Public policy futures: A Left trilemma?

The long term

The OBR Fiscal Sustainability Report (OBR, 2011b) attempts to project future government spending and revenues over a fifty-year period. Such projections are subject to considerable uncertainty, since it is impossible to predict growth, interest rates, changes in productivity growth, migration and other factors with certainty. A number of different scenarios under differing assumptions are developed. The basic finding is that ‘public finances are likely to come under pressure over the longer term, primarily as a result of an ageing population … Government would end up having to spend more … on … pensions and healthcare. But the same demographic trends would leave government revenue roughly stable’ (OBR, 2011b: 3). The central (and optimistic) assumption (Figure 1) shows that the 8 per cent gap between spending and revenue in 2010–11 is removed by 2020–21 as a result of stringent cuts that bring spending as a proportion of GDP to 1998 levels, assuming a return to pre-crisis growth. Spending then returns to long-term trend levels and the balance deteriorates, with a gap of 0.6 per cent of GDP in 2030–31 widening to 3.2 per cent by 2060–61. This implies net borrowing of 7.7 per cent of GDP and a debt of 107 per cent of GDP (calculated from OBR, 2011b: Table 3.6). In other words, even after the cuts, economic pressures continue.

OPEN IN VIEWER

Figure 1.

Office for Budget Responsibility central long-term projection: spending, revenue and the balance (%GDP).

Four points should be noted, concerning benefit uprating, state sector productivity, restructuring costs and inequality.

Benefit cuts

First the report generally assumes current Coalition plans will be realized, except that short-term benefits will be uprated by earnings rather than the Consumer Prices Index (CPI), the latter being the current policy. Maintaining CPI uprating would cut spending by about 1.6 per cent of GDP and benefits by about 31 per cent by 2030–31, with further cumulative savings (OBR, 2011c: Annex C, p. 34). This would resolve the problems of the spending gap, at severe cost to very poor people, by reducing benefits for those of working age by about 14 per cent compared with general inflation and nearly three times that compared with projected earnings (OBR, 2011c: Annex C, p. 34).

Productivity

Secondly, the projections assume real improvements in productivity across the public sector in line with those in the rest of the economy. If public sector productivity falls behind (and public sector workers do not have corresponding pay cuts), the real cost of maintaining standards in the public sector, and the proportion of GDP it absorbs, rises (OBR, 2011c: Annex D, pp. 39–40). Productivity is hard to measure in the public sector, mainly because outputs include qualitative as well as quantitative aspects (for example the dignity with which patients are treated and the mortality rate; the quality of the school environment and test results). Improvements have proved very hard to achieve (Judd, 2011). NHS productivity has fluctuated between 1996 and 2009, with a very slight net fall, mainly due to increases in the drugs bill and staff pay (Hardie et al., 2011). For education, productivity also fluctuated, mainly due to changes in the school population, but it has shown no overall improvement (Ayoubkhani et al., 2010). More recently efforts to improve productivity have redoubled. A drive to generate 4.4 per cent annual savings between 2010 and 2015 for reinvestment and to meet spending pressures has so far achieved cost-efficiencies at about half that rate, half of which are not permanent (HoC Public Accounts Committee, 2011; HoC Health Committee, 2012).

The NHS is probably a particularly difficult area in which to achieve cost-efficiencies because standards are so politically salient, high-skilled staff are needed and other costs such as drugs and medical technology may also rise. An alternative possibility analysed by OBR considers NHS productivity rising at 1 per cent below that of the whole economy. This would require a further 1.7 per cent of GDP to be directed to health care spending by 2030–31 and an extra 5 per cent by 2060–61 to maintain standards.

These points make clear that there is considerable uncertainty in the public spending predictions, but that real pressures are likely to continue beyond the current period of austerity. Adjustments to spending in such areas as pensions, bus passes, fuel allowances and military procurement, and to revenue in energy taxes, vehicle duties, inheritance and capital taxes are possible (see for example Glennerster, 2011). However, a progressive programme, which does not envisage the cuts in living standards for working age claimers and state sector workers with which the current government implicitly plans to balance the books, and which expects to achieve real improvements in health care and elsewhere, must respond to these pressures.

Restructuring costs

One further point: current policies involve a profound restructuring of public provision which requires extra upfront spending and additional longer-term commitments (Taylor-Gooby, 2012). Would it be possible to square the triangle of rising demand, diminished resources and weakened capacity to meet needs by directing cutbacks differently? The government provides estimates in relation to NHS reform of £1.7 billion (rejected by the HoC Health Committee, 2010: para 92, who suggest a figure closer to £3 billion, see also Walshe, 2010). The Universal Credit reforms are estimated at £3 billion by OBR (2011b: 65) and the student loan system at 30 per cent of loans written off, or about £3 billion at current prices (OBR, 2011b: 70). In the first two cases, the plans simply assume that restructuring will deliver NHS efficiency savings faster than ever before and that benefit reforms will increase employment. For student loans, spending will not be recouped within the 30-year time horizon examined, and possibly never. However, these sums amount to less than 0.3 per cent of public spending annually over the period. They are unhelpful, but abandoning the plans would not resolve the problem.

Inequality

Society has grown more unequal during the last three decades (NEP, 2010: S3; Atkinson et al., 2011; Joyce, 2012). The distribution of incomes and wealth seems likely to fan out further. The modest post-war trend towards greater equality of opportunity may well be in reverse (Blanden and Machin, 2007). These points matter, because greater inequality appears to reduce willingness to support social provision, because of the impact on interests and on social values: those who hold the majority of resources are better able to meet their own needs privately and have less interest in provision for the poor (Horton and Gregory, 2010); inequality undermines collective solidarity and public trust in state institutions (Uslaner, 2008). Progressive social policies will have more to do, with lower public support.

A helpful analysis distinguishes between the trend to greater fanning-out of incomes for the mass of the population (as better-off groups improve their position relative to the median, while that of lower-income groups deteriorates) from the tendency of small very rich minorities to gain large relative improvements (Bailey, 2011). The first trend seems to apply to most developed countries during the past three decades. The latter is more a feature of the Anglo-Saxon world, most notably of the US and the UK (Atkinson, 2007). Public policy in the UK appears to have arrested the deterioration of the relative position of those at the bottom during the early and mid-2000s (Hills et al., 2009: 28). How easy it will be to pursue similar strategies in the future is unclear.

Overall the general tenor of debate is that there are likely to be real but not insuperable additional costs to maintaining welfare standards, even when stable growth returns. Current estimates of those costs are, if anything, on the low side. This raises the bar for any attempts to develop a more progressive policy approach than that pursued by government. Satisfactory policies must avoid the damage of the current cuts, which bear very heavily on those least able to cope. They must take into account the real resource losses resulting from a deep and continuing recession. They must set a course that will meet needs in the longer term at least as well as they are met at present, and do so in a way that accommodates the shift to greater inequality. They cannot rely on assumptions about sharp improvements in the cost-efficiency of the restructured services.

Traditional vs. aspirational Labour voters

Middle-class Labour voters are often seen as more aspirational and concerned with greater opportunities, while the core working class values basic public services and policies to address inequalities, but at the same time endorses work-ethic distinctions between deserving and undeserving groups. New Labour succeeded in attracting votes from both middle-class and working-class people from 1997. In 2010 support declined among most social groups. One interpretation of election polls supports the view that the loss of support was most marked among semi and unskilled working class voters (Miliband, 2010: 56–57, see Diamond and Radice, 2010 for a detailed statement of this view). Prominent commentators (Gann, 2011: 146–147; Whitty, 2011: 25–26) refer to statistics from Ipsos-Mori which show a striking collapse of support among the groups most likely to vote Labour in 2005: an 11 per cent fall among C2s and 8 per cent among the semi and unskilled DE core working-class group. However, these statistics are generated by amalgamating all Ipsos-Mori polls taken during the six-week campaign period to produce a combined sample of 10,000 (Table 1). They may reflect events at different stages in the campaign, the changing contexts in which questions were asked and the problems of mixing and reweighting poll data. It is noteworthy that a second Ipsos-Mori poll taken after the election and given in alternate columns of Table 1 in italics, shows a different picture: the collapse in Labour support was about half that estimated in the widely quoted combined polls among the DE group and more evenly spread across middle-class AB and C1 groups. Labour voters in the C2 group are often the focus of concerns about how well Labour appeals to ‘aspiring’ people on lower to middle incomes. Support among this group fell by only 1 per cent as against the 11 per cent shown in the earlier combined surveys.

OPEN IN VIEWER

Table 1.

Ipsos-Mori combined pre-election and post-election poll statistics.

	Conservative		Labour		Liberal Democrat
	Pre-election combined	Post-election	Pre-election combined	Post-election	Pre-election combined	Post-election
AB	39 (+2)	40 (+3)	26 (-2)	23 (-5)	29 (0)	31 (+2)
C1	39 (+2)	40 (+3)	28 (-4)	28 (-4)	24 (+1)	28 (+5)
C2	37 (+4)	25 (-8)	29 (-11)	39 (-1)	22 (+3)	28 (+9)
DE	31 (+6)	29 (+4)	40 (-8)	42 (-6)	17 (-1)	22 (+4)

Notes:

Percentage point change since 2005 in parentheses.

Pre-election combined amalgamates all Ipsos-Mori polls between 19 March 2010 and 5 May 2010 (10,211 adults) [http://www.ipsos-mori.com/researchpublications/researcharchive/poll.aspx?oItemId=2613].

Post-election: 12–13 May 2010 national survey (1023 adults) [http://www.ipsos-mori.com/Assets/Docs/Polls/NoTW%20Post%20Election_website.pdf].

The poll figures are not conclusive, but indicate that concerns about alienating either middle- or working-class Labour supporters may be misplaced. This is confirmed in the available analyses of the academic 2010 British Election Survey: those in the ‘working class’ (supervisors and manual occupations) ‘were slightly more likely to vote Labour … the decline [between 2005 and 2010] was no larger among manual than among routine non-manual and professional workers’ (Johnston and Pattie, 2011: 287–288).

Compatibility with public discourse

Quantitative and qualitative attitude studies and analyses of media discourse point to five general patterns in public opinion which are sustained over time:

In addition to these five directions identified in attitude surveys, a sixth current in ideas is also relevant. This is to do with a growing sense of disillusionment with or distrust of the state and the public sector (Taylor-Gooby, 2009: chapters 6, 7). It emerges in the growing political disengagement termed ‘anti-politics’ (Stoker, 2006; Hay, 2007).

These points indicate difficulties in promoting an inclusive and generous programme. They also imply real opportunities. The fourth point makes it hard to address inequalities as opposed to basic need, and the first is particularly powerful in limiting the scope of progressive policy. The second and third constrain the capacity to raise revenues and to redistribute, leading to a policy bias towards cutbacks that is strengthened by concerns about the deficit. The fifth creates further difficulties in marshalling support for progressive change.

The opportunities lie in the possibility that notions of desert may be expanded to cover a broad range of needs (point one), that the possible linkage of entitlement to contribution legitimates contribution-based policies (point two) and that growing inequalities may strengthen the appeal of redistribution (point three). The task of developing policies to cope with continuing pressures will be easier if basic trust in government returns (point six).

The uncertainty over election polls adds to these concerns. The more commentators emphasize a distinctive collapse of core working-class support, the more they are inclined to place weight on the issues of desert (point one) and of tax resistance (point three).

Leading in a progressive direction

As well as gaining public acceptance, progressive policies must help lead the way people understand welfare towards more inclusive and generous provision. Inclusiveness relates to the identification of net recipients and net givers and to reciprocity between them. A strong tradition in social psychology (Tajfel, 1978) and sociology (for example, Parkin’s ‘exclusionary theory’, 1979) stresses the power of in-group and out-group cleavages. To overcome these barriers, services should endeavour to include as many as possible and should seek to ensure greater recognition of the needs and capacities of others. Complementary work in social anthropology (Mauss, [1922] 1990), political science (Ostrom and Walker, 1997), decision theory (Gintis et al., 2005: chapter 1) and social policy (Titmuss, 1970) demonstrates the importance of reciprocity in social relationships. This may strengthen links between those able to help each other, but confirm divisions between those engaged and those defined as non-contributors.

Generosity concerns recognition of need and willingness to address it. Analysis of opinion surveys and media discourse indicates a shift in the established pattern of greater generosity in hard times occurring about 2007–8. Figure 2 shows that for most of the 1990s and early 2000s agreement with the statement that ‘benefits for unemployed people are too low and cause hardship’ from British Social Attitudes surveys, and media references to ‘scroungers’ tended to mirror unemployment rates. Higher unemployment was associated with greater concern about benefit levels and fewer references to scroungers. However, the tendency to greater punitiveness across the period of declining unemployment appears to continue as unemployment starts to rise after 2007. Whether this is simply an attitudinal lag or indicates a secular shift in attitudes is currently unclear. If it is the former, public opinion may move slowly to accept greater redistribution, as need increases. If the latter, the current campaign to label claimers as undeserving will have produced a real shift in attitudes, creating further problems for progressives.

OPEN IN VIEWER

Figure 2.

Unemployment as per cent of labour force (World Economic Outlook harmonized rates), media discourse (Nexis) and benefit generosity (British Social Attitudes).

Macro-level

At the macro-level the main themes are to do with slowing or deferring cutbacks to maintain the level of demand, expanding infrastructure spending, especially on environmentally friendly projects and tax reform. All the programmes that include a macro-economic element mention these themes with different degrees of emphasis (see Dolphin, 2011). The underlying concern is to create the conditions for real growth. The Coalition argues that private enterprise will generate growth provided intervention is limited. Progressives follow Keynes in pointing to a major role for the state during the downswing of the economic cycle, although they differ on the extent of intervention during an upswing.

Compass’s Plan B (2011: 5, 19–24) probably puts the strongest emphasis on the importance of maintaining demand, followed by the Institute for Public Policy Research (for example Cooke et al., 2011), but the theme also emerges in The Red Book (Clarke and Gardner, 2011), The Purple Book (Philpot, 2011) and What Next for Labour? (Scholes-Fogg and Hamid, 2011). All these programmes include a substantial state investment bank with a bias towards environmentally friendly projects in such areas as energy generation, public transport and home insulation, and towards projects that boost employment. Plan B again develops the case most strongly with estimates of the return on such investments and details of the environmental case. The brief pamphlet In the Black Labour (Cooke et al., 2011) expresses the greatest concerns about how commitments on public spending will damage electoral perceptions of fiscal responsibility, but supports infrastructural spending. The Social Market Foundation, pursues a similar approach, making a case for further specific cuts, mainly on tax reliefs for better-off people, to finance an extra £15 billion infrastructural spending (Mulheirn, 2012: 7).

Plan B also argues for international financial regulation and management of banking and a tax on financial transactions. It provides a detailed account of income and wealth inequalities and argues for more progressive taxation and plans for a gradual redistribution of wealth (Compass, 2011: 22–23). Many of the programmes also stress the importance of making the tax system more progressive but this is not generally worked out in detail. The Glasgow Media Group estimates that a 20 per cent tax on the wealth of the top 10 per cent would raise some £800 billion, enough to pay off the national debt (Philo, 2010). It points to 80 per cent support for such a measure in a YouGov poll, reflecting public endorsement of redistribution from the distant rich. The plan is briefly stated and does not consider the obstacles to levying such a tax, retrospective changes to legislation in relation to pensions, trusts and other areas and problems of avoidance.

Micro-level

At the micro-level the range of initiatives is greater. Five major themes may be identified in the debate: social spending as social investment; building social solidarity; responding to the deserving/undeserving distinction; the importance of engaging users directly in service provision; and the logic of pre-distribution as against redistribution. Some policies build on existing programmes such as Sure Start or Tax Credits.

Social investment

The clearest and most developed argument for social investment is contained in IPPR work on returns from child care (Ben-Galim, 2011). This shows a possible return of £20,000 over four years for each woman returning to work after one year’s maternity leave, reducing to £5,000 when income figures are based on the pay rates of women returners rather than workforce averages. In either case the return is impressive. Plan B refers generally to social investment, giving the uncosted examples of obesity reduction and other behavioural changes to reduce pressures on the NHS (Compass, 2011: 32). The returns on investment, for example in improvements to the quality of social housing or in public health, are often diffuse, long term and dependent on factors outside the state sector. Previous research shows that the most impressive returns on social sphere investment are found in education (Dickson and Smith, 2011), but that it is more difficult to establish a clear picture in other areas.

Building solidarity

A substantial stream of work on the factors that have contributed to successful redistributive and integrative welfare in the UK, drawing on historical analysis, cross-national comparison and attitude surveys, concludes that welfare policies must be inclusive if they are to succeed in meeting the needs of the poor (Horton and Gregory, 2010: xix). Highly targeted provision typically fails to enlist the support of the mass of the population and leads to a downward spiral of increasing restriction and exclusion. These authors seek to reconcile commitment to universalism with the realities of deserving/undeserving distinctions through a welfare system that incorporates as many groups as practicably possible. To achieve the first objective they propose an expanded tax-credit system that is effectively a citizen’s income for all low income people, plus a combined housing benefit that incorporates subsidies to all tenure groups and reduces the divisions between them (Horton and Gregory, 2010: 216). To address the second they suggest basing benefit entitlement for working age claimers on social participation (p. xiii); and real sanctions for claimers who refuse to engage in activities that make a social contribution (pp. xxiv–xxv).

The objective is to ensure that the welfare state includes citizens and does so by reinforcing relationships of reciprocity rather than moralistic divisions or individual interest. One problem lies in ensuring that all groups can engage in activities that will be realistically understood by most people as making a social contribution. The research carried out by Horton and colleagues indicates that full-time caring for children or frail elderly or sick and disabled relatives, full-time studying for useful qualifications and some kinds of full-time voluntary activity are recognized in this way. These activities would form the basis of an inclusive welfare state.

Responding to social divisions

An alternative response to public opinion is to seek to devise policies that work with existing divisions. This is attractive to those who believe that the Labour Party became detached from its traditional working-class roots during the early 2000s (for example, Byrne, 2011: 140–141; Glasman et al., 2011). The resulting logic of ‘something-for-something’ welfare (Miliband, 2011) is designed to link the notions of reciprocity and inclusiveness to the theme of desert. Social contribution justifies the receipt of benefit. Many proposals centre on the idea of social insurance, a theme which has had some success in protecting welfare systems in corporatist European countries (Field, 2001).

This approach is attractive in many ways. One problem is illustrated by the National Salary Insurance scheme of Purnell (Cooke, 2011), which is often seen as a key example. This scheme allows those whose employment is interrupted by illness or redundancy to claim a relatively high benefit which will then be recouped through a supplement to tax when they return to work. This is essentially a loan system underwritten by the state rather than a benefit or insurance. As Baumberg (2012) points out, the approach is of limited application. It cannot address the needs of those who have a weak employment record, who face extended periods of unemployment or who are low paid and unable to finance their own benefits. In short, it entrenches a deserving/undeserving distinction among claimants, the issue that recognition of a wider range of socially desirable activities seeks to address in the Horton and Gregory model.

User engagement

User engagement emerges in a range of ideas from co-production, where services users are actively involved in providing, managing and sometimes financing the services they use, through to the expansion of mutuals and co-operatives. Plan B draws heavily on the work of the New Economics Foundation (for example, 2008) in its discussion of co-production. Current work in this area has applied the idea with some success in health and social care (Glasby et al., 2010). How it would be extended to other areas, especially those involving substantial professional inputs or cash transfers, is not worked out in detail. Co-operatives are discussed in a number of the essays in The Purple Book (Richards, 2011: 50–51). Provider co-operatives would operate to stabilize firms and to moderate pay inequalities, a theme that has recently been taken up by mainstream politicians (Clegg, 2012). Co-operatives that included users in the areas of social care for example, might lead to more responsive services and a stronger trust in social provision. If co-operatives succeed in engaging the public, they may help to rebuild trust in the state that will facilitate further progressive reform.

The co-operative approach offers a management model for such areas, but can only address issues of inclusion if specifically designed to facilitate entry by outsiders. Mutuals which provided services or support for members might be subject to all the in-group biases identified by social psychologists, anthropologists and sociologists and be unwilling to extend the service to new members who did not have demonstrable resources to contribute. This requires careful attention to how co-operatives recruit members and interact with outsiders.

Pre-distribution

The grand tradition of state welfare has rested on redistribution through taxation of the better-off or of individuals at life-cycle stages when income exceeds resources, to provide welfare when they are in need. Pre-distribution addresses inequalities at source, through state interventions into the operation of market systems to reduce income inequalities and shift power towards the lower paid. The most powerful arguments for this approach rest on the claim that state welfare was most successful in the UK when the institutions to reinforce pre-distribution were at their strongest, in the 1950s and 1960s (Coats et al., 2012: chapters 2, 3; Pearce, 2012). Hacker (2011) points out that the massive increase in inequality in some of the most developed countries is associated with erosion of the protective institutional framework.

Pre-distribution might include institutional changes, such as a strengthening of the bargaining position and influence of workers through stronger trade union rights and representation on works councils, legislative interventions such as enforcement of higher minimum wages or a living wage, better working conditions or shorter working hours, measures to curb wages at the top end through reforms to remuneration systems and possibly maximum wage legislation and interventions to control the prices of items of common consumption such as utilities, transport or food.

This approach is attractive for three reasons: it addresses the issue of redistributing power towards groups that have grown relatively weaker in the market; it increases the influence of institutions that might reinforce solidarity and promote greater reciprocity across groups that may then perceive a stronger common interest; and implementation costs are low. The problems are to do with the impact on the structure of social relationships and on market interactions. Some of the proposals are directed to strengthen the bargaining power of organized labour, while some are directed to those at the margin of or outside work. At various times these groups have failed to recognize common interests and much of the tradition of a deserving/undeserving distinction derives from this division. One contribution is to direct interventions to investments that increase the supply of jobs while strengthening the protection of wages and standards in those jobs, the reverse of the current spending cuts plus Work Programme strategy. At minimum, pre-distribution needs to engage with redistribution to protect those outside the labour market.

Pre-distribution confronts the operation of the capitalist market system directly and may affect international competitiveness. This leads to concerns about how well an approach which developed during the heyday of nation-state capitalism is fitted to advance collective interests in a more globalized world (Glynn, 2006). This takes the argument in the direction of the greater cross-national regulation of inequalities stressed by Plan B in order to ensure that competitive pressures do not damage a high-wage economy.

This brief overview of progressive strategies indicates areas of consensus and of conflict. Consensus is found at the macro-level in relation to a stronger role for the state in moderating the economic cycle and in economic investment to secure growth. Damage to Labour’s reputation for economic prudence may constrain willingness to spend. At the micro-level, disagreements are more marked and focus on issues of balance between different policies. This is particularly evident in two areas: first between policies which direct their efforts to working with the deserving/undeserving distinction (expanded notions of contribution, social insurance-based logics) as against those which seek to side-step it (pre-distribution, social investment); and secondly between those which stress the engagement of ordinary people (co-production, mutuals and co-operatives) as against those which rest on top-down initiatives (child-care spending, progressive tax, better social housing). Social investment and pre-distribution are strongly supported.

Macro-level

There is a considerable measure of agreement at the macro-level on the themes of infrastructural spending to encourage growth and of more progressive taxation. The Coalition introduced a modest £6.3 billion infrastructure package in the 2011 Autumn Statement, with the expectation of some £5 billion in the longer term and a possible £20 billion leveraged from private sector pension funds of which only £2 billion has been achieved. The IFS (2012) calls for a further £10 billion and NIESR (2012) suggests £15 billion. Compass (2011: 26) proposes a much larger programme, with a £100 billion fund for the investment bank. Tax reforms are also not described in detail but include new taxes directed at the reduction of income and wealth inequalities. Similar ideas are included in most programmes.

It is unclear how far these policies would be more successful than current government initiatives in addressing the immediate problems of sluggish growth, mainly because developments in the UK are intertwined with those elsewhere. Apart from the arguments of In the Black Labour (Cooke et al., 2011), there is little work examining in detail how spending plans would relate to budgetary constraints. If we assume that these constraints are, at least in the short term, severe, it is necessary to identify areas for cut back.

Non-welfare-state areas, especially defence, face sharp cuts, which could continue. A generous and inclusive policy would seek to spread the savings burden more evenly. IFS analysis (Browne, 2010; Crawford, 2010; O’Dea, 2010) shows that the cuts and tax changes bear most on those at the bottom and on a small group at the top, facing higher tax rates, and on families and children rather than benefits for older people. Further taxes on the distant wealthy (a capital value or land tax, perhaps or higher inheritance taxes) would fit with public discourse. One problem is how far it is possible to bring higher taxes down the income distribution, given public resistance to taxing the middle class. A further issue concerns the desire to mitigate benefit cuts. Some resources could be saved in the short term by terminating the restructuring of benefits and public services. It is difficult to see how far further savings could be made without addressing state pensions. This would affect a large group in the electorate typically seen as deserving.

These considerations imply that new policies at the micro-level are essential if the macro-level problems are to be addressed. These policies must at least constrain spending. They must also fit with public opinion and build the public trust necessary to support future progressive reforms.

Micro-level

Social investment is attractive because it offers a return, so that net spending commitments are contained; areas like child care fit the public opinion constraint because they focus on an obviously deserving need group (young children) and are directed at supporting paid work. They also help support an inclusive solidarism since they can be universal across the group. For these reasons they have been vigorously promoted by progressives in Europe (Vanderbroucke et al., 2011). However, the case for investment is only persuasive in a limited range of areas, mainly around education and public health.

Solidarity policies based on principles such as contribution include a wider range of social groups, provided the public will accept a relatively broad range of activities and not simply a social insurance payment or paid work as a relevant contribution. Spending commitments may generate problems, since inclusion is likely to involve extra spending on higher benefits for those who are unable to make equivalent cash contributions. The deservingness barrier creates difficulties in including people, typically from the most deprived groups, who are unable to make any contribution. Horton and Gregory (2010) address this by postulating activities that might enhance future contributions such as attendance on training and work-readiness programmes for unemployed people. Benefit entitlement for those unwilling to accept these requirements would be reduced. The policies offer a progressive contribution because they build reciprocity and help channel debate towards the idea that citizens contribute and derive benefit from participating in society.

Policies that operate within the constraints of the deserving/undeserving distinction, such as National Salary Insurance, are likely to be financially viable, since their range is relatively small. They are electorally feasible and attractive to public opinion, but do not build inclusiveness.

User engagement programmes can use a wide range of techniques, involve little extra spending, are often attractive to a public which feels ignored by remote decision-makers, and may help build trust in public policies to facilitate difficult spending decisions in the future. There are real limits to how far a sense of inclusion and participation can be promoted, for example across a national education system or health service. Identification with, for example, a particular school in a more participative or co-operative system may strengthen social divisions. These considerations may limit participation to the more restricted forms such as personal budgeting.

Pre-distribution is also highly attractive in terms of low direct public spending. It does not obviously involve extra spending and may well have longer-term benefits in reducing inequalities. The approach may encounter electoral and public opinion obstacles if it is seen to strengthen sectional interests. Again it may fail to strengthen inclusivity and solidarity between those in secure work and those in less stable jobs.

None of the proposed policies offers a complete solution to the problems outlined earlier. Any progressive future is likely to rely on a combination, so that issues arise in reconciling conflicts between groups addressed by different approaches. The budgetary pressures require policies that constrain current spending. By the criterion of economic sustainability, social investment, deserving-centred, user engagement and pre-distribution policies seem attractive, because cheaper. Conversely, the social contribution approach seeks to offer a way of presenting substantial generous and inclusive welfare so that it is electorally feasible. The Left trilemma reflects a concern to demonstrate economic competence, a desire to gain public support and a commitment to progressive policies that is particularly intransigent when resources are limited. Policies which lead to savings are hard to reconcile with generosity and inclusiveness, which are hard to promote when public discourse is suspicious of the undeserving and lukewarm about taxing the better-off.

This suggests that the problem is not so much one of designing the right programme, as of shifting public discourse through measures which build solidarity and inclusiveness. The five themes identified earlier have remained powerful over time but do fluctuate. In general they have moved against generosity and inclusiveness. This may indicate that the Right is more successful in leading public debate. The theme of social solidarity through a broad-based contribution system, reinforced with pre-distribution and social investment policies, may contribute to an attractive electoral platform. If contributory welfare involves increased spending this may limit its practicality, so that it is only practicable through gradual introduction over time. Provided that the systems are designed to foster inclusiveness and to minimize barriers between groups, the potential for leading public debate in a progressive direction is strong.

These policies must be embedded within an overall approach that fosters a return to growth through investment in infrastructure, production and social provision, and seeks higher taxation of the better-off. The resources available will be limited in the medium term, so that some cuts will be necessary to enable support for the most vulnerable. It is hard to see how generosity to pensioners can escape such constraints, because of the size of the programme.

This article has set out the difficulties any progressive social programme faces as a trilemma between economic viability, public acceptability, and inclusiveness and generosity, which does not confront the Right. It shows that no suggested programme is wholly satisfactory on all counts and that there is a conflict at least in the short term between policies that seem most likely to satisfy harsh spending pressures and an ungenerous public discourse, and those with the strongest potential to generate adequate and solidaristic welfare. This points to a broad and conjoint policy platform, and one which pays attention to ensuring that policies do not reinforce divisions and put obstacles in the path of any move towards greater inclusiveness. The trilemma can only be resolved through policies which build solidarity and inclusiveness as a continuing and cumulative process.