Comments on Paul Krugman and Alwyn Young on The Myth of Asia's Miracle - why 'quantity' may be more important than 'quality' in economics

Preparing for a panel discussion with Paul Krugman at Jiao Tong University in Shanghai led to reflection on how different the parameters of practical policy making are from those of academic economics. The questions asked and point of approach are frequently divergent

In policy making all theoretical and other arguments have to be aligned and concentrated around one settling one decisive issue - ‘what should be done’. That is, what is involved is a synthetic decision – assembling issues, and giving them their specific weights, around one point. In academic discussion exploration of distinctions and points can be pursued without settling the decisive practical question of what difference it makes to what should be done.

This particular reflection was reinforced by re-reading, to prepare the debate, Paul Krugman’s well known 1995 paper, ‘The Myth of Asia’s Miracle’. This analysis, arguments from which are still frequently used today, drew heavily on two quantitative papers by Alwyn Young on growth in the four Asian Tigers/Newly Industrialised Economies (NICs) of South Korea, Singapore, Taiwan and Hong Kong.[1]

In analysing the Asian Tiger economies Young/Krugman were attempting to deal with a theoretical/analytical issue. Was the rapid rate of growth of the South East Asian Tigers based on, or substantially contributed to, by a particularly high rate of growth of productivity – whether of labour, capital, or total factor productivity? Or to what degree was it based on quantitative growth of factors of production – i.e. accumulation of labour and capital?

It should be noted that the quantitative results of Young’s work has come under criticism - notably from Chang-Tai Hsieh. However, for the moment, leave statistical criticism aside and assume, for the sake of argument, that Young’s quantitative conclusions were correct – although, to be clear, this is done as a hypothesis and is not an acceptance of Young’s calculations per se. Then what follows?

Writing in 1995 Young noted that for the period 1960-85 the four Asian Tigers constituted four out of the five countries with the fastest growth of GDP per capita in the world - the fifth, Botswana, was an economy sufficiently small that no general conclusions would be drawn from it. However, after subtracting growth due to the increase in labour input (including increased participation in the workforce, higher educational achievement etc) and the rate of additions and improvements to capital, Young concluded that the growth of total factor productivity in the Asian Tiger economies was not remarkable. Summarising his article, Young wrote that he:

‘presents estimates of “total factor productivity” in the sample economies... the ranks of Taiwan and South Korea [among economies placed in descending order of growth of total factor productivity] are now reduced to 21st and 24th, respectively. While this remains a strong performance, it is no longer dramatically differentiated from that of the rest of the world economy. Fully 81 of the 118 sample economies lie within one standard deviation… of Taiwan and South Korea. Surprisingly, economies such as Bangladesh, Uganda, Iceland and Norway are now seen to have outperformed Korea and Taiwan, whose productivity growth is only 0.5% greater than that of a renowned laggard, the United Kingdom. Singapore, where participation and investment rates have risen faster than any of the NICs, is reduced to a rank of 63rd in the world economy.’

So, therefore, Young finds the growth of productivity in the NICs was average or slightly above average and their rapid growth was not primarily due to extraordinary growth in total factor productivity but was due to large scale quantitative inputs of capital and labour. To which the appropriate answer, from the point of view of economic growth, is: ‘yes, that is quite adequate, even very encouraging. For it shows that if it is possible to combine average productivity growth with very large quantitative inputs, then the economy’s rate of growth will be far higher than the average and very rapid in absolute terms – enough to industrialise a country in a single generation (which is what the NICs achieved).’

The point is a simple arithmetic one. The effectiveness of the contribution of investment, for example, to economic growth depends on the combination of its quantity and how efficiently the economy utilises it. This blog has noted on numerous occasions that there is a fundamental logical error in judging an economy’s growth potential by economic approaches which concentrate only on the efficiency of the use of investment rather than also analysing the quantity of investment. The quantitative relation of the relative scale of investment and the relative efficiency of investment is the critical one. If, for example, economy A utilises investment 20% more efficiently from the point of view of generating growth than economy B, but nevertheless economy B invests 50% more as a proportion of GDP, then economy B will grow more rapidly than A despite the fact that economy A uses its investment more efficiently.[2]

Young/Krugman demonstrate that the rate of productivity growth of the Asian Tiger economies is not below average, but only average, as a result of which these economies quantitative advantage in growth of inputs of investment and labour ensures much more rapid growth that economies with higher rates of total factor productivity growth but much lower rates of input growth.

This is why, for example, criticisms that countries such as South Korea, during their phases of rapid growth, allegedly allocated capital inefficiently compared to more ‘liberal’ economies such as Britain or the United States entirely miss the point. An economy such as South Korea invested so much more as a proportion of GDP, almost double the rate of the US, that unless, from the point of view of growth, its' efficiency of investment was only half that of the US the South Korean economy would still have grown more rapidly than the US.

Put in properly formulated economic terms the quantitative level of macro-economic allocation of resources to investment may be more important from the point of view of economic growth than the marginal efficiency of investment. Put crudely, when it comes to investment and growth, 'quantity; may simply be more important than 'quality'. That, for example, would by itself be enough to vindicate the present very high rates of investment in India and China.

Whether it has proved in practice a more viable growth strategy to have an average rate of growth of productivity, combined with very high quantitative inputs of investment and labour, or whether it is more effective to aim at the highest rate of growth of total factor productivity, with much smaller quantitative inputs of investment and labour, may be illustrated rather graphically by showing the rank order of countries produced by Young’s calculation.

Young found that the top five countries in terms of growth of total factor productivity, after he has carried out his adjustments, were as set out in Table 1.

Table 1

09 05 21 Young TFP Growth

In short, if highest possible growth in total factor productivity is the variable that should be targeted, then Egypt, Pakistan, Congo and Malta, together with Botswana, should be taken as the most successful economies in the world – the economic models to be emulated.

If, however, the key criteria of success is increase in GDP per capita, achieved, according to Young’s calculation, by the Asian Tiger economies combining average rate of growth in total factor productivity with massive quantitative inputs of investment and labour, then in contrast Table 2 shows the world ranking of economies.

Table 2

09 05 21 Young GDP Per Capita Growth

Which economic variable is in practice decisive in determining real economic outcomes may be shown graphically by taking the case of by far the worst performing case of total factor productivity according to Young/Krugman’s account – Singapore.

Singapore, poorly performing in terms of total factor productivity, has today, in Parity Purchasing Power terms, the 5th highest GDP per capita in the world – a level 9% higher than the United States. Egypt, which is better performing in terms of growth of total factor productivity, ranks 101st in the world with a GDP per capita only 13% that of the United States. While the second ranking, from the point of view of total factor productivity growth, Pakistan ranks 130th in the world with a GDP per capital 6% that of the US.

In short, taking for arguments sake Young and Krugman's calculations as entirely correct, then the route to actual economic success, in terms of economic growth and a high living standard, lay in the average rate of increase of total factor productivity, combined with massive quantitative inputs of capital and labour, of Singapore rather than in the high total factor productivity, combined with far lower quantitative growth of inputs, of Egypt, Congo and Pakistan. Or, put in deliberately shocking terms, 'quantity' (growth of factor inputs) was much more successful in determining growth in GDP per capita than 'quality' (growth in total factor productivity)!

It is, of course, possible to have a rate of growth of total factor productivity that is so low (potentially a negative number) that even the greatest increases in quantitative inputs cannot produce viable growth – the USSR in its final period represents such a case. But the case of the Asian Tiger economies showed that provided close to average increases in total factor productivity can be achieved then quantitative increases were the decisive ones. Put formally, the evidence is that provided an average, or near to
average, rate of total factor productivity growth can be achieved then
ensuring very large quantitative inputs proved a more viable growth
strategy than aiming to maximise efficiency – i.e. total factor
productivity growth.This is simply the arithmetical outcome of multiplying the rate of growth of factor productivity by the rate of growth of inputs. The criteria which must decide the strategy chosen is therefore that which maximises the rate of growth of GDP per capita, not the abstract theoretical one of maximising rate of growth of factor productivity.

Turning to India and China this has an immediate practical consequence. It means that even if it were to be assumed, for the sake of argument, that the efficiency of their use of investment were average, or even somewhat below average, then they might well be right to concentrate on massive inputs – to take the Singapore route. That, in turn, evidently raises the question of whether investment in India and China actually is inefficient – which goes beyond the scope of the present article, but will be returned to in a future article. But it should be noted from the above that even if, for the sake of argument, it were assumed that Krugman and Young’s quantitative premises are correct then this does not constitute a valid argument, from the point of view of the key variable of maximising the rate of growth of per capita GDP, against the effectiveness of the growth model followed by either the South East Asian Tiger economies or current policies pursued by India or China.

As stated at the beginning of this article, in economics quantity in some cases may simply be more important than quality.

* * *

This article by John Ross originally appeared on the blog Key Trends in Globalisation


[1] The argument of all three papers by young and Krugman was that the rapid growth of the NICs was based on quantitative accumulation of inputs of labour and capital and not on any productivity growth that was remarkable by international standards. The same analysis was then applied to China in Young’s 2003 paper ‘Gold into Base Metals: Productivity Growth in the People’ Republic of China during the Reform Period’.

[2] Ideally, of course, a combination of the maximum level of efficiency of investment from the point of view of economic growth and the maximum level of inputs would be achieved. However, while this is optimal in a purely abstract theoretical model in practice it may be necessary to chose between the two. An evident case of this is heavily state influenced financial systems aimed to maximise savings, as for example existed in Japan and South Korea during periods of rapid growth, versus those which are aimed to maximise the efficiency of use of savings. General discussion of this point, however, goes beyond the scope of this article.

No comments:

Post a Comment