Policy Papers

Politics and pensions in post-war Britain

Hugh Pemberton |

  • RSS Feed Icon

Executive Summary

  • A notable feature of the current pensions debate is a persistent failure to look back as well as forward, to consider the roots of the pensions crisis as well as possible solutions to it. But we cannot hope completely to understand the present crisis, let alone devise effective solutions to it, without understanding its history.
  • There are important lessons to be drawn from the development of British pensions over the past 60 years.
  • A history of repeated 'reforms' to the UK's overall system of pensions suggests that further reform is possible.
  • However, each attempt at reform faced the problem that long-term contracts between workers and their pension provider (public or private) meant that any break would incur substantial financial and/or political transition costs.
  • Politicians, faced with the prospect of incurring such short-term costs, preferred to achieve change that would incur costs only in the long-term.
  • This combination of long-term contracts and short-term political horizons encouraged governments to achieve 'reform' not by replacing an existing element within the system but by creating a new sub-system, or by allowing a new sub-system to emerge.
  • British pensions therefore embodied a systemic tendency towards growing complexity in the system as a whole, and this contributed to decreasing system efficiency.
  • In the present context, history suggests that change is again likely to be achieved by adding a new element to the system, thus incurring a further increase in the level of system complexity.
  • In addition, the failure of Labour's 1957 proposal for a PAYE based 'national superannuation' scheme with funds invested in the stock market does not augur well for the Pensions Commission's current proposal for a National Pensions Savings Scheme.


The problem of maintaining incomes in old age has become one of the most urgent political issues facing British policy makers today. It seems now to be generally accepted that we face a major crisis over the next few decades, the product in large part of our ageing population but also of the complexity and inefficiency of the current arrangements. The Pensions Commission, set up by the government to examine the problem, recently concluded in its second report that 'major reform of the UK pension system is needed to create a new settlement for the twenty-first century'. But if we are to craft and implement such a settlement we would be well advised to understand how the present crisis has arisen. And to do that, we need to understand the history of British pensions over at least the past half-century.

The importance of history

In its first report, the Pensions Commission remarked that the UK has one of the most complex pensions systems in the world and noted that this was the product of the way that the system had developed over more than a century. Then, in its second report, it was clear that the Pensions Commission had itself found history highly constraining in crafting its proposals for change in the state pension system. So it is more than surprising that that report devoted only 2 of its 460 pages to any consideration of how the present system, both public and private, developed or of the consequences of that development for any putative reforms. In fact, a notable feature of the current pensions debate is a persistent failure to look back as well as forward, to consider the roots of the pensions crisis as well as possible solutions to it.

The only real attempt to explore the developmental history of British pensions has been by historical institutionalists in public-policy analysis - who have argued that reform of the system is well-nigh impossible because the system has become locked-in by feedback effects that make further developments 'path dependent'. I do not propose to discuss path-dependency theory in detail here. It derives from the argument of economists and economic historians that the deployment of a new technology by an early entrant to the market may allow it to corner that market even though better technologies come along later - a phenomenon known as 'lock-in'. The concept was pounced on by historical institutionalists in political science and there seems general agreement amongst them that 'pensions policy is a locus classicus for the study of "path-dependent" change' as a consequence of its very long time-horizons. Whether or not the pension system is 'locked-in' matters because it has important implications when we come to explain why the complexity of the overall system seems to increase inexorably; and it also has important implications for any attempts to reform the system.

In the remainder of this paper I examine developments in British pensions since the publication of the Beveridge report in 1942, focusing particularly on key decisions made in the next two decades. In doing so I emphasise:

  1. The dynamics of continual change, which suggests the system is far from 'locked-in'.
  2. The discrepancy between the long-term nature of pensions and short-term political horizons.
  3. The fact that individual elements within this system are locked-in - by long-term contracts, whether explicit or implicit.
  4. The way in which a combination of 2) and 3) has created a tendency for the complexity of the system as a whole to increase.

The Beveridge settlement

When the Beveridge Committee started work in 1941, Britain's pensions system comprised:

  • The 1908 non-contributory and means-tested scheme.
  • The 1925 contributory scheme, which paid pensions for only 5 years from age 65, after which the pensioner proceeded to the 1908 scheme but without means-testing.
  • A back-up system of supplementary pensions paid to approximately 250,000 of the poorest pensioners.

Beveridge's task was essentially to unify the 1908 and 1925 systems, and to a large extent he succeeded. That he was able to do so was largely due to the exceptional conditions of the time. The war had two important effects:

  1. It reduced the political costs of moving to the new national insurance pension by virtue of the solidaristic environment and faith in the capabilities of the state produced by victory.
  2. It reduced the financial transition costs via wartime inflation which eroded the value of pre-war pensions and made it easier to offer a higher nominal pension for all.

In 1946, Britain shifted to a state-run universal contributory pension paid at a flat-rate until death; a system backed by National Assistance payments which Beveridge expected would largely wither away after a transitional period during which employee contributions would build up to yield a full pension.

Administratively, however, the new national insurance pension did not entirely transcend the preceding arrangements. More importantly, it did not entirely achieve its objectives. The causes of this failure lay in the need further to minimise both the financial costs of transition and the political costs. Three factors were at work:

  1. Although Beveridge's main task was to bring the 1908 and 1925 schemes together under a single administrative umbrella, the 1908 scheme continued for those who had retired without ever contributing to the 1925 scheme.
  2. In the straitened conditions of post-war Britain, Beveridge's desire to bring actual pensions paid up to 'subsistence' level - i.e. pay a pension that would be enough to live on and thus eliminate the need for supplementation - was not achieved.
  3. Beveridge's concept of 'National Insurance' implied a funded scheme in which an employee's contributions would purchase specific future benefits. In fact, even under Beveridge's proposals this was not entirely the case.

Moreover, it was made even less so by a political decision by the incoming Labour government to drop Beveridge's 20-year transition period, or 'golden staircase', to full pension rights. For Ernest Bevin and others on the Cabinet Committee overseeing the translation of the Beveridge plan into legislation, the long delay envisaged by Beveridge to allow contributions to build up was simply not practical politics in the context of sacrifices made by workers during the depression and the war. Instead, full pensions would be paid immediately to those insured since 1925, and after only 10 years to those who had joined later. But abolishing the 'golden staircase' had several deleterious consequences:

  • It severely weakened the finances of the new state pension by making it essentially an unfunded pay-as-you-go (PAYG) scheme with current pensions paid out of current contributions.
  • It helped make the pension inadequate because, with the long-term finances of the National Insurance Fund now in doubt, the Treasury ruled out plans to increase the pension up to 'subsistence'.
  • The politics of the decision created the appearance of a long-term financial contract between individual contributors to the system and the state. In effect, a political sleight of hand had been employed to allow the immediate payment of benefits, the cost of which would not have to be redeemed for many years. This, however, was concealed from contributors, who continued to view National Insurance as a financial contract between them and the state in which their contributions purchased rights to a fully-funded future pension.
  • This implicit financial contract assumed both flat-rate contributions and benefits. But a flat-rate contract tied the level of the contribution to that which the lowest-paid worker could afford. This was to prove a major constraint which shaped the development of pensions policy during the decade and a half after 1945.

The growth of occupational pensions in the 1950s

Post-war full employment to some extent lessened the pressure on the National Insurance Fund because the cost of unemployment benefits proved to be significantly lower that that envisaged by Beveridge. Nevertheless, by 1952 the Treasury's view was that the pension costs associated with increasing longevity, the abandonment of Beveridge's 'golden staircase' to full pension rights, the shift to an unfunded system, and the fact that improvements in the pension would themselves be unfunded, would combine to produce a steadily rising burden on the exchequer. Consequently, although the new national insurance pension was raised several times during the 1950s the Treasury conceded each rise with great reluctance and the pension steadily fell below National Assistance scales.

Not surprisingly, the Treasury was even more reluctant to accede to demands that the pension should reflect not just the increase in post-war prices but also the increase in real average earnings that was such a notable characteristic of the 'golden age'. Instead, in a very tight labour market, the demand was filled by top-up occupational pension schemes offered by employers. Before the war, occupational pensions had been the preserve of a minority - around 13% of workers, mainly public servants and higher paid private sector employees. By 1954, however, the Phillips Committee was noting that employers were becoming major providers of pensions. By 1956, according to the Government Actuary, the proportion of workers in occupational schemes had leapt to 33 per cent, most of this growth having occurred since the end of the war. By 1967 it had reached 53 per cent. By the 1960s, therefore, occupational pensions had moved from a 'niche' position to a central role in Britain's system of pension provision.

Like the 1946 national insurance pension, this new element in mass pension provision was 'locked-in':

  • Partly this was because the value of occupational pension funds, and the number of contributors, grew rapidly, giving the funds considerable market and political power which they were not afraid to exercise.
  • But pension benefits under an occupational scheme were also effectively accrued as a matter of contractual right under the rules of the scheme. As long as the employee remained in the scheme until retirement, and as long as the fund remained solvent, s/he was effectively guaranteed a pension. Thus, in the boom conditions of Britain's post-war 'golden age', the commitment by a robust scheme seemed to employees to be both firm and long-term.

Because the 1946 settlement had resulted in a minimalist state pension locked in by contractual obligations, it had created the conditions for the development of a parallel system of occupational pension provision on a vastly greater scale than Beveridge had envisaged. As a consequence, by the mid-1950s Britain effectively had two systems of mass pensions provision rather than the one envisaged by Beveridge; each locked-in by contractual obligations.

The adoption of a state graduated pension in 1961

Although the rapid growth of occupational schemes filled a gap in the market for many employees, the 1954 Phillips Report had noted that two-thirds of workers still remained wholly dependent on a flat-rate state pension which was not keeping pace with rising earnings. Recognising the electoral appeal of extending access to an earnings-related pension to all workers, in 1957 the opposition Labour Party proposed a system of 'national superannuation'. This would pay a state pension equal to approximately half income at retirement. Higher contributions would be collected from workers through the PAYE system and the state would invest the resulting funds in the stock market in order to build up a very large fund out of which the considerably improved pension would be paid. There are some interesting resemblances between this scheme and the National Pensions Saving Scheme proposed by the Pensions Commission in its second report. And, as we shall see, the ultimate failure of Labour's proposals does not auger well for the success of the scheme proposed by the Commission.

Labour's proposal was unusually well thought-out and envisaged a scheme that would be at least as good as any under discussion in social-democratic parties on the Continent. It could be expected to play well electorally and the Conservative government responded by bringing forward its own proposals. In theory, one way of satisfying the desire for earnings-relation in pensions was simply for the Conservatives to encourage the private sector to provide occupational pensions to all - a solution pushed aggressively by John Boyd Carpenter, the Minister for Pensions. Yet, as the life companies themselves privately acknowledged, it was never going to be possible for the private sector to provide occupational pensions for all workers because it would be uneconomic to collect contributions from workers employed in the multitude of small firms in the British economy. This inability of the private sector to deliver earnings-related pensions for all employees implied an extension of the existing state pension. The question was how extensive should this be and how feasible was reform of the existing national insurance pension?

Ian Macleod, then Minister of Labour, with the support of the Prime Minister, Harold Macmillan, produced initial proposals to compel employers either to set up and contribute to a private occupational pension scheme, or to enter a new state scheme of equivalent generosity - a radical reform which Macleod intended would encompass both the flat and supplementary pension. After considerable debate within government during 1957 and 1958, however, the intention to instigate a radical reorganisation of the 1946 state scheme was dropped. Instead, the flat-rate pension was retained and a separate, and not particularly generous, earnings-related or 'graduated' state pension was introduced. This was a critical moment in the development of Britain's system of state pensions and, by extension, in the development of the private sector.

Why was a generous state earnings-related pension not implemented?

The limited nature of the new graduated state pension was in part the product of a continuing anti-state culture in sections of the Conservative Party. Opponents of a generous state earnings-related scheme, such as Boyd Carpenter and Peter Thorneycroft, the then Chancellor, by creating deadlock-in the cabinet at a time of pressing need to defuse the threat posed by Labour's proposals, were able to secure a decision to adopt a less ambitious scheme, even if this was clearly going to be inadequate in the long-term. As had been the case in 1945, therefore, short-term political expediency proved more important than long-term effectiveness. As Macmillan noted to ministers, 'In the long run we shall all be dead...So do not let us bother too much as long as we do not spend too much for the next two or three years'.

But an important motive for the government's adoption of a limited scheme of earnings-relation was provided by the way in which an implicit contract with contributors had effectively locked-in the flat-rate state pension. There were two factors at work here:

  1. The fact that the 1946 settlement had created an apparent flat-rate contract between each contributor and the state. This at once fed the desire for 'reform', by undermining the finances of the National Insurance Fund, and made reform of the flat-rate pension all but impossible, because it would expose the link between contributions and benefits as a sham.
  2. The pressing need to solve the 'financial crisis' in the National Insurance Fund, coupled with the flat-rate design of the 1946 pension, which fed Treasury opposition to a generous state earnings-related pension.

Consequently, the focus in Whitehall rapidly shifted from how to improve the level of the state pension to the question how to devise a scheme that would appear to offer improved benefits, whist in reality using the higher contributions earnings-relation would produce to bail out the National Insurance Fund. Thus the enduring nature of the flat-rate 'contract' meant that all the plans considered by the government were forced, either explicitly or implicitly, to assume its continuation. Likewise, Labour's proposals also envisaged a continuation of the flat-rate scheme.

It is sometimes claimed that the 1961 implementation of a state graduated pension represented a decisive break with two of the four key principles of Beveridge's 1942 proposals - flat-rate contributions and benefits. However, whilst there was the appearance of such a break, in reality policy makers were forced to accept that the 'contract' represented by the 1946 settlement could not be broken. Consequently earnings-relation had to be implemented as a new element in the state scheme. The flat-rate pension continued, as did flat-rate contributions. The break in respect of the flat-rate principle was confined to the raiding of the new earnings-related scheme (the 'graduated pension') to subsidise the flat-rate pension.

The fact that this new element did not offer contributors a radically improved earnings-related pension reflected the market and political power enjoyed by the life and pensions companies as a consequence of the enormous post-war growth of occupational schemes. By 1956 there were over 8 million employees in such schemes and around 300,000 pensions in payment. With this came large funds to spend on policy research, on lobbying opinion formers, briefing the press, advertising, and mobilising trade union opposition to Labour's proposals. The interest of the industry, as both major parties came to realise, had to be taken into account. It could no longer be ignored as it had been in 1946. Indeed the ability of the industry to mobilise public and political opinion after 1957 and thereby to derail Labour's proposals, which represented a considerable threat to their business, owing to the cheapness of administration and the power it would give the government in the stock market, suggests that the Pensions Commission proposed National Pensions Saving Scheme may face similar obstacles.

The fact that the proposals of the Life Offices Associations formed the basis of both the Conservative government's plan for earnings-relation in the state system and Labour's final proposals for National Superannuation is therefore highly significant. Under both plans provisions were made to allow occupational schemes to 'contract-out' of the new earnings-related state pension, effectively encouraging the continued growth of such schemes, despite the fact that allowing them to do so creamed off higher-paid workers and thus significantly weakened the finances of the state scheme.

The construction of a state earnings-related pension was therefore highly constrained by the fact that contractual obligations and feedback effects served to lock-in each of the existing system elements. The result was a 'graduated state pension' that was limited in scope, used to bail out the National Insurance Fund over the short-term, and installed on top of the existing state scheme in order to respect the flat rate national insurance contract. Rather than 'reform', therefore, we again see a further element added to the pensions system.

Developments since 1961

It might be argued that the development of these three parallel sub-systems was simply a product of filling in what has come to be seen as the standard three tiers of pension provision - a basic pension provided by the state, an additional earnings-related element, and finally an additional voluntary component. Yet, a brief survey of subsequent developments shows that this process of sub-system accumulation has continued. Each of the elements in place by 1961 endures today, despite a number of subsequent major 'reforms' (see Figure 1).

At best, reform has involved the 'cold storage' of rights under a given pension when a new element has been introduced. In 1978, for example, the state graduated pension was apparently superseded by Labour's state earnings-related pension (SERPS). In fact, this 'reform' was unable to integrate the Conservatives' graduated pension into the new scheme. Contributors to the graduated pension retain an entitlement to its benefits to this day and will continue to do so until the last contributor dies.

More commonly, 'reform' involved the installation of a new system element in parallel with the old. In the 1980s, for example, the Conservatives' desire to abolish SERPS was thwarted by the attachment of its contributors to the earnings-related pension it produced, and by worries, not least in the Treasury, about the costs involved in crystalising future pension entitlements under SERPS. Such was the opposition that the Thatcher administration found it much easier in 1988 simply to introduce a new system of mass personal pensions in addition to, rather than as a replacement for, SERPS. The introduction of free-standing additional voluntary contributions (FSAVCs) in 1987 added a further element to the system.

Figure 1

Figure 1 - Contract path dependence in UK pensions

In the current decade, Labour's introduction of stakeholder pensions in 2000 was an addition to personal pension plans, not a replacement. And more recently, whilst Labour's introduction of the State Second Pension in 2002 brought an end to contributions under SERPS, it did not end SERPS pension rights - which were put into cold storage and will live on until all the scheme's contributors are dead.


There are important lessons to be drawn from the developmental history of British pensions over the past 60 years.

  • A pattern of continual change over more than half a century suggests that Britain's pension system as a whole is not 'locked-in' and indicates that further change is possible.
  • However, the short-term horizons and adversarial nature of British politics are profoundly unsuited to crafting pensions policy. The pattern of contracts between workers and their pension provider (public or private), rather than between one generation and the next, has meant that any break would incur substantial political and financial transition costs.
  • In effect, whilst the system as a whole is not 'locked-in', individual elements within that system tend to be. Coupled with short-term political horizons, this encouraged governments to achieve a series of 'reforms' not by amending the existing system but by creating a new system element, or by allowing a new element to emerge.
  • The result is that British pensions demonstrate a systemic tendency to growing complexity in the system as a whole. This has implications for decreasing system efficiency.
  • So, whilst change is possible, unless the long-term benefits can be seen to be worth the potentially very high financial and political costs of eliminating an existing element, it is likely to be achieved by adding a new element to the system.

In the context of the present pensions crisis these findings suggest that an attempt to create a new pensions settlement for the twenty-first century is likely to involve a further increase in the level of complexity in Britain's overall system of pensions.

More generally, they suggest that the key to a successful solution to the pensions crisis is to recognise both the very poor fit between the long-term contractual nature of pensions and short-term political horizons. In 1997, John Denham, the incoming Labour government's Minister of Pensions, acknowledged the cumulative failure of British pensions, the mounting complexity and decreasing efficiency of the system, the malign tendency to 'reform' the system every ten years or so, and the tendency for such reform to be dominated by short-term political horizons rather than long-term needs. The implication was that Labour would do things differently. It singularly failed to live up to this promise.

Yet the government's decision to set up the Pensions Commission and to encourage a national debate about how best to address the pensions crisis is encouraging. David Blunkett's professed intention to 'reach out to the other major political parties because we need a lasting solution for the decades ahead - not a quick fix' was also welcome. Mr Blunkett has gone, of course, but the principal lesson of history is that his instincts were right. Unfortunately, the whispering campaign mounted against the Pensions Commission's proposals for reform rather suggests that this lesson has not been learnt.

Further Reading

G. Bonoli, The politics of pension reform: institutions and policy change in Western Europe (Cambridge: Cambridge University Press, 2000).

B. Ellis, Pensions in Britain, 1955-1975 (London: HMSO, 1989).

H. Fawcett, 'The Beveridge strait-jacket: policy formulation and the problem of poverty in old age', Contemporary British History, vol. 10, no.1 (1996), pp. 20-42.

L. Hannah, Inventing retirement: the development of occupational pensions in Britain (Cambridge: Cambridge University Press, 1986).

J. Harris, William Beveridge (Oxford: Clarendon Press, 1997).

J. Hills et al, Beveridge and social security (Oxford: Oxford University Press, 1994).

Labour Party, National Superannuation: Labour's policy for security in old age (London: Labour Party, 1957).

R. Lowe, 'A prophet dishonoured in his own country? The rejection of Beveridge in Britain, 1945-1970' in J. Hills et al, Beveridge and social security (Oxford: Oxford University Press, 1994).

J. Macnicol, The politics of retirement in Britain, 1878-1948 (Cambridge: Cambridge University Press, 1998).

Pensions Commission, A new settlement for the twenty-first century. The second report of the Pensions Commission (London: TSO, 2005).

Pensions Commission, Pensions: challenges and choices. The first report of the Pension Commission (London: TSO, 2004).

P. Pierson, 'Increasing returns, path dependence and the study of politics', American Political Science Review, vol. 94, no. 2 (2000), pp. 251-67.

N. Timmins, The five giants (London: Harper Collins, 2001).

Related Policy Papers


Papers By Author

Papers by Theme


Sign up to receive announcements on events, the latest research and more!

To complete the subscription process, please click the link in the email we just sent you.

We will never send spam and you can unsubscribe any time.

About Us

H&P is based at the Institute of Historical Research, Senate House, University of London.

We are the only project in the UK providing access to an international network of more than 500 historians with a broad range of expertise. H&P offers a range of resources for historians, policy makers and journalists.

Read More