Global imbalances and the paradox of thrift

W. Max Corden
Abstract.
This paper analyses the relationship between the global imbalances and the financial crisis.
The imbalances were connected with the increase in world saving emanating from the ‘saving
glut’ countries, notably China. This increase in saving led to a decline in world interest rates,
and thus to greater borrowing, especially in the United States. This borrowing was for
financing consumption, wars, and unwise rather than fruitful investment, especially in
housing. The failure to invest in fruitful investment led to the financial crisis, and this the
decline in US and world-wide aggregate demand.
This was the indirect paradox of thrift. It is to be contrasted with Keynes’s paradox of thrift,
where the decline in aggregate demand and output would have resulted directly from the
rise in the world savings.
The paper also discusses why there was not more borrowing for fruitful investment,
especially in developing countries, hence avowing the financial crisis.
I
Introduction
The Global Imbalances are fashionably and officially perceived as a problem, especially by the
US government and US economists. There have been proposals that they should be
internationally regulated and, above all, reduced. These “Imbalances” are national current
account surpluses and deficits offset by net capital flows between countries, such flows
including changes in foreign currency reserves. It is often stated that they are a cause –
possibly even the principal cause – of the global financial crisis that began in 2008. Particular
criticism is directed at surplus countries. The aim of this paper is to analyze these views
rigorously.
I start with the simple proposition that such imbalances reflect international intertemporal
trade, an idea developed in Obstfeld and Rogoff (1996, chapter 1) and Corden (2007). Since
one normally assumes that trade has benefits and possibly that “free trade is best” (subject
to well-known qualifications) one needs to ask: what is so special and bad about cross-border
flows of capital and hence international intertemporal trade? Since we are dealing here with
a particular kind of “trade” an insight into the issues can be obtained by applying ideas from
normative trade theory.
In sections I and II below I expound in some detail the international intertemporal trade
argument and how international equilibrium is obtained, leading to the provisional
conclusion that there is no problem about the Global Imbalances, this being parallel to the
argument that “free trade is best”. I shall call this the neoclassical approach. The rest of the
paper develops some qualifications which help to explain why there is a widely held concern
about actual or at least “excessive” global imbalances. This analysis is parallel to various well-
known arguments that qualify the case for free trade. The whole of this paper is concerned
with making explicit what is often just left implicit.

II
Intertemporal Trade and the Return Journey
The countries with current account surpluses are net exporters of goods and services in
exchange for imports of financial instruments (call them “bonds” for short), while deficit
countries are net importers of goods and services and exporters of “bonds”. But such trade is
not sustainable because the surplus countries are foregoing goods and services today but
expect, in return to receive net goods and services tomorrow. This is, what I shall call, “the
return journey”. They expect a return in the form of interest and dividends, and possibly,
return of capital. The required return journey is integral to the intertemporal feature of this
kind of trade. In the case of a country that has been a net exporter of capital, like Japan or
China, this return journey involves eventually a shift from a surplus in ordinary (non-interest)
trade to a deficit.
The exports of capital (purchases of bonds) by the surplus countries are not foreign aid. It is
thus quite reasonable that some of the surplus countries – the net capital exporters, like
China or the oil exporters – are relatively poor while deficit countries, notably the United
States, are rich. Furthermore, flows of capital reflect not only differences in savings
propensities but also differences in investment opportunities, and these depend on many
factors, including differences in total factor productivity.
Thus a capital-rich country may well have a higher marginal productivity of capital even
though the ratio of capital to labour is much higher than in capital-poor countries. Hence a
capital-rich country like the US is importing capital from capital-poor countries.

What is wrong with international, intertemporal trade? Are there not “gains from trade”?
Within countries intertemporal trade goes on all the time. Some regions of a country are net
capital exporters (hence have current account surpluses) and others are net capital
importers. Individual savers in the form, for example, of contributors to pension funds, are
lenders and hence capital exporters while corporations that borrow for investment or issue
stock are capital importers. If such a flow of funds crosses international borders it may
contribute to the much-maligned global imbalances.
There can also be surprising cross-border flows. Corporations in China are high savers and
households in the United States have been low savers. Thus funds flow from Chinese lenders
to US borrowers. Other cross-border flows are not at all surprising. Funds flow from Japanese
households and corporations that are high savers, partly for demographic reasons, to
Australia where population growth is higher and where perceived investment opportunities
(in the mineral industry or in housing) are greater.
III
Free Intertemporal Trade and International Equilibrium
I use as a reference point the current situation where, with respect to current account
imbalances, every country does essentially what it wants – or its government, corporations,
and households want – and the countries interact through the international general
equilibrium system. Essentially, what I am presenting here is an international neo-classical
story.
In each country decisions are made by numerous private agents and by the central bank and
the government, the latter two acting through monetary and fiscal policies and often through
exchange rate intervention. If a government wishes to reduce a current account deficit, for
example, while maintaining the nation’s internal balance it can do so through fiscal
contraction combined with monetary expansion, one effect being depreciation of the
exchange rate. Depreciation can also be brought about by direct intervention in the foreign
exchange market, as well as by controls on capital inflows. Mostly governments do not
actually target their current account balances.
These are by-products of a variety of independent decisions by private and public actors,
influenced also by the decisions of other countries.
International equilibrium – which ensures that the sum of current account surpluses equals
the sum of deficits – is brought about through the capital market. For example, if country
group “A” increases savings, with investment unchanged, so that its current account surplus
is increased, lower world interest rates or increased credit availability will increase spending
in country group B and so increase the latter’s current account deficits.
This system has two benefits. I use here the same arguments that are used to describe the
benefits of a system of free trade. First, decision-making is decentralised, so that the usual
problems of central planning – especially international central planning or coordination – are
avoided. Since such central planning or coordination is difficult to bring about, it is indeed an
advantage that it may not be necessary. Second, the benefits of the “gains from trade”
resulting from different comparative advantages, are realised. For example, there are gains
from intertemporal trade when Japanese savings finance, to some extent, investment in
Australia rather than Japan because of initial differences in expected investment returns in
Australia compared with Japan.
Similarly, there may be gains from intertemporal trade when country A has an age
distribution that yields a high rate of national savings while country B has a distribution that
yields, at least temporarily, a lower rate of saving. Both countries may gain when excess
savings are exported from A to B. Market forces, if unrestricted, will bring this about and
residents of both countries may gain.
I now turn to the many possible qualifications to this simple free trade or neo-classical
argument which I have applied to international intertemporal trade. Here one may find some
rationales for the common concern with global imbalances.
IV
Borrowing for Consumption, for Wars and for Unwise Investment
Consider the following simplified story which describes what happened internationally in a
period beginning approximately in 2003 and culminating with the 2008 crisis. Savings go up
more than investment in a group of savings glut countries – the group consisting of Japan,
China, Germany, the oil exporters, and some other smaller European and Asian countries. In
some countries (notably Japan and some other East Asian countries) private investment
actually declined. This increase in savings relative to investment lowered world real interest
rates and (backed up by central bank policies) made credit more readily available all over the
world, notably in the United States. So there was a borrowing boom, especially in the United
States.
The borrowing was, above all, through mortgages and led to a housing boom. It financed
housing construction, which is a form of investment, and it lead to an asset boom which, in
turn, stimulated private consumption. Housing construction was excessive, so that it can be
described as “unwise investment”. Easy credit also stimulated private demand in other forms.
Encouraged and supported by government agencies, “sub-prime” housing loans became
common in the United States. In addition, and in fact before this private sector boom, the US
government borrowed to finance the Bush tax cuts and the Iraq war. All this is a familiar story
by now.
The internationally free flow of capital thus eventually created a debt crisis. This was initially
a private sector crisis in the United States and some other developed countries. In the case of
the US government, it also helped to finance and thus sustain for a time a potential well-
recognized government fiscal problem. Perhaps more important, the eventual recognition of
the situation lead to a crisis for the world-wide, but especially the US, financial sector. In the
financial sector the low interest rates encouraged a “search for yield”, in fact a willingness to
run more risks in the hope or even belief that this will yield higher returns.
The heart of the problem to which it gave rise was that borrowing financed increased current
consumption, unwise investment (housing) and, in the case of the US government, current
warfare. If borrowing had been for sound investment – with good prospects of being fruitful
– it would have been expected to provide for the return journey – for the payment of
interest, dividends and future repayment of debt. For the international neo-classical system
which I have described earlier to work smoothly there has to be an expectation that the
borrowers will be able to pay interest or dividends, and gradually to repay their loans.
Investment in itself is not enough; it must be investment with reasonable expectations of
good returns. When the US housing market crashed it became evident that the investment
had been largely unsound. Hence a private sector debt crisis resulted, and spread from the
United States to other countries, primarily in Europe.
One must add that borrowing for consumption may be acceptable under certain
circumstances, notably if it is expected to be temporary and if the borrowing country’s
savings are expected to increase in due course. Thus borrowing for consumption is sensible if
the needs are seen to be temporary (as during 2009-10 when there was a recession), or
during a war, or if there is a sufficiently high underlying productivity growth rate.
In this recent episode, culminating in 2008, the effect of borrowing for consumption (or for
unsound investments, especially in housing) was to generate an eventual expectation that
debt service could not be maintained, and thus caused a debt crisis, with damage especially
to the financial sector itself.
The savings glut countries made available more resources to the rest of the world, notably
the United States. One might regard this as a benefit to the rest of the world.
But these resources were loans and not gifts. Thus their acceptance by the potential
borrowers and their financial intermediaries implied an awareness of the required return
journey.
V
Did Global Imbalances cause the Crisis?
The crisis was caused by the interaction of two factors. The first was the sharp but quite
prolonged decline in world real interest rates and increase in credit availability.
The second was the inadequacy of the US – or perhaps, better, North Atlantic – private
financial sector. I would assert that without any one of these two factors there would have
been no crisis. With regard to the second, I will elaborate on it shortly.
Let me first analyze the first.
It was a particular global imbalance that contributed to the crisis. This was the significant
excess of the increase in savings over the increase in investment in the substantial group of
savings glut countries. This led to the fall in the world real interest rate and increased credit
availability that contributed to the crisis.
Suppose there had been a significant group of countries outside the United States where a
change had caused investment to exceed savings. This would also have led to a global
imbalance but this time associated with a rise in world real interest rates. It might have led to
a reduction in the US current account deficit and possibly even a surplus. That would not
have led to a crisis, or at least not a crisis of the kind we have had. Alternatively, suppose
that, starting with a US deficit there had been an increase in US savings (as since 2008) with
no initial change in other countries. This would have actually reduced the global imbalance
affecting the United States, but would still have led to a reduction in world real interest rates
both within the United States and in the world as a whole. It might therefore still have
provided the conditions for a crisis by leading to over-borrowing.
Coming now to the second causal factor of the crisis, namely the inadequacy of the US
financial sector this has been widely discussed and I have little to add, except to relate it to
the basic trade theory approach in this paper.
Basically the world’s financial sector (primarily the US sector) misallocated the additional
resources made available by the world’s savers. The sector showed inadequate risk aversion,
excessive short-term thinking, and (arguably) general incompetence1. Perhaps individual
agents in the sector were perfectly rational, but faced incentives that led to damaging results
for the sector as a whole, and indeed for whole economies. New financial instruments were
developed which were barely understood.
If there had been no capital imports (no international intertemporal trade) into the potential
deficit countries, notably the United States, savings that originated domestically would still
have been poorly allocated, given the financial sector’s inefficiency. Restricting capital inflow
would have raised domestic interest rates, led to increased domestic savings by some
elements in the economy, and these additional funds would still have been misallocated to
finance dissaving by others. In addition some gains from intertemporal trade would have
been lost.
In trade theory language, this inefficiency of the financial sector was a “domestic
distortion2”. First best policy would require that the distortion or inefficiency in the financial
sector be reduced or eliminated. This would have benefited resource allocation of savings
originating both from abroad and from home. Just restricting intertemporal international
trade would have been second best. The policy focus since the crisis on the reform and
improved regulation of the financial sectors in the United States and the United Kingdom has
thus been correct and (in trade-theory terms) first best.
In arguing for free trade one assumes that buyers and sellers in different countries know
what is good for them – or at least one leaves it to them to decide, thus achieving the
advantages of decentralized decision-making. Yet here we may have a case where key agents
in the borrowing country, especially the United States, did not in hindsight follow optimal
policies from the point of view of the country’s own aggregate national interest. Thus the
standard assumption of the free trade argument does not apply. In the US the private and
the public sectors borrowed while inadequately considering the future implications – i.e. the
need for a “return journey”.
Of course, this is just one possible point of view. Perhaps, on balance, it was sensible to
borrow when the interest rate was so low and credit was so easily available. The Bush
administration was strongly committed both to a war and to tax cuts – in historical terms an
unusual combination – as well as to fostering widespread housing ownership. Perhaps some
decision-makers, private or public, expected their employer or even the government to
default, in which case the funds provided by the savings glut countries would turn out to be
gifts rather than loans. Furthermore, some individuals in the financial sector extracted
personal gains at the expense of their employers, their creditors, or their governments.
VI
The Paradox of Thrift
I now come to an alternative model, seemingly very different from the neo-classical one I
have been using so far. This is Keynes’ model or idea of “the paradox of thrift3”.
It makes sense of much popular discussion of the global imbalances and especially of the
criticisms of surplus countries. The idea was applied by Keynes to a single country. Here it
needs to be extended to the world economy.
An increase in savings is motivated by some people, corporations or governments wishing to
consume less today, for the sake of more tomorrow. This manifests the admirable Victorian
virtue of prudence – providing for the future – a virtue currently very prevalent in East Asia
and in Germany. But, Keynes pointed out that, on its own, an increase in savings only reduces
current aggregate demand. Savings in themselves do not increase a country’s capacity to
produce, and hence do not supply the resources needed to provide the extra future
consumption that the savers expect. This provision for the future requires not just saving but
also extra investment, which is induced in the neo-classical model through the decline in the
interest rate. Hence savings must be channeled into investment, this being the responsibility
of the financial sector.
In Keynes’ view investment depended on many factors, notably expectations, “animal
spirits”, and also current consumption (seen as a guide to future demand), but not much on
the rate of interest. Thrift is unlikely to lead to more output in the future, but instead – by
reducing aggregate demand – would lead to less output in the present.
Thus savings that are expected to make people eventually richer will actually make the nation
currently poorer, without any benefit for the future. That is the paradox.
The key feature of this Keynesian model is the failure of the interest rate to equilibrate the
system by stimulating investment when savings rise. Keynes, of course, was influenced by the
situation in the nineteen thirties when a lack of aggregate demand was obvious. During that
period animal spirits were lacking.
Let us now expand this approach to the world economy, with an international capital market.
The paradox of thrift requires two crucial assumptions. The first is that there is no direct link
between worldwide saving and investment through the rate of interest or, if there is some
link, it is inadequate. Secondly, maintaining aggregate demand is a world-wide problem. One
then arrives at the following conclusion. Spending, whether on consumption or investment, is
good, while saving (not spent directly on investment) is bad. Spending will increase aggregate
demand while saving will reduce it.
From an international point of view, high savers are to be disapproved of, unless they spend
their savings on domestic investment. In other words, it is the excess of saving over
investment – the current account surplus – that is bad. This explains why there is a concern
not just about saving but about “global imbalances.” If a country’s bad policies (high savings)
are offset by good policies (domestic investment), then there is no adverse international
effect. Furthermore, if a country’s investment exceeds its saving, or indeed if the country
dissaves as well as invests, it will have a current account deficit and that, presumably is a
service to the world from this narrow point of view.
Given this theory or paradigm, current account surplus countries – notably China, Japan and
Germany – where people feel virtuous because of their prudent saving for the future –
should be told that it is investment, not saving, that raises future output.
When they rely on other countries to do their investing for them, they are really outsourcing
the difficult part of the story. Of course, these countries, notably China, are also big, indeed
very big investors, either now or in the past, and here we are only concerned with the excess
of saving over investment. In China saving has been about 50% of GDP and investment has
been about 40%, so that the current account surplus in 2008 was about 10%, which, of
course, is a high figure.
This Keynesian “paradox of thrift” theory or set of assumptions makes sense of the popular
view – held even by leading economists, notably Martin Wolf and Paul Krugman – that
China’s current account surplus is harmful to the world economy.
How does the paradox of thrift story relate to the story earlier of borrowing for consumption
and for unwise investment? During the savings glut period there was no aggregate demand
(or Keynesian) problem in the world until the crisis of 2008. High net savings coming from the
savings glut countries did lead to reduced world interest rates, and this led to borrowing
primarily for housing construction and for consumption in other countries, especially the
United States. I shall focus on the consumption increase here. In effect, increases in saving in
some parts of the world led, through the adjustment process of the interest rate, to reduced
saving or dissaving in other parts of the world. Since aggregate demand was maintained
there was actually no paradox of thrift in the strict Keynesian sense. Nevertheless, something
that was required by our basic neo-classical model was indeed missing. Thus the “paradox”
model is indeed relevant to our discussion.
It can be shown that there was an indirect paradox of thrift. In an indirect way, the saving glut
did lead to an aggregate demand problem. This to be contrasted whit the Keynesian or direct
paradox of thrift. Because there was not a sufficient increase in investment, defined as sound
or fruitful investment, there was insufficient provision for the return journey – and a series of
debt crises, particularly in the private financial sector, resulted. There was, of course,
investment in housing construction, notably in the United States and in Spain, but there was
a surprising failure for investment more generally to increase. I shall come back to this crucial
matter later. Worldwide aggregate demand was maintained primarily at first by US
government fiscal expansion (to finance a war and tax cuts) and then by US private sector
growth in consumption demand plus housing construction.
The eventual decline in worldwide aggregate demand – especially in the United States – was
caused essentially by the financial sector debt crisis, and this in turn was caused by increased
savings in one part of the world being channeled to borrowing for consumption or for
unfruitful investment in another part. Thus the initial effect was not for aggregate demand to
decline – the Keynesian paradox of thrift – but for a financial sector debt crisis to be
incubated.
The eventual debt crisis to a decline aggregate demand (and indirectly to a decline in sound
investment) for two reason. First, the losses of the private financial sector to reduce the
willingness of that sector (particularly banks) to act as an intermediary between the
monetary authorities and the potential non-financial borrowers. In others words, the supply
of sound to the non-financial sector severely declined. Second, the heavily non-financial
sector (especially the household sector in the United States) reduced its demand for new
funds owing to its desire to reduce its indebtedness to improve its balance sheets. This
created a ‘balance sheet recession’ (Koo, 2008).
The blame must be put primarily on the inefficiency of the world’s, especially the US’s –
financial sector. More of the funds saved in the savings glut countries should have gone into
equity financing rather than into debt financing, and, above all, more should have financed
fruitful investment rather than consumption. Much of financing of housing in the United
States must also be regarded as unfruitful. It failed to provide for the return journey because
the sub-prime mortgagees would not be able to maintain their mortgage payments.
To summarize, there are two versions of the paradox of thrift paradigm. There is the
Keynesian direct version and there is the indirect version that operates through the
intermediation of a debt crisis. The events of recent years are best explained by the indirect
version. Greater thrift in the savings glut countries failed to lead to increased investment -
fruitful or sound investment - that would provide for the return the journey. This then led to
the debt crisis, and thus a drastic decline in aggregate demand.

With regard to the use of borrowed funds for financing a war by the US government, it might
be argued that it has been normal in history for governments to borrow massively during
wars. But they have not normally reduced taxes at the same time.
This experience raises a further important historical or counter-factual question. Let us take
the savings glut effect as given. Suppose the increase in borrowing in the United States for
consumption and for housing had been much less, would the gap have been filled by
borrowing for fruitful investment in the United States and elsewhere? Perhaps this is the
central question about this historical episode which ended in 2008. Were suitable investment
projects, private or public, available, or could they have been developed? Could USA and
European private corporations have borrowed more? In particular, was there an excessive
reluctance to borrow for investment in emerging market countries? I come to this issue now.
VII
Aversion to Current Account Deficits: Instability of Capital Flows

Some of the best investment opportunities may well exist in developing countries.
Some of these countries should generate the capital inflows and hence current account
deficits that would balance the surpluses of the savings glut countries – or they should have
balanced them in the period up to 2008. There are two reasons for their reluctance to allow
substantial current account deficits to develop. These may explain why such a large
counterpart to the surpluses has been the US deficit, rather than the deficits of developing
countries, and why the possibility of a worldwide Keynesian lack of aggregate demand
problem resulting from an initial savings glut cannot always be ruled out.
The first reason is the sad experience of instability of international capital flows, especially
into developing countries. However justified initially the investments that are the reasons for
the capital inflows, the inflows tend so often to overshoot and then suddenly come to a stop,
creating a crisis. The 1997 Asian crisis is the best example of this story. In particular, bank
lending tends to be extremely volatile and subject to herding behaviour. Many governments
are now wary of capital inflow booms when such booms can suddenly – and perhaps
irrationally – come to an end and force contraction in demand and real depreciations, all in
crisis conditions. These features reflect, again, the inefficiency of the world’s financial sector.
This instability of capital flows into developing countries also explains why so much of the
flows from the savings glut countries went to the United States. The memory of the Asian
crisis, and indeed also various Latin American crises, must have encouraged caution in many
emerging market countries. By contrast the United States was seen as a safe haven in spite of
its obvious potential fiscal problem. There were dire forecasts before 2008 of a likely dollar
crisis because of the US current account deficit, but the first reaction to the crisis was for the
dollar actually to go up in value. Thus there was indeed a crisis, but not the one that had
been widely predicted. The reasons why the United States was seen as a good country to
lend money to were discussed in Cooper (2007), a point of view that was criticized in Wolf
(2008). Clearly it helped that the US dollar was the world’s key currency. Unlike most
developing countries it could borrow in terms of its own currency.
VIII
Aversion to Current Account Deficits: Unpopular Real Appreciations
The second reason for the reluctance of many countries to allow substantial capital inflows –
which would inevitably generate current account deficits – is that such inflows are inevitably
associated with real appreciations. These have adverse effects on the tradeable sectors of
economies. Real appreciation is particularly undesirable when, as often, it is likely to be
short-term owing to the capital market volatility just mentioned, and thus will soon have to
be reversed. Real appreciation resulting from private sector capital inflows can be avoided or
at least modified by sufficient budget surpluses at the same time. Alternatively capital
controls can reduce or slow up the inflows. Exchange rate intervention requires to be
sterilized, otherwise inflation would result, so that there would still be real appreciation even
when the nominal exchange rate is kept fixed by the intervention. In practice all these
counteracting measures have problems.
The main point is really this. Potential gains from international intertemporal trade may well
justify capital inflow. The inflow may finance investment that yields a high rate of return, so
that there are mutual gains from intertemporal trade – i.e. from foreign savings financing the
country’s investments. Both foreign savers and domestic industries where the investments
take place gain. Nevertheless, in the capital-importing country there can still be losers from
such intertemporal trade. These losers are the industries and workers in the export and
import-competing industries adversely affected by real appreciation. This is the Dutch
Disease effect.
On the basis of the standard theory of the gains from trade, given the relevant assumptions,
the country as a whole does gain from free intertemporal trade, in the sense that gainers
could compensate losers, and a net gain would remain. But in the absence of compensation
there are inevitably losers, and this applies to intertemporal trade as much as to “ordinary”
trade.
Through real appreciation capital inflow thus reduces employment in import-competing and
export sectors. But there is not necessarily an overall decline in employment but only a
redistribution of it. Capital inflow increases employment in the non-tradable sectors. This
increase in non-tradable employment is generated by the extra domestic spending brought
about by the capital inflow. The crucial mistake is often made to focus exclusively on the
adverse effects in the import-competing and export sectors, and to ignore the offsetting
favorable effects in the non-tradable sectors.
This analysis applies not only to developing countries but also to the United States. It was
widely, but falsely, believed in the United States that “China” (meaning really all the savings
glut countries) caused unemployment in the United States) during a period before 2008
when there was actually no increase in overall unemployment.
I would put the matter differently. As a result of the international general equilibrium
adjustment to the savings glut countries the US developed a current account deficit that was,
in part, the mirror image of the savings-glut countries’ surpluses. The US deficit was financed
principally by capital inflow from China and other savings-glut countries. The by-product
within the United States was a loss of employment in some sectors and a gain in others. Of
course the overall employment effect was also influenced by US monetary and fiscal policy,
and other much discussed factors (such as technological advances). The mistake I have just
referred to was also made in US discussion. The focus was on the negative employment
effects in import-competing industries while ignoring the positive employment effects in
areas financed or stimulated by capital inflow.
IX
Savings too high, or Investment too low?
Let us list the four ways in which an ex-ante excess of world savings over world investment
can be resolved.
1. Aggregate demand declines until aggregate savings decline to the given level of
investment. This is the Keynesian “paradox of thrift” case. To some extent this has happened,
but only since the 2008 financial crisis. It did not happen during the savings glut period.
2. The interest rate falls, and so investment rises, maintaining aggregate demand. This is the
hypothetical outcome of the neo-classical model and has not actually happened since the
beginning of the “savings glut”. (I refer here only to “fruitful” investment).
3. The interest rate falls, leading to increased borrowing for consumption. This thus leads to a
decline in the world savings propensity which offsets the effects of the initial savings glut. A
decline in aggregate demand is avoided. This did actually happen but eventually led to the
private sector debt crisis. (I also include “unfruitful investment” under this rubric.)
4. The decline in aggregate demand is moderated by Keynesian fiscal expansion. Thus a
Keynesian paradox-of-thrift problem is handled with a Keynesian solution. Depending on the
extent of existing public debt and whether public borrowing is used for consumption or
investment, this may (or may not) lead to a sovereign debt crisis.
Solutions 1. and 3. are clearly not satisfactory. Solution 2. should have happened, but did not.
Solution 4. may be satisfactory in the short-run, and even in the long-run if budget deficits
finance sound investment.
And so we come to Solution 5. It is to urge or pressure some of the high savings countries,
above all China, to save less. This has certainly been tried by the US government and US
economists, but not with success7. It is a policy proposal that seems to follow naturally from
Keynes’ paradox of thrift and assumes that options 2 and 4 are ruled out.
Here I wish to highlight – without really answering – a central question about recent
experience. Why did option 2 (the neo-classical solution) not happen? With interest rates so
low and credit so readily available why was there not more private and public investment in
both developed and emerging market countries? I refer to the period from roughly 2000 to
2007 or 2008. I am thinking particularly of infrastructure investment. Such investment,
financed by foreign borrowing, would have provided the neo-classical solution and perhaps
avoided so much “borrowing for consumption” and for “unfruitful” investment (notably for
housing) in the United States.
Given the needs for long-term investment for both demographic and environmental reasons,
it was an opportunity missed. Part of the answer is contained in sections VI and VII above,
namely the aversion of many developing countries to incurring current account deficits. I
exclude China from this question since it actually had a massive investment boom (even
though domestic investment was less than savings, hence yielding its current account
surplus).
X
Conclusion: The Main Point and the four Qualifications
To conclude, the central argument developed at the beginning of this paper must be
emphasized. The imbalances represent international intertemporal trade, and normally one
would expect this form of trade to yield gains from trade benefiting both the lenders (the
saving glut countries, primarily) and the borrowers (above all, the United States). The
common view that the imbalances are, in some sense, basically or a priori undesirable, hardly
makes sense.
Yet there are qualifications. Firstly, the borrowers must be prepared for the return journey –
that is, the inevitable need for repayment, or at least payments of interest and dividends.
Otherwise a debt crisis will, or may, result. In effect, most of US private borrowing was to
finance consumption, or housing construction in excess of likely demand. On the other hand,
to the extent that US borrowing financed a Federal government budget deficit that was
politically determined and was not a response to the availability of foreign purchases of
Treasury bonds, there was a clear benefit to the United States through having to pay lower
interest rates on its bonds than if foreign funds had not been available.
A second qualification is that the inflow of capital into the borrowing or potential borrowing
countries would tend to bring about real appreciations of the currencies of the borrowing
countries, and this has often been thought undesirable. It has certainly been perceived as a
problem. I have discussed this effect at length. There is sometimes a misunderstanding
(notably in the United States) that reduced employment in the export and import competing
sectors represented a net loss of national employment.
Thirdly, if some countries have current account surpluses others have to run deficits, the
latter brought about by capital inflows. But such capital inflows can be very unstable, and
quickly reversed. Hence many policy-makers are, for good reasons, averse to deficits, and
hence seek to control or limit inflows. The fault here is basically with the international
financial sector that creates the instabilities.
A fourth qualification is that high savings may lead to a paradox of thrift, in the sense that
reduced interest rates caused by higher savings may fail to stimulate sufficient demand from
potential borrowers, so causing unemployment. If the return journey problem is not to arise
this demand must be for financing fruitful investment and not consumption. Here I have
noted that, for various reasons, in recent years the demand for financing fruitful investment
has been inadequate. Perhaps this has been the heart of the problem.