Civil Servant

Civil Service Pensions

Key Facts

There have been a number of recent changes to the main civil service pension scheme, generally in advance of similar changes to other public sector schemes. The key changes are as follows:

The old final salary pension scheme was closed to new entrants from July 2007. From that date, new entrants joined a career average scheme - see below for an explanation of what this means. The Government has also moved pre-July 2007 members to a new career average scheme from 2012 - but for future accruals only, and only for those more than 10 years from Normal Retirement Age.

Incoming Chancellor George Osborne announced in June 2010 that civil service pensions would in future increase each year in line with the Consumer Prices Index (CPI), not the Retail Prices Index (RPI). This immediately reduced the value of all scheme benefits - including those benefits already accrued from past contributions - by around 15%. (For instance, a comparison of historic RPI and CPI shows that if this rule had applied in the past then the pension of someone who had retired in 1988 would by 2010 have been 15% lower than it was. See notes 2 and 3 for further information about the differences between CPI and RPI.)

The Chancellor also announced that he believed that the cost, to the taxpayer, of public sector pensions needed to be reduced, and the most effective way to make short-term savings was to increase member contributions. The Chancellor announced that the discount rate used to estimate the future return on current pension contributions should be reduced, so requiring higher contributions - 'though partly from employers. The Chancellor said that he chose the rate at which the economy is expected to grow which he took to be 3%. If that was indeed his logic, then 3% seemed pretty optimistic, and he reduced the rate further to 2.8% in the 2016 Budget, although the resultant increase in pension contributions was this time taken only from public sector employers.

On the other hand, the Government announced in November 2013 that all public sector pensions would continue to increase after retirement in line with inflation (as measured by the CPI) for at least the next 25 years. This is not always the case for private sector pensions.

(Incidentally, civil servants (and many other public servants) are generally worse off from 1 April 2016 following the changes to the new higher 'flat rate' state pension. This is first because their national insurance contributions will rise as there is no longer any reduction for those who have opted into paying into their employer's scheme. And, second, those who have been in the civil service scheme will have their state pension significantly reduced because they have a second pension, so they do not benefit from the new higher rate pension. Eventually, of course, their increased contributions will lead to increased state pensions, but this will take many years to work through completely.)

The post-2012 pension scheme accordingly has the following features:

Benefits already accrued under previous schemes will be protected. In particular, those who entered the civil service before July 2007 will have the proportion of their pension that is accrued up to April 2012 calculated by reference to the salary they are on when they eventually retire.

Pensions will be uprated each year by reference to the CPI, not the RPI. Pension contributions have increased - in part as a result of the 'burden sharing' summarised below - but with further increases on top. The higher paid have had their contributions increased by up to nearly 5% more than the lower paid. The average increase was 3.2%.

The 'normal pension age' for civil servants retiring in the next few years is now 66, and it will eventually rise to 68, alongside similar changes to the age that older people can start to receive the the state pension.

Most civil service pensions are hardly excessive, although some 'high-fliers' who entered before career averaging was introduced in July 2007 do much better than most. The following table (extracted from Hutton's interim report) summarises all public sector pensions in payment in 2009-10. See the notes at the end of this web-page for information about the differences between the public sector pensions paid to men and women.

(Mean figures are total pensions divided by number of pensioners. The median is the figure above (and below) which half of pensions may be found. Mean is much higher than median where there is a relatively small number of high paid staff - and hence high pensions - with a long 'tail' of low paid staff, and hence lower pensions. There are around 300 public service pension schemes, but more than 95 per cent of members are in the six largest categories of scheme - local government, NHS, teachers, civil service, armed forces and police.)



There are big problems with both public and private sector pension schemes throughout the developed world, as people live longer and birth rates fall. In the UK, life expectancy at birth is increasing at an amazing one year every four years. It is estimated that the ratio of UK pensioners to workers will therefore increase from 27% in 2004 to 48% in 2050 - unless something changes. The inevitable consequence is that employers and/or their employees now need to set aside a much higher proportion of salaries to meet pension costs, or else employees have to retire much later, or a combination of both. Sharp falls in share prices around the beginning of this century added to the pressure on funded "final salary" schemes and so caused attention to be drawn to the special circumstances of the unfunded civil service pension scheme.

The problem for the Government is that many civil servants regard their pension as a hugely important part of their overall remuneration package. Senior Civil Servants, in particular, are paid substantially less than their private sector counterparts, and get none of their perks, such as company cars and private health insurance. Their pension really does matter to them. So the Government is faced with a large pension cost, and a workforce determined not to have their employment contract re-written. The old final salary pension scheme was accordingly closed to new entrants in 2007, but it remains to be seen whether the government will seek to reduce the cost of the pensions that remain payable to those who were employed as civil servants before that date.

It does not help that there are a number of misunderstandings about the civil service pension scheme. This note therefore seeks to inform the debate by providing further information about the scheme, and recent developments.

Historical Background

Pensions were introduced by employers - in particular in the armed forces and then the civil service - in order to facilitate the retirement of older less efficient workers so as to make way for younger fitter men. The very existence of such schemes then encouraged employers to recruit younger fitter people who would work for many years before drawing their pensions.

The first pensions appeared in the late 1600s in the public sector. Before then, it was the custom for serving naval, Customs officers etc. to sell their office for a lump sum or annuity. The first known civil service pension was awarded in 1684 when a senior Port of London official became too ill to carry on, and his successor was appointed on a salary of £80pa on condition that £40pa of this was paid to his predecessor. And so began the first 50% pension, paid out of the salary of a younger employee.

Significant further developments occurred in the 1760s and 70s when contributions and pensions were extended to junior officers in the armed services etc., and again in the 1847 when the Admiralty decided to face the fact that 200 senior captains were never going to sea again, promoted them to Rear Admiral, and put them on half pay as a form of retirement pension. Shortly afterwards, in the 1850s, Northcote and Trevelyan recommended that "good service pensions" should be extended into "the ordinary Civil branch of the public service" (Northcote Trevelyan Report - see in particular pp 21-22). A Royal Commission reporting shortly after Northcote & Trevelyan then separately recommended that retirement from the civil service should be possible at age 60 and compulsory at age 65. And there were parallel developments in the private sector, as railway, gas and other large companies developed similar schemes to attract and retain better staff.

A more detailed history of occupational pensions may be found in chapter 12 of Pat Thane's "Old Age in English History".

The pre-July 2007 Scheme

Those who joined the civil service in or before July 2007 remain members of the previous version of the Principal Civil Service Pension Scheme - "the PCSPS" - which had a number of interesting features:-

The scheme is unfunded, in the sense that there is no pension fund which is built up to pay future pensions. But it is funded in other ways:

The budgets of individual departments include employers' contributions to the PCSPS, and these contributions are now not only large in themselves, but have recently risen very quickly, up to an average of 19.4% in 2007. (Employer contributions are tiered, depending on salary paid. The details are in a document known as EPN 148.)

All civil service salaries are fixed by reference to those gross salaries that would be paid for similar work in the private sector, and then reduced to allow for notional pension contributions. This leads to the strange effect that, as pensions are calculated on civil servants' net salaries (after they are reduced by these notional contributions) the reduction is carried through to the pension eventually paid. Civil servants and their dependants therefore continue to contribute getting on towards 20% a year towards their pensions after retirement!

Many civil servants also have real pension contributions of up to 3.5% deducted from their salaries.
All these savings and contributions are pocketed by the Treasury and used to reduce current Government expenditure. Pensions are then paid out of taxation when they fall due. This arrangement has a number of advantages, for example in avoiding the kind of investment risk faced by funded schemes, but Ministers and officials are all too aware of the long term liability that has been built up. (The Government Actuary estimated that the unfunded liabilities of the civil service, teachers, NHS & emergency services schemes totalled around £770 billion as at 31 March 2008.) It is also possible that past contributions (notional or real) were in practice set at levels which were - at least with the benefit of hindsight - far too low. The same phenomenon can of course be found in the funded sector, but it appears that the private sector did move more quickly to address the issue, in particular by closing 'final salary schemes' - see further below.

It follows that, when comparing civil service pension contribution figures with contributions made by/for employees of other organisations, it is important to take account of both other employees' contributions (deducted from gross pay) and other employers' contributions (in addition to gross pay).

Pensions are inflation-proofed - that is they increase each year in line with prices (not wages). This feature is common in the funded sector, but those increases are not guaranteed if, for instance, inflation rises rapidly and/or the scheme runs into financial trouble. In contrast, the state's backing for the indexation of civil service pensions, which is in turn assisted by the absence of a specific pension fund, provides special security for civil service pensioners.

Note that the Chancellor announced in June 2010 that civil service pensions will in future increase each year in line with the less advantageous Consumer Prices Index (CPI), not the Retail Prices Index (RPI).
The pre-July 2007 scheme was a "final salary scheme" which means that pensions are initially set at a certain percentage of the final year's salary. For instance, a civil servant who opted to join the Premium Scheme (which means they pay 3.5% real contributions in addition to their employer's 19-22% notional contributions) would benefit from an increase in their pension of 1/60th of their salary for every year that they work - assuming they do not 'commute' any part of their pension into an up-front lump sum. This is equivalent to putting 1/60th of annual salary into a pension pot, and revalorising the pot each year in line with the employee's own salary. After 30 years, for instance, the pension would be 50% of final salary, assuming no lump sum. This means that 'high fliers' get a lot more back - in terms of pensions - than those on more modest earnings for the same amount of pension contributions. (Hutton noted that high flyers can receive almost twice as much in pension payments per £100 of employee contribution than low flyers.) This seems unfair unless - as is to a great extent the case - there is a counteracting factor in terms of 'high fliers' earning much lower salaries than their opposite numbers in the private sector:- see my separate note on civil service pay.

Final salary schemes used to be common in the private sector but have fallen out of favour as many firms switch either to "career average" schemes or to schemes which do not guarantee specific benefits but instead pay pensions by reference to the financial performance of the pension fund.

Civil servants who had worked for 40 years could typically retire, under the 'Classic' pension scheme, on a pension of one half of their final salary, and also take a lump sum of 1.5 times their final salary. Later variations of the scheme, called 'Classic Plus' and 'Premium' offered somewhat better benefits, including greater flexibility between pension and lump sum, in return for higher contribution rates.

The standard retirement age, for these schemes now closed to new entrants, is 60. Those who choose to retire earlier not only receive a lower pension because they will have worked for fewer years, and failed to receive their last few annual pay rises, but the pension will also be "actuarially reduced" because the pension is taken prematurely and so will be paid for longer. Those who retire after age 60 benefit from their further years (up to a maximum of 40) and from their annual pay rises, but there is no "actuarial enhancement" to allow for the fact that they will draw the pension for fewer years.

Was it a good scheme?

So how do these features stack up in practice? Well, it all depends on grade, and length of service.

Let's begin by comparing the pension of a civil servant in one of the lower grades with that of his/her equivalent in the private sector. Here, the main advantage is that civil service pay tends to increase with length of time in the job (even ignoring pay inflation), whereas many non-civil servants earn much the same just before they retire as when they are in their twenties. Other things being equal, therefore, many civil service employees will receive somewhat better pensions - based on their final salaries - than many in other final salary schemes. And of course the vast majority of private sector employees do not benefit from final salary schemes anyway. But it is important to remember that we are still talking about relatively low paid officials, so their pensions will hardly keep them in luxury. As noted above, the average pension was £7,600pa in 2009.

The balance of advantage is rather different for the 4000 or so in the Senior Civil Service (SCS). On the face of it, their pension looks very generous. These civil servants will typically have worked in the civil service for many years, and will earn much more towards the end of their careers, following several promotions, than than in their early years. So their low early contributions (whether notional or real) will buy very high eventual benefits. But - and it is a big "but" - these people regard their pensions as a very important part of their overall remuneration package, which is distinctly ungenerous compared with that of their private sector counterparts. (The median total Grade 3 remuneration package was in 1999 worth only 40% of that of private sector comparators. Follow this link for more detail.) SCS pensions are therefore a major retention factor, and a big reason to seek promotion, as well as a very attractive feature when the Government is seeking to meet its target of attracting 30% of the Senior Civil Service from outside its current ranks. Any diminution in the value of the pension will probably need to be matched by commensurate increases in base salaries.

It is worth noting, at this point, that the mandarin's pension also plays a large part in the implicit contract under which the official offers total loyalty to the Government of the day - and keeps his/her mouth shut! - in return for being well looked after in the long term, with good holidays and sickness and pension benefits etc. Many civil servants would no doubt happily trade their eventual pension benefits for the more immediate return of a better salary, and more job mobility, but is far from clear that Ministers would be better served by a less tied workforce.

Employees' contributions are sure to rise: 'Burden Sharing'

The Government announced in 2005, when introducing the new post-2007 pension scheme - see further below - that there would be a cap on any future increase in taxpayers' liabilities if costs were to rise beyond what was then anticipated. Initially any extra cost would be split 50:50 with employees, but with an absolute cap that might prevent the employers' contribution rising above 20 per cent of salary from the then current average of 19.4 per cent. Therefore, if the first actuarial valuation suggests that total contribution rates need to rise by 0.8%, employers' average contributions will rise from 19.4% to 19.8%, and employees' contributions will rise from zero to 0.4% (the Classic scheme) and from 3.5% to 3.9% (the Premium and post-2007 schemes). But any rise over 1.2% (that is 0/6% for employers and 0/6% for employees) would likely fall entirely on employees - depending on negotiation, and possible legal action, at that time. It was not expected that any adjustment to contribution rates would be needed until 2012 at the earliest, and much would then depend upon recent and forecast changes in life expectancy for the over 60s.

The Government subsequently announced, in December 2009, that the dire state of public sector finances meant that they intended to "cap the contribution of employers [to public service pensions], thereby limiting the liability of the taxpayer". It did not at first sight appear that this went any further, in the case of civil service pensions, than had been announced in 2005.

But the new Coalition Government mandated further contribution increases - 3.2% on average - from April 2012. Those earning up to £15,000 pa saw their contribution rates frozen - usually at 3.5%. But others saw increases, with the highest paid (over £60,000pa) now typically paying over 8.25%. This differential represents a significant additional income tax on the higher paid. (Employers' contributions averaged 18.9% in 2013.)

What changes apply to those who joined the civil service after July 2007?

Career Averaging

Post July 2007 entrants to the civil service have joined what is in effect a totally new pension scheme, although it remains unfunded or "pay-as-you-go" with the government pocketing all contributions.

The fundamental change is that pensions are no longer based on final salaries, but on career averages. In other words, pensions will be based on a proportion of the pay earned in each and every year of service, rather than on a proportion of the pay earned shortly before retirement. Employers will each year calculate an amount of pension based on that year's salary, and then freeze it. They will then increase that figure by inflation each year so that lots of little bits of pension will eventually make up the total pension.

Other things being equal, there have been clear "winners" and "losers" as a result of the move to career average pensions - see the following paragraphs. And, although the change is said to be cost neutral overall, this only applies in the short term. In the longer term, the cost will depend upon the way in which civil service salaries react to the fundamental changes in the pension scheme. Equally, the 2007 changes will likely lead to a significant alteration to the current implicit contract under which the more senior civil servants are offered postponed remuneration, in the form of a generous pension, as long as they remain employed by the civil service - and hence forced to remain loyal and discreet right through to their 60s.

The effect on individuals depends on changes to (a) the "accrual rate" - i.e. the proportion of pay that will each year become a frozen pension, (b) the figure used to increase the frozen pension each year, and (c) the individual's career profile. The Government announced in January 2007 that (a) the accrual rate will be a much more generous 2.3% of salary (compared with an effective 1.67% in the old "premium" scheme - see above). But (b) the annual revalorisation will be by reference to increases in the Consumer Prices Index, and so much less generous than under the old scheme, where pensions were increased in line with the recipient's salary. And then, so far as "c" is concerned, the change will benefit staff who stick in one grade throughout their careers, and/or come into the civil service for a short time and then leave. But it will disadvantage loyal "high fliers" who come in from school or university on low salaries, and leave many years later after achieving high salaries. It will also disadvantage women with children whose careers are put on hold for several months or years. (See also Note 1 below.)

The Government put it this way in a communication with staff: "While most workers will be asked to retire later and pay more towards their pension, at the same time, most low and middle earners working a full career will receive pension benefits at least as good, if not better, than they get now. Those less than ten years from their Normal Pension Age on 1 April 2012 will continue to be protected from these changes".

The change therefore sends a strong message to the next generation of young ambitious civil servants whose future rewards will be more closely related to career-average salary than to final salary. Future long-service high salary civil servants will no longer face retirement on one-half pension or more. Instead, their pensions, based mainly on their early/mid-career salaries, will be a relatively low proportion of their final salaries. This is bound to encourage them to leave in their forties or fifties if they feel they could build up a sizeable retirement fund out of much higher private sector salaries.

An alternative outcome would be for "high fliers" to be able to command higher salaries to compensate them for their reduced pensions, in which case the net cost of the new arrangements will be higher than the old scheme. This would particularly apply in those departments, such as the economic regulators, which are competing for staff with the private sector. But maybe other staff (those whose salaries remain relatively steady and whose pensions will - other things being equal - be higher than in the past) will find that they are expected to accept lower salaries in return for their improved pensions?


1. Male public service pensioners typically receive more than female pensioners. Hutton noted that the median male public sector pensioner receives just over £8,000 per annum, while the median female pensioner receives just under £4,000 per annum. This gap can be explained in part by a combination of more fragmented female careers (particularly connected to caring responsibilities), differential rates of part-time employment (there are currently more than seven times as many female part-time public service workers as male ones) and historic differences in careers and consequently in pensionable pay. More information about the issues facing women civil servants can be found here.

2. This ONS Consultation Document‌ explains why CPI is generally lower than RPI. This is not (as has often been reported) because one uses geometric averaging and one uses arithmetical averaging. The real reasons are much more complex and include the way in which clothing prices are tracked, and the fact that the RPI (unlike the CPI) includes housing costs but excludes purchases more often made by low and high income households.

This - from John Kay - neatly explains why there is no one 'correct' measure of inflation:

The stock market goes up and down. Suppose its level alternates each year between 50 and 100, with no upward or downward trend. What is the capital return on the market? Common sense suggests the answer is zero. But is that common sense right? In the good years you obtain a return of 100 per cent. In the bad, the yield is minus 50 per cent. The average return is therefore 25 per cent. There is logic to that. If you invested the same amount every year, and sold at the end of a year, you would indeed make - on average - a very attractive return of 25 per cent a year. What if you bought and sold at random? In that case, four options are equally likely: buy and sell at 50, buy and sell at 100, buy at 50 and sell at 100, buy at 100 and sell at 50. The overall annual expected gain is 12.5 per cent.

If you are by now thoroughly confused, you are not alone. The average historic return on the volatile equity market is central to calculations of the cost of capital and provision for future pension liabilities. But the figure has been debated for decades. The dispute is less about the underlying data than about the way you make the calculation. The question is often framed as the choice between arithmetic and geometric means. But there is no right or wrong answer. In all problems of this kind, the relevant measure is specific to the particular purpose you have in mind.

3. The 2011 Hutton Report provides a good summary of the issues surrounding public sector pensions.

4. Those looking for more detailed information, including about their own pensions, should try one or more of these options:

4. This is a complex subject and I would be glad to be corrected if I have misunderstood or over-simplified anything. But please note that I cannot answer questions about the civil service pension scheme, or help trace pensions. You should instead use the resources listed above.

Martin Stanley