Intergenerational fairness Contents

3The state pension triple lock

State pension uprating policy and its effects

43.The ‘triple lock’ or ‘triple guarantee’ was announced by the Coalition Government in the June 2010 Budget75 and implemented in 2012. It uprates the basic state pension (BSP) annually by the highest of:

The triple lock also applies to the new state pension (NSP), which was introduced for people reaching state pension age from 6 April 2016.77 The Government confirmed its commitment to retaining the triple lock until 2020 in the June 2015 Budget.78

44.An earlier statutory link between the BSP and earnings was removed by the Social Security Act 1980.79 It was restored by the Pensions Act 2007 and came into effect in 2010.80 This followed the 2005 Pensions Commission recommendation that restoration of the link was required to “stop the spread of means testing which would occur if present indexation arrangements were continued indefinitely”. Means testing undermined incentives to save privately and therefore the prospect of people saving adequately for retirement.81 In the interim period, the BSP was uprated in line with the Retail Prices Index (RPI), a further measure of price inflation. This indexation was supplemented in 2001 by a minimum annual increase of 2.5 per cent.82 This followed controversy over a 75p increase in the weekly pension prompted by low inflation in 1999.83

Figure 13: Trends in the state pension compared to earnings

Source: Work and Pensions Committee calculations based on ONS AnnualSurvey of Hours and Earnings / New Earnings Survey, DWP Annual Abstract of Statistics, OBR Economic and Fiscal Outlook

45.Figure 13 shows trends in the BSP and NSP relative to average earnings. As prices tend to rise more slowly than earnings, the value of the BSP declined relative to average earnings during the period of price indexation. The BSP was 26 per cent of mean full-time earnings in 1979, falling to 16 per cent in the period 2000–2008.

46.A combination of economic factors and policy changes in recent years have resulted in the state pension making up much of the ground it had lost relative to earnings during the 1980s and 1990s. Historically, earnings growth has tended to exceed both inflation and 2.5 per cent. This was not, however, the case in the aftermath of the 2008 financial crisis. Exceptional economic circumstances resulted in the BSP increasing relative to average earnings at a rate that would not have been anticipated when either uprating regime was instigated. Steve Webb described the rapid growth in the state pension relative to earnings as a “windfall”, as his concern as Pensions Minister had been that such progress would be “glacial”.84 RPI exceeded average earnings for the purposes of state pension uprating in 2008–09, 2009–10 and 2011–12. In the intervening year, the Government applied a discretionary increase of 2.5 per cent,85 which exceed average earning growth and both measures of inflation. With the triple lock then in place, the BSP increased by more than average earnings in each year from 2012–13 to 2015–16 as either CPI inflation or 2.5 per cent exceeded earnings growth (figure 14). As a consequence of these combined increases the BSP is expected to be 18.6 per cent of average full-time earnings in 2016–17, a level not seen since the 1980s.

47.The full rate of the NSP is currently £155.65, substantially higher than the BSP at approximately 24 per cent of average full-time weekly earnings.86 This percentage is near the peak of the BSP’s value relative to average earnings (26 per cent in 1979). Under the triple lock, the NSP will still be worth at least 24 per cent of average full-time earnings by the end of this Parliament (see figure 13). In time, the vast majority of new pensioners—between 80 and 90 per cent—will receive the full rate of NSP, which will place their incomes above the threshold for the means tested Pension Credit guarantee (currently £155.60).87 When asked whether the state pension is still too low, Baroness Altmann told us it:

is now recovering to a level that is decent and the new state pension on top of this will provide a good base for people to make their own private provision on top. That is the purpose of the policy.88

Now the former Pensions Minister, she argued recently that “the triple lock will have fulfilled its purpose by 2020”.89

48.The triple lock, allied to the introduction of the flat rate new state pension, has succeeded in increasing the value of the headline state pension relative to average earnings to a level not seen since the original earnings link was removed in 1980. Low rates of earnings growth following the 2008–09 recession mean this process has occurred faster than was expected. Provided the new state pension is maintained at this proportion of earnings the work of the triple lock, to secure a decent minimum income for people in retirement to underpin private saving, will have been achieved.

Sustainability of the triple lock

The ratchet effect

49.As the state pension rises each year by the highest of three factors, it is structurally more generous than its individual components. It sets the state pension on a permanently divergent upward trajectory relative to both earnings and prices. Lord Willetts described this as a ratchet effect:

It really is a ratchet, where regardless of the state of the economy, regardless of what is happening to other groups’ earnings and incomes, the pension just rises substantially.

50.The charts below show the effect of the triple lock since it was first applied in respect of 2012–13 state pension rates. The BSP was uprated in line with CPI in 2012­–13 and 2014–15, earnings growth in 2016–17 and the 2.5 per cent minimum in 2013–14 and 2015–16 (figure 14).

Figure 14: Indexation factors applicable to fiscal years 2012–13 to 2021–22

Indexation factors that apply to State Pension rates in each financial year:

(a) CPI inflation (annual growth) in September of preceding year

(b) Average Weekly Earnings (AWE) index - annual growth in total pay (3-month average) as at July of previous year (ONS series KAC3)

(c) 2017–18 onwards are OBR forecasts (Economic and Fiscal Outlook March 2016)

On current forecasts, the triple lock will result in the basic state pension being worth nearly seven per cent more than it would have been under a simple earnings link by the end of this Parliament (figure 15). It will be worth nearly 4 per cent more than if the previous system of RPI indexation had applied over the period.

Figure 15: Indexation trajectories for basic state pension from 2011–12 onwards

Work and Pensions Committee analysis; indexation factors from same sources as figure 14

51.Reform, a think tank, estimated that the triple lock will in 2016 result in annual state pension expenditure £4.5 billion higher than it would have been had a simple earnings link been in place.90 This gap can only grow. The most recent OBR projection of the long-term impact of the triple lock on state pension expenditure was contained in its 2015 Fiscal Sustainability Report. Given that in some years earnings growth is expected to fall below the rate of price inflation or the 2.5 per cent minimum, the OBR estimated that the triple lock delivers an average annual increase of 4.9 per cent in the long term. That is 0.39 percentage points higher than average annual earnings growth (4.5 per cent) and 2.89 percentage points higher than CPI inflation (assumed in the long term to be 2.0 per cent in line with the Bank of England’s target).91 Based on this assumption, the OBR estimated that:

Figure 16: OBR projections of state pension expenditure, % of GDP

Source: based on OBR Fiscal Sustainability Report June 2015 chart 3.7

52.The long-term effect of the triple lock ratchet on state pension expenditure was the backdrop for many commentators calling for its abandonment. The IEA said it had “little economic justification” and created “an ever-inflating cost dynamic for the state pension”. 93 Philip Booth added “it is crazy to have a system by which the pension rises by the greater of three different figures and it just leads to a totally arbitrary level of the state pension”.94 Baroness Altmann argued that the 2.5 per cent minimum annual increase was “arbitrary” and had “no specific relationship to society or the economy”.95 The IFS concurred, noting it applied “regardless of changes in meaningful economic variables like earnings and prices”.96 Its Director, Paul Johnson, said the triple lock:

adds £15bn to costs by 2050 relative to earnings indexation; introduces an element of pure randomness into the pension level [and] makes no sense as a policy.97

The IFS summarised that the triple lock is not sustainable indefinitely and that “there is no plausible objective to which the triple-lock is the best solution”.98

53.The Intergenerational Foundation argued against the triple lock on grounds of intergenerational fairness:

From an equity standpoint, it makes sense to uprate the BSP in line with either inflation (to maintain its purchasing power) or earnings (so that it increases at the same rate as the working population’s capacity to pay for it), but increasing it by more than either of these measures is unfair, especially at a time when many working-age benefits have been either frozen or indexed by less than the value of inflation.99

Professor James Sefton added that the triple lock prevents “intergenerational risk sharing” in the event of shocks to the economy.100 It shares increased prosperity but concentrates the costs of an economic crisis on younger groups.

54.The triple lock is inherently unsustainable. In the absence of reform the state pension would inevitably grow at a faster rate than the rewards of work and would account for an ever-greater share of national income. In particular, we find no objective justification for the 2.5 per cent minimum increase.

State pension age increases

55.As was demonstrated by OBR analysis, if the triple lock is to be maintained then the long-run sustainability of state pension expenditure will be dependent on further increases in the state pension age. The state pension age for men and women is currently scheduled to rise to 67 by 2028 (affecting those currently aged 38–54) and to 68 by 2046 (affecting those currently aged under 38). The Pensions Act 2014 provides for periodic longevity-based reviews of the state pension age.101 The first such review, led by John Cridland, will report in early 2017. The ‘guiding principle’ of these reviews is to maintain the proportion of adult life spent in receipt of a state pension at no more than a third. This means that, on the basis of current projections, “the increase in the state pension age to 68 is likely to come forward from the current date of 2046 to the mid-2030s, and that the state pension age is likely to increase further to 69 by the late 2040s”.102 Even with SPA increases along these lines, however, the OBR projects that expenditure on the triple-locked state pension will continue to rise markedly as a share of GDP.103

56.Steve Webb, who introduced the triple lock as Pensions Minister, defended its long-term viability by arguing that it should be viewed not in isolation but as part of a package of structural changes to the state pension system. The new state pension, he said, acted to “strip out long-term cost pressures from the system.”104 The new state pension is contributory but people cannot accumulate more entitlement than the full flat rate for which 35 qualifying years of national insurance contributions or credits are required. Someone who leaves school at 18 may have a working life of 50 years or more, but further contributions will not result in higher pensions. Steve Webb acknowledged, “today’s teenagers will be lucky to get a pension before they reach the age of 70”.105 In summary, he described his state pension policy as “we will not pay you a pension until a much later age [ … ] but when you get it, it will be worth having”.106

57.Age UK similarly defended the triple lock as being of long-term benefit to today’s younger generations:

During retirement, most other parts of income such as private pensions and annuity income, are likely to fall in real terms. The triple lock will ensure that at least one element of retirement income maintains its value over time.107

The Pensions Policy Institute noted that the triple lock greatly increases the probability of today’s younger workers achieving income replacement ratios in retirement that adequately maintain their living standards.108

58.We heard concerns, however, that the trade-off of an ever higher state pension age to compensate for generous uprating disadvantaged certain groups. Baroness Altmann said:

Many low earners, people in heavy industrial jobs, those living in particular areas of the country have much lower life expectancy than average. These significant variations in life expectancy mean that continually increasing state pension age will further disadvantage particular groups of society.109

In its interim report, the state pension age review considered some of these points. It found significant geographical differences in life expectancy. For example, life expectancy at birth in Greater London is 2.6 years higher for women and 2.4 years higher for men than in Greater Manchester. There are, however, even wider differentials between local areas within those urban areas, as shown in figure 17, which is based on transport networks:110

Figure 17: Resident male life expectancy at birth by public transport stops

Source: Independent review of the state pension age: Interim report, October 2016

59.The state pension age review also found significant differences in life expectancy by local deprivation levels and socio-economic group. For example, life expectancy at birth was 5.9 years higher for men and 4.4 years higher for women in the managerial and professional classification than for those in routine occupations.111 People living in deprived areas or in lower socio-economic groups tend to spend far fewer years in retirement than their affluent peers and are less likely to be healthy in that retirement.112 People who leave school in their teens may well also contribute more years of work than their university-educated peers for no higher new state pension. The review has invited responses on whether the state pension age is the appropriate mechanism for ensuring the fiscal sustainability of the state pension.113

60.There is, to some extent, a trade-off between the uprating of the state pension and the state pension age. The cost of a more generous state pension can be offset by restricting its availability to fewer people. Increases in the state pension age, however, disproportionately affect younger people. They also risk further skewing receipt of the state pension towards people in areas of the country, and socio-economic groups, in which life expectancy is high. People with low life expectancies, who may have been disadvantaged in their early years and working lives, would be further disadvantaged in their later years. We do not doubt that further increases in the state pension age will be required as welcome increases in life expectancy continue. They should not, however, be the sole means of ensuring the long-term affordability of the state pension. We welcome the wide-ranging approach the ongoing review of the state pension age has taken to its work.

Alternatives to the triple lock

Removing one of the elements of the triple lock

61.The IEA argued that a return to price-indexation of the state pension would be “particularly helpful” to achieving fiscal sustainability.114 Saga identified the earnings component of the triple lock as “potentially problematic” on the grounds that “should earnings shoot up in future the triple-lock will prove unsustainable”. It suggested that the triple lock could be replaced by a link to a pensioner-specific price index, underpinned by a 2.5 per cent minimum increase.115

62.We heard however that an earnings link was important not only to enable state pensioners to share in rising prosperity, but to incentivise private saving. The full rate of the NSP is currently set just above the level of the means tested Pension Credit standard income guarantee, which is itself subject to statutory earnings uprating.116 In announcing the NSP in 2013, the then Government said it would “simplify the state pension and better support saving for retirement”.117 The DWP estimates that by 2030 the NSP system will reduce eligibility for Pension Credit guarantee credit by nearly a third for new pensioners compared with the legacy system.118 If the new state pension were to lag behind earnings, it would soon fall below guarantee credit level, making increasing numbers eligible for a means tested top-up. This would erode the incentive for additional pension saving, undermining the central purpose of the reform.119 In our Report on pensions automatic enrolment we concluded that the Government should consider means of promoting further private saving to ensure more people had adequate incomes for retirement.120

63.Baroness Altmann argued for the abandonment of the “arbitrary” 2.5 per cent minimum increase from 2020, making the triple lock a “double lock” with the state pension being increased by the higher of prices and earnings growth in any given year. This, she argued, would protect pensioner incomes while delivering “billions of pounds lower long-term pension costs”. She added that governments could still opt to apply a higher discretionary increase in an individual year if circumstances dictated.121

64.Andrew Hood of the IFS cautioned, however, that a double locked pension would still tend to account for an ever great share of national income:

If you took the 2.5% out of the triple lock and said the state pension will rise with the higher prices and earnings, you still have the problem that over the long run it would rise faster than both. During the recession it would have risen in line with prices, therefore becoming a higher share of average earnings, and then because it goes up in line with average earnings after that, that has now got baked in, it has got locked in.

Smoothed earnings link

65.The IFS suggested an alternative system which would ensure that growth in the state pension:

That combination of policy objectives could be achieved by linking the state pension to a fixed minimum proportion of average earnings, a system in operation in Australia. The state pension would be uprated with earnings, but with temporary price-indexation when inflation exceeded wage growth. Price indexation would continue once earnings growth again exceeded inflation, but only for as long as the value of the state pension remained above the original fixed minimum proportion of average earnings. Indexation would then revert to earnings.122 Pensioners would not be hit with real falls in income, but the state pension would not rise ever higher relative to earnings. Andrew Hood of the IFS told us, therefore, that “you can have your cake and eat it”.123

66.Figure 18 illustrates how an ‘Australian-style’ smoothed earnings uprating mechanism might work compared to other indexation methodologies. It shows a notional 30-year period during which average earnings grow twice as fast as prices in most years but each decade is punctuated by a brief period of below-inflation earnings growth, as might occur in a recession. The double lock is slightly less generous than the triple lock in the long run, as there is no fixed minimum increase of 2.5 per cent. While offering transitional protection during the recession periods, an Australian-style uprating system is less generous still, tracking earnings growth in the long run. In doing so, it would avoid the structurally higher long-term costs associated with a triple lock or double lock.124

Figure 18: Illustration of different indexation trajectories over 30 years

67.The triple lock should not continue beyond 2020. By then, the value of the new state pension relative to average earnings will be close to the historic high for the headline state pension rate. If maintained, the arbitrary boost the triple lock gives to the state pension relative to both earnings and prices will become ever harder to justify both in fiscal terms and from the perspective of intergenerational fairness. We urge political consensus before the next general election on a new earnings link for the state pension.

68.We recommend the Government benchmark the new state pension and basic state pension at the levels relative to average full-time earnings they reach in 2020. The triple lock should then be replaced by an earnings link. In periods when earnings lag behind price inflation, an above-earnings increase should be applied to protect pensioners against a reduction in the purchasing power of their state pension. Price indexation should continue when real earnings growth resumes until the state pension reverts to its benchmark proportion of average earnings. Such a mechanism would enable pensioners to continue to share in the proceeds of economic growth, protect the state pension against inflation and ensure a firm foundation for private retirement saving. The new state pension and basic state pension it replaced would track average earnings growth in the long term. That is more fiscally sustainable and more intergenerationally fair.

75 HC Deb 22 June 2010, vol 512 cols 166–80 [Budget statement]

76 In the legacy state pension system (which applies to those who reached state pension age before 6 April 2016), the triple lock only applies to the Basic pension while the additional State Pension (SERPS/S2P) is raised in line with CPI inflation.

77 HL Deb 28 April 2016 col 1235 [Lords Chamber]

78 HM Treasury, Summer Budget 2015, 8 July 2015, HC (2015–16) 264;

79 Social Security Act 1980, section 1 amended section 125 of the Social Security Act 1975 to link long-term benefit increases to prices, not earnings.

80 Social Security Administration Act 1992, s150A - inserted by Pensions Act 2007 s5(1), which restored the earnings link, and amended by Pensions Act 2014 which introduced the new State Pension.

81 Pensions Commission, A New Pension Settlement for the Twenty-First Century, Second Report, November 2005, Executive summary, page 10–12.

83 State Pension Uprating, House of Commons Library Briefing Paper SN-05649, 31 Aug 2016, para 2.1

84 Q4 [Steve Webb]

85 This discretionary increase was announced in the 2009 Pre-Budget Report (HC Deb 9 Dec 2009 col 362).

86 Based on average full-time weekly earnings of £627.40 as at April 2015 (source ONS Annual Survey of Hours and Earnings 2015 provisional results table 1), increased in line with average earnings growth of 2.4% in 2015/16 as forecast by OBR (March 2016 Economic and Fiscal Outlook)

89Triple lock for state pensions could move to a double lock”, Baroness Altmann, blog, 30 July 2016

90 William Mosseri-Marlio, “Cost of the triple lock is set to surpass £20 billion”, Reform, 28 October 2016

91 OBR Fiscal Sustainability Report June 2015, table 3.4, p 63. The 2016 FSR publication has been cancelled in the wake of the EU Referendum.

93 Institute of Economic Affairs (IGF0031)

94 Q231 [Philip Booth]

95 Baroness Altmann, Triple lock for state pensions could move to a double lock, blog, published 30 July 2016; “Government to keep the triple lock on state pension until 2020,” Financial Times, 31 July 2016

96 Institute for Fiscal Studies (IGF0023)

97Pension policy – where have we been, where are we going?” Presentation by Paul Johnson, Institute for Fiscal Studies, 20 October 2015

98 Institute for Fiscal Studies (IGF0023)

99 Intergenerational Foundation (IGF0010)

100 Qq 2 6, 272 [James Sefton]

102 HM Treasury,Autumn Statement 2013, Cm 8747, December 2013, paras 1.119–1.125; see also Autumn Statement announcement on a core principle underpinning future State Pension age rises: DWP background note December 2013

103 OBR Fiscal Sustainability Report June 2015, para 3.56 and chart 3.7

104 Steve Webb, “Yes, pensioners have never had it so good – but they still deserve the ‘triple lock’”, Telegraph, 22 Oct 2015. See also Work and Pensions Committee, Eighth Report of Session 2015–16, Communication of the new State Pension, HC 926, para 29

105 As above

106 Q5 [Steve Webb]

107 Age UK (IGF0044)

108 Pensions Policy Institute (IGF0055)

109 Baroness Altmann, “Triple lock for state pensions could move to a double lock,” blog, 30 July 2016

111 Independent review of the State Pension age: interim report, 13 October 2016, pp 58–60. Figures related to England and Wales.

114 Institute of Economic Affairs (IGF0031)

115 Saga (IGF0069)

116 Social Security Administration Act 1992 section 150, as amended by Pensions Act 2007 section 5

117 Department for Work and Pensions, 2013, The single-tier pension: A simple foundation for saving, CM 8528

119 TUC (IGF0042), para 5.4

120 Work and Pensions Committee, Eleventh Report of Session 2015–16, Automatic enrolment, HC 579

121 Baroness Altmann, “Triple lock for state pensions could move to a double lock,” blog, 30 July 2016; “Government to keep the triple lock on state pension until 2020”, Financial Times, 31 July 2016

123 Q287 [Andrew Hood]

124 The Australian Age Pension differs from the UK State Pension in that it is means tested and eligibility is based on years of residence rather than contribution record. The Age Pension for a couple is currently benchmarked at 41.76 per cent of male total average weekly earnings whereas the single person rate is 66.3 per cent of the couple rate (27.7 per cent of average male earnings). Source: Australian Dept of Social Services and Australian Parliamentary Library Research Papers Pension indexation: a brief history(April 2014)

© Parliamentary copyright 2015

4 November 2016