Savings, Investment and Growth: Theory and Reality

Neoclassical economic models are based on the assumption that investment is financed from household savings.  Accordingly, capital accumulation will be maximized by policies aimed at increasing household savings rates and capital imports (“foreign savings”).  These models also predict that capital should flow from rich to poor countries, attracted by higher rates of return.

However, facts do not support the predictions of these models.  Following the Asian financial crisis of 1997-1998, many developing countries have registered strong improvements in their external balances and their aggregate current account has swung into surplus.  As a result, they have become net exporters of capital to developed countries.  The fact that capital is flowing “uphill”, from poor to rich countries, is not consistent with mainstream neoclassical economic theory.  What is even more surprising, in light of mainstream theory, is that developing countries that are net exporters of capital also tend to grow faster and to have a higher investment ratio than developing countries that receive net capital inflows.

These facts have been considered “puzzles”, but they are no longer puzzling if one recognizes the limitations of the underlying theories.  The 2008 TDR argues that one main problem with neoclassical economic theory is what Kaldor (1983) called its “anthropomorphic nature”.  That is, the tendency of using a representative agent model to describe the behaviour of a complex economy, with millions of agents with diverging interests and many fallacies of composition.  For example, and very often forgotten, prices only clear markets if supply and demand are determined independently.

This is not the case for one of the most important prices, that of labour.  Wages are a cost factor, and thus influence the supply of goods and services, but they also determine the income of the largest segment of the population and thereby influence the demand for goods and services.  In the same vein, an individual agent may reduce its consumption in order to invest more.  However in a complex economy, where investment and savings decisions are made independently by different actors, higher savings (in the form of putting money aside) do not lead to an increase in investment, but rather may trigger a fall of investment.

In neoclassical models, the assumption of full employment prevents a fall in demand brought about by an increase in the savings rate, which amounts to saying that employment cannot fall because it is assumed that it cannot fall.  In the closed economy version of the model, higher savings lead to an immediate reduction in the interest rate, and firms react by increasing their investment despite falling profits since they supposedly have perfect foresight and anticipate higher growth in the future.  This implies that firms increase investment even as involuntary inventories rise and their capacities are not fully utilized.  It is hard to think of a real world entrepreneur who would behave in this way.  On the contrary, unless companies have accurate information about the future, they will react to lower demand and falling profits by reducing investment, thereby reducing income.  Keynes (1936) convincingly argued that the decision “not to have dinner today” may depress the business of preparing dinner today without immediately stimulating any other business.

An alternative view, based on the works of Schumpeter and Keynes and consistent with the experiences of post-war Western Europe and the successful catching-up experiences in East Asia, suggests that the financing of investment depends primarily on savings from corporate profits and the possibility of the banking system to create credit.  Strong enterprise profits simultaneously increase the incentive of firms to invest and their capacity to finance new investments from retained earnings.  This view better reflects uncertainty and the complexity of the real world.  The overall income of an economy is not a given but is determined by the process of decision making of all agents under uncertainty.  That is why entrepreneurial profits immediately adjust to changes in demand, and entrepreneurial decisions based on profit expectations in an uncertain world determine the level of investment in real productive capital.  That is why a fall in the savings ratio does not lead to a fall in investment.  On the contrary, since it implies an increase in consumer demand, it will increase profits, stimulate investment and spur growth of overall income.

By the same token, an improvement of the current account, as a result of changes in relative prices in favour of domestic producers, does not represent a reduction in the inflows of foreign savings that causes a fall in investment.  On the contrary, it is equivalent to an increase in aggregate demand and in the profits of domestic producers, and tends to lead to higher investment.  Thus, an increase in savings is not a prerequisite for either higher investment or an improvement in the current account.  Causality works in just the opposite direction.

The consequences of the different theoretical approaches for economic policy could not be more different.  When investment, output growth and employment are determined largely by profits of enterprises, economic policies have an important role to play in absorbing shocks and providing a stable environment for investment.  By contrast, in the neoclassical model, there is little room for economic policy and, where it does offer economic policy options, they point exactly in the opposite direction to those suggested by the Keynes-Schumpeter model.  Where the neoclassical model sees the need for households “to put aside more money” or for developing countries to attract more “foreign savings” to raise investment in fixed capital, the Keynes-Schumpeter model emphasizes pro-growth monetary conditions, positive demand and profit expectations as incentives for domestic entrepreneurs, and the need for reliable and affordable financing for enterprises.  “Sound macroeconomic policies”, as prescribed by the Washington Consensus, combined with financial liberalization seldom led to the desired result of higher investment and faster growth, whereas alternative policy approaches helped the newly industrializing economies of East and South-East Asia to accelerate their catch-up processes.

In Asia, accommodative and stimulating monetary policies, with low policy interest rates and government intervention in the financial markets, have been accompanied by undervalued exchange rates since the financial crisis in 1997-1998.  Fiscal policy has been used pragmatically to stimulate demand whenever that was required to respond to cyclical developments.  In South, East, and South-East Asia, the policy interest rate (in real and nominal terms) has been, on average, consistently lower than the growth rate (in real and nominal terms) of the economy (except during the Asian financial crisis).  By contrast, policy interest rates have been considerably higher in Latin America, where monetary policy has focused entirely on avoiding inflation, with the result that investment ratios and growth rates remained low.  It is only since 2003 that more accommodative monetary policies and an overall good growth performance have prevailed in the majority of the countries in that region.

This evidence suggests that sustained income growth needs proactive economic management so that there is a permanent tendency for planned investment to exceed planned savings.  Such an environment enables vigorous economic expansion, even if the propensity of households to save remains unchanged.

By the same token, if growth and fixed investment are constrained by tight monetary conditions, efforts aimed at improving governance or strengthening market institutions may not generate the expected results.  In pursuing the agenda of the Washington Consensus, which aimed at “getting prices right”, many countries got two of the most important prices — the exchange rate and the interest rate — wrong.  This may explain why the Washington Consensus was not applied in Washington: the United States, after flirting briefly with monetarist orthodoxy at the beginning of the 1980s, returned to fine-tuning the interest rate and to an extraordinarily accommodative monetary policy stance over the past two decades.

To be sure, a stable environment conducive to investment in productive capacity must include price stability.  Countries that are prone to high and accelerating inflation may find it more difficult to start and sustain a process of development and catching up than countries with a history of price stability.  In other words, without a sufficient number of policy instruments available to effectively dampen inflationary pressure, attempts to spur development by expansionary macroeconomic policy are more likely to fail if inflation soars.  But appropriate wage and income policies could help countries to maintain price stability so that monetary policy can be used to support an investment-led development process without risking an acceleration of inflation.

The recent positive performance of several developing countries also highlights the role of supportive exchange rate policies.  In the past, both exchange-rate pegging and flexible exchange-rate regimes often led to real currency appreciation and growing current-account deficits.  Over time, a worsening of the current-account balance increased the perception among international investors of greater currency risk and, at a certain point, triggered a sudden and strong capital outflow.  An increase in the current-account deficit — as a result of an appreciation of the real exchange rate and a concomitant loss of competitiveness of domestic producers — may be temporarily financed by a net capital inflow.  But it will sooner or later require some form of adjustment, normally in the form of a real depreciation.  Indeed, overvaluation has been the most frequent and most “reliable” predictor of financial crises in developing countries over the past two decades.

By contrast, the new strategy of many countries has been to aim at defending favourable competitive positions created by undervalued exchange rates and avoiding dependence on the international capital markets associated with a current-account deficit.  A competitive real exchange rate is a key factor in increasing aggregate demand in the short run and achieving higher employment in the long run.  However, there is a risk that governments will use exchange-rate manipulation in the same way as wage compression, subsidies and lower corporate taxation to artificially improve the international competitiveness of domestic producers.  This kind of “new mercantilism” in the competition for higher market shares cannot achieve the desired results.  All countries can simultaneously boost productivity, wages and trade to improve their overall economic welfare, but not all of them can simultaneously achieve current-account surpluses.  Successive rounds of competitive devaluations are therefore unproductive and likely to cause considerable damage.  Therefore, the world economy needs a new code of conduct going far beyond the existing framework of international rules of trade policy as agreed to in the World Trade Organization (WTO).



Kaldor, Nicholas (1983).  The Economic Consequences of Mrs. Thatcher.  London, Gerald Duckworth & Co. Ltd.

Keynes, John Maynard (1936).  General Theory of Employment, Interest and Money.  London, Macmillan, Cambridge University Press, for Royal Economic Society.

UNCTAD (2008).  Trade and Development Report 2008 “Commodity Prices, Capital Flows and the Financing of Investment,” UNCTAD Geneva.

Heiner Flassbeck is the Director of the Division on Globalization and Development Strategies of UNCTAD.  The text above is G-24 Policy Brief No. 35, based on Chapter 3 of UNCTAD 2008 Trade and Development Report (TDR) titled Commodity Prices, Capital Flows and the Financing of Investment; it is reproduced here for non-profit educational purposes.

| Print