One of the most publicised theories regarding the current state of the world economy, and the causes of the international financial crisis, is that regarding 'global imbalances'. This thesis has been presented at book length by Martin Wolf in Fixing Global Finance, as well as in numerous articles in his position as the Financial Times chief economics commentator, by Stephen Green of Standard Chartered, by David Cohen of Action Economics, by Brad Setser, by Michael Pettis, Professor of Finance in Peking University, and numerous other authors.
This theory has however been described as erroneous by Asian leaders. That they are right in this opinion will be shown below. Specifically there is a logical incoherence in the theory in that its policy prescriptions are not entailed by its analysis. This in turn leads to it not being in accord with the facts of the world economy as the global imbalances it describes are shrinking by quite other mechanisms than the ones it outlines. As a result it is wrong in its policy proposals - as shown by the fact that the most rapid economic growth is being enjoyed by an economy, China, which is pursuing policies which are the opposite of those prescribed by this theory.
A theory which is deficient in coherence, and which fails to foresee the possibility of reducing global imbalances by mechanisms which are actually occurring, is evidently a theory which is erroneous.
Given the wide range of those supporting variants of what will be termed the 'global imbalances hypothesis' this theory naturally has a number of secondary variations. But they have a common core consisting of an analysis coupled with a policy conclusion which it is alleged is entailed by it.
On this theory, the well known problems of the US financial system are outward signs of a much more fundamental malaise of imbalances in the world economy. In the words of Martin Wolf: 'Nothing that has happened has been a product of Fed folly alone. Its monetary policy may have been loose too long. The regulators may also have been asleep. But neither point is the heart of the matter…. It is also a symptom of an unbalanced global economy. The world economy may no longer be able to depend on the willingness of US households to spend more than they earn.'
The core analysis of this theory is that the world economy has been characterised by two global imbalances, the first being the US balance of payments deficit – ascribed to US lack of saving/over consuming, the second being China's balance of payments surplus – allegedly caused by China 'over saving'. The policy conclusion which allegedly follows from this is that the US save more and China save less.
As Brad Setser put it: 'the global economy prior to the crisis was characterized both by high levels of both savings and investment in Asia and the oil exporters and by high levels of consumption and low levels of savings in the US. 'Therefore, as Robert Skidelsky phrased it: 'emerging market economies need to spend more and save less, and mature market economies need to spend less and save more.'. As Martin Wolf put it specifically regarding China: ''does it make sense for China to save so much or, for that matter, to invest so much?... Higher consumption today would surely be desirable.'
It is unclear in some variants whether this theory is being put forward merely prescriptively or whether it is stated that the trends it proposes as desirable are actually occurring – or a combination of the two. However it is clear that a number of its supporters believe that the proposed policy prescriptions of this theory are actually occurring. Thus for example Robert Skidelsky argued in July: 'In fact the present financial meltdown is producing the market-led adjustment that has eluded policy makers. Willy-nilly Americans are having to spend less and save more.'
The 'global imbalances hypothesis' has become so prevalent it may be described, in the phrase popularised by JK Galbraith, as a 'conventional wisdom'. Unfortunately, like many other previous conventional wisdoms it is not true. There are a number of further aspects of the 'global imbalances hypothesis' which are wrong, and which have earlier been analysed on this blog. But if a theory, as will be shown, is inconsistent and factually wrong that constitutes adequate grounds why it should not be maintained. A new, more correct, conventional wisdom is required.
Inconsistency of the theory
The economic inconsistency of the 'global imbalance hypothesis', that is that the policy prescription does not logically follow from the analysis, is easily demonstrated. A country's balance of payments is, by basic accounting identity, equal to the difference between its domestic savings and its domestic investment - the US deficit is exactly equivalent to the degree that US savings are lower than its domestic investment, China's surplus is exactly equivalent to the degree that its savings are higher than its domestic investment.
It immediately follows from this identity that the only way to correct such imbalances, if that is set as the overriding goal of policy, is not at all, as the 'global imbalances hypothesis' suggests, for the US to save more and China to save less. It is equally possible to solve these imbalances, again if this is set as the policy goal , by the US investing less and China investing more. Furthermore it is this latter process affecting investment, not the ones foreseen by the 'global imbalances hypothesis', that is actually occurring and leading to the lessening of the global imbalances.
To take this in detail, factually global imbalances are indeed rapidly declining – the US balance of payments deficit is shrinking and China's balance of payments surplus is declining. But they are doing so for reasons that are the the opposite to those outlined in the theory above.
Far from US saving rising it is declining. The US balance of payments deficit is therefore not shrinking because US saving is rising but because US investment is falling even more rapidly than US saving. China's balance of payments deficit is not declining primarily because it is consuming more but because it is investing more – that is, because China is moving its investment level up to its savings level, and in so doing generating the most rapid rate of growth in the world.
These factual trends have previously been outlined both regarding the US and China on this blog. The latest data which allows a major testing of the different theories is the publication of the second quarter US GDP figures together with data for the equivalent period for China's trade and GDP. All US data below therefore, unless specifically stated otherwise, is calculated from the tables accompanying the US second quarter 2009 GDP figures published by the US Bureau of Economic Analysis. The data shows clearly that the trends taking place in the world's two largest economies are not those in the 'global imbalances hypothesis'.
The rise in US consumption
Taking first the fundamental trend in US consumption, it is clear that under the impact of the international financial crisis US consumption has not fallen but risen further as a percentage of GDP. This is shown in Figure 1.
As may be seen total US consumption, the sum of personal and government consumption, rose rapidly as a proportion of GDP from 1997 until the end of 2003, then stabilised until the end of 2007, and then began to rise sharply again from the beginning of 2008 under the impact of the developing financial crisis.
Taking precise figures, between the last quarter of 2007 and the second quarter of 2009 US consumption rose from 85.8% of GDP to 87.6% - an increase of 1.8% of GDP. Between the second quarter of 2008, the last before the opening of the entirely open financial crisis with the collapse of Lehman's, and the second quarter of 2009 US consumption rose from 87.0% to 87.6% of GDP. Between the first and second quarters of 2009 US total consumption rose from 87.2% to 87.6% of GDP. The trend of rising US consumption under the impact of the financial crisis is therefore clear. More detailed breakdown of this rising share of consumption in US GDP may be found in the footnote. 
If US consumption has risen as proportion of GDP why, therefore, has the US balance of payments deficit been shrinking? By accounting identity this deficit is necessarily equal to the shortfall of US savings compared to domestic investment. Therefore shrinkage of the deficit means the gap between saving and investment is narrowing despite consumption rising. The explanation is that US savings are not increasing but falling, but US investment is falling even more rapidly than US saving.
The decline in US investment and saving
Analysing first investment, for which the necessary detailed data is available in the US second quarter GDP figures, these show that US total investment fell between the fourth quarter of 2007 and the second quarter of 2009 from 19.1% of GDP to 14.8% - a decline of 4.3% of GDP. In the period between the second quarter of 2008 and the second quarter of 2009 US investment fell from 18.3% to 14.8% of GDP.
Turning to savings, ideally one would wish to have direct measurement of total savings for the second quarter of 2009, which were not published with the GDP figures, and figures for the US balance of payments for the same period – which are also not yet published. Fortunately, however, the shifts in the US trade balance are so large that it is relatively easy to work out the trends.
Savings equal the sum of total investment, for which full US figures are available, minus the balance of payments deficit – for which second quarter figures are not yet available. But shifts in the US balance of payments are dominated by changes in the trade balance. Provided, therefore, that it is being used to establish a qualitative direction of change, and is not projected as an exact statistical calculation, it is perfectly possible to use the major shift in the US trade balance to show the change in the direction of US savings.
If 2008 is taken as the year in which the financial crisis unfolded then the US balance on trade in goods and services fell between the last quarter of 2007 and the second quarter of 2009 from 4.9% of GDP to 2.5% - an improvement of 2.4% of GDP (see Figure 2).
It follows from the data above that US investment has fallen by 4.3% of GDP since the beginning of the financial crisis while the US balance of trade has improved by only 2.4% of GDP - a difference of 1.9% of GDP. If all other components of the balance of payments had remained the same then this would necessarily mean that US savings had also declined by 1.9% of GDP – if savings had remained static, and other components of the US balance of payments had remained constant, then a 4.3% of GDP fall in investment would have translated into a an equivalent 4.3% improvement in the balance of payments figures. The 1.9% of GDP gap between the 4.3% decline in investment and a 2.4% fall in the balance of payments would necessarily mean that US savings had fallen by 1.9% of GDP.
Evidently no such precise quantitative assertion can be made as what is being calculated above is the trade balance and not the overall balance of payments. But it means that for US savings not to have fallen components of the US balance of payments other than trade would have had to improved by 1.9% of GDP or $269 billion. This is completely implausible and it is therefore evident that US savings have been falling.
This is confirmed by taking the latest figures for which there is measured data on total savings, that is for the first quarter of 2009. Between the fourth quarter of 2007 and the first quarter of 2009, US total savings fell from 13.9% of GDP to 11.5% of GDP. These trends of falling US savings are shown in Figure 3.
The reason some media commentators have claimed US saving is rising, when it is actually falling, is because they confuse household saving (which rose from 1.1% of GDP to 4.0% of GDP between the last quarter of 2007 and the second quarter of 2009) with total saving (the sum of household, company and government saving). The rise in US personal saving is however being more than offset by the decline in company and government saving – hardly surprising given the scale of the US budget deficit. Robert Skidelsky's statement, cited above, that US saving is rising is therefore inaccurate - it appears he may be making an incorrect generalisation from personal saving to total saving.
Michael Pettis claims that US consumption is falling more rapidly than GDP, which would imply saving is rising, but unfortunately makes two errors – first he confuses personal consumption with total consumption, and second he fails to note that shifts in relative prices mean that although in volume terms US personal consumption fell more rapidly than GDP in the second quarter of 2009 it increased as a percentage of GDP. From the point of view of global imbalances, that is the US balance of payments deficit, it is the proportion of GDP devoted to consumption which is determining and not movements in volume.
It is therefore clear that the first part of the 'global imbalances hypothesis' regarding the US is wrong. The imbalance of the US balance of payments deficit is not falling because US saving is rising but because US investment is falling even more rapidly than US saving is falling. Now consider China.
China's rising investment and declining trade surplus
While the macro-economic data available for China for the second quarter of 2009 is not as detailed as for the US nevertheless, again, the shifts are so large it is relatively easy to ascertain the trends.
The first trend is that China's investment is rising as a percentage of GDP. Second China's balance of payments deficit is declining. These will be considered in that order.
First, taking the rise in China's investment as a percentage of GDP, the components of the 7.1% rise in GDP in the first half of 2009 were 6.2% rise in investment, 3.8% rise in consumption, and minus 2.9% fall in net exports. Unless China's savings were increasing equivalently such a rise in the percentage of investment in the economy necessarily means that China's balance of payments surplus must fall.
As with the US balance of payments figures for China for the second quarter of 2009 are not yet available. But the drop in the trade surplus, which dominates China's balance of payments position, is of sufficient magnitude that it is clear that China's balance of payment surplus is falling.
Figure 4 shows China’s monthly trade surplus since 1992 up to June 2009. Figure 5 shows the same data calculated as a three monthly moving average in order to avoid any purely short term distortions.
The trend is clear. China#s trade surplus rose steadily from 2005 onwards and then temporarily rose even further under the impact of the onset of the international financial crisis in September 2008. The peak was reached in January 2009 with a monthly surplus of $42.1 billion. Since then China's surplus has fallen steadily. The surplus for June was $8.25 billion.
Expressed in terms of three monthly moving averages China’s monthly trade surplus was $22.5 in August 2008, immediately before the collapse of Lehman brothers, rose to $38.1 billion in January 2009, and has since dropped to $12.5 billion.
Shifts in other components of China's balance of payments sufficient to offset the rapid decline in the trade surplus are not credible so it is clear that China's overall balance of payments surplus has shrunk - meaning the gap between China's savings and investment has narrowed.
While savings figures are not available for China at present it is clear from the data above that they have not risen to match China's rise in investment. There are reasons to believe China's savings may have declined somewhat – China's budget is projected to move into a 3% of GDP deficit, profitability of export and other industries is under pressure from the international financial crisis, and given a 15% rise in retail sales there is no reason to believe household saving has risen significantly (if at all).But it is clear that China is fundamentally responding to the financial crisis by raising its rate of investment. The result has been the most rapid rate of growth in the world.
The logical lacunae in the global imbalances hypotheses, that it did not point out that China could just as much reduce its balance of payments surplus by increasing investment as by reducing saving, is therefore the actual course of China's economic policy - with the most successful results of any country in the world.
Indeed it is quite probable that this year, in net terms,the whole of world growth will be accounted for by the expansion of China's economy. Compared to this level of economic success all discussion of 'green shoots' in other economies is insignificant.
The errors of the 'global imbalances hypothesis'
For the reasons set out above it is therefore not material that the 'global imbalances hypothesis' is conventional wisdom - many things that are conventional wisdom turn out to be false, nor that a number of those supporting it are outstanding economists, nor that Martin Wolf is one of the world's outstanding economic journalists with a deep knowledge of economic statistics etc. A theory which lacks internal coherence, which is factually wrong, and which leads to wrong policy prescriptions is a theory that does not meet the test of scientific rationality. It should therefore be set aside.
Instead the realities of the world economy should be recognised. The US balance of payments deficit has been shrinking not because US saving has been rising but because US investment has been declining more rapidly than US saving has been falling. China's balance of payments surplus has been declining primarily because its investment level has been rising. That is, the logical gap which existed in the 'global imbalances hypothesis' corresponds to the actual course taken by the world economy. As always when the facts and a theory do no coincide it is the theory which should give way.
* * *
This article by John Ross originally appeared on the blog Key Trends in Globalisation.
 The wide circulation of this analysis may be traced to a speech by Ben Bernanke, now Chairman of the Federal Reserve,
 Breaking these figures down into their detailed components US personal consumption rose from 69,9% of GDP to 70.6% between the fourth quarter of 2007 and the second quarter of 2009. In the period from the second quarter of 2008 to the second quarter of 2009 US personal consumption rose from 70.3% of GDP to 70.6%. US personal consumption increased between the first and second quarters of 2009 from 70.4% of GDP to 70.6%.
To calculate precisely how much US government consumption has risen it is necessary to note that the US is unusual in that in its main aggregate GDP statistics it groups together government consumption and government investment – most countries statistically treat government investment simply under overall investment. If the two components (consumption and investment) of US government expenditure are taken together they increased from 19.2% of GDP in the last quarter of 2007 to 20.7% of GDP in the second quarter of 2009. It is necessary to eliminate from this the rise of government investment from 3.3% of GDP to 3.6% of GDP in the same period. Government consumption rose from 15.9% of GDP to 17.0% of GDP. In the period since the second quarter of 2008 US government consumption has risen from 16.4% to 17.0% of GDP.Government consumption rose from 20.3% to 20.7% of US GDP between the first and second quarters of 2009.
Considering, therefore, both total consumption and its breakdown the rising proportion of consumption in US GDP since the beginning of the financial crisis is clear – indeed consumption has risen as a proportion of US GDP both as regards personal consumption and government consumption.
 Private fixed investment declined from 15.8% to 12.3% of GDP, inventories fell from plus 0.1% of GDP to minus 1.1.% of GDP, and government investment rose from 3.3% to 3.6% of GDP.