Accepted on 27 Oct 2020
Submitted on 26 Jun 2020
Over the past century, social security has become, by far, the largest government programme in most advanced economies of the world. How has this come about? Where are we now? And what are the possibilities for the future? These are the questions addressed in this paper.
Social security (or social protection) is not one cohesive system that has been consistent over time or across nations. For example, although the taxation system is not traditionally thought to be in the realm of social security, tax credits often have similar objectives and are an increasingly important form of redistributing income. In some countries, social security refers only to contributory social insurance schemes and does not include means-tested ‘welfare’ schemes. This paper adopts a broad definition of ‘social security,’ embracing all cash benefits based on social objectives. Most of the examples and data are drawn from the British experience, but some international comparisons are made, and most of the issues discussed are common to all advanced, industrial or post-industrial economies.
In this paper, Part 1 describes the growth of social security and explains the reasons for this growth. Part 2 discusses the achievements of these systems, alongside analysing some of the limitations of the current system. Part 3 considers different approaches and possible reforms to social security. One of the challenges facing social security is ensuring its political sustainability, an issue often ignored, but one that is discussed in the final section.
In considering these issues, it might be simpler to discuss only the underlying philosophy of social security, or only the political issues that arise, or only the economics of this major component of modern societies. However, this would ignore the entanglement of philosophy, politics and economics that explains the development and functioning of this key area of social policy.
Every society’s social security system is uniquely shaped by their history, attitudes and expectations of the future, of which their political, economic and philosophical experiences play a significant part. While ‘security’ connotes ideas of stability and continuity, the radical changes that have taken place globally over the past century have equally affected the development and operation of social security. As these changes to social security take place, they, in turn, go on to cause further shifts in broader society, and the cycle continues. There is no reason to suppose the next century will be any different. We have certainly not reached the end of history as far as social security is concerned.
At the start of the 20th century, government spending was everywhere a small fraction of what it has become. Yet, even within the low rates of government spending, social security was sparse, with few nations having any substantial social security systems. Instead, the poor and destitute relied largely on families, charities, and religious institutions.
The first half of the 20th century, however, saw the beginnings of a dramatic change in the UK, with government expenditure growing across the board, from 14 to 41 percent of gross national product (GNP). Social services, including health and education, was a particular beneficiary, growing from 18 percent of government spending in 1900 to 43 percent of spending by 1951, as shown in Table 1.
Table 1
UK Government Expenditure 1900–1951.
| Total Government Expenditure | Social Services | ||
|---|---|---|---|
| As % of GNP | Index (1900 prices) | As % of Total | |
| 1900 | 14.4 | 100 | 18.0 |
| 1910 | 12.7 | 145 | 32.8 |
| 1920 | 26.2 | 599 | 25.9 |
| 1930 | 26.1 | 823 | 42.3 |
| 1938 | 30.0 | 815 | 37.6 |
| 1951 | `40.7 | 1354 | 42.9 |
Source: Tables A-6, A-11, A-15, The Growth of Public Expenditure in the United Kingdom, Alan T Peacock and Jack Wiseman, London, George Allen and Unwin, 1967 (Revised Second Edition).
Since the Second World War, this growth has continued, with social security spending having increased more than ten-fold in real terms and nearly tripled as a proportion of gross domestic product (GDP), as shown in Table 2. This upward trajectory has largely continued, and while there has been a slight fall since the recession of 2009, this type of spending is still above where it was 20 years ago. Indeed, throughout the period shown in Table 2, social security spending has constituted a greater percentage of government spending than health or education combined. This is despite, as Table 3 shows, the dramatic increases in health spending that have taken place over the past two decades.
Table 2
Growth in Social Security Expenditure, UK, 1949–50 to 2018–19.
| Index of Expenditure | Expenditure | |
|---|---|---|
| in Real terms | as % of GDP | |
| Benefit Expenditure (a) | ||
| 1949–50 | 100 | 4.7 |
| 1959–60 | 150 | 5.3 |
| 1969–70 | 264 | 7.0 |
| 1979–80 | 436 | 9.0 |
| 1989–90 | 580 | 9.5 |
| Social Protection (b) | ||
| 1989–90 | 100 | 11.4 |
| 1999–2000 | 143 | 11.8 |
| 2009–10 | 220 | 14.4 |
| 2014–15 | 238 | 14.1 |
| 2018–19 | 233 | 12.9 |
Sources: (a) Table 1, The growth of social security, Dept of Social Security. HMSO, 1993.
(b) Table 4.4. Public Expenditure Statistical Analyses, 2010 (Cm 7890) and 2019 (CP143).
Table 3
Growth of social and total public spending as % of GDP, UK.
| Social Protection | Health | Education | Total Managed Public Expend | |
|---|---|---|---|---|
| 1989–90 | 11.4 | 4.5 | 4.8 | 39.2 |
| 1999–2000 | 11.8 | 4.7 | 4.0 | 34.1 |
| 2009–10 | 14.4 | 7.6 | 5.7 | 44.9 |
| 2014–15 | 14.1 | 7.2 | 4.6 | 40.5 |
| 2018–19 | 12.9 | 7.2 | 4.2 | 37.9 |
Source: Table 4.4. Public Expenditure Statistical Analyses, 2010 (Cm 7890) and 2019 (CP143).
This growth in social security spending has not been confined to the United Kingdom. If we survey a wider range of nations, cash benefits in 2015 were a higher proportion of gross domestic product in all six of the seven major OECD nations than in 1980, as shown in Table 4. The range of spending in 2015 varied from nine percent in the USA to twenty percent in Italy (although differences of definition, charges and exemptions for essential services and the taxing of benefits make direct comparison difficult). Thus, while there is a wide (and growing) range in social security spending across nations, in all advanced economies it has become a major sector of the economy.
Table 4
Cash Benefits as % of GDP, G7 Countries, 1980–2015.
| 1980 | 1990 | 2000 | 2010 | 2015 | |
|---|---|---|---|---|---|
| Canada | 7.0 | 9.4 | 8.6 | 8.9 | 9.3 |
| France | 13.4 | 15.2 | 16.1 | 18.6 | 19.3 |
| Germany | 14.8 | 13.3 | 15.0 | 14.7 | 13.5 |
| Italy | 11.8 | 14.5 | 15.8 | 18.6 | 20.2 |
| Japan | 5.1 | 5.8 | 8.5 | 11.5 | 11.2 |
| UK | 9.5 | 8.3 | 8.7 | 11.0 | 10.3 |
| USA | 8.2 | 7.5 | 7.3 | 9.6 | 9.0 |
Source: Social Expenditure Database, Paris: OECD Downloaded 30.12.19.
Within social security, spending on pensions is overwhelmingly the largest component in all seven nations, as shown in Table 5. Pensions were more than twice the sum of all incapacity, unemployment and family benefits in every nation included here except the UK- but there it still dwarfed the spending in every other category.
Table 5
Public Spending by Type as % of GDP, G7 Countries, 2017.
| Pensions | Incapacity | Unemployment | Family | |
|---|---|---|---|---|
| Canada | 4.7 | 0.8 | 0.6 | 1.6 |
| France | 13.9 | 1.7 | 1.6 | 2.9 |
| Germany | 10.1 | 2.1 | 0.9 | 2.2 |
| Italy | 16.2 | 1.8 | 1.0 | 2.0 |
| Japan | 9.4 | 1.0 | 0.2 | 1.3 |
| UK | 6.2 | 1.9 | 0.2 | 3.5 |
| USA | 7.1 | 1.3 | 0.2 | 0.6 |
Source: OECD Social Expenditure Database, Downloaded 30.12.19.
To summarise, all the advanced economies have experienced a huge growth in social spending since 1900. This growth was not a one-off, quantum leap from minimal provision for the poor to the levels of provision established by the New Deal in the USA or the post-war welfare states of European countries. Rather, there has been fairly steady growth, not only in real terms, but also as a share of gross domestic product, and this growth has been largely sustained, even during periods of austerity. Notably, however, this spending has not been distributed throughout all age groups in society, but has been focused primarily on the elderly.
The reasons for this growth of social security are, at one level, simple, but at another, complex and multi-faceted. At the simple level, changes in societies and economies altered social needs, while changes in political and economic structures affected the ability and willingness of governments to meet such needs. Alongside this, changes in social and political attitudes changed people’s expectations and the demands placed upon governments. However, to understand these changes, they must be examined at a deeper level, taking into account historical changes and past political conflicts and outcomes.
Clearly, examining in detail the development of social security in any one country, let alone in many countries, is beyond the scope of this paper, but helpfully others have considered the issue. Peter Lindert’s review of the growth of social spending since the 18th century [1] concluded that:
‘Five leading forces help to explain the rise of social transfers…
On social affinities, Lindert explains: ‘The more the middle-income voter looks at the likely recipients of public aid and says “that could be me” (or my daughter or my whole family), the greater that voter’s willingness to vote for taxes to fund such aid.’ [1: 187]
In short, the growth of social security has been the product of changing political power and beliefs in the face of changing economic and social circumstances.
If there is a single objective of social security by which its success can be judged, it is the ending of poverty. Beveridge’s plan [2], for example, was primarily directed at delivering freedom from want, and the justification for developing social security in all countries is always linked to the relief of poverty. This goal was expressed internationally in Article 25 of the UN Declaration of Human Rights in 1948, which stated that:
‘Everyone has a right to a standard of living adequate for the health and well-being of himself and his family.’
Politically, objectives of social security have included nation-building, the avoidance of revolutionary tensions, and achieving a social security system perceived to be just. In terms of redistribution, objectives have not only been to tackle poverty but to smooth income over the life cycle, to promote equity between households in different circumstances and to reduce inequality. Other concerns have been the impact of the system on peoples are social and economic behaviour and with the way those claiming benefits have been treated. At an organisational level, objectives concern operational effectiveness and efficiency in directing resources to those in need.
Here, because of limitations of space, the emphasis is primarily on social security and its role in addressing poverty.
In order to answer the question of how social security can relieve poverty, a definition of poverty is required. It is widely accepted that poverty must be defined in relative terms so that the poverty standard should rise as living standards have generally risen. A relative standard does not, however, imply, as some erroneously imagine, that poverty can never be abolished; it merely requires that no one should fall below a specified relative level.
Relative poverty rates in the UK – the proportions of people living in households below 60 percent of median household disposable income, adjusted for family size, after housing costs – are shown over nearly 60 years in Table 6, which distinguishes between the circumstances of pensioners, those of children, and those of the whole population.
Table 6
Poverty Rates, UK, 1961–2018/9.
| Pensioners | Children | All | |
|---|---|---|---|
| 1961 | 37.8% | 13.4% | 13.5% |
| 1981 | 19.8% | 19.5% | 14.4% |
| 2001/2 | 25.6% | 30.7% | 22.7% |
| 2011/2 | 13.4% | 27.2% | 21.0% |
| 2018/9 | 15.8% | 29.8% | 22.0% |
Notes: Poverty is where household income is below 60% of median equivalised income after housing costs. Data up to 2001–2 relate to GB only.
Source: Living Standards, Inequality and Poverty Spreadsheet, London: Institute for Fiscal Studies Downloaded 27.6.20.
Has social security, then, abolished poverty? Clearly not. Rather, the proportion living in poverty has increased and is now over 20 percent, compared to around 14 percent in 1961 and 1981. The experience of different groups varies, with a fall in the proportion of pensioners in poverty but a rise in the proportion of children.
A rather different question, and one that is perhaps more relevant, is whether social security has reduced poverty. Evidence on this is shown for six countries in Table 7, which provides poverty rates with and without social transfers in six European countries. Poverty after social transfers is lower than before social transfers in all six countries. However, while the social security system reduces poverty, it comes nowhere near eliminating it, with all the countries covered having between 13 and 20 percent of their populations still in poverty after social transfers.
Table 7
Poverty Before and After Social Transfers, 2018.
| Before Transfers | After Transfers | |
|---|---|---|
| France | 24% | 13% |
| Germany | 24% | 16% |
| Ireland | 31% | 15% |
| Italy | 26% | 20% |
| Netherlands | 22% | 13% |
| UK | 29% | 19% |
Source: At-risk-of-poverty rate before/after social transfers, EU-SILC Survey, Eurostat, Downloaded 27.6.20.
Counterintuitively, then, the growth of social security has also been accompanied by an increase in relative poverty. Why has this occurred?
First, because social security benefit rates have not reached and do not reach the poverty level. For example, in the UK, universal credit payments to families with children only provide for, at most, two-thirds of their poverty level, as defined by the government. In contrast, for the elderly (whether living alone or in a couple), the levels of minimum pension are higher relative to their poverty level – in part explaining the lower poverty rates of pensioners.
Second, social security benefits in the UK have, since 2013, been deliberately capped. The purposes of the benefit cap are explicitly to limit public expenditure, to increase the incentive on claimants to find employment and to discourage larger families, which are most likely to be poor. This cap marked the abandonment of any comprehensive attempt to meet any level of minimum requirements.
Third, there are social security benefits that are unclaimed and unpaid, so that not all are receiving the benefits to which they are entitled. This may occur because the costs of claiming are deemed too high because of stigma or humiliation in the claim process, or because the benefits available are unknown or incomprehensible to potential claimants. Alternatively, it may be due to administrative failure to assess eligibility correctly or to properly deliver benefit entitlements.
Limitations of the social security system are not, however, the main cause of growing poverty. Instead, there are a plethora of interconnected factors that have served to diminish household net income. Net income depends on many factors, the most obvious being jobs and earnings. The availability of and access to jobs depends at a macro level on economic policies and at a micro level on accessibility of housing, transport and child care. Alleviating poverty also relies on these jobs being sufficiently well-paid, which depends on the levels and availability of education, training, and health. The income of the household also depends on the rates of income tax and the generosity of the social security system, alongside the household composition, specifically the number of earners and the hours they work. Finally, the levels of rent and other housing costs are relevant factors. Changes in all of these factors have contributed to the growth of poverty.
Thus, in assessing any social security system, it is important to recognise that it is only one factor determining the overall distribution of incomes and in the extent of poverty.
Having reviewed the current set of remaining needs in terms of poverty alleviation, a natural next question concerns what form of social security best meets these needs.
The organisation of social security has taken innumerable forms. It can embrace anything from reliance on family support, to private charity, to a planned economy requiring full employment. Within this paper, the focus is on four government systems or approaches. Three of these have been widely adopted and the fourth is one that has been extensively advocated in recent years to either supplement or replace existing systems. This section reviews all four approaches, highlighting their relative advantages and disadvantages.
The purpose here is not to consider the minutiae of such approaches in theory or in practice, but to consider the broad features of each, their advantages and their disadvantages.
This approach is one of poverty relief. Benefits are restricted to the poor or destitute, and so are dependent on means-testing of income, and often also of capital assets. Many schemes also require further tests or impose conditions for receipt of the benefit.
One example of this approach was the English Poor Law that lasted from the 16th century to the middle of the 20th century. Its successors – national assistance, supplementary benefit, income support, and, most recently, universal credit – retain many of its features. Underpinning this system was the belief that poverty was a sin, and so draconian measures were imposed. ‘Indoor relief’ saw elderly couples split up and separated mothers from their children in the workhouse. The provision of relief was through local parishes, meaning that parishes often sought to direct or transport claimants to neighbouring parishes, seeking to reduce their costs.
A more modern example of a social assistance approach is the British Universal Credit (UC) scheme, which was announced in 2011 [3] and was intended to be fully rolled-out by 2017. UC sought to unify a number of means-tested benefits1 in order to simplify the welfare system, to reduce the effects of overlapping means tests (which had led to very high effective marginal tax rates, often described as the poverty trap), and to improve the financial gains from taking low-paid jobs. The UC system also extended the range of sanctions and conditions that could be imposed, with the intent that it would encourage or force people to take a job. In-keeping with this employment-oriented approach, UC paid benefits monthly, replicating a salary, and sought to integrate the welfare provision of those in and out of work, in order to smooth the transition from unemployment to employment (or vice versa).
The goals of UC were widely welcomed in principle, and this latest form of social assistance was probably well-intentioned. However, in practice, it has not been a success. Timmins [4] has lucidly described the profound problems that have waylaid the system. Within the Department for Work and Pensions there was little cohesion between policy makers, technical system operators, and front office staff, exacerbated by the fact that frontline workers’ numbers were being cut just as their responsibilities were dramatically increasing.
As a result of this disorganisation, the development of a huge IT programme was initiated before key aspects of how UC was to work had been decided, which made putting UC into practice problematic. The head of the National Audit Office castigated the implementation of UC, writing that those responsible for its implementation had shown ‘weak management, ineffective control and poor governance [5].’ Three years after UC was supposed to be fully operational, it remains only partially implemented, with full roll-out put back to 2024. Most seriously, cuts in the social security budget imposed by the Treasury undermined any prospect of UC giving significant help to those in low-paid jobs and of mitigating the poverty trap.
Universal Credit, despite its name, was never genuinely universal: it was only for those in poverty, whether in or out of work. Even if it had been successfully implemented, it would not have meaningfully tackled any of the main causes of poverty – low pay, unemployment, the cost of children or high housing costs – in a comprehensive manner. Instead, it attempted to alleviate the kind of acute poverty arising from these multiple causes retrospectively, rather than reducing the risk of poverty prospectively. Furthermore, the later restrictions on benefits, imposed as a result of the Treasury’s cuts to the social security budget, meant that it left many worse off rather than better off. The Institute for Fiscal Studies [6] estimated that of the 11 million adults who would be entitled to UC – one-third of all working age adults – 1.6 million would gain more than £1000 per year but 1.9 million would lose more than this; and those ‘persistently poor’ would lose an average 1 percent of their income- more than any other income group. All of this is exacerbated by the fact that the scheme originally envisaged a five-week delay before benefits would be paid. Even after this flaw was pointed out, the government did not seek to expedite the payment of benefits, but instead introduced a system of repayable advances. Thus, the new system has caused many applicants to fall into unmanageable and poverty-enhancing debt.
There are circumstances in many parts of the world in which social assistance is the only feasible approach to tackling poverty, as discussed, for example, by Barrientos [7]. But for the reasons set out in 1909 by Beatrice Webb in the Minority Report on the Poor Law [8], and in the Beveridge Report [2], and as evidenced by the current Universal Credit debacle, a social assistance approach is limited, ineffective, and, in many respects, harmful.
Social insurance was the major social security development in the 20th century, introducing a compulsory, collective, employment-related system that mitigated risk for all of society. It provides principally for retirement pensions, but also for periods of sickness, industrial injury and, in some cases, unemployment. In general, the idea is to ensure that everyone is provided for in periods without earnings, realising that the causes, onset and duration of such periods can be unpredictable and devastating. This system, generally financed by a combination of contributions from employees, employers, as well as through general taxation, was, in Britain, proposed by and implemented as a result of the Beveridge Report.
The main limitation of this approach is that it is based on a model which assumes a 1950’s style economy, where industry-employed men are the primary breadwinners in the household. For those in self-employment, those with employment interrupted by child rearing, those who have disabilities, and those who are never employed, they are generally excluded or disadvantaged, as are new entrants into the labour market, such as young people or immigrants. Although various credits have been used to extend coverage, contribution conditions still exclude many from receiving benefits.
Despite these limitations, social insurance has its advantages. In many countries, it has provided a politically attractive and sustainable basis for redistributing income over the life cycle for a large proportion of the population. The issue is whether it will hold the same appeal, or be as politically or economically sustainable, in the future?
Occupational pensions, dependent on a model of secure, lifetime employment are already being rolled back, and in the future the nature and distribution of paid work may be transformed. As Yuval Noah Harari wrote: ‘In the twenty-first century we might witness the creation of a massive, new unworking class: people devoid of any economic, political or even artistic value, who contribute nothing to the prosperity, power and glory of society. This “useless class” will be not merely unemployed – it will be unemployable [9: 379].’ Such apocalyptic forecasts have been made before, without being borne out. Yet the enormous uncertainty now surrounding the future labour market means that many are concerned that this could be the time that such a prediction comes to pass. This is alarming for those governments who would have to find a way to fund future pension payments without a substantial working population to support them. With this prospect on the horizon it is no wonder that employment-based contributory social insurance is genuinely under threat for the first time.
To remedy the gaps in social insurance coverage, social security systems in many countries have increasingly moved towards universal, non-means-tested, benefits for particular groups, such as for children, the elderly and the disabled. The case for such a policy was most cogently made by George Akerlof in 1978 [10], where he wrote that ‘some types of programs, either by their eligibility requirements or by the self-selection of the beneficiaries, identify (tag) people who are in special need.’ [10: 17], and provide benefits only for the ‘tagged.’
In this system, ‘tagged’ groups are at greater risk of experiencing poverty, having lower or zero incomes and higher costs than the ‘untagged’, so redistributing income to them reduces overall poverty and increases equity across households. Because there are no contribution conditions nor means tests involved, the system of benefits is simpler and the take-up of benefits is higher. In addition, recipients are less likely to fall into the poverty trap by the withdrawal of the benefit as other incomes increase. On the other hand, the cost of universal benefits inevitably requires higher tax rates.
In the case of children, child benefit was extended to all children in 1976 and was fully universal until 2013, when it was taxed back from high income recipients. However, the recent decline in its value has been concerning. Between 2010 and 2020, child benefit fell by 21 percent in real terms, meaning that it now only provides a fraction of the poverty level requirement of a child. This has meant, despite the availability of means-tested benefits, that children tend to live in poverty at much higher rates than other age groups.
Shifting to targeted provision can create an incentive to be identified as a member of a target group. Because tagging is binary – a person is either deemed in need of support or not, with no partial measures – people may go to great lengths to attract the relevant ‘tag.’ For instance, the provision of benefits to lone-parent families may encourage parents to separate in order to acquire the ‘lone-parent tag.’ Universally provided child benefit reduces these kind of incentives. Returning to the example, tagging all children regardless of their family circumstances, as current child benefit does, would clearly do less to distort family structures.
‘Targeting’ is often associated with only means-tested benefits. But, as Beveridge proposed, contingency-based benefits are also targeted. The case for such targeted, universal benefits remains very strong both as an effective means of tackling poverty and of promoting a more equitable distribution of incomes.
A universal basic income – UBI – is now enthusiastically advocated by many as a replacement for, or an addition to, the entire social security system. Here, the focus is on replacement, where UBI would aim to provide a full basic income replacing most or all of the current social security system.
Much of the appeal of the replacement model is its simplicity. If the goal is to guarantee everyone a poverty level income, then the best way of doing this is to give everyone that income and pay for it out of general taxation. This would require a UBI adequate to meet basic needs, which was paid to all individuals (and parents on behalf of children) in the form of a regular income, without any conditions or any test of means. According to its proponents, such a system could end poverty, be socially just, and be politically feasible. All these claims are open to question.
First, UBI may be simple but human needs are not. Needs depend on household composition, the health of individuals, and the level of housing costs. A simple UBI would not take such variations into account. Ironically, much of the complexity of social security systems, criticised by advocates of UBI, has developed precisely to take account of such variations and to make the resulting system fairer.
The strongest case for the unconditionality of a UBI was put forward by Phillipe van Parijs, in his 1991 paper ‘Why Surfers Should Be Fed: The Liberal Case for an Unconditional Basic Income [11],’ and developed in his latest book [12], with the argument based upon Rawls’ Theory of Justice [13]. Van Parijs argues that someone who chooses not to work, and who therefore has a low income, falls among the least-advantaged members of society. From this, Van Parijs’ argues that these people (dubbed ‘surfers’ in the paper’s title) should be granted a Basic Income – a ‘wealth-distributing, power-conferring, self-respect-preserving unconditional Basic Income. Indeed, that one should introduce such an income at the highest sustainable level.’ [11: 105]
Rawls did not agree with Van Parijs’ conclusion, writing that ‘those who surf all day off Malibu must find a way to support themselves and would not be entitled to public funds.’ [14: 257] A key reason for the dispute is that in arguing that such individuals need support, Parijs failed to recognise the diversity of ‘leisure’ – most crucially, the distinction between its desired and undesired forms.
The provision of social security to the unemployed is generally premised on the assumption that most periods of unemployment are involuntary. Proponents of this view do not see ‘leisure’ as something that the unemployed value, at least not as much as they would value earnings. In contrast, those who voluntarily choose to give up well-paid jobs, whether temporarily or permanently, for the delights of the Malibu surf are doing so because of the value they attach to their leisure. They calculate the experiential gains as being at least equal to the earnings forgone.
In light of this distinction between voluntary and involuntary leisure, van Parijs’s justification for unconditionality fails. (For a fuller explanation of van Parijs’s error, see [15]). It is unjust that people who have the capability and opportunity to work, yet choose not to do so, can be categorised as part of society’s ‘least advantaged,’ and so enjoy a publicly funded basic income.
Beyond the issue of equity is the issue of cost. If a UBI came free then no one could be opposed. But it does not. The problem was well illustrated by Tony Atkinson:
‘In his 1972 campaign for the US presidency, George McGovern proposed a $1000 a year demogrant (basic income) financed by a broad-based income tax. The story goes that he made the announcement on the campaign trail, before returning to ask his economic advisor what tax rate would be required. The advisor, James Tobin, is said to have replied that, if you need an x percent tax rate to finance the rest of government, then a demogrant equal to y percent of average income means that the tax rate has to be (x+y). … It means that, with a 20 percent rate of tax needed to finance other government purposes, a flat-tax rate of 33 1/3 per cent would finance a basic income of 13 1/3 per cent of average income, which seems scarcely adequate to replace existing transfers.’ [16: 218]
To finance a UBI at a poverty level defined as half of the average income would require a flat rate tax of about 70 percent. Few realists can imagine that such a cost is feasible.
Thus, when the case for a full UBI is put under scrutiny, its attractions as some sort of economic panacea that should or could replace social security rapidly fade.
Perhaps recognising this, some propose UBI as an addition to the social security system, arguing that this could be done by replacing tax allowances with tax credits worth the same amount to all, in contrast to tax allowances whose value increases with the marginal tax rate paid. Making this kind of replacement could have real attractions while falling far short of claims made for a full UBI.
There may be other, more incremental, ways to make progress that could be seen as progress towards a UBI. For example, if advocates of UBI promoted the universal child benefit discussed in the previous section, then they might make common cause with those who see greatly increased universal child benefit as a secure basis for tackling child poverty and promoting equal opportunity.
This review of the current system’s limitations has shown that if poverty is to be overcome, there are reforms that need to be made to the social security system as well as within society more broadly. While replacement UBI-style systems may be a misguided revolutionary fantasy, there are many more feasible reforms that would reduce the deficiencies in the current system and go further towards relieving poverty.
First, it must be recognised that poverty can never be eliminated by social security alone. When Beveridge made his plan, the great majority of the poor were in households without any employed member. Now that over half of the households living in poverty in Britain have at least one employed member, low pay has become a major cause of poverty. Increasingly social security and tax credits have been used to supplement the incomes of the low paid. Despite the minimum wage, low pay has become more prevalent and zero hours contracts and insecure employment have spread, so that the gap that social security has sought to bridge has become ever wider. Genuine social security requires not just an effective system of social security benefits but also requires secure and decently paid employment – which in turn requires affordable housing, transport and childcare.
Social security that is primarily means-tested – as it now is for 80 percent of working age recipients in the UK (compared to 25 percent in 1979) – has serious disadvantages compared to universal benefits. It is more costly to administer both for claimants and government, its take-up is lower and it can stigmatise claimants. Furthermore, the means tests involved result in high marginal effective tax rates which can distort incentives to work: as people earn more and start to exceed the cut-off points for eligibility of their various benefits, their extra income earned can be strongly offset by the loss in benefit grants.
Instead, moving towards a targeted universal benefits system, with certain groups or contingencies ‘tagged’ as being eligible for benefits, would reduce such costs and equity concerns. Such ‘tagging’, such as that advocated for by Akerlof, and expounded by Tony Atkinson, works. Inevitably, there is some over-inclusion, with some people given benefits that they do not need. Such over-inclusion was seen as a failure by Milton Friedman in the 1960s, which is what led him to embrace wholly means-tested support, taking the form of a negative income tax [17]. What Friedman failed to recognise was that this system, as with UBI, means either having a wholly inadequate minimum income or hugely increasing tax rates for everyone.
These arguments in favour of targeted universal benefits do not mean that, where such benefits currently exist, there are no further reforms needed. The UK’s universal benefit for children, for example, still falls far short of adequacy. Child poverty cannot be eradicated until this changes.
In addition to this, the process for determining eligibility needs to be reconsidered. As the Ken Loach film I, Daniel Blake depicted, many of the rules have become Kafkaesque, or impose too great a burden on individuals already facing difficult circumstances. While every benefits system must have rules on eligibility – and this is true of UBI whether based on nationality or residence 2 the focus should be on making recipients’ lives easier, not creating more challenges. Such rules and their enforcement must recognise the diversity of the ‘crooked timber of humanity,’ rather than seeking to suspect all claimants of being scroungers. No one should lose the right to be treated fairly and humanely just because they are poor or dependent on benefits.
Finally, if social security is to be part of an integrated approach that tackles the causes of poverty as well as its consequences, its animating philosophy must change, moving from a basic system of income redistribution that we currently have towards a focus on poverty in all economic and social policies. For example, there must be more effective connections between the provision of financial benefits and basic services, as Coote and Percy have recently discussed [18]. In many developing countries, eligibility for conditional cash transfers benefits is linked to participation in education and/or health programmes designed to improve opportunities. The UK should move away from the negative approach of past welfare systems, like the workhouses, and towards a more positive, development-orientated approach.
Most advanced economies now have well-established social security systems, with these systems commanding the largest amount of government spending in most of these countries. This combination of longevity and breadth means that many assume social security’s survival is without question. Such a sanguine approach neglects real uncertainties and risks, in particular failing to account for the vagaries of public opinion – on which nearly all government programmes ultimately depend.
Political attitudes depend substantially on perceived self-interest: on whether individuals see themselves as current or potential payers or recipients of social security, and also on how likely and for how long they expect to be in each category. For those currently paying into the system, they may question the probability of the system existing when they come to retire. There are different societal attitudes, with some being more collectivist, viewing children and the elderly as a collective responsibility of society, and with others more individualist, viewing them as purely private responsibilities.
In recent British general elections, social security received very little attention, with most domestic focus put on other issues such as the NHS, immigration and crime. While inattention may suggest political consensus, it may also indicate disinterest among the public – and so enable governments to raid social security in order to either fund more electorally valuable programmes or cut taxes, as has to a substantial extent occurred since 2010.
Reversing this apathy among the electorate should therefore be the focus for any political parties or interested groups that want to rejuvenate social security. Following the publication of his plan in 1942, Beveridge launched a personal programme of lectures and newspaper articles promoting it. The Labour Party joined in this campaign, and so the implementation of Beveridge’s programme, alongside the establishment of the NHS, became key issues in the 1945 election. Such electoral interest in social security has never since been repeated.
How then could the survival and strengthening of social security be promoted?
First, to promote social security, the public must first have a more informed understanding of how social security operates, and how all are likely to depend on it at different stages of their life.
Second, in order to aid such understanding, the system must be comprehensible and efficient, while treating recipients as cooperating citizens who deserve respect, courtesy and clarity. The consequences of austerity, telephone and on-line processing have made the benefits of social security increasingly opaque, and have dehumanised what should be, at heart, a human service.
Third, social security should be allied with other programmes that tackle causes of poverty and promote behaviour beneficial to individuals and their children – positive conditionality. By using these programmes with the social security framework, positive, independent actions taken by individuals to move out of poverty can be directly rewarded.
Fourth, there should not be the expectation that social security alone can solve everything. Broader strategies must be in place that offer individuals genuine security against macro-economic risks like inflation and unemployment, against social risks like insecure employment and the costs of children and family break-up, and against personal risks, such as disability, chronic indebtedness, crime, gambling and other addictions, all of which can contribute to long-term poverty.
Fifth, while social security can and does achieve a great deal, it must be recognised and emphasised that it cannot achieve everything. It must be a key component of a broader strategy that provides genuine security against macro-economic risks of inflation and unemployment, against social risks posed by insecure employment and the costs of children and family break-up, and against personal risks of disability, chronic indebtedness, crime, gambling and other addictions, all of which can contribute to long-term poverty.
Finally, the primary goal of social security being to end poverty must be emphasised, defended and promoted as fundamental to the rights of individuals and to the opportunities of children – and in turn to the welfare of society as a whole. One starting point in this direction would be to increase child benefit to track the poverty level; this would have amounted to about £54 per child per week, based on the latest 2018-19 data [19] – approximately three times the current average level of child benefit, with an annual net cost of around £25 billion. By contrast, paying a full Universal Basic Income to all which was set at the poverty level for a single person would have an annual gross cost of about £600 billion, more than double current social security spending. The reduction in child poverty, the reduction in dependence on Universal Credit, the contribution to children’s health, education and security, as well as to equal opportunities and to ‘levelling up,’ all indicate that this would be a sound investment in a sustainable social security system and in society as a whole.
In any country, the social security system reflects the society. Over the past century social security has grown exponentially into the biggest single activity of government in most advanced economies – growth that has reflected changes in political values and power, and changes in economic and social circumstances. Yet the goal of ending poverty is far from being achieved, which leads many to question the future of social security.
It has been argued here that social security remains a crucial component of a socially just society. Focussing only on those already in poverty through social assistance has serious disadvantages and disincentives. Universal benefits for children, disabled people and the elderly extend equal opportunities more effectively than assistance systems do for those for whom things have already gone wrong. But it is neither necessary, fair nor feasible to replace social security with a universal basic income.
Social security is an essential complement to other state services not secondary to, nor replaceable by them. It cannot solve all the ills of the economy or society – policies to ensure access to secure employment and fair pay are essential, as are policies to protect the vulnerable against exploitation. Yet, if this core element of a just and civilised society is to be maintained, it is essential actively to promote public understanding of social security and build awareness of how fundamental it is to a just society.
1Housing benefit, jobseekers’ and employment allowance, child and working tax credits, and income support.
2This would be the case even for a system of UBI, where questions over nationality or residence, for instance, would still require difficult decisions to be made.
The author has no interests to declare other than a concern with reducing poverty and promoting a more just society.
The author is Professor Emeritus of Social Policy at the London School of Economics. He is most grateful to Fran Bennett and Jonathan Bradshaw for many helpful corrections and suggestions, most of which he has heeded.
This paper was largely prepared before the Coronavirus Pandemic: given the present uncertainty about its social and economic consequences, no attempt has been made to consider its possible implications for social security.
Lindert P. Growing Public: Social spending and economic growth since the eighteenth century. Cambridge: Cambridge University Press; 2004. DOI: https://doi.org/10.1017/CBO9780511510717
Beveridge W. Social Insurance and allied services, Cmd 6404. London: HMSO; 1942.
Department of Work and Pensions. Universal Credit: Welfare that works, CM 7957. London: HMSO; 2010.
Timmins N. Universal Credit: from disaster to recovery? London: Institute for Government; 2016.
National Audit Office. Universal Credit: Early Progress. HC 621. London: Stationery Office; 2103.
Brewer M, Joyce R, Waters, T, Woods J. Universal Credit and its impact on household incomes: the long and the short of it. IFS BN248. London: Institute for Fiscal Studies; 2019.
Barrientos A. Social Assistance in Developing Countries. Cambridge: Cambridge University Press; 2013. DOI: https://doi.org/10.1017/CBO9781139856058
Webb B. Royal Commission on the Poor Laws and Relief of Distress 1905–09, Minority Report, Cmd. 4499. London: HMSO; 1909.
Harari YN. Homo Deus: A brief history of tomorrow. London: Penguin; 2016. DOI: https://doi.org/10.17104/9783406704024
Akerlof G. The economics of ‘tagging’ as applied to the optimal tax, welfare programs, and manpower planning. American Economic Review. 1978; 68(2): 38–42.
Van Parijs P. Why surfers should be fed: The liberal case for an unconditional Basic Income. Philosophy and Public Affairs. 1991; 20(2): 101–131.
Van Parijs P, Vanderborght Y. Basic Income. Cambridge, MA: Harvard University Press; 2017. DOI: https://doi.org/10.4159/9780674978072
Rawls J. A Theory of Justice. Cambridge, MA: Harvard University Press; 1971.
Rawls J. The priority of right and the idea of good. Philosophy and Public Affairs. 1988; 17(4): 251–276.
Piachaud D. Basic income: confusion, claims and choices. Journal of Poverty and Social Justice. 2018; 26(3): 299–314. Policy Press. DOI: https://doi.org/10.1332/175982718X15232797708173
Atkinson AB. Inequality: What shoud be done? Cambridge, MA: Harvard University Press; 2015. DOI: https://doi.org/10.4159/9780674287013
Friedman M. Capitalism and Freedom. Chicago, IL: University of Chicago Press; 1962.
Coote A, Percy A. The Case for Universal Basic Services. Oxford: Polity Press; 2020.
Department of Work and Pensions. Households below average income: an analysis of the income distribution 1994/95 to 2018/19: Summary Results. 2020, (accessed 19 June 2020).