Friday, 4 May 2012

Time and money

I was reading a piece earlier today on what could be done about youth unemployment, which incidentally rehearsed the hoary old lump of labour fallacy fallacy (so good they named it twice). This got me to thinking about how the benefits of retirement (i.e. not having to work) and pensions (i.e. not having to starve) are allocated across society.

There was some discussion last year about the class-bias of a universal pension age, specifically around two socio-economic factors. The first is that many manual workers will have started working at 15, and can expect to make NI contributions for 50 years, while many non-manual workers will not have started working till 21 or later (i.e. post-further education). The second feature is the difference in longevity between the classes, which means that the number of years you can hope to enjoy beyond a universal retirement age will differ. According to the Office of National Statistics, this difference amounts to about 5 years for men, though analysis by region shows that this average masks differences of 10 years or more.

Using the ONS data (2006 actuals and trend growth from 1982 onwards), I created a quick model to show the impact of the proposed increase in the state retirement age. To remind you, the plan is for the age to increase from 65 to 67 in 2026/8, and to 68 in 2044/6. I further included a more pessimistic scenario (for which there already appears to be a  growing lobby) that sees the latter increase revised up to 70.

What this shows is that the current retirement age of 65, if unchanged (scenario A), would be progressive in terms of the relative gain in retirement years for the bottom of the earnings scale compared to the top. Though the well-off gain more years in retirement due to increased longevity, they're starting from a higher base. The poor see larger gains because their average longevity is closer to the state pension age. What this highlights is the original expectation, dating from Lloyd George's 1908 Pensions Act, that the beneficiaries of state pensions would be few and short-lived. The Act provided for people over 70 at a time when 63% of the population died before 60 (pg 5). The pension was seen as insurance in case you had the "misfortune" to live on into old age, rather than as an entitlement that most could expect to enjoy. Increased longevity means that most people now clear that hurdle.

Interestingly, the planned increases in the retirement age (scenario B) result in a relative growth of retirement years that is the same for the top and the bottom, though again the absolute benefit is greater for the well-off. If the retirement age is pushed further, as some are already advocating, then we end up in a situation (scenario C) where the better-off gain both in absolute and relative terms. In fact, the years in retirement then start to decline at the bottom of the scale.

This highlights the fact that we are looking at two distinct problems here. To date, the focus has been on the pressure that increased longevity will create on pensions. This is often expressed in the hopeful phrase "we're all living longer". The second point, which pension reformers and government have been less vocal about, is that some of us are living longer than others and that this disparity is getting worse. In other words, the second issue is widening inequality.

If longevity continues to increase (and there's every reason to believe that it will), and this happens in parallel with a widening gap in living standards, reflected in bigger differences in longevity by social class, then there is a good chance that the rate of increase of the retirement age, which is universal and therefore a notional average, may be greater than the rate of increase in the longevity of the poorest. In simple terms, the historic growth in the number of years in retirement for the poorest, which has benefited from the male retirement age staying static at 65 since 1925 as longevity increased, will go into reverse.

De facto this means a transfer of wealth (in terms of pension payments) from the poor to the rich. On the face of it, this is pretty repellent, however it's worth noting the degree to which the contributory principle has become diluted over time, and may well be finessed out of existence in the coming years. Consequently, the idea that you are "owed" more may become difficult to establish.


  1. It seems inevitable to me that over time, the retirement age may even need to rise higher, to perhaps 67-68. Ironically enough, this whole pensions issue is actually a backdoor argument in favor of immigration. Immigrants tend to be younger, and as the native population increasingly ages, this may be part of the solution to the pensions issue.

    1. You're right. There is an irony here that the best defence against the increase in the retirement age that disgruntled middle class voters could come up with is support for increased immigration. Terry and June should be demanding fewer resources for the UK Border Agency, not flirting with UKIP.

      On a more serious note, increased immigration is just a stiking plaster as those young immigrants in turn become oldies and their children opt for a lower birth rate. Our ultimate problem (a global systems issue), is that population growth (which accounts for much economic growth) must eventually peter out.

  2. I'd contest the "good chance that the rate of increase of the retirement age, which is universal and therefore a notional average, may be greater than the rate of increase in the longevity of the poorest."

    Your model suggests that the number of years in retirement for that group will increase by about 40% between 2006 and 2046. I acknowledge your points that the data feeding the model may be flawed, but a swing from 40% increase to any decrease sounds like a pretty big swing to me. What numbers would you need to see in order to have that result? Is there any data to back up the notion that it's likely?

    Also, I'd like to see numbers on the percentage of population that would be regarded as 'poorest' against 'richest' in your model. I suspect that there was a much larger proportion in the 80s when manufacturing (and therefore manual labour) still had some power in the economy, never mind in comparison to 1908 when the pensions act was introduced. Low paid office workers may have a similar life expectancy to well paid office workers.

    I strongly suspect that the vast majority of the population fall into the gap between the two with an ever decreasing proportion at each pole. I'm sure you can dig out the numbers to prove that theory. This might mean that the number of people affected in the way you suggest may actually be very small.

    And what are the other benefits in place that protect the poorest from falling into the poverty regarless of their age? What other benefits will be available at the point at which the state pension kicks in? When those are taken into account, what is the gap in terms of money paid?

    I understand that this isn't about money, but rather about that period of time beyond which you have 'paid your dues' and can now relax and let the rest of society give you back some of what you paid - but that's a notion based on the idea that NI is paid into a pension pot for you to draw on in the future. That may have been the way it was set up, but fundamentally isn't taxation actually about the forced centralisation of capability weighted payment for services for which the individual may not otherwise choose (or be able) to pay - for the good of the whole society - rather than a long term investment for the individual's benefit?

    The notion of NI in its original form doesn't really sit well with me (which I guess is strongly related to your final point) - you paid in, so you deserve it back - you didn't pay in, so you deserve nothing.

    In which case, ensuring that all people in the society are above a certain standard of living at all times might be enough..?

  3. Rob,

    Look at scenario C in the above table. If the retirement age is increased between 2036 and 2046 by 3 years, then the years in retirement for the bottom cohort drops from 14.9 to 14.2. The latter figure is still an increase on the 10.4 enjoyed in 2006, but an inflexion point has been reached.

    My pessimism is based on two assumptions. First, now that the principle of an increase in the retirement age has been accepted, I expect periodic upward revision to become the norm. Second, the inequality in life expectancy will continue to widen not just in absolute but in relative terms - i.e. the percentage growth for the better-off will be higher, year on year, than that of the poorest. It is for this reason that the increase in the universal pension age will be faster than the increase in the life expectancy of the poorest, causing their years in retirement to decline.

    The top/bottom categories I've used were taken from the ONS report (linked above), page 5. Top is "managerial & professional" occupations, while bottom is "manual & routine". I agree that all of this is speculative, and that changes in work and health, not to mention the impact of other benefits, may invalidate my conclusions, but based on the limited evidence we have today, I think what I've outlined is a credible scenario.

    Re your more fundamental point about taxation, there is an argument that the state does not provide centralised services (mandatorily funded by all) for the good of society but for the good of capital. Where state provision doesn't exist, we must either save our income (and so spend less) and make our own provision, or we piss our wages away and end up reliant on charity.

    The first is inefficient as individuals aren't very good at saving the exact amount needed (consider the inadequate rate of pension saving today). The second is dangerous as you end up with desperate people attracted to expropriation. As we see today in China, the lack of public healthcare has fuelled excessive savings, which in turn funded the orgy of debt in the West, and has served to suppress domestic demand. I suspect they will shortly announce plans for an NHS, specifically to free up savings and produce a domestic boom as exports flatten.

    For capital, it makes sense to socialise welfare provision. Businesses don't have to get involved, a central state has the authority to gather taxes, and a national scale solution is more efficient. We shouldn't forget that across Western Europe the welfare state, and specifically contributory insurance schemes, were first advanced by conservatives (Bismark) and Liberals (Beveridge), not by socialists.