The concept of economic growth has been with us for many centuries, but for much of this period little in the way of pure theorising was done on the subject. Prior to the European enlightenment, it was widely thought that growth was a matter of exploiting resources to one’s own advantage – i.e.
buying or acquiring them cheaply and processing the raw materials to add value, before selling the end product at a profit. Notions of the benefits of open competition were to come much later, and for a long period growth was
broadly seen as being at the expense of another party; either those deprived of the raw material or those sold the expensive final product. That is, growth was, at least in part, a zero sum game, where to ‘beggar your neighbour’ was the surest route to success.
This ‘mercantilist’ approach was to be turned on its head, however, by the work of first Adam Smith, and then David Ricardo. Smith (1776) argued that growth was a product of increasing productivity, which in turn was the result of the division of labour and the increased specialisation – and thus higher productivity – that this process produced. Furthermore, not only did the division of labour increase productivity in the production of goods (and services), it also allowed a surplus to be generated (i.e. beyond the immediate needs of the individual producer) that could be profitably traded with other producers specialising in the production of other goods.
Ricardo (1817) developed a rigorous, and in some ways pessimistic, but in other ways prescient, theory of growth. Ricardo was concerned that the ‘law of diminishing returns’ would eventually lead to the end of growth and a
‘steady state’ economy. However, this logical end to his economic system could be avoided by increasing technical progress and/or international trade.
For Ricardo, the second of these was the most important, and his famous theory of ‘comparative advantage’ encapsulated his powerful insight that trade was far from a zero sum game. On the contrary, Ricardo showed that free trade was of benefit to all parties. Differing levels of productivity of the factors of production – land, labour and capital – between countries ensured that if each country specialised in the production of goods in which it had a comparative advantage, total output would be maximised and mutually beneficial trade could be established.
Thus, for both Smith and Ricardo, specialisation and the division of labour leads to increasing productivity and growth – though for Ricardo this is subject to diminishing returns – and trade both within and between coun- tries enables the benefits of this specialisation to be diffused throughout society, raising output, welfare and growth.
Modern growth theory9 is, in large part, an extension of the work of these originators of classical economics. The following sections will sketch out the main developments that occurred in growth theory in the twentieth century, before linking this to recent findings on the relationship between financial sector development and growth, and concluding with a summary of what we know about the relationship between financial sector development and growth.
2.3.1 The Harrod-Domar growth model and its long-term developmental impact
The Harrod-Domar model (HDM) was developed independently by R. F.
Harrod and E. D. Domar in the l930s, and was the dominant growth model in development economics until the late 1950s. Furthermore, as we shall see,
aspects of the HDM have retained their importance even to the present day, despite the fact that the model itself is now widely discredited.
The fundamental prediction of the HDM is that national growth rates are directly proportional to the level of investment in the economy. The model makes a number of important assumptions to arrive at this prediction:
• Production results from the combination of labour and capital, which are in fixed proportions – i.e. more capital cannot be substituted for labour.
• There is no constraint on the supply of labour (particularly in developing countries with abundant supplies of workers, not least those moving from rural to urban areas).
• Physical capital is constrained (particularly in developing countries, which tend to have low savings rates and therefore low rates of invest- ment in physical capital).
• The productivity of capital (the ‘capital-output ratio’) is fixed. Conse- quently, there are ‘constant returns to scale’.
• More physical capital is therefore the only means to generate growth, and net investment leads to more capital accumulation, which generates higher output and income.
• Ultimately, higher income allows higher levels of saving, which feeds back into higher investment.
Technically, the HDM is very straightforward. The growth rate of GDP can be written as:
G(Y) =∆Y/Y Where:
Y = GDP.
To estimate the growth rate [G(Y)] the first step is to estimate the ‘incremental capital-output ratio’ (ICOR), which measures the productivity of capital.
That is, how much additional output is achieved by employing one additional unit of capital?
ICOR =∆K/∆Y Where:
K = capital stock.
A high ICOR value implies that large increases in the capital stock are needed to produce small increases in output. Or, put another way, that the marginal productivity of capital is low.
As described in the HDM assumptions above, physical capital is taken to
be the only constraint on increased production, so that investment (I) equates with the change in the stock of capital:
I =∆K
However, investment in the HDM is equal to national savings (S), which in turn is equal to the savings rate (Sr) multiplied by national income (Y):
I = S = Sr*Y Therefore:
ICOR = Sr*Y/∆Y Rearranging the terms gives:
G(Y) =∆Y/Y = Sr/ICOR
Therefore, as we can see, the rate of growth of the economy is simply the savings rate (and therefore the rate of investment) divided by the ICOR (the measure of the productiveness of the capital that these savings and invest- ment produce). In basic terms, a country with a savings rate of 20% and an ICOR of 4 would be expected to grow at a rate of 5% per year.
A key implication of the HDM is that, in order to kick-start growth, coun- tries with low savings rates should borrow to invest in physical capital, which will trigger a virtuous circle of growth, rising incomes, higher savings, higher investment in physical capital and, therefore, further increases in growth rates.
In the immediate post-war period, development economists began to look in earnest at means of reducing poverty in the developing world, with their preferred mechanism being to raise rates of economic growth. The HDM assumption that there was no constraint on the supply of labour looked eminently plausible at this time, where underemployment was the norm.
Furthermore, as Easterly (1997) points out, the rapid industrialisation of the Soviet Union had been driven by forced savings and investment in physical capital, seeming to confirm the validity of the HDM.10
Taking the lead from Sir Arthur Lewis (1955), it was assumed that, to achieve meaningful growth (i.e. real per capita growth where GDP growth is greater than population growth by a significant margin) investment in the economy, the rate of investment would have to rise to somewhere around 12%
of GDP. Assuming an ICOR of 4 – which was long the general assumption – this would result in growth of 3% per year.
However, if the savings rate (and therefore the investment rate) of the economy was just 4%, then growth would be 1%: below the rate of population growth and therefore producing falling per capita incomes. The difference between the 12% investment rate needed and the 4% actual savings and
investment rate was termed the ‘financing gap’. And it is here that we find the birth of the system of international aid (now termed ‘official development assistance’ [ODA]).
It was assumed that there was no prospect of developing countries being able to mobilise sufficient domestic resources to reach the 12% investment rate, which led to the conclusion that external resources from industrialised coun- tries could fill this financing gap, until a self-reinforcing process of faster growth, higher incomes, higher savings rates and more domestically financed investment could be established. Aid, from the outset, was therefore seen as a transitional mechanism to ‘kick-start’ the process leading to this virtuous circle.
The notion that at some point growth would become self-reinforcing relies on Rostow’s (1960) formulation in The Stages of Economic Growth. Rostow argued that the key stage was one of ‘take-off’ where growth became self- sustaining, and the key variable determining when ‘take-off’ would be reached was the level of investment in the economy approaching 10% of GDP (Easterly, op. cit.).
From a political economy perspective, it is important to remember – though hard to imagine today – that in the post-war era, many economists feared that capitalist economies would be unable to match the growth rates of the communist Soviet Union, not least because of its ability to marshal high levels of domestic savings and employ them as investments in physical capital on a mass scale. Politically, therefore, those such as Rostow who argued for high levels of aid to developing countries were able to play on fears that these countries would turn to communism rather than capitalism. As we shall see in later chapters – particularly regarding the birth of the regional development banks – many aspects of ‘development’ at this time were as much, if not more, a result of the need to prevent poor countries turning to communism, as they were the result of any particular altruism.
However, the arguments, whatever they were, worked. In the US, Kennedy increased the aid budget substantially, and by the presidency of Lyndon Johnson, US aid had reached its historic high of 0.6% of GDP (Easterly, op. cit.).
Box 2.4 Aid, East Asia and the Cold War
Aid is sometimes referred to as a form of ‘soft power’, as opposed to
‘hard’ military power. This is certainly not a new phenomenon, however.
After the Second World War, the United States became increasingly alarmed at the rise of communism, as one country after another appeared to fall under its sway. As well as concerns over the ideological appeal of communism, a key factor was its then economic success, which was exemplified by the growth and technical advances of the Soviet Union.
The Cold War was perhaps hottest in East and South East Asia, the locations of both the Korean and Vietnam Wars. The US with its regional ally Japan saw it as essential to create successful capitalist economies in the region, to act as a bulwark, and overseas aid was a key component of this. As can be seen from the figure, aid flows to South Korea in the 1960s were huge, contributing more than 70% of the cost of total fixed investments in the early 1960s – this compares with a figure closer to 5% for low- and middle-income countries as a group.
Furthermore, this intervention was not restricted to South Korea, nor just to aid, as Benedict Anderson (1998) points out:
In every important country of South-East Asia, with the exception of Indonesia, there were major, sustained Communist insurrec- tions, and Indonesia, in the early Sixties, had the largest legal Communist party in the world outside the socialist bloc. In all these states, except Malaysia, which was still a colony, the Americans intervened politically, economically, militarily and culturally on a massive scale. The notorious domino theory was invented specific- ally for South-East Asia. To shore up the line of teetering dominos, Washington made every effort to create loyal, capitalistically pros- perous, authoritarian and anti-Communist regimes – typically, but not invariably, dominated by the military. Each disaster only encouraged Washington to put more muscle and money behind its remaining political allies. No world region received more aid.
Furthermore, the export-led growth strategy employed in East and Southeast Asia could not have worked if the major export markets (particularly the US) had not been prepared to grant access to their Aid as a proportion of total capital investment, 1960–1970.
markets, without insisting on reciprocal access to the Asian markets.
This willingness to subordinate immediate economic interests to the strategic goal of creating capitalist ‘paragons’ in Asia was clearly an important factor in determining their success.
As we have seen, the ‘financing gap’ approach based on the HDM was sup- posed to be a time-limited injection of funds for investment to boost growth and achieve ‘take-off’. Consequently, there was little thought put into the prob- lems of the debt burden that developing countries were taking on, as it was assumed that growth in the future would be more than sufficient to cover this.
In practice, however, growth rates in much of the developing world did not respond as predicted to the influx of external financing, leading to the worst of all worlds: sluggish or even negative growth in per capita incomes, coupled with rising levels of indebtedness. The demise of the HDM as the pre- eminent growth model was therefore a result of the policies based on it to achieve their objectives. The model was seen to have failed.
Despite this, however, development economists – particularly those at the World Bank – continued to base their estimates of required external financing needed in developing countries on the ‘financing gap’ described above. Despite the growing complexity of the Bank’s models – culminating in the Revised Minimum Standard Model (RMSM) – the financing gap was effectively calculated in exactly the same way as above (Easterly, op. cit.) To the best of my knowledge, the RMSM remains in use today, despite the manifest failings of the growth model from which it is derived.
Although it was ultimately to founder on practical grounds – i.e. it didn’t work – the HDM had long been superseded on theoretical grounds. The seminal contribution in this regard was the neo-classical growth model (NCM), developed in the 1950s by Robert Solow.
Box 2.5 Robert Solow
The American economist Robert Solow was born in 1924 and is princi- pally known as the founding father of neo-classical growth theory, which is discussed below. Solow developed his early interests in statistics and probability while studying under Leontief at Harvard, but his interests moved more towards macroeconomics when he joined MIT in 1949. During the next half century Solow worked – often in conjunc- tion with Paul Samuelson – on a range of subjects, including capital theory and the famous Philips curve, which sought to describe the stat- istical relationship between the rates of unemployment and inflation as a trade-off where unemployment above a certain threshold leads to increased rates of inflation.
2.3.2 Solow’s neo-classical growth model
An important assumption in the HDM is that capital and labour are employed in fixed proportions (more capital cannot be substituted for labour) and that the capital-output ratio (i.e. the quantity of output resulting from the employment of a given quantity of capital) is fixed. That is, there are
‘constant returns to capital’.
In the 1950s Robert Solow adapted this model by relaxing these unrealistic assumptions of fixed ratios of capital-labour and capital-output. In Solow’s model it is possible to vary the proportions of capital and labour employed, and, crucially, he assumed diminishing returns to capital (not constant returns as in the HDM).
Therefore, as each unit of capital is added, the marginal increase in output falls. Ultimately, the increase in output will fall to the extent that the economy’s growth rate will equal the rate of growth of the popula- tion. If population growth is zero, economic growth will therefore also be zero.
In the Solow model, the production function can be written as follows:
Y = A*F(K,H,E*L) Where:
Y = income/output A = technological progress F = a function of
K = capital H = human capital E = productivity of labour L = labour force.
And the evolution of the capital stock is determined by:
∆K = Sr*F(K,H,E*L) – dK Where:
d = depreciation rate.
As can be seen from the equations above, Solow’s model was able to avoid reaching this ‘end-state’ by incorporating technological progress (A). Thus, in the absence of technological progress, output (growth) is a function of phys- ical and human capital accumulation – which inevitably face diminishing – and the size of the population, or labour force (L). However, Solow allows technological advances to raise the productivity of capital – the capital output ratio – therefore enabling growth to be maintained at a level above the rate of population growth.
As is clear from the first equation, however, Solow assumed that techno- logical progress was exogenous to the model. That is to say, technological progress was viewed as something that is not the result of internal economic forces in society (i.e. endogenous), but is an external input.
The Solow model is therefore able to explain the aspects of growth which result from increases in the stock of physical and human capital and the labour force, but not that resulting from technological change. This unexplained portion was termed the ‘Solow residual’ and was an increasing source of dissatisfaction with Solow’s model amongst economists. This is not surpris- ing, as estimates suggested that the share of growth accounted for by the Solow residual was not small. Quite the opposite. Economists increasingly came to see technical change as the key determinant of divergent growth rates between countries, and were understandably dissatisfied with a growth model that was unable to explain the determinants of this progress.
It is interesting to note that both the HDM and the NCM predicted that, under the right circumstances, higher growth in developing countries would enable them to catch up with the industrialised economies. In the HDM this would be achieved by raising investment rates sufficiently to achieve ‘take-off’ growth. In the NCM, in contrast, developing countries can inherently grow quicker as they start from a lower position, so that the process of diminishing returns to capital takes longer to kick in.
From a development perspective, the replacement of the HDM with the Solow model brought things to something of a dead-end, however. Whereas growth could be ‘created’ in the HDM through investment in physical capital, the NCM had very little to say about how growth could be increased. Indeed, as growth was largely viewed as a product of technological progress, which was exogenously determined, policy-makers could do little more than sit and wait for this ‘manna from heaven’ to fall into their laps.
Unsurprisingly, given that neither the HDM or the NCM could adequately explain the divergent growth experiences of developed and developing coun- tries, the focus shifted to developing models that were able to do this. The key aim, in all of this, was to discover how to make technological change inherent – or endogenous – to the growth process.
2.3.3 Endogenous growth models and new growth theory
Pritchett (2006) explains how the impasse reached in growth theory by the early 1980s coincided with the findings of the state-of-the-art macroeconomic models of the time, which suggested that policy-makers could do nothing to influence the rate of technical progress in the economy.
The immediate appeal of the approach that was to be termed New Growth Theory (NGT) was therefore that it opened up the possibility that the drivers of growth were endogenous to the economy, and could indeed be influenced by policy. Given this, it is unsurprising that national policy-makers and policy