The political economy of UK welfare: institutions and reform trajectories

There are a number of alternative ways to conceptualise the most important determinants of welfare reform processes. Perhaps most importantly, we can distinguish institutionalist explanations that centre the role of structures and institutions in potentiating or precluding alternative reform trajectories from agent-centred explanations which focus on the motivations and strategies of actors pursuing conflictual goals through the electoral system. We follow Beramendi et al.’s (Reference Beramendi, Haüsermann, Kitschelt and Kriesi2015) attempts to integrate these approaches together in a model of ‘constrained partisanship’, which distinguishes the demand and supply sides of public policy. ‘Demand’ corresponds to the ‘policy problems’ that different actors identify and seek to be addressed through political processes, and ‘supply’ to the feasible set of policy ‘solutions’ that politicians and parties propose to address them. On the demand side, countries vary with respect to the electoral significance of different interest groups (their size, organisation and the power resources they command) as well as the nature and extent of different policy problems that these groups face. The supply side, in contrast, reflects constraints on state capacities as well as sunk costs and vested interests arising from inherited policies and institutions. In other words, this approach incorporates an institutionalist interpretation into analysis of the electoral dynamics underpinning welfare reform trajectories, thus bridging an important gap in the literature and providing a more holistic understanding of the combination of factors that determine the feasibility of specific policies being introduced in different country contexts.

The literature on comparative welfare state reform commonly points to the persistence of so-called welfare ‘regimes’ characterised by path dependency (Esping-Andersen, Reference Esping-Andersen1990; Pierson, Reference Pierson1998; Bonoli and Palier, Reference Bonoli and Palier2000). However, welfare states are not ‘frozen landscapes’, and, under pressure from a series of socio-economic challenges, have changed profoundly in recent years. Long-term structural transformations – among them deindustrialisation, globalisation and technological change – have eroded the industrial working class as an electoral force underpinning support for generous compensatory social security (Gingrich and Häusermann, Reference Gingrich and Häusermann2015). Socio-demographic changes of population ageing, increased female labour force participation, and changing family forms – coupled with increasing numbers of disadvantaged workers ‘left behind’ by the demands of the knowledge-intensive, service-focused economy – have been conceptualised as ‘new social risks’ (Bonoli, Reference Bonoli2005). As a result, over a time period in which fiscal pressures have intensified and attitudes to welfare have hardened across Europe – an era of ‘permanent austerity’ in Pierson's (Reference Pierson1998) phrasing – there have emerged new demands for increased social spending. Governments across Europe have reacted to these joint challenges in diverse ways, with the direction and pace of reform partly determined by prevailing policy and institutional conditions.

Esping-Andersen's (Reference Esping-Andersen1990) famous typology placed the UK within the ‘liberal’ grouping of largely Anglophone countries, characterised by low replacement rates, extensive means-testing and residual coverage in the welfare state. While Esping-Andersen's framework has been subject to extensive debate over many years, it is generally accepted that the UK remains a liberal welfare regime, despite significant reforms during the New Labour era of government (1997-2010). Indeed, according to Van Kersbergen et al. (Reference Van Kersbergen, Vis and Hemerijck2014), ‘the welfare regime's market-orientation has become stronger and more pronounced during and since the Thatcher-era’. Means-testing has become increasingly prominent, emerging to eclipse contributory and universal payments in magnitude. Between 1978 and 2013, spending on means-tested benefits as a proportion of total working-age welfare expenditure tripled from 24 per cent to 79 per cent (Hood and Oakley, Reference Hood and Oakley2014). Comprehensive universalism in the UK is now largely restricted to the provision of public services, such as the National Health Service and schooling, rather than social security benefits.

More generally, and in keeping with the interests of dominant actors in the productive economy, the UK's response to deindustrialisation has been to promote labour market flexibility and the growth of employment in services (Wren, Reference Wren2013). This has served to promote relatively high levels of employment, including in highly-skilled service sector occupations such as finance, professional services and ICT, but has also contributed to extensive low-skilled employment, increased wage inequality and relatively high rates of poverty amongst the working age population.

This, in turn, has generated pressure for the introduction of in-work benefits (alongside stringent and increasingly punitive active labour market conditionality) to alleviate poverty caused by low pay, and to reduce poverty and unemployment traps. Investment in human capital formation for the low and middle tiers of the labour market, particularly in employment-specific vocational training, remains weak, in keeping with a ‘work-first’ strategy of labour market activation. High levels of employment reduce political pressures for social insurance as wage replacement, while expansive participation in higher education (and the earnings that flow from graduate occupations) provides a form of income protection for the professional middle classes. Thus we observe ‘institutional complementarities’ between the UK's underlying political economy and its recent welfare reform trajectory, as predicted by the varieties of capitalism and associated comparative welfare state literatures (Iversen and Soskice, Reference Iversen and Soskice2009; Schröder, Reference Schröder2013).

The reforms of the New Labour era were broadly indicative of a shift towards a ‘social investment’ strategy that is observed, to a greater or lesser extent, in welfare states across Europe (Hemerijck, Reference Hemerijck2017). An important pillar of these reforms in the UK was the introduction of Tax Credits, which combine an in-work wage subsidy element with support for families with children. Tax Credits have been described as a form of ‘progressive universalism’ – i.e. targeted at those on low incomes via means-testing, but with greatly expanded coverage in comparison to the system that Labour inherited in 1997. They contributed to impressive reductions in child poverty rates and the number of workless households in the decade before the financial crisis (Millar, Reference Millar and Millar2009). Extensive investments were also made in early years education and childcare, although coverage of these services and fiscal transfers remains partial, rather than universal, and participation is characterised by social class inequalities (Ghysels and Van Lancker, Reference Ghysels and Van Lancker2011).

Another important aspect of New Labour reforms was state pension reform. In response to unacceptable levels of pensioner poverty, in 1999 the government introduced the means-tested Minimum Income Guarantee (subsequently replaced by Pension Credit in 2003) to bring pensioners’ income levels up to a specified minimum. Following the final report of the Pensions Commission in 2005, a new architecture for pension policy was introduced, combining a higher, single-tier Basic State Pension (BSP) indexed to earnings, auto-enrolment into occupational pension schemes, and increases in the state pension age.

Since 2010, when a Conservative-Liberal Democrat Coalition government came to office, followed by Conservative administrations elected in 2015 and 2017, central institutional features of the UK's liberal welfare regime have been consolidated, whilst significant demographic inequalities – between young and old – have emerged in welfare state spending and entitlements. Tax credits and benefits for the working age population have been cut or frozen, entitlement to Child Benefit means-tested, and benefit conditionality for claimants intensified. Under the new system of Universal Credit, which integrates a number of different benefits into one monthly payment, conditionality will be extended to households in work as well as the unemployed. Means-testing and residualisation of social security have increased, whilst universalism and contributory entitlements in the welfare state have diminished.

At the same time, the BSP has been protected by a ’triple-lock’ of annual uprating by inflation, average earnings or 2.5 per cent, whichever is higher. As a result of these measures, and wider socio-economic changes, there has been significant divergence between the real household incomes of pensioners and those of the working age population in the period since the financial crisis (Corlett and Clarke, Reference Corlett and Clarke2017). This is significant for both fiscal and political reasons: state pensions form by far the biggest element of UK social security spending, and political support for the BSP is high. Older people also have much higher rates of turnout at general elections, ensuring a solid protective bloc of political support for pensioners’ welfare entitlements.

In contrast, according to Larsen (Reference Larsen2008), liberal welfare systems ensure that public support for working age welfare recipients remains low. The logic is that because unemployment levels are low, the unemployed and inactive are seen as responsible for their plight (and thus undeserving of generous support). Furthermore, because replacement rates and other social transfers are low, welfare recipients are observably poor and can be identified and stigmatised as belonging to an inferior social class. Politicians can more readily frame those on out-of-work benefits as ‘undeserving’. Indeed, public support for working age welfare has declined significantly, especially for able-bodied claimants (Pearce and Taylor, Reference Pearce, Taylor, Park, Bryson, Clery, Curtice and Phillips2013). Although the most recent attitudinal data (Clery et al., Reference Clery, Curtice and Harding2017) shows evidence of increased opposition to austerity measures, and a softening in attitudes to the unemployed, the broad trajectory over recent decades has been one in which social solidarity and public support for the working age welfare state has weakened.

Turning to recent developments in welfare state theory, it is conceivable to situate BI within a social investment theoretical framework that is consistent with an institutional evolution of the UK welfare state (Hemerijck, Reference Hemerijck2017). For example, a BI could be offered as minimum income guarantee, something that has arguably been lacking in the social investment paradigm, given its primary focus on ‘new social risks’ and associated employment and human capital investment policies. A BI might also constitute a platform for lifelong learning participation, and more flexible, responsive job search activities and/or self-employment. Thus conceived, BI could take its place within a capacious social investment state (SIS) strategy, as a policy well suited to a more inclusive welfare state in a liberal market economy. Furthermore, in the UK, the residual nature of the welfare state has one important political-economic consequence for a BI style scheme: it would not generate opposition from political actors committed to the defence of generous provisions of the kind that typify Nordic welfare states, and which a BI might threaten. Just as significantly, the relative dearth of contributory benefits in the UK welfare system (old age pensions excepted) means that it avoids the profound difficulties experienced by the ‘dualised’ continental European corporatist welfare states in their efforts towards reform; in those countries, policy makers face strong resistance from labour market ‘insiders’ to any attempts to reduce or withdraw accrued welfare entitlements, as well as the de facto veto powers of ‘social partners’ involved in the management of social insurance funds (Palier, Reference Palier and Palier2010). Yet, if these institutional features provide some cause for cautious optimism, the strength and cohesion of political support for BI remains uncertain, as examined in the next section.

The politics of basic income as welfare reform

BI has been positioned by many of its advocates as benefitting those that Guy Standing (Reference Standing2011) has termed the ‘precariat’ – workers lacking access to secure, full-time jobs and adequate social security protection. Although the concept of the ‘precariat’ is contested in the sociological literature, there is plentiful evidence that the numbers of working age adults who are self-employed (voluntarily or not) and/or on zero hours or short-term contracts and other forms of irregular employment, has grown in the past decade (Pennycook et al., Reference Pennycook, Cory and Alakeson2013; Broughton and Richards, Reference Broughton and Richards2016). The UK's high employment rate is mirrored by a significantly deregulated labour market in which millions of workers lack access to employment rights, contributory social security, and the protections afforded by trade union membership.

Yet whether this is a large and coherent enough group to form a constituency in favour of BI is far more debatable. Limited trade union coverage and sharp inequalities in social class rates of voting make it hard to mobilise low-skilled, low-waged workers into politics without the support of middle class voters in the UK. At the same time, different groups within the precariat, broadly conceived, may have very different welfare state preferences. Individuals such as graduates in temporary or self-employment – ‘high-skilled labour market outsiders’ (Häusermann et al., Reference Häusermann, Kurer and Schwander2015) – might support a basic unconditional income payment, particularly during their early years in the labour market, but stand to lose out financially as a result of tax rises detailed in specific policy proposals to pay for a BI. Their welfare state preferences will thus diverge from precariously employed low-skilled workers who are likely to support tax increases on the middle classes to pay for government redistribution. Similarly, existing welfare claimants might support reduced bureaucratic intrusion into their lives – especially in the context of increasingly intrusive and punitive conditions – but would oppose any BI scheme in which they were to experience a fall in income from means-tested benefits. Thus, even if dissatisfied with the status quo, it is by no means clear that each group would unambiguously favour BI vis-à-vis alternatives, let alone support it sufficiently to mobilise politically. In any case, many of those who might stand to benefit from BI in material terms might oppose it on normative grounds of fairness or reciprocity, particularly if it were extended to ‘undeserving’ groups. Welfare policies are not just differentiated based on the extent to which they ‘decommodify’ labour through generosity of payment levels and comprehensiveness of coverage; equally crucial are reciprocal obligations imposed on welfare recipients (e.g. relating to behavioural conditions) and the manner in which those ‘undeserving’ of support are excluded from entitlement. It is plausible that low income demographics would favour BI's distributive implications in the abstract but oppose the erosion of the reciprocity principle, the lack of conditions, and the implied extension of the political community to which welfare rights should extend. Indeed, this ‘multidimensionality’ of welfare preferences is well reflected in a tendency towards nationalist ‘welfare chauvinism’ – a desire to exclude migrants from social security and to erect barriers to free movement of labour – now found in the core constituencies of social democratic parties.

Multidimensionality of welfare preferences is one reason that BI sits uneasily in the UK policy landscape in terms of its reception within the main political parties. Another important consideration is that the policy actually represents a wide variety of diverse schemes. Given that the level of BI is not specified as a definitional attribute, it is easy to see that it could appeal equally to progressive movements favouring an expansive, redistributive role for the state in welfare provision, or to more conservative elements whose primary concerns are to reduce bureaucracy, tackle poverty and unemployment traps, and minimise the fiscal burden of welfare. In this respect the basic income movement is indeed a broad church.

But while in the abstract BI may appeal to a broad range of political persuasions, adopting specific features to propose a conservative or progressive form of BI is likely to lead to erosion of support (De Wispelaere, Reference De Wispelaere2016). So what are the concrete possibilities of either of the UK's major parties (Labour and the Conservatives) supporting BI? Both would face difficulties winning round a majority of their own supporters to BI, due to inherent opposition to some of its key features. Furthermore, a party system that grants the governing party a considerable degree of policy autonomy also generates incentives to retain the welfare policy status quo; ‘opportunities to transfer the blame to other groups for policy failures are limited and the party of government must ensure that policy directions are acceptable to a sufficient proportion of the electorate to be confident that it will retain power’ (Taylor-Gooby and Larsen, Reference Taylor-Gooby, Larsen and Taylor-Gooby2004). Martin (Reference Martin2016) interviews party activists regarding the potential for BI to be adopted. He finds cause for ‘pessimism among basic income advocates about internal party debates leading to basic income support’ (p. 120). And although the Labour Party established a working group to examine BI before the 2017 general election, there was no proposal to pilot BI schemes in its manifesto, as many had expected. Enthusiasm for BI appears restricted to regional leaders, party activists and some trade unionists – what De Wispelaere (Reference De Wispelaere2016) calls ‘cheap support’.

Our scepticism about the prospects for a powerful and coherent BI constituency are borne out by recent polling commissioned by the University of Bath's Institute for Policy Research (Ipsos Mori, Reference Mori2017). It is perhaps surprising to learn that approximately half of those surveyed expressed support for BI, with 26 per cent opposed. However, when respondents were asked to consider a BI funded through increased tax levels, net support dropped to 30 per cent (with 40 per cent opposed); when they were asked to consider a BI funded through cuts to other benefits, support dropped to 37 per cent (with 30 per cent opposed); and when asked to consider a BI funded by both tax increases and benefit cuts, net support fell to 22 per cent (with 47 per cent opposed). Recent microsimulation studies have shown that ‘fiscally feasible’ schemes require a combination of significant tax rises and cuts to welfare expenditure, suggesting that any realistic proposition would fall short of garnering the requisite public support.

The polling has other interesting implications. Not only do high levels of support dissolve when respondents are asked to consider ‘realistically funded’ schemes, but the results confirm our intuitive understanding that supporters of BI are not homogenous with respect to their preferences and priorities. 40 per cent of Conservative-leaning respondents support BI in general terms, compared to 63 per cent of Labour-leaning respondents, with 41 per cent and 17 per cent respectively opposing the idea. But besides being differently inclined to support BI – and perhaps more pertinently – Conservative and Labour voters are interested in different types of BI. Support among Conservative-leaning respondents grows to 49 per cent when asked to consider a BI funded through benefit cuts. While support among Labour-leaning respondents drops when asked to consider concrete funding models, this group has a clear preference for schemes which rely on tax increases – particularly on wealth – compared to those that require benefit cuts.

All this is suggestive of the argument (following De Wispelaere, Reference De Wispelaere2016) that BI's support base is characterised by persistent political division, making constituency- and coalition-building a very challenging endeavour. Adding to political difficulties, microsimulation models have generated concrete data indicating the intractable nature of the trade-offs in policy objectives that arise when comparing alternative BI schemes. In particular, there is a three-way trade-off between the goals of a) meeting need, b) controlling fiscal cost, and c) eliminating means-testing (and thus retaining the administrative simplicity that is a hallmark of the BI concept). Martinelli (Reference Martinelli2017b) modelled scenarios in which a modest BI replaces most social security benefits. The unavoidable conclusion was that despite significant fiscal effort, such schemes would lead to unacceptably large income losses for large numbers of households, and significant increases in poverty levels. Partial schemes, such as those proposed by Reed and Lansley (Reference Reed and Lansley2016) and Torry (Reference Torry2016), result in negligible numbers of households becoming worse off, and some improvements in poverty rates. As Martinelli (Reference Martinelli2017a: 74) notes, such schemes ‘clearly balance the goals of controlling fiscal expenditure and ensuring the minimisation of adverse distributional consequences more effectively’; however, at the same time it ‘is not clear that partial models will give rise to full complement of advantages on which BI is sold. . . [and such schemes] would have limited appeal to proponents whose priorities are to reduce bureaucratic complexity and reduce high withdrawal rates, thus reducing the prospects of a broad coalition in its favour’.

Possible trajectories for reform

The preceding paragraphs have arguably painted a pessimistic picture of BI's political prospects. But there remain several relatively promising avenues for reform that capture at least some of the characteristics of a full BI.

Jordan (Reference Jordan2012) considered the possibility that Universal Credit might be a route to BI ‘through the back door’. The argument is that Universal Credit's core features – consolidation of numerous working age benefits, and the harmonisation of administrative procedures and conditions for working and out-of-work claimants – are motivated by many of the same arguments that support BI. However, Universal Credit is a means-tested household benefit, not a universal individual income payment, and receipt is dependent on significant and extensive conditions (which apply to those in low-paid work, as well as those out of work). The ‘work allowance’ within Universal Credit – which might have functioned as a quasi-basic income allowance – has also been cut back significantly from its original design. It would therefore appear unlikely that a BI scheme can be built out of the Universal Credit platform.

Ironically, it is the growing numbers of claimants affected by the worst excesses of the existing benefit conditionality from which basic income advocates might draw support to their cause. Thus one feasible direction of reform – falling short of BI per se – might be towards a softening of the increasingly stringent labour market conditions that are attached to benefits. This could include a reduction in the scope and severity of sanctions, or the broadening of the scope of behavioural requirements to include forms of valuable non-labour market activities such as those mentioned above. As noted in the introduction, there is some appetite to look more closely at BI within devolved parliaments and local authorities in the UK. The trials that may result from ongoing consultations might end up falling short of advocates’ hopes, as a result of the limited powers invested in the Scottish Parliament and Welsh Assembly to vary tax allowances or set the levels and entitlement rules for the main social security benefits. But there may be scope for the authorities to experiment with the conditionality and sanctions regimes imposed on claimants, and to use their tax and benefit powers at the margins to facilitate experimentation. Opposition to sanctions and their effect on minimum living standards for social security claimants appears to have driven much of the recent upsurge of interest in BI from sub-national policymakers and grassroots organisations.

Another possible reform trajectory is the introduction of a BI for specific age groups. The potential advantages of such schemes are that they are significantly cheaper than truly universal schemes, and can be targeted at demographics with persuasive cases for public support. The most obvious would be the introduction of a ‘citizen's pension’. The past decades of reform have already seen the erosion of the contributory principle (via a reduction in the number of years required to qualify for a full BSP, and the introduction of National Insurance credits for parents and carers) and the introduction of a (means-tested) minimum income guarantee for pensioners. As mentioned above, pensioners are frequently seen as more ‘deserving’ than working age welfare recipients, and comprise a disproportionately powerful electoral constituency (Berry, Reference Berry2014). But despite the announcement of a more generous single-tier BSP, a universal citizen's pension would still represent a significant fiscal effort for relatively little benefit: according to Martinelli (Reference Martinelli2017b) it would cost £24bn (in 2016/17 prices) – increasing existing pension spending by around a third – while only leading to a very marginal reduction in household poverty levels (a fall of around 3 per cent), and even contributing to a small increase in inequality. Nonetheless, in the context of an ageing society, in which millions of pensioners will not have built up a reasonable private or occupational pension and therefore must rely primarily on the BSP, the prospects of a citizen's pension must be considered better than for many other options.

Another age-specific BI might extend to young adults. Rising awareness of intergenerational inequalities has focused political attention on the young and their welfare state entitlements. Martinelli (Reference Martinelli2017b) modelled a version of a Young Adult's Income in which the main working age benefits were withdrawn alongside the introduction of a BI of equivalent value to basic payment rates. The scheme implied a net fiscal cost of £23bn. It reduced household poverty by around 8 per cent, with a greater reduction in child poverty of 11 per cent. Given normative opposition to unconditional benefits and the lack of public support for working age welfare, this scheme could be given greater political legitimacy if presented as a variant of Atkinson's (Reference Atkinson1996) participation income – that is, attached to conditions to engage in socially useful activities such as education, voluntary work, and care. The likelihood of this option attracting support will also depend on how it is seen to interact with tuition fees and maintenance support for young adults in further and higher education: as a starting point it is possible to envisage a young persons’ basic income as a maintenance grant for all young people, whatever form of post-compulsory education and training they undertake (this would be similar to the Australian Youth Allowance for sixteen to twenty-one year olds in education and training or looking for work).

One final reform trajectory worthy of note is to frame BI as a more prosaic tax policy reform. The UK has a relatively generous personal income tax allowance. (The personal allowance is currently set at £11,500 per annum, which implies a ‘tax expenditure’ for working people equal to £2,200.) As Painter (Reference Painter2016) notes, this could readily be repurposed as a universal, fully refundable tax credit. Martinelli (Reference Martinelli2017b) found that this would have a net fiscal cost of around additional £36bn, after eliminating the personal tax allowance as well as Child Benefit, and recalculating means-tested benefits to take the new BI into account. Such a reform would be consistent with the development of tax policy in recent years, but would radicalise it in the direction of universality.

This overview of possible reform trajectories for BI in the UK suggests that the most likely is incremental reform and experimentalism made possible by measures of tax and social security devolution to Scotland and Wales (and in the future, perhaps, to English cities). These will keep debates about BI on the public policy agenda and may generate useful empirical results. Age-related BI schemes also offer promising political-economic constituencies and can plausibly be aligned with developments in existing welfare state institutions and entitlements.