On the frequency of economic surprises
26 February 2017
The central theme in these notes is that to better interpret the economic and political noise of a changing world, it is useful to think about the underlying structural dynamics – drawing on history, politics, as well as economics. So how does economic history suggest we should think about themes such as the Trump reflation trade, the prevailing scepticism about the outlook for large parts of Europe, or the outlook for China?
One simple but powerful idea in this regard is regression to the mean. In a 1994 paper, Larry Summers and co-authors noted the absence of growth correlations across decades: countries that did well in a particular decade were no more likely to out-perform in the subsequent period. Examples abound: the US did better in the 1990s than in the volatile 1970s and 1980s; Japan was feted as a superpower in the 1980s before sliding into sustained stagnation; the ‘Trente Glorieuses’ (thirty glorious years) from 1945 in Europe ended abruptly in the 1970s; and the Swedish model was written off in the 1980s before the Nordics became the world’s favourite countries. And so on.
Subsequent analysis shows that this remains the case. Germany, for example, was regarded as the sick man of Europe in the 1990s before coming back strongly in the 2000s. And there was strong services-based growth in the UK from the late 1980s after two grim decades.
There are some persistently successful economies – Australia and the Netherlands have gone over 25 years without a recession, for example – but the distinctiveness of this experience serves to make the point. And sometimes growth dynamics can play out for longer than expected: for example, Paul Krugman called the end of the Asian Miracle in 1994 on the basis that input-driven growth would hit diminishing marginal returns. Growth rates in the East Asian tigers have now slowed, but these countries sustained above average growth rates across several decades.
This international experience speaks to the contingency of growth. Particular growth models work well in particular environments or for particular reasons – and then need to adapt. There is no growth model that is right for all time, and there is an element of good luck in many economic success stories. Countries often struggle to adapt to changed circumstances. This was evident in many developed countries after the shocks of the 1970s. As Harold Macmillan, former UK PM, said when asked about the greatest challenge facing governments, “Events, dear boy, events”. High growth countries should not be complacent.
Indeed, it is often the most successful countries that struggle the most to respond. In some cases this is because there is less urgency around the need to adapt to disruptive developments (the Kodak phenomenon). And in other cases it is because successful growth models carry the seeds of their own demise: deep investments are made in particular sectors and markets, or the economic structure becomes tightly organised around a particular mode of operation (such as natural resources, FDI attraction, or labour force growth).
In periods of significant external change, movement in the distribution of economic performance is particularly likely. Such transitions were seen in the post-war period, in the 1970s, in the 1990s, and are likely again now given the structural political and economic changes underway. Economies that have prospered in the environment of the past decade or two need to make adjustments to position themselves for a regime change; and some poor performers may find that they now have the right capabilities to perform well.
Of course, regression to the mean is not an iron law and much depends on how specific countries respond to emerging challenge and opportunities. But it is a powerful tendency. And so it is important to think broadly about the distribution of possible economic outcomes, rather than anchoring on recent relative performance and extrapolating forward. Indeed, my sense is that there is fragility in some high performing countries and latent potential in some laggards.
Consider the following groups of countries. First, although deeply challenged, the sluggish large economies of France and Japan (and Italy at a stretch) have underlying capabilities and strengths – from human capital and technology, to leading global firms. There is upside potential over the next several years under the right political and economic conditions, perhaps like Germany 15 years ago. Contrast that with some Anglo economies (US, UK, Australia) that have performed well over the past couple of decades, but which face growing political and economic risks. There is the potential for a reversal in fortunes of these country groups.
Second, the historical record suggests that high growth emerging markets such as China and India will struggle to sustain even moderate versions of their current growth rates. A potential for a disruptive downward shift in growth rates is very possible. Indeed, of the BRICS grouping, Brazil, South Africa and Russia have all seen a marked worsening of their economic performance.
And third, small advanced economies have prospered on the back of intense globalisation over the past 25 years but now face a more challenging international economic and political environment. I remain broadly positive on small countries because they have strong intrinsics, but I expect the distribution of outcomes across the small economy group to widen. Growth will not come as easily, and some may falter.
I expect that the growth rankings in 2020 – and certainly 2025 – will look materially different from the current rankings. The historical record suggests that these growth performance transitions can occur quickly, and that we should expect to be surprised. Past performance is no guarantee of future success.
John Kenneth Galbraith once observed that “a nicely conformist nature, a good tailor, and the ability to articulate the currently fashionable clichés have usually been better for personal success than an excessively inquiring mind”. But given the shifting tectonic plates, some excessive inquiry – and perhaps some contrarianism – will be useful rather than accepting the conventional wisdom.