This article focuses on evidence confirming long-term slowdown, as well as cyclical recession, of the US economy as indicated in the latest release of the 2^nd quarter 2010 US GDP figures.

Introduction

As widely reported, the second estimate of 2^nd quarter 2010 US GDP revised annualised US growth down from 2.4% to 1.6% – i.e. US GDP grew by 0.4% during the 2^nd quarter. The main changes compared to the first GDP estimate, in constant and annualised 2005 price terms, were a downward revision of net exports by -$19bn, due primarily to an upward re-estimation of imports by $14bn, and a revision of inventories downwards by -$13bn. Fixed investment remained essentially unchanged compared to the earlier first GDP estimate, with a revision downwards of -$1bn, and personal consumption was recalculated as $8bn higher than in the first GDP estimate (Bureau of Economic Analysis, 2010b) (Bureau of Economic Analysis, 2010a). An earlier article made a detailed examination of 2^nd quarter US GDP data and therefore only the implications for long-term trends are dealt with here. (Ross, 2010)

Slow recovery

The downward revision of 2^nd quarter GDP naturally highlights how much slower present US recovery is than in previous post-World War II business cycles. Ten quarters into the downturn US GDP still remains 1.3% below its peak in the 4^th quarter of 2007 – see Figure 1. In the previous worst post-World War II business cycle, that following 1973, recovery to the previous peak level of GDP was complete after eight quarters. Unless there is a significant acceleration of growth, US GDP will not regain its peak level until 2011 – meaning at least three years of net zero percent growth.

Figure 1

This slow recovery is, however, in line with a gradual but clear deceleration of long-term growth in the US economy – see Figure 2. The moving 20 year average of US GDP growth has now fallen gradually to 2.5% – significantly below its 3.5% historical average. Reasons the US is unlikely to reverse this trend in the foreseeable future are analysed below.

Figure 2

Fixed Investment fall

The new GDP figures also cast clear light on the issues of whether the recession in the US is primarily created by trends in consumption or investment. A number of analyses suggested that the core of the US economic crisis would be deleveraging by US consumers– see for example (Roach, 2009). If so the decline in US GDP would be centred in US consumption. The present author has consistently argued that this analysis is in error and that the core of the recession in the US is the decline in fixed investment. (Ross, 2010a) This is again strongly confirmed by the new revision of US GDP data.

Due to the significant downward revision of the US GDP figures, and the small upward revision of the consumer expenditure figures, consumption as a percentage of US GDP clearly remains well above its pre-financial crisis level – see Figure 3. Between the peak of US GDP, in the 4^th quarter of 2007, and the 2^nd quarter of 2010, US personal consumption has risen from 69.9% of GDP to 70.5% and total US consumption has risen from 85.8% of GDP to 87.6%.

Figure 3

The 1.8% of GDP increase in consumption as a percentage of US GDP is accounted for by a 0.8% of GDP increase in the share of military expenditure, a 0.6% of GDP increase in the share of personal consumption, and a 0.4% of GDP increase in the share of Federal non-military consumption.

In contrast the share of fixed investment in US GDP has fallen sharply by 3.6% of GDP. The share of non-residential fixed investment has fallen by 2.1% of GDP and the share of residential fixed investment by 1.5% of GDP.

The changes in components of US GDP, in terms of fixed price annualised 2005 dollars, are shown in Figure 4. US GDP remains $172bn below its previous peak level. However net exports, inventories, and government consumption are already above their 4^th quarter 2007 level – by $116bn, $51bn and $112bn respectively. Personal consumption is below its 4^th quarter 2007 level but only by $72bn. The US recession is entirely dominated by the $410bn decline in fixed investment.

Figure 4

The US economy, therefore, has not responded to the financial crisis primarily by reducing consumption, through personal debt deleveraging or other means, but by sharply reducing fixed investment.

Implications for long term US growth rates

A severe decline in US fixed investment, however, does not have only short term effects. As confirmed in the latest data of Jorgenson and Vu, capital investment continues to account for more than fifty percent of US GDP growth – the percentage for the latest period they analyse, in 2004-2008, is 61%. (Jorgenson & Vu, 2010) Under such conditions a severe decline in US fixed investment, of the type seen during the current recession, in practice excludes a rapid resumption of US GDP growth.

The slowdown that has been witnessed in long term US economic growth is therefore likely to continue. The present recession confirms a pattern of not simply cyclical downturn but structural slowing.

In that context the marked acceleration of US GDP growth which took place in 1995-2000 would appear to be a temporary upward fluctuation, financed by large scale import of capital, within an overall context of a long term structural slowdown of the US economy. It would not appear to mark the beginning of a more rapid US growth period.

The above trends therefore indicate that not only short but medium and long term projections for US economic growth should be assumed to be lower than historical averages. The US economy has been gradually slowing in not only a cyclical but a structural fashion.

Bibliography

Bureau of Economic Analysis. (2010b, August 27). National Income and Product Accounts Gross Domestic Product, 2nd quarter 2010 (second estimate). Retrieved August 27, 2010, from Bureau of Economic Analysis: http://www.bea.gov/newsreleases/national/gdp/2010/gdp2q10_2nd.htm

Bureau of Economic Analysis. (2010a, July 30). National Income and Product Accounts: Gross Domestic Product: Second Quarter 2010 (Advance Estimate). Retrieved July 30, 2010, from Bureau of Economic Analysis National Economic Accounts: http://www.bea.gov/newsreleases/national/gdp/2010/gdp2q10_adv.htm

Roach, S. (2009). The Next Asia. Hoboken, New Jersey: John Wiley and Sons.

Ross, J. (2010a, February 11). The myth of the decline of the US consumer. Retrieved August 28, 2010, from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2010/02/the-myth-of-the-decline-of-the-us-consumer.html

Ross, J. (2010, July 31). US 2nd quarter GDP figures – investment remains the key issue for US recovery. Retrieved August 28, 2010, from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2010/07/us-2nd-quarter-gdp.html

The most fundamental issue involved is that of the tendency of the contribution of investment to economic growth to increase with economic development. This trend was first analysed by Adam Smith and affirmed by other economists including Keynes.(3)However, it was rejected in analyses of economic growth put forward by various mid 20th century theorists. The latter asserted, against the analysis flowing from Adam Smith, that the division of the economy between investment and consumption remained constant with economic development – this is, of course, a central assumption of the widely employed Cobb-Douglas production function, of the economic growth model put forward by Solow (Solow, 1957), and continues to be repeated in many(4), although no longer all,(5) economic textbooks.

As discussed in detail elsewhere (Ross, 2009), the present author from the early 1970s concluded, on the basis of the long term historical data on growth that was then beginning to be published, than Adam Smith and those who followed his analysis were clearly correct. Despite the fact that the theories of Solow et al of a constant contribution of investment to GDP growth were the prevailing orthodoxy they were clearly contradicted by the historical data. Jorgenson has outlined more general reasons for the breakdown of such econometric models. (Jorgenson D. W., 2009) The theory that the proportion of the economy devoted to investment remained constant, rather than rose with economic development, was erroneous and classical economics was correct.

Initial statistical unclarity in this discussion was undoubtedly aided by the fact that the US is untypical in that, unlike the great majority of other economies, the proportion of the US economy devoted to fixed investment has indeed not risen for the approximately 150 year period for which reliable statistical data exists. (Ross, 2008a) (Barro & Sala-i-Martin, 2004).(6) Analyses based on generalisations from the US, therefore, arrived at the erroneous generalisation of a constant, rather than rising, share of investment in GDP.(7)

A consequence of the difference between the pattern in the US economy and the general international trend of a rising share of investment in GDP, is that the US share of fixed investment in GDP, which was above the international average in the 19th and first halt of the 20th century, has now fallen below the international average – in particular the US level of investment in GDP has fallen below the level of rapidly growing Asian economies. This contributes to the slow growth of the US economy compared to Asian competitors, the US balance of payments deficit with Asia, and present international financial developments. (Ross, 2008b)

The question of whether the role of fixed investment in economic growth rises with economic development, or remains constant, has numerous practical economic implications. The key pieces of evidence demonstrating the rising contribution of fixed investment to GDP growth with economic development, prior to the publication of the database of Jorgenson and Vu, included:

1. Trends in leading international growth economies. The analysis of leading growth economies, in successive historical periods of economic development from the 18^th century, carried out by the present author from the 1970s onwards, showed that each such leading economy – in chronological succession the UK, the US, West Germany, Japan, South Korea and now China – was characterised by a higher percentage of fixed investment in GDP than the preceding lead growth economy. This trend is shown in Figure 1.

Figure 1

2. The comprehensive comparative analyses, published by Angus Maddison, of economic growth in developed economies since World War II, including his Phases of Capitalist Development (Maddison, 1982) and Dynamic Forces in Capitalist Development (Maddison, 1991), demonstrated that capital investment was the largest contributor to post-World War II GDP growth in a wide range of advanced economies. Maddison demonstrated over a longer time frame than the post-World War II period that both non-residential fixed investment rates and savings rates, which were correlated with fixed investment rates, rose with time in most advanced economies – the US, as already noted, being an exception and not a rule. (Maddison, 1992)

3. Studies by Jorgenson and his co-authors demonstrated that capital investment was the largest single contribution to GDP growth in the US, the G7 and a number of developed economies.(8)

While these studies analysed a wide range of developed, and a number of leading developing, economies they were however not entirely comprehensive in coverage – in particular they did not include a comprehensive range of developing economies. The publication by Jorgenson and Vu (Jorgenson & Vu, 2007b) of a comprehensive growth accounting base for up to 122 economies, constituting more than 95% of world GDP, therefore provides an opportunity to fill this major gap.

A striking trend that may be analysed in its data, in addition to those drawn attention to by Jorgenson and Vu themselves, is that GDP growth in developed economies is capital-intensive while GDP growth in most developing economies is labour- intensive. The path of economic development is therefore a transition from labour intensive growth to capital intensive growth. The East Asian developing economies may be seen as an intermediate group between the majority of developing economies and the developed economies.

Such a pattern is, of course, consistent with the other evidence showing the rising contribution of investment to GDP growth with economic development. However, such a comprehensive database for evaluation of these trends has not previously been available – in particular it allows an integration of trends in developing economies with those in developed economies. It is for this reason that these three analyses of trends in the data of Jorgenson and Vu for capital and labour inputs in different stages of economic development have been published.

Summary

It is clear on the basis of the above data that there is overwhelming evidence of a tendency for the contribution of investment to GDP growth to rise with economic development. In light of such clear evidence it is evident that the assertion that the contribution of investment to GDP growth remains constant with time, rather than rising, is erroneous.

Economic models and theories therefore must take into account that the contribution of investment to GDP growth increases with economic development – Adam Smith and those who followed him on this issue were correct.

The implications of this for economic theory and practice of such a trend are numerous as has been analysed elsewhere (Ross, 2008a).

The normal disclaimer must, of course, be made that while these papers utilise calculations based on data produced by Jorgenson and Vu they do not bear responsibility for the conclusions drawn.

* * *

Since this analysis was carried out Jorgenson and Vu have extended their data to 2008. (Jorgenson & Vu, 2010) The new data does not alter the main trends analysed above. A detailed analysis from the angle of approach in this article will be published.

Notes

(1) (Ross, 2010a)

(2) (Ross, 2010b)

(3) In The Wealth of Nations, Adam Smith analysed that the role of capital and intermediate inputs, which he jointly termed ‘stock’, would increase as an economy developed. Smith noted: ‘As the accumulation of stock must, in the nature of things, be previous to the division of labour, so labour can be more and more subdivided in proportion only as stock is previously more and more accumulated… As the division of labour advances, therefore, in order to give constant employment to an equal number of workmen, an equal stock of provisions, and a greater stock of materials and tools than what would have been necessary in a ruder state of things must be accumulated beforehand.’ (Smith, 1999, p. 372) Keynes, arrived at the same conclusion of an increasing role of capital investment in economic development via a somewhat different chain of reasoning related to savings behaviour: ‘the richer the community, the wider will be the gap between its actual and its potential production… For a poor community will be prone to consumer by far the greater part of its output, so that a very modest measure of investment will be sufficient to provide full employment; whereas a wealthy community will have to discover much ampler opportunities for investment if the saving propensities of its wealthier members are to be compatible with the employment of its poorer members.’ (Keynes, 1983, p. 31) For a wider discussion see (Ross, 2009).

(4) Romer for example asserts ‘The growth rates of output and capital has been about equal (so that the capital-output ratio has been approximately constant).’ (Romer, 2006, p. 17) Blanchard asserts: ‘the savings rate does not appear to systematically increase or decrease as a country becomes richer.’ (Blanchard, 2006, p. 226)

(5) It is clearly rejected, as noted below, in (Barro & Sala-i-Martin, 2004)

(6) The initial data on which the present author arrived at the conclusion of confirmation of a rising share of investment in GDP was based on calculations from (Deane & Cole, 1967), (Feinstein, 1972), (Mitchell, 1980), (Economist, The, 1982), (Lister, 1989). Barro and Sala-i-Martin note: ‘For the United States, the striking observation… is the stability over time of the ratios for domestic investment and saving… The United States is, however, an outlier with respect to the stability of its investment and saving ratios; the data for the other seven countries [analysed] show a clear increase in these ratios over time… The long-term data therefore suggest that the ratios to GDP of gross domestic investment and gross national savings tend to rise as an economy develops, at least over some range. The assumption of a constant gross savings ratio, which appears… in the Solow-Swan model, misses the regularity in this data.’ (Barro & Sala-i-Martin, 2004, p. 15) Baro and Sala-i-Martin do not, however, draw out all the implications of this.

(7) Similar assertions were, however, also made by those who were centred on the UK economy – despite the fact that the UK economy showed a clear tendency for the proportion of investment in GDP to rise with time. Kaldor for example, in a widely cited paper, claimed as one of his ‘stylised facts’ on economic growth: ‘Steady capital-output ratios over long periods; at least there are no clear long term trends, either rising or falling, if differences in the degree of utilisation of capacity are allowed for. This implies, or reflects, the near identity in the percentage rates of growth of production and the capital stock – i.e. that for the economy as a whole, and over long periods, income and capital tend to grow at the same rate.’ (Kaldor, 1961, p. 178)

(8) See for example (Jorgenson & Yip, 2001), (Jorgenson D. W., 2003). Other authors have, of course, also noted the rising proportion of investment in GDP growth – see for example (Jones, 1995) and (De Long & Summers, 1992).

Bibliography

Barro, R. J., & Sala-i-Martin, X. (2004). Economic Growth. Cambridge, Massachusetts, US: MIT Press.

Blanchard, O. (2006). Macroeconomics. Upper Saddle River: Pearson Prentice Hall.

De Long, J. B., & Summers, L. H. (1992, revised October 1992). Equpment investment and economic growth. Retrieved August 22, 2010, from http://j-bradford-delong.net/pdf_files/Brookings_Equipment.pdf

Deane, P., & Cole, W. (1967). British Economic Growth 1688-1959. Cambridge: Cambridge University Press.

Economist, The. (1982). World Business Cycles. London: The Economist.

Feinstein, C. H. (1972). Statistical Tables of National Income, Expenditure and Output of the UK 1855-1965. Cambridge: Cambridge University Press.

Jones, C. I. (1995). Time series tests of endogenous growth models. The Quarterly Journal of Economics , 110 (2), 495-525.

Jorgenson, D. W. (2003). ‘Information technology and the G7 economies’. World Economics , 4 (4), 139-169.

Jorgenson, D. W. (2009). ‘Introduction’. In D. W. Jorgenson (Ed.), The Economics of Productivity (pp. ix-xxviii). Cheltenham, UK: Edward Elgar.

Jorgenson, D. W., & Vu, K. M. (2009). ‘Growth accounting within the International Comparison Programme’. The ICP Bulletin , 6 (1).

Jorgenson, D. W., & Vu, K. M. (2007b). ‘Information technology and the world growth resurgence – updated tables’. Retrieved from Dale Jorgenson: http://www.economics.harvard.edu/faculty/jorgenson/recent_work_jorgenson

Jorgenson, D. W., & Vu, K. M. (2010). Potential growth of the world economy. Journal of Policy Modeling (doi:10.1016/j.polmod.2010.07.011).

Jorgenson, D. W., & Yip, E. (2001). ‘Whatever happened to productivity growth’. In C. R. Hulten, E. R. Dean, & M. J. Harper (Eds.), New Developments in Productivity Analysis (pp. 509-540).

Kaldor, N. (1961). Capital Accumulation and Economic Growth. Retrieved August 22, 2010, from http://www.fep.up.pt/docentes/joao/material/macro2/Kaldor_1961.pdf

Keynes, J. M. (1983). The General Theory of Employment, Interest and Money. London: MacMillan.

Lister, T. (1989). One Hundred Years of Economic Statistics. London: The Economist.

Maddison, A. (1992). ‘A long run perspective on saving’. Scandanavian Journal of Economics , 94 (2), 181-196.

Maddison, A. (1991). Dynamic Forces in Capitalist Development. Oxford, UK: Oxford University Press.

Maddison, A. (1982). Phases of Capitalist Development. Oxford: Oxford University Press.

Mitchell, B. R. (1980). European Historical Statistics 1750-1975. London: MacMillan.

Romer, D. (2006). Advanced Macroeconomics. Boston: McGraw-Hill Irwin.

Ross, J. (2008b, September 25). Fundamental driving forces of the financial crisis. Retrieved August 22, 2010, from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2008/09/it-is-superfluous-to-note-on-this-blog-that-the-world-economy-ispassing-through-the-most-severe-financial-crisis-since-1929.html

Ross, J. (2010b, August 22). Labour inputs in stages of economic development – trends revealed in the data of Jorgenson and Vu. Retrieved August 2010, 2010, from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2010/08/labour_inputs.html.html

Ross, J. (2009, September 8). ‘The Asian and Chinese economic growth models – implications of modern findings on economic growth’. Retrieved from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2009/09/the-issue-of-whether-chinas-economic-stimulus-package-and-the-asian–growth–model-in-general-is-correct-and-therefore-its.html

Ross, J. (2010, August). ‘The Transition From Labour-Intensive to Capital-Intensive Growth During Economic Development – Trends Revealed in the Data of Jorgenson and Vu’. Retrieved from Key Trends in Globalisation.

Ross, J. (2010a, August 22). ‘The transition from labour-intensive to capital-intensive growth during economic development – trends revealed in the data of Jorgenson and Vu. Retrieved August 22, 2010, from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2010/08/capital_intensive_growth.html.html

Ross, J. (2008a, September 17). Why Asia will continue to grow more rapidly than the US and Europe – a historical perspective. Retrieved August 22, 2010, from Key Trends in Globalisationi: http://ablog.typepad.com/keytrendsinglobalisation/2008/09/data-on-long-term-trends-in-investment-and-economic-growth–this-post-deals-with-the-historic-trend-of-investment-and-econo.html

Smith, A. (1999). The Wealth of Nations. London: Penguin.

Solow, R. M. (1957). ‘Technical change and the aggregate production function’. Review of Economics and Statistics (3), 312-320.

One of the most important databases and analyses regarding the international economy released in the recent period is that calculated by Jorgenson and Vu to evaluate the relative contributions of capital, particularly information technology (IT), and labour inputs compared to that of total factor productivity (TFP) in GDP growth ((Jorgenson & Vu, 2007a), (Vu, 2007)) This has formed part of the World Bank’s International Comparison Programme (Jorgenson & Vu, 2009)). Because of the comprehensive nature of the published data trends can be identified in addition to those emphasised by Jorgenson and Vu themselves.

One of the most important of such trends is clear evidence that as economies become more developed the contribution of capital inputs to GDP growth increases relative to that of labour inputs, i.e. ‘capital-intensive’ growth replaces ‘labour-intensive’ growth. Such a pattern of transition from labour-intensive to capital-intensive growth with economic development would be important in itself and indicate a confirmation of analyses in classical economic theory.

The present article outlines these trends. It should be pointed out that while this paper utilises calculations based on data produced by Jorgenson and Vu they do not bear responsibility for conclusions drawn in the present article.

Source of economic growth in developed and developing economies

Table 1 sets out annual average GDP growth, together with the percentage contributions of capital, labour, and TFP, for 109 economies, 22 developed and 87 developing, for the ten year period 1995-2005. The calculations are from the data set out by Vu (Vu, 2007). Subgroups for developed and developing economies are shown. Table 2 sets out similar calculations from the data for periods defined by Jorgenson and Vu – 1989-1995, 1995-2000 and 2000-2005 (Jorgenson & Vu, 2007b).

In addition to periodisations, other differences between the two tables should be noted. Table 1 shows unweighted means and medians – i.e. the significance of each country is treated as equal, whereas the data of Jorgenson and Vu in Table 2 is weighted. In Table 1 improvements in labour quality are included in TFP, as in the data presented by Vu (Vu, 2007), whereas in Table 2 improvements in labour quality are included in labour inputs (Jorgenson & Vu, 2007b). There are 109 countries in Table 1 compared to 122 countries in Table 2. As will be seen, however, such differences do not alter the qualitative trends found.

The dominance of factor inputs in GDP growth

One of the fundamental findings of Jorgenson and Vu, shown in both Table 1 and Table 2, is the dominance of factor inputs of capital and labour compared to TFP in international economic growth. Taking all economies in Table 1, the mean percentage of growth accounted for by increases in inputs of capital and labour is 74.8% and the contribution of TFP growth is 25.1%. Taking the three periods defined by Jorgenson and Vu in Table 2, the contribution of factor inputs of capital and labour to GDP growth is 83.7%, 76.8% and 64.3%. From their data Jorgenson and Vu conclude: ‘We allocate the growth of world output, as measured in the World Bank’s International Comparison Program, between input growth and productivity. We find… that input growth greatly predominates. ‘(Jorgenson & Vu, 2009)

Role of capital inputs

Turning to a more detailed breakdown, the greatest contribution to GDP growth is from increase in capital inputs. Taking all economies in Table 1, the mean percentage contribution of increases in capital to GDP growth is 39.6% and the median contribution is 42.6% – which is above the mean of 35.2% and median of 35.1% for the contribution of increase in labour hours, which itself exceeds the mean contribution of TFP growth of 25.1% and median contribution of 25.1%.

Taking the periods defined by Jorgenson and Vu, the contribution of capital inputs to GDP growth is greater than either labour hours or TFP in all periods – at 54.1%, 46.4% and 40.7%. Jorgenson and Vu conclude: ‘About 40-45% of world growth can be attributed to the accumulation and deployment of capital and another 25-33% to the use of labour input… productivity accounted for only 20-35% of growth.’ (Jorgenson & Vu, 2009)

The different pattern of growth in developed and developing economies

Taking the data in Table 1, however, there is a clear contrast in the pattern of growth between developed and developing economies. The contribution of increases in capital inputs is significantly greater in the developed economies than in the developing ones – i.e. developed economies follow a ‘capital-intense’ path of development compared to developing economies. Considering all economies in Table 1, the mean contribution of capital inputs to GDP growth in developed economies is 52.9%, significantly above the 36.3% in developing economies. The median contribution of capital inputs to growth in developed economies is 50.6% compared to 39.3% in developing economies.

Table 1

Table 2 

In contrast to developed economies ‘capital-intense’ path of growth, the contribution of labour inputs to GDP growth is significantly higher in developing economies than in developed ones – i.e. in contrast to developed economies, developing economies have a ‘labour intense’ path of growth. The mean contribution of inputs of labour hours to GDP growth is 38.3% in developing economies compared to only 23.1% for developed economies, while the median contribution of increase in labour hours to GDP growth is 40.6% in developing economies compared to 23.8% for developed economies.

Summary of trends

Consolidating the above data, capital inputs are the dominant contribution to GDP growth in developed economies whereas the role of labour hours exceeds capital inputs in developing economies even if improvements in labour quality are included in TFP. The mean contributions to GDP growth in developed economies are, in order of percentage contribution, capital 52.9%, TFP 24.0%, and labour hours 23.1%, compared to, in developing economies, 38.3% labour hours, 36.3% capital, and 25.4% TFP. The median contributions to GDP growth in developed economies are capital 50.6%, TFP 26.0%, and labour hours 23.8%, compared to 40.6% labour hours, 39.3% capital, and 25.0% TFP in developing economies.

To see these trends visually, the mean percentage contribution of capital, labour hours and TFP to GDP growth for developed and developing countries is shown in Figure 1 and the median contributions in Figure 2.

Figure 1

Figure 2

Periodisation of Jorgenson and Vu

Jorgenson and Vu in their papers do not present a consolidated figure for all developed and all developing economies but the same pattern as in Table 1 may be seen from the data in Table 2:(1)

• In Jorgenson and Vu’s data, the percentage contribution of capital inputs to GDP growth in the G7 economies is the highest for any group in all periods. In all three periods capital investment contributed more than 50% of GDP growth in G7 economies. The capital-intensive path of development of G7 economies is therefore evident.

• The contribution of capital to GDP growth for the non-G7 developed economies is lower than for the G7 in all periods, but it is also higher than in the non-Asian developing countries in all periods. The contribution of capital to GDP growth for the non-G7 developed economies is higher than in the developing Asian economies for two periods and lower in one. This overall pattern confirms that, after the G7, the group of economies most dependent on capital investment for GDP growth is the non-G7 developed economies – although the gap with the East Asian developing economies is not great.
• The contribution of capital to GDP growth in the developing Asian economies is higher than for all other groups of developing economies in all periods – i.e. among developing economies the Asian economies most resemble the developed economies in the high intensity of capital investment in GDP growth.

• For the Latin American, Sub-Saharan African, and North African and Middle Eastern developing country groups, with only one exception, Sub-Saharan Africa in 2000-2006, the contribution of labour inputs to GDP growth exceeds the contribution of capital inputs to GDP growth in all periods.

• Eastern Europe(2), undergoing transition from Communism to capitalism, differs sharply from all other groups in that growth has weak inputs of capital and labour and relies primarily relied on TFP increases.

Therefore, although Jorgenson and Vu do not present a consolidated figure for developed and developing countries, the more capital-intensive character of economic growth in the developed economies compared to most developing economies is clear in their periodisation. Developing Asian economies constitute an ‘intermediate’ group between the majority of developing economies and the developed economies. The percentage contribution of capital to GDP growth in the periods and for the country groups defined in Jorgenson and Vu is shown visually in Figure 3.

More detailed analysis by country group within the above overall trends will now be considered.

Figure 3

The G7

The ‘capital-intensive’ pattern of development of the G7 economies is clear.(3) Taking the period 1995-2005 the G7 is the most capital-intensive in terms of its pattern of growth of any economy group. The mean contribution of capital inputs to GDP growth is 60.3% for the G7 compared to 39.6% for all economies, and the median contribution of capital inputs to GDP growth in the G7 is 50.9% compared to 42.6% for all economies – see Table 3.

Taking the periodisations of Jorgenson and Vu, shown in Table 2, the percentage contribution of capital inputs to GDP growth in the G7 is 60.0%, 53.4% and 56.3% – the highest of all groups for all periods.

The relative contribution of increases in labour hours to GDP growth in the G7 economies is low – a mean of 11.3% in the G7 economies compared to 35.2% in all economies and a median of 15.9% in the G7 economies compared to 35.1% for all economies. Taking the periodisation of Jorgenson and Vu the percentage contribution of labour inputs to GDP growth in the G7 is below the average for all economies in all periods.

The G7 economies therefore have a clear ‘capital intensive/low growth of labour inputs’ pattern of development.

Table 3

Non-G7 developed economies

The pattern of development of the non-G7 developed economies also shows a capital intensive path of development compared to developing economies.(4) Taking the period 1995-2005, the mean contribution of capital inputs to GDP growth is 49.5% for non-G7 developed economies compared to 36.3% for developing economies – the median contribution is 50.3% compared to 39.3% for developing economies. Compared to the G7, the contribution of capital inputs to GDP growth for non-G7 developed economies is either slightly lower or the same as for G7 economies – the mean for G7 economies being 60.3% and that for non-G7 developing economies being 49.5%. The median for the G7 economies is 50.9% and the median for non-G7 developed economies is 50.3%.

Taking the periodisation of Jorgenson and Vu, the non-G7 developed economies have a higher percentage contribution to GDP growth of capital inputs than all developing economy groups in all periods with the one exception of the East Asian developing economies in 2000-2005 – i.e. the more capital-intensive path of economic development in the non-G7 developed economies compared to developing economies is clear.

Given both G7 and the non-G7 developed economies have a more capital-intensive pattern of growth than developing economies, the more capital intensive growth of developed economies compared to developing economies is clear.

Table 4

East Asian and Asian developing economies

Jorgenson and Vu analyse the developing Asian economies as a single group – see Table 2. It may be preferable to divide them into an East Asian and a South Asian group – although the distinction is not vital from the point of view of the trends considered in this article. First the East Asian group will be considered and then the developing Asian economies as a whole.

The pattern of growth of the East Asian developing economies in the period 1995-2005 is shown in Table 5.(5) The East Asian group of developing economies are slightly less capital intensive in their path of development than the G7 and non-G7 developed economies – although the difference is not great. The mean percentage contribution of capital to GDP growth is 60.3% in the G7, 49.5% in the non-G7 Developed Economies and 47.2% in the East Asian Developing Economies, while the median contribution is 50.9% in the G7, 50.3% in the non-G7 Developed Economies and 48.4% in the East Asian economies. Overall, however, the East Asian developing economies clearly show a relatively similar capital-intensive path of GDP growth as the developed economies.

Taking the periodisation of Jorgenson and Vu, as shown in Table 2, and considering the developing Asian economies as whole, then as already noted the developing Asian economies have a lower percentage contribution to GDP growth of capital inputs compared to the G7 for all periods and compared to the non-G7 developed economies for two out of three periods. However, the developing Asian economies have a higher percentage contribution of capital inputs to GDP growth than all other groups of developing economies for all periods. The intermediate situation of the Asian/East Asian developing economies, in terms of capital-intensity of growth, between the developed economies and the other developing economies is therefore clear.

The developing East Asian/developing Asian economies are, however, not equidistant between the developed and the developing economies in their pattern of growth. Their pattern of growth, while not as capital-intensive as the developed economies is nevertheless closer, in its capital-intensity, to the developed economies than to the majority of developing economies.

Table 5

Other developing economies

Turning to wider groups of developing economies, Table 6 shows the pattern of GDP growth for the period 1995-2005 for developing economies excluding East Asia, which have already been examined, and Eastern Europe – which is analysed below. This covers 63 developing economies in South Asia, (6) Latin America,(7) Sub-Saharan Africa(8) and the Middle East and North Africa(9). This constitutes the majority of developing economies. The pattern of growth of these developing economies, compared to the developed economies and the developing East Asian/developing Asian economies, is clear. Unlike the developed economies and East Asian developing economies, GDP growth in these other groups of developing economies is dominated by labour inputs.

Taking the period 1995-2005 the mean contribution of inputs of labour hours for the 63 developing economies is 50.1% – compared to 28.8% for the East Asian developing economies, 28.5% for the non-G7 developed economies, and 11.3% for the G7 economies. The median contribution of inputs of labour hours is 44.2% compared to 26.9% for the East Asian developing economies, 24.5% for the non-G7 developing economies and 15.9% for the G7 economies.

Taking the periodisation of Jorgenson and Vu, as set out in Table 2, and taking their groupings of Sub-Saharan African, Latin American, and North African and Middle Eastern developing economies, the percentage contribution of inputs of labour is the highest input to GDP growth in all groups in all the periods up to 2006 except for Sub-Saharan Africa in 2000-2006.

The more labour-intensive path of growth of the majority of developing countries is clear.

Table 6

Eastern Europe and the former USSR

It may be seen from Table 2 that the pattern of economic changes in Eastern Europe in the period studied differed fundamentally from the rest of the world economy.(10)
The East European economies suffered severe falls in production in the early 1990s – this lasting until 1998 in the case of the former USSR.(11) The percentage contribution to GDP growth of input of labour hours in Eastern Europe and the former USSR was negative. The contribution of increase in capital inputs to GDP growth was very weak by comparison to the rest of the world economy. Taking the periodisation of Jorgenson and Vu, as shown in Table 2, after being negative in 1989-95, TFP accounted for 136.1% of GDP growth in Eastern Europe in the period 1995-2000, and for 88.9% in the period 2000-2005 – i.e unlike the rest of the world economy growth in Eastern Europe was overwhelmingly due to TFP increases.

This unique situation in the East European economies statistically raises the percentage contribution of TFP to GDP growth for the total economies in Table 1 and Table 2. However as Eastern Europe accounted for only 6.6% of the GDP of all countries analysed in 1989-1995, and only 5.5% in 1995-2000 and 2000-2005, the combined size of these economies is too small to alter substantively either the overall balance between inputs of capital and labour and TFP, or the capital intensive pattern of GDP growth of the developed economies.

The dependence of economic development on TFP growth in Eastern Europe may be regarded as either a unique one off event, due to the collapse of the former economic system in Eastern Europe and the former USSR, therefore not giving general lessons for economic growth, or treated as that in any period there will be statistical outliers. A case can therefore be made for excluding Eastern Europe and the former USSR from comparisons, on the grounds of their undergoing a unique experience, which would lower the average international contribution of TFP to GDP growth. However, in order to avoid the suggestion of selective inclusion of data, the case considered here is that of all economies including Eastern Europe.

Relevance to the path of development in classical economic theory

The finding that developed economies follow a capital-intensive path of growth compared to most developing economies, with the East Asian/Asian developing economies forming an intermediate group between the developed economies and other developing economies would be, of course, important in itself. It however would also cast light on an important issue in economic theory.

In the classic founding work of modern economics, The Wealth of Nations, Adam Smith analysed that the role of capital and intermediate inputs, which he jointly termed ‘stock’, would increase as an economy developed. Smith noted: ‘As the accumulation of stock must, in the nature of things, be previous to the division of labour, so labour can be more and more subdivided in proportion only as stock is previously more and more accumulated… As the division of labour advances, therefore, in order to give constant employment to an equal number of workmen, an equal stock of provisions, and a greater stock of materials and tools than what would have been necessary in a ruder state of things must be accumulated beforehand.’(Smith, 1999, p. 372) Other economists, including Keynes, arrived at the same conclusion of an increasing role of capital investment in economic development via a somewhat different chain of reasoning.(12) (For a wider discussion see Ross, 2009).

Smith, of course, had no systematic econometric data of the modern type with which to verify his findings – his conclusion was based on theoretical reasoning drawn from particular observations. Jorgenson and his collaborators have, however, already found that intermediate inputs, one element of Smith’s category of ‘stock’, grow more rapidly than capital, labour or TFP.(13) The finding in the data of Jorgenson and Vu that the percentage contribution of capital inputs to GDP growth is higher in developed than in developing economies is therefore also in line with, and casts important light on, Smith’s analysis and that of his successors.

Contrast of classical economic formulations with others

The pattern in the data calculated by Jorgenson and Vu is in line with classical economics. It however does not support other more recent theories regarding economic development. These contrasting views of the pattern of economic development are:

• Classical economic theory, as originally formulated by Adam Smith, foresaw a dynamic of transition from labour-intensive growth to more capital-intensive growth during economic development.

• Alternative theories, for example popularised by Krugman in regard to the East Asian economies (Krugman, 1994), instead suggested that the dynamic in economic development is one from growth dominated by factor inputs of capital and labour to one dominated by TFP growth in the most developed economies. (14)

The data presented by Jorgenson and Vu provides substantial evidence for the analysis of classical economics that economic development is accompanied by a transition from labour intensive to capital intensive growth. It however provides no evidence for the view that economic development leads to a greater role being played by TFP rather than factor inputs. Taking the six comparisons in the periodisation of Jorgenson and Vu (the two sets of developed economies, the G7 and the non-G7, times the three periods 1989-1995, 1995-2000, 2000-2005) TFP growth makes a lower percentage contribution to GDP growth in the developed economies than the average for all economies in four periods and a higher contribution in only two.

Conclusion

Jorgenson and Vu, in analysing their data, have emphasised:

• that capital and labour inputs predominate over TFP in GDP growth,
• the increasing role of IT investment in GDP growth,

However, a further significant trend is that their data shows a pattern whereby capital is the predominant input to GDP growth in developed economies – i.e. as economies develop they make a transition from labour-intensive to capital-intensive growth. Such a finding is of considerable importance:

• It indicates that, given such a pattern, an economy and its policy makers should anticipate growth becoming more capital intensive as an economy moves towards developed economy status.
• East Asian developing economies form an ‘intermediate’ group between the developed economies on the one side and the majority of developing economies on the other – although in greater reliance on capital inputs for growth than most developing economies they more resemble developed economies.
• Most developing economies are more dependent on labour inputs compared to capital inputs for GDP growth compared to developed economies.
• Such a transition from labour-intensive growth to capital-intensive growth as economies develop is in line with classic formulations of economic theory flowing from Adam Smith.
• Jorgenson and Vu’s data provides clear evidence in line with the classical economic view that as an economy develops its pattern of growth becomes more capital intensive, but it provides no evidence supporting views that as economies become more developed the role of factor inputs in GDP growth declines relative to TFP.
• Such a transition from labour intensive to capital intensive growth with economic development, while foreseen by classical economics, is not theorised in a number of standard contemporary treatments of economic growth – for a brief historical review see (Ross, 2009).(15)

The data produced by Jorgenson and Vu, and the trends it reveals, is therefore of great importance both from the practical point of view of view of study of economic growth and policy making, and from the point of view of economic theory.

* * *

Since this analysis was carried out Jorgenson and Vu have extended their data to 2008. (Jorgenson & Vu, 2010) The new data does not alter the main trends analysed above. A detailed analysis from the angle of approach in this article will be published here.

Notes

(1) Jorgenson and Vu note that for Latin America ‘The contribution of labour input was 1.77 before 1995, 1.70 from 1995-2000 and 1.82 after 2000, accounting for the lion’s share of regional growth.’ They also note for Sub-Saharan Africa: ‘As in Latin America, the contribution of labour input predominated throughout the period 1989-2004’ (Jorgenson & Vu, 2007a, p. 15).They do not however, make a generalisation to an overall pattern for developed and developing countries. It should also be noted that their data in the revised tables they have published (Jorgenson & Vu, ‘Information technology and the world growth resurgence – updated tables’, 2007b) differs from the original data in their article (Jorgenson & Vu, 2007a) in that its shows capital was the main percentage contributor to GDP growth in the non-G7 developed economies whereas the original data in their article indicated that labour input exceeded capital input in these economies and they had noted: ‘the non-G7 economies maintained rapid growth after 2000… The contribution of labour input predominated over capital input throughout the period 1989-2004.’ (Jorgenson & Vu, 2007a, p. 14)

(2) For conciseness Eastern Europe in this paper is to be taken as including the former USSR unless otherwise stated.

(3) Economies in the G7 are Canada, France, Germany, Italy, Japan, the UK and US.

(4) Countries in the Non-G7 Developed group are Australia, Austria, Belgium, Denmark, Finland, Greece, Ireland, Israel, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, and Switzerland.

(5) Countries included in the East Asia developing economies group are Cambodia, China, Hong Kong China, Indonesia, Malaysia, Philippines, Singapore, South Korea, Thailand, Vietnam.

(6) Countries in the South Asia group are Bangla Desh, India, Nepal, Pakistan and Sri Lanka.

(7) Countries in the Latin American group are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Trinidad & Tobago, Uruguay, and Venezuela.

(8) Countries in the Sub-Saharan Africa group are Benin, Botswana, Burkina Faso, Cameroon, Central African Republic, Chad, Republic of Congo, Cote d’Ivoire, Ethiopia, Gabon, Ghana, Guinea, Kenya, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Senegal and South Africa.

(9) Countries in the Middle East and North Africa group are Algeria, Egypt, Iran, Jordan, Lebanon, Morocco, Syrian Arab Republic, Tunisia, Turkey and Yemen.

(10) Countries in the East European group are Albania, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Russia, Slovak Republic, Slovenia and Ukraine.

(11) The latter states collectively experienced the largest declines in GDP in peacetime in the history of any modern economies

(12) Keynes derived the tendency of a rising role of investment with economic development from savings behaviour. It formed a cornerstone of his analysis of effective demand and crisis: ‘the richer the community, the wider will tend to be the gap between its actual and is potential production; and therefore the more obvious and outrageous the defects of the economy system. For a poor community will be prone to consume by far the greater part of its output, so that a very modest measure of investment will be sufficient to provide full employment; whereas a wealthy community will have to discover much ampler opportunities for investment if the saving propensities of its wealthier members are to be compatible with the employment of its poorer members. If in a potentially wealthy community the inducement to invest is weak, then in spite of its potential wealth, the working of the principle of effective demand will compel it to reduce its actual output, until, in spite of its potential wealth, it has become so poor that its surplus over its consumption is sufficiently diminished to correspond to the weakness of the inducement to invest.’ (Keynes, 1983, p. 31)

(13) For example analysing US industrial sectors Jorgenson, Gollop and Fraumeni found that intermediate inputs were the largest source of growth. They noted:

‘the contribution of intermediate input is by far the most significant source of growth in output. The contribution of intermediate input alone exceeds the rate of productivity growth for thirty six of the forty five industries for which we have a measure of intermediate input… the predominant contributions to output growth are those of intermediate, capital and labour inputs. By far the most important contribution is that of intermediate input.’[xiii] (Jorgenson, Gollop, & Fraumeni, 1999, p. 200)

Considering such findings for the US in more detail, Jorgenson concluded:

‘The analysis of sources of growth at the industry level is based on the decomposition of the growth rate of sectoral input into the sum of the contributions of intermediate, capital and labour inputs and the growth of sectoral productivity… The sum of the contributions of intermediate, capital, and labour inputs is the predominant source of growth of output for 46 of the 51 industries…

‘Comparing the contribution of intermediate input with other sources of growth demonstrates that this input is by far the most significant source of growth. The contribution of intermediate input exceeds productivity growth and the contributions of capital and labour inputs. If we focus attention on the contributions of capital and labour inputs alone, excluding intermediate input from consideration, these two inputs are a more important source of growth than changes in productivity… The explanatory power of this perspective is overwhelming at the sectoral level. For 46 of the 51 industrial sectors… the contribution of intermediate, capital and labour inputs is the predominant source of output growth. Changes in productivity account for the major portion of output growth in only five industries. (Jorgenson D. W., 1995, p. 5)

Regarding studies of rapidly growing Asian economies, Ren and Sun found for China that in the period 1981-2000, subdivided into 1984-88, 1988-94 and 1994-2000: ‘’Intermediate input growth is the primary source of output growth in most industries.’ (Ren & Sun, 2007). For Taiwan, analysing 26 sectors in 1981-99, Chi-Yuan Liang found regarding intermediate material inputs: ‘Material input is the biggest contributor to output growth in all sectors during 1981-99, except… seven’. (Liang C.-Y. , 2007). For South Korea Hak K. Pyo, Keun-Hee Rhee and Bongchan Ha found: ‘The relative magnitude of contribution to output growth is in the order of: material, capital, labour, TFP then energy.’ (Pyo, Rhee, & Ha, 2007)

(14) For an attempt to apply such an analysis to China see (Zheng, Bigsten, & Hu, 2009).

(15) For standard surveys of theories of economic growth see (Barro & Sala-i-Martin, 2004) or (Acemoglu, 2009).

Bibliography

Jorgenson, D. W. (1995). ‘Productivity and postwar US economic growth’. In D. W. Jorgenson, Productivity (Vol. 1, pp. 1-23). Cambridge, Massachusetts: The MIT Press.

Jorgenson, D. W., & Vu, K. M. (2007a). ‘Information technology and the world growth resurgence’. German Economic Review , 8 (5).

Jorgenson, D. W., & Vu, K. M. (2007b). ‘Information technology and the world growth resurgence – updated tables’. Retrieved from Dale Jorgenson: http://www.economics.harvard.edu/faculty/jorgenson/recent_work_jorgenson

Jorgenson, D. W., & Vu, K. M. (2009). ‘Growth accounting within the International Comparison Programme’. The ICP Bulletin , 6 (1).

Jorgenson, D. W., Gollop, F. M., & Fraumeni, B. M. (1987). Productivity and US Economic Growth. Harvard University Press.

Keynes, J. M. (1936). The General Theory of Employment, Interest and Money. London: MacMillan.

Krugman, P. (1994). ‘The myth of Asia’s miracle’. Foreign Affairs , 62-78.

Liang, C.-Y. (2007). ‘Industry-wide total factor productivity and output growth in Taiwan, 1981-1999’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia: Economic Growth and Competitiveness (pp. 146-184). Cheltenham: Edward Elgar.

Pyo, H. K., Rhee, K.-H., & Ha, B. (2007). ‘Growth accounting and productivity analysis by 33 industrial sectors in Korea (1984-2002)’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia: Economic Growth and Competitiveness (pp. 113-145). Northampton: Edward Elgar.

Ren, R., & Sun, L. l. (2007). ‘Total factor productivity growth in Chinese industries, 1981-2000’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia, Economic Growth and Competitiveness (pp. 76-112). Northampton: Edward Elgar.

Ross, J. (2009, September . ‘The Asian and Chinese economic growth models – implications of modern findings on economic growth’. Retrieved from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2009/09/the-issue-of-whether-chinas-economic-stimulus-package-and-the-asian–growth–model-in-general-is-correct-and-therefore-its.html

Smith, A. (1999). The Wealth of Nations. London: Penguin.

Vu, K. M. (2007). ‘Determinants of economic growth over the period 1995-2005′. Retrieved from http://docs.google.com/viewer?a=v&q=cache:s7odRXb7owsJ:www.spp.nus.edu.sg/Handler.ashx%3Fpath%3DData/Site/SiteDocuments/wp/wp0907.pdf+%22Determinants+of+Economic+Growth+Over+the+Period+1995-2005%22+Vu&hl=en&gl=uk&pid=bl&srcid=ADGEESj9omj1MsLK6PX0-ns9×4Gvz

Zheng, J., Bigsten, A., & Hu, A. (2009). ‘Can China’s growth be sustained? A productivity perspective’. World Development , 37 (4), 874-888.

One of the most important databases and analyses regarding the international economy released in the recent period is that calculated by Jorgenson and Vu to evaluate the relative contributions of capital, particularly information technology (IT), and labour inputs compared to that of total factor productivity (TFP) in GDP growth ((Jorgenson & Vu, 2007a), (Vu, 2007)) This has formed part of the World Bank’s International Comparison Programme (Jorgenson & Vu, 2009)). Because of the comprehensive nature of the published data trends can be identified in addition to those emphasised by Jorgenson and Vu themselves.

One of the most important of such trends is clear evidence that as economies become more developed the contribution of capital inputs to GDP growth increases relative to that of labour inputs, i.e. ‘capital-intensive’ growth replaces ‘labour-intensive’ growth. Such a pattern of transition from labour-intensive to capital-intensive growth with economic development would be important in itself and indicate a confirmation of analyses in classical economic theory.

The present article outlines these trends. It should be pointed out that while this paper utilises calculations based on data produced by Jorgenson and Vu they do not bear responsibility for conclusions drawn in the present article.

Source of economic growth in developed and developing economies

Table 1 sets out annual average GDP growth, together with the percentage contributions of capital, labour, and TFP, for 109 economies, 22 developed and 87 developing, for the ten year period 1995-2005. The calculations are from the data set out by Vu (Vu, 2007). Subgroups for developed and developing economies are shown. Table 2 sets out similar calculations from the data for periods defined by Jorgenson and Vu – 1989-1995, 1995-2000 and 2000-2005 (Jorgenson & Vu, 2007b).

In addition to periodisations, other differences between the two tables should be noted. Table 1 shows unweighted means and medians – i.e. the significance of each country is treated as equal, whereas the data of Jorgenson and Vu in Table 2 is weighted. In Table 1 improvements in labour quality are included in TFP, as in the data presented by Vu (Vu, 2007), whereas in Table 2 improvements in labour quality are included in labour inputs (Jorgenson & Vu, 2007b). There are 109 countries in Table 1 compared to 122 countries in Table 2. As will be seen, however, such differences do not alter the qualitative trends found.

The dominance of factor inputs in GDP growth

One of the fundamental findings of Jorgenson and Vu, shown in both Table 1 and Table 2, is the dominance of factor inputs of capital and labour compared to TFP in international economic growth. Taking all economies in Table 1, the mean percentage of growth accounted for by increases in inputs of capital and labour is 74.8% and the contribution of TFP growth is 25.1%. Taking the three periods defined by Jorgenson and Vu in Table 2, the contribution of factor inputs of capital and labour to GDP growth is 83.7%, 76.8% and 64.3%. From their data Jorgenson and Vu conclude: ‘We allocate the growth of world output, as measured in the World Bank’s International Comparison Program, between input growth and productivity. We find… that input growth greatly predominates. ‘(Jorgenson & Vu, 2009)

Role of capital inputs

Turning to a more detailed breakdown, the greatest contribution to GDP growth is from increase in capital inputs. Taking all economies in Table 1, the mean percentage contribution of increases in capital to GDP growth is 39.6% and the median contribution is 42.6% – which is above the mean of 35.2% and median of 35.1% for the contribution of increase in labour hours, which itself exceeds the mean contribution of TFP growth of 25.1% and median contribution of 25.1%.

Taking the periods defined by Jorgenson and Vu, the contribution of capital inputs to GDP growth is greater than either labour hours or TFP in all periods – at 54.1%, 46.4% and 40.7%. Jorgenson and Vu conclude: ‘About 40-45% of world growth can be attributed to the accumulation and deployment of capital and another 25-33% to the use of labour input… productivity accounted for only 20-35% of growth.’ (Jorgenson & Vu, 2009)

The different pattern of growth in developed and developing economies

Taking the data in Table 1, however, there is a clear contrast in the pattern of growth between developed and developing economies. The contribution of increases in capital inputs is significantly greater in the developed economies than in the developing ones – i.e. developed economies follow a ‘capital-intense’ path of development compared to developing economies. Considering all economies in Table 1, the mean contribution of capital inputs to GDP growth in developed economies is 52.9%, significantly above the 36.3% in developing economies. The median contribution of capital inputs to growth in developed economies is 50.6% compared to 39.3% in developing economies.

Table 1

Table 2 

In contrast to developed economies ‘capital-intense’ path of growth, the contribution of labour inputs to GDP growth is significantly higher in developing economies than in developed ones – i.e. in contrast to developed economies, developing economies have a ‘labour intense’ path of growth. The mean contribution of inputs of labour hours to GDP growth is 38.3% in developing economies compared to only 23.1% for developed economies, while the median contribution of increase in labour hours to GDP growth is 40.6% in developing economies compared to 23.8% for developed economies.

Summary of trends

Consolidating the above data, capital inputs are the dominant contribution to GDP growth in developed economies whereas the role of labour hours exceeds capital inputs in developing economies even if improvements in labour quality are included in TFP. The mean contributions to GDP growth in developed economies are, in order of percentage contribution, capital 52.9%, TFP 24.0%, and labour hours 23.1%, compared to, in developing economies, 38.3% labour hours, 36.3% capital, and 25.4% TFP. The median contributions to GDP growth in developed economies are capital 50.6%, TFP 26.0%, and labour hours 23.8%, compared to 40.6% labour hours, 39.3% capital, and 25.0% TFP in developing economies.

To see these trends visually, the mean percentage contribution of capital, labour hours and TFP to GDP growth for developed and developing countries is shown in Figure 1 and the median contributions in Figure 2.

Figure 1

Figure 2

Periodisation of Jorgenson and Vu

Jorgenson and Vu in their papers do not present a consolidated figure for all developed and all developing economies but the same pattern as in Table 1 may be seen from the data in Table 2:(1)

• In Jorgenson and Vu’s data, the percentage contribution of capital inputs to GDP growth in the G7 economies is the highest for any group in all periods. In all three periods capital investment contributed more than 50% of GDP growth in G7 economies. The capital-intensive path of development of G7 economies is therefore evident.

• The contribution of capital to GDP growth for the non-G7 developed economies is lower than for the G7 in all periods, but it is also higher than in the non-Asian developing countries in all periods. The contribution of capital to GDP growth for the non-G7 developed economies is higher than in the developing Asian economies for two periods and lower in one. This overall pattern confirms that, after the G7, the group of economies most dependent on capital investment for GDP growth is the non-G7 developed economies – although the gap with the East Asian developing economies is not great.
• The contribution of capital to GDP growth in the developing Asian economies is higher than for all other groups of developing economies in all periods – i.e. among developing economies the Asian economies most resemble the developed economies in the high intensity of capital investment in GDP growth.

• For the Latin American, Sub-Saharan African, and North African and Middle Eastern developing country groups, with only one exception, Sub-Saharan Africa in 2000-2006, the contribution of labour inputs to GDP growth exceeds the contribution of capital inputs to GDP growth in all periods.

• Eastern Europe(2), undergoing transition from Communism to capitalism, differs sharply from all other groups in that growth has weak inputs of capital and labour and relies primarily relied on TFP increases.

Therefore, although Jorgenson and Vu do not present a consolidated figure for developed and developing countries, the more capital-intensive character of economic growth in the developed economies compared to most developing economies is clear in their periodisation. Developing Asian economies constitute an ‘intermediate’ group between the majority of developing economies and the developed economies. The percentage contribution of capital to GDP growth in the periods and for the country groups defined in Jorgenson and Vu is shown visually in Figure 3.

More detailed analysis by country group within the above overall trends will now be considered.

Figure 3

The G7

The ‘capital-intensive’ pattern of development of the G7 economies is clear.(3) Taking the period 1995-2005 the G7 is the most capital-intensive in terms of its pattern of growth of any economy group. The mean contribution of capital inputs to GDP growth is 60.3% for the G7 compared to 39.6% for all economies, and the median contribution of capital inputs to GDP growth in the G7 is 50.9% compared to 42.6% for all economies – see Table 3.

Taking the periodisations of Jorgenson and Vu, shown in Table 2, the percentage contribution of capital inputs to GDP growth in the G7 is 60.0%, 53.4% and 56.3% – the highest of all groups for all periods.

The relative contribution of increases in labour hours to GDP growth in the G7 economies is low – a mean of 11.3% in the G7 economies compared to 35.2% in all economies and a median of 15.9% in the G7 economies compared to 35.1% for all economies. Taking the periodisation of Jorgenson and Vu the percentage contribution of labour inputs to GDP growth in the G7 is below the average for all economies in all periods.

The G7 economies therefore have a clear ‘capital intensive/low growth of labour inputs’ pattern of development.

Table 3

Non-G7 developed economies

The pattern of development of the non-G7 developed economies also shows a capital intensive path of development compared to developing economies.(4) Taking the period 1995-2005, the mean contribution of capital inputs to GDP growth is 49.5% for non-G7 developed economies compared to 36.3% for developing economies – the median contribution is 50.3% compared to 39.3% for developing economies. Compared to the G7, the contribution of capital inputs to GDP growth for non-G7 developed economies is either slightly lower or the same as for G7 economies – the mean for G7 economies being 60.3% and that for non-G7 developing economies being 49.5%. The median for the G7 economies is 50.9% and the median for non-G7 developed economies is 50.3%.

Taking the periodisation of Jorgenson and Vu, the non-G7 developed economies have a higher percentage contribution to GDP growth of capital inputs than all developing economy groups in all periods with the one exception of the East Asian developing economies in 2000-2005 – i.e. the more capital-intensive path of economic development in the non-G7 developed economies compared to developing economies is clear.

Given both G7 and the non-G7 developed economies have a more capital-intensive pattern of growth than developing economies, the more capital intensive growth of developed economies compared to developing economies is clear.

Table 4

East Asian and Asian developing economies

Jorgenson and Vu analyse the developing Asian economies as a single group – see Table 2. It may be preferable to divide them into an East Asian and a South Asian group – although the distinction is not vital from the point of view of the trends considered in this article. First the East Asian group will be considered and then the developing Asian economies as a whole.

The pattern of growth of the East Asian developing economies in the period 1995-2005 is shown in Table 5.(5) The East Asian group of developing economies are slightly less capital intensive in their path of development than the G7 and non-G7 developed economies – although the difference is not great. The mean percentage contribution of capital to GDP growth is 60.3% in the G7, 49.5% in the non-G7 Developed Economies and 47.2% in the East Asian Developing Economies, while the median contribution is 50.9% in the G7, 50.3% in the non-G7 Developed Economies and 48.4% in the East Asian economies. Overall, however, the East Asian developing economies clearly show a relatively similar capital-intensive path of GDP growth as the developed economies.

Taking the periodisation of Jorgenson and Vu, as shown in Table 2, and considering the developing Asian economies as whole, then as already noted the developing Asian economies have a lower percentage contribution to GDP growth of capital inputs compared to the G7 for all periods and compared to the non-G7 developed economies for two out of three periods. However, the developing Asian economies have a higher percentage contribution of capital inputs to GDP growth than all other groups of developing economies for all periods. The intermediate situation of the Asian/East Asian developing economies, in terms of capital-intensity of growth, between the developed economies and the other developing economies is therefore clear.

The developing East Asian/developing Asian economies are, however, not equidistant between the developed and the developing economies in their pattern of growth. Their pattern of growth, while not as capital-intensive as the developed economies is nevertheless closer, in its capital-intensity, to the developed economies than to the majority of developing economies.

Table 5

Other developing economies

Turning to wider groups of developing economies, Table 6 shows the pattern of GDP growth for the period 1995-2005 for developing economies excluding East Asia, which have already been examined, and Eastern Europe – which is analysed below. This covers 63 developing economies in South Asia, (6) Latin America,(7) Sub-Saharan Africa(8) and the Middle East and North Africa(9). This constitutes the majority of developing economies. The pattern of growth of these developing economies, compared to the developed economies and the developing East Asian/developing Asian economies, is clear. Unlike the developed economies and East Asian developing economies, GDP growth in these other groups of developing economies is dominated by labour inputs.

Taking the period 1995-2005 the mean contribution of inputs of labour hours for the 63 developing economies is 50.1% – compared to 28.8% for the East Asian developing economies, 28.5% for the non-G7 developed economies, and 11.3% for the G7 economies. The median contribution of inputs of labour hours is 44.2% compared to 26.9% for the East Asian developing economies, 24.5% for the non-G7 developing economies and 15.9% for the G7 economies.

Taking the periodisation of Jorgenson and Vu, as set out in Table 2, and taking their groupings of Sub-Saharan African, Latin American, and North African and Middle Eastern developing economies, the percentage contribution of inputs of labour is the highest input to GDP growth in all groups in all the periods up to 2006 except for Sub-Saharan Africa in 2000-2006.

The more labour-intensive path of growth of the majority of developing countries is clear.

Table 6

Eastern Europe and the former USSR

It may be seen from Table 2 that the pattern of economic changes in Eastern Europe in the period studied differed fundamentally from the rest of the world economy.(10)
The East European economies suffered severe falls in production in the early 1990s – this lasting until 1998 in the case of the former USSR.(11) The percentage contribution to GDP growth of input of labour hours in Eastern Europe and the former USSR was negative. The contribution of increase in capital inputs to GDP growth was very weak by comparison to the rest of the world economy. Taking the periodisation of Jorgenson and Vu, as shown in Table 2, after being negative in 1989-95, TFP accounted for 136.1% of GDP growth in Eastern Europe in the period 1995-2000, and for 88.9% in the period 2000-2005 – i.e unlike the rest of the world economy growth in Eastern Europe was overwhelmingly due to TFP increases.

This unique situation in the East European economies statistically raises the percentage contribution of TFP to GDP growth for the total economies in Table 1 and Table 2. However as Eastern Europe accounted for only 6.6% of the GDP of all countries analysed in 1989-1995, and only 5.5% in 1995-2000 and 2000-2005, the combined size of these economies is too small to alter substantively either the overall balance between inputs of capital and labour and TFP, or the capital intensive pattern of GDP growth of the developed economies.

The dependence of economic development on TFP growth in Eastern Europe may be regarded as either a unique one off event, due to the collapse of the former economic system in Eastern Europe and the former USSR, therefore not giving general lessons for economic growth, or treated as that in any period there will be statistical outliers. A case can therefore be made for excluding Eastern Europe and the former USSR from comparisons, on the grounds of their undergoing a unique experience, which would lower the average international contribution of TFP to GDP growth. However, in order to avoid the suggestion of selective inclusion of data, the case considered here is that of all economies including Eastern Europe.

Relevance to the path of development in classical economic theory

The finding that developed economies follow a capital-intensive path of growth compared to most developing economies, with the East Asian/Asian developing economies forming an intermediate group between the developed economies and other developing economies would be, of course, important in itself. It however would also cast light on an important issue in economic theory.

In the classic founding work of modern economics, The Wealth of Nations, Adam Smith analysed that the role of capital and intermediate inputs, which he jointly termed ‘stock’, would increase as an economy developed. Smith noted: ‘As the accumulation of stock must, in the nature of things, be previous to the division of labour, so labour can be more and more subdivided in proportion only as stock is previously more and more accumulated… As the division of labour advances, therefore, in order to give constant employment to an equal number of workmen, an equal stock of provisions, and a greater stock of materials and tools than what would have been necessary in a ruder state of things must be accumulated beforehand.’(Smith, 1999, p. 372) Other economists, including Keynes, arrived at the same conclusion of an increasing role of capital investment in economic development via a somewhat different chain of reasoning.(12) (For a wider discussion see Ross, 2009).

Smith, of course, had no systematic econometric data of the modern type with which to verify his findings – his conclusion was based on theoretical reasoning drawn from particular observations. Jorgenson and his collaborators have, however, already found that intermediate inputs, one element of Smith’s category of ‘stock’, grow more rapidly than capital, labour or TFP.(13) The finding in the data of Jorgenson and Vu that the percentage contribution of capital inputs to GDP growth is higher in developed than in developing economies is therefore also in line with, and casts important light on, Smith’s analysis and that of his successors.

Contrast of classical economic formulations with others

The pattern in the data calculated by Jorgenson and Vu is in line with classical economics. It however does not support other more recent theories regarding economic development. These contrasting views of the pattern of economic development are:

• Classical economic theory, as originally formulated by Adam Smith, foresaw a dynamic of transition from labour-intensive growth to more capital-intensive growth during economic development.

• Alternative theories, for example popularised by Krugman in regard to the East Asian economies (Krugman, 1994), instead suggested that the dynamic in economic development is one from growth dominated by factor inputs of capital and labour to one dominated by TFP growth in the most developed economies. (14)

The data presented by Jorgenson and Vu provides substantial evidence for the analysis of classical economics that economic development is accompanied by a transition from labour intensive to capital intensive growth. It however provides no evidence for the view that economic development leads to a greater role being played by TFP rather than factor inputs. Taking the six comparisons in the periodisation of Jorgenson and Vu (the two sets of developed economies, the G7 and the non-G7, times the three periods 1989-1995, 1995-2000, 2000-2005) TFP growth makes a lower percentage contribution to GDP growth in the developed economies than the average for all economies in four periods and a higher contribution in only two.

Conclusion

Jorgenson and Vu, in analysing their data, have emphasised:

• that capital and labour inputs predominate over TFP in GDP growth,
• the increasing role of IT investment in GDP growth,

However, a further significant trend is that their data shows a pattern whereby capital is the predominant input to GDP growth in developed economies – i.e. as economies develop they make a transition from labour-intensive to capital-intensive growth. Such a finding is of considerable importance:

• It indicates that, given such a pattern, an economy and its policy makers should anticipate growth becoming more capital intensive as an economy moves towards developed economy status.
• East Asian developing economies form an ‘intermediate’ group between the developed economies on the one side and the majority of developing economies on the other – although in greater reliance on capital inputs for growth than most developing economies they more resemble developed economies.
• Most developing economies are more dependent on labour inputs compared to capital inputs for GDP growth compared to developed economies.
• Such a transition from labour-intensive growth to capital-intensive growth as economies develop is in line with classic formulations of economic theory flowing from Adam Smith.
• Jorgenson and Vu’s data provides clear evidence in line with the classical economic view that as an economy develops its pattern of growth becomes more capital intensive, but it provides no evidence supporting views that as economies become more developed the role of factor inputs in GDP growth declines relative to TFP.
• Such a transition from labour intensive to capital intensive growth with economic development, while foreseen by classical economics, is not theorised in a number of standard contemporary treatments of economic growth – for a brief historical review see (Ross, 2009).(15)

The data produced by Jorgenson and Vu, and the trends it reveals, is therefore of great importance both from the practical point of view of view of study of economic growth and policy making, and from the point of view of economic theory.

* * *

Since this analysis was carried out Jorgenson and Vu have extended their data to 2008. (Jorgenson & Vu, 2010) The new data does not alter the main trends analysed above. A detailed analysis from the angle of approach in this article will be published here.

Notes

(1) Jorgenson and Vu note that for Latin America ‘The contribution of labour input was 1.77 before 1995, 1.70 from 1995-2000 and 1.82 after 2000, accounting for the lion’s share of regional growth.’ They also note for Sub-Saharan Africa: ‘As in Latin America, the contribution of labour input predominated throughout the period 1989-2004’ (Jorgenson & Vu, 2007a, p. 15).They do not however, make a generalisation to an overall pattern for developed and developing countries. It should also be noted that their data in the revised tables they have published (Jorgenson & Vu, ‘Information technology and the world growth resurgence – updated tables’, 2007b) differs from the original data in their article (Jorgenson & Vu, 2007a) in that its shows capital was the main percentage contributor to GDP growth in the non-G7 developed economies whereas the original data in their article indicated that labour input exceeded capital input in these economies and they had noted: ‘the non-G7 economies maintained rapid growth after 2000… The contribution of labour input predominated over capital input throughout the period 1989-2004.’ (Jorgenson & Vu, 2007a, p. 14)

(2) For conciseness Eastern Europe in this paper is to be taken as including the former USSR unless otherwise stated.

(3) Economies in the G7 are Canada, France, Germany, Italy, Japan, the UK and US.

(4) Countries in the Non-G7 Developed group are Australia, Austria, Belgium, Denmark, Finland, Greece, Ireland, Israel, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, and Switzerland.

(5) Countries included in the East Asia developing economies group are Cambodia, China, Hong Kong China, Indonesia, Malaysia, Philippines, Singapore, South Korea, Thailand, Vietnam.

(6) Countries in the South Asia group are Bangla Desh, India, Nepal, Pakistan and Sri Lanka.

(7) Countries in the Latin American group are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Trinidad & Tobago, Uruguay, and Venezuela.

(8) Countries in the Sub-Saharan Africa group are Benin, Botswana, Burkina Faso, Cameroon, Central African Republic, Chad, Republic of Congo, Cote d’Ivoire, Ethiopia, Gabon, Ghana, Guinea, Kenya, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Senegal and South Africa.

(9) Countries in the Middle East and North Africa group are Algeria, Egypt, Iran, Jordan, Lebanon, Morocco, Syrian Arab Republic, Tunisia, Turkey and Yemen.

(10) Countries in the East European group are Albania, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Russia, Slovak Republic, Slovenia and Ukraine.

(11) The latter states collectively experienced the largest declines in GDP in peacetime in the history of any modern economies

(12) Keynes derived the tendency of a rising role of investment with economic development from savings behaviour. It formed a cornerstone of his analysis of effective demand and crisis: ‘the richer the community, the wider will tend to be the gap between its actual and is potential production; and therefore the more obvious and outrageous the defects of the economy system. For a poor community will be prone to consume by far the greater part of its output, so that a very modest measure of investment will be sufficient to provide full employment; whereas a wealthy community will have to discover much ampler opportunities for investment if the saving propensities of its wealthier members are to be compatible with the employment of its poorer members. If in a potentially wealthy community the inducement to invest is weak, then in spite of its potential wealth, the working of the principle of effective demand will compel it to reduce its actual output, until, in spite of its potential wealth, it has become so poor that its surplus over its consumption is sufficiently diminished to correspond to the weakness of the inducement to invest.’ (Keynes, 1983, p. 31)

(13) For example analysing US industrial sectors Jorgenson, Gollop and Fraumeni found that intermediate inputs were the largest source of growth. They noted:

‘the contribution of intermediate input is by far the most significant source of growth in output. The contribution of intermediate input alone exceeds the rate of productivity growth for thirty six of the forty five industries for which we have a measure of intermediate input… the predominant contributions to output growth are those of intermediate, capital and labour inputs. By far the most important contribution is that of intermediate input.’[xiii] (Jorgenson, Gollop, & Fraumeni, 1999, p. 200)

Considering such findings for the US in more detail, Jorgenson concluded:

‘The analysis of sources of growth at the industry level is based on the decomposition of the growth rate of sectoral input into the sum of the contributions of intermediate, capital and labour inputs and the growth of sectoral productivity… The sum of the contributions of intermediate, capital, and labour inputs is the predominant source of growth of output for 46 of the 51 industries…

‘Comparing the contribution of intermediate input with other sources of growth demonstrates that this input is by far the most significant source of growth. The contribution of intermediate input exceeds productivity growth and the contributions of capital and labour inputs. If we focus attention on the contributions of capital and labour inputs alone, excluding intermediate input from consideration, these two inputs are a more important source of growth than changes in productivity… The explanatory power of this perspective is overwhelming at the sectoral level. For 46 of the 51 industrial sectors… the contribution of intermediate, capital and labour inputs is the predominant source of output growth. Changes in productivity account for the major portion of output growth in only five industries. (Jorgenson D. W., 1995, p. 5)

Regarding studies of rapidly growing Asian economies, Ren and Sun found for China that in the period 1981-2000, subdivided into 1984-88, 1988-94 and 1994-2000: ‘’Intermediate input growth is the primary source of output growth in most industries.’ (Ren & Sun, 2007). For Taiwan, analysing 26 sectors in 1981-99, Chi-Yuan Liang found regarding intermediate material inputs: ‘Material input is the biggest contributor to output growth in all sectors during 1981-99, except… seven’. (Liang C.-Y. , 2007). For South Korea Hak K. Pyo, Keun-Hee Rhee and Bongchan Ha found: ‘The relative magnitude of contribution to output growth is in the order of: material, capital, labour, TFP then energy.’ (Pyo, Rhee, & Ha, 2007)

(14) For an attempt to apply such an analysis to China see (Zheng, Bigsten, & Hu, 2009).

(15) For standard surveys of theories of economic growth see (Barro & Sala-i-Martin, 2004) or (Acemoglu, 2009).

Bibliography

Jorgenson, D. W. (1995). ‘Productivity and postwar US economic growth’. In D. W. Jorgenson, Productivity (Vol. 1, pp. 1-23). Cambridge, Massachusetts: The MIT Press.

Jorgenson, D. W., & Vu, K. M. (2007a). ‘Information technology and the world growth resurgence’. German Economic Review , 8 (5).

Jorgenson, D. W., & Vu, K. M. (2007b). ‘Information technology and the world growth resurgence – updated tables’. Retrieved from Dale Jorgenson: http://www.economics.harvard.edu/faculty/jorgenson/recent_work_jorgenson

Jorgenson, D. W., & Vu, K. M. (2009). ‘Growth accounting within the International Comparison Programme’. The ICP Bulletin , 6 (1).

Jorgenson, D. W., Gollop, F. M., & Fraumeni, B. M. (1987). Productivity and US Economic Growth. Harvard University Press.

Keynes, J. M. (1936). The General Theory of Employment, Interest and Money. London: MacMillan.

Krugman, P. (1994). ‘The myth of Asia’s miracle’. Foreign Affairs , 62-78.

Liang, C.-Y. (2007). ‘Industry-wide total factor productivity and output growth in Taiwan, 1981-1999’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia: Economic Growth and Competitiveness (pp. 146-184). Cheltenham: Edward Elgar.

Pyo, H. K., Rhee, K.-H., & Ha, B. (2007). ‘Growth accounting and productivity analysis by 33 industrial sectors in Korea (1984-2002)’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia: Economic Growth and Competitiveness (pp. 113-145). Northampton: Edward Elgar.

Ren, R., & Sun, L. l. (2007). ‘Total factor productivity growth in Chinese industries, 1981-2000’. In D. W. Jorgenson, M. Kuroda, & K. Motohashi (Eds.), Productivity in Asia, Economic Growth and Competitiveness (pp. 76-112). Northampton: Edward Elgar.

Ross, J. (2009, September . ‘The Asian and Chinese economic growth models – implications of modern findings on economic growth’. Retrieved from Key Trends in Globalisation: http://ablog.typepad.com/keytrendsinglobalisation/2009/09/the-issue-of-whether-chinas-economic-stimulus-package-and-the-asian–growth–model-in-general-is-correct-and-therefore-its.html

Smith, A. (1999). The Wealth of Nations. London: Penguin.

Vu, K. M. (2007). ‘Determinants of economic growth over the period 1995-2005′. Retrieved from http://docs.google.com/viewer?a=v&q=cache:s7odRXb7owsJ:www.spp.nus.edu.sg/Handler.ashx%3Fpath%3DData/Site/SiteDocuments/wp/wp0907.pdf+%22Determinants+of+Economic+Growth+Over+the+Period+1995-2005%22+Vu&hl=en&gl=uk&pid=bl&srcid=ADGEESj9omj1MsLK6PX0-ns9×4Gvz

Zheng, J., Bigsten, A., & Hu, A. (2009). ‘Can China’s growth be sustained? A productivity perspective’. World Development , 37 (4), 874-888.

The real situation in China’s trade

The last two years in China’s international trade, including the July 2010 data, have provided a clear factual test between two theories regarding China’s trade surplus.

• The first, strongly promoted by US ‘neo-con’ circles, argues that the key factor in China’s trade balance is the RMB’s exchange rate. Therefore, it states, the decisive step to keep down China’s trade surplus is to revalue the RMB. The factual evidence clearly contradicts this view, as will be shown.
• The second theory, argued by this blog and others, is that the key factor in China’s trade balance is China’s relative rate of GDP growth compared to other economies. Therefore, on this analysis, what is required to keep down China’s trade surplus is that China’s economy should grow rapidly.

As most analysts would evidently agree that both exchange rates and relative growth rates affect trade balances, the key question is which of the two is most quantitatively important. It is this that the trade data clearly answers. The decisive factor has been shown to be the relative growth rate of growth of China’s economy not the exchange rate.

July’s trade surplus, as well the dramatic shrinkage of China’s surplus during 2009 and the first quarter of 2010, did not follow changes in the RMB’s exchange rate but did follow extremely closely the acceleration and deceleration of China’s economy.

These trends are examined in more detail below. An analysis is given of the angle of approach to dealing with China’s trade surplus and dealing with inflationary pressures in the real estate sector.

Test of the two analyses

The present author has repeatedly pointed out the errors in economic theory and fact of those who have argued that revaluation of the RMB is the key to reducing China’s trade surplus so it is not required to list these again – a couple of responses to the new July trade data can be noted below.

For a detailed examination of trade trends prior to the 2008 financial crisis, which are consistent with those analysed here, readers are referred to an earlier article on this blog and Figure 3 below. The crucial factual point shown in these is that as the RMB’s exchange rate rose, between 2005 and 2008, China’s trade surplus did not fall – as would be predicted by the analysis that RMB revaluation would lead to a shrinking trade surplus. On the contrary, a rise in the RMB’s exchange rate was accompanied by an increase in China’s trade surplus.The reasons for this were dealt with previously.

Turning to July’s trade figures, Sophie Leung and Li Yanping note on Bloomberg: ‘China’s trade surplus reached an 18-month high as exports rose to a record and import gains slowed, adding pressure on officials to allow faster appreciation of the yuan.’

Geoff Dyer in the Financial Times wrote: ‘The pace of increase in exports actually fell last month to 38.1 per cent year-on-year, down from 43.9 per cent in June. However, import growth slowed even more, moving up 22.7 per cent against 34.1 per cent in June. The rising trade surplus will increase the political pressure on Beijing to appreciate its currency more rapidly.’

These authors therefore consider the key issue to mention in relation to trade is the RMB’s exchange rate. They do not draw equal attention to the quantitatively dominant trends shown below.

China’s trade pattern

In the last three months China’s trade surplus has begun to expand significantly – see Figure 1 below, which shows the monthly trade balance, and Figure 2 which shows the monthly balance calculated as a three monthly moving average.

Taking the three monthly moving average, to eliminate the effects of purely short term fluctuations, China’s trade surplus fell steadily from a peak of $39bn in January 2009 to a low of $1bn in April 2010. Since then it has expanded again to $23bn in July 2010.

None of these shifts, however, can be explained by changes in the RMB’s exchange rate as throughout this period, until 18 June 2010, the RMB’s exchange rate against the dollar was entirely stable. After this the RMB’s exchange rate moved up marginally, by less than one percent – although this change is too recent to have significantly affected the figures. In short, China’s trade surplus fell, practically to the point of disappearance, and then rose again despite a stable exchange rate.

While China’s trade surplus did not track changes in China’s exchange rate at all, it did track extremely closely the acceleration and deceleration of China’s economy. During the whole of 2009, and right up to the first quarter of 2010, China’s economy was accelerating strongly. The year on year GDP growth rate speeded up progressively from 6.1% in the 1^st quarter of 2009, to 7.9% in the 2^nd, to 8.9% in the 3^rd, to 10.7% in the 4^th and finally to 11.9% in the 1^st quarter of 2010. Throughout this period China’s trade surplus fell.

In the 2^nd quarter of 2010 China’s GDP growth slowed to 10.3%, while China’s trade surplus began to rise.

In short, while China’s trade surplus did not follow its exchange rate it did follow closely the GDP growth rate. China’s growth, not its exchange rate,therefore controlled China’s trade balance.

This analysis that currently China’s trade surplus was controlled by the economy’s rate of growth and not the exchange rate was set out in detail in an article that appeared on this blog in March commenting on a Ministry of Commerce press conference. It noted: ‘the only way to continue reducing China’s surplus, as [Ministry of Commerce spokesperson] Yao Jian says, is to take “measures to stimulate imports.” This however requires rapid economic growth. China’s imports rose faster than exports in 2009 because its economy grew more rapidly than others. While other markets stagnated or declined, China grew at 8.7 percent, and sucked in imports. The most effective way to maintain the import surge is rapid growth.’

The slowing of China’s economy in the second quarter of 2010 shows the same driving force from the opposite perspective – the economy slowed, imports fell, and therefore the trade surplus increased. But the same mechanism, whereby it is China’s growth rate and not the exchange rate that is controlling the trade surplus, continues to operate.

The answer to how to keep down China’s trade surplus is therefore equally evident. It requires rapid economic growth. That in turn, evidently raises the issue of why it was necessary for the Chinese authorities to start slowing the economy from the 1^st quarter onwards?

What will happen?

The reason for the recent steps to cool China’s economy are are well known – inflationary risks and pressures. However China did not face generalised inflationary pressure, indicating an overall and fundamental disequilibrium between demand and supply, but a concentrated inflationary pressure in two areas – food and housing.

The 3.3% increase in China’s consumer price index in July was primarily driven by food prices which rose by 6.8%. Non-food inflation remained subdued at 1.6%. The key factor in asset price inflation is housing – the share market in contrast remains lower than its level a year ago. The annual rate of increase of housing prices in July 2010 fell to 10.3% year from June’s 11.4% but this was still too high – hence economic policy makers determined action to cool the housing market.

Therefore the direction of causation is clear. The situation in the housing market is the key factor in inflation. Inflation is the reason China had to slow its economy – and therefore its trade surplus began to rise. A key to controlling China’s trade surplus, therefore, lies in China’s housing market. The chief domestic issue facing China’s economy, and the chief international one, are not separate but interconnected.

This is a far from unique situation – both the UK and the US have frequently seen situations in which severe fluctuations in housing prices were correlated with shifts in the balance of payments.

In such situations macroeconomic fundamentals and the immediate economic situation coincide. A balance of payments surplus, of which the trade surplus is the dominant part, is by accounting identity equal to the deficit of China’s domestic investment compared to its domestic savings. To narrow the trade surplus therefore requires either reducing saving – or raising consumption which is the same thing, or increasing investment. Under conditions of inflationary pressure simply raising consumption demand, without an increase in supply, is not called for – adding to demand and not supply will increase inflationary pressure. The macro-economic way to deal with this situation, under conditions of threatening inflation, must therefore be to raise investment.

The situation in the housing market shows both where supply is lacking compared to demand and the target for such investment. Overall China has experienced a major rise in the percentage of fixed  investment in GDP since 2008. A further general rise therefore does not appear called for. But overheating in the housing market is sufficiently severe that it clearly needs to be addressed not only by dampening demand, which the Chinese authorities have been attempting since early in the year, but also by increasing supply – that is by a large house building programme. Such a policy has been embarked on, but only recently.

Housing and the trade surplus

The above data therefore shows the angle of approach for a key policy by which China’s trade surplus needs to be tackled. To permit resumption of more rapid growth inflation must be contained. A decisive key to containing inflation is the situation in the housing market. An increase in investment in the housing market will help contain inflationary pressures, thereby putting downward pressure on the trade surplus via more rapid growth, and by increasing investment it will act directly to reduce the trade surplus by shifting the balance between investment and saving.

One of the most important ways for China to tackle its trade surplus is to therefore house building. As large scale housing programmes involves financial risk, experience in other countries is that state action is required to underpin very large scale housing construction. Such a housing programme must therefore be underpinned by the state.

Conclusion

To summarise, the facts show the RMB’s exchange rate does not control China’s trade surplus – indeed, due to J curve effects if the RMB is revalued this will increase China’s trade surplus in the next period. If anyone outside China wants to help reduce China’s trade surplus, therefore, they would urge China to accelerate its economic growth and, to achieve this, to embark on a large scale housing programme to help contain real estate asset inflation. Such a policy, not an increase in the RMB’s exchange rate, which has again been shown during the last two years not to control China’s trade balance, is the real way to keep down China’s trade surplus.

Why the correct policy to deal with the trade surplus meets US neo-con opposition

China’s economic policy makers do indeed appear to have decided to embark on large scale house building. This is evidently linked to the issue of accelerated urbanisation. Such a programme therefore needs to be seen not only from the housing but from the macroeconomic perspective. A prolonged programme of housing construction is key not only for China’s economic development but also for dealing with its trade surplus.

Such a programme of maintained or accelerated growth, to a greater degree powered by housing construction, is however naturally unacceptable to ‘neo-con’ circles in the US who, above all, want to ensure that the US’s economy remains larger than China’s and for whom, therefore, rapid growth in China is anathema. That is why they continue to promote a course, RMB revaluation, which would not reduce China’s trade surplus but would lower China’s growth rate, rather than one that actually would hold down China’s trade surplus.

Those who are interested in co-ordinated international economic growth, as opposed to playing political games, will therefore hope China successfully increases supply in its housing market and, by co-ordinating this with demand side measures, allows rapid economic growth be maintained. That is a key policy to deal with the trade surplus.

So far China’s economic policy makers have shown a markedly clearer understanding of economic fundamentals and required policy than the US – to which the success of China’s economic stimulus package and the weakness of the recovery in the US bears graphic witness. It is to be hoped that this clear grasp of economic fundamentals continues.

Charts

Figure 3

The data confirms the US recovery is weak

Unsurprisingly, because it was anticipated, and as has been widely reported, the data confirmed the slowdown in US economic recovery. Taking the latest revised figures, annualised US GDP growth decelerated from 5.0% in the 4th quarter of 2009, to 3.7% in the 1st quarter of 2010 to 2.4% in the 2nd quarter.

US GDP remains 1.1% below its peak level in the 4th quarter of 2007. At the 2nd quarter’s rate of growth previous peak US GDP will not be regained until the 4th quarter of 2010.

Such figures have essentially decided the debate between those who argued that because the US downturn was very severe its economy would spring back strongly from recession, and those, such as the present author, who pointed to the underlying structural situation of the US economy and therefore argued recovery would be weak compared to previous US post-war business cycles.

Figure 1 illustrates how much weaker the present US economic recovery is than in previous post-war cycles. In the previous most serious post-war cyclical downturn, that following 1973, the US economy regained its previous peak level of production after eight quarters. In this recession after 10 quarters the US economy has still not recovered its peak GDP level.

As also widely reported, the new GDP data now calculates the US recession was deeper, and started earlier, than previously estimated. Peak US GDP is now analysed as having occurred in the 4th quarter of 2007 rather than the 2^nd quarter of 2008 as previously estimated. The trough of US GDP in the 2nd quarter of 2009 is now calculated to have been 4.1% below the cyclical peak level – the previous deepest fall in a US post-World War recession, that after 1973, was 3.2%.

The fall in US investment

The driving force of the depth of the US recession is also clear. It was due to the decline in US fixed investment. As shown in Figure 2, measured in constant 2005 prices, US GDP in the 2nd quarter of 2010 was $147 billion below its 4th quarter 2007 level. However several components of US GDP are already above their 4th quarter 2007 levels – inventories are up $63 billion, government consumption up $112 billion, and net trade up $135 billion. Consumer expenditure was below its 4th quarter 2007 level but only by $80 billion. But US private fixed investment was down $412 billion – dwarfing all other contributions to the recession.

Figure 2

Fixed investment and inventories

A further feature indicating the specific pattern of US recovery is the financing of gross domestic investment – i.e. fixed investment plus inventory accumulation. Although, as noted above, US fixed investment remained severely depressed, nevertheless for the first time for four years there was a small upturn, of 0.5%, in the percentage of US GDP devoted to fixed investment in the 2nd quarter – fixed investment rose to 15.6% of GDP from its low of 15.1% in the 1st quarter of 2010. This reflected a stabilisation of the share of residential investment in GDP and a slight increase in the share of non-residential investment – see Figure 4.

Figure 4

Nearly two years into the financial crisis it is possible to show clearly in figures which of these two contrasting analyses is correct – they evidently lead to different conclusions as regards policies. As will be shown below the overwhelming driving force of the Great Recession is a collapse in fixed investment, not a decline in US consumption – or consumption in other economies. This also casts clear light on why China has been the country which has come the most successfully through the financial crisis. To summarise the statistical conclusions below:

‘Decline in fixed investment accounts for approximately 96% of the fall in GDP in the OECD area as a whole and for 76% of the decline of GDP in Europe. In three countries – the US, Spain, and Portugal – the decline in fixed investment was greater than the decline in GDP. In Japan, France and Greece the proportion of the fall in GDP due to the decline in fixed investment was over 70%, 80% and 90% respectively. In every country except Germany the fall in fixed investment was the single biggest component of the decline in GDP. In short the decline in fixed investment entirely dominates the Great Recession’

The focus of this article is therefore a detailed factual account of what has actually occurred during the Great Recession. These facts leave no doubt. The ‘Great Recession’ is actually ‘The Great Investment Collapse’. Policies for dealing with the Great Recession must therefore primarily address reversing the investment decline.

The OECD as a whole

Taking first aggregate changes in the components of GDP in the OECD area as a whole, i.e. all advanced economies, Figure 1 shows these since the beginning of the economic downturn after the first quarter of 2008 up to the latest available aggregated OECD data. As may be seen the GDP fall is entirely dominated by the decline in fixed investment.

During the period from the first quarter of 2008 to the fourth quarter of 2009 OECD GDP fell by $1.04 trillion dollars in constant parity purchasing power (ppp) terms – the form in which the OECD aggregates data. Of this fall $0.99 trillion, equivalent to approximately 96 percent, was accounted for by a decline in fixed investment. In contrast the decline in personal consumption expenditure was $0.25 trillion, only one quarter of the decline in investment, government consumption rose by $0.23 trillion and the balance of trade of the OECD economies improved by $0.23 trillion.

Figure 1

The US

Turning to the US, the changes in the components of GDP in the downturn after the second quarter of 2008 to the first quarter of 2010 are shown in Figure 2. Again, as may be seen, the fall in US GDP is entirely dominated by the decline in fixed investment. During this period US GDP in constant price terms, 2005 dollars at annualised rates, fell by $177bn. However most components of US GDP actually rose over the period as a whole – consumer expenditure by $8bn, government expenditure by $58bn, inventories by $78bn, and net trade by $108bn.

The entire decline of US GDP is therefore due to the $420bn decline in fixed investment.

Figure 2

In order to avoid any suggestion that this investment decline is due simply to the fall in residential investment, propelled by the sub-prime mortgage crisis, Figure 3 divides the decline in US fixed investment in the period into residential and non-residential. The decline in non-residential fixed investment is $310bn and the decline in residential fixed investment is $110bn – i.e. the decline in US fixed investment is overwhelmingly accounted for by the fall in non-residential investment.

Figure 3

Europe

Taking Europe as a whole, the changes in the components of GDP in the downturn after the first quarter of 2008, up to the latest available OECD data, are shown in Figure 4. The fall in GDP is again dominated by the decline in fixed investment.

In constant price ppp dollars, the form in which the OECD calculates aggregated data, the GDP of the OECD area in Europe fell by $583.5bn. Of this $445.7bn, or approximately 76%, was due to the decline in fixed investment. Personal consumer expenditure fell by $145.6 bn, while government consumption rose by $123.3 bn and net trade improved by $47.4bn.

Figure 4

Japan

Turning to Japan, the changes in the components of GDP in the downturn after the first quarter of 2008, up to the latest available OECD data, are shown in Figure 5. The downturn in almost all components of Japan’s GDP, except government consumption, is severe. However by far the largest decline is accounted for by the fall in fixed investment.

During the period since the start of the economic downturn Japan’s GDP, in constant price terms, fell by ¥8.3 trillion. Government consumption rose by ¥0.3 trillion while personal consumption fell by ¥1.1 trillion, and net trade worsened by ¥2.2 trillion. However fixed investment fell by ¥5.7 trillion – i.e. approximately 69% of the fall in Japan’s GDP was due to the decline in fixed investment.

Figure 5

Germany

Turning to the individual major European economies, the components of GDP in the downturn in Germany’s economy after the first quarter of 2008, up to the latest data for the first quarter of 2010, are shown in Figure 6. Germany is specific in that, as will be seen, it is the only major economy in which the worsening of the net trade balance is greater than the decline in fixed investment in terms of its impact on GDP. The combination of the fall in investment plus the worsening of the net trade balance accounts for the severity of the German recession – GDP in Germany in the first quarter of 2010 was still 5.3% lower than in the first quarter of 2008.

In constant price terms German GDP fell by €30.4bn between the first quarter of 2008 and the first quarter of 2010. Government expenditure rose by €6.0bn, personal consumption fell by €5.6bn, fixed investment fell by €14.7bn and net trade worsened by €21.6bn.

Figure 6

France

For France Figure 7 shows the changes in the components of GDP in the downturn after the first quarter of 2008 up to the first quarter of 2010. As may be seen, the fall in GDP is dominated by the decline in fixed investment. During this period France’s GDP in constant price terms fell by €11.7bn. Net trade worsened by €1.0bn while personal consumption rose by €3.6bn and government consumption by €4.6bn. However fixed investment fell by €11.7bn – i.e. approximately 87% of the fall in France’s GDP was due to the decline in fixed investment.

Figure 7

The UK

Figure 8 shows the changes in the components of UK GDP in the recession after the first quarter of 2008 up to the first quarter of 2010. Again, as may be seen, the largest component of the fall in GDP is fixed investment.

In constant price terms UK GDP fell by £18.6bn. Net trade improved by £2.0bn, government consumption rose by £3.6bn, and personal consumption fell by £8.3bn. However fixed investment fell by £10.5bn – i.e. approximately 56% of the fall in UK GDP was due to the decline in fixed investment.

Figure 8

Italy

Turning from the major north European economies to the southern European states, the so called PIGS (Portugal, Italy, Greece, Spain), the situation is equally clear.  Figure 9 shows the changes in the components of GDP in Italy in the recession following the first quarter of 2008 up to the first quarter of 2010. In constant price terms Italy’s GDP fell by €19.6bn. Government consumption rose by a marginal €0.3bn, net trade worsened by €4.4bn and personal consumption fell by €4.8bn. Fixed investment however fell by €10.0bn – i.e. 51% of the fall in Italy’s GDP was due to the decline in fixed investment.

Figure 9

Spain

Turning to Spain the changes in the components of GDP in the downturn after the first quarter of 2008, up to the first quarter of 2010, are shown in Figure 10. The fall in GDP is dominated by the decline in fixed investment.

During this period Spain’s GDP in constant price terms fell by €9.2 billion. Government consumption rose by €3.0 billion and net trade improved by €10.6 billion as Spain began to reverse its wide balance of payments deficit. There was a significant fall in personal consumption of €7.1 billion but by far the dominant element was the €13.7 billion fall in fixed investment. The fall in fixed investment in Spain was greater than the entire decline in GDP.

Figure 10

Portugal

Figure 11 shows the changes in the components of GDP in Portugal in the recession after the first quarter of 2008 up to the first quarter of 2010. In constant price terms GDP fell €0.8bn, net trade improved by €0.2bn while personal consumption increased by €0.3bn and government consumption by €0.4bn. Fixed investment however fell by €1.5bn – more than the entire decline in GDP.

Figure 11

Greece

Finally Figure 12 shows the changes in the components of GDP in Greece from the beginning of its recession, which commenced in the third quarter of 2008, up to the first quarter of 2010. In constant price terms Greece’s GDP fell €2.0bn. Net trade improved by €0.2bn while personal consumption fell by €0.1bn and government consumption by €0.3bn. Fixed investment fell by €1.8bn – i.e. approximately 90% of the fall in GDP was due to the decline in fixed investment.

Figure 12

China

It is evident from the above data that whether considering the advanced economies as a whole, or looking at individual economies, the overwhelmingly dominant element in the economic downturn is the fall in fixed investment.  Decline in fixed investment accounts for approximately 96% of the fall in GDP in the OECD area as a whole and for 76% of the decline of GDP in Europe. In three countries – the US, Spain, and Portugal – the decline in fixed investment was greater than the decline in GDP. In Japan, France and Greece the proportion of the fall in GDP due to the decline in fixed investment was over 70%, 80% and 90% respectively. In every country except Germany the fall in fixed investment was the single biggest component of the decline in GDP. In short the decline in fixed investment entirely dominates the Great Recession. The policy conclusions which follow from this are evident. The decisive question is to reverse the decline in investment.

It is equally clear from this data why China has come most successfully through the financial crisis. China’s government carried out its stimulus package not via an increase in the budget deficit, which has remained at less than 3% of GDP, but by a major increase in infrastructural and other fixed investment.The comparative paths of fixed investment in China and the US under the impact of their stimulus packages in 2009 are shown in Figure 13. Whereas in the US fixed investment fell by twenty percent China’s urban fixed investment rose by more than thirty percent due to the stimulus package. The impacts, in terms of the changes in US and China’s GDP in 2009, are shown in Figure 14.

China’s stimulus package dealt directly with the central issue in the Great Recession. China focussed on  investment – rather than attempting primarily to influence this indirectly via the hope that a stimulus to personal and government consumption, maintained by a large budget deficit, would induce a reversal of the investment decline. Consequently China was able to launch a large stimulus package without running a large budget deficit. The GDP growth induced by the investment rise in turn produced large scale tax revenue – China’s fiscal income in 2009 rose by 11.7%, compared to major declines in tax revenue in the US and European economies. China therefore currently does not face the choice faced by the US and European governments of whether to maintain large scale budget deficits, to attempt to sustain economic stimulus, or whether to engage in fiscal consolidation. In light of the actual character of the Great Recession China’s stimulus package was therefore significantly better designed than those in the US and Europe.

Understanding what has been the real core of the Great Depression will therefore better aid the policy response in recovering from it.

From Paul Krugman Blog – NYTimes.com

June 28, 2010, 11:37 am

‘The Invisible Bond Vigilantes Continue Their Invisible Attack

via krugman.blogs.nytimes.com

Paul Krugman is not an economist I always agree with – despite his being a nice person and a generally liberal voice. His interpretation of Keynes as being primarily about budget deficits is not an accurate presentation of Keynes. And, a related issue, his claim that the main driving force of the US economic downturn was not a decline in investment, which he made in discussion when I was on a panel with him in Shanghai, has been confirmed to be wrong. However he is entirely right on one key issue – that the US budget deficit, in present economic conditions, would not lead to a rise in US interest rates. The idea that in the circumstance of a deep recession in the US the budget deficit would lead to ‘crowding out’ of other demand for capital, which is what the idea that the budget deficit will lead to high interest rates is based on, is nonsense.The above comment on his New York Times blog is therefore entirely valid regarding an issue on which he has carried out a long polemic with, and been vindicated, against Niall Ferguson and others.

Such arguments consist of two sentences. ‘China runs a trade surplus. Therefore to eliminate it China should increase the exchange rate of the RMB.’

Unfortunately elementary economic reflection will show that the second sentence does not necessarily follow from the first. Consideration of supply and demand reminds us that an increase in the exchange rate of the RMB will only reduce China’s export earnings if demand for China’s exports is elastic – that is any percentage fall in sales is greater than any percentage rise in price resulting from revaluation. Equally China’s imports in value terms will only rise if any increase in their volume is greater than any fall in their price due to revaluation.

The question of whether China’s trade surplus will fall or rise in response to RMB revaluation is therefore a matter of fact, not of logic, which therefore has to be examined empirically – as the paper below notes. It quite simply does not follow that an increase in China’s exchange rate will logically necessarily lead to a fall in China’s trade surplus. Indeed it is quite possible logically, for example if demand for China’s exports is inelastic, and the volume of its imports is not particularly price sensitive, that a rise in the RMB’s exchange rate will lead to an increase in China’s trade surplus.

Given the seriousness of the issue one would have thought that if this matter were being dealt with objectively the US administration would have produced a mountain of material to justify its claim that an increase in the RMB’s exchange rate would lead to a fall in China’s trade surplus – China has certainly produced abundant data, including directly by the Commerce Minister, showing the opposite. But no such material has been forthcoming from the US administration. Instead there is the intoning of a literal non-sequitur.

The reason evidence has not been produced by either the US administration or by those in agreement with it is that at least as regards the immediate and medium term economic situation their argument is factually false. As is shown in the paper below an increase in the RMB’s exchange rate would immediately lead to an increase in China’s trade surplus and not to a fall – and this is one of the last things which the world requires while attempting to emerge from the international financial crisis.

This paper was written in April and published in Chinese. It deals with a wider range of issues than simply bilateral China-US trade. The opening paragraph has been amended to remove purely contemporary references and altered to read in the past tense. The rest of the article is unchanged – its fundamental arguments do not require revising.

*   *   *

At the meeting between President Obama and President Hu Jintao earlier this year the Chinese side noted that RMB appreciation would not balance Sino-U.S. trade. This conclusion reiterated other studies published by Chinese ministers and economic specialists on bilateral US-China trade. This conclusion is in line with the data below.

However it should also be noted that 84% of China’s foreign trade is with countries other than the US. A change in the RMB’s exchange rate would therefore have not only bilateral consequences between the US and China but wider effects on the world economy. This article therefore examines these. The most important conclusion it demonstrates is that:

• The immediate effect of an increase in the RMB’s exchange rate would be to increase China’s trade surplus and not decrease it – contrary to apparent expectations of the US administration;

It is important to understand this dynamic so that wrong anticipations of events do not exist in the US, China or elsewhere.

In addition:

• RMB revaluation would produce limited but distinct inflationary pressure not only in the US but in the international economy due to China’s position as the world’s largest exporter;
• As 84% of China’s trade is with countries other than the US the consequences for the rest of the world economy of early RMB revaluation would be greater than for the US.
• At present China is in trade balance or deficit with the rest of the world apart from the US – limiting trade frictions to a relatively few countries. RMB revaluation would probably shift China to running a trade surplus with the rest of the world apart from the US, widening the range of countries with which trade frictions are possible.

While Chinese economic policy makers may have valid overriding domestic reasons for an early increase in the RMB’s exchange rate, such as the struggle against inflation or to aid cooling the economy, this will not resolve trade issues. Indeed it should be recognised that in the field of trade there is a significant risk of negative trends created by an RMB revaluation which must be taken into consideration.

The aim of this paper is, therefore, to clarify factual understanding of what would actually occur if an increase in the RMB’s exchange rate occurred.
The empirical evidence on the consequences of an increase in the RMB’s exchange rate

The argument of those favouring an early increase in the exchange rate of the RMB is that as China runs a trade surplus it should revalue the RMB to reduce it. Unfortunately the second part of the argument only follows from the first if factually RMB revaluation would lead to a reduction in China’s surplus. This is an issue of economic fact – of the degree of the elasticity of demand for China’s exports and imports and of their independence of movement – and it does not follow as a matter of logic. Furthermore it is necessary to distinguish between effects in the short term and the long term.

If for example, other things remaining equal, over any given period of time the RMB’s exchange rate went up 10% but China’s export volume fell by only 5% then China’s trade surplus would actually increase due to RMB revaluation. For China’s trade surplus to fall after revaluation changes in the volume of China’s exports and imports would have to be sufficient to offset the fact that RMB revaluation will produce an increase in China’s export prices relative to import prices.

Most of those arguing for short term RMB revaluation make no attempt to demonstrate this factual linkage – they illegitimately claim it without proof. The reason they do not attempt to prove it is that particularly in the short term, which is crucial for the world as it emerges from the financial crisis, it is false.

The consequences of RMB revaluation

Factually, as a number of Chinese authors have pointed out, over the short to medium term it is unnecessary to rely purely on theoretical models to examine the consequences of an increase in the RMB’s exchange rate on bilateral trade with the US. As they note, the increase in RMB’s exchange rate from July 2005 to August 2008 led not to a decrease of China’s bilateral trade surplus with the US but to its increase. However this trend of China’s trade surplus increasing as the RMB’s exchange rate rose was general and not simply a bilateral one with the US.

Taking US figures for its own trade and the IMF’s for total trade, to avoid any suggestion of relying on China’s data which might be alleged to be biased, and to remove secondary differences on the calculation of bilateral US-China trade balances, the US trade deficit with China increased from $162 billion in 2004, the year before RMB revaluation, to $268 billion in 2008 – a rise of $106 billion or 65%. However over the same period China’s trade balance with the rest of the world moved from a deficit of $103 billion to a surplus of $93 billion – a movement of $196 billion in China’s favour. The factual evidence over a three year period therefore showed that an increase in the RMB’s exchange rate led to an increase in China’s trade surplus, and not to a decline, not only with the US but with the rest of the world economy.

This trend is shown clearly in Figure 1 – which graphs the way in which the increase in China’s exchange rate was accompanied not by a fall in China’s trade surplus but by an increase.

Figure 1

The trade surplus was due to a fall in the relative value of China’s imports

The explanation of why China’s trade surplus increased as the RMB’s exchange rate rose is clarified by Figure 2 – which shows China’s exports and imports as a percentage of GDP. As may be seen, prior to 2004, during the period of RMB exchange rate stability, China’s exports and imports rose in parallel as a percentage of GDP. Consequently no trade surplus developed.

From 2005 onwards, however, China’s exports continued to rise as a percentage of GDP but its imports began to fall. Factually, therefore, the emergence of China’s trade surplus was not due to acceleration of exports, as is frequently claimed, but to a relative fall in the value of China’s imports.

To judge from their statements US commentators, and others, advocating an early increase in the RMBs exchange rate do not appear to have studied this factual trend.

Figure 2

Why an increase in the RMB’s exchange rate led to an increase in China’s trade surplus

There is a clear economic explanation of the factual trends shown above – i.e. that as long as the RMB’s exchange rate remained stable exports and imports moved in parallel in value terms, and no trade surplus developed, but when as an increase in the exchange rate took place the total value of imports fell relative to the total value of exports and therefore a trade surplus develop. This trend would necessarily be the case if China’s exports and imports moved together in volume terms, or if the effects of the shifts in relative volumes were lower than the effects of the shifts in relative prices. In that case the effect of RMB revaluation would put up China’s export prices relative to its import prices, the changes in export and import volumes would be less that the effect of the rise in export prices relative to import prices, and therefore an increase in the RMB’s exchange rate would increase the trade surplus. The explanation of this tendency of exports and imports to relatively move together in volume terms would be if a large proportion of imports were inputs into exports – as is the case with China.

That this is the explanation of why the increase in the RMBs exchange rate after 2004 led to an increase in China’s trade surplus, and not to its reduction, is indicated in Figure 3, which shows the UN’s calculations for the movement of China’s exports and imports in fixed price, i.e. volume, terms from 1998-2008. As may be seen the volume of China’s exports and imports moved essentially in parallel throughout this period. Under those circumstances an increase in the RMB’s exchange rate would necessarily lead to an expansion of China’s trade surplus – as occurred.

The ‘J curve’ effect

To analyse the effect of RMB revaluation over an even shorter time frame, it will be assumed, simply for the sake of argument, that the long term effect of an increase in the RMB’s exchange rate would gradually break down the tendency of the volume of China’s exports and imports to move relatively in parallel – although, it may be noted that after July 2005 this did not occur over a three year period, and under conditions of a 21% revaluation, and therefore a considerable delay in adjustment should be anticipated. However even in this case the short term effect on an RMB revaluation would be to increase China’s trade surplus.

The reason for this is the well established short-term ‘J curve’ effect regarding currency exchange rate changes. Under this effect, when a currency’s exchange rate changes, it takes time for demand to adjust. Therefore in the short term, whatever the long term shifts, the change in volumes relative to the effect of price shifts is relatively weak. Revaluation increases export prices and reduces import prices. The ‘J curve effect’ of an RMB revaluation would increase China’s trade surplus in the short term.

However even a short term increase in China’s trade surplus would reverse the most significant present source of international demand and be negative for the world economy as it attempts to emerge from the financial crisis.

It may be noted that purely modelling studies on the long term effects of an increase in the RMB’s exchange rate on China’s trade surplus are mixed. Some find that it would lead to a fall in China’s trade surplus and some that it would lead to an increase. The empirical evidence of a three year rise with a 21% revaluation after 2005 is, however, that it led to an increase in China’s trade surplus and not a fall.

Unless the evidently implausible assumption is made that China’s economy can remain competitive no matter how high its exchange rate, at some point in a very rapid and very large revaluation the trend seen in 2005-2008 would be reversed, and a rising RMB exchange rate would lead to a fall in China’s trade surplus. However the empirical evidence is that such an increase in the RMB’s exchange rate would have to be of a very high percentage and of exceptional duration to reduce China’s trade deficit. Such a very sharp and very prolonged increase in the RMB’s exchange rate would, of course, do great damage to China’s economy and under such circumstances China’s trade surplus would only decline after it had continued to increase for possibly a significant period.

On any reasonable assumptions, therefore, the short and probably medium term consequences of an increase in the RMB’s exchange rate would be to increase China’s trade surplus and not reduce it.

The structure of China’s trade

Turning from the direction of change that would be produced in the short to intermediate term of an increase in the RMB’s exchange rate to its consequences, a picture is frequently presented by critics of China’s exchange rate policy which creates the impression that China runs a massive trade surplus with virtually the entire world. It is important to understand that the factual situation is the opposite. The trade data produced by the US itself shows that China’s trade has frequently been in deficit with the rest of the world apart from the US, that the surplus in China’s non-US trade which did appear after 2006 is declining rapidly, and that China has almost certainly returned to a situation where its trade with the rest of the world, apart from the US, is once again in balance or in deficit.

It should be again made clear that this situation is clear from US data and does not require acceptance of China’s own data or resolution of the statistical differences between US and China. In the following IMF figures are used for China’s total trade.

China’s trade is in deficit with the rest of the world excluding the US

To understand a difference between widely cited US and Chinese trade figures It should be noted that the most frequently quoted US trade statistics differ from the trade data usually quoted in China in that they exclude indirect trade costs such as transport, insurance etc. – i.e. US data is usually cited on a balance of payments basis whereas China’s statistics are usually cited on a balance of trade basis. Calculated on the US basis Figure 4 shows China’s total trade balance, its trade balance with the US, and China’s balance on non-US trade.

The pattern of China’s trade with the rest of the world, apart from the US, is evident. Until 2006 China ran a deficit on non-US trade. In 2007 and 2008 a surplus appeared – of $57 billion and $93 billion respectively, and in 2009 this again fell to around $15 billion. As China’s trade surplus has continued to fall in 2010 it is almost certain that China’s trade is now once again in balance or deficit with the rest of the world apart from the US.

The geographical distribution of China’s trade

Turning to the more specific geographical structure of China’s trade, US statistics indicate that 76% of China’s exports went to countries other than the US – China’s trade figures show 83% of its exports going to non-US destinations. Calculated from US figures, 94% of China’s imports come from countries other than the US – approximately the same figure as in China’s trade data. In total, even on US figures, 84% of China’s trade is with countries other than the US.

An immediate consequence of the US argument on RMB revaluation is evident from this geographical distribution of China’s trade. The US is proposing to address a bilateral issue, that of the US-China trade balance, with a solution, an increase in the RMB’s exchange rate, which would affect not only the US but all countries. Indeed by far the greatest part of the direct impact of an increase in the RMB’s exchange rate, that is 84% of it, would be outside the US while only 16% would be on trade with the US.

It may also be noted from the graph above that the effect of the increase in the exchange rate of the RMB on China’s balance on non-US trade was much greater than its effect on trade with the US. Between 2004, the year before RMB revaluation commenced, and 2008 the increase of China’s trade surplus with the US was $106 billion. The shift in China’s favour in the balance of non-US trade was however $196 billion – from a deficit of $93 billion to a surplus of $103 billion. This indicates a danger that the short/intermediate term increase in China’s trade balance with countries other than the US which would result from an increase in the RMB’s exchange rate would be great than with the US – creating a potential for widening of trade frictions.

International impact of RMB revaluation

Analysing this quantitative structure of China’s trade, the general effect of an increase in the RMB’s exchange rate on inflation and living standards in other countries may be gauged. RMB revaluation, other things remaining equal, puts up China’s export prices. While the indirect inflationary impacts of such export price increases are difficult to calculate, as they tend to be self-reinforcing, the approximate magnitude of the direct inflationary effect of an increase in RMB prices due to revaluation is evident from this data.

A number of China’s commentators have pointed out that an increase in China’s export prices would reduce living standards and add to inflationary pressure in the US. To put figures on the scale of this effect, China’s exports to the US in 2007-2009 were slightly above 2% of US GDP in each year ($321 billion in 2007, $337 billion in 2008, and $296 billion in 2007). A rough guide to the direct inflationary effects, other things being equal, is therefore that a 10% increase in the RMB’s exchange rate would add 0.2% to US inflation, a 20% revaluation of the RMB would add 0.4% to US inflation etc. Other factors remaining equal, this would translate into a reduction of the same order of magnitude in US living standards. Given that China’s imports are largely in the lower part of the price spectrum, the effect on the least well off US consumers would be greatest.

However due to the relatively low percentage of trade in US GDP the inflationary impact, or put in other terms the effect in lowering living standards, on the rest of the world of an RMB revaluation rate would actually be somewhat greater than for the US.

Assuming China’s exports are priced in dollars, as in the majority of cases, in 2008, the latest year for which full figures are available, China’s merchandise exports to non-US destinations were equivalent to 2.6% of the world’s GDP excluding the US and China. Other things being equal, therefore, the direct effects of a 10% increase in the RMB’s exchange rate would add approximately 0.26% to world inflation, a 20% increase in the RMB exchange rate would add 0.52% to world inflation etc. A 40% increase in the RMB’s exchange rate, as demanded by some in the US, would add in terms of direct effect more than 1% to world inflation. Other factors remaining equal, there would be equivalent reductions in world living standards.

Naturally, given the complex interactions involved in determining inflation and consequent reductions in living standards, these figures on direct effects only give guides to orders of magnitude. But they are sufficient to confirm that the greater part of the effect of any increase in the exchange rate of the RMB would be felt outside the US – inevitably given the geographical distribution of China’s trade. China’s position as the world’s largest exporter means that changes in its exchange rate have a perceptible effect on international inflation and living standards.

It is superfluous to note that such pressures to inflation and reduction in living standards are undesirable – particularly in conditions where the world is recovering from the worst financial crisis for eighty years.

Short term consequences of RMB revaluation

Turning from such structural effects to more short term consequences the latter might be roughly defined as covering a 6-18 month period. Normally it might be possible to ignore such short term effects, but this cannot be done under conditions where the world economy is still recovering from severe financial meltdown.

To evaluate short term consequences it is necessary to recall that China is not only the world’s largest exporter but also its fastest growing importer. Since the outbreak of the international financial crisis, China’s imports have provided the world’s single biggest boost in external demand for other economies.

In annualised terms between July 2008 and December 2009 OECD data shows US imports fell by $550 billion and China’s imports rose by $100 billion, while in the same period the US trade deficit shrank by $364 billion and China’s surplus fell by $180 billion. On an annualised basis the US therefore was subtracting $364 billion from international net demand while China was adding $180 billion.

Since more than half the goods China imports are inputs to exports, a cut in the volume of China’s exports, due to RMB revaluation, would lead to reductions in its imports – evidently hitting countries such as Japan, South Korea, Germany and Australia especially hard.

Furthermore the US explicitly aims to reduce its trade deficit – thereby reducing net international demand. China’s policy is to expand international demand by boosting imports and cutting its trade surplus. A reversal of the trend whereby China’s trade surplus was falling, which would be caused by RMB revaluation, would therefore have an undesirable effect in reducing world external demand.

Growth of imports and domestic economic policy

Finally, as RMB appreciation, that is a movement in the relative price of exports and imports, would for the reasons outlined lead to an immediate increase in China’s trade surplus, therefore, what means are available to reduce the trade surplus? As this cannot be achieved in the short to intermediate term by price changes the answer necessarily has to be found in effects on volumes. Given that reduction in exports is undesirable, as China’s export industries have still not recovered their previous volumes of output prior to the financial crisis, the only workable solution to narrowing China’s trade gap further, as Commerce Ministry spokesperson, Yao Jian stated, is to take “measures to stimulate imports.”

This of course occurred during 2009 and 2010 and narrowed China’s trade surplus. The pre-financial crisis peak of China’s exports and imports was reached in July 2008. By March 2010, compared to the pre-crisis peak level, China’s exports had fallen by 18% but its imports had risen by 7%.

Stimulating imports, however, is not primarily a matter of trade delegations – although these are of course useful. It is primarily a function, first, of measures to stimulate domestic demand and second of the rapid growth of China’s economy.

China’s imports rose faster than exports in 2009 in major part because its economy grew more rapidly than others. While other markets stagnated or declined, China grew at 8.7 percent, and sucked in imports. The most effective way to maintain the import surge is rapid economic growth.

Inflation and domestic economic policy

At this point trade intersects with domestic economic policy. The most immediate threat to growth in China is inflation. Inflationary capacity constraints have already emerged. Fortunately China’s 2009 growth pattern puts it in better shape to deal with inflation than if it had followed some of the advice it was offered from abroad. Many commentators who called for revaluation also favoured expanding domestic consumption, but not investment – citing the threat of “overcapacity”. If this policy had been pursued, China would be facing much greater inflationary capacity constraints – threatening economic growth and imports. Happily, China expanded both domestic consumption and investment. The increased capacity that resulted will allow faster growth than would otherwise have been the case, increasing demand for imports.

Mutually beneficial policies

A coherent strategy both from China’s point of view and for creating a “win-win” situation with other economies therefore means avoiding an excessively early RMB revaluation. This will allow China’s exporters to recover and avoid a short term increase in China’s trade surplus. Simultaneously rapid economic growth is not just good for China but also boosts imports. And to make rapid growth possible without inflation China needs to boost domestic investment as well as consumption. Investment, in turn, boosts imports of machinery and equipment. Foreign countries should argue for China to maintain the existing RMB exchange rate in the short run, maintain rapid growth, and, to underpin growth, undertake high levels of domestic investment. This is the combination that will create a “win-win” situation. By contrast, an excessively early RMB revaluation would create a “lose-lose” scenario. It would hit China’s exporters, lead to a short term increase in China’s trade surplus, withdraw an economic stimulus in a recessionary international economic situation and, linked to reducing investment, would exacerbate inflationary capacity constraints. It may be that other economic considerations, notably the fight against inflation, will force China into a premature increase in the RMB’s exchange rate, but from the trade point of view, and that of international economic recovery, this would be undesirable.

Conclusion

Holding the RMB’s exchange rate steady in the short term, while letting it rise in the medium/long term, has so far been the position of China’s government. This seems to have been based primarily on domestic economic considerations. But it also happened to be in the best interests of the world economy.

It may be noted that, of course, trade cannot be the only consideration in setting the RMB’s exchange rate. RMB revaluation may be adopted for reasons such as the struggle against inflation or cooling an overheated economy. These considerations may have to take priority over trade policy. However such decisions should be taken recognising that, in the short term, RMB revaluation will put upward pressure on China’s trade surplus.

Timing

Timing is crucial for the world economy and therefore for RMB revaluation. In the medium to long term evidently the exchange rate of the RMB should and will go up – the increasing productivity of China’s economy makes it competitive at progressively higher exchange rates. But because an increase in the RMB exchange rate would put upward pressure on China’s trade surplus in the short term, from the point of view of world trade it would be better if revaluation did not take place until the world recovery is more firmly established – on current trends likely to be closer to the end of this year.

Appendix – a note on some false arguments by foreign critics of China’s exchange rate policy

In the paper above a number of serious arguments relating to RMB revaluation have been dealt with. It should, however, be noted that a number of simply false claims regarding China’s trade are made by international critics. Two of these, notably that by the influential US economist Paul Krugman, who called for US trade action against China in a New York Times column which has been widely quoted, are therefore dealt with here.

Paul Krugman’s erroneous claim

It was analysed above that China’s trade with the rest of the world, apart from the US, has frequently been in deficit and has probably passed back into balance or deficit in 2010. It may be noted from this reality why some critics make false and exaggerated claims regarding the size of the China’s surplus – a statement of the actual facts would reveal the actual situation that China has not typically, and is not currently, running a surplus in non-US trade. For example Paul Krugman, talking about China’s balance of payments, including service and investment income as well as trade, in his New York Times column claimed that ‘the International Monetary Fund expects China to have a 2010 current surplus of more than $450 billion.’

Of course if China were factually running a $450 billion dollar balance of payments surplus this is sufficiently large it would entail that a trade surplus was being run not only with the US but with the rest of the world. No source is given but if any IMF statistician made such a claim it is inaccurate – as Krugman could have found out by checking the figures.

Once China’s balance on services and income is included, and the standard statistical correction for carriage and insurance made, China’s annual balance of payments surplus is about $120 billion over its unadjusted trade surplus. The latter was $198 billion in 2009 and falling – in the 12 months to February 2010 it was $180 billion. Therefore there is no way China’s balance of payment surplus will reach anything like $450 billion this year. The actual falling trend of China’s trade surplus is shown in Figure 5.

Figure 5

To take a similar example, the British economist Will Hutton claimed in The Observer that China is ‘increasing its reliance on exports.’ Again this is factually false. China’s exports declined to 25 percent of GDP in 2009 from 35.7 percent in 2006 – as shown in Figure 2. In reality the increase in China’s domestic demand has led to a sharp decline in the weight of exports in China’s GDP.

Figure 6

More significant than the marginal $238 million increase in the deficit between March and April, from $40.05 billion to $40.29 billion, was the trend in the three month moving average for the monthly deficit – shown in Figure 2. This widened from $28.4 billion to $40.2 billion. Taking this average since its low point in June 2009 the monthly deficit has increased from $26.8 billion to $40.2 billion. This average deterioration of $13.3 billion equates to an annualised rise in the US trade deficit of $160 billion.

Figure 2

Ian Shepherdson of High Frequency Economics did note that the April data confirmed a weak start to US trade in 2010: ‘The headline number is flattered by a rebound in the aircraft surplus and an unexpected dip in the oil deficit, which offset a $1.2B increase in the core deficit. Goods exports ex-oil and aircraft fell 2.0%, after a 4.8% jump in March. Stepping back from the noise, the post-Lehman rebound in trade does now seem to be slowing. This is disappointing, given that the strengthening of the dollar cannot yet be playing a part.’ But Shepherdson failed to to note the longer term deteriorating trend.

RDQ Economics missed the main point, arguing, ‘The small declines in imports and exports are nothing to worry about (yet) as the three-month trends in import and export growth remained robust.’ In reality the US trade balance is clearly deteriorating.

Zach Pandl, of Nomura Global Economics, had clearly not been following trends clearly, writing, ‘we had expected a significant improvement in the real trade balance this month’ – a view that could not be arrived at if the widening deficit which has been developing for ten months was being followed.