By Stephen Roach
The global debate is endless (fortunately), but it’s also very simple. The key question is whether the current US slowdown has broader cross-border consequences. For financial markets, which are still discounting relatively sanguine global growth prospects for 2007-08, there is great enthusiasm for the ever-optimistic decoupling scenario – whereby the rest of the world miraculously untethers itself from the US. That remains a real stretch, in my view.
On the surface, the latest global trends seem quite consistent with a decoupling scenario. America has slowed but the rest of the world has picked up. In particular, there seems to have been a meaningful shift in the mix of growth in the industrial world. The US economy has downshifted from 3.4% growth over the 2003-05 period to only about 2% over the past year while trend growth in Europe and Japan has accelerated from around 1.5% to 2.5%.
Never mind that the improved pace in Europe and Japan is only a scant faster than the weakened trend now evident in the US. The decoupling crowd rests its case on the “second derivatives” – the juxtaposition of a deceleration in the US compared with acceleration elsewhere in the industrial world. China and India are the icing on the cake – emblematic of a seemingly open-ended boom in the developing world that remains unscathed by the US slowdown. The case for global decoupling concludes that world GDP growth – which surged at a 30-year high of 4.9% over the past four years – will barely skip a beat in 2007. Little wonder that financial markets are priced for a continuation of what many call the best global economy in a generation.
World economy yet to face a legitimate test
The fly in the ointment in this debate is that it may well be that an increasingly integrated world economy has yet to face a legitimate decoupling test. The US may have slowed but the downshift hardly represents a major derailment of the world’s major growth engine. Moreover, the deceleration has been concentrated in one of the least globalized pieces of the US economy – homebuilding activity.
Over the final three quarters of 2006, a steep contraction in residential construction expenditures knocked an average of 1.0 percentage point off real GDP growth in the US – a swing of -1.5 percentage points from the positive growth contribution of 0.5% over the preceding three years and enough of a drag to have accounted for all the downshift in real GDP growth over the same period. While the housing recession has undoubtedly reduced US demand for foreign sourced construction materials, this is hardly a major challenge to growth elsewhere in the world economy.
So far, the rest of the US economy has been relatively resilient in the face of this steep contraction in residential construction activity. That’s especially the case for personal consumption – more than 70% of US GDP and the one sector of aggregate demand that has the tightest linkages to America’s trading partners. During the final three quarters of 2006, when homebuilding activity hit the skids, annualized real consumption growth still averaged 3.2% - down only 0.2 percentage point from the growth pace of the preceding three years and fully 33% faster than overall GDP growth over the final three quarters of last year; moreover, in the first period of 2007, our latest tracking estimates suggest consumption growth held at this same impressive 3.2% pace. Business capital spending has started to weaken a bit in recent months. But the weakening has been concentrated in the equipment piece – only 7% of US GDP, or one-tenth the size of the personal consumption sector. Needless to say, as long as the American consumer continues to hold its own as a source of relative resilience, the US economy can shrug off a capex hit – and the global economy will hardly be tested.
Internal spillovers and external linkages
This outcome underscores a major source of confusion over the global decoupling call – the distinction between internal spillovers and external linkages. The former, in my view, pertain to the interconnectedness within an economy – the relationships between sectors. An obvious case in point is the lack of any spillovers between homebuilding and consumption in the US – at least, so far. I would define linkages as more of a cross-border phenomenon – in effect, the transmission of shifts in one economy to the broader global economy through global trade flows. Internal spillovers are a necessary – but not sufficient – condition for cross-border linkages. But if there have been no internal spillovers, the external linkage debate – and therefore, the global decoupling call – is all but meaningless. That remains very much the case today, in my view.
This same point recently has been made by the research staff of the IMF in the prepublication of one of the chapters in the April 2007 issue of the World Economic Outlook (see Chapter 4 on the IMF website, “Decoupling the Train? Spillovers and Cycles in the Global Economy”). Notwithstanding erroneous press accounts of this research, the IMF staff throws cold water on the notion of a global decoupling from the US. To the contrary, they stress that the “…potential size of spillovers from the United States has increased with greater trade and financial integration.” They underscore the same point I stressed above – that as long as the US slowdown remains confined to sector-specific developments such as housing, the less the chances of a more severe stalling out of the American growth engine and, as a consequence, the lower the probability of a more broadly based global slowdown.
Who has the greatest export exposure the the US?
The IMF research also provides a comprehensive ranking of the cross-border linkages to the US. Based on export exposure to the US, America’s NAFTA partners – Mexico and Canada – are at the top of the vulnerability list; for both of these economies, goods shipped to the US account for around 25% of their GDP.
By contrast, Japan has reduced its dependence on America, with US-bound exports averaging just 2.9% of GDP over the 2001-05 interval – well below the 4.0% portion some 20 years earlier. For the Euro area, US dependency ratios remain quite low, although they have inched up from 1.5% in the first half of the 1980s to 2.4% in the first half of 2000s. Similar modest increases in US exposure have been evident in Brazil and Argentina, and because of oil and resource linkages, US dependency ratios have also risen for Sub-Saharan Africa – from 3.0% in 1981-85 to 5.9% in 2001-05.
The results of the IMF staff research are not surprising. They are, in fact, nearly identical with similar conclusions that I and others have stressed in considering the repercussions of a US slowdown on the broader global economy (see my 30 October 2006 dispatch, “The Fallacy of Global Decoupling”). As I noted at the time, the “decouplers” – economies that can stand on their own in the event of a major growth shortfall in the US – must satisfy three conditions: They need to have a broadening base of self-sustaining domestic demand, a diversified export mix, and policy autonomy. In my view, progress is still quite limited on all three counts. Private consumption continues to lag in Europe and Asia. Moreover, the US is still the dominant global export destination; by IMF estimates, the US accounted for 20% of global merchandise exports over the 2001-05 period – a record high for the US and larger than the Euro area as the biggest portion of global trade. Nor is there much leeway for global policy makers to ride to the rescue in the event of a US growth shock; that’s especially the case in developing Asia, which is constrained by currency considerations, but it is also true in Japan, where policy rates are still very close to “zero.”
The outlook for the US consumer
In the end, this debate boils down to the one big call that has always weighed most heavily on the macro outlook – the fate of the American consumer. If US consumption growth remains brisk in the face of pressures building elsewhere in the economy – especially housing, but also business capital spending and autos – then a globalized world will, in effect, have nothing to decouple from. The surprisingly strong March labor market surveys – brisk employment and falling joblessness – underscore the ongoing resilience of labor income generation and consumer purchasing power.
Yet as Dick Berner, our resident consumption bull, recently conceded, consumers will need all the help they can get in the face of higher energy and food costs, decelerating housing wealth creation, adjustable-rate mortgage resets, and a tightening of lending standards in the aftermath of the sub-prime mortgage fiasco (see his 2 April dispatch, “Perfect Storm for the US Consumer?”). But if the US labor market continues to display extraordinary staying power in the face of adversity elsewhere in the economy, the overly-indebted, saving-short American consumer could squeak by once again – and so, too, would the rest of a still-coupled world. I remain highly dubious of such an outcome but concede that the burden of proof remains on me.
I have long been struck by the inherent inconsistency of a macro call that extols the virtues of integration and globalization, on the one hand, while celebrating the resilience of a decoupled world, on the other hand. Don’t kid yourself – if the lead engine of the global growth train goes off the tracks, the rest of the world will be quick to follow. So far, that hasn’t happened – underscoring my basic conclusion that there has yet to be a meaningful test of the global decoupling thesis. It’s up to the American consumer as to whether that test will ever occur.
By Stephen Roach, global economist at Morgan Stanley, as first published on Morgan Stanley’s Global Economic Forum