The Cautious Consumer
Address to The Anika Foundation Luncheon
Supported by Australian Business Economists and Macquarie Bank
Sydney - 26 July 2011
Thank you for coming out once again in support of the Anika Foundation. I want also to thank in particular Macquarie Bank and the Australian Business Economists for their continuing support of this annual series.
In last year's Anika Foundation address, I talked about the fiscal difficulties being faced by governments of some of the world's largest countries in the wake of the financial crisis. A theme of that talk was that a number of major advanced economies had been facing for a while the need to address long-term structural issues in their fiscal accounts. In large part these stemmed from the inevitable collision of long-run trends in demographics and entitlements. A deep recession and the prospect of a slow recovery have brought forward the pressure to face these issues.
Over the past year, we have seen the focus on fiscal sustainability continue to increase. The problems have been most acute in Greece, though unfortunately not confined to it. Greece is a small country that has nonetheless assumed considerable significance. The citizens of Greece are now experiencing an austerity regime of historic proportions, which is a pre-condition for access to the foreign official funding that will allow them to meet their near-term obligations. But the longer-term solution surely will involve the taxpayers of Europe accepting part of the costs of restoring Greece to sustainability. The recent agreement appears to be a further step in that direction, with risk being shifted from the private sector onto the European public sector. I would view this as a step on the road to an eventual solution, though European policymakers continue to face a very delicate task in preserving the combination of fiscal sustainability and the single currency.
Concerns about the US fiscal position have also increased, though this has not been reflected, at this point, in market prices for US debt. The immediate need is for the US authorities to lift the debt ceiling, then for them to work towards longer-term sustainability.
In both the US and European cases, the process of allowing things to go right to the brink of a very disruptive event before an agreement is reached on the way forward has been a source of great uncertainty and anxiety around the world.
That anxiety has extended to Australia, even though, as I am sure people are sick of hearing me say, Australia is in the midst of a once-in-a-century event in our terms of trade. I won't recite the facts yet again. Suffice to say that this is, at least potentially, the biggest gift the global economy has handed Australia since the gold rush of the 1850s.
Yet it seems we are, at the moment, mostly unhappy. Measures of confidence are down and there is an evident sense of caution among households and firms. It seems to have intensified over the past few months.
There are a number of potential factors to which we can appeal for an explanation of these recent trends. The natural disasters in the summer clearly had an effect on confidence, for example. Interest rates, or intense speculation about how they might change, are said to have had an impact on confidence – even after a period of more than a year in which the cash rate has changed only once, the most stable outcome for five years. Increasingly bitter political debates over various issues are said by some to have played a role as well. The global outlook does seem more clouded due to the events in Europe and the United States. We could note, on the other hand, that the Chinese slowdown we have all been anticipating seems to be relatively mild so far – that country has continued to expand at a pretty solid pace as measured by the most recent data. But these days, mention of the Chinese expansion reminds people that the emergence of China is changing the shape of the global economy and of the Australian economy. And structural change is something people rarely find comfortable in the short term, even though a capacity to adapt is a characteristic displayed by the most successful economies.
So the description of consumers as ‘cautious’ has become commonplace. It is not one I disagree with. Indeed the RBA has made such references on numerous occasions over the past couple of years.
Nor do I wish to dismiss any of the concerns that people have. People want to make sense of the disparate information that is coming at them.
I want to suggest that to do that – to make sense of it all – it is worth trying to develop a longer-run perspective, particularly in the area of household income, spending, and saving. That is my task today.
I have two charts that I think help us to understand the story.
These figures are from the quarterly national accounts (Graph 1). The top panel shows the level of household disposable income, and household consumption spending. Income as shown here takes account of taxes, transfers and household interest payments. Both series are measured in real per capita terms, and shown on a log scale. In the lower panel is the gross household saving ratio, which is of course the difference between the other two lines expressed as a share of income.
Also shown is a trend line for each of the income and consumption series, estimated using ordinary least squares, over the period 1995-2005. The trend is then extrapolated for the 21 quarters since the end of 2005.
Notice that the trend growth for real per capita household income over the period 1995 to 2005 was 2.0 per cent per annum. That's a pretty respectable rate of growth for an advanced country. It was more than double the growth rate seen in the preceding two decades from 1975 to 1995. Growth at that pace means that the average real income doubles about every 35 years.
Notice also that the trend line for consumption was steeper than that for income over the same period. These two lines were on their way to meeting. Real per capita consumption growth averaged 2.8 per cent per annum. This was a full percentage point higher than in the preceding ten-year period, and similar to the sorts of per capita growth rates for consumption seen in the late 1960s and early 1970s.
For ten years up to 2005, then, consumption growth outpaced income growth, which was itself pretty solid, by three-quarters of a percentage point, on average, every year. In fact this convergence of income and consumption was the continuation of a trend that had already been in place for about a decade. As the bottom panel shows, the flow of saving fell as a share of income.
As you can see, things began to change after that. From about 2006, real per capita income began to grow faster. Over the five years to the end of 2010, it rose by 2.9 per cent per annum. It may not be entirely coincidental that that period is when the terms of trade really began to rise in earnest.
Yet from about the end of 2007, even as income was speeding up, household consumption spending slowed down. In per capita terms, real consumption today is no higher than three years ago. It's no wonder that people are talking about consumer caution, and no wonder that retailers are finding things very tough indeed. Coming after a period in which real consumption had risen by 2.8 per cent a year for a decade, and had outpaced income growth for two decades, no net growth in consumption for three years is quite a big change.
But these figures suggest that lack of income growth is not the reason for lack of consumption growth. It's not that the income is not there, it's that people are choosing, for whatever reason, not to spend it in the same way as they might have a few years ago.
Why is that? To find an answer we need to look to the financial accounts of the household sector. It is now time to introduce the second chart (Graph 2). It shows gross household assets, also measured per capita, in real terms (i.e. deflated by the consumption deflator). The two components are financial assets (including superannuation assets) and non-financial assets, the bulk of which is dwellings. These data say that gross assets across Australian households presently average about $800,000 per household, or about $300,000 per capita.
Between 1995 and 2005, assets rose at an average annual pace of 6.7 per cent in real, per capita terms. Completely comparable figures for earlier periods are hard to come by. But it's pretty clear that this increase stood out.
Using the Treasury's series for private wealth, from 1960 to 1995 the annual average per capita rate of increase in total wealth, in real terms, was 2.6 per cent. That is, it was broadly similar to the per capita growth rate of real GDP, which is what one would expect. So the growth from 1995 to 2005 was at a pace well over double the average of the preceding three or four decades. A large part of the additional growth was in the value of dwellings. The extent of leverage against the dwelling stock also tended to increase, with the ratio of debt to total assets rising from 11 per cent at the start of 1995 to around 17½ per cent by the end of 2005. It has tended to rise a little further since then.
Had we really found a powerful, hitherto unknown route to genuine wealth? Or was this period unusual?
Looking back, it appears the latter was the case. In 2008, the trend changed. Real assets per person declined for a period during the financial crisis. Given the nature of that event and the potential risks it presented, that is not surprising. Real asset prices have since risen again but, so far, have not resumed the earlier trend rate of increase, and at this stage they show no signs of doing so. They look very much like they are on a much flatter trend.
This adjustment has been considerably less abrupt than those seen in some other places. Nonetheless, it is a very substantial change in trend. If people had been banking on a continuation of the earlier trend, they would be feeling rather disappointed now. Of course if that earlier trend in gross wealth owed something to a tendency to borrow to hold assets, then a continuation would have exposed households to increased risk over time.
Casual observation suggests that this change of trend in the growth of assets, or ‘wealth’, roughly coincided with the slowing in consumption spending relative to its earlier very strong trend. It seems fairly clear that these financial trends and the real consumption and saving behaviour of households were closely connected.
I would argue that the broad story was as follows. The period from the early 1990s to the mid 2000s was characterised by a drawn-out, but one-time, adjustment to a set of powerful forces. Households started the period with relatively little leverage, in large part a legacy of the effect of very high nominal interest rates in the long period of high inflation. But then, inflation and interest rates came down to generational lows. Financial liberalisation and innovation increased the availability of credit. And reasonably stable economic conditions – part of the so-called ‘great moderation’ internationally – made a certain higher degree of leverage seem safe. The result was a lengthy period of rising household leverage, rising housing prices, high levels of confidence, a strong sense of generally rising prosperity, declining saving from current income and strong growth in consumption.
I was not one of those who felt that this was bound to end in tears. But it was bound to end. Even if one holds a benign view of higher levels of household debt, at a certain point, people will have increased their leverage to its new equilibrium level (or, if you are a pessimist, beyond that point). At that stage, debt growth will slow to be more in line with income, the rate of saving from current income will rise to be more like historical norms, and the financial source of upward pressure on housing values will abate. (There may be other non-financial forces at work of course.)
It is never possible to predict with confidence just when this change will begin to occur, or what events might potentially trigger it. But an international financial crisis that envelops several major countries, which has excessive borrowing by households at its heart, and which is coupled with a major change in the global availability of credit, is an event that would be likely to prompt, if nothing else did, a reassessment by Australian households of the earlier trends. It would also prompt a re-evaluation by financial institutions of lending criteria. This is precisely what has occurred over recent years.
What are the implications of these changes?
An important one is that, as I said at a previous Anika Foundation lunch two years ago, the role of the household sector in driving demand forward in the future won't be the same as in the preceding period. The current economic expansion is, as we all know, characterised by a very large build-up in investment in the resources sector and expansionary flow-on effects of that to some, but not all, other sectors of the economy. It is certainly not characterised by very strong growth in areas like household consumption that had featured prominently in the preceding period.
That is partly because the change in the terms of trade, being a relative price shift, will itself occasion structural change in the economy: some sectors will grow and others will, relatively speaking, get smaller. That is particularly the case if the economy's starting point is one that is not characterised by large-scale spare capacity.
But those pressures for structural change are also coinciding with changes in household behaviour that are associated with the longer-run financial cycles I have just talked about. Just as some sectors are having to cope with the effects of changes in relative prices – manifest to most of us in the form of a large rise in the exchange rate – some sectors are also seeing the impacts of a shift in household behaviour towards more conservatism after a long period of very confident behaviour.
It would be perfectly reasonable to argue that it is very difficult for everyone to cope with both these sets of changes together – not to mention other challenges that are in focus at the same time. However, if we were to think about how things might have otherwise unfolded – if households had been undergoing these shifts in saving and spending decisions without the big rise in income that is occurring, to which the terms of trade have contributed – it is very likely that we would have had a considerably more difficult period of adjustment.
What then about the future? Can we look forward to a time when these adjustments to household saving and balance sheets have been completed? We can. To return to the first of my two charts, the current divergent trends between income and consumption spending are no more sustainable than the previous trends ultimately were. At some point, the two lines are likely to stop moving apart. That is, the saving rate, debt burdens and wealth will at some stage reach levels at which people are more comfortable, and consumption (and probably debt) will grow in line with income, with a relatively steady saving rate. We could then reasonably expect to see consumption record more growth than it has in the past few years. After all, it is very unusual for real consumption per person not to grow.
We cannot really know, of course, when that might happen. Doubtless it will depend on what else is occurring. We can note that the rise in the saving rate over the past five years has been much faster than its fall was in the preceding decade. In fact it is, at least as measured, the biggest adjustment of its kind we have had in the history of quarterly national accounts data. So the adjustment in behaviour to what should be a more sustainable relationship between spending and income has in fact proceeded pretty quickly (which is presumably why it has become such a prominent topic of discussion). That in turn means that the time when more ‘normal’ patterns of consumption growth recur is closer than it would have been with a more drawn-out adjustment. Viewed in long-run perspective, it is not unreasonable for a nation to save a good deal of a sudden rise in national income conferred via a jump in the terms of trade, until it becomes clearer how persistent that new level of income is. As a better sense of the degree of persistence is gained, people will probably be more confident to spend than perhaps they are just now. It is entirely possible that, were some of the current raft of uncertainties to lessen, the mood could lift noticeably, so I don't think we need to be totally gloomy.
But what is ‘normal’? Will the ‘good old days’ for consumption growth of the 1995‑2005 period be seen again?
I don't think they can be, at least not if the growth depends on spending growth outpacing growth in income and leverage increasing over a lengthy period. A rapidly rising saving rate isn't normal, but nor is a continually falling one. While the rise in the saving rate has been unusually rapid, the level of the saving rate we have seen recently looks a lot more ‘normal’, in historical perspective, than the much lower one we saw in the middle of last decade.
A return to those earlier sorts of growth rates for consumption would instead require, and could only really be sustainably based on, a continuation of the faster pace of income growth we have seen since about 2006. To the extent that that income growth has been a result of the increase in the terms of trade, however, it probably won't be sustained at the same pace. The level of income will probably stay quite high – above the level implied by the earlier trend – unless the terms of trade collapse. But the rise in the terms of trade has probably now come to an end. So the rate of growth of per capita income could be expected, all other things equal, to moderate from its recent unusually strong performance.
If we want to sustain the rate of growth of incomes, and hence lay the basis for a return in due course to the sorts of growth of spending seen in the golden period of 1995-2005, we will have to look elsewhere. As everyone in this room would know, there is only one source of ongoing higher rates of growth of real per capita incomes, and that is higher rates of growth of productivity. Everyone here also knows that it is now just about impossible to avoid the conclusion that productivity growth performance has been quite poor since at least the mid 2000s.
So everything comes back to productivity. It always does. It has been observed before that past periods of apparently easy affluence, conferred by favourable international conditions, probably lessened the sharpness of our focus on productivity. Conversely, the will to reform was probably most powerful when the terms of trade reached a long-term low in the mid 1980s. Those reforms ushered in a period of strong productivity growth.
The thing that Australia has perhaps rarely done, but that would, if we could manage it, really capitalise on our recent good fortune, would be to lift productivity performance while the terms of trade are high. The income results of that would, over time, provide the most secure base for strong increases in living standards. That sort of an environment would be one in which the cautious consumer might feel inclined towards well-based optimism, and reopen the purse strings.
- The Anika Foundation supports research into adolescent depression and suicide. See: <http://www.anikafoundation.com/>.
- The measure here differs from the commonly cited net saving rate, which deducts an estimate of depreciation of the stock of fixed assets (such as dwellings) owned by households.
- I am conscious that this explanation has not made explicit reference to demographics – in particular the decisions of ‘baby boomers’ in the years leading up to their planned retirements. No doubt these factors also played a role. But there is enough complexity to grapple with here already for today's purposes. I have spoken before about population ageing and finance. See: <http://www.rba.gov.au/speeches/2005/sp-dg-161105.html>.
- ‘… the prominence of household demand in driving the expansion from the
mid 1990s to the mid 2000s should not be expected to recur in the next upswing. The
rise in household leverage, the much lower rate of saving out of current income, and
the rise in asset values we saw since the mid 1990s, are far more likely to have been
features of a one-time adjustment, albeit a fairly drawn-out one, than of a permanent
trend. Moreover the risks associated with those trends going too far are apparent
from events in other countries. These risks have been reasonably contained so far
in Australia – but it would be prudent not to push our luck here.’