Sunday, 10 November 2013

Demographics and dependency

This is probably a pretty rubbish post, but then it was a rubbish debate, and I'm not proud of my own contribution to it. I did get things slightly wrong. But not as wrong as others.

Once again Fraser Nelson has tweeted graphs he doesn't really understand and drawn the wrong conclusions from them. Here's the first graph he tweeted, with his comment:

It's from this article by The Economist. The source matters, as will become apparent shortly.

And he then went on to say this:

The first chart is the dependency ratio - the ratio of working to non-working people in the population. And the second shows the rising proportion of over-65s in the working population. On the face of it, the second should influence the first. But actually they aren't related - and neither of them is much use.

The usual calculation of the dependency ratio assumes that people are working between the ages of 16 and 64. The Economist's chart doesn't say this, but it is reasonable to assume that it has been created on this basis. This of course is wrong at both ends. Fraser Nelson rightly points out that a rising number of over-65s are remaining in the workforce, or re-entering it. But neither Fraser Nelson nor The Economist mention that the number of 16-25 year olds in the workforce is falling rapidly as more and more of them spend longer in education and training. This more than offsets the increase in over-65s working.

However, the main reason for the rising dependency ratio is that people are living longer. The Economist points out (my emphasis):
Between 2013 and 2037 the nation’s population is expected to grow by 10m, to 73m. But the growth is not evenly spread between the generations—which means that a demographic crunch is coming. The number of people aged 75 and over will increase by 87%, to 9.3mThe number of people of prime working age (between 30 and 60) will increase by just 3%.
This is The Economist's main point - that the proportion of non-working people in the population is set to rise significantly due to increasing longevity. This rests not only on the assumption that immigration and fertility levels don't change, but also on the assumption that over-65s remain economically dependent. The Economist has actually left this question open: there is a large gap between its "prime working age" and the age at which it assumes no-one is working, which is 75. It's probably fair to assume, therefore, that The Economist knows that some over-65s work and is expecting more to do so. The question, therefore, is whether over-65s working would make a significant difference. We know that more of them are choosing to work. Does this mean that fewer over-65s are, or will be, economically dependent?

Well, not really. Fraser Nelson's second chart shows that just over 250,000 over-65s have re-entered the workforce since 2010 - hardly a majority. But what the chart doesn't show is that most of these are working part-time, they all receive a state pension and other benefits, and many are also receiving corporate and private pensions. They may be working, but overall they are still dependent. To eliminate their economic dependence there would have to be far more of them and they would collectively have to be earning enough for their pensions and benefits to be taxed away. That is a simply huge cultural shift. The fact is that the vast majority of over-65s expect to live primarily from pensions, including the state pension, and work is simply a means of topping up their pensions. The currently planned rise in the state pension age is not enough to make a huge difference to the hours collectively worked by over-65s, especially as many people gradually reduce their working hours long before they actually retire. The state pension age would probably have to be at least 70 for there to be a significant reduction in over-65s' economic dependence.

But actually the dependency problem is far more widespread than the dependency ratio chart shows. Those of working age who are economically inactive because they are students, or incapacitated, or full-time carers are counted in the working population, distorting the ratio considerably. And even more worrying, many people who ARE working are net recipients of benefits, because their wages are so low that they have to be "topped up" with state benefits which effectively wipe out their tax and NI contributions. These people are not contributing to the support of the elderly. On the contrary, they are competing with them for a share of tax revenues. And there are a growing number of them due to the bifurcating labour market and the decline of middle-income skilled jobs.

So Fraser Nelson is correct that the first chart is nonsensical - but not for the reason that he gives. The dependency ratio is a crude measure that takes no account of the actual economic contributions made by people in different circumstances and at different stages in their lives. A few over-65s working mainly part-time to top up their state pensions doesn't invalidate the ONS's dependency ratio calculation. But a large number of people dependent on state benefits to top up their wages does. We don't just have a demographic problem. We have a low wage problem.

Related reading:

Britain's demographic crunch will arrive suddenly - The Economist
Global greying - FlipChart Rick, Pieria
Bifurcation in the labour market - Coppola Comment
Analysis on the living wage - IFS
Workers kept their jobs but one third faced nominal wage freezes or cuts - IFS


  1. Is the likely fall off in pensioner income also a reason for concern? For instance my father has a fully index linked pension. Not many under 50s have that (pessimists who see lost generations, zombies and ZIRP forever might be even more downhearted).

    Does raising the pension age just push more people into lower benefits? Either way I'm unclear as a lay person if this will save a great deal as expected (savings will get burned sooner which is not awful but means a deeper dependency?).

    I agree with your point though an underemployed or low wage 30 year old is a dependent as much as someone with minimal pension.

    1. Yes, I agree. Present pensioners who have index-linked and/or final salary pensions are better off than any of tomorrow's pensioners will be. Sadly I do see ZIRP forever, because long-term interest rates are linked to population growth and the fact is that in the West, population growth is levelling off. The UK is something of an exception at the moment, so we may be able to get interest rates off the floor in a few years' time maybe, but don't count on it.

      The future does appear to be dependency for a growing number of working-age people, including many of those who would in past years have been retired. Therefore I agree that raising the retirement age may save much less than government would like - it will simply move the problem to a different part of the budget.

  2. Sadly theres far too much rubbish written about pensions v dependancy v age

    pensioners only get what can be described as a pittance of state pension if they have worked and paid in for 39 years

    wheras far too many layabouts, large families and working age get all manner of benefits thanks to GB which current pensioners certainly never got

    add on health tourists and illegal and legal immigrants gaining a disproportionate share of benefits the current pensioners never got but are sure paying for

    if anyone actually bothers to calculate everything correctly would prove todays pensioners are receiving a very raw deal indeed especially if they did the right thing and saved

    what is indisputable is the total change of careers and jobs for the young which has contributed to mass unemployment and destruction of hope for those leaving school or college

    banking and insurance etc used to employ millions of young and now all those jobs are in india

    that is what needs sorting very fast indeed

    1. Deborah,

      I don't accept comments that are off-topic. Please read my rules regarding comments - you will find them on the About This Blog page, tab at the top of the screen. This post is not about the problems of pensioners.Nor is it about the employment problems of the young. It's about demographic changes and the low-wage economy.

      I will not accept further comments from you that are off-topic. If you have something useful to contribute to the subject of the post, that is fine. But my blog is not the place to air your views about how unfair life is.

  3. Brick says

    The graphs assume that life expectancy will rise or at least remain at current levels. Low wages, employment stress, lack of employment can contribute to changing life expectancy. For example men in the Calton area of Glasgow have a life expectancy is 54 so inequality in life expectancy is rife. With the bifurcation of the labour market the assumptions that life expectancy will continue to climb should be questioned.

    Getting back on subject the ONS forecasts that people of state pension age is forecast to rise 28 percent from 12.2 million to 15.6 million by 2035. The population is forecast to rise from 62 million to 73 million. So i think the ratio of retirement age people is set to rise from 19.5 percent to 21.2 percent. The ratio we should be concerned with is the ratio of employed people to state pension age persons. Roughly 30 million people are employed in the UK between the ages of 16 and 64 which could be extrapolated out to around 35 million in 2035. If you work through the numbers you might find that if your unemployment rate halves then the demographic ratio problems go away.

    Low wages and low business tax collection mean that the government would still have an issue. Rather than a demographic issue, we should perhaps look at it in the light of globally low wages, globally low corporation tax, poor global growth, poor investment opportunities and general lack of demand. You could argue that if the safety net was better then wages would begin to rise, some potential retirees would make way for the youth of the country forcing up wages at the lower end and more immigrants would arrive. I can however foresee so many unintended consequences that it might structurally damage the economy (house prices for instance).