More on The Idea of 'World GDP'

Ever since Dave raised the question of Stephen Roach's problematic use of the idea of world GDP (and here ) I haven't been able to get the thing out of my head. Yesterday it came back again:

In all my years in this business, I’ve never come across such a worrisome and potentially lethal confluence of imbalances. For starters, they are global in scope. A lopsided world economy has never been so dependent on one growth engine — the United States. Over the seven-year 1995 to 2002 interval, revised figures now indicate that the US accounted for fully 96% of the cumulative increase in world GDP (at market exchange rates); that’s nearly three times America’s 33% share in the global economy.

In other words, outside of the United States, the rest of the world accounted for only 4% of the cumulative increase in global GDP over the seven years ending in 2002. While the strength of the dollar has exaggerated America’s contribution to world GDP growth over this period, there can be no mistaking the extraordinarily narrow base of this US-centric global growth dynamic.
Source: Morgan Stanley Global Economic Forum
LINK

Now that I am alerted to it, there is clearly something wrong here. The US responsible for 96% of global growth, this is absurd. The problem I think lies with the dubious use of the market exchange rates concept of World GDP (again see here). But the more I look the more I find problems. Take the following, for example:

In a US-centric world, export support is the rule not the exception. The more economies suffer from a deficiency of domestic demand, the more dependent they become on external support from world trade and on the currency underpinnings of such a growth dynamic. Make no mistake — this is a global phenomenon. Over the 1995 to 2002 period, worldwide growth of exports (goods and services) accounted for fully 50% of the cumulative increase in global GDP.

This seems fine? Well think about it for a moment. Apparently, collectively, economies suffer from a deficiency of domestic demand, yet manage to export. How is this possible? If we are all short on demand, how can we sell to the others, who also lack domestic demand? Well the answer is pretty obvious really: globalisation. The globalisation process means that export/import penetration is becoming more important in each and every economy. But that is why esport lead growth is becoming more and more important: because your exports must be someone else's domestic demand. So what Roach is really saying is that we are becoming more and more dependent on the extension of globalisation to drive growth and increasing economic welfare. Fine, then let's say that and not something else. Am I too much of a stickler here? I don't know. I'm a Roach admirer, but I think we should have the highest - not the lowest - standards with those we admire. The real problem is that Roach is good on instinct, but short on theory and explanation. As he says, if he feels the way he does, it's because he's a 'card carrying skeptic'. The global growth slowdown is real, but he hasn't got through to the explanation yet. Europe and Japan are getting stuck in the mud but what exactly are the "reforms that are needed to unshackle domestic demand". China and India will eventually bring about a 'global rebalancing', but this, alas, may not take the form which Roach himself is anticipating.