by Gavyn Davies, Department of Macroeconomic Research, Fulcrum Asset Management.
Research on secular stagnation, and its effects on equilibrium real interest rates (r*) continues apace in the macro-economic community. In the past couple of years, researchers at the Federal Reserve and the Bank of England have focused on the role of demography in reducing r* in the US and other advanced economies. According to the results of this research, three different demographic forces have together reduced r* by around 1-1.5% since 1980. These forces are: slower growth in the labour force, the aging of the population and the lengthening in life expectancy. These forces are expected to persist for quite a while. The Federal Reserve seems to be paying increased attention to these demographic factors, which means that they see less role for an early decline in the temporary economic “headwinds” which they had previously believed might disappear in the near future. This reduces the risk that the Federal Reserve will impart a hawkish shock to financial markets in the period ahead.
James Carville won the Presidency for Bill Clinton in 1992 with a sign in the campaign’s headquar- ters saying “The economy, stupid”. Maybe there should be a sign in the Federal Reserve saying “Demography, stupid”.
Central bankers, like investors, have usually tended to ignore or underplay the influence of demographic factors over the short and medium term. The size and age distribution of the population changes very gradually, and in a fairly predictable manner, so sizable shocks to asset prices from demographic changes do not happen very often.
That does not mean that demography is unimportant. The cumulative effects can be very large over long periods of time. Apart from technology, there is a case for arguing that demography is the only thing that matters in the very long run. But demographic changes usually emerge very slowly, so they do not trigger sudden fluctuations in the determinants of asset prices, notably the economic cycle and monetary policy.
However, there are exceptions to this rule, and we may be living through an important exception at the present time. It seems that the Federal Reserve is starting to recognise that the decline in the equilibrium interest rate in the US (Davies,2016)(r*) has been driven not by temporary economic “headwinds” that will reverse quickly over the next few years, but instead has been caused by longer term factors, including demographic change.
Because these demographic forces are unlikely to reverse direction very rapidly, the conclusion is that equilibrium and actual interest rates will stay lower for longer than the Fed has previously recognised. Of course, the market has already reached this conclusion, but it is important that the Fed is no longer fighting the market to anything like the same extent as it did in 2014-15. This considerably reduces the risk of a sudden hawkish shift in Fed policy settings in coming years.
Furthermore, greater recognition of the permanent effects of demography on the equilibrium real interest rate has important implications (see Carvalho et al.,2016) for inflation targets, the fiscal stance and supply side economic policy. These considerations are now entering the centre of the debate about macro-economic policy.
The relationship between demography, growth and interest rates has been studied by economists ever since the days of Malthus, but it has played relatively little role in mainstream macro-economic discussion in the last few decades. Recently, however, several important studies (summarised below) have emerged from central bank economists, emphasising the link between demography, GDP growth and r*.
These links are obviously related to the work of Lawrence Summers on “secular stagnation”(see Rachel and Smith, 2015), and more particularly to the work of Alvin Hansen (see Brown, 1989) on population growth in the 1930s. The different forms of secular stagnation have become increasingly influential among policy makers. Last week, Fed Vice Chairman Stanley Fischer1 accorded an important role to demography in an important speech on the causes of the decline in r*. Since Fischer was among the most hawkish members of the Fed’s Board when he wanted to “normalise” interest rates last year, this could mark a significant change in the thinking of the FOMC.
Why is there a link between r* and demography? Remember that the equilibrium real rate of interest is that which ensures that savings and investment in the economy are equal in the long run. If ex ante savings exceed investment, r* declines, and vice versa. Since the savings behaviour of households is clearly affected by the age distribution of the population, and the investment behaviour of the corporate sector is affected by the labour supply, it is obvious that demography matters a lot for the determination of r*.
In recent work, three aspects of the population statistics have emerged as important in explaining the decline in r* in the developed economies. These are:
- The growth rate in the labour supply. Most models (though not all) produce a relationship between real GDP growth and r*, and also allow GDP growth to be impacted by a change in the supply of labour. The labour force is now slowing down rapidly in most advanced economies. Since the capital stock is fairly fixed for lengthy periods, this will increase the capital/labour ratio in the economy, and the “abundance” of capital will both reduce the rate of return on capital, and the attractiveness of new investment projects. This reduces real interest rates (Figure 1).
- The dependency ratio within the population. When the number of dependents (young and old people) relative to those of working age is low, the savings rate in the economy tends to rise, because workers save more than retirees. When the bulk of the baby boomers were in the labour force before 2000, this caused a large rise in savings in the ad- vanced economies, which triggered a drop in r*. This will shortly start to reverse as the baby boomers retire (Figure 2).
- The life expectancy of the population. If lifespans are expected to lengthen, while the retirement age remains unchanged, then people will choose to save more while they are in employment (or delay expenditure when retired, which is more difficult) in order to remain comfortably off until they die. This increases the savings ratio and reduces r* (Figure 3).
Although recent economic studies do not completely agree about the relative importance of these three factors2, there is a consensus that, together, they have accounted for a significant part of the decline in r* since 1980. For the world as a whole (including emerging markets), Bank of England authors (Rachel and Smith, 2015) calculate that demographic composition and labour supply growth has reduced r* by about 1 per cent in the past 30 years. Carvalho et al.(2016) estimate that the demographic transition has reduced r* by 1.5 percentage points in developed economies since 1990. And Federal Reserve authors (see Gagnon et al., 2016), in a significant recent paper, conclude that their demographic model accounts for 1.25 percentage points decline in r* and trend GDP growth3 since 1980. They say this is “essentially all” of the decline in these variable in the US over this period. Stanley Fischer quoted this estimate with approval in his speech last week. Although the retirement of the baby boomers may soon start to cause a drop in the US savings ratio, other demographic factors are expected to keep r* abnormally low for a long time to come. The Federal Reserve authors calculate that the current level of the underlying equilibrium real interest rate, based on the state of demography alone, is only 0.5 per cent. This compares with the latest FOMC estimate4 of 0.9 per cent for r*. That official estimate includes several economic forces other than demography that are also keeping interest rates down, so r* may well be reduced further in coming FOMC meetings. In any event, as investors and policy makers absorb the latest macro-economic research, demography may assume an increasingly important role in their thinking about fiscal and monetary policy. I will return to the policy and financial market implications on another occasion.
Brown, E. (1989): “Alvin Hansen’s contributions to business cycle analysis,” MIT working paper.
Carvalho, C., A. Ferrero, and F. Nechio (2016): “Demographics and Real Interest Rates: Inspecting the Mechanism,” Federal Reserve Bank of San Francisco.
Davies, G. (2016): “What investors should know about R star,” Financial Times [Online; posted 11-September-2016].
Gagnon, E., B. K. Johannsen, and D. Lopez-Salido (2016): “Understanding the New Normal: The Role of Demographics,” Fed Board Discussion Series.
Rachel, L. and T. D. Smith (2015): “Secular drivers of the global real interest rate,” BoE Working Papers.
1) “Why Are Interest Rates So Low? Causes and Implications,” The Economic Club of New York, October 17, 2016
2) There are also other ways in which characteristics of the population can affect r*. For example, ever since Kuznets, it has been suggested that a larger, faster growing or younger population can result in faster productivity growth, which is uniformly regarded as a key determinant of r*. Some economists therefore believe that the aging of the population in recent years has reduced productivity growth and consequently r*.
3) “There’s a devastatingly simple explanation for America’s economic mess,” The Washington Post, October 7, 2016
4) “Summary of Economic Projections,” The Federal Reserve, September 2016