The Ladder of Competitiveness HOW TO CLIMB IT
League tables of national competitiveness give an easily comparable ranking of the winners and losers of global economic competition. But they leave a number of questions unanswered. Why are the “poor” countries four times less productive than the “rich” ones?
Why are some rich countries twice as productive as others? How important is human capital compared with other factors such as financial capital or infrastructures? Using empirical data from over 50 countries, this book shows how even small differences in a number of factors combine to boost or block productivity. Governments need such information to set priorities. Investors need it too, and two new rankings are proposed as alternatives to a simple comparison of industrial productivity. The first, called the “investor ranking”, is based on infrastructure, human capital and total factor productivity. The second, “exporter ranking”, is for investors whose prime concern is for a production platform well-integrated into world trade. Combining the new rankings with a more traditional one produces three groups of countries, termed balanced, high potential, and vulnerable. Group membership reserves some surprises: you may be rich, but that doesn’t mean you’re not vulnerable.
Development Centre Studies
The Ladder of Competitiveness
HOW TO CLIMB IT
ISBN 92-64-02826-9
www.oecd.org
The full text of this book is available on line via these links:
http://www.sourceoecd.org/development/9264028269 http://www.sourceoecd.org/generaleconomics/9264028269 Those with access to all OECD books on line should use this link:
http://www.sourceoecd.org/9264028269
SourceOECD is the OECD’s online library of books, periodicals and statistical databases.
For more information about this award-winning service and free trials ask your librarian, or write to us at SourceOECD@oecd.org.
This work is published under the auspices of the OECD Development Centre. The Centre promotes comparative development analysis and policy dialogue, as described at:
www.oecd.org/dev
« Development Centre Studies
The Ladder of Competitiveness
HOW TO CLIMB IT
by Orsetta Causa and Daniel Cohen
The Ladder
of Competitiveness
HOW TO CLIMB IT
by
Orsetta Causa and Daniel Cohen
DEVELOPMENT CENTRE OF THE ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT
The OECD is a unique forum where the governments of 30 democracies work together to address the economic, social and environmental challenges of globalisation.
The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies.
The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of the European Communities takes part in the work of the OECD.
OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and research on economic, social and environmental issues, as well as the conventions, guidelines and standards agreed by its members.
Also available in French under the title:
L’échelle de la compétitivité COMMENT LA GRAVIR
© OECD 2006
No reproduction, copy, transmission or translation of this publication may be made without written permission.
Applications should be sent to OECD Publishing: rights@oecd.org or by fax (33 1) 45 24 99 30. Permission to photocopy a portion of this work should be addressed to the Centre français d'exploitation du droit de copie (CFC), 20 rue des Grands-Augustins, 75006 Paris, France, fax (33-1) 46 34 67 19, contact@cfcopies.com or (for US only) to Copyright Clearance Center (CCC), 222 Rosewood Drive Danvers, MA 01923, USA, fax (978) 646 8600, info@copyright.com.
THE DEVELOPMENT CENTRE
The Development Centre of the Organisation for Economic Co-operation and Development was established by decision of the OECD Council on 23rd October 1962 and comprises 21 member countries of the OECD: Austria, Belgium, the Czech Republic, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Korea, Luxembourg, Mexico, the Netherlands, Norway, Portugal, Slovak Republic, Spain, Sweden, Switzerland and Turkey, as well as Brazil since March 1994, Chile since November 1998, India since February 2001, Romania since October 2004, Thailand since March 2005 and South Africa since May 2006. The Commission of the European Communities also takes part in the Centre’s Governing Board.
The purpose of the Centre is to bring together the knowledge and experience available in member countries of both economic development and the formulation and execution of general economic policies; to adapt such knowledge and experience to the actual needs of countries or regions in the process of development and to put the results at the disposal of the countries by appropriate means.
The Centre is part of the “Development Cluster” at the OECD and enjoys scientific independence in the execution of its task. As part of the Cluster, together with the Centre for Co-operation with Non-Members, the Development Co-operation Directorate, and the Sahel and West Africa Club, the Development Centre can draw upon the experience and knowledge available in the OECD in the development field.
✒
THE OPINIONSEXPRESSEDANDARGUMENTSEMPLOYED INTHISPUBLICATIONARETHE SOLERESPONSIBILITYOFTHEAUTHORSANDDONOTNECESSARILYREFLECTTHOSEOFTHE OECD, ITS DEVELOPMENT CENTREOROFTHEGOVERNMENTSOFTHEIRMEMBERCOUNTRIES.Foreword
This study was written as part of the Development Centre’s work on
“Trade, Competitiveness and Adaptive Capacity”. It is intended to underpin work elsewhere in the OECD on competitiveness issues and to suggest policy options under the general umbrella of the Centre’s 2003/2004 work programme entitled, “Adaptive Capacity and Inclusive Development”.
Table of Contents
Acknowledgements ... 6
Preface ... 7
Executive Summary... 9
Introduction ... 13
Chapter 1 Manufacturing Productivity ... 17
Chapter 2 Explanatory Factors ... 25
Chapter 3 The Overall Productivity of the Economy ... 37
Chapter 4 Social Infrastructure ... 53
Chapter 5 International Trade ... 63
Chapter 6 New Approaches to Competitiveness... 77
Conclusion ... 83
Annex I Data: Definitions and Sources ... 85
Annex II The Econometrics of Productivity ... 95
Annex III Econometric Analysis of Relative Prices and the Lucas Paradox ... 103
Annex IV The Econometrics of Trade ... 111
Annex V Theoretical Framework ... 125
Bibliography ... 129
Acknowledgements
The Development Centre thanks the governments of Switzerland and the United Kingdom for their financial support.
The authors thank Isabelle Chort, Marcel Soto, Cornelius Schaub and Cécile Valadier for their invaluable assistance in the preparation of this report, as well as all of their colleagues at the OECD Development Centre for many useful comments at various stages in the writing.
Preface
Understanding the productivity of a nation is one of an economist’s most important tasks. It is also the subject of the book that founded modern economic science, Adam Smith’s The Wealth of Nations.
Industry is unique as an economic sector in that it is responsible for producing goods and equipment, thus favouring the accumulation of capital.
Internationally it is the most widely traded asset: industrial goods account for 80 per cent of world commerce.
It is clear that, for both these reasons, determining the factors involved in the industrial productivity of nations is important. Hence this study.
Infrastructures, education, transport costs and private investment are analysed to determine the principal causes of strengths and weaknesses of industrial productivity within a country. To do this the authors use extensive data from the OECD, UNIDO (the United Nations Industrial Development Organization), the World Bank and UNESCO.
This book allows a preliminary classification of the 51 countries under study. According to a first ranking of countries based on total productivity, Japan comes in first, while Bangladesh is last. The industrial productivity ratio of these two countries is 1:30! Countries can use this method of classification to measure themselves against others, but also (perhaps more importantly) to understand their own shortcomings.
The book presents two more rankings which will allow international investors to compare the same countries using different criteria. The second ranking is aimed at firms wanting to invest in the country, be it for internal trading or exportation. According to this, the United States comes first and India last. The third ranking is specifically destined for investors interested in exporting from a country: Sweden ranks first and Central African Republic last.
These rankings are not meant to encourage countries to try to outdo each other, but to help them understand the interdependent factors that shape industrial productivity and to compare themselves against others so as to identify their strong points and the areas where they need to concentrate their efforts.
We hope the book will thus contribute towards debate in each country on their future priorities, and help them to improve, in the words of Adam Smith, the “productive powers of labour”.
Louka T Katseli Director
OECD Development Centre September 2006
Executive Summary
The industrial productivity of poor countries is much lower than that of rich countries. According to the results presented herein, it is on average four times lower but, on either side of this average, there are considerable differences between countries.
Among the rich countries, productivity can vary by as much as a factor of two, for example between Japan and Australia. Similar differences can be observed among European countries: Greece, for instance, is twice as productive as Portugal. Among the countries described as “emerging”, Singapore is nearly twice as productive as Mexico, while Mexico is more than twice as productive as Ecuador, itself twice as productive as Egypt, Indonesia and India. The productivity of the last three countries is less than 10 per cent of that of the most productive group.
The productivity gaps between countries are governed by a law of
“multiplicative handicaps”. To study the causes of these handicaps, this volume considers five factors: three country-specific factors — physical capital, infrastructure and human capital — to which we will add a term measuring the country’s degree of integration in international trade and another measuring the (residual) net productivity of each economy.
Some countries will be found to be deficient mainly in infrastructure (particularly in Africa), others in human capital (notably in the Middle East), others in trade integration (often in Latin America) or in overall efficiency (India, Indonesia and Egypt).
The approach to industrial productivity presented in this volume also enables us to link up with the more general notion of competitiveness.
Indicators of international competitiveness of the kind published by the World Economic Forum regularly attract attention. The fact that France, for example, is ranked at such and such a level is never indifferent to the economic press and politicians.
Despite their defects, these indices are of clear interest to investors. As Lall (2001) points out, they contain useful information on the conditions facing international investors. On the basis of the analysis we have chosen for this report, one can understand that international investors are less interested in the level of industrial productivity than in some of its constituent determinants.
For example, investors may attach greater importance to data on infrastructure and education (what we call “social infrastructure”). The reason is that industrial productivity indicators include the contribution of capital, about which investors hardly need information. Taking this idea further, it is possible to develop other indices than those derived solely from cross-country comparison of productivity levels.
We propose here two rankings as alternatives to a simple comparison of industrial productivity, although the latter is retained as our first ranking.
Our second ranking is obtained by subtracting the contribution of capital from the calculation of industrial productivity. This choice is the direct result of what was said above, namely that international investors can decide for themselves what the appropriate level of capital is. We will call this ranking the “investor ranking”. In concrete terms, the second index is constructed as the product of three factors: infrastructure, human capital and total factor productivity (TFP).
Our third ranking is useful for an investor in need of a production platform that is well integrated into world trade. If the investor is capable of providing not only capital but also vocational training, and also has control of the productive efficiency of the company, the only terms which will be of interest to that investor are physical infrastructure and access to the world market. In Chapter 5, we will show that these two terms are in fact highly complementary. In practical terms, then, we will define this third index by calculating the product of the infrastructure and trade integration terms. We call it the “exporter ranking”.
The results are presented on the next page.
Rankings
Productivity Ranking Investor Ranking Exporter Ranking 1 Japan United States Sweden 2 United States Japan Norway 3 Finland Denmark United States 4 France Sweden Belgium 5 Sweden United Kingdom Canada 6 Belgium Canada France 7 Netherlands France Spain 8 Austria Finland Finland 9 Canada Norway Netherlands 10 Norway Austria United Kingdom 11 Denmark Netherlands Singapore 12 United Kingdom Belgium Italy
13 Korea Korea Austria 14 Singapore Australia Denmark
15 Italy Singapore Portugal 16 Australia Italy Korea
17 Spain Spain Hungary 18 Chile Chile Australia 19 Brazil Brazil Mexico 20 Greece Mexico South Africa
21 Mexico Greece Greece 22 Turkey Colombia Venezuela
23 Thailand Uruguay Malaysia 24 Peru Turkey Japan
25 Cyprus Cyprus Costa Rica 26 Colombia Peru Brazil
27 Uruguay Jordan Cyprus
28 Venezuela Venezuela Trinidad and Tobago 29 Portugal Panama Uruguay
30 Panama Portugal Morocco 31 South Africa South Africa Colombia
32 Malaysia Malaysia Egypt 33 Bolivia Thailand Fiji
34 Trinidad and Tobago Philippines Honduras 35 Zimbabwe Bolivia Panama 36 Senegal Zimbabwe Philippines 37 Ecuador Trinidad and Tobago Thailand 38 Zambia Zambia Jordan 39 Philippines Costa Rica Chile 40 Jordan Hungary India 41 Central African Republic Ecuador Zambia 42 Cameroon Morocco Peru 43 Morocco Fiji Ecuador 44 Fiji Cameroon Zimbabwe
Rankings (contd.)
Productivity Ranking Investor Ranking Exporter Ranking 45 Hungary Honduras Indonesia 46 Costa Rica Central African Republic Bolivia
47 Egypt Senegal Senegal
48 Indonesia Indonesia Cameroon 49 Honduras Egypt Turkey
50 India Bangladesh Bangladesh 51 Bangladesh India Central African Republic Note: See chapter 6 for the calculation of the new rankings.
Among the emerging countries, there are many changes in rank from one ranking to the next. The insights gained from these three rankings can be summarised by identifying three groups of countries:
Balanced countries, whose positions do not change more than seven places between ranking 1 and rankings 2 and 3. This group comprises the following 29 countries: United States, Finland, France, Sweden, Netherlands, Austria, Canada, Norway, Denmark, United Kingdom, Korea, Singapore, Italy, Australia, Brazil, Greece, Mexico, Cyprus, Colombia, Uruguay, Venezuela, Panama, Trinidad and Tobago, Ecuador, Zambia, Philippines, Cameroon, Indonesia and Bangladesh.
High-potential countries, whose positions in rankings 2 and 3 are at least seven places higher than in ranking 1. This group comprises the following 13 countries: Belgium, Spain, Portugal, South Africa, Malaysia, Jordan, Morocco, Fiji, Hungary, Costa Rica, Egypt, Honduras and India.
Vulnerable countries, which belong to neither of the preceding groups, and whose positions in rankings 2 and 3 are at least seven places lower than in ranking 1. These are the following nine countries: Japan, Chile, Turkey, Thailand, Peru, Bolivia, Zimbabwe, Senegal and the Central African Republic (CAR).
This typology encapsulates the main points of the analysis presented here. Most of the countries are balanced, in the sense that the three handicaps generally point in the same direction. Studying them in one or other way changes their rankings hardly at all. Some atypical countries stand out, however: on the one hand, high-potential countries, which could become more prosperous if they managed to convince international investors to believe in their capabilities; on the other, vulnerable countries, which are poorly integrated in world trade or face the risk of capital flight. For the latter group, despite productivity results which might seem satisfactory, work on “fundamentals”
is urgent if they wish to consolidate their place in the world rankings.
Introduction
A country’s industrial productivity is one of the key determinants of its prosperity. This point is generally acknowledged, although it has often given rise to misunderstandings in the past.
For an emerging country, industry is in the first place an essential lever for gaining access to the world market. Nearly 80 per cent of world trade today takes place in the industrial sector. As a result, an emerging country with low productivity will have greater difficulty in entering the world market. In economic parlance, an industrial good is essentially an internationally tradable good.
The second reason for the critical importance of industry is that it is the sector which produces capital goods. Where industrial productivity is low, the cost of manufacturing capital equipment increases, and the higher relative price for industrial goods discourages investment.
These two closely linked factors alone constitute sufficient grounds for emerging countries to attach great importance to increasing their industrial productivity. Other, much less convincing explanations have also been given for the primordial role played by the industrial sector. It is sometimes argued that, unlike other sectors, industry exerts spillover effects on other sectors (positive externalities) that justify its privileged status. For example, the notion of “industrialising industrialisation” was put forward in the 1960s as a rationale for the sometimes monumental projects undertaken in the industrial sector.
According to this argument, a country must industrialise in order to initiate a process of self-sustaining growth. This conclusion is unjustified: many agricultural countries, such as Australia and Denmark, have prospered even though they remained exporters of agricultural products. The related argument that agriculture, industry and services are obligatory stages in the development of an economy has not been proved either. The two-fold nature of tradable goods and capital goods justifies the attention given to industry, but not the extreme idea that the sector is the necessary path to growth.
This volume will seek to analyse the determinants of industrial productivity while keeping in mind the scale and limits of its importance.
This analysis will enable us in particular to understand what is meant by a country’s “competitiveness”, as it is understood by, among others, the analyses of the World Economic Forum, by providing us with specific points of comparison of countries’ performance.
The main findings of our study may be summarised as follows. The first fundamental result concerns the figures themselves. Produced from UNIDO data and checked against many other studies (World Bank, Groningen Growth and Development Centre, etc.), our figures show the industrial productivity of poor countries to be much lower than that of rich countries — four times lower on average. The differences observed between countries are much greater than those between continents, which offer a much less contrasting picture.
To study the causes of these handicaps, we consider five terms: three country-specific factors — physical capital, infrastructure and human capital — to which we will add a term measuring the country’s degree of integration in international trade and another measuring the (residual) net productivity of each economy. If the level of each of these terms in the rich countries is set by convention at one, the productivity deficit in poor countries can be explained as follows. The contribution of human capital is 0.64; that of the four other terms — physical capital, infrastructure, trade integration and net efficiency — an average of 0.8. However, 0.8 to the fourth power is 0.40, and when this is multiplied by 0.64 we obtain 0.26. Although each of these five handicaps may seem moderate in itself, when they are multiplied together a country is soon found to have very low productivity. Poor countries are subject to what may be called the tyranny of numbers (see Young, 1995). Their low productivity is due to the multiplicative interaction of five terms which, taken individually, are never very far from the levels reached by rich countries.
Beyond this general picture, the situations of individual countries often display specific bottlenecks that handicap the countries’ productivity. To understand the logic at work here, it may be useful to recall briefly the O-ring production theory put forward by Michael Kremer. Kremer (1993) borrows the term from a part of the space shuttle Challenger that failed to resist the high temperatures to which it was exposed and caused the shuttle’s explosion in 1986. Kremer’s idea is that the quality of a production chain as complex as that of the space shuttle depends on the quality of its weakest link. The best law firms are those which recruit the best secretaries as well as the best lawyers.
The quality of each part must be at the quality level of the whole. Applied to developing countries, this theory explains why the production factors in a
developing country are all of low, average or good quality according to whether the country is poor, emerging or rich.
As with the O-ring that caused the loss of the space shuttle, we will talk of bottlenecks when it appears that one part of the production chain is not up to the standard of the others. One of the aims of our analysis is to describe the nature of these bottlenecks country by country. In this case, the countries must set specific priorities to remove these blockages caused by a few weak sectors.
When considering two countries with identical productivity, analysis of the weak links of each will shed light on the specific handicaps of each. Some countries will be found to be deficient mainly in infrastructure (particularly in Africa), others in human capital (notably in the Middle East), others in trade integration (often in Latin America) or in overall efficiency (India, Indonesia and Egypt).
We also examine the link between industrial productivity and the overall productivity of GDP. Most emerging countries have lower industrial productivity. As we will demonstrate, this difference in productivity fully explains why the relative price of investment goods is nearly 75 per cent higher in poor countries than in rich countries, as well as why capital accumulation is lower in poor countries.
We subsequently make the connection between our approach to industrial productivity and the more general notion of competitiveness. Paul Krugman’s celebrated remark that, when poorly defined, a nation’s competitiveness is a meaningless concept, can be explained very simply. A nation is not in the same situation as a company. A company can be less competitive than another and be under threat of bankruptcy because, for example, its cost structure prevents it from keeping up with its direct competitors. A nation, in contrast, can always devalue its currency to restore its “competitiveness” in relation to its costs.
What counts is its productivity, which is to say the volume of goods produced by its inhabitants. From this point of view, the only ranking that counts, if any ranking is really required, is that of productivity.
From the standpoint of an outside investor, however, it is essential to know whether low industrial productivity in a given country is due to a particular bottleneck. If a country is less productive than another because it is short of capital, for example, the external investor may be able to make good the shortfall. If its lower productivity is due to lack of domestic or commercial infrastructure, the task is more difficult, although, as we demonstrate, investors can remedy the situation through private investment.
Our analysis will thus — and this is the main purpose of this study — help each country to evaluate the scale of the priorities to which it must respond.
Chapter 1
Manufacturing Productivity
Data
The data on which we base our analysis are derived from two sources:
the OECD’s STAN database and the Industrial Statistics Database of the United Nations Industrial Development Organisation (UNIDO). We have cleaned the data, eliminating aberrant values via a procedure described in Annex I which makes use of sectoral data supplied by UNIDO. In Annex I we also compare our data to other available sources. With only a few exceptions, this comparison does not alter the overall analysis proposed here.
All the data presented will refer to a reference group consisting of the sub-group of rich countries for which data are available in both the OECD database and the UNIDO database1, covering the years 1990, 1995 and 1999.
The reference countries thus selected are Canada, the United States, Japan, Austria, Belgium, Finland, France, Italy, the Netherlands, Spain, Sweden, the United Kingdom and Australia. We will take the average of these countries’
results and use it as a common denominator, the figures for each country being expressed as a ratio of the reference group.
We have calculated the data for each of the three years considered. In the great majority of cases, the results obtained (relative to the reference group) are quite stable from one period to another. As exchange rate variations can affect the result obtained in the short term, however, we have elected to present a three-year average, which provides a more reliable result than data for a particular year, even the most recent one (in this case, 1999).
Data by Country
Table 1.1 presents the ranking obtained from our data, country by country.
Countries are listed in descending order of productivity, beginning with Japan and ending with Bangladesh.
Table 1.1. Industrial Productivity In relation to the reference countries (y)
Reference 1.00 Japan 1.31 United States 1.19
Finland 1.12 France 1.11 Sweden (*) 1.10
Belgium 1.08 Netherlands 0.99 Austria 0.98 Canada 0.96 Norway 0.96 Denmark 0.92 United Kingdom 0.88
Republic of Korea 0.87
Singapore 0.84 Italy 0.79 Australia (*) 0.70
Spain 0.70
Chile 0.61 Brazil 0.60 Greece 0.50 Mexico 0.43 Turkey 0.42 Thailand 0.38 Peru 0.36 Cyprus 0.36 Colombia 0.33 Uruguay 0.33 Venezuela 0.27 Portugal 0.26 South Africa 0.26
Panama 0.24 Malaysia 0.22 Bolivia 0.21 Trinidad and Tobago 0.18
Zimbabwe 0.18
Table 1.1. (contd.)
Senegal (**) 0.17
Ecuador 0.17 Zambia 0.17 Philippines 0.17 Jordan 0.17 Central African Republic 0.15
Cameroon 0.14 Morocco 0.13 Fiji 0.12 Hungary 0.12 Costa Rica 0.12
Egypt 0.09 Indonesia 0.08 Honduras 0.08 India 0.06 Bangladesh 0.04 Note: Average for the 1990s.
(*) Estimate for 1995 (Australia) and 1999 (Sweden) for calculation of the reference average on the basis of UNIDO data.
Estimate based on the two other years available from UNIDO.
(**) Includes ISIC sector 311 (“food manufacturing”), despite the econometric results concerning capital/value added ratios (see Annex I), because the sector accounts for 80 per cent of manufacturing employment according to UNIDO data.
Source: INDSTAT 2001; INDSTAT 2003; STAN 2004 database; WDI 2003; OECD ITCS database; Cohen and Soto (2001).
Japan tops the ranking, followed by the United States, with industrial productivity respectively 31 and 19 per cent higher than the reference group.
At the other extreme is a group of five countries — Egypt, Honduras, Indonesia, India and Bangladesh — with productivity less than 10 per cent of the reference level.
Regional Averages
We have also produced regional averages. To do this, we grouped the countries as follows:
1) “Other European countries”: Cyprus, Denmark, Greece, Hungary and Portugal;
2) Southeast Asia and Pacific: Bangladesh, Fiji, India, Indonesia, Republic of Korea, Malaysia, Philippines, Singapore and Thailand;
3) Sub-Saharan Africa: Cameroon, Central African Republic, Senegal, South Africa, Zambia and Zimbabwe;
4) Latin America and Caribbean: Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Honduras, Mexico, Panama, Peru, Trinidad and Tobago, Uruguay and Venezuela.
5) Middle East and North Africa: Egypt, Jordan, Morocco and Turkey.
The results are as follows:
Table 1.2. Industrial Productivity – Regional Averages
Reference 1.00 Other European 0.43
Poor 0.26 Poor excluding sub-Saharan Africa 0.28
Sub-Saharan Africa 0.18 Southeast Asia and Pacific 0.31 Middle East and North Africa 0.20 Latin America and Caribbean 0.30 Source: UNIDO, INDSTAT 2001, INDSTAT 2003, 3-digit ISIC Rev. 2; STAN 2004
database; WDI 2003; OECD ITCS database; Cohen and Soto (2001).
We can see that on average the “poor” countries are nearly four times less productive than the reference countries. Sub-Saharan Africa and the Middle East/North Africa are nearly five times less productive, while Latin America and Asia about three times less productive. The figures suggest, too, that the differences between poor regions are much smaller than the differences between individual poor countries. The most spectacular differences are found within the regions.
Review of Existing Analyses
Without proposing here an exhaustive review of the plentiful literature on the subject, let us recall the main lines of analysis that have been put forward to explain the productivity differential between rich countries and poor countries.
At the risk of simplifying to the extreme, it can be said that there are two schools of thought, which, though not always in conflict, propose two different interpretations. The first, which is currently the prevailing view among macroeconomists, puts the emphasis on institutional differences between countries to explain their different performance. The second explains productivity differences more simply in terms of insufficient accumulation of factors of production. There has been incessant to-ing and fro-ing between these two theoretical camps; a brief summary of some of these shifts is therefore in order.
The most recent and influential articles of the neo-institutionalist school are Hall and Jones (1999) and Acemoglu et al. (2000). Hall and Jones’ article shows that the main differences between countries can be explained by an index comparing levels of rule of law and openness. The second article shows that these indices are themselves the legacy of the colonial history of the countries concerned. According to Acemoglu et al., two scenarios have emerged. In the first, the conquistadores exterminated the indigenous population, as was the case in North America, and literally transplanted political and judicial institutions from European countries. In the other, they suffered from the rigours of the climate or resistance from the native population and were obliged to give up their plans to establish themselves permanently in the colonised countries. They therefore preferred to exploit the indigenous population rather than build a new society, leaving behind an imperfect constitutional framework which the countries concerned were unable to patch up after their independence.
Recently, however, an incisive article by Glaeser et al. (2004) has challenged the empirical relevance of the “neo-institutionalist” theories.
Glaeser et al. show that the “institutional quality” indicators, which are supposed to explain economic growth, have in reality no predictive power in this respect. Institutional differences are measured via several indicators, often published in guides destined for international investors, which measure respect for rule of law and the extent of corruption and protectionism. Although their correlation with the country’s economic performance at a given date is very close, Glaeser et al. show that an indicator measured over decade t has no predictive power regarding the growth observed in decade t+1. The authors deduce from this that the rating agencies are easily misled by success. China, to take the most obvious example, is today very well rated, but in the event of an economic crash similar to the one that struck Asia in 1997, no-one would be surprised if the quality ratings of the country’s institutions suddenly fell. In Asia, denunciations of “crony capitalism” were heard immediately — after the crisis.
At the other end of the intellectual spectrum, a number of studies have stressed the intrinsic importance of the accumulation of production factors in explaining different levels of wealth. The first versions of endogenous growth theory, produced by Romer (1986) and Lucas (1988), had as their central intuition the idea that the accumulation of capital (physical for Romer, human for Lucas) generated externalities capable of drawing countries into the process of self-sustaining growth.
These studies as well, however, quickly came under critical fire. Following the work of Benhabib and Spiegel (1994), Bils and Klenow (1998) and Easterly and Levine (2001), econometric analysis of growth cast doubt on the idea that human or physical capital accumulation could have a lasting impact on a country’s growth. For these authors, the only thing that counts is technical progress and its propagation, which brings us back to the analysis that institutions are the primary cause of under-development.
The statistical methods used to challenge the importance of factor accumulation are not without their disadvantages, however. Where education is concerned, the quality of the data — which is very low if one wishes to assess the rate of increase in the number of years of education — plays an important role in the statistical analysis. Domenech and de la Fuente (2002) and Cohen and Soto (2001) have shown that, when better data are used, education does play a significant role in economic growth. Moreover, growth is very volatile from one decade to another, and it is not surprising that production factors, which are much less so, explain these variations with difficulty.
In contradiction to the theories of Lucas, there seem to be hardly any positive externalities that would generate a social return to education higher than its private return. In the same way, the initial idea of Romer (1986), according to which the accumulation of physical capital could generate positive externalities, was rapidly called into question, notably by Mankiw et al. (1992).
These authors proposed coming back to an enhanced Solow model in which physical capital and human capital play a role in determining income that is roughly identical to the share explained by their private return. Neither physical capital nor human capital generates externalities; taken together, however, their explanatory power is, according to Mankiw et al., substantial.
The error of the proponents of the “institutionalist” approach is to have concluded that any explanation of the inequalities between nations depends on technical progress, while proponents of the production factors approach committed that of thinking that the variability of technical progress from one
country to another could be ignored (Mankiw et al., for example, assume it to be identical in all countries). If, like Cohen and Soto (2002), we consider all three factors — physical capital, human capital and technical progress — the disparities between economies are much more easily explained. For most countries, if the analysis is confined to income per capita adjusted for purchasing power parity, the three terms play roughly identical roles in explaining differences of wealth.
This analysis forms the basis for the approach used here. To the three factors of production — physical capital, human capital and technical progress — we will add two additional terms: the role of infrastructure and an indicator measuring countries’ integration in world trade. In analysing industrial production more specifically, rather than just aggregate GDP, we will place the emphasis on one of the key terms determining countries’ positions in world trade. We will return in Chapter III to the question of how to link industrial productivity and per capita income.
Note
1. More precisely, the UNIDO Industrial Statistics Database, ISIC Rev. 2, 2001 edition (CD-ROM); UNIDO Industrial Statistics Database, ISIC Rev. 2, 2003 edition (CD- ROM) (see Annex I). To save space, we use the abbreviation INDSTAT.
Chapter 2
Explanatory Factors
The Model
We will assume that industrial output is the product of four terms:
physical capital (K), infrastructure (Z), human capital (H) and a term called total factor productivity (TFP).
We therefore can write:
(1)
Y
ijTFP
ij. K
ijDH
ijEZ
ijJwhere we assume that
D E J 1
, which amounts to assuming constant returns to scale: by doubling K, H and Z simultaneously, we can double industrial production (this hypothesis is not rejected by the data; see Annex II).We will also specify that TFP is partly explained by the countrys degree of integration in world trade, measured by an index T, as follows:
(2) TFPij = AijTiø
where the subscript i designates a country and the subscript j a sector. The index used to measure Ti is the so-called Grubel-Lloyd (1975) index, which calculates the intensity of intra-industry trade. This index, as we will show in Chapter 5, is the one which appears best correlated to various indices of a countrys trade integration. The index A measures economies residual productivity once the four other factors, K, Z, H and T, have been taken into account.
The sources for the data are the following. Physical capital is obtained from UNIDO data on investment flows, broken down by sector in manufacturing industry (see Annex I for the details concerning construction of the capital stock). The human capital data are taken from the education data bank of Cohen and Soto (2001). The data used to represent infrastructure come from Canning (1998) and the World Bank (World Development Indicators).
What they actually measure is electric power production. The reasons for this choice will be discussed in the next chapter. Here, suffice it to say that these are the data that best represent the various dimensions of the overall infrastructure problem.
The weightings given to the factors stem from an econometric analysis which is presented in an annex (see table A.II.3, column (7)). They are in fact very close to those generally used in publications in this field. These weightings are:
47 . 0
; 14 . 0
; 1
; 30 .
0 E J T
D
.In Annex I, we put our results in perspective by comparing them with some comparable recent scientific contributions.
Results by Country
The results are presented in Table 2.1. All figures are expressed as ratios of the corresponding levels in the reference countries. All their interactions are multiplicative.
¢
¢
¡
¢
¢
£
¡
¢
¢
¢
£
¢
¢
¢
¡
¢
¢
¡
¢
Main Results
On average, as we said earlier, poor countries are four times less productive than rich countries. This considerable difference is due to the multiplicative interaction of terms which are quite moderate when taken separately. The relative values of the five terms capital, infrastructure, human capital, trade integration and productive residual range from 65 per cent for human capital, the weakest link, to about 80 per cent for the others.
To simplify, we can say that the productivity of poor countries is the product of a principal term, human capital, which stands at about two-thirds of the levels reached in rich countries and which activates four levers, each of which has a return equivalent to 80 per cent of that of the rich countries. However, 0.8 to the fourth power is 0.40, and multiplying this by human capital of 0.64 gives productivity of only 0.26. Each lever brings a return only slightly lower than that obtained in the rich countries, but the force of multiplication the
tyranny of numbers in the words of Alvyn Young (1995) is such that the countries are ultimately much less productive than each of the levers individually would lead us to expect.
Regions
We will now analyse the results for each poor region.
The Southeast Asian and Pacific countries are the closest on average to the model with the levers which interact closely with each other in multiplicative fashion. As for the other poor regions, human capital is the weak link, at two- thirds the levels observed in the reference countries. It moves four levers that are roughly equal in level capital at 0.80, infrastructure at 0.79, trade at 0.88 and productive residual at 0.84. We should note that the trade integration index of the Southeast Asia and Pacific region is the highest among the poor countries.
The Latin American and Caribbean countries invite a similar diagnosis.
The level of human capital is identical at 0.67. The capital index of Latin America is slightly lower than that of the Southeast Asia and Pacific region, the infrastructure index slightly higher and trade integration slightly lower, contributing to a total factor productivity of 0.70, compared to 0.74 for Southeast Asia.
The Middle East and North Africa region is close to the two preceding cases, despite a lower productivity level. The human capital level is lower at 0.59 of the reference level and moves four levers that, though their levels are lower as well, are all at roughly the same level 0.73 for capital, 0.76 for infrastructure, 0.83 for trade integration and 0.76 for net productivity. More than for the two preceding zones, therefore, human capital is seen to be the regions weak link.
Sub-Saharan Africa is the sole truly atypical region. Its capital index is very close to that of the reference level. In fact, in relation to manufacturing output, Africa is the continent with the highest capital intensity. We will return to this point in more detail below. It should be noted that, preconceived notions notwithstanding, poor countries do not seem to suffer from a particularly marked capital deficit.
At this general level, we reach the same result as that obtained for aggregate GDP by Cohen and Soto (2001). Poor countries must face not a single challenge, such as institutions or trade, but a variety of challenges interact in multiplicative manner. None of these levers is strong enough by itself to have a significant effect on the overall situation. Countries must take action on all fronts if they are to succeed. The differences between continents appear slight in relation to this general lesson.
Analysis by Diminishing Productivity
We now analyse individual countries according to their productivity levels. We divide the countries into four sub-groups, defined such that productivity differences within each do not exceed a ratio of one to two.
Group 1: Rich Countries
The first group is that of the most productive countries, which include the reference countries. It ranges from Japan to Spain, the latter being half as productive as the former, and includes two emerging countries: Korea and Singapore. The data on Korea may be over-estimated (see Annex I), but those for Singapore seem much more reliable and are in keeping with other sources.
In the case of Singapore, TFP is excellent, and the weak point seems rather to be human capital. This picture disagrees with that of Young (1995), who found Singapore to have low TFP, but his calculations were made for GDP as a whole.
Korea does not seem to suffer from any notable human capital handicap, according to our statistics, but appears to be much less well endowed with infrastructure than Singapore.
Group 2: Emerging Countries
This group is defined as the one in which productivity is higher than half the level of Chile, which is the first country excluded from group 1. It comprises ten countries, with Chile at the top and Uruguay at the bottom, and its members are countries generally regarded as emerging, including Brazil, Mexico, Turkey and Thailand, as well as a European Union member, Greece.
The average breakdown of the groups productivity is as follows:
The industrial productivity of Group 2 is thus 43 per cent of that of the most productive countries. The weak link in the chain is human capital, which is 31 per cent lower than in the reference countries. Apart from this term, all the other factors are balanced. The residual term A is in fact at the same level as the reference countries. The message for these countries seems to be clear:
continue to invest in education and improve the other factors at the same time.
Within this group, certain atypical examples stand out. Chile has a very low trade integration index but a very high residual A. This suggests that the country is managing to compensate for the negative effects of poor trade integration in the industrial sector. We should note here the similarity with Australia, which is also poorly integrated as a result of the tyranny of distance. Australia, however, has not succeeded in compensating for its poor trade integration by higher productivity, making the performance of Chile all the more remarkable.
Brazil and Turkey have certain traits in common. They are both in a situation in which a low level of human capital seems to be offset by an excellent level of TFP. In the case of Brazil, the product of the human capital and TFP terms is 0.73, while for Turkey, the product of the two terms is 0.81. There is thus considerable room for progress, which could be obtained through better education.
£ ¢
£
¢
The comparison between Turkey and Greece is also interesting. Greece is more productive than Turkey, but in a more balanced way than that of its neighbour. The terms which contribute to its productivity are all equal to at least 0.78. Apart from education, Turkey suffers from a second handicap with respect to Greece, namely a very low infrastructure level. This explains why Turkey is less attractive to foreign investors than Greece: its social and physical infrastructure is less developed. We will come back to this point below.
¢
¢
Mexico and Thailand offer another interesting comparison. Their industrial productivity levels are close (0.43 for Mexico and 0.38 for Thailand), but the breakdown is very different. Thailands productivity is highly intensive in physical capital, while that of Mexico suffers from lower capital accumulation. It should be recalled that the Asian crisis was triggered by a high current account deficit in Thailand. There was talk of over-investment, which the present data confirm.
¡
¡
Group 3: Weak Countries
The third group ranges from Venezuela to Cameroon. The average breakdown is as follows:
This group is on average half as productive as the preceding one. Taking together the contribution of K, Z, H and T on the one hand and the net residual A on the other, the average productivity of the group can be expressed as the product of 0.34 x 0.58, where 0.34 measures the contribution of all the factors except A, which equals 0.58. Factor accumulation thus remains an essential priority for this group, even though the contribution of individual factors appears less marked than in the preceding group.
The situation within this semi-peripheral group displays strong contrasts.
It includes a European Union country, Portugal, which is in fact very representative of the group average. Portugal suffers from two bottlenecks, namely a low level of capital and low residual efficiency, despite its very good trade integration.
The group also includes African countries such as Senegal and Cameroon, which suffer from poor trade integration and a low level of infrastructure, for which they compensate in part by a high level of accumulation. It should be noted nevertheless that neither Cameroon nor Senegal nor the Central African Republic suffers, according to our data, from particularly low residual productivity, with levels of 0.67, 0.65 and 0.85 respectively. We can see also that South Africa, which has the same level of industrial productivity as Portugal, is much closer to this latter country than to the other two African countries in the group.
Group 4: Fragile Countries
The last group comprises all countries less productive than Morocco, which means they have productivity of less than 0.13. India, Egypt and Indonesia, the three heavyweights of the group, have productivity of less than 10 per cent of the levels reached by the reference countries. For this group, the average breakdown is as follows:
The group is less than half as productive as Group 3, and its residual productivity is much lower. Here too, however, it is the multiplication of handicaps which causes poverty. The productivity level of 0.09 is the product
of 0.23 x 0.39, the first term corresponding to the factors K, Z, H and T together and the last term corresponding to residual productivity. Among the factors of production, low capital accumulation appears to be the major obstacle.
The three types can therefore be presented in summary form as follows:
¢¢
¢
¢
One can see that even for the most fragile countries where efficiency is very low, handicaps linked to insufficient production factors remain the dominant constraint, representing on average a handicap twice as heavy as that of overall efficiency1.
Note
1. To summarise the contribution of production factors and residual efficiency, we have written x=kzht and Var (Log y)= var(Log x) + var(Log a) + 2cov(Log x, Log a).
We find:
¡
¢
¢ ¡
¢
This confirms the greater importance of production factors.
Chapter 3
The Overall Productivity of the Economy
GDP per worker
The preceding analysis dealt with productivity in the industrial sector alone. We now turn to the link between industrial productivity and aggregate productivity. As we will see, apart from its intrinsic interest, this comparison plays an essential role in the analysis of the formation of the capital equipment prices and hence of the accumulation of physical capital in the country under consideration.
Using the same methodology as that presented above but excluding the trade integration index, let us first examine income per capita in the countries in our sample on the basis of the four terms: physical capital, infrastructure, human capital and TFP.
In Table 3.1, we present the geographical averages before examining the data country by country.
Table 3.1. The Determinants of GDP per Worker
GDP Capital
(K)
Infrastructure (Z)
Human Capital (H)
TFP
Reference 1.00 1.00 1.00 1.00 1.00 Other European countries 0.69 0.91 0.95 0.84 0.95
Poor 0.32 0.65 0.88 0.64 0.87 Poor excluding sub-Saharan Africa 0.35 0.68 0.89 0.66 0.86
Sub-Saharan Africa 0.18 0.49 0.87 0.54 0.80 Southeast Asia and Pacific 0.35 0.69 0.86 0.66 0.90 Middle East and North Africa 0.26 0.59 0.88 0.48 1.03 Latin America and Caribbean 0.35 0.69 0.90 0.67 0.84 United States 1.27 0.99 1.02 1.12 1.12
Belgium 1.12 1.06 1.00 0.90 1.16 Italy 1.09 1.08 0.96 0.84 1.24 Netherlands 1.06 1.02 0.96 0.96 1.13 Canada 1.04 0.97 1.05 1.13 0.91 Norway 1.01 1.05 1.09 1.10 0.81 Australia 1.00 0.95 0.99 1.16 0.92 France 0.99 1.05 1.00 0.91 1.04 Austria 0.99 1.02 0.98 0.97 1.02 Finland 0.96 1.03 1.02 0.97 0.94 Sweden 0.91 0.95 1.05 1.04 0.89 Denmark 0.91 0.96 0.94 1.04 0.98 Spain 0.90 1.04 0.98 0.78 1.13 United Kingdom 0.87 0.85 0.95 1.13 0.96
Singapore (*) 0.87 1.18 0.99 0.66 1.12
Japan 0.79 0.98 0.95 1.08 0.78 Greece 0.71 0.93 0.97 0.81 0.98 Portugal 0.67 0.97 0.97 0.61 1.17 Republic of Korea 0.65 0.86 0.94 1.02 0.79
South Africa 0.54 0.76 1.03 0.58 1.19 Trinidad and Tobago 0.52 0.73 0.94 0.82 0.93
Mexico 0.49 0.81 0.92 0.67 0.98 Hungary 0.48 0.78 0.93 0.91 0.73 Venezuela 0.47 0.88 0.99 0.56 0.96 Malaysia 0.47 0.77 0.91 0.75 0.88 Chile 0.45 0.71 0.91 0.82 0.85 Uruguay 0.43 0.69 0.93 0.71 0.94 Brazil 0.40 0.79 0.92 0.63 0.86 Costa Rica 0.33 0.70 0.91 0.59 0.88