptations were only widely available in Trains; this was for a range of reasons, including the fact that Trains was relatively insulated from financial pressures and able to deliver job and financial security for older workers. As many older workers will continue to be employed by organisations similar to Local Government and Hospitality, we argue that policy makers cannot rely solely on employers to make adaptations.

Multiple welfare states are re-emphasising the need for street-level bureaucrats’ (SLBs) discretion to stimulate responsive service provision. However, little is known about how SLBs with diverse backgrounds in inter-departmental settings deliberate what it means to use discretion well when different rules, eligibility criteria, and interpretations apply to a client. We address this gap by investigating the stories that participants of a Dutch policy experiment told each other to justify which clients should be granted a flexible interpretation of entitlement categories amid scarcity. We found that ‘caretakers’ used the ‘victim of circumstances’ and ‘good citizen’ plot-type to convince ‘service providers’ that the use of discretion was the right thing to do, whereas the latter used the ‘not needy enough’ or ‘the irresponsible citizen’ plot-type for contestation. Our analysis shows that storytelling helped SLBs to make sense of and bring cohesion to complex situations. Moreover, the analysis shows how stories can have a strong emotional appeal and create a sense of urgency to act collectively, yet can also create divisions and opposition among SLBs. As such, storytelling influences how SLBs think and feel about the client, themselves, and each other, and influences how discretion is used at the front-line of public policy.

Since the onset of the COVID-19 pandemic, both viral infection and the corresponding economic turmoil have wreaked havoc across the globe, highlighting the imperative function of the state as a social protection provider. The pandemic has seemingly created favourable political circumstances for rapidly expanding social protection, but its influences on public welfare attitudes remain unclear. In this study, I argue that the impact of pandemic-driven economic risk is too limited to spur strong public support for social protection. The employed empirical analyses using panel data collected in South Korea show that unemployment induced by the pandemic is conducive to higher degrees of individual support for social protection measures, but the impact is only short-lived. Further analyses show that, once individuals are re-employed and as the time spent in economic difficulties becomes more distant, personal unemployment experiences are no longer positively associated with support for social protection. Finally, pandemic-induced unemployment experiences have a lasting impact primarily on young adults. The evidence therefore suggests that significant institutional changes in the welfare state are hard to achieve by solely relying on the impact of economic risk.