Demographics: greyer needn’t mean poorer

  • Written by 

    Written by Kevin Gardiner 12 April 2014

The UK Budget has put the pensions debate back on the front pages. However, while the changes to the way in which defined contribution pension funds can be used are welcome, they don’t really affect the collective resourcing of our pensions. We needn’t worry about that: received wisdom is far too pessimistic.

Here we remind readers why our greyer society can easily afford its collective pensions; why financial assets – capitalised or annuitized – have less to do with it than you might think; and why an explicit demography ‘theme’ should not feature prominently in investment portfolios. Valuations permitting, a substantial holding in equities remains the best way of acquiring the long-term cashflow that finances individual pensions.

Cattle to coins: the evolution of modern money 5 of 5 China reforms: a progress report 3 of 5

Residents of the United States, please read this important information before proceeding

Demography needn’t mean dependency

There are few things that can be predicted 30 years ahead with any confidence, but a greyer population is one of them. Most of the people who will be alive then are already born. Life expectancy is rising while birth rates have been declining, and the result is that the over-65 group is poised to almost double, from 8% of the population to around 15%. The trend will be most pronounced in the BRIC quartet and Japan, but Europe faces big increases too.

The resultant ‘demographic time-bomb’ has long been worrying economists. How can aging societies maintain living standards? How will they pay for their collective pensions? But the worries are overdone.

Societies will get greyer – but economists may be worrying too much

Economists have form here. They have often proclaimed – prematurely – that the world is about to run short of a key resource. Malthus worried that a growing population would not be able to feed itself. The Club of Rome worried we’d run out of oil. Fears about the aging population are driven by just such a concern, although this time the idea is that we are about to run out of workers. They are likely just as premature.

Before explaining why, there is a conceptual point to clarify first, and it relates to the role played by savings in “funding” pensions.

The pensions debate has been given added urgency by the financial crisis. Not only are we set to run out of labour, but our cumulative savings, and the income from them, seem disappointingly small, hence the UK chancellor George Osborne’s decision to liberalise the way in which pensions can be taken. However, in aggregate, financial markets and savings do not directly make available the real resources that pay for society’s collective pensions.

Financial markets don’t fund pensions – the economy does

Forget something you may take for granted at the personal level: the idea that a pool of savings directly provides your pension. For society as a whole, this is not the case. There are no collective stockpiles to be consumed in retirement – no warehouses full of tinned food and clothing. What pensioners consume is made shortly beforehand by those still in work. Pensions in aggregate are resourced on a pay-as-you-go basis by the productive capacity of the economy. That capacity is linked to past aggregate saving and investment decisions, but only indirectly and loosely. Imagine what would happen if we all tried to double the amount we save (or Google ‘The paradox of thrift’). The economy would shrink, because the immediate result of us spending less would be that business dried up and total incomes would fall – and with them, most likely, the actual flow of savings (hence the paradox).

The collective picture is thus more subtle than realised. If there is nothing in the shops, our carefully accumulated investment portfolios will be of little use. Ownership of financial assets gives us a place in the queue for a share of that production, but for society as a whole it doesn’t change what’s available for distribution between those in work and those not.

How does this help with the demographic problem? Because it focuses attention on what is important. Economic growth matters most, and the obstacles to growth posed by an aging society are routinely and materially overstated, for three main reasons:

(1) The elderly are not the only dependents

For the UK, the ratio of elderly to working-age people is poised to rise by almost one-half in the next 30 years. For some other countries, the increase is even more dramatic: roughly two-thirds in France, Germany, Italy and the US; a doubling in Japan and Russia; a more-than-doubling in India and Brazil; and a more-than-trebling in China.

Demography is not the only driver of economic dependency

The pessimists assume that this is a good guide to the increase in economic dependency, the burden carried by the working population. But it isn’t. The output of those in work has to support all those who are not working. Pensioners are not the only non-working age group. More importantly, not everybody of working age is in work.

Time-bombers routinely overlook both the unemployed (people of working age who are part of the workforce, but can’t find a job), and the economically inactive (people of working age who are not looking for work, and so not participating in the workforce, for whatever reason). When we allow for the wider non-working population, the increase in total dependency ahead is much smaller than the age mix alone suggests. The UK experienced a bigger burden back in the early 1980s: there were fewer pensioners then than there will be, but unemployment was very high, and participation rates were lower – which brings us to…

(2) Labour utilisation can change

It gets better (or worse, if you worry about the quality of the public debate in this respect). Time-bombers don’t simply forget to place pensioners in their wider economic context, they also ignore the possibility that the context can change. This is perverse. Why rush to worry about a shortage of labour when we’re so visibly not using what we’ve got to begin with?

Worry about running out of labour when we’re using fully what we’ve got to begin with

Most people who are unemployed would like a job if one were available. So too would many people who are economically inactive and not looking for a job, such as ‘discouraged’ workers, early retirees, family carers and some students (circumstances permitting, of course). Lower unemployment and higher participation can make a big difference. Compared to the naïve ‘time-bomb’ analysis, the proportionate impact of more older workers in the numerator of the dependency ratio is diluted by the inclusion of the other dependents; then, over time, moving adults out of unemployment and non-participation, and into employment, further cuts the numerator and raises the denominator. Plausible changes in labour utilisation could even deliver lower dependency, suggesting that current pensioner living standards could be financed at lower average taxes than today! For some countries, a prospective fall in total labour supply could become an increase (Figure 1).

This is not the end of the story. If they need to, pension ages and working hours can rise a little. For some, this is an admission of economic failure, but a retirement age fixed when life expectancy was much shorter – and when physically demanding manual work was more prevalent – is an expensive anachronism.

Figure 1 shows a bar chart displaying a comparison of the change in labour supply with the change in the elderly/working-age ratio across different countries.

Today’s retirees typically have a quarter of their lives still ahead of them, in contrast to those (for example) in Victorian times for whom retirement amounted to a few insecure and unhealthy years after a lifetime of hard, industrial effort. This assumes of course that older workers can play a productive role; and why shouldn’t they? If the average person is living for longer, and more healthily, they may be willing and able to work an extra year or two. Similarly, the length of the working week has been falling steadily. Professor David Miles has noted that the proportion of a typical male’s waking hours spent at work has fallen by three-fifths in the last century and a half. Slightly longer working hours now may not be that contentious – particularly since more jobs are part-time in nature.

What should governments do to help all this happen? They should pursue policies aimed at promoting labour market flexibility – as opposed to, say, raising the aggregate savings rate, which on its own could be counterproductive, as noted above.

Much growth comes from new technology and learning by doing, not labour input

(3) Growth is not driven by labour alone

What if changes in the labour market fail to occur – or if their impact is offset by a declining population – and total supply falls anyway? It may not matter as much as you might think.

Labour is not the only input. And much growth comes not from extra input, but from total factor productivity (TFP). It sounds too good to be true, but an estimate from the National Institute for Economic and Social Research suggested TFP accounted for most UK GDP growth in the second half of the twentieth century. Technology is a big source of such gains: life without spreadsheets (and IT generally) was less productive. More prosaically, think about a task done for the first time. Do it a second time and it’s faster; a third time and it’s faster still. You are moving up the learning curve. These changes are going on all the time across the economy, and can deliver more from less for as long as invention and learning are possible.

Portfolio implications

An older population need not be a threat: we are not fully using the labour we’ve have to begin with, and prosperity is, in any case, not dependent solely on labour input. The room for manoeuvre is greatest in the developed West; in China and Japan the aging is particularly pronounced and there is less slack in their labour markets. In the case of Japan, this is probably more important than the deflation that currently has so many pundits worried, and the likely result is that, in many ways, the Japanese economy will become more like the West – not the other way round, as many seem to think. But I digress.

As noted, pensions worries have been amplified of late by the poor investment performance of many formal pension schemes. Individuals with money-purchase schemes retiring recently have been disappointed by investment returns and by historically low annuity rates. UK Chancellor George Osborne’s 2014 Budget has provided a welcome attempt to increase the flexibility of defined contribution schemes in this respect, by relaxing the requirement to buy an annuity. However, individuals still have to decide how best to accumulate and then disinvest their funds from here.

As noted, the financial assets held by pension funds don’t directly provide the real resources consumed by pensioners. They do however offer one mechanism (the tax and benefit system is another) for redistributing resources away from workers and towards pensioners: members’ spending power gives them a place in the collective checkout queue at the supermarket, as it were. They decentralise administration, and create a link between individual contributions and the financial benefits paid on retirement. Some economists suggest they can also foster better-functioning capital markets, but the evidence is mixed.

You might think that since economic growth ultimately pays for pensions, the most natural pension asset might be equities, which share directly in that growth. For defined benefit pension plans, where payouts are fixed in advance, stock ownership has been made difficult by accounting policies introduced – with the best of intentions – in recent years, and most such schemes are closed to new entrants. Those accounting policies, and the ‘liability-driven investment’ they have fostered, favour bonds ahead of stocks. But for defined contribution plans, in which the value of pensions depend on the investment portfolio (and, until the 2014 Budget, annuity rates) on retirement, a more growth-oriented approach is feasible.

Stocks offer longer-duration cashflow than bonds

The growing cashflow generated by a diversified portfolio of stocks is contrasted with that available from a long-dated bond in Figures 2 and 3. A basket of typical stocks has longer duration than a basket of typical bonds, which might, ironically perhaps, make them a better match for the liabilities of defined benefit funds too if accounting guidance was altered.

We think today’s low interest rates are unlikely to be permanent; yield curves have been pricing-in a substantial rebound, and annuity rates might well rise alongside other longer-dated bond yields from here. Stock market returns have rebounded – and more importantly, valuations and prospective earnings growth suggest to us that further positive risk-adjusted returns are still likely on a tactical and strategic time horizon. Stocks are volatile, of course, and the older a particular individual’s fund is, the more it might make sense to allocate a higher proportion to more stable bonds. But stocks’ dividends are less volatile than their prices (Figure 4), and with no pressing need to purchase an annuity it may make sense, if risk appetite and financial circumstances permit, to continue to favour stocks for longer.

Figure 2 displays a bar chart showing the present value of cashflow: conventional 10-year bond, price 100, discount rate 2.5%. Figure 3 shows the present value of cashflow in terms of the equity index.

There are, in fact, few investment implications of an aging population, other than the general advice to keep an open mind about growth, and to avoid allowing today’s gloom to suck you into expensive bond markets. Almost everything that can affect security prices and interest rates is in play in the longer term. A confident forecast for one projected trend is of much less use than you might think, because there are so many other moving parts. Moreover, the fact that this trend has been visible for so long means that any specific investment conclusions that could have been drawn are likely long since priced into forward-looking markets.

Figure 4 shows a graph of US and UK stock prices and dividends, logged & indexed: 1965-2013

Demographic data can be correlated with stock market valuations. Some conclude that baby boomers’ savings were responsible for a big surge in stock prices, and that, as the boomers retire, lower valuations and poor performance is likely. Many funds do switch their members into more stable bonds as retirement dates approach, as noted. But this analysis ignores all the other drivers of stock market value, including growth, profitability and real interest rates – each of which could move unpredictably and markedly, and independently of demography, in the next 30 years. And if stocks are inexpensive to begin with, other investors may be happy to step in. You might think too that an older population will squeeze interest rates higher simply because it is less willing to wait: time gets more precious as you get older. This might lead to higher real rates, because collective ‘time preference’ is one of the underlying long-term drivers of interest rates. Again, however, there will be many other influences on rates. If real rates do go up – but partly because corporate profitability is strong – then the investment conclusion would be different (and might favour stocks ahead of bonds).

The most important investment ‘theme’ should be the maximisation of risk-adjusted returns – not demography

Convincing investment ideas are hard to find even at the sector level. An older population will favour healthcare, and leisure and entertainment sectors, but this has been largely priced into company valuations for a while. And the beneficiaries’ relative performance might yet be undermined by new, as yet unknown, industries and ideas. Remember, 30 years back, few pundits were tipping mobile telephony, digital media or China’s economic surge. More subtly, labour-intensive companies will benefit from more flexible labour markets as participation rates and retirement ages adapt to the altered pattern of labour supply. This is a less widely appreciated point, and so stands a chance of moving security prices, but it may do so across such a wide range of industries that it may not be of much practical use.

Bear these points in mind when you are offered the opportunity to invest in any ‘theme’ other than that of maximising risk-adjusted returns. For example, climate change, inflation or disinflation, emerging market growth and Western lifestyle worries, are, like demography, just some of the many influences that will shape investment performance – and they are not new ideas.