The Chairman (Mr S. C. Yates, F.I.A.): Firstly, I will introduce the authors.
Rob Hammond chairs the Institute and Faculty’s working party on state pension age. He is a partner of First Actuarial and a scheme actuary with over 14 years of experience in the industry. Rob Hammond specialises in public service pension schemes and has advised a large number of trade unions during the recent reform of public service pensions, including the NHS Pension Scheme, the Teachers’ Pension Scheme, and the Local Government Pension Scheme. Public sector pension schemes have a link between their normal retirement age and state pension age, so the impact of state pension reform will be all the more important for them.
Mark Sadler on the working party is also from First Actuarial. He previously worked for Prudential and Hazell Carr and, after some years as a scheme actuary, he currently works on First Actuarial’s pensions technical team.
Steven Baxter provides strategic advice to insurers, pension schemes, and banks with a particular focus on managing longevity risk. He leads the longevity research programme at Hymans Robertson and was the architect of the longevity analytics used by Club Vita.
Steven has a keen interest in the equity and fairness of state pension changes and has been a regular commentator in the pensions press on the latest government proposals.
Sonel Mehta’s background is in pension consulting, having started out in the Government Actuary’s Department. She then moved on to PWC and KPMG. She now heads the actuarial resource team at Hazell Carr.
Mr R. J. Hammond, F.I.A. (introducing the paper): The State Pension Age Working Party was established in 2013. It was set up in 2013 and the purpose is to consider the state pension age, what factors should be taken into account when setting the state pension age, and what areas could and should the Institute and Faculty of Actuaries (IFoA) be involved in in the future? After we were set up, the government started to release information in this area in the Autumn Statement followed by more detail in December 2013. There were details about how the DWP foresaw how government could link state pension age in the future to future longevity. The Pensions Act 2014 introduced the requirement for state pension age to be reviewed periodically and we had a remit for what we should look at.
Two people who are not here but who were on the working party are Mr Richard Bramley and Mr Arti Kakkad to whom we extend our thanks. They were instrumental in the writing of the report and formulating the discussion. Equally we thank Mr Tim Sharples, who was assigned to us as a mentor for the Institute, and to the peer reviewers of the report.
I am going briefly to talk about the perceived need for change to the State Pension Age, SPA. I will then pass over to Steven Baxter who is going to talk about the DWP formula that GAD will adopt to carry out their review of SPA, look at some of the projections if it is carried out in the way laid out, projections of how SPA might change in the future, how that formula might be sensitive to some of the assumptions in the formula, and point out some of the possible limitations of the formula.
Mark Sadler will then talk us through the role that the panel will play. I will talk a little bit about that as there are two sides to the periodic review, one coming from the DWP formula, and another coming from an independent panel.
Then I will briefly touch on some of the alternative designs for state pensions and SPA that the working party considered. The remit of our report was mainly on the current system and how it is perceived to happen in reality.
At the end there will be time to have a debate on what we see and what you see as the role of the IFoA; what they could do and should do perhaps to influence the future SPA.
Our paper includes a summary of the role of the SPA and the history of state pensions. The role of the SPA is currently the earliest age at which a UK citizen can draw their state pension, but it also affects lots of other things: occupational pension schemes sometimes refer explicitly or implicitly to SPA. Individuals’ retirement behaviours are affected by their SPA. Lots of people think that they will retire when they reach SPA. There are also other benefits: for example, National Insurance contributions cease at SPA. It has a bigger role than just the age at which you get your state pension.
In terms of history, the first state pensions were paid in 1909 in the UK to around 490,000 people aged 70 or over who fell below a certain annual income eligibility, and so a means tested benefit paid only to the very old with a life expectancy in those days of about 5 years.
Through time the objective and the form of state pension have changed frequently in the UK. It has gone from being non-contributory to contributory, from means tested to a combination of means tested and non-means tested, and an element of earnings-related benefit; SPA has decreased from those early days of 70 and now obviously it is coming back up again.
There is lots going on in the pensions landscape, not just changes to SPA. One such change is the introduction of the flat rate state pension from April 2016. This will be a universal flat rate for everybody. This stems from the recommendation in the Pensions Commission report from 2005, in particular, the recommendations in the second report.
Pensions auto enrolment has kicked in now. The state is encouraging private pension savings by setting minimum contribution rates that employers and employees should pay.
Freedom and choice: the new pension freedoms effective April 2015 affect the pensions landscape. Two particular ways where it is related to state pension perhaps are that it could mean that individuals are able to retire before their SPA and use the new pension freedom to bridge the gap between retiring early and their SPA, particularly if the SPA increases.
On the flip side, it could also increase the reliance on state pensions if the new pensions freedoms result in individuals spending their private pensions before they die.
Although there are a lot of other things going on that are related to this topic, we are going to concentrate on the government’s intention to link SPA to longevity.
The Pensions Commission 2005 identified concerns with the existing pension arrangements and it picked out some key concerns.
The first one is complexity and uncertainty. How much will people get from the state pension and at what will be SPA? The concern was that the current state pension system was too complicated for individuals to understand. Secondly, there was concern that it might be inadequate for the lower earners but possibly too generous for the higher earners. The commission recommended that that was looked at.
Thirdly, disincentive for private pension savings: this is a criticism of the means testing of other benefits that people can apply for, perhaps meaning that people do not want to have their own private savings. A recommendation was to address that.
Finally, sustainability of the current state pension system. Would it be sustainable now and in the future to UK governments? Is there an element of inter-generational unfairness? The working population of today paying for the pensions for the current pensioners of the day?
Many of the recommendations were accepted and have been implemented in stages since 2007, the introduction of auto enrolment being one example and, the single-tier state pension being another example of trying to avoid disincentive in private savings.
We are looking at things like linking SPA to longevity, building on those initial recommendations.
The Pensions Act 2014 has introduced a structure for how SPA will be linked to longevity, and how it will be periodically reviewed. The structure effectively is that the Secretary of State for Work and Pensions is required to produce a written report. The first one must be produced by 7 May 2017, and then periodic reports must be carried out within 6 years of the previous report. The Secretary of State also has to commission two pieces of work. The formula which I referred to earlier, the DWP formula, will effectively be a review carried out by the GAD. They will look at whether a person reaching a SPA within a specified period can be expected to spend a specified proportion of their adult life in retirement using that formula; and, if not, they need to suggest how the SPA could be increased to mean that the formula is met.
On the other side, the Secretary of State will appoint a person or persons to act independently from the GAD report and prepare a report on the other factors that the Secretary of State should consider in deciding whether to increase SPA or not. These other factors are to be specified by the Secretary of State.
The aim is for these two reports to be equally balanced but it could well be that one area is more important than the other. We do not suggest at this point which one will be more important.
Back to those deadlines: 7 May 2017 will be when the first report by the Secretary of State must be produced, bearing in mind that these two previous reports have also got to be completed before then for the Secretary of State to consider. That means that there is not much time between now and then for those reports to be done and, indeed, for the Institute, if it wants to influence those reports or influence the Secretary of State on what factors to consider.
Mr S. D. Baxter, F.I.A. (introducing the paper): The formula proposed by the DWP as the means by which to link SPA to any longevity increases is simple. It looks at the number of years each birth cohort is expected to live beyond SPA as a proportion of the total years of adult life.
When this proportion is projected to exceed a certain level, then the SPA will be increased to ensure it remains below that level. As currently proposed, these increases in SPA will be phased in over 2 years, and everyone is expected to have 10 years’ notice of any changes in his or her SPA.
The threshold at which the SPA needs to rise has currently been set at 33.3% – or equivalently for every 2 years spent prior to SPA, 1 year will be spent in receipt of state pension. We can also think of this as for every year of increase in life expectancy that we are fortunate enough to enjoy in the future, 4 months of it will be spent in receipt of state pension.
A few points of actuarial detail: the life expectancies used in this formula are unisex on a cohort basis, and adult life is deemed to start at age 20, consistent with OECD conventions. The formula ignores any chance of death prior to reaching SPA.
Let’s start by looking at the anticipated progression of SPA under the formula. In the chart in Figure 6.1 of the paper, the grey line is the currently legislated SPA – although for simplicity prior to 2020 we have shown the SPA for men rather than muddy the picture with the current programme of equalisation. We can see the rise to 66 happening between 2018 and 2020, and then the rise to 67 between 2026 and 2028, as brought forward by the 2014 Pensions Act.
The blue line is what we can expect legislated SPA to be if it were set purely by reference to the DWP formula, and assuming that the life expectancies are calculated with reference to the current ONS Principal Projections for the UK.
The blue line starts in 2029. This purely reflects the practicalities of how soon the first increase could happen. Rob Hammond referred to the first reports to the Secretary of State being due in May 2017. By the time that we allow for the 10-year notice period for everyone and the 2-year phasing, it will be 2029 before we can realistically see any steep changes in SPA.
One thing that we can immediately see from this chart is that the formula accelerates the increase in SPA to 68 by some 10 years – to between 2034 and 2036.
Further, assuming no other demographic changes and no change in government policies, SPA would further increase to 69 between 2047 and 2049.
All of these changes are driven by the proportion of adult life being spent in receipt of state pension. Let us overlay that on the chart (Figure 6.2) as bars. The values corresponding to the bars are to be read off from the right-hand axis. For ease, we have shown as the red line, the DWP’s limit of 33.3%.
The grey bars correspond to the grey line, i.e., the currently legislated progression of SPA. They start in 2012 – the first year from which the current ONS projections can be used. We can see how these bars broadly average out at one-third.
In contrast, the blue bars show the impact of the one-third limits in the DWP formula. We can see how this leads to the upwards revisions in the SPA (given by the blue line).
As per the blue line, the blue bars start in 2029. They initially track the grey bars, as longevity increases. But come 2036, they would hit 33.3%, triggering a change in SPA from 67 to 68 between 2034 and 2036. As a consequence the blue bars drop down in 2034 and 2035 to a lower level. After that, the proportion rises, as longevity is projected to continue to increase. By 2049, the blue bars would reach 33.3% again, triggering a further increase in SPA to 69.
The key thing that I would like you to take away from the chart is that the one-third proportions is ultimately acting as a cap. We are never expecting to spend one-third of adult life in receipt of state pension; it is always going to be a little bit below that.
That leads us to a natural question of where does this one-third come from. And how sensitive is the progression of SPA to that choice?
The working party asked the DWP to clarify the reasoning behind the one-third. This was their response.
“This is broadly in line with the proportion of adult life spent in retirement over the last decade (assuming retirement age of 65 for men and women) and also consistent with the proportion of time people would have expected to spend in retirement when the original SPA timetable was set in 2007, given life expectancy projections at the time (2004 ONS principal cohort projections)”.
So, in effect, the one-third was the “run-rate” at the time when the changes to the SPA were first mooted.
This also enables a nice simplicity of communication by policymakers. It is rather neat to be able to refer to 1 year in receipt of the state pension for every 2 years working; and I have heard Steve Webb make reference to this at a number of events.
Importantly, the actual proportion is subject to specification by the government of the day. There is no certainty that it will remain at 33.3%; and, ultimately, the progression of SPA is sensitive to that choice.
In the chart (Figure 6.3 of our paper), we have chosen to nudge up and nudge down the limit to the proportion of adult life spent in receipt of state pension very slightly. The dotted green lines are if we nudge it down to 31%. The dotted purple lines are if we nudge it up to 35%.
Starting with the green lines; these are looking at a lower cap to the proportion of adult life being spent in receipt of state pension. Since that cap will be breached sooner, changes to SPA are brought forward. When the first review is carried out in the next year or so, we would expect it to identify that the SPA would need to increase to 69 in at the earliest possible opportunity. In practice, to avoid birthday lotteries, we suspect that this would be implemented as a series of changes between 2026 and 2032, resulting in a period of rapid increase in SPA. We can also see how the SPA then goes on to increase to 70 in 2039 and reaches 71 by the 2050s.
In contrast, a small increase to the maximum proportion of adult life to be spent in receipt of state pension to 35%, delays the point at which SPA needs to change for the first time from 66 to 67. Indeed, we could effectively unwind the 2014 Pensions Act changes. Whether or not the Secretary of State would repeal that legislation is unclear, but it does highlight the sensitivities.
What is the right proportion? There is no real magic answer to this. But as we go into this new regime, it is important for us to be aware of how changes to what may appear to be small details of the formula dramatically change the output.
The formula also relies heavily on the longevity projections used. In the chart in Figure 6.4 in our paper, we have illustrated the sensitivity of the formula to the longevity projections by using two alternative ONS projections for the UK population. The dotted purple lines use the low life expectancy projections. The dotted green lines use the high life expectancy projections.
The ONS’s projections, are very similar to the CMI model in terms of starting with the current rates of improvement and blending these to a long-term rate. For the projections shown, the central principal projections (the blue lines) are effectively on a long-term rate of 1.2% p.a. The lower life expectancy projections (the lower purple lines) are based on a 0% p.a. long-term rate. The high life expectancy projections (the top green lines) use a 2.4% p.a. long-term rate.
The results of the formula are very sensitive to the choice of projection. If we look at the extremes, we either retain a state pension age of 65 or, alternatively, SPA needs to reach 74 by 2060, a spread of 9 years. Ultimately, the formula is reliant on cohort life expectancies for individuals reaching 65 in future years and decades. By virtue of using cohort life expectancies, it is particularly sensitive to the most subjective part of longevity projections – the long-term rate.
One thing we do know is that actuarial projections of longevity are subject to revision in light of new data and new methodologies. They have changed in the past and undoubtedly will change again in the future. With that in mind, we asked ourselves how robust the formula would have been to historical changes in longevity projections. If we had implemented this some 20 or 30 years ago, what would the public have seen in terms of changes to SPA?
Unfortunately, the data is not available to do this based on the historical ONS projections, but we have been able to do it using the prevailing actuarial projections of the day. So for the moment, let us imagine we are in 1980, and the formula is just being implemented.
In this context, 1980 is perhaps a slightly ironic choice. It was the early days of Margaret Thatcher’s government and the then Chancellor of the Exchequer, Geoffrey Howe, was suggesting to “the Iron Lady” that the finances of the country were not great and it would be rather good if we could reduce the state pension. State records show a conversation to the effect of “Let’s try and take 3% off the state pension this year because we could really do with the money in the public coffers”. Margaret Thatcher overruled Geoffrey Howe though. She did not want to be seen to be going back on a televised personal pledge. There was no way she was going to reduce the state pension.
If we go back to 1980, we also need to make some assumptions. The first is that we have assumed that SPA has already been equalised at 65. The second key assumption that we are going to make is that the then Secretary of State, Patrick Jenkin, would have looked at the current run-rate for the proportion of adult life spent in receipt of SPA and used that in the formula. So instead of the 33.3%, we base our calculations on the 28% that was prevailing at the time.
The results of our calculations are set out in Figure 6.5 of our paper. The key thing to draw your attention to is the top red line. This is the progression of SPA which we would have seen as a consequence of the series of successive changes in longevity projections.
It arises from a series of changes. For example, starting with the bottom dotted blue line; this is the 1980 series projections. They would have required a single increase in SPA to 66 in the early 2000s, following their publication in the early 1990s. The dotted lines above this show the succession of other projections that we have seen since.
But rather than trying to follow each of those through, it is easiest to follow this chart if we look at an example individual.
Let’s consider a young person, aged 20 in 1980. He or she would have been anticipating a SPA of 65 that is to start receiving their state pension in 2025.
Ten years into their job, in 1990, the 80 series projections were published – this would have nudged his or her SPA up to 66. That alone is perhaps not too painful.
But then the 92 series projections were published in 1999 – this would have resulted in his or her SPA leaping up to 69.
The publication of the cohort effect around 2002 would have caused another year to be added to SPA. The adoption of underpins in our projections around 2007 – another year on to SPA. And most recently the CMI model, typically used with higher long-term rates than the preceding underpins, would have nudged the SPA up yet another year.
Thus, in just over 20 years between 1990 and 2012, this person has been told of five separate increases to his or her SPA, and receipt of state pension has been delayed 7 years until 2032.
Further, there is still time for at least two more SPA reviews before he or she is within the 10 years of SPA when they can have certainty on when state pension will commence.
For “Joe Public”, these successive changes would have done little to instil confidence in the state pension system.
So the formula, whilst simple, does have some limitations. It is highly sensitive to a number of factors including the target proportion chosen, the projections used, and changes in view as to what is the appropriate longevity projection.
Relying on the formula could lead to successive material changes in SPA, which for younger workers could dent their confidence both in the state pension and pensions saving more generally, unless carefully communicated. Indeed, the communication aspect may well be one of the factors which the panel considers as a balance to the formula output.
Whilst the formula is simple, it does overlook a number of important aspects. I mentioned at the start that it ignores death before SPA. To put that into context, consider the generation reaching 65 today. Around 80% of those alive aged 20 in 1970 have survived from age 20 to 65 now; 20% of that generation have not reached 65.
The formula does not factor in affordability or fairness constraints, nor does it consider the wide variations in life expectancy which we see across society.
This leaves plenty for the panel to consider in their report.
Mr M. C. Sadler, F.I.A. (introducing the paper): Steven has covered the formula, and that is what we are expecting that the GAD report will be producing and considering. In addition to the GAD report, there is a second report that gets commissioned, which is a report produced by a panel, single person or persons, which considers factors beyond just that formula.
There are various steps to that. The Secretary of State first of all decides what factors the panel needs to consider. The panel considers it and then they get fed into the Secretary of State when he is setting the final results of what the SPA will be in the future. The profession’s point of view is there may be a role both in trying to influence the Secretary of State on what factors he wants to specify for that report, and then either being directly involved or producing some of the evidence for that panel to consider going forward.
The other thing that is important to say is that the report is considered entirely independently of the GAD report. So at the same time as the GAD is doing their report, this report gets considered.
There are many things they could consider, but within our report we break it down into three distinct areas: the first one is healthy life expectancy and ability to work; the second is variation in life expectancy by different categories of people; and, finally, financial considerations.
The first bit we looked at was the healthy life expectancy and working at older ages. The blue lines at the top of Figure 7.1 in the paper show life expectancy from birth, which progress as we go forward. A reasonably steady improvement of people living longer.
The grey bars are healthy life expectancy. Healthy life expectancy is a self-assessed measure. There is some element of lots of different ways of measuring this but at least it gives us some influence. We can see that is improving over time. Actually, it does not improve as fast as life expectancy improves. Healthy life expectancy is going up by about 0.8 years for each year that life expectancy improves. The gap between them in terms of number of years is gradually widening.
The other point to note on this chart is that healthy life expectancy is never actually getting up even to current SPA. At the moment there are people who are not healthy by the time that they reach SPA. These people may already be drawing some type of disability benefit or even just unemployment benefit. So any increase in SPA for those people may actually just result in shifting costs from state pension to disability. One element that the panel will need to consider is maybe the SPA should go up, but actually are you just moving costs from one area of government to another and what do those total costs look like?
There are then some follow-on questions from this. The first question is: Do you have an ability to work in later life? For some people, as life expectancy goes up, you can continue to work. You should be able to do your job as you go forward. There are some professions which have doubts about their ability to work as they get older. Firemen and nurses have been campaigning to stop their retirement ages going up. Their argument is that you cannot keep doing the job as you get older. So if state pension is increased, there becomes a gap. Some people may not be able to do what they currently do, but is there going to be an alternative job for these people to do or can they actually extend current life within their current job? It will be the job of the panel to decide the balance on this. It may be that if state pension is increased, the government needs to find some role that these people can perform.
Then finally within this is voluntary work. Around 1.3 million people in retirement are estimated to provide some type of care, be that to a loved one, it could be their partner who may already be disabled or to grandchildren. If you start to increase SPA and people continue to work, then they may not be able to do some of these other unpaid supporting roles and the State may have to step in and take over. So it is not just a question of SPA increasing and saving some money. You end up with the possibility of spending money elsewhere.
The next big question is variation in life expectancy. There are lots of variations including regional variations and socio-economic class. We have picked a socio-economic class one partly because it was on the DWP’s list.
This one simply plots the improvement in life expectancy over the years for different socio-economic classes. Higher management and professionals are the top class as shown in the chart Figure 7.2 of the paper. Their life expectancy is steadily improving. At the end of the chart it is well over 80 but with routine occupations it is down at the bottom. Theirs is improving but the actual amount by which they are improving is still a long way below the senior management. So towards the end of the chart they are just over 75.
The key point is that the number of years people are spending in retirement varies enormously. So for people in routine occupations, they have perhaps 10 years of retirement, if they are retiring at 65. For those in senior management or professional class you may be looking at 16 years. Just a single year of increase for people in a routine class is perhaps 10% of their post-retirement years that they have lost on this. The higher management professional does not make anywhere near as big an impact.
When we see that number as 33%, for some people the figures are very different. For those routine occupations, it is going to be a lot less than 33% of their lives that they are spending in retirement. For higher managerial they are getting a lot more and the impact of any change is not as much.
It will be for the panel to decide whether some people are being unduly penalised and, in particular, when you look at these are those people in routine occupations being treated unfairly in any change, and should there be a different approach, or should the increase be reined back a bit?
There are others. Life expectancy changes by region; but actually when you look at the report there is a table for London, and within London healthy life expectancy varies by 14 years. So even a simple figure, where perhaps we might do something for a particular region, does not work that well because it masks the variations within the region. There are many people who will not achieve the average life expectancy and there are other people who will be above it. We should not lose sight of those people who fall a long way below the average and have an even bigger impact in the change.
Beyond that, issues arise on affordability, sustainability, and these sorts of issues. Some of these issues are political but it is worth looking at the things that the panel could consider. The first one is the actual cost as a percentage of GDP. Figure 4.3 in the paper from the 2011 Office for Budget Responsibility report and shows that state pension, as a percentage of GDP, was expected to rise. This is based on reaching 67 in 2036 and 68 in 2046.
What it does show is that state pension as a percentage of GDP was expected to increase and is still expected to increase; a slight dip to 2020 and then a big, steady rise up from 5.5% up to nearly 8%. It is for the panel to decide how much more GDP will be spent.
The chart set out in Figure 7.4 of the paper shows a slightly more up-to-date chart from DWP figures in 2014. Again, we can see that the percentage of GDP is increasing. Cost is still similar to the OBR as far as 2033.
In a similar vein to that, we also have the sustainability issue as set out below.
This one covers inter-generational fairness. What it shows is the dependency ratio; the number of pensioners to the number of people of working age in the population.
On the left-hand side, the line shows the observed figures. It is relatively stable from 1970 up to about 2011 at around 300. Then we can see that after 2011, if there was no change and state retirement age had remained at 60 and 65, that the dependency ratio really began to increase, perhaps reaching around 500 by 2051.
At that point you effectively have only two workers for each pensioner. At the moment it is around three workers for each pensioner. This is one of those big questions about inter-generational fairness. How fair is it to make any worker support more and more pensioners as you go forward?
So the government made some changes up to and including the Pensions Act 2011. That is the equalisation at 66 in 2020. That has made a difference. It still rises by a little bit but it is a lot closer to that 300 going forward. Any change they make as a result of the panel and this new formula is likely to bring forward the next increase to 68 in about 2036.
But it is for the panel to decide by how much and what they consider fairness is.
The panel’s conclusions are likely to be that there is one size but it might not fit all people. For some people it is going to have a very different impact from others. The panel’s decision will ask “how fair is it?”. Is it going to be fair for everybody generally or do some people get unduly penalised? Steven Baxter has shown that the actual projections vary and it is highly sensitive to the DWP formula. It is for the panel to decide how they are going to react to that. The problem is that the GAD report is going to be produced alongside this report. It will be quite hard for the panel to do it. They can comment on this sensitivity; but without actually seeing the final numbers, will it be possible for them to decide on what action should be taken?
Finally, the panel have to decide if SPA is to increase, will there be questions about whether people will want to work longer and will they be able to work longer? You can increase SPA but are you just going to defer your costs elsewhere and are people going to follow it?
The other question for the panel is, if you are increasing the SPA, will employers still want to employ those people. That is for the panel to conclude and that will sit alongside the GAD report.
I am now going to hand over to Rob Hammond who is going to look at some of the alternative methods that could be used for state pension.
Mr Hammond: I am focussing on alternative designs: the Pensions Commission’s concerns are complexity and uncertainty, inadequate and overgenerous perhaps to some, disincentives for private pension saving, and then sustainability and affordability and inter-generational unfairness.
As part of our report we wanted to look at some alternative designs that government could consider to address some of those concerns of the commission, but also some of the limitations of the current and the proposed state pension system. We looked at these only at a fairly high level. We did not carry out any detailed analysis on the alternatives. That is perhaps an area where you think the IFoA should carry out some further research.
If we take the first one, and this touches on what Mark Sadler has highlighted, does one size really fit all? It will be unfair to those with the shorter life expectancies. The changes will have a much bigger impact on some people than others if SPA is increased. What we might find is that those people most in need of state pensions will be the ones that are disproportionately affected.
That is an issue that we highlighted and perhaps that the proposed system will not address.
What are possible things that could be done and how could they be addressed? There was a report carried out by PricewaterhouseCoopers (PwC) in 2014 that looked at allowing a SPA window. This is the idea that, instead of having a SPA that is fixed where you can retire only at that age or later, you could take your state pension earlier or later than that age.
The SPA at the moment is not fixed completely, because people can defer the age at which they take the state pension, but they cannot take it any age earlier. PwC suggested some kind of pivotal age that would be linked to longevity. You would set an age at which you could retire before that pivotal age or above that pivotal age. As the pivotal age moved, let us say it went from, say, 65 to 66, the individual who was hoping and planning to retire at 65, could still do so; they would just have their SPA reduced by 1 year.
Other countries that do offer early retirement before SPA are actually moving away from that. Denmark is one recent example where they have got rid of the ability to retire early from SPA. Perhaps PwC’s suggestion of a state pension window is therefore out of kilter with the rest of Europe.
Other ways would be to vary SPA by region or occupation or earnings meaning different SPAs for different parts of the country. There might be some element of devolving setting SPA to different regional governments. Or you could even vary SPA by the number of years that an individual has worked. So there are different ways that the pension age could be set that are not in the proposed system. Or, rather than the actual SPA, we thought you could even look at varying the actual pension amounts; perhaps those with shorter life expectancies might get more than those with higher life expectancies.
With all these changes, and indeed with the proposed system, there are limitations to any of these alternative designs. Limitations to the ones that I have suggested here are communicating the SPA and indeed retirement planning. Introducing some kind of SPA window introduces more complexity which goes against the aim of making it simple to understand. How do you pick the SPA if it varies by region and occupation and earnings? If somebody throughout their career changes where they live, changes their occupation, changes their salary, do you need to come up with some kind of career average SPA? Again, more complexity and harder again for the individuals to understand.
There are also perhaps some selection effects. The individual could actually move just before retirement to an area that has a lower life expectancy just to get the higher SPA. Again, how do you deal with career breaks, maternity, adult education, and that kind of thing?
Another issue that we thought we could look at is targeting the needy. Do the proposed changes to state pension really target the needy? Is a single-tier state pension going to be enough for those who really need it? And is it still too much for those who do not need it?
What we thought you could do here is to vary the pension amounts. Or you could reintroduce some degree of means testing. There are two ways of introducing means testing. It could be that individuals have to claim means tested benefit, which is the way that most people think about means tested benefits; or it could be that the benefit is provided to everybody and then the means tested benefit is systematically removed if somebody’s income or assets are above a certain level. So introducing it in that way perhaps means that some of the criticisms of means testing, whereby some people might feel embarrassed about claiming it or are not be able to fill in the form might facilitate that the right people get the right benefit.
Other ways are to work out how we can get some more income to provide state pensions. Mark Sadler touched on the increase of GDP that is paid on pensions. So further income examples might be taxing or having higher tax on pensioners or having national insurance contributions continuing after SPA. After all, national insurance was brought in not just for pensions but for NHS provision. Pensioners use the NHS as much as, if not more than, people who are under SPA.
Or you could introduce different SPAs for the better off or even for people starting their working life later. They could perhaps have their SPA from a later date. The theory there is that somebody starting their working life later might have gone to university and may also have a higher life expectancy if they then go on to have a better occupation and a better lifestyle. So you could have different SPAs based on when you actually start in the working population.
But there are challenges such as communication. Anything that is more complicated than the proposal goes against the simplicity aim. Means testing does not work if people will not actually claim, which is politically disliked by some parties. It is difficult to sell higher taxes on pensioners. Also integration with other state benefits may actually lessen some of these issues of targeting the needy.
Sustainability and affordability are concerns of the Pensions Commission. It is not so much the individual that we are worried about, where we were targeting the needy; it is more about the actual cost of the benefits. Things that could be done do include increase income by increasing national insurance for those working above SPA, higher tax, means test some of the other state benefits – perhaps winter fuel allowance should not be paid to everybody but only to those that actually need it. Or you could even reduce the amount of state pension over time as auto enrolment starts to have a bigger impact so that you do not actually have to change SPA but just change the amount of state pension.
The key thing is: Will it be enough if we do any of these things? Will it be enough to sustain state pension and make it affordable for a UK government?
Some other alternatives that we considered, and just to highlight again that the current regime and the proposed regime may well change in the medium to long term if different parties get into power. Different political parties have different views on state pensions. Also to mention that I said you could increase tax on pensioners or introduce national insurance for those above SPA, but there will be political reasons why that might not happen.
We could rip up the rules book and start again. Let us have a completely different state pensions system from what has been proposed. One example would be an individual state pension calculated for every individual. Some kind of a notional fund based on defined contributions, what that individual had actually paid in throughout their career. This is used in Italy, Poland, and Sweden so there are precedents for this kind of state pension provision. Again, would it target the needy and how will transitional arrangements work if we do completely rip up the current system and move to a new system? If we have a radically different system, how would that be brought in? How will individuals understand what is going on?
We might see that auto enrolment makes a positive impact on individuals’ retirement savings. These are some more radical things that could be done on state pension: we have only highlighted a few and we have not gone into any detailed analysis.
Finally, we wanted to think about what the role of the IFoA is in setting the SPA. What should the IFoA be doing or what could the IFoA be doing? For example, the IFoA, in our view, should be influencing the Secretary of State when specifying the assumptions for the Government Actuary’s report and the assumptions used in the DWP formula. We could also influence the Secretary of State in setting the factors that the panel should consider. We could influence the Government Actuary when projecting life expectancies – I do not mean tell him what the answer is but I mean consider the detail of what the mortality assumptions might be. Again, we could influence the panel by being on the panel. Perhaps the profession should be on that panel.
We prepared a list of what we thought the IFoA could do.
1. We should be independently checking the GAD’s results.
2. Provide a critique of the underlying longevity assumptions. Steven Baxter showed how sensitive the formula is to the assumptions.
3. We could lobby the Secretary of State on factors and seek representation on the panel.
4. And taking a wider responsibility as a profession by letting the wider population and insurance markets and occupational pension schemes understand the potential impact of SPA proposals, and to help generally with communication of SPA to the UK population.
Finally, the profession should build on the research that we have done and carry out some more research.
The Chairman (opening the discussion): May I ask the speakers how they feel about the 10-year notice period, particularly with longevity moving on as quickly as it has done? Do they view that 10-year lead-in period as too long, too restrictive?
Mr Baxter (responding): As a group we had a variety of views as to what might be appropriate. Generally, we thought 10 years was probably close to the minimum that could be used and give people a reasonable chance to adjust their plans for retirement. But this does mean that we lock people in to a SPA, and in turn our cost of providing the SPA, some 12 years in advance of each generation reaching SPA.
We did a survey amongst the profession at the pensions conference last year. A wide range of views was expressed by our peers (see Appendix of the paper), ranging from 5 years being the shortest acceptable notice period through to some people suggesting we should allow 15 or 20 years’ notice in order to adequately adjust retirement plans. So, no single right answer we can all agree on, but as a working party we were fairly comfortable with 10 years as the broadly minimum practical solution.
Ms S. Mehta, A.I.A.: My concern with the 10 years is, as was shown in Steven Baxter’s presentation, by the time someone starts working it is possible that they will have several changes to their SPA before they start receiving the benefits. That is not going to instil confidence in pensions.
Mr A. C. Martin, F.F.A.: I pick up on almost the final point in terms of issuing guidance or challenging debate. That is quite simply extrapolating the excellent paper to occupational pension schemes.
If UK society is happy to move the goalposts on state pensions, then they should equally be happy to move the goalposts in respect of defined benefit promises from employers. As most here will know, defined benefit promises are disappearing rapidly in the private sector. Putting private pension schemes on a similar footing to state pensions, that is scrapping Section 67, would be an excellent challenge to put to politicians.
Just to be helpful, in a similar way to the timing in changes on advanced corporation tax and changes on indexation RPI to CPI, could I suggest the politicians might make this change just after the next general election?
Mr Hammond: As a working party when we first started we wanted to be very careful and not accidentally fall into any kind of political area. You have made some good points there. Perhaps some more flexibility could be introduced to defined benefit occupational pension schemes that could make them more sustainable in the future.
Mr Sadler: The proposals for defined ambition are designed to pick up that element of, say, varying the retirement age according to state pension.
But I do have a bit of sympathy with the DB schemes which cannot vary their pension age. Once they have done it, it is locked. I am not sure I would go so far as scrapping Section 67, but maybe some flexibility in there does not seem unreasonable. The rest of the panel may disagree.
Mr R. K. Sloan, F.F.A.: I am a retired pensioner of 7 or 8 years’ duration and probably fall into the higher socio-economic group.
I would like to pick up on a few points that I thought were very interesting. You talked about the possibility of varying pensions by socio-economic group. But I would have thought the fact that the higher earners had contributed more by way of national insurance contributions probably balanced that one out, so that the current situation is not too bad from that standpoint.
Next, if you were to try to implement different SPAs for different socio-economic groups, that would be totally divisive and pretty much a non-starter, as well as too difficult to implement.
Coming on to the final issues including the role of the IFoA, many of the alternative structures that you briefly looked at in effect described SERPS, career-averaged revalued pensions, which are really a jolly good idea and always have been a good idea. One problem was how to fit that in with the minimum subsistence level of benefits. And I hope I am not speaking out of turn here but I have to say that I do not think the profession covered itself in too much glory back in 1974–1978 when SERPS was coming in. There was far too much lobbying for contracting-out rather than accepting the good common sense of the two-tier state pension system. Certainly most of my clients contracted in to SERPS and sat on top of that. But that is by the way.
What I would like to say now is a big “yes” to the point about lobbying the Secretary of State, although I do not know to what extent it is our role to do that. What I would like to see is the result of your research coming down to some positive suggestions, not a sole suggestion, but a number of possible alternatives that could be put forward and lobbied for. We have been too silent on many of the changes that have been imposed on us in recent years. We certainly moan about it afterwards, as we have to cope with all its complexities.
So, I think that we need to be more proactive. I am not sure how far we can take that. But if we can come up with a positive set of potential alternatives that are well-grounded in our research, I would like to see the lobbying idea being advanced.
Mr Hammond: The working party agrees that the profession needs to be playing a much bigger role in being seen as a profession that the government should seek their views on for these kinds of things. We have highlighted some things that could be done. We could lobby the Secretary of State on the factors for the panel to consider. We also wondered whether the profession should be on that panel, the reason being that actuaries have a very complementary skillset. We know about mortality; we know about economics; we like to think that we understand about pensions. Perhaps actuaries are well-placed to be on that panel.
Ms Mehta: I agree varying it by socio-economic class will not be all that practical. However, how you vary the age really depends on what you see as the purpose. You did mention that those who are in the highest socio-economic class will have paid more into the system. There is, however, an upper earnings limit that is used to calculate what each person will pay in, which I do not think allows for the fact that the more you earn, the longer you are going to live and therefore the longer you are going to receive the state pension.
If you see the purpose as delayed remuneration, then perhaps saying let us not have any variants at all is fine; but if you do see it as targeting the needy, then you need to decide whether you will allow for the fact that some people who will have been working from the age of 16 will still have to wait just as long to start receiving state pension, and they are the ones who, on average, are going to die earlier.
My concern with the whole state pension system is how it integrates with other welfare benefits. I am not sure whether the government is just changing this to then push the bill to something else like pension credit.
Mr M. A. Potter, F.I.A.: I echo Ronnie Sloan’s point about the profession speaking out on these issues. That is an excellent idea and we should continue doing so. I can give you an example. At this time last year we had our panel debate on the Scottish independence referendum. Shortly before that I had given evidence at a Westminster Select Committee on some of the Scottish government’s proposals, which at the time you might recall included the ability for Scottish governments to set their own SPA, the argument being that of course we in Scotland do not all have the health and wealth that Ronnie Sloan has, and should have lower SPAs.
I found myself with the enjoyable task of explaining dependency ratios to MPs. So there are some challenges there for the IFoA to continue to work on.
One of the explanations I gave at the time was a dependency ratio has a top and a bottom; you will all appreciate how a pay-as-you-go system works and you need to think about the top and the bottom. That is where the politics comes in.
It struck me that all this science and factors around the target proportion of a third can quickly be undone by some politics on who is eligible for it in the first place, what levels of national insurance they pay or indeed what state pension they get at the end of it. I would encourage the IFoA to get involved in all those aspects rather than just one ratio of it.
Mr Hammond: I agree fully that the IFoA should do more, can do more, and is well-placed to do more. Where the politics come into it was shown at the start of my presentation with the balancing scales with the DWP formula on one side and the independent panel reports on the other. It remains to be seen which of those two reports the Secretary of State believes is the most important. You may have your own views on which one is the most important.
Ultimately, it might be that the Secretary of State ignores both of them and just decides that the cost of not increasing SPA outweighs the cost of leaving it where it is to the extent that it has to be increased.
Ms Mehta: By the time that many of us actually reach retirement, we will probably see several more overhauls of the state pension system, depending on who the government of the day is.
Mrs S. G. Mathieson, F.F.A.: I am a member of the IFoA executive and am involved in research. Some of the points you raised about what else the IFOA could do are ultimately dependent on having the evidence base and having the research to be able to go out and inform the public affairs debates.
What does the panel see as the key priorities in terms of research under this particular heading for the IFoA?
Are there particular areas that you think the IFoA should be researching and prioritising for SPA? Is it about the alternative models or is it about perhaps the longevity assumptions, for example?
Mr Hammond: Coming back to what Ronnie Sloane was saying, he was suggesting that we should be looking more at alternatives.
When we started out on this project we gave alternative options the back seat. We concentrated on what was actually proposed going forwards, which is where we have concentrated our research.
It is interesting that more than one of you have already suggested that further research should be into the alternatives rather than just looking at what has been proposed. Perhaps that is something that would be the next research project.
Mr Sadler: I see it in two parts. We have these reports coming up that have to be in by May 2017. There is perhaps a role for the profession on some parts of that, perhaps looking at the assumptions in the formula or helping to develop better models of healthy life expectancy. I see that as something we should be doing now because those reports are due in the next couple of years. Then perhaps once those have gone through, people will be more willing to look at alternatives.
I do not see that this is an end here. There is an immediate job to decide which parts we can do with some of the immediate models and then a longer term looking at those alternatives.
Mr Baxter: I echo Mark’s comments. When the panel reports, and the Secretary of State weighs up the evidence, a number of issues will come to the fore. It would be very unfortunate if we found ourselves in the position at that point in time where, as a society, we do not have credible alternatives that address any issues the panel raises. We feel that, as a profession, actuaries are uniquely placed to take a holistic view and identify, given the prevailing political will and perceived role of the state pension, which alternatives might best fit. For example, depending whether or not the government of the day feels the state pension is a safety net, or a universal benefit, very much influences the suitable alternatives.
Gearing ourselves up as a profession to be on hand to offer credible alternatives that reflect the holistic pensions framework that Ronnie Sloan was alluding to is something that we should be doing.
The Chairman (closing): I should like to thank all those who participated in the discussion and invite you all to thank the panel for their excellent presentation.
