The Wealth Of Nations

THE WEALTH OF NATIONS A collection of papers on the theory, empirics and policy implications December 2004 Tilburg University

With contributions from Ament, J. Bresser, H. de Broersma, J. Chen, Y. Erp, P.F. van Gevel, B.J.C. van de Gielissen, R. Gille, P. Gils, B.W.H. van Glowacki, Jakob F Heijboer, M.R. Hekken, J.C. van Hendrix, H.J.J. Jonker, M.F. Kerstholt, W.J.N. Ketelaars, M.E.M.A. Kitzen, S.W.C.M. Kok, R. Moorsel, P van Mulken, M.A.H. van Oudeman, J.R. Oudheusden, P. van Pei, N. Peters, N.K.A. Rooijackers, M.M. Smulders, S. (editor) Spellen, G. Urlings, N.M.A. Verkooijen, L. Verschuren, L.B.C. Zhou, B.

Reading Guide The following text introduces the main concepts, sketches the connections among the different papers and summarizes the main arguments of the papers.

I. Why are some countries so rich and some so poor? Introduction Income differences between countries are huge. For example, in 1988 per capita income in The Netherlands was only 80 per cent of that in the USA. Even more extreme, as a percentage of USA per capita GDP, this figure equals 32 % in Brazil, 19% in Egypt, 6% in India and 3% in Central African Republic. (See G15 for further details) Growth rates differ widely among countries. For example, over the period 1960-2000, Singapore grew at an average annual rate of 7%, the USA at a rate of 2.8%, the Central African Republic at a rate of –2%. Economists have tried to explain this kind of differences at various levels of abstraction. Table I.1 gives an overview of the types of explanation. Table I.1 Sources of growth and income differences - Classification Proximate sources Production factors (quantity, quality, composition, utilization) Physical capital Labor, human capital Natural resource inputs Productivity Efficiency Technology Ultimate sources (“fundamentals”) Geography Policy Institutions Some countries are richer than other countries, because they have more capital, a better educated labor force, they work longer, or they have more natural resources. Economists call these elements production factors. Countries that accumulate more production factors per capita can be expected to have higher income per capita. A second source of income differences is productivity: access to more advanced technologies, more efficient use of technology or more efficient use of production factors in general leads to higher production levels. These two classes of sources behind income differences, viz production factors and productivity, are the so-called “proximate causes”. They follow directly from a production function approach to national income in which GDP is a function of a nation’s production factors and the productivity (total factor productivity) of these inputs. Ultimate sources of growth or income differences explain why productivity is high or why a country has many production factors. Thus, ultimate sources are the fundamentals behind the proximate sources. The proximate sources we are going to

discuss are geography (related to geographical and ecological characteristics of the country), institutions, and policies. 1. Proximate sources of growth and income differences To study the proximate sources, economists perform growth or level accounting exercises. Furthermore, a number of important economic growth models have been developed to link factor accumulation and productivity to economic growth. Empirical studies: growth and level accounting. In G16, Europe is compared to the USA. First it is shown that Europe’s per capita income diverged from the USA’s per capita income from 1820 to 1950, then caught up until 1990. From then on, the income gap remains more or less constant. Next, per capita income differences are decomposed into labor productivity, hours worked, and participation. Systematically, the Europeans work less than the Americans but labor productivity (GDP per hour) is not always lower. G16 also studies the differences between R&D investments in Europe and the USA. The differences might explain the differences in total factor productivity growth. In G15, developed countries (DCs) are compared to less developed countries (LDCs). In the level accounting exercise it turns out that poor countries not only have low per capita capital, but also low levels of human capital and low levels of total factor productivity. Hence, the proximate sources are positively correlated. In the growth accounting exercise, it turns out that the growth rates of total factor productivity (TFP) do not differ systematically across DCs and LDCs. Combined with the existence of large TFP level differences, this implies that these differences are persistent over time. Models of factor accumulation, productivity, and growth The Solow model explains how capital accumulation and exogenous technological change might lead to long-run per capita growth and how income levels of rich and poor countries might converge. G11 discuss the Solow model, as well as the extended Solow model (the Makiw/Romer/Weil model). Notice that both models take total factor productivity as an exogenous variable. In contrast, in the models by Romer and Jones, TFP follows from investment in R&D. G11 investigates how realistic this assumption is. Indeed, empirical studies find a strong and significant link between R&D and output. Other models have stressed the fact that growth develops in stages. This is elaborated in G17. In early periods of history, many economies have experienced growth in GDP, but not in per capita GDP. Higher GDP growth was accompanied with higher population growth. This is called the Malthusian trap. A challenges for economists and historians has been to explain why for example the UK and France escaped the Malthusion trap in the 19th century. At that time, per capita income started to rise and population growth to fall. There was a transition from growth a la Malthus tot growth a la Solow. For development economists the question has been what determines when a country takes off in terms of starting to invest sufficiently to increase economic growth. The Solow model models an economy that invests at a given rate. However, in practice, long periods of low investment precede the stage in which investment is high. This is Rostov’s theory of growth in stages.

2. Ultimate sources of growth and income differences What determines high rates of factor accumulation or high rates of productivity? This is the question of ultimate causes behind growth. G14 distinguishes between geography, institutions and policy as fundamental determinants of growth. Empirical research is presented in order to sort out which of these three factors is most important. It turns out that the role of institutions is so important, that not much is left for policy or geographical conditions to affect growth or income in a way that is independent of institutions. In contrast, geographical conditions may be an important determinant of institutions, and thus only indirectly affect growth through institutions. For example, in colonies where climatic conditions caused high mortality rates under European settlers, extractive institutions were introduced and property right were not established. These countries are left with bad economic institutions until now, which results in low income levels. 3. Endogenous institutions Economic institutions, like the protection of property rights, not only have a major impact on economic growth and income levels, they are also likely to be shaped by economic conditions. Thus, institutions are endogenous. Modern political economy models study the consequences of differences in institutions on economic growth. [At the time of Adam Smith, the term political economy had a different meaning: it comprised the study of national income and distribution in general]. In particular, a stream of economics literature called the “New Institutional Economics” devotes special attention to the role of institutions and the endogenous nature of institutions. G13 explains how Nobel prize winner Douglas North has contributed to this literature. In the theory of Acemoglu, Johnson and Robinson (2004), the interaction between political institutions (the institutions that determine de facto and de jure political power) and economic institutions determines income distribution and the evolution of institutions over time. In G12, this theory is used to explain and compare economic and institutional developments in UK, France, USA and Argentina. In G13, European history is reviewed. It is shown how changes in economic conditions led to changes in institutions and the other way around.

I. 1 (G15) Proximate sources Noortje Urlings (statistician) Lona Verkooijen (manager) Marielle Ketelaars (analyst Growth accounting) Hilde de Bresser (analyst Level accounting)

GROWTH AND LEVEL ACCOUNTING: DC’S VERSUS LDC’S Introduction There are various ways to compare income across countries, such as growth accounting and level accounting. Growth accounting is a set of theories used in economics to explain economic growth. The total national income in an economy may be modelled as being explained by various factors. In a simple model, these might be: • • •

the total stock of capital available, K the size of the labor force, L the total factor productivity, A

The most used equation to measure growth of national income is as follows:

Yˆ = Aˆ + αKˆ + (1 − α ) Lˆ

An increase in national income can be explained by an increase in the capital available, an increase in the labor force, or an improvement in the technology used. In level accounting studies the basic objective is to divide the differences in income levels to differences in levels of total factor productivity and factor inputs. This method focusses only on the levels of economic variables. It will provide a better view on the difference of the values of income between the less developed and the developed countries. In this section we will deal with the question: Why do less developed countries perform less than developed countries? We also make a comparison of the growth rates and the levels of economic variables between LDC’s and DC’s.

Level accounting In this section we will try to get a better view on the difference of the values of income between the less developed and the developed countries. Figure 1 shows productivity levels across countries plotted against output per worker. The figure illustrates that differences in productivity are very similar to differences in output per worker. This means that countries with a low level of productivity also have a low output per worker and countries with a high level of productivity also have a high output per worker. The productivity levels of the countries are ratios to U.S. values. The relationship in Figure 1 is positive and significant.

Figure 1. Productivity and output per worker

Source: Hall and Jones, Why do some countries produce so much more output per worker than others?

The countries with the highest levels of productivity are Italy, France, Hong Kong, Spain and Luxembourg. The position of Puerto Rico is remarkable, but in short it is a result of an overstatement of real output.1 The figure shows that most OECD countries are on the right side; this means that these countries have higher productivity and higher levels of output per worker. There are also a couple of LDC’s on the right side, such as Saudi Arabia and Jordan, but these countries belong to the OPEC. This is an explanation for their high positions in the figure. Those countries with the lowest levels of productivity are Zambia, Comoros, Burkina Faso, Malawi and China. These countries are less developed countries and therefore they can be found on the left side of the figure. The difference in output per worker between developed and less developed countries might be partially explained by the difference in productivity levels. Comparison of OECD with LDC
Differences in output per worker The measure we use for calculating economic performance is the level of output per worker. In this section we make a comparison of the determinants of output per worker between OECD’s and LDC’s. We have chosen five developed countries and twelve developing

1

Further explanation, see Baumol and Wolff, 1996.

countries to show the differences in the level of output more clearly. To make the comparison easier, all terms are expressed as ratios to U.S. values. Table 1 Productivity calculations: Ratios to U.S. values Y/L (K/Y)a/(1-a) United States 1,000 1,000 Netherlands 0,806 1,060 France 0,818 1,091 Italy 0,834 1,063 UK 0,727 0,891

H/L 1,000 0,803 0,666 0,650 0,808

A 1,000 0,946 1,126 1,207 1,011

Brazil Colombia Peru Venezuela

0,319 0,264 0,237 0,495

0,873 0,818 0,935 0,994

0,482 0,544 0,618 0,593

0,758 0,593 0,409 0,839

China India Iran Malaysia

0,060 0,086 0,295 0,267

0,891 0,709 0,981 1,004

0,632 0,454 0,467 0,592

0,106 0,267 0,642 0,450

Central Africa R. 0,033 0,582 0,339 0,159 Egypt 0,187 0,454 0,576 0,716 Madagascar 0,041 0,299 0,514 0,264 South Africa 0,250 0,959 0,568 0,460 Source: Hall and Jones, Why do some countries produce so much more output per worker than others?

The determinants of output per worker are physical capital intensity ((K/Y)^a/(1-a)), human capital per worker (H/L) and total factor productivity (A). The total factor productivity is calculated as a residual. For example, this table shows that the output per worker of the Netherlands is about 81 percent of that in the United States. The Netherlands have about the same amount of capital per worker, but the human capital per worker is only 80 percent of that in the United States. The productivity is also somewhat lower in the Netherlands compared to the United States. The difference between the level of output per worker in the Netherlands and the United States is primarily explained by the difference in inputs and partly by the level of productivity. The output per worker in the developed countries (France, Italy, Netherlands, UK) is lower than in the United States, because of the differences in human capital per worker. In the UK differences in capital intensity also play a determinant role. For some developing countries in the table (Peru, China, India, Malaysia, Central Africa and Madagascar) differences in productivity are the most important factor in explaining differences in output per worker. The differences in output per worker for several other developing countries (Brazil, Venezuela, Egypt, Iran) are primarily explained by differences in human capital per worker and partly by differences in productivity. For the remaining developing countries (Colombia and South Africa) the differences in output per worker are explained by both human capital per worker and productivity. By comparing the levels of output per worker between the developed and the developing countries, we can conclude that developing countries have very low levels of output per worker relative to the developed countries. In most developing countries the determinants of output per worker are much lower than in the developed countries, which explains their lower performance. The main cause of the differences in output per worker is the difference in total factor productivity. In most of the developing countries the capital intensity doesn’t differ much from the capital intensity in developed countries, except for Central Africa, Egypt and

Madagascar. There are also differences in the level of human capital per worker between the developing and developed countries, but these differences are less important than the differences in productivity.2 Explanation for differences in output per worker between LDCs and developed nations There are several reasons why the levels of the determinants of output per worker are lower in the less developed countries than in the developed countries. First of all there is a lack of physical capital and a low savings rate. They also have little human capital because the labor force in LDC’s is often illiterate and lack necessary training. The technology that is used in LDC’s is mostly old fashioned. Developed countries use more modern and more efficient technology. But in the LDC’s there is no money for modern technology and the labor force doesn’t have the skills to work with this kind of technology. It is commonly known that technology and human capital are complements. The simple reason for the limited physical and human capital and low technology level is that there is not enough money. The reason therefore is a low savings rate. Thus, there is a vicious circle that is very difficult to get out off. Another important reason why less developed countries perform less than developed countries is the political instability and the government policies that discourage production and investment in less developed countries. Less developed countries are sometimes plagued by political insurgence and political corruption which create the wrong environment for investment and production. All these problems are linked with each other and may require assistance from international organizations.3

Growth Accounting Growth accounting is a way to explain economic growth. The growth of national income is determined by several factors, such as growth of capital, growth of labor and growth of total factor productivity. We can derive the widely used growth accounting equation from the level accounting equation as follows: Y = AKα L 1-α Ln Y = Ln A + α Ln K + (1- α) Ln L dLn Y = dLn A + α dLn K + (1- α) dLn L

in levels in logarithms in growth rates

The term dLn Y can be written shorter, namely Yˆ . The whole growth accounting equation can be written as:

Yˆ = Aˆ + αKˆ + (1 − α ) Lˆ Using growth accounting one can measure what explains the growth of national income. It can be explained by the growth of input factors, capital and labor, or by the growth of total factor productivity, also known as the residual. A change in the total factor productivity represents the change in national income that is not explained by changes in the level of inputs (capital and labor) used. This is normally taken as a measure of the level of technology employed. To detect the growth of the residual, growth accounting has to be done. The levels of national income, capital and labor are known for several years, so one can determine the growth of these variables. If these growth rates are known, one can determine the growth rate of the total factor productivity by rewriting the growth accounting equation: Hall 3

and Jones, Why do some countries produce so much more output per worker than others?, 1999

http://www.globalchange.umich.edu/globalchange2/current/lectures/dev_pov/dev_pov.html, University of Michigan, 2001

Aˆ = Yˆ - α Kˆ - (1- α) Lˆ To calculate the growth of the total factor productivity we have to substract the growth of input factors from the growth of national income. If the growth of national income is explained by the growth of input factors, than we say that national income grows because of factor accumulation. If national income grows because of factor accumulation we speak of extensive growth. This means the growth is not caused by an improvement in technology, but by putting more input factors in the production. If national income grows because of an increase in total factor productivity we speak of intensive growth. This means there has been an improvement in technology and so production becomes more efficient. This means that we can produce more for the same costs or we can produce the same amount against a lower price. Growth accounting in practice To get a better look at growth accounting we take a sample of 16 countries, four OECD countries and twelve LDC’s. We apply growth accounting on those countries. This section explains the method used. The OECD countries are the Netherlands, France, UK and Italy. The LDC’s are: - India, Iran, China and Malaisia from the Asian continent - Brazil, Colombia, Venezuela and Peru from the South American continent - Madagascar, Central Africa, South Africa and Egypt from the African continent In the sample we took a period of 40 years, from 1960 to 2000 and a period of 18 years, from 1980 to 1998. For each country we calculated the growth of national income, the growth of the population, the growth of capital, the growth of labor and the growth of the total factor productivity for the period of 1960 to 2000 and the period 1980 to 1998. To calculate the growth of capital, we had to determine the growth of the investment in those years. Subsequently we calculated the naive estimate, the estimate with correction and the final growth of the total factor productivity growth. The naive estimate of the total factor productivity is based only on the observation of the growth of national income per capita. It is the labor share times the growth of national income per capita. The naive estimation with correction is based on the growth of national income per worker. It is the labor share times the national income per worker. The final estimate is also corrected for the capital intensity, it presents the percentage of the growth of total factor productivity. Results of growth accounting in practice for the period 1960-2000 This section shows the results of the growth accounting for the 16 countries of the sample. In the end of this section we can say which countries have grown by factor accumulation or by technological progress. The next tables show the estimates of the growth of total factor productivity of each country from the sample in the period 1960-2000, the numbers are in percentages. Table 2.1. Results of Europe

Europe: Naïve estimate corrected tfp Netherlands 1,619 1,171 France 1,676 1,587 UK 1,365 1,244 Italy 1,789 1,816 Source: Penn world data

0,834 0,496 0,402 1,650

Table 2.2. Results of Asia

Asia: Naïve estimate corrected India 1,755 Iran 1,342 China 3,027 Malaysia 2,510 Source: Penn world data

tfp 1,887 1,268 2,775 2,437

0,549 0,571 1,100 0,526

Table 2.3. Results of South America

South America: Naïve estimate corrected Brazil 1,688 Colombia 1,211 Venezuela -0,535 Peru 0,506 Source: Penn world data

tfp 1,440 0,501 -0,821 -0,093

0,919 -0,798 -2,497 2,031

Table 2.4. Results of Africa

Africa: Naïve estimate corrected tfp Madagascar -0,637 -0,386 Central Africa -1,426 -1,020 South Africa 0,586 0,627 Egypt 1,676 1,588 Source: Penn world data

-0,719 -2,112 0,082 0,198

Our data shows that the growth of national income of all 16 countries is only by a very small part determined by growth of total factor productivity. Put another way, our data suggests that no country in the sample has experienced intensive growth. What we have to take into account is that the growth of TFP is calculated over a period of 40 years. Every economy fluctuates, sometimes there is a high growth of TFP and sometimes there is a low growth of TFP. So the growth rates of TFP in our results can be seen as averages over a period of 40 years. This can explain why the growth rates aren’t that high. If we compare the TFP growth of the different continents then we can see that the African and South American countries clearly have a lower TFP growth over this 40 year period. The Asian countries have quite similar results as the European countries. So these particular Asian countries can’t be qualified as LDC’s in TFP growth terms. Growth accounting in practice for the period 1980-1998 This section shows the results of the growth accounting for the 16 countries of the sample. In the end of this section we can say which countries have grown by factor accumulation or by technological progress. The next tables show the estimates of the growth of total factor productivity of each country from the sample in the period 1980-1998, the numbers are in percentages. Table 3.1. Results of Europe 1980-1998 Naïve estimate corrected TFP Netherlands 1,289 0,587 France 0,995 0,891 UK 1,469 1,344 Italy 1,235 1,071 Source: Penn world data

-0,061 -0,594 0,578 0,511

Table 3.2. Results of Asia 1980-1998 Naïve estimate corrected TFP India 2,519 2,428 Iran 1,179 0,978 China 4,149 3,791 Malaysia 2,444 2,760 Source: Penn world data Table 3.3. Results of South America 1980-1998 Naïve estimate corrected TFP Brazil 0,199 0,148 Colombia 0,963 -0,511 Venezuela -0,553 -0,824 Peru -0,275 -1,599 Source: Penn world data Table 3.4. Results of Africa 1980-1998 Naïve estimate corrected TFP Madagascar -1,024 -0,723 Central Africa -2,200 -1,853 South Africa -0,247 -0,354 Egypt 1,804 1,460 Source: Penn world data

0,782 0,896 0,535 0,132

-0,663 -2,465 -2,135 -0,376

-1,879 -3,525 -0,406 0,373

The data above shows that the growth of national income of all 16 countries is only by a very small part determined by growth of total factor productivity. The growth rates of TFP in the 16 countries in a particular period can be seen as an average over that period. The smaller the period, the more accurate the average will be. So a time period of 18 years says more than a time period of 40 years. A lot of countries experienced a negative TFP growth during 19801998, especially the African and South American countries. These LDC’s clearly have lower TFP growth than Asia and Europe during this period, except for Egypt. What also can be said is that the Asian TFP growth has been stable and reasonably well over the 40 years. The TFP growth of the other countries have been fluctuating much more over time.

Conclusion By comparing the levels of output per worker between the developed and the less developed countries, we can conclude that developing countries have very low levels of output per worker relative to the developed countries. In most developing countries the determinants of output per worker are much lower than in the developed countries, which explains their lower performance. The main cause of the differences in output per worker is the difference in total factor productivity. We took a sample of four OECD countries and twelve less developed countries over two time periods: 1960-2000 and 1980-1998. One would expect that the OECD countries would have intensive growth, but this isn’t the case. The four OECD countries do have positive TFP growth during the period 1960-2000, but the Netherlands and France had negative TFP growth during 1980-1998. The four Asian countries experienced similar TFP growth rates to the four European countries. Therefore, these four Asian countries cannot be seen as less developed countries, in our view. The countries in the other two continents (Africa and South America) clearly experienced lower or negative TFP growth in both time periods. So these countries can be defined as less developed countries. The growth rates of TFP clearly show

similar rates for the European and Asian continent, but the levels of TFP differ significantly. The TFP levels of the European continent are higher than those of the Asian continent.

References Hall and Jones, Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics, 1999 University of Michigan: Global Change, 2001 http://www.globalchange.umich.edu/globalchange2/current/lectures/dev_pov/dev_pov.html Penn world Data: http:// www.bized.ac.uk/dataserv/penndata/pennhome.htm

I. 2 (G16) Proximate sources

Jeffrey Oudeman (statistican) Marcel Mulken (analist) Jasper van Hekken (analist) Paul Gille (manager)

GROWTH AND LEVEL ACCOUNTING: USA VERSUS EUROPE

Introduction This paper is about the gap in income and growth between the US and Europe. When you look at the economy, America has been the world leader for over a century now. Theory suggest that Europe or at least Western-Europe should be catching up with the US. But when we look at GDP per capita ( see figure) we see that this is only threequarters of that in the United States.

In this paper were trying to find reasons for the GDP difference between US and Europe. While doing this we look at two subjects which we find very important and in which we hope to find some answers. First we take a look at the labour market of both countries and try to find differences that could have an effect on growth. Second we take a look at the research and development sector which has become more and more important the past few decades. R&D can improve the productivity of countries and in that way can positively affect the welfare of a country.

Labour One of the proximate sources of growth is labour. So when we want to know why the US is richer than the EU we have to look at differences with respect to labour between the US and Europe. The income gap between Europe and the US is large. Table 1 shows the income gap with the US for several European countries. The US is in terms of GDP per capita richer then any country in the EU. GDP per hour, which is a way of noting productivity, is the same or higher (relative to the US) in Northern and Western European countries (except for Finland and the UK) and (sometimes much) lower in Southern European countries. From table 1 can also be seen that US workers work a lot more hours than European workers. This is not true for Spain, Portugal and Greece, where people also work a lot. But there’s more. The employment rate in the US is higher than in the EU, so more people are working in the US. Figure 1: The income gap with the US explained

Source: De Groot, Nahuis and Tang, Is the American Model Miss World? Choosing between the Anglo-Saxon model and a European-style alternative, CPB So the EU is less productive, people in the EU work less hours and the employment rate is also lower in the EU. Of course there are differences within the EU. Some countries are more productive than the US (take a look at Norway), but others are far less productive (Portugal!). The low productivity in Southern European countries is the main reason why GDP per capita is very low in this countries. Figure 2 on the next page shows the facts about labour in another way. This graph shows that European countries are doing a lot worse then the US is doing. Again it is clear that in some countries productivity is low and the figure points out that participation in Europe is very low compared to the US and people work less hours.

Figure 2: Income gap of various European countries explained.

Source: CPB, Four Futures of Europe. So how do we close the income gap? Hours The main reason why the richest countries are behind the US is the number of hours. By increasing the number of hours income per capita will increase. But will working more hours also increase welfare? Working more hours automatically means enjoying less leisure. This social cost should also be taken into account. Participation We have already seen that participation in Europe is very low and unemployment is rather high in most European countries. Many European countries have large pools of hidden unemployment in, for instance, the disability and sickness schemes. In Europe social protection is at a high level. The incentives for people to work and to keep working are much higher in the US. This is proved by the fact that the participation of people above 55 years old is very low in European countries (see figure 3 on the next page) compared to the US. So to close the income gap, schemes for early retirement should be less generous. In fact the European population is becoming older than the US population. While the population in Europe in the eighties and nineties remained stable, immigration and higher fertility has caused the population in the US to grow. This demographic change makes it necessary that people in Europe will work longer in the future or the ageing problem will have serious consequences for our economic growth potential.

Figure 3: Employment of people aged 55-64

Source: Speech minister Hoogervorst, Europe: the challenges ahead.

Productivity Productivity is a problem in some southern European countries as we have already seen, while in other European countries labor productivity is very high. The gap in skills and technology is small for these countries. This does not apply for the countries like Greece, Portugal and Spain and for the Eastern European countries that have recently joined the union. You can see that there is a huge difference between the productivity. Labour market institutions An important problem for Europe is the fact that the European labour market is not as flexible as the US labour market is. The European welfare state has limitations, especially when it comes to adjusting to economic shocks. The US has a less developed welfare-state. Unemployment benefits are lower and the average time when the benefit is received is shorter. So by making the labour market more competitive like it is in the US and raising the incentives to work, the EU could catch up with the US when we look at growth. Off course we have to notice that inequality in the US is much larger. There seems to be a trade-off between participation and income-equality. We can see this relation in figure 4 on the next page, but looking at this data one could say this isn’t a perfect linear relation. There might be different lessons to be learned from the different social models.

Figure 4: Equality vs Participation

Source: De Groot, Nahuis and Tang, Is the American Model Miss World? Choosing between the Anglo-Saxon model and a European-style alternative, CPB

Conclusion GDP per capita in the US is higher than it is in Europe. Productivity is low in some European countries, which of course is a reason for this income gap. But there are also countries in the EU that already have high levels of productivity. Another reason why GDP per capita in whole the EU is relatively low, is that participation in the EU is very low and people work fewer hours. GDP per capita will increase a lot when participation rises and workers work more. By removing these differences with the US, most of the income-gap disappears. But we have to keep in mind that inequality might rise due to the actions that will be necessary to achieve this. In the next part of this paper we’re taking a look at the research and development sector. We examing how R&D can affect growth difference between the US and Europe.

Research and Development In the long run technological development is seen as the most important factor for competitive strength and in that way for economic growth. The most important factor influencing this technological development is Research and Development (R&D). Characteristic for R&D is that in the research there is a strive for originality and renewal, mostly with the application of information technology, physics and medical science. Europe lags behind the US in R&D expenditure, the R&D in Europe is scattered and despite the high level of science in Europe there are few innovations. This has attracted a lot of attention from policy makers. They claim that the EU needs to narrow down these differences to be able to catch up with the US. At The European Council in Lisbon in 2000 the strategic goal of transforming the European Union by 2010 into the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion were set. They set some clear and specific targets and agreed that Member States should try to get 3% of GDP spent on R&D by 2010. Because R&D is very important for moving towards a knowledge based economy, the EU has put a lot of attention on R&D in order to help meet the goals set at Lisbon. But little progress to this 3% target has been made in the years following, as can be seen by the 1,99% spent on R&D in 2002. Some data on past R&D Business expenditure on R&D is rising everywhere. It increased substantially in the EU, by about 50%, between 1995 and 2001. But growth was even higher in the US. There, R&D expenditure increased by about 130% between 1995 and 2001. In 2000 the European Union and the individual Member States invested a combined total of 164 billion euros in research, compared with 288 billion euros by the USA. In other words, R&D expenditure growth in the US and the EU is very different. Research and Development expenditure represented 1.99% of the European Union’s Gross Domestic Product (GDP) in 2002 against 1.95% in 2000. The gap relative to the United States and Japan however, still is very large, as these countries spent respectively 2.80% and 2.98% of their GDP on R&D. But not in all countries of the EU are the R&D expenditures as a percentage of GDP lower than in the US. Some countries like Sweden and Finland for instance, with respective shares of their GDP of 4.27% and 3.49%, made the greatest research effort. the lowest R&D ratios were registered in the southern countries (Greece, Portugal, Spain and Italy). Figure 5 shows that there are big differences between the US and the European countries. Only Sweden and Finland spend a higher percentage of their GDP on R&D than the US. Greece, Portugal, Spain and Italy have levels far below 1% and in this way they bring the EU average down fast. The EU goal of 3% will be even harder to achieve now, because a lot of new countries have joined, in which R&D expenditures are even lower and in this way the EU average will go down even further.

Figure 5:

Source: Ernst & Young, ‘Beyond Borders, The Global Biotechnology Report 2003’ R&D expenditure by sector The R&D expenditure is usually broken down among four sectors of performance: 1) business enterprises, 2) higher education, 3) government 4) private non-profit institutions serving households If we look at the R&D expenditure by performing sector we see that in the US the business enterprises and private non-profit sector contribute a proportionally much bigger part to total R&D expenditures than in the EU. In the bottom figure we can again see that the R&D expenditures as a percentage of GDP are much higher in the USA than in the EU. It is also visible that both in the US and in the EU the percentages stayed pretty much the same for both sides between 1981 and 1999. And in this way the gap between the US and EU has remained fairly constant over the past 2 decades. It is also visible that the part of the government declined in the US in combination with a growing part of the business enterprises, whereas in the EU there hasn’t been a lot of change between the division between government and business enterprises. This could therefore partly explain why the US is doing so much better than the EU: business enterprises in the US started putting more effort in investing in R&D, whereas the EU has had the same division for almost 20 years now and is very much depending on governments for R&D expenditures.

Figure 6: R&D expenditures by sector

Source: OECD, MSTI database, May 2001, R&D financing and performance

R&D Performance by Industry The United States, the EU, and Japan are the three largest economies in the industrialized world, and their industries have been leaders of innovation in the international marketplace. An analysis of US and EU R&D trends can explain past success, provide insight into future product development, and highlight shifts in national technology priorities. In the US, as can be seen in figure 7 on the next page, in 1987 the total service sector industries accounted for less than 9 percent of all U.S. industrial R&D. But in years after that, the amount of R&D performed in the service sector caught up and even got ahead of that performed by other U.S. manufacturing industries until 1991, when the service sector accounted for nearly 25% of all U.S. industrial R&D. Manufacturers regained their position; however, their share declined to 81 percent of total U.S. industrial R&D by 1996, led by industries making computer hardware, electronics equipment, and motor vehicles

Figure 7: R&D performance US and European Union

Source: Organisation for Economic Co-operation and Development, EAS, ANBERD database 2002 The data for the late 1990s and 2000 show a new rise of the U.S. service sector as the most important performer of industrial R&D. After 1997 a turnaround occurred, which was followed by large increases each year thereafter. The service sector's share of total R&D was less than 19 percent in 1996 but 34 percent by 2000. U.S. manufacturing industries collectively perform nearly two-thirds of the nation's industrial R&D and include most of the nation's top R&D-performing industries. In 2000, the electronic equipment industry was the most important in the industrial R&D. This industry has for a long time been among the top five performers, but its rise to the top coincided with rapid growth in the telecommunication industry during

the late 1990s. Other top R&D performers in 2000 were producers of chemical products, scientific instruments, motor vehicles and the industries providing computer services. Computer and office hardware manufacturers fell out of the top five. As well as the aerospace industry. The R&D performance in the U.S. aerospace industry grew more slowly during the 1990s than in other U.S. industries. It accounted for 19 percent of total U.S. R&D in 1990, but its share dropped nearly every year throughout the decade. By 2000, the U.S. aerospace industry accounted for just 5 percent of total R&D. The European Union: As in the United States and Japan, manufacturing industries account for most of the industrial R&D in the EU. The EU's industrial R&D however, is less concentrated in specific industries than R&D in the United States. Manufacturers of chemicals and chemical products have historically been the most important industry in the EU. Electronics equipment, and motor vehicles have always been among the top five industrial R&D performers in the EU. The top 5 in the EU has consisted of the same industries for a reasonable amount of years now. Only the aerospace industry and the total services have swapped places in between 1995 and 1999. The most striking difference between the EU and the US is that the service sector comes at the 4th place in the EU whereas in the US it is by far the most important. Conclusion: The EU still has a lot of work to do and a long way to go if it wants to catch up with the US in reaching the same level of R&D expenditure, and in this way becoming more competitive with the US and creating more economic growth. This will now even be harder with the expansion of the EU with a large number of ‘’less developed’’ countries. The weakness of R&D and the slow accumulation of knowledge in the EU is probably a major reason why Europe has failed to catch up with the US productivity performance during recent decades. Yet putting all the attention on the spending target for R&D could be misplaced, because it’s not only about increasing the level of R&D, but also about enhancing the efficiency of R&D in Europe. Further, actively subsidizing investment may not necessarily deliver the desired results, as a large part of European capital finds it way to the American capital market and thus does not necessarily benefit innovation in Europe. Consequently, policy meant to enhance the efficiency and productivity of R&D in Europe should focus on the level of knowledge of workers and the capacity of entrepreneurs to come to technological innovation. The relative inefficiency of EU R&D is to a certain degree the result of the segmentation of public research efforts and overlapping of competing research programs, which leads to double work and thus not to the full utilization of the available human resources. The time has now come to create an integrated EU market for research and researchers.

References De Groot, Nahuis and Tang (2004), “Is the American Model Miss World? Choosing between the Anglo-Saxon model and a European-style alternative”, CPB De Mooij and Tang (2003), “Four Futures of Europe”, CPB Ahn and Hemmings (2000), “Policy influences on economic growth in OECD countries: an ecvaluation of the evidence”, economics department working papers no.246 Gordon (2002), “Two centuries of Economic Growth: Europe Chasing the American Frontier”, Northwestern University and NBER “Organisation for Economic Co-operation and Development”, EAS, ANBERD Brandon Shackelford “U.S R&D projected to have grown marginally in 2003’’ Eurostat “European R&D Limps Behind The Rest Science and technology spending in Europe still under 2% of GDP’’ Charles F. Larson “R&D and Innovation in Industry’’ Aswath Damodaran “Research and Development Expenses: Implications for Profitability Measurement and Valuation’’ Simona Frank “R&D expenditure in the European regions’’ Simona Frank “R&D expenditure and personnel in the EU’’

I. 1 (G11) Proximate sources Marc Rooijackers, Ben van Gils, Peter van Oudheusden Bart van de Gevel

SOLOW MODEL VERSUS NEW GROWTH MODELS In this chapter we look at proximate sources of growth. Which sources explain why some countries are richer than others? To answer this question we look at some of the most used growth models in economics. By studying these theoretical models we can deduce which of the proximate sources are important for growth and which not. Growth models can be quite different in how they describe the process of economic growth. To decide which of these models describes the real world best is foremost an empirical question. This is why we will compare some empirical studies to see if predicted source of growth by the models are also in real life decisive for growth. In the first section Marc Rooijackers will compare the Solow model (1956) with the human capital extended version created by Mankiw, Romer and Weil (1992). The mother of all models is the Solow model. This model explains how capital accumulation causes growth. Technical progress in this model is exogenous and falls like manna from heaven. This model explains different levels of growth through different saving rates between countries. The MRW model extended the Solow model by human capital. First the models are explained theoretical. Secondly there is an empirical comparison between the two models. Thirdly he will discuss some alternative growth models with human capital. Lastly this section will also contain a subsection where Peter van Oudheusden gives some critique on the textbook Solow model. In the second section Ben van Gils will treat R&D and Growth. The central question in this section is if R&D can count for Total Factor Productivity growth? He will answer this question by looking at two models, which explain R&D as a proximate source of growth. The first model he deals with is the Romer model (1990). In this model in contrast with the Solow model technological change is endogenous. The second model is the model by Aghion and Howitt (1992) in which growth is caused by ongoing innovations. A question which comes to mind here is if there can be too little or too much R&D. Lastly he will look at the empirical evidence given by Cameron(1998). The last section will contain a short conclusion. Were we will try to give some insights given by models and conclude which proximate sources are important for growth.

-1-

§1.1 The Solow growth model and human capital by Marc Rooijackers

I Introduction The purpose of this paper is to compare the Solow-model with new growth models. After Robert Solow introduced his famous growth theory in 1956, many researchers extended his production formula. In this paper we describe an important extension: the adding of human capital. Many researchers have found contradicting empirical evidence in answering the question: is growth dependent on human capital? We begin with a description of the Solowmodel. Then we describe an extension of this model with human capital, known as the MRWmodel. In chapter three, contradicting empirical evidence will be given of these two models. We conclude our paper with a short description of other growth models. II. The Solow Model This chapter is based on Robert M. Solow’s classic article on growth theory: “A contribution to the theory of economic growth”, published in 1956. The basis of this theory is a standard neoclassical production function. In this production function, there is only one commodity, output as a whole: Y(t). This is the real income of the community. Output is produced with two factors of production: capital (K(t)) and labour (L(t)). The accumulation of capital depends on investment (which is equal to savings), population growth n, technological progress g and depreciation δ. An assumption of full employment at the available stock of capital is made. At a moment of time there is a given amount of labor supply and capital stock. Because the real return to factors will adjust to achieve full employment of labour and capital, we find the current rate of output by using the following production function: Y(t) = K(t)α (A(t)L(t)) 1-α [1] In this production function, A denotes the level of technology. Because the value of α is assumed to be between zero and one, there are decreasing returns to scale. In the Solowmodel, the rates of saving, population growth, and technological progress are exogenous. A fraction of output is saved at a constant s, so that the rate of saving is sY(t). When we define k = K/AL and y = Y/AL, then the change in capital stock can be formulated as follows: dk/dt = sY(t) - (n+g+δ)k(t) [2] Formula two indicates that there is a steady state level k* According to the following formula, this steady state level reached when k equals k* = (s/(n+g+δ)1/(1-α)) [3] The steady state of capital stock k* is the stock of capital at which investment and depreciation just offset each other (∆k=0). This reasoning implies that Solow predicts conditional convergence between countries. One of the shortcomings of the model is that, in the real world, this has not happened. The Solow-model can be tested empirically by two methods. The first is growth accounting, which describes different possible growth patterns of one country. Second, level accounting tries to explain the differences between multiple countries. This paper compares different growth models, the method used is mostly level accounting. III. The MRW-Model For a long time, economists have stressed the importance of human capital to the process of growth as a explanatory variable. The very beginning of this work has been done by Krueger (1968). The first empirical evidence was found from 1988 by Rauch. Azariadis and Drazen (1990) find that no country was able to grow quickly during the postwar period without a highly literate labour force. Romer (1989) finds that literacy in 1960 helps to explain -2-

investment. Also, his findings include that literacy has no impact on growth beyond its effect on investment. A famous model in which human capital is incorporated, is the model designed by Mankiw, Romer and Weil, or the MRW-model. The following is merely based on their article: “A contribution to the theory of economic growth”, published in 1992. Basically, this model is equal to the simple Solow-model described in chapter II, with a third production factor: human capital. This leads us to the function: Y(t) = K(t)α H(t)β(A(t)L(t)) 1-α-β [4] In this model H is defined as the stock of human capital. The MRW-model assumes free substitution of production factors, one unit of consumption can be produced by either one unit of physical or human capital. The MRW-model assumes α + β < 1, which implies that there are decreasing returns to both physical and human capital. It’s important that s, n, g, and A are all exogenous. Because it is now possible to invest in physical and human capital as well, the amount that the economy saves has to be split. Therefore, sk is the fraction of income invested in physical capital and sh is the fraction that is invested in human capital. When we define h = H/AL and y=Y/AL, the accumulation of physical capital is determined by the following equation: dk/dt = sKy(t) - (n+g+δ)k(t) [5] And for the accumulation on human capital counts: dh/dt = sHy(t) - (n+g+δ)h(t) [6] IV. The Solow model compared to the MRW-model As we have seen in the two previous chapters, the difference between the Solow and the MRW-model is only the addition of human capital in the production function. When we compare these models with reality it will be clear which model fits best. There is contrasting empirical evidence of which model fits reality best. The MRW-paper finds a strong positive relationship between human capital and growth. Benhabib & Spiegel (1994) have not found any empirical relationship. Krueger & Lindahl (1999) have found a relationship between human capital and growth but these results are arbitrary. Section 1: Empirical evidence supporting the addition of human capital in the production function This section will summarize the work of MRW regarding to comparing the Solow and MRWmodel. According to Solow, growth depends two important factors. The first is savings, which account for investment and therefore capital accumulation. This results in a larger stock of capital and a larger production (thus income) follows. The second is population, a larger population results in less capital accumulation and therefore in less production. The following table presents the results of testing the Solow-model in three groups of countries: non-oil states; intermediate states and OECD countries.1

1

The first group (non-oil) consists of all countries excluding those for which oil production accounts for most of GDP. It’s not to be expected that standard growth models integrate this production of oil. The second group, intermediate, includes only countries with a population larger than one million. Small countries’ GDP is dependent on influential factors and measurement of data is often not precise. Therefore, the small countries are omitted in the second group. The third sample consists of 22 OECD-countries with population > 1mn. The data are consistent and of high quality, disadvantages are that the sample size is small. For more information, see Mankiw, Romer, Weil (1992) p.413.

-3-

Sample

Non-oil

Intermediate

OECD

98 5.48

75 5.36

22 7.97

1.59

1.55

2.48

Savings

1.42

1.31

0.50

s.e. savings

0.14

0.17

0.43

-1.97

-2.01

-0.76

0.56

0.53

0.84

0.59

0.59

0.06

Observations Constant 2

s.e. constant 3

Population

4

s.e. population

Adjusted R

2

Table 1 : estimation of the Solow-model. Dependent variable: log GDP per working-age person in 1985.5 The main conclusion from this table is that differences in saving and population growth, account for a large fraction of the cross-country variation in income per capita. The adjusted R-square is equal to 0.59. Important to keep in mind is that these results are an example of growth accounting and not level accounting. Other conclusions are that: - savings account positively to growth while population growth has the predicted negative effect; - the effect of savings and populations are highly significant: the standard error is low compared to the actual value coefficient. Besides this theory supporting empirical results, is the model not completely successful. The implied α by the coefficients is: 0.60 for non-oil, 0.59 for intermediate and 0.36 for OECDcountries. Actually, the implied α should be equal to the capital’s share in income, which is about one third. The data strongly contradict this prediction. The same test, using the same data sets and the same countries is performed using the MRWmodel. These results are shown in table 2. Sample

Non-oil

Intermediate

OECD

98

75

22

Constant

6.89

7.81

8.63

s.e. constant

1.17

1.19

2.19

Savings

0.69

0.70

0.28

Observations

s.e. savings

0.13

0.15

0.39

Population

-1.73

-1.50

-1.07

0.41

0.40

0.75

0.66

0.73

0.76

0.07

0.10

0.29

0.78

0.77

0.24

s.e. population

Human capital s.e. human capital

Adjusted R

2

6

Table 2 : estimation of the MRW-model. Dependent variable: log GDP per working-age person in 1985. 2

S.e. stands for the standard error. MRW used ln (I/GDP) which is the growth rate of investment divided by GDP. Basically, this is s. 4 MRW used ln (n+g+δ) which is the growth rate of gross growth. (n+g) was assumed to be 0.05. Population growth remains. 5 Actually, the real OLS regression is performed by estimating GDP per working-age person, according to the following formula which is derived from the basic Solow-model: Y α α
n = α + +
n( s ) −
n(n + g + δ ) + ε L 1−α 1−α 6 The proxy for human capital is a school variable. 3

-4-

The main conclusions from this table are: - Adding the human capital variable is highly significant. We identify a human capital coefficient of 0.66 ranging to 0.76, with standard errors ranging from 0.07 to 0.29; - The adjusted R-square value is much higher compared to the Solow-model. Hence, human capital is important for explaining growth; - In the Solow-model, the implied α was deviating from the predicted one third. Here, the implied α is consistent with this prediction, MRW found α to be 0.28, 0.30 and 0.17 respectively. Section 2: Empirical evidence contradicting the addition of human capital in the production function Some researchers have found empirical evidence that human capital does not contribute to growth. The research done by Benhabib & Spiegel (1994)7 is an example. Their research points out that human capital is never significant in explaining growth. This is tested in six different models, using growth accounting as well as level accounting.8 In three of the models schooling has a linear effect, in the other three a logarithmic effect. This seems to be surprisingly contrasting with the paper of MRW. But Benhabib & Spiegel found also a positive relationship between human capital and physical capital. This means that human capital influences physical capital which influences growth. Thus, human capital has an indirect effect on growth. Krueger and Lindahl (1999)9 have found different results. To start with, they extended the research of Benhabib & Spiegel. They present six different models (regression equations, see their table 1, p. 1112). In two of the six models, the effect of schooling (as a proxy of human capital) on growth is significant. Unfortunately, production input factors as capital and labor are not included in these particular models. This means that the results are doubtful. Therefore, we have to conclude that human capital is not significant for explaining growth. Besides the research and results stated above, they did find other evidence that human capital does not contribute to growth. Their research indicates that initial schooling has no effect on growth of GDP per capita in OECD-countries. Their results are shown in table 3: Variable Sample size Initial schooling s.e.

5-Year growth 138 .000

10-year growth 69 .000

(.001)

(.001)

20-year growth 23 .000 (.001)

R2 .43 .55 .35 Table 3: Effect of schooling on Economic growth in the OECD. Dependent variable: annualized change in log GDP per capita, various time periods. Clearly, coefficients and standard errors indicate that schooling is not useful explaining growth.

7

J. Benhabib, M.M. Spiegel, The role of human capital in economic development. Evidence from aggregate cross-country data. Journal of Monetary Economics (1994). 8 Six different models are used, performing the same test at different groups of countries. 9 Krueger, Lindahl, Education in Sweden and the world. (1999)

-5-

V. Other Growth models The simplest model is the AK-model. Here, production (Y) is dependent on technology (A) and capital (K). The production function is: Y=AK. In the Solow-model, we saw decreasing returns to scale (DRS) unless α is equal to 1. In this case Solow predicted constant returns to scale (CRS). In the AK-model, there is always CRS. Other empirical growth models are the Baumol-type and the Barro-type equations. The most important conclusion from the Baumol-type equations is that one (new) variable explains growth for the biggest part. This variable is the country’s initial GDP. Other variables have only minor influence. Barro-type equations were introduced one year before MRW. These equations contained more than one proxy for human capital. School enrollment (similar to MRW) formed the basis, but adult literacy rates and initial student-teacher ratio were added.10 This equation was subject to the critique of Levine and Renelt who found that a large variety of economic and political indicators, other than the initial enrollment rate, are not robustly correlated with growth. Conclusion The adding of human capital to the Solow-model in theoretical formulas is easy. But when testing the formula in the real world, multiple contradicting evidence is found whether human capital should be added to growth production functions or not. Unfortunately, we cannot give a standard solution to this question. Probably, the actual growth ratio’s of countries are dependent on many factors, including human capital, but cannot be integrated in one standard formula which applies for all countries.

10

Barro, R.J. “Economic growth in a cross-section of countries”, Quarterly Journal of economics (1991).

-6-

§1.2 Critique on the Solow model by Peter van Oudheusden

In this box we will look at the standard textbook Solow model and try to deliver some critique on the assumptions that this model makes. We will do this to look at the variance of the explanatory variables and the variance of the productivity and then calculate the elasticity of capital such that the variance of the first can explain the variance of the latter. Theory: The explanatory variables of the Solow model are the investment ratio [s] and the population growth [n], the assumptions made are that the growth rate of technology [g] and the depreciation rate [d] are exogenous and are the same for every country. In addition the assumption of the textbook Solow model is that the level of technology [A] is given and is also the same for every country. Our starting point is the formula that calculates the steady state level of productivity: α

1.

 1−α s Y  = A ∗  L n+ g +d 

When we take the natural logarithm we get the following formula: Y α (ln s − ln(n + g + d )) ln = ln A + 2. L 1−α For calculating the variance we use the general formula:

3.

VAR(aX + bY ) = a 2VAR( X ) + b 2VAR(Y ) + 2abCOV ( X , Y )

Applying this to formula 2 we get the following: 2

4.

 Y  α  VAR ln  = VAR(ln A) +   [VAR(ln s − ln (n + g + d ))] +  L 1−α   α  2 COV (ln A, (ln s − ln (n + g + d ))) 1 −α 

Because the Solow model makes the assumption that the level of technology is exogenous and the same for every country this results in a variance of zero. Because the level of technology is given we may assume that there is no covariance with any variable. Taking this in account this results in the formula: 2

5.

 Y  α  VAR ln  =   [VAR(ln s − ln (n + g + d ))]  L  1−α 

We can rewrite this as: 5a.

 Y  α  VAR ln  =    L  1 −α 

2

  − ln(n + g + d ) +   VAR(ln s ) + VAR  2COV (ln s,− ln(n + g + d ) 

-7-

We mentioned before that the growth of technology and the depreciation rate are given and the same for every country, we can say now that: 6.

VAR (ln(n + g + d ) = VAR (ln n)

Results data MRW: For the data we used the data of Mankiw, Romer, Weil (1992). We only have taken the non-oil producing countries for the reason that the productivity in those countries cause outliers and are significant contributors to a larger variance in the productivity. Further we have taken two situations, in the first we took the covariance between de investment ratio and the population growth from the data. In the second case we set the covariance equal to zero. In table 1 are the results for the data MRW. Table 1 ln s (-) ln n ln Y/L ln g + d ln A

Variance 0.260 0.338 1.165 0.000 0.000 Covariance ln s, (-) ln n 0.091 ln s, (-) ln n 0.000

alfa 0.550 0.583

The alfa in the table is the elasticity of capital under the assumptions made when testing the model. The outcome says that if the elasticity of capital is 0.550 the variance of the variables (here the log of investment ratio and the log of population growth) can explain the variance in the log of growth in per capita income. As we can see the elasticity of capital in these cases are larger than the predicted 1/3 of the textbook Solow model. Thus, taking the 1/3 the variance in the explanatory variables can’t explain the variance in the productivity.

Critique on the model: The first general critique we can give is that for calculating the elasticity of capital we use the formula for the steady state in the Solow model implementing that all countries are in their steady state, of course this isn’t the case. Further critique can be given to the assumption that growth of technology and the depreciation rate are given. A rather unorthodox but good way of looking at the influence of those variables is taking a random sample of data. This can be done by taking random numbers between 0.04 [g = 0.01, d = 0.03] and 0.125 [g = 0.025, d = 0.10] and add those numbers to the data of MRW11. In the table 2 are the results for the adjusted data.

11

For convenience we assume that there is no correlation between the growth of technology and the depreciation rate.

-8-

Table 2 ln s X (-) ln X ln Y/L ln A

Variance 0.260 n+g+d 0.407 1.165 0.000 Covariance ln s, (-) ln X 0.055 ln s, (-) ln X 0.000

alfa 0.550 0.569

As we can see the elasticity in these cases is hardly influenced by the randomly chosen data. It would be better to take real data to get the influence of these variables. However, this method can say that even if technology and the depreciation rate are randomly distributed over the world it could not explain the variance in productivity according to the Solow model. This brings us to our last critique on the Solow model. As we have seen the model assumes that the level technology is exogenous and the same for every country. When we look at the empirical evidence Jones (1995) says that there is no evidence that there is any variance in the level of technology. However McGrattan (1996) says there is a link between the level of technology and the investment ratio. When we apply his thinking on the data of Jones12 we get a variance of the level of technology of 0.154. As we can see in table 3 the elasticity of capital is also hardly influenced by including this measure of level of technology variance. Table 3 ln s (-) ln n ln Y/L ln g +d ln A

Variance 0.260 0.338 1.165 0.000 0.154 Covariance ln s, (-) ln n 0,091 ln s, (-) ln n 0,000

alfa 0,532 0,565

However, we can say that if we include measures for variance in level of technology, growth of technology and the depreciation rate the elasticity of capital shift to the predicted level of 1/3 in the Solow model. Conclusion: When we look at the textbook model of Solow and the assumptions that are made we can conclude that the model fails to predict the variance in the productivity with the explanatory variables chosen. Even with adding imperfect measures for variables we get results that indicate that these additional variables can explain more of the variance of the productivity.

12

The investment ratio is taken with a lag of one time period on the growth rate per worker.

-9-

§2 Can R&D account for TFP growth? by Ben van Gils

Nowadays output per hour worked is 10 times as valuable as output per hour worked 100 year ago [Maddison 1982]. According to the economists in the 1950’s (Abramovitz 1956, Kendrick 1956 and most important Solow 1956) almost all the change in output per hour worked is attributed to technological change. But research and development plays an important role in the process of technological change and the translation of new technologies into new products and services [Ahn and Hemming 2000]. Therefore I want to investigate the new or endogenous growth model. First I describe some important theoretical work on this topic. Then I focus on the empirical work done by Cameron (1998). At the end of the paper I try to answer the question “Can R&D account for TFP growth?” A pioneer on the endogenous growth is Paul Romer. Romer (1990) states that R&D plays a central role in the production of knowledge. He designed a model that is an extended version of the Solow model (1956). Romer puts technological change into the Solow model. In contrast with the Solow model the Romer model assumes endogenous rather than exogenous technological change. An important characteristic of technology is that developing new and better instructions is equivalent to incurring a fixed cost. Romer’s answer to the question is yes because increases in the size of the market induce more research and faster growth. Larger markets get a larger total stock of human capital that lead to faster growth. In his model Romer uses an A for the stock of technology. Growth in this stock increases the productivity of human capital in the research sector because the total stock of knowledge increases when a new design is created. Romer assumes here that anyone engaged in research has free access to the entire stock of knowledge. This finding suggests that free international trade can act to speed up growth. Another suggestion of the model is that low levels of human capital may help explain why growth is not observed in underdeveloped closed economies but observed in underdeveloped open economies through economic integration with the rest of the world. Another important observation made by Romer (1990) must be mentioned before I go through with the analysis. Ideas are very different from most other economic goods. In contrast, ideas are nonrivalrous. Once an idea has been created, anyone with knowledge of the idea can take advantage of it. Most economic goods share the property that use of the good by someone excludes your use of that same good. But ideas share a characteristic with most economic goods. They are, at least partially excludable. To what extent the good is excludable depends on the degree to which the owner of the good can charge a fee for its use. For this there exists patent systems and copyright laws. Two other important economists on the field of endogenous growth are Philippe Aghion and Peter Howitt (1992). They concluded that growth results from technological progress and this results from competition among research firms that generate innovations. The result of an innovation is that you can produce more efficiently. In a growth process there is obsolescence; the newly improved products replace the old ones. This progress creates losses too so Schumpeter’s idea of creative destruction is born. Because monopoly gains are probably in the future, present firms were encouraged to develop a new and better product so they (the innovator) can access the market for the existing monopolist [Canton 2002]. So the economic growth is determined through the level of innovations. Thus the theory of Aghion and Howitt also explains how R&D contributes to growth. Only the economic growth can be too high or too low according to them. On the one hand knowledge spillovers lead to lower R&D. Because entrepreneurs only look at the revenues of innovation in their business cycle and

- 10 -

when a new innovation is present than the firm is driven out the market. On the other hand an innovation leads to a better product and drive existing products out the market. This phenomenon is known as business stealing. The concept of how much R&D should be used is described by Jones and Williams (1998). They investigate the over- or under investment in R&D. According to them the optimal R&D spending as a share of GDP is more than two to four times larger than actual spending. In their research they make a distinction between the private and social marginal product of research. This distinction is because of three reasons. First there is a stepping on toes effect. Doubling the number of researchers does not double the number of new ideas discovered. It can happen that two or more of them came with the same idea. Another reason is the standing on shoulders effect. An innovation is based on knowledge of the previous product. Because of some flaws of the existing product the researchers can come up with a better product. The last reason is creative destruction and is described above in the piece of Aghion and Howitt (1992). If you want to know if R&D can account for TFP growth you must also look at the returns of R&D. According to the research done by Cameron (1998) there exists a strong and significant link between R&D capital and output. In contrast to the above studies he found empirical evidence for the yes answer. A one percent increase in the R&D capital leads to a rise in output of between 0,05% and 0,1%. This conclusion is obtained by using two different ways of estimating the effect of innovation. The first way is to use a measure of stock of R&D capital (Dt) in a regression of the level of TFP. He uses the equation: log TFPt = log A + β log Dt + µt The second way is to use a measure of R&D intensity in a regression of the change in TFP, shown in the following equation: dlogTFP = ρ (Rt /Y)+ µ The important difference between these two ways is that the first is based on a level estimate and the second on a change estimate. But there are a number of problems with these two approaches, both theoretical and empirical. The various studies still have not proven that knowledge is separable in the production function. Empirically there are the usual measurement problems. R&D data are a problem because of problems of definition, inflation, depreciation and the treatment of time lags. Cameron finds that it is also important to adjust for the cyclical nature of the TFP data. I think this is a very good point because we are dealing with economics and therefore we must take economic trends into account. The results of the two approaches are summarized in table 2 for the first approach and in table 4 for the second approach (Appendix). Table 2 presents the estimate of the output elasticity of R&D (β). Table 3 presents the estimate of the rate of return to R&D (ρ) in the USA and table 4 for other important countries. When these parameters are known you can determine the effect of R&D on growth. But the outcomes of the studies are too different that this is almost impossible. The output elasticity of R&D ranges from 0 to 0,5 and the rate of return to R&D range from 0 to 1,43 (direct rate of return, because of too many empty spaces for the indirect rate of return). The important point that you can make about the outcomes that they are all positive and this means that there is a positive relationship between R&D and growth. The conclusion of the paper is that R&D can account for TFP growth. Theoretically this is explained by Romer (1990) and Aghion and Howitt (1992). Empirically this is explained by Cameron (1998). He founds a positive relationship between R&D and growth. - 11 -

§3 Conclusion In this chapter we compared the Solow model with several new growth models. The reason we did this was that these models give us the answer which proximate sources are important for growth and which not. We have seen that the Solow model explains growth through accumulation of capital. Where the level of growth is explained by the saving rate of a country and the growth of its population. However the empirical evidence doesn’t supports the traditional Solow model. Our own research with respect to the textbook Solow model shows that the model fails to predict the variance in the productivity with the explanatory variables chosen. Mankiw Romer and Weil have shown that the inclusion of human capital can improve the empirical results and that there is strong evidence that human capital is a source of growth. Economist like Benhabib and Spiegel however found contradicting evidence on the role of human capital on growth. This means there is still no unanimous consensus on the role of human capital. Other economists have looked at R&D and innovation as sources of economic growth. In the Romer model growth in R&D is endogenous. By investing in R&D the human capital stock of a country broadens and growth follows. Aghion and Howitt have made another endogenous technological growth model. In their model growth is been caused by innovations. Firms that compete for a monopoly in the market generate these innovations. Cameron gives the empirical evidence that R&D can account for TFP growth. He concludes that there is a strong and significant link between R&D capital and output. Our conclusion is that there are several proximate sources of growth. Traditional theory predicts that the saving rate and population growth are important sources. Though we think that these still account for economic growth we stress that there are more sources. In this chapter we have looked at two other sources: human capital and R&D. We think that these two sources are probably connected to each other and complementary. We think there is evidence that they account for a significant part of growth. Besides these two sources there may be other sources that are important in explaining growth. One that we didn’t discuss in this paper is social capital. With social capital we mean stocks of social trust, norms and networks that people can draw upon to solve common problems.13 Recent research suggests that this source could also account for growth.

References Aghion, P. and Howitt, P. (1992), A model of growth through creative destruction Ahn, S. and Hemmings, P. (2000), Policy influences on economic growth in OECD countries: An evaluation of the evidence Barro, R.J. “Economic growth in a cross-section of countries”, Quarterly Journal of economics (1991) Benhabib J., M.M. Spiegel, The role of human capital in economic development. Evidence from aggregate cross-country data. Journal of Monetary Economics (1994). 13

This explanation is based on the Civic Dictionary, http://www.cpn.org/tools/dictionary/capital.html

- 12 -

Cameron, G. (1998), Innovation and growth: A survey of the empirical evidence Canton, E. (2002), Onderwijs, R&D en economische groei Jones, Charles I. (1995). Time series tests of endogenous growth models. Quarterly Journal of Economics 110 (MAY): 495-525 Jones, C. and Williams, J. (1998), Measuring the social return to R&D Kroeger, Lindahl, Education in Sweden and the world. (1999) Maddison, A. (1982), Phases of capitalist development Mankiw, Gregory N.; Romer, David; Weil, David N.(1992) A contribution to the empirics of economic growth. The Quarterly Journal of Economics 107 no: 2 407-437 McGRattan, Ellen R. (1996). A defense of AK growth models. Federal Reserve Bank of Minneapolis Quaterly Review 22no: 4 13-27 Romer, P.(1990), Endogenous Technological Change

- 13 -

Appendix

- 14 -

- 15 -

- 16 -

I.4 (G17) Proximate sources Zhou Beichun Nanfei Pei Jacob Glowacki (manager) Ying Chen

STAGES OF ECONOMIC GROWTH “It is not the strongest of the species that survive, nor the most intelligent, but The one most responsive to change” [Charles Darwin]. Introduction While there are numerous economic theories that devote time towards explaining how growth can be achieved, how growth can be sustained and why some situations and developments are detrimental to a sustained growth process, only relatively few resources are devoted to the task of outlining the path a newly developing economy will take throughout its growth process. While most theories allow to discern likely developments and key moments from the underlying functions and resulting phenomena, key development stages as one would could offer, for example in the realm of biology or astrophysics, are usually not touched upon. One reason for this avoiding behavior might be the scope the economist has to consider: the development of an economy from undeveloped to an industrial economy is a lengthy, complicated and not entirely clear-cut process, where substantial and convoluted developments in various economic and social spheres have to be taken into account, forcing the economist to look beyond purely economic data and analyze trends and developments as regards demographic variables and social change. It is a challenge to the economists to study the development and growth from the point view of the population, technology and output. How can we explain the evolution of human society from the agricultural economy to the industrial regime, and now entering the modern growth stage? Why do some countries escape the agricultural economy earlier than other countries? Some studies on the relationship between the population, technology and output cast light on those questions. The Malthusian model explains one thousand years of the development of human history; the latest studies by Galor, Weil and David try to explain how human society evolves from the Malthusian regime to the Post Malthusian, and finally enters the Modern Growth era.

1. Malthusian growth theory General overview The earliest Malthusian model developed by Thomas R. Malthus in 1798 (Malthus 1798) can explain most of the history of mankind predating modern times. The Malthusian model includes two key components (Galor and Weil 2000). The first component is the existence of some factor of production such as land: this factor is in fixed supply, implying a decreasing rate of return on all other factors. The second key component is the positive relationship between the standard of living and the growth rate of the population. The Malthusian model postulates that in the absence of technological progress or the availability of land, the population is self-

equilibrating. The increasing standard of living will be the underlying cause of the growth of the population. As a result, there is no big difference in the standard of living across countries under the Malthusian regime. Malthusian growth theory and human history The prediction of the Malthusian model fits well with most of the documented human history. The following study on population growth and output in the western counties gives a nice picture of the Malthusian regime: from figure 1 below, we can see that the population growth rate is about 0.1 percent during the years 500 A.D. to 1500 A.D.; according to Massimo Livi-Bacci (Livi-Bacci 1997), the average growth rate through year 1 A.D. to 1750 A.D. is about 0.064 percent. The resulting growth rate of output per capita is almost zero in the time interval from 500 A.D. to 1500 A.D.. Furthermore, the studies from Easterlin (Easterlin 1981), Pritchett (Pritchett 1997), and Lucas (Lucas 1999) show that the differences in the standard of living among countries were quite small prior to the year 1800, yet there existed some difference concerning the level of technology: For instance, China’s sophisticated agricultural technology before the nineteenth century did improve the output per acre, but failed to raise the standard of living above the subsistence level. A similar situation occurred in Ireland where a new productive technology in potato growing caused an increase in the total population but no improvements in the standard of living. Therefore, under the Malthusian regime, the development process in the economy did not improve the living standard, but rather resulted in low yet steady population growth. In other words, the advantage of any new and improved production process was offset by a corresponding burst of growth in the population under the Malthusian regime. As a consequence, the rate of technological change is extremely low. Kremer (1993) argued that the change of the size of the population can be seen as a direct result of changes in technology. Society development-escaping the Malthusian trap After a thousand years of development, the population in Western Europe maintained a stable growth rate. Furthermore, the output per capita started to increase after 1500 A.D., with a significant jump at the beginning of the nineteenth century. See figure 1 and figure 2.

Figure 1: Output growth in Western Europe 500 – 1990 (Galor & Weil 2000)

Figure 2: Output per capita in western Europe from year 0 to 2000 (Galor and Moav 2002)

2. Evolution of the Human society Galor and Weil Galor & Weil 2000) develop a unified model to describe the evolution of the economy from the Malthusian regime to the Post-Malthusian regime, and finally to the Modern Growth Era. The analysis focuses on two important issues from a macroeconomic viewpoint: first, the behaviour of the income per capita; second, the relationship between income per capita and the growth rate of the population. As can be seen in figure 1, both the growth rates of the population and output increased after 1820 A.D.. Indeed, the growth rate of the population began to fall after it reached its peak in the nineteenth century whereas the output per capita kept increasing. Population growth accounts for about 40 percent of the output growth during 1820 A.D. to 1870 A.D., but for only 20 percent during 1925 to 1990, while now the current forecast for the population growth in Europe predicts a negative growth rate. The unified endogenous-growth model tries to account for the stated facts and show that the developments in population growth, technology and output growth can be described and explained consistently. The core point in this model is the interaction between human capital and technological progress: technological progress raises the rate of return of human capital while in turn human capital influences the rate of technological progress, changes in attitude towards one variable will allow for dynamics to arise. In the following pages, we will explain what effects the technological progress had on the evolution of the human society; of particular focus will be the role of mortality. We also give an overview of the latest research from other points of view. Effects of Technological progress The model by Galor and Weil produces a Malthusian pseudo steady state that remains stable for a long period but disappears in the long run. In this Malthusian regime, output per capita is quite low and technology progresses slowly. Because of the modest technology progress, the return to human capital limited. As described earlier, any economic development induced by technological progress was offset by corresponding growth in the population. In other words, the growth of the population can be regarded as the indicator of the technology

development. As a result of the limited return to human capital and the corresponding features of the economy, parents would focus more on the quantity of children rather than quality in a Malthusian regime. The pseudo steady state vanishes because of a steady accumulation of technology development and population growth. At a sufficiently high population level, the technology progress is high enough to provide parents with an incentive to invest in some human capital, i.e. in their children. Subsequently, the increasing level of human capital induces an acceleration of the technological progress, which in turn raises the value of human capital. In fact, the technological progress has two effects on the growth of the population: on one hand, a higher level of technology awards families higher income. As a direct result, the families become able to raise more children. This is what happened in the post Malthusian regime, when the population experienced a growth spur. On the other hand, higher income and return to human capital induces parents to allocate more resources to the quality of child instead of quantity. Eventually, rapid technology development resulting from the increase in the return of human capital triggers the demographic transition: wages and the return to the quality of children increase, but the population growth rate decreases. This phenomenon can be observed the Western European countries. The role of mortality As the development dynamics in the unified model of endogenous growth hinge on the accumulation of human capital, changes in mortality may play a pivotal role in describing the transition from a Malthusian regime to a Modern Growth Regime: Galor and Weil draw an example where an exogenous shock to health technology causes a change in mortality that is not immediately observable. Consequently, as fertility is not adjusted immediately, population growth will increase. However, this rise in population growth will combine with the lower mortality to generate higher returns to human capital, in turn spurning the rate of technological change that in turn will lower mortality. Advanced Malthusian models A later study from Galor and Moav (Galor and Moav 2002) furthered the development of a unified model to capture the interactions between the evolution of the human society and the economic development by illustrating the relationship between population, technology and output. Their study endeavours to depict the interaction between four key elements: The first contains the main ingredients of the Malthusian world - the economy is characterized by land as the fixed factor of production, and a subsistence consumption constraint below which survival becomes impossible. The second element incorporates a number of Darwin’s attributes of evolution such as variety, natural selection, and an overlapping process of evolution, into a Malthusian economic environment. The Malthusian pressure imposed is an important influence to the evolution of human species. Galor and Moav propose that although the individual does not operate consciously in order to assure an evolutionary advantage, the existence of a variety of types enable the selection process to select those who fit the economic environment. The third element links the evolution of the human species to the economic growth process - based on the hypothesis that the development of human capital has a positive effect on the technology progress, and ultimately results in economic growth.

The fourth element links the rate of the technological progress to the demographic transition and to sustained economic growth. The final argument is consistent with that of Galor and Weil (Galor and Weil 2000): The increase in technological progress increases the return to human capital, in turn inducing parents to substitute for quantity by quality as children are concerned. Hence, the growth of the rate of the technological progress increases the average quality of the population, and in turn accelerates the rate of technological progress even further. Ultimately, technological progress reaches a sufficient level in order to induce a reduction of the fertility rate, generate a demographic transition and sustained economic growth.

3. Rostow’s theory Next to the theory based on the Malthusian regime, other approaches also are proposed: Rostow (Rostow 1960) originally proposed stages of growth model of rather descriptive nature. Within his framework, he identifies five categories of development stages: the traditional society, the preconditions for take-off, take-off phase, the drive to maturity and finally the age of high mass-consumption. a) The traditional society: this state is characterized by a society with preNewtonian science level and attitudes. The economy features only limited production functions; innovation is possible but usually does only increase total output while output per capita remains unchanged. The productivity of capital is low; labour is used primarily in agriculture. This stage corresponds roughly to the Malthusian steady state. b) The preconditions for take-off: this is a transition period during which the conditions for take-off are developed: science features prominently and begins to induce technological progress. At the same time, capital mobilization becomes commonplace, allowing for investment and private ventures. The society and economy do not change during this transition period and traditional structures and production methods prevail. c) Take-off: the society receives a decisive stimulus of usually technical nature. Capital per worker increases due to higher savings rates and thus higher capital mobilization. Agriculture becomes increasingly industrialized, necessitating less labour per unit of output and allowing for a greater population to be sustained. The structure of society changes: entrepreneurs enter the scene, while farmers with small farmsteads are reduced in number. d) The drive to maturity: this is a period of prolonged growth and technological progress. The economy benefits from a high investment rate and international trade dynamics with other developed societies. Industry structures are being established, leading to a well-defined modern economy. e) The stage of high mass-consumption: the growth in income per capita experienced in the last stages allows for universal consumption patterns beyond necessities, skilled workers are prominent in the workforce. The economy performs a switch towards durable consumption goods and services, economic growth is technology driven. While Rostow’s framework serves descriptive purposes primarily, the expressions are used more or less universally throughout the literature. Some researchers try to

move closely along the stages within their theoretical concepts. One such model is proposed by Zilibotti (1993). In this model, Zilibotti tries to implement the known stages by Rostow into a theoretical model of economic growth. He does this by matching the stages proposed by Rostow to growth models and subsequently develops transition mechanisms. Thus, a Solow-type model featuring a steady state is employed to model the corresponding steady-state stage while a Romer-type model is employed with the beginning of the take-off stage to model the following sustainable rate of economic growth generated by technological change. The model thus builds on a number of ‘engines’ to provide a fitting description to the respective development stage.

Bibliography Darwin, Charles R. “On the Origin of Species by Means of Natural Selection: Or the Preservation of Favoured Races in the Struggle for Life”, London: John Murray 1859 Easterlin, Richard “Why Isn’t the Whole World Developed”, Journal of Economic History, XLI (1981), 1-19 Galor, Oded & Weil, David N. “From Malthusian Stagnation to Modern Growth”, The American Economic Review Vol. 89 (1999), No. , 150-154 Galor, Oded & Weil, David N. “Population, Technology, and Growth: From Malthusian Stagnation to the Demographic Transition and beyond”, The American Economic Review Vol. 90 (2000), No. 4, 806-828 Galor, Oded & Moav, Omer “Natural Selection and the Origin of Economic Growth”, The quarterly journal of economics Vol. 117 (2002), No.4, 1133-1191 Kremer, Michael “Population growth and technological change: one million B.C. to 1990”, Quarterly Journal of Economics CVII (1993), 681-716 Livi-Bacci, Massimo “A Concise History of World Population”, Oxford: Claredon Press 1997 Lucas, Robert “The Industrial Evolution: Past and Future”, University of Chicago 1999 Malthus, Thomas R. “An Essay on the Principle of Population”, London: J. Johnson 1798 Pritchett, Lant “Divergence, Big Time”, Journal of Economic Perspectives 11(3) (1997), 3-17 Rostow, Walt W. “The Stages of Economic Growth: A Non-Communist Manifesto” London: Cambridge University Press 1960 Zilibotti, Fabrizio “A Rostovian model of endogenous growth and underdevelopment traps”, European Economic Review Vol. 39 (1995), 1569-1602

I. 5 (G14) Ultimate sources Sjoerd Kitzen Jeroen Broersma Statistician Rianne Heijboer Robert Gielissen

GEOGRAPHY VERSUS INSTITUTIONS AND POLICY

1. The Influence of Geography and Institutions on Economic Growth By: Jeroen Broersma1 and Rianne Heijboer2

Economic growth is fuelled by the accumulation of production factors or improvements in productivity, through technological change and efficiency improvements. More fundamental forces, or “ultimate sources”, in turn drive these proximate sources of growth. Examples of ultimate sources are institutions, market imperfections, policies or geography, which affect economic growth in an indirect way through affecting accumulation and productivity. In this chapter we discuss the influence of institutions and geography on economic growth. Does either of these two factors affect economic growth? Is there any connection between those two? First we address the role of geography. We investigate whether institutions affect economic growth. Having considered these two sources separately, we discuss how the two are connected. Is there causality between these sources? Are they direct or indirect sources of economic growth?

Geography and economic growth It is widely argued that geography affects economic growth and development in the context of understanding cross-country differences and economic performance. Acemoglu et al. (2004) approach focuses on differences in geography, climate and ecology. These factors determine both the preferences and the opportunities of each individual agent in different economic societies. They refer to this broad approach as the geography hypothesis. We will briefly look at the three variations of his hypothesis, which each emphasize a different aspect. Firstly, climate is one of the important determinants of work effort, incentives or productivity. This idea has been around for a long time. The heat of climate can be so 1

Statistician

1

excessive to the human body, that it drains the strength and causes unproductiveness. As Acemoglu et al. put it: “the inhabitants of warm countries are, like old men, timorous; as opposed to the people in cold countries, who are like young men, brave.” (2004:13) Second, geography can determine the technology available to society, especially in agriculture. (Acemoglu, 2004:13) We should always take in to account the climate and its impact on soil, vegetation, animals, humans and physical assets, in other words we should take into account living conditions in economic development. Sachs is another proponent of the view that geography is important in agricultural activity. He states: “By the start of the era of modern economic growth, if not much earlier, temperate-zone technologies are more productive than tropical-zone technologies....” (2001:2). Sachs also argues that geography exerts an independent effect on the public health environment and on transport costs. The third variant of geography hypothesis links poverty in many areas of the world to their ‘disease burden’. Sachs emphasizes: “The burden of infectious disease is similarly higher in the tropics than in the temperate zones” (2000:32) Furthermore in an earlier paper he claims that prevalence of malaria reduces the annual growth rate of Sub-Saharan African economies by more than 1.3 percent a year. This is a very large effect considering if malaria had been eradicated in 1950 income per capita would be double of what it is today. Olsson and Hibbs (2004) formalize a slightly different argument. They focus on biogeography as being the fundamental impact on economic development all the way back to prehistoric times. Countries with favourable bio geographic conditions, such as the prevalence of plants and animals suitable for domestication were much faster in the transition from hunter-gatherer to sedentary agriculture society. This would ultimately lead to the rise of civilization, which conferred a development head start of thousands of years over the less well-endowed areas. They claim three contributions: firstly, their paper is one of the first attempts to explain the Neolithic transition from hunting gathering to agriculture. Secondly, they formalize a study in the spirit of Diamond (1997) of the causal links between bio geographic endowment, technological progress, and current levels of prosperity. The last claimed contribution is that their study confronts the theory with biogeographically data. This means that their empirical results demonstrate a relationship between initial biogeography and income levels that does not run through institutions which contrasts to the work of other scholars, who argue that geography only affects current prosperity indirectly via institutions. All of the different geography hypothesis variations mentioned above show us that there is a link between endowments and economic growth and development. We will analyze the empirical aspect of the geography hypothesis further down after discussing the effects of institutions on economic performance.

2

Analyst

2

Institutions and economic growth North (1990: 3) defines institutions as follows: “Institutions are the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction.” Economic institutions are considered one of the most profound causes of economic growth and cross-country differences. They ensure a good structure of property rights and give equal opportunities to everyone as well as ensuring the presence and perfection of markets. There must be enforcement of property rights for a broad section of the society so that everyone has an incentive to invest, innovate and take part in economic activity. A certain degree of equality of opportunity is also necessary so that those with good investment opportunities can take advantage of them. Acemoglu et al. believe that the structure of markets is endogenous, and therefore partly determined by property rights. Equality of opportunity and secure individuals’ property rights together create private incentives to maximize social welfare. Differences in market performance thus must be an outcome of differing systems of property rights and political institutions (Acemoglu, 2004:13). Acemoglu et al. find that institutions are among the most important causes of long run growth. They set up a framework in which today’s economic institutions influence future economic performance and distribution of resources; this in turn together with future political institutions affects political power, which is the determinant of both economic and political institutions. (2004:6)

Empirical evidence on determinants of economic growth Through out this section we will discuss the empirical evidence on the determinants of economic growth found by different studies. We will also look at the association between geography and institutions. Are they direct or indirect sources of economic growth? Many different studies have focused on the role geography and institutions play in determining economic growth. Acemoglu et al. (2004), Easterly/Levine (2003) all of these find empirical evidence that is in favour of the institutional hypothesis. Acemoglu et al. (2004) find in their paper that institutions are the main determinant of economic growth. They argue that there is convincing empirical support for the hypothesis that differences in economic institutions are the cause of differences in incomes per capita. They outline this by using a scatter plot showing the cross-country bi-variate relationship between the log of GDP per capita in 1995 and a broad measure of property rights. The outcome is imperfect as a measure of economic institutions but the findings are robust to using other available measures of economic institutions. The outcome of the plot shows us that countries with more secure property rights have higher average incomes. Nevertheless other aspects should be taken in account as well. We cannot automatically assume that there

3

is a casual link between secure property rights and prosperity. There can be a problem of reverse causation or of omitted variable bias. To avoid these problems of inference we should use natural experiments. They use the Korean and the colonial experiment. The clearest one goes over the subdivided Korean peninsula. After the division two independent countries came about. They organised in different ways and adopted different sets of institutions. The South attempted to use markets and private incentives in order to develop the economy. In the North however the economic decisions were made by the communist state rather than mediated by markets. South Korea maintained the system of private property of land and capital as in North Korea they were abolished. Both countries had the same history and cultural roots. They showed remarkable resemblance on ethnic, linguistic geographic and economic factors. So any difference in economic growth should be attributed to differences in institutions. Consistent with the institutional hypothesis the Koreans have experience dramatically diverging paths of economic growth. South Korea has developed in one of the Asian Tigers and has been a member of the OECD since 2000, while North Korea has a level of per-capita income about the same as a typical sub-Saharan African country. The only plausible explanation for this difference in economic development is the different institutions, which the country used. (Acemoglu, 2004:18-21) The Korean case is not as helpful as it is perhaps too extreme, the difference between a market-orientated economy and a communist one provide insufficient evidence of the importance of economic institutions. The colonial experiment shows that colonies, which have inherited good institutions from their colonizers, are performing well. The countries, which were sparsely inhabited, had low mortality rates or had less extreme climates when the colonial period started, have developed good institutions and therefore good economic performance. (Acemoglu, 2004:20 and 2001) Easterly and Levine (2003) focus their analyses of the determinants of economic growth, they evaluate four main questions regarding endowments and institutions using a sample of 72 former colonies. For tests 1 and 2 the authors used simple OLS regression, whereas for the 3rd and 4th tests two stage least squares model were used. •

First, do institutions explain development? From appendix A3 it is clear that the institutions index has a positive association with both economic development and better endowments.

•

Second, do endowments explain cross-country variations in economic development? The results we refer to can be found in appendix B.4 There is clear indication that endowments explain economic development. Each of the four endowment indicators

3 4

Figure 3 and Table 1 in Easterly and Levine (2003). Table 2 in Easterly and Levine (2003).

4

which are used (settler morality, latitude, landlocked and the crops/minerals indicator) significantly explain cross-country variation in the logarithm of GDP per capita •

Thirdly,

do

endowments

explain

cross-country

variations

in

institutional

development? The result of the regression (appendix C5) indicates that endowments help explain variation in institutional development. Each of the indicators is significantly related to the aggregate institutions index. •

Finally, do endowments explain cross-country variations in economic development beyond their ability to explain cross-country variations in institutions? This is the most important test of their paper. In the second stage of the regression the authors find that the dependent variable for geography does not differ from zero. The results (referring to appendix C3) provide strong support for the institutional hypothesis but no evidence for the geography hypothesis. Geography in this model is an indirect source of economic growth.

Olsson and Hibbs (2003) have a rather different view than previously discussed; they set up a theoretical framework over three major stages in history; the hunter-gatherer stage, the agricultural stage, and the industrial stage. Their main aim was to provide a formal representation of the link between initial bio geographic conditions and the present level of economic well-being. The essence of their model implies that the earlier the transition from hunter-gatherer to agricultural production, the longer the period of endogenous growth of knowledge, and as a result, the earlier the transition to industrial production, and the higher the level of economic development will be. The empirical analyses of the study focuses mainly on contemporary levels of economic development, the regressions accounted that up to half of the 1997 international variation in log output per person could be explained by their exogenous measures. Olsson and Hibbs interpret this evidence as strong support of their central prediction, which is that current variation in economic prosperity, is significantly affected by prehistoric productive potentials of various environments. They also claim that the geographic and biogeographic signals that were detected in the current levels of income per person were robust to the controls for political and institutional variables that are known to exert powerful, proximate statistical influence on international variations in economic prosperity. (Olsson and Hibbs, 2003:28) In Olsson and Hibbs (2003) the framework set out makes the principle prediction of initial bio geographic endowments being positively related to contemporary levels of economic growth. They set out using measurements of some key geographic and bio

5 3

Table 3 in Easterly and Levine (2003). Table 3 in Easterly and Levine (2003).

5

geographic variables: the size of a continent, major axis of continent, climate and the number of animals and plants that are suitable for domestication. The relationship is shown in figure 1 (in appendix D) When the regression is set out the exogenous geographic conditions are measured by four variables: Climate, latitude, axis and size. Biogeographic endowment is measured separately with two variables, plants and animals. There are also variables for political environment and social infrastructure. From the results they gather that there is much evidence in line with Acemoglu et al. (2001) and Hall and Jones (1999): they conclude that political and institutional measures have a clear proximate and statistical effect on economic performance. Olsson and Hibbs (2004) retain the common view that institutional arrangements have much influence on the wealth or poverty of nations. However, ‘good institutions’ cannot themselves be regarded as a fully independent variable, unaffected by geography, biogeography and the level of economic development. Although some studies such as Acemoglu et al. have claimed that the effects of geography are mediated entirely through institutions, the results in Olsson and Hibbs (2004) show that biogeography as well as geography retain statistical significance and remain important even if a measure for institutional quality is included in the multiple regression. They conclude by emphasizing that geography and biogeography (as the prime mobiles leading to current prosperity) have an undisputed prior causal status over both institutions and current per capita incomes.

Conclusion Acemoglu et al. (2004) state that geography influences economic growth. Geographic aspects influence the type of institutions a country develops. In particular, economic institutions that protect property rights and give equal opportunities to everyone boost growth. Acemoglu et al. present evidence for the hypothesis that geographical elements affect growth indirectly through institutions, whereas institutions directly affect economic growth. Economic institutions determine incentives and constraints, and thus shape economic outcomes. Their analysis relies on historical examples, political economy models, and econometric testing. Easterly and Levine (2003) show that geographical variables significantly correlate with growth in a simple OLS regression. However, the direct link between geography and growth vanishes when a two-stage least squares method is applied. In the first stage, institutions are significantly explained by geographical variables. In the second stage, the authors find an insignificant coefficient for institutions, which are instrumented by geographical variables. The authors conclude that geographic indicators, tropics, crops and germs, shape institutions, which in turn explain economic growth. Thus geographical

6

elements play an indirect role on growth only, which is similar to the conclusion by Acemoglu et al (2004). Olsson and Hibbs (2004) reach different conclusions. Although they support that institutions are a main source of economic growth, they find evidence that geography and biogeography also directly affect growth, besides affecting growth through institutions. The view elaborated by Acemoglu et al and Easterly and Levine can be graphically summarized by the following scheme:

Geography

Institutions

Economic Growth/Development

While Olsson and Hibbs’s view can be summarized by:

Geography

Institutions

Economic Growth/Development

Summarize, we conclude that Acemoglu et al. and Easterly/ Levine find empirical evidence in favour of institutions as the ultimate explanation of income differences, while Olsson/Hibbs find evidence in favour of biogeographical determinants. What explains the differences in results by these two groups of researchers? •

Acemoglu et al. and Easterly/ Levine only use data on former colonies; Hibbs and Olsson use a sample of 112 countries. In particular, they do not take the ‘neoEuropean’ countries (Australia, New Zealand and North America) in account, because the food and technology package have not developed indigenously in these countries but have been transferred by the European colonizers. In Acemoglu et al., in contrast, these countries play a crucial role in their argument, providing examples of colonization processes that brought good institutions that translated in high growth.

•

Acemoglu et al. start their analysis at the onset of colonization 1500 AD, but Hibbs and Olsson use the period of the Neolithic revolution, 11.000 BC as starting point for their analysis. “In historical time, geography and biogeography are the primes mobiles of current prosperity.” (Hibbs and Olsson, 2003) Hibbs and Olsson distinguish three stages of growth: the hunter-gatherer economy, the agricultural economy, and the industrial economy. Acemoglu et al., in contrast, confine the analysis to the period after the colonization.

•

Hibbs and Olsson give a causal link between biogeographical determinants and the present level of income and development today. “Regions with a well-endowed natural environment, which consequently made the transitions to agriculture and

7

industry comparatively early, should other things equal have higher income per capita today than more poorly endowed regions where the transitions came later.” (Hibbs and Olsson, 2003:19)

Bibliography •

Acemoglu, D., S. Johnson and J.A. Robinson (2001). “The Colonial Origins of Comparative Development: An Empirical Investigation”, American Economic Review, December, 91, 5, 1369-1401.

•

Acemoglu, D., S. Johnson, and J.A. Robinson (2004). “Institutions as the Fundamental Cause of Long-Run Growth”, NBER working paper #10481. .

•

Easterly, W. and R. Levine (2003). “Tropics, germs and crops: How endowments influence economic development”, Journal of Monetary Economics 50, 3-40.

•

Hall, R. and Jones, C. (1999). “Why Do Some Countries Produce so Much More Output per Worker than Others ", the Quarterly Journal of Economics, Vol. 114, 83-116.

•

North, D.C. (1990), Institutions, institutional change and economic performance, New York, Cambridge University Press.

•

Olsson, O. and Hibbs, D. (2003). “Biography and long-run economic development”, Elsevier econbase, European Economic Review.

•

Olsson, O. and Hibbs, D. (2004). “Geography, biography, and why some countries are rich and others are poor”, Proceedings of the National Academy of Sciences (US), vol. 101, no. 10: 3715-3720.

•

Sachs, J. D. (2000). “Notes on a New Sociology of Economic Development” in L.E. Harrison and S. P. Huntington eds. Culture Matters: How Values Shape Human Progress, Basic Books, New York.

•

Sachs, J.D. (2001). “Tropical Underdevelopment,” NBER Working Paper #8119.

8

Appendix A Table 1: Correlations and summary statistics: selected variables

9

Appendix B

10

Appendix C

11

Appendix D

Figure 1: Biography and long-run economic development (Olsson and Hibbs 2003:20)

12

2. The influence of policies and institutions on economic growth

In the studies discussed in the previous section, institutions were found to have a strong direct effect on economic growth. Institutions are the result of a long historic and cultural process. For policy relevant questions, one would like to turn to ultimate factors behind growth that can be changed over a shorter time period. We capture these factors by the term “policy”. In this section we investigate the relationship between policies, institutions and economic growth. First we will define what we mean by policies and institutions. Then we discuss two opposing views: the one that policies affect economic growth and the view that policies have no effect on growth. Eventually, we will conclude this section with our view.

Definition By “policies” we mean the set of decisions made by the government. For example, the decision of the government to keep inflation low (monetary policy) is an example of a good policy. Other examples are the decisions on how high the tax-rate in a country will be and the degree of openness to international trade. There are many characteristics of the economy that a government can influence, so the term ‘policies’ is a rather broad term. The policy-view, as this view is called in the literature, is therefore a view with many possible explanations. The term ‘institutions’ has been defined in the previous section, but the distinction between policies and institutions is sometimes blurred. By institutions we mean the way a society is organised, which can only change slowly over time. Contrary, policies can be changed very quickly, for example a government setting the tax-rate a bit higher.

The view that policies do affect economic growth As mentioned in the previous sector, Acemoglu (2004) concludes that institutions are the fundamental cause of long-run growth. Economic institutions influence economic performance and the distribution of resources. This will, together with political institutions, influence the distribution of political power. And finally, the distribution of political power will influence economic and political institutions. Acemoglu describes in his paper that institutions, formed hundred of years ago, determine whether policies are “good” or “bad”. So in this view, policies do affect economic growth, but these policies are determined by the institutions formed long ago.

Bassanini et al (2001) discuss different ways in which policies can influence economic growth. They claim that policies have a direct influence on growth. The research in this paper is based on growth-regression studies of OECD countries.

13

First, they claim that monetary policy to control inflation is important. The evidence found in this article suggests that high inflation is negatively associated with the accumulation of physical capital in the private sector and, through this channel, has a negative bearing on output.

Average growth and median inflation in equal-sized samples of annual data for OECD countries

Moreover, a high variability of inflation affects growth, possibly because it leads to a shift in the composition of investment towards lower return projects. The authors present a regression for 21 OECD countries over the 1971-1988 period. In this regression it indeed becomes clear that a high variability of inflation has a negative effect on economic growth, as the coefficient is significant and negative. Second, the empirical evidence also confirms the importance of financial markets for growth, both by helping to channel resources towards the most rewarding activities and in encouraging investments. The authors of this paper don’t make a distinction between policies and institutions; they see these terms as being essentially the same.

Also Ahn and Hemmings (2000) conclude that policies affect growth. Their paper is a literature review of many theoretical and empirical papers, which is far too broad to fully describe here. Their main-conclusions on macro-economic policy are that low and stable inflation affects economic growth positively. According to the authors there are several reasons why high variability of inflation affects economic growth. Besides the uncertaintyeffect described earlier, there is the relative price distortion. Tommasi (1993) for example 14

finds that price distortions increase with inflation, which will have a negative impact on growth. Another policy they examine is the policy on international trade. Theoretically, they state that more international trade will lead to exploitation of comparative advantages, economies of scale and more competition. This will eventually lead to more economic growth. The empirical evidence on these theories is mixed. There are studies that find a positive link between openness and growth (for example Barro, 1991), but there are also a lot of studies that find a negative link (Levine and Renelt, 1992).

Summarizing, we can conclude this part by saying that there are theories that state that economic policies (like aiming at a stable inflation) affect economic growth. These theories are confirmed by empirical evidence en regressions.

The slowdown in growth However, there is some empirically evidence against the view explained above. The most cited fact is that in developing countries the growth rates in the past two decades was much lower than in the 1960s and 1970s.

Although the Washington Consensus which aimed to introduce neo-liberalistic policies in developing countries, median per capita growth in these countries was around zero (as compared to 2,5% in the sixties and seventies). If we assume that governments learn what policies are best for their countries, we would expect higher growth-rates in the last decades, instead of lower rates. Easterly (2001) has made a study on this topic. He states that if one uses a standard growth model, with ‘the usual’ variables such as financial depth and real

15

overvaluation, initial conditions like health, education, fertility and infrastructure, these models would predict economic growth for developing countries in the eighties and nineties. He shows that the real deposit rates improved in the 80s, the overvaluation of the real exchange rate diminished in the 80s and 90s and that life-expectancy increased. Overall, this means that the economic policies indeed improved in time, what can be expected. But, these improved policies contrasts with the slowdown in growth. This evidence would give rise to the idea that policies do not affect economic growth. Easterly then tests whether he disappointing growth-rates are caused by some economic shocks, like an interest shock or a shock in the term of trade. These shocks don’t seem significant in explaining the slowdown. The only explanation of the slowdown is the slowdown in growth in OECD-countries, because they have a huge impact on the growth of developing countries. This can be seen in the regression:

Per capita growth in developing countries regressed on policies, initial conditions, and trading partner growth rates

16

According to Easterly, this result means that policies are less important for growth than stated in the previous view. The paradox of improved policies in combination with lower growth is one of the reasons that another view has been given attention in the literature, namely the view that policies don’t affect economic growth.

The view that policies do not affect economic growth We will now describe some papers that state this view. We will begin with Mukand and Rodrik (2002). They also point out, as Easterly, that economic policies in developing countries have been improved. As an example, they give some figures on Latin America. The

17

index on structural reform created by Morley et al. rises in Latin America from 0,47 in the 70s to 0,82 in 1995 (with 1 as maximum). But, in contrast to this figures, there are but a few countries that have a higher growth in the 90s. Only Chile really does it better. Another example that policies don’t seem to have much influence is the rise of India and China. The authors state that these countries, whose growth-rates have been remarkable, have introduced policies that have been regarded as unorthodox (like limited deregulation, absence of clear property-rights, partial liberlization). India was after these reforms still the one of the worlds most protected countries. In contrast, countries that introduced the so called ‘good’policies, like Latin America, didn’t experience higher growth-rates. Mukand and Rodrik then present a model that will explain these differences. They state that it policies are specific for a region. What will be a good policy in one region, can be disastrous in another region. Furthermore, they state that countries will follow a country that has introduced a good policy, either by implementing the same policy or by experimenting their own policy. If one country experience high growth-rate due to a good policy, neighbouring countries will adopt this policy. How further away, how less well the policy will work. So countries at intermediate distance will at average have fewer gains than direct neighbouring countries. Countries far away see the success of the leading country. But because they are far away, they know that the policy isn’t suitable for them. In reaction, they experiment with an adapted policy. This will result in another remarkable gain, or in a disastrous fall, because the effect of the experiment isn’t known. So this theory explains that countries in the neighbourhood of a leading county have a bigger chance on success, and more importantly, that countries far away can have either great successes or big failures. The authors test their model with 32 formerly socialist countries and they conclude that the success of some Asian republics is indeed explained by this model So these authors don’t say that policies don’t affect economic growth, but they state that there are some constraints on the implementation of policies. So the same policies don’t have to work for different countries. Easterly (2003) further elaborates the arguments. He gives some examples why policies could have a smaller impact on growth than is stated by authors of the previous view. Although theory predicts that taxes on capital should reduce investment on hence growth, in practice many types of capital (e.g. human capital and capital in the informal sector) are not taxed. Hence increases in taxes on conventional capital can be avoided by substituting to nontaxed capital goods, thus reducing the effect of taxation on growth. In the literature the author finds evidence that policies don’t affect growth. First he describes the paradox explained above. Second, he claims that the literature has failed to find a link between the most obvious policy that would influence economic growth, tax-rates, and

18

economic growth. Third, data on policies start in the 60s, and a lot of important data from before the 60s is then not incorporated. Eventually, Easterly conducts some empirical research. He regresses six policyvariables for the years 1960 – 2000 on economic growth. It seems that they have a positive and significant effect on growth.

Efffect of one standard deviation improvement in each policy variable on economic growth

However, if the extreme cases are left out, it turns out that none of the variables is significant anymore. So, Easterly concludes that if extremes are left out, policies don’t affect economic growth. Only in extreme cases policy’s matter, and he explains this by stating that it’s quite logical that very bad policies affect growth negatively. But that doesn’t mean that good policies will affect growth positively. This can be seen as “it is a lot easier to cut down a tree than to grow one” (Easterly, 2003).

So we can conclude this part by saying that there are authors who say no connection between policies and growth, or at least a connection with some conditions.

The interaction between policies and institutions We now come to our last part, the interaction between policies and institutions. As is already said, Acemoglu’s view is that institutions determine the policies of today. Easterly also has conducted a research on this topic, by regressing policies and institutions on growth. He finds that de macro-economic policies are never significant if institutions are being regressed too (Easterly 2003). This would suggest that policies are indeed determined by institutions.

19

Another research, by Rodrik (2004), examines the question whether economic reforms (which can be seen as improved policies that had a lot of impact on the economy) caused economic growth. This study identified 80 periods since 1950 where a country had a growth of 2% or more per year for a period of at least 7 years. The majority of these cases were unrelated to major economic reforms (as liberalization and opening up trade). This is a quite pessimistic view, because countries with bad institutions will not improve. We find it logical that it is possible that good policies can improve institutions. For example, if the institutions are of low quality, but a shock takes place (a revolution, or an very good president), then it would be possible that good policies undertaken by this president could improve the institutional quality. For example, the economic and political changes by Gorbatsjov in the Soviet Union improved the institutions. Also the more gradually changes in China moved the economy to a more capitalistic system. Notice that it are new policies that influences the institutions, although it happens rather slowly (Li, 1998). Nowadays China grows at a remarkable speed. This discussion is rather difficult, however, because the distinction between policies and institutions becomes narrower. Furthermore, it is more reliable to trust cross-country studies then to look at some examples, because we don’t know which other factors are responsible for the changes in Russia and China.

Conclusion Based on our review of the recent literature, we have to conclude that it is not clear if good policies affect economic growth positively. There is evidence that there is a positive link, but there is also a view that states that there isn’t. The idea that bad policies are bad for growth, is wider accepted. The idea that policies are determined by institutions is confirmed by other authors, although there isn’t either any consensus here. And we think furthermore that there can exist a positive effect from policies on institutions. Graphicaly, this can be depicted as:

Institutions

Economic growth/development

Policies

20

Literature: Acemoglu (2004): Institutions as the fundamental cause of long-run growth Bassanini (2001): Economic growth: the role of policies and institutions Tommasi (1993): Inflation and relative prices: evidence from Argentina. Barro (1991): Economic growth in a cross section of countries. Levine and Renelt (1992): A sensitivity analyses of cross-country growth resgressions. William Easterly (2001): The lost decades: Developing countries' stagnation in spite of policy reform 1980-1990, World bank, February 2001 William Easterly (2003): National Policies and economic growth: a reappraisal, New York University, March 2003 D. Li, (1998), The dynamics of Institutional Change in China: The role of the bureaucracy. (Geciteerd uit: http://ieas.berkeley.edu/shorenstein/1998.03.html) Rodrik (2003), Getting institutions right. Mukand and Rodrik (2003), In search of the holy grail: policy convergence, experimentation and economic performance.

21

I. 6 (G13) Ultimate sources Joris Ament Willem Kerstholt (manager) Renske Kok Peter van Moorsel

POLITICAL ECONOMY AND ENDOGENOUS INSTITUTIONS MODELS Introduction

Political Economy - Endogenous Institutions Models I - Introduction Modern political economy models study the consequences of differences in institutions on economic growth. [At the time of Adam Smith, the term political economy had a different meaning: it comprised the study of national income and distribution in general]. In particular, a stream of economics literature called the “New Institutional Economics” devotes special attention to the role of institutions and the endogenous nature of institutions. II - Major Approaches in Political Economics There are various ways of approaching political economics. The major approaches are: State, Justice, Economic, and Power-centered approaches. Below these different kind of views will be discussed: State centered approaches The classical theory in political economy presumes that the state only fulfils a responsive role. The state only responds to the economic processes. State-centered approaches shift the balance between market and state. The role of politics in the economy does not start with the identification of the market failures. The politics have their own goals they wish to achieve, and they accomplish that by implementing policies that influence economic processes and institutions. So state centered approaches refer to the imposition of political agendas to the economy (Caparosa and Levine, 1993). Justice centered approaches The justice centered approaches do not cope with the subject of market’s success or failure, but with the rights that come forward of the market. A market is, ultimately, a system of property rights (Caparosa and Levine, 1993). The political process can define and alter these rights. These rights often concern the distribution and the equality of the recourses. The question is, of course, what is fair? What is a fair distribution? There are two kinds of justices: substantive justice and procedural justice (Craren, 2004). Substantive justice concerns the outcome of the process. It concerns with the amount of recourses everybody has at the end of the process. In contrary procedural justice is not about the outcome, but about the process itself. For example

in general is in Western countries procedural equality high and substantial equality low. For communistic countries it is reversed. In general the procedural equality is low and the substantial equality is high for that countries. This approach is subjective and dependent on the ideology of a person. Economic centered approaches At this approach economic theories are applied to policy. It makes individual maximizing behavior and rational choice the theme of political process and political decisions. This method is inspired by the neo-classical theory, but that theory doesn’t imply a great role on the government. In this approach that is just the case, there is a great role for the government. Yet the role for the government doesn’t come down to the public responsibility for economic affairs, or by trying to prevent market failure. The government should interpreted politics as one focus to applied economic logic. Power centered approaches This approach identifies politics with the use of power and, by finding power in the economy, claims to have established that the economy is political. There are two main problems, which make this approach complicated. The first problem lies in the way power and politics are related. Those who use power within the economy do so in the pursuit of their private interests. This does not (directly) need to involve a political struggle over the instruments and institutions of power (in other words; the government). If it is not involved with that struggle, then power-centered approaches to political economy have to do with politics only in a very limited sense of the word. The second problem, which occurs at this approach, is the consideration whether the use of power is positive or negative. Examples like the power to improve technology are of course positive. Naturally enough examples of negative power exist. A good distribution of power makes an economy more efficient. So unlike the other approaches to political economy, power-centered approaches do not focus on the relation between private interests and public decision-making, but focuses more on the allocation of efficiency in the society (Caparosa and Levine, 1993). III - Endogenous institutions approaches A group of approaches links the performance of economies to the incentives the institutional setting provides to the participants in the economy. North (1990) defines institutions as: “[…] humanly devised constraints that shape human interaction.” Institutions thus are the rules of the game to which each economic actor should obey. In many countries the rules of the game are the outcome of a political process, hence the institutional approach is a part of political economics. It borrows from several of the previously treated approaches. Political actors have influence on the decision process in so far as that they have the power to change the rules. Neoclassical, non-political, economics treats institutions as exogenously given, usually not of any influence on the results of the market or on the behavior of the atomized actor. The institutional approach (also known as New Institutional Economics) on the other hand sees the actor as embedded in social relationships. Institutions govern the behavior between human beings, and gradually change over time. This last point is an important point because it is the step between institutions and the development of economies, and thus the link between politics and economics (Yeager, 1999). Different groups in societies have different interests. In an ideal democracy each groups can promote the benefit of the own group only by gaining enough support for the idea to let it be implemented by politicians of the own choosing. In this sense, the institutional setting will

reflect, or gradually change to come to reflect, the interests of vested powers. When social settings change, and certain groups rise in power and other fall, the institutional setting may reflect archaic social stratification. The old holders of power, still in charge of the politicians in case of a democracy, or for example in charge of the army in less free societies, will refrain from changing the institutional setting towards the interests of the new holders of power. The new holders of power will nevertheless gradually be able to alter the situation. In certain cases this can lead to civil wars, but in democratic situations the changes can occur peacefully. The key proposition is that the relation between different categories in a civilization may, via political processes, change the relative price structure, which might, depending on the direction of the changes, better or worsen the economic position of a country (North, 1990). In the remainder of this chapter the endogenous institutions model will be applied to the last two thousand years of European history. In this history several examples of clashing groups in society will be apparent, with differing results for the local economies. IV - A Practical Application – Europe, From Rome to Mercantilism After the decline of the Roman Empire, Europe underwent several drastic changes to both its political and its economic structure. Former Roman holdings became lands governed under the feudal system, during which agriculture played a crucial role in daily life. Increased demand for a wider variety of goods, as well as new discoveries abroad delved Europe into several centuries of mercantilism, where each nation-state involved attempted to maintain a trade-surplus in order to fund successful military campaigns to protect both their colonial holdings and their trade routes. This increased level of nationalism also involved the colonization of Asia, Africa and the New World, as well as the strengthening of national borders within Europe. The face of Europe was altered again when, first in England and later the rest of Europe, the Industrial Revolution moved many former agricultural workers into urban areas to work in factories. These drastic changes also affected the class system in Europe. There always was a rich minority, just as there has always been a poor majority. However, as the political and economic climates of Europe changed, so did the relationships between the upper, middle and lower classes. From Rome to Feudalism The roots of European feudalism lie in the political and economic structure of the Roman Empire. Rome’s economy was based on agriculture, and the poor peasant farmers who worked the land were required to pay a land tax, irrespective to frequent bad harvests or their inability to work the land while engaged in military service. Those farmers who defaulted on their debts could be sold into slavery, at the same time losing their land. The virtually powerless peasants, in the face of frequent raids from northern European Barbarian groups, as well as the demands of the Roman tax collector, turned to powerful figures for protection. Often, the peasants gave up ownership of their land in return for protection (Jupp, 2000). It is this practice that caused the European feudal system to develop after Rome’s decline. Like the Roman Empire, the feudal system in Europe depended massively on agriculture. The peasants, under the protection of their king and his army, farmed their individual tracts of land. The peasant’s entire family would help, as this was the family’s occupation. In return for his protection, the ruler/owner of the land would require a tribute from his peasants, whether it be in the form of agricultural product, money or military service (Jupp, 2000). Everything the peasants required was either produced on the farm, received from their ruler, or obtained through barter with local artisans such as blacksmiths or weavers. The king, in turn, received

his revenue from his subjects. The political and economic institutions in this case are very easy to understand, as a solely agricultural economy requires little trade, so no revenue can come from this source. Therefore, the governing party, the rich landowners, sold protection in return for tribute. There existed a very small, extremely wealthy ruling class as well as a very poor class of peasants and artisans. Also better off than the peasants were the ruler’s fighting men, and these enforced the class system. A change in either political or economic institutions was unlikely at this point because the king, using his feudal knights, could easily put down any peasant revolt. There was also little incentive for the king’s knights to attempt to seize power, because they were richer and more powerful than all but the king himself. The only group that would possibly want to challenge the king, the peasants, lacked the power to do so. A considerable increase in the level of trade eventually caused the decline of the feudal system. This increase in trade introduced wider variety of available goods. Merchants set up their shops in towns, and almost all trade was conducted there. The economic institutions that were in place at the time were ill designed to profit from an increase in the level of trade. Landowners only taxed their peasants, and the farmlands they worked, and therefore did not collect any rent on the business conducted inside towns (Jupp, 2000). Farmland declined in value relative to towns, causing wealthy landowners to lose revenue, and with that, power. Merchants, in turn, only gained wealth and influence. The ruling classes, who became determined to reap the benefits of this new economic institution, did not ignore the increase in importance of trade. To Mercantilism Mercantilism became Western Europe’s major form of economic and political institution in the seventeenth century. (Irwin, 1991) Trade’s rise to importance during the decline of feudalism elicited in many nation-states the desire to out-perform other nations. Countries, or rather the ruling classes of these countries wanted not only to reap the benefits of trade, but also that other nations performed worse as a result. During this time, the smaller feudal powers consolidated into larger and more competitive nation-states (LaHaye, 2004). Herein lies the idea of mercantilism; a nation through its trade policies, as well as its military actions, would attempt to outperform its rivals. The fierce competition caused this time period to be characterized by nearly constant warfare between the major participating nations. The forces of each nation were charged to both protect its own mercantile interests, as well as disable that of other nations. To finance these constant military campaigns, trade-surpluses were required, specifically the in-flow of gold and silver resulting from the trade abroad (Cameron, 1989). There were opposing views on whether mercantilism was beneficial or detrimental to an economy. Smith’s Wealth of Nations portrayed mercantilism in a bad light, pointing out that through protectionist policies, a nation forwent the potential gains from trade. Several German historians and economists, particularly Gustav von Schmoller, opposed Smith’s views. Schmoller believed that the mercantile system was “above all a policy of state-making carried out by wise and benevolent rulers,” which at the same time helped in strengthening national economies. (Cameron, 1989) The political, and therefore also the economical institutions in the major mercantile nations differed significantly, and these differences, in turn, led to varying levels of success. In the absolute-monarchies of France, Spain and Portugal, the royalty used its absolute power to reap the profits of its merchant fleet. The royalty spent all of the profit from trade on their own lavish lifestyles and frequent military campaigns. The great majority of the populations of these nations lived in extreme poverty. Because the royalty, and not better informed trade experts were making all of the decisions, these absolute monarchies did not develop as successful a trade empire as did the Netherlands and England. The success of these two nations relative to the absolute-monarchies arises from the increased

competency of the decision-makers. The monarchs of England and the Netherlands did not have absolute power. Instead, the power was in the hands of a wider, albeit still small, range of the population. These people, such as the ‘stadhouders’ in the Netherlands, had closer ties to trade than did the monarchs themselves, and were therefore better equipped to develop more efficient economic institutions. These merchants convinced their rulers to grant them outright monopolies on the national trade, and were also given protection by the nation’s military. Thus, the much more efficient system of trading abroad, paired with the exploitation of their colonial holdings allowed the wealth of these nations to flourish. (Irwin, 1991) Because not only the aristocracy of these countries was profiting from the mercantile ventures, this greater wealth was also spread over a larger fraction of the population. In the cases of the Netherlands and England, the specialized and highly skilled merchant class used its seafaring expertise and its influence to allow its power rose drastically in relation to royalty. In both England and the Netherlands, the political institutions, constitutional monarchies, resulted in greater wealth, and also allowed for the lower merchant class to achieve even more political power. The royalty of these two nations was willing to concede this political power, because they too were earning massive sums of money thanks to the expertise of the merchants. Advances in seafaring led to the rise of mercantilism in the seventeenth and eighteenth centuries, and it would be another type of advance that would help to rearrange the political, economic and class systems in the eighteenth and nineteenth centuries. This time technological advances in England greatly increased the productivity of capital, in a period that came to be called the industrial revolution. With the numerous advances in technology, production of goods, such as textiles, that was once time-consuming and expensive, could now be done quickly and at lower costs in factories. The industrial advances centralized production in factories and thereby necessitated the move of a large fraction of the population into urban environments. This centralization, though miraculous for the factory-owners and other rich industrialists, also “gave rise to an industrial proletariat,” the poorly paid, massively overworked employees of the new factories. (Habakkuk, 1965) These people were the former farmers and artisans whose work were displaced by the more efficient machines of the industrial age, and had little influence and few rights. The pay was incredibly low, and the work was difficult and often dangerous. The use of child labor was abundant in this period, and workers had very few political and economic rights. There loomed the threat that this incredibly poor working class would cause social unrest, which was the cause for several “strategic concessions, which were made in England to adult men over an 86 year period.” (Acemoglu, Johnson and Robinson, 2004) Starting in 1832, these concessions were made by the ruling class, knowing that they were giving up political power. However, the concessions were deemed worth the loss because they served to appease the working class thereby avoiding potentially disastrous revolts. Thus the initial stages of the industrial revolution served to drive England’s population into the factories and into population. However, by organizing, the poor masses were able to force concessions of more and more poiltical and economic powers and freedoms, because they possessed de facto power in their ability to revolt. Towards the beginning of the 20th century, however, there remained a small group of wealthy rulers in each country, extracting the majority of the rents from their business interests. There also remained nations, the majority of whose population was now an unskilled, extremely poor factory worker. Though the proletariat did earn gradual increases in both compensation and rights, they still did not achieve the political power to be able to earn any significant income.

V - Conclusion Political economy, the interrelationship between the political and the economic affairs of the state, is a very complex and broad subject. Over the years different theories have been made about this subject. The subject also can be viewed out of different perspectives, out of which different approaches originate. It should be clear that it is an area that keeps on developing. Endogenous institutions models give a synthesis between economic reasoning and more sociological discourse. It gives an explanation of differences in the attained income and how these result from distribution of power and the way changes in this distribution are handled. If there is anything to be learned from these 1500 years of class struggle in the face of changing political and economic institutions it is that the real rents to be collected from any successful economic venture fall into the hands of those with political power. In cases where a small group possesses the political power, such as in the absolute monarchies of France and Spain in the 17th and 18th centuries, any wealth gained due to economic prosperity will fall into this ruling minority. In many ways, this is still the same, despite the fact that political power today is not concentrated among nearly as few people as in the past. The use of protectionist methods, such as tariffs and voluntary price floors protects from international competition those groups who have the political influence to acquire this type of protection. Politicians, in an effort to maintain their political power, attempt to appease these influential groups in exchange for their support. Thus, it is important to realize which parts are played by whom when analyzing a change in political or economic institutions. The course of the last 1500 years also makes it evident that for a lower class to rise in status, this class must acquire either economic or political power. Unrivaled expertise in shipping and trade were what propelled the merchant class into prominent roles, and the seizure of de facto power in the form of threatening social unrest forced rulers to concede more political representation to the industrial-age English proletariat. Despite many major institutional changes, the distribution of power has long determined, and still determines the distribution of wealth. This must be kept in mind when assessing present-day and future changes in government or international strategy, on both domestic and international levels. Bibliography 1. Acemoglu, Johnson, Robinson. “Institutions as the Fundamental Cause of Long-Run Growth” NBER Working Paper Series. Working Paper No. 10481 (May, 2004). 2. Basgen, Brian and Andy Blunden. “Encyclopedia of Marxism”, 1999, http://www.marxists.org/glossary 3. Cameron, Ronaldo. “A Concise Economic History of the World.” Oxford: Oxford University Press, 1989. 4. Caporaso, J.A. and D.P. Levine, “Theories of Political Economy”, Cambridge University Uress, 1993 5. Craren, S. 2004. “Justice based Political economy”, University of Texas, Dallas 6. Gabriel, S. 2004. “Introduction to Political Economy, Course Outline” http://www.mtholyoke.edu/courses/sgabriel/political_economy_main.htm 7. Gamble, A. and A. Payne. 1996. “Editorial policy statement from: New Political Economy” Vol. 1, No. 1 8. Habakkuk, H.J. and M. Postan, Eds. “Cambridge Economic History of Europe VI part I.” Cambridge: Cambridge University press, 1965. 9. Hartwig, J. 2004. “Introduction to Political Economy, Course Outline” (now offline) http://darkwing.uoregon.edu/~jhartwig/classes/ps321

10. Irwin, Douglas A. “Mercantilism as Strategic Trade Policy: The Anglo-Dutch Rivalry for the East India Trade.” The Journal of Political Economy, 99.6 (1991): 1296. 11. Jupp, Kenneth. “European Feudalism from its Emergence through its Decline” The American Journal of Economics and Sociology, (Dec, 2000). 12. LaHaye, Laura. “Mercantilism.” Library of Economics and Liberty. 20 September http://www.econlib.org/library/enc/Mercantilism.html 13. North, Douglas C. 1990. “Institutions, Institutional Change and Economic Performance.” Cambridge University Press. 14. O’Brien, Patrick. “Mercantilism and Imperialism in the Rise and Decline of the Dutch and British Economies 1585-1815.” De Economist 148.4 (2000): 469-501. 15. Steele, G.R. “The money economy: mercantilism, classical economics and Keynes' general theory.” The American Journal of Economics and Sociology, Oct, 1998. 16. Sidwick , Henry. 1887. “The Principles of Political Economy”. MacMillan and Company, 2nd edition (taken of the internet). 17. Yeager, Timothy J. 1999. “Institutions, Transition Economies, and Economic Development”. Boulder: Westview Press.

I. 7 (G12) Endogenous Institutions Pieter van Erp Dirk Hendrix Marcel Jonker (manager) Luc Verschuren

EXPLANING ECONOMIC AND POLITICAL INSTITUTIONS IN FOUR COUNTRIES 1. France vs. Great Britain Introduction In this paper a comparison is made between the development of the institutions in Britain and in France. Britain and France are both West-European countries and have a population of 60 million people. Britain and France were two rivaling countries, which have fought several wars. Both countries are now rich and powerful, but got in this position trough different paths.

France 1.1 Historical development of institutions before the Revolution Until the beginning of the sixteenth century France was not a stable country. The Romans, Germans, Muslims and the Britain’s invaded France. In France a lot of civil wars were fought, but in the early 16th century the France Crown was strengthen and Francois I became the king of the Frankish monarchy. In this period the first political institutions became clear. In the beginning the France Crown had the de jure and de facto power. In this period property rights were introduced but the nobility still ruled strongly. The economic institutions could not be taken too serious. At the end of the 16th century the de facto power began to change. The number of Protestants began to grow and tried to take over the power. This led to the Wars of Religion. Henry IV ended this Wars of Religion and became the first Bourbon King. Henry IV introduced the Edict of Nantes, which guaranteed religious property rights for Catholics and Protestants and political rights. (Source: Website Landenweb) In the 17th century King Louis XIII and Cardinal Richelieu made France from a feudal monarchy to an absolute monarchy. The Feudal parliament and the nobility were almost never summoned for political decisions. They started a rebellion. When this rebellion was suppressed the power of the French Monarchy was at his highest. This absolute monarchy caused a lot of different political and later economic institutions King Louis XIV, the Sun King, took over and allocated power to Versailles where al the princes and lords lived. France then became a very centralized ruled country. France becomes the most powerful country in Europe. Louis XIV promoted industries, trading and colonization oversees. The influence of France in the colonized countries was relatively small because France brought mainly slaves to this countries and only little of their own inhabitants. Many institutions for the higher culture were established like police protection. (Source: Engerman, Sokoloff) Louis XIV decentralized France and divided his government into almost forty smaller governments. This way France had a more incentive full tax system. The bureaucratization of Louis the XIV started out well through stimulation of the France market and subsidizing the 1

transport cost of French merchants. The overseas trade was directly controlled because of the strong monarchy. The absolutist monarchy could also increase their revenues by granting trade monopolies. However the reign of Louis XIV failed in several areas. For example, Louis XIV revoked the Edict of Nantes and got France in financial troubles. When he died, Louis XV took over. Louis XV was not able to rule as strongly as Louis XIV did. The de facto power of the Bourgeoisie was rising and England became the largest colonial power. The economic institutions could not be maintained. The Bourgeoisie, feeded by the Enlightment, made the commoners question the principles of the old regime and absolution. The commoners wanted equal rights and the abolition of the class system. In 1789, the crisis came to a head and the Bourgeoisie wanted more rights. The French revolution had begun. (Source: Website Landenweb, Website UNCG). In Figure 1.1 the changes within the French system are showed by theory of Acemoglu. (Source: Acemoglu 2004) Figure 1.1

The initial political situation of an absolute monarchy was charactarized by certain de facto and de jure power. These powers, on their turn, affected the economic and political institutions in France. These institutions and the influence of the de jure and de facto political power realized new institutions and a new distribution of resources. This is shown in the last two boxes. 1.2 Historical development: The Revolution until the present This revolution has been very important for the institutions France has today. The revolutionaries issued the Declaration of the Rights of Man. This declaration was meant to end the class system and embodied freedom, equality and brotherhood. The Declaration of the Rights of Man consisted a lot of things that are important today like liberty, property rights for the people in the Bourgeoisie and security. During the revolution, people gained property rights, because the Bourgeoisie was now able to own their own firms and did not have to hand over their profits to the nobility. The Revolution ended when Napoleon entered Paris and became First Consul. In 1804 he crowned himself emperor. He thereby challenged and diminished the power of the church. Napoleon reformed the education system and justice system and started a powerful central administration and expanded his empire with many military campaigns. A new period of different political institutions had started. In this period France got some very important new economic institutions, like more secure property rights and the Patent Law promulgated in the Code Civil.(Source: Acemoglu). The military campaigns got France in North-Africa where France became a colonial power in North-Africa. In Europe he conquered a large empire. For example, Napoleon conquered Italia, Germany and Poland. His plans of war got him increasingly bigger problems. Russia defeated Napoleon and England pointed out to be to 2

strong at sea. Napoleon was defeated in Waterloo in 1815. (Source: Website Landenweb, Website UNCG) The French Crown was restored. The centralized administration of the Republic was conserved but the nobility and the church took over power. This reminded people too much of the old regime and this system got not be maintained. In 1848 Louis Napoleon, the nephew of Napoleon Bonaparte was elected the first president of the Second Republic. After having an argument with the legislative power Louis Napoleon made himself emperor of France. Louis Napoleon started to reshape the French economy. He realized for example a free trade pact with England Under his power the trade and industry flourished. This started the industrial revolution in France. Louis Napoleon got into a war with Germany, Italia and Russia and was defeated. The empire ended in 1871 when the war with Germany ended and Louis Napoleon was defeated. (Source: Website Landenweb, Website UNCG) In this period France became a democracy. The industrial expansion continued during this time and the French economy grew fast. This growth was mainly fed by several Pacts made with other European countries. This growth had slowed down in the 20th century when France got into the First World War. One of the pacts was the Triple Entente with Russia and England. When one of these countries, Russia, got involved in the First World War France had to participate. Until 1917 France wasn’t very successful in their war against Germany. In 1917 the defense of France was structurally reorganized and France defeated Germany. (Source: Website Landenweb, Website UNCG) After the war the relation between France and a lot of European countries worsened, because France wanted too much reparations. The economic situation in France worsened, because of the postwar problems and the unstable government. A new cabinet was chosen and this cabinet tried to strengthen the alliances to stop the pressure of Hitler’s Germany. But when Germany violated the Pact of Munich, France started the Second World War against Germany. The Germany troops pulverized the France army. Until September 1944, the Germans suppressed France. The invasion in Normandy made sure that France was a free country again. (Source: Website Landenweb, Website UNCG) The postwar years where characterized by economic growth, consumerism, technological advancement and political instability. At fast speed the France cabinets succeeded each other. In 1958 De Gaulle became the new president of France. He changed a lot of political institutions. He issued a new fundamental law, which centralized the power of the government. The period of De Gaulle is characterized by restoring the position of France between the influencing countries in the world. De Gaulle ruled till 1969 when he resigned but his policy was maintained. France grew to become one of powerful countries of the European Union and the world. In the coming decades France stayed an influencing and strong economic country. The power of the president decreased a little in the 20th century but France stayed a centralized ruled country. (Source: Website Landenweb, Website UNCG) 1.3 Conclusions Over time a lot of different thing happed in France and France was ruled by a lot of different nations and people. In the 17th century the power of nobility started to grow and in the 18th century, France became an absolute monarchy. A lot of new institutions were created but almost all of them favored the nobility. However, the power of the Bourgeoisie was growing.

3

This lead to a Revolution, which caused power to shift in the direction of the Bourgeoisie. In this period, France was ruled by Napoleon. Napoleon established a lot of important economic and political institutions, which are still very important nowadays. When Napoleon was defeated at Waterloo a lot of things changed. After a short period of changes in leadership, France became a relatively stable democracy. Over time France grew to become the country as it is today. When we look at the institutions and the theory of Acemoglu, the period between 1789 and 1815 had been important for France. In Figure 1, it is shown how the institutions and distribution of resources have changed over time. During this period of changes, the Bourgeoisie got securer property right and the Code Civil was issued. After this period of changes, the Industrial Revolution caused France to grow and become a rich and powerful country as it is today.

Great Britain Historical development of institutions The history of Britain is exciting and far from stable. Over a long period of time, the Romans, the Anglo Saxon and the Vikings entered Britain and conquered parts of the country. After the Norman invasion in 1066, England became a unified country for the first time since the Romans left 600 years earlier. The Norman kings consolidated their hold on England, and then took control of Wales and Ireland. Britain became a monarchy. In the beginning the kings had the de jure and de facto power. (Source: UK information guide, British History) After Henry II, England started running into problems. People and pretenders to the throne tried to gain political rights. Upcoming parliaments where developed and from that moment on, kings failed to maintain an absolutist monarchy. This started when King John (1215) was defeated by the barons and only kept the throne by signing the Magna Carta, which stated that the king was not above the law, that he only ruled by the will of the people, and that if he broke his part of the contract, then the people had the right to overthrow the king. This did limit the king’s powers of taxation and require trials before punishment. It was the first time that an English monarch came under the control of the law. This reduced power of the king (monarchy) and the increased power of the people can be notified as a change in de facto power. (Source:Acemoglu 2004) Non-stable monarchies followed during the Middle Ages. There have been several fierce contests for the Crown and several (continental) wars continued to cost England more money than it could afford. England soon lost all its French territories, apart from Bordeaux. The Battle of Bosworth (1485) with the victory of Henry VII resulted in a new royal house, the Tudors and England was to enter a new period of history. The rule of the Tudors, including Henry VIII, Mary and particularly Elizabeth brought in one of the most glorious eras of British history; exploration, colonization, Atlantic trade, victory in war, and growing world importance. Britain discovered and collected several overseas colonies in taking part in the conquest of the New World. (Source: UK information guide, British History). However the Tudor monarchs, followed by the first Stuart kings failed to build an absolutist monarchy, mostly because of the Parliament, which blocked attempts to concentrate power. In England the crown was not strong enough to control overseas trade (by for example creating monopolies) so individuals and small partnerships carried on most trade. There was not a high entry barrier into the merchant class and the interests of the merchants were now directly opposed to those of the kings. Namely, changing to more efficient economic institutions 4

would decrease their political power (losing also de jure power to the merchant class) (Source: Acemoglu 2004). This rule would have a central role in the political changes that would follow later on. Also the fact that the Stuart kings believed that they had a divine right to govern caused increasing resentment in the changing world. This brought him into a bigger conflict with the English Parliament. The struggle for supremacy between Parliament and the King (monarchy) led to Civil War in 1641. The parliament won and the king, Charles I was executed in 1649. Oliver Cromwell became head of state, which was the start of England's only period of dictatorship. Cromwell was unable to find anything to replace the monarchy and when he died, his son, Richard, became head of state. He was not a man to rule Britain and was not a popular choice. Parliament invited the son of the dead king to re-take the throne his father had died for. So Britain resumed a monarchy under Charles II in 1660. (Source: UK information guide, British History) The Civil War did have an influence on the great expansion of the British colonization and the trading merchant groups in the Atlantic. The so-called constitutional outcome in England that was settled because of the Civil War and later on the Glorious Revolution (1688) had enormous consequences. The de facto power was successfully used to reform political institutions so that de jure political power was acquired. This new form of government led to secure property rights, a favorable investment climate and rapid multiplied effects on other economic institutions, particularly financial markets. The change in balance of (de jure and de facto) political power in Britain led to a set of economic institutions favoring the interests of merchants. The merchants were able to control the powers of the monarchy in order to protect their property and trade opportunities. These changes in political power, political institutions and thus economic institutions seemed to be efficient, and resulted in an industrial advantage. Because of this, England moved ahead economically compared to the countries that moved further towards greater absolutism. (Source: Acemoglu 2004) Further changes in institutions after 1688 resulting in a Parliament that was in control of the fiscal policy. The English monarchy was now able to borrow huge amounts of money because the Parliament guaranteed that it would not default, which was crucial to the success of the English war machine. There also was a huge expansion of financial institutions and markets because of the greater security of property rights. Because of the huge successes of the changes in economic institutions, the Parliament became more and more successful which made its power permanent. (Source: Acemoglu 2004) Britain was fast becoming the crucible of the Industrial Revolution as steam power, steam trains, coalmines and waterpower began to transform the means of transport and production. The industrial revolution brought about a more urban society. New territories in Canada and India expanded the Empire, but there was also the loss of the American Colonies with the declaration of independence (1776). (Source: UK information guide, British History) Later on, the situation in Britain cannot be described as stable. For a period of 22 years, from 1793 to 1815 Britain waged war with France. In spite of the great power of the Britain Empire, the French troops controlled Europe. The grandiose ambitions self-crowned Emperor, Napoleon was finally defeated by the British sea power, under command of Wellington’s Peninsular army in 1814. (Source: UK information guide, British History) Britain, the dominant industrial and maritime power of the 19th century, played a leading role in developing parliamentary democracy. As the century and Industrial Revolution progressed, small elites were taking over the power in the parliamentary monarchy. The franchise was restricted to males with relatively large amounts of assets, income or wealth. The existing economic institutions, particular in the labor market, disadvantaged the poor and the organization of trade unions by workers was illegal. The poor, disenfranchised workers had

5

little de jure power and also limited de facto power. In later periods, this all caused collective actions by (disenfranchised) workers. These masses were relatively well organized and therefore difficult to repress by military actions or by making concessions to buy off opposition. Elites now faced a trade-off between losing political power and the prospect of, possibly radical, redistribution versus the risks of destroying assets and health and the threat of revolution. The de facto power of the disenfranchised had increased and because of the threats democracy was, in many ways, forced on the elites. This resulted in political and institutional changes. This was the first step to the future allocation of de jure power. The rise of democracy had begun. (Source: Acemoglu 2004) The First Reform act of 1832 removed some of the worst inequalities under the old electoral system. However the Reform act did not create mass democracy and is was rather designed by the elites as a strategic concession. The elites where still in electoral advantage and there was also evidence of continued corruption. The Second Reform Act (1867) and Third Reform Act (1884) increased the electorate and working class voters became the majority in all urban constituencies. The Redistribution Act of 1885 removed many remaining inequalities in the distribution of seats. From now on economic institutions also began to change. (Source: Acemoglu 2004) Civil services where opened to public examination. The introduction of a huge amount of labor market legislations favored workers position in relation with the industry. During 19061914, the Liberal Party introduced the modern redistribute state into Britain. The introduction of health and unemployment insurance, government financed pensions, minimum wages, and a commitment to distributive taxation resulted in more equal Britain after the 1870’s. Also the education system was extended and opened up for all classes. The People Act of 1918 gave the vote to all adult males and later on to all women (1928). These changes in economic institutions had a great, positive, influence on the British economy. (Source: Acemoglu 2004) The twentieth century also has seen Britain fight two world wars at considerable human and crippling economic cost. The first half of the 20th century saw a serious depletion of strength of Britain. The second half witnessed the dismantling of the Empire but Britain rebuild itself into a world leading, modern and prosperous European nation. (Source: UK information guide, British History)

Conclusions Overall, it can be concluded that there has been a clear shift in power that stared in 1832. In the beginning Britain was governed by the relatively rich. The de jure and de facto power was in hands of the rural aristocracies. However, under pressure by the threat of revolution, elites where forced to introduce economic and social changes, which changed the de facto power of the disenfranchised. Workers now also demanded political rights, which would allocate future political power to them. Changes in political institutions influenced the economic situation, which had led to a different distribution of resources. Changes in the labor market, in government policy, in the educational system reduced inequality over time. Mass democratization, which reallocates the durable de jure power away from the elites to the masses, was a fact. From now on also the poorer segments of society could vote to implement economic institutions and policies consistent with their own interests. A general result of the democratization is that the economic institutions changed radically in favor of those newly endowed with the de jure political power, mostly the relatively poor. (Souce: Acemoglu 2004)

6

Comparison Over a long period, Britain and France were ruled and invaded by several nations. In the Middle Ages several kings took over power. In the 17th century Louis XIII ruled France and Cardinal Richelieu made France from a feudal monarchy to an absolute monarchy. The power of the monarchy was now at his strongest level. This absolute monarchy remained under the reign of Louis XIV. Louis XIV changed a lot of economic institutions to stimulate and control trade, industries and colonization. The bureaucratization of Louis XIV started out well through stimulation of the France market and subsidizing the transport cost of French merchants. But the reign of Louis XIV also failed in several areas. When he died, the de facto power of the Bourgeoisie was rising. The commoners wanted equal rights and an abolition of the class system. In 1789 the crisis was at his highest and the lead to the French revolution. In Britain, the Tudor monarch was not able to create an absolute monarchy as in France. Mainly because the Parliament blocked attempts to concentrate power. The Crown wasn’t strong enough to control overseas colonies so individuals and small enterprises developed. Britain needed these enterprises, because it had a lot of inhabitants living oversees, while France had mainly slaves living in the oversees colonies. (Source: Engerman, Sokoloff) There were not high entry barriers to the merchant class, so the group of merchants grew rapidly. The Kings thought that they could keep power, despite of the growing power of the merchants. This led to a Civil War between parliament and the kings. After this war, Britain grew even faster because of the reform of the political and economic institutions, forced by the parliament. In France, in this period the absolute monarchy was still at its strongest. The French merchants had less rights then the merchants in Britain and so the France economy was not growing as fast as in Britain. While France was struggling and waiting for the revolution, Britain was changing enormously. This happened, because of the change of the distribution of de facto power. In Britain, the property rights became more secure through the huge expansion of financial institutions and the Industrial Revolution. Despite the great power of the Britain Empire, the French troops controlled Europe. The Revolution of 1789 in France ended when Napoleon entered Paris. Napoleon reformed France by changing the justice, educational, military systems. In France, during this period, 160 years later as in Britain, the economic institutions changed and the inhabitants got more secure property rights. But Britain remained economically stronger than France; despite of Napoleon started invading Europe and Northern Africa. These military campaigns cost France a lot of money. Napoleon conquered a large Empire, but got himself in trouble by fighting in Russia and later against England. In Britain, small elites were taking over power in the parliamentary monarchy when the Industrial Revolution progressed. The industrialization in combination with the presence of bad economic institutions for the poor, especially in the labor market, played a huge role in the change towards democracy. Upcoming, relatively well-organized, disenfranchised masses were gaining de facto power from the parliamentary monarchy that was mostly in the hands of a small number of elites. Reform acts led to a more equal distribution of seats, which caused a shift in political power (de facto and later on de jure) and changing economic institutions. This led to the democracy as there still is today. In France, after the Revolution, the Crown was restored but this could not be maintained, because of the pressure of the merchants and entrepreneurs, so a Republic was created. In 1848, France started growing towards a democracy. This democratization leads to a catch up with Britain. Trade and industry began to flourish. Despite of a war with Germany, the French economy grew fast and France became a democracy. 7

In the beginning of the 20th century, Britain and France started the Triple alliance with Russia. When Russia got involved in a war with Germany, France and Britain started the First World War. Until the end of the Second World War the growth was slowed down. The slowdown was the largest in France, because France was captured and Britain could trade with the United States. After the wars, both countries were characterized by economic growth, consumerism and technological advancement. De Gaulle centralized the power of the government in France even more. This gave the president a lot of power. In Britain, the government remained decentralized. Britain and France grew to become powerful countries within the European Union and in the world. Although the history of institutions is very different between France and Britain, they are almost equally rich nowadays. Different institutions caused similar incentives and similar growth. This does not mean that institutions are not important. Important for the growth in France and Britain was the Industrial Revolution. The important changes in institutions in France and Britain had been introduced before and in the beginning of the Industrial Revolution. At the beginning of the Revolution, Britain was a stable monarchy and the institutions helped the industry to develop. The Industrial Revolution in France started later on. It started at the time when France became a stable democracy. Thus, it is important for a country to have stable institutions. These institutions are important apparatus to help a country to develop. Overall, can we concluded that Britain moved ahead economically after the middle ages. In Britain, the presence of the Parliament and its increasing political influence resulted in a favorable trade climate and good economic institutions. France was still an absolute monarchy, while Britain starts to industrialize and democratize. It took until after the French Revolution and the Napoleon era for France to catch up. In this period France became a Republic. In France, from this period on, the good economic institutions also had been introduced. In the period of the two world wars, the growth in Britain and France had slowed down. After the Second World War, both countries grew rapidly. Britain remained a relatively decentralized ruled country, while France, under the influence of the Gaulle, had become more centralized. Nowadays, France and Britain are rich countries with respectively a GDP per capita of $27,600 and $ 27,700.

8

References Acemoglu, D., Johnson, S. and Robinson, J. "Institutions as the Fundamental Cause of LongRun Growth." NBER working paper 10481 (2004)-LVIII – nr. 1 Engerman, Stanley L. and Kenneth L. Sokoloff. “Factor Endowments, Institutions, and Differential Growth Paths among New World Economies” (1997) Great Britain - UK information guide, British History Britannia, British History by Hampton R. and Fox S. Frankrijk Bronnen: Bailey, R. 2000, Lonely Planet, 2001. Van Reemst. 1999. French History Timeline

9

2. Argentina vs. United States of America Introduction In the 16th century the Aztecs and Incas in the America’s were among the richest civilizations in the world, yet the nation states that now coincide within the boundaries of these empires are among the poorer societies of today. At the same time, the countries occupying the territories of the less-developed civilizations in (North) America are now among the very richest countries in the world. Acemoglu, Johnson, and Robinson (2004, p.21) claim that this “reversal of fortune” can be explained by the impact of European colonialism on the economic institutions in these countries. This paper will perform two case studies in order to verify whether this analysis is correct. Firstly it will examine the impact of colonization on Argentina, secondly on the United States of America, and finally it will come to the main conclusion that the fundamental cause of the differences in institutional developments between the Argentina and the United States is the historically used type of colonization. The initially established political institutions by the colonizers have been very important determinants for the following institutional developments in both Argentina and the United States.

Argentina 1. Introduction During the end of the 19th century and the beginning of the 20th century, Argentina had been an economically emerging country with exceptional comparative advantages. It was one of the largest and best- endowed countries in Latin America, with economic, social, political and institutional developments comparable to those of developed countries. Although, in the 16th century, the Spanish colonizers had not established very elaborate and favourable institutions in the country. Even though many countries experienced strong economic growth in the years after the second World War, Argentina has lagged far behind in the second half of the 20th century. (Source: Véganzonèz & Winegrad) In this case political and institutional developments in Argentina are discussed, which are, for a large part, responsible for these, rather exceptional, economic developments in Argentina. 2. Historical development of institutions A milestone in the history of Argentina was the colonization of the country by the Spanish. After the Inca’s, the powerful Spanish came to the county of Argentina in search of silver and gold. They established a set of institutions aimed at extracting resources rather than respecting the property rights of the majority. After the colonization by the Spanish and until the independence of Argentina by the first constitution in 1816, the political institutions were primarily consequences of shifts in de facto political power of internal or external authorities. These shifts were dependent on the changes in distribution of resources. Thus, because of the extractive form of colonization by the Spanish, there were no incentives for them to establish elaborate political and economic institutions. Nonetheless, the Spanish placed Buenos Aires on the map as an important mercantile town of the world. (Source : Van Reemst, van der Doef, Frank, Hotwijk & Thielen, 2001/2002)

10

At the time of the independence of Argentina there was a struggle for power between centralists and federalists. This initiated between the so-called ´Unitario´s´, supporters of central political authority, and the leaders of the provinces and landowners, ´gaucho´s, who did not want to submit to the central authority in Buenos Aires. Internal physical struggle between federalists and centralist ultimately leaded to the victory of Urquiza, a supporter of central power, over the federalists. The party with most de facto power thus saw the opportunity of altering the political institutions in her favor and thereby gained more de jure political power; in 1853 a new constitution and a federal presidential system were founded and Urquiza became the first president of Argentina. The president gained an enormous amount of power, because he became the political leader of the country and he gained the right to veto. Under the new political institutions, economic institutions were improved during the 19th century and the beginning of the 20th century; the educational system was improved, there was state-induced industrialization and the economy was modernized. During this period of economic prosperity, Argentina was pretty much commercially integrated with Europe. This was an important reason why, at that time, the country was experiencing rapid economic growth, like in western countries. Paradoxically, the big landowners, financial elites and most important exporters actually had most of the power during these times of economic prosperity. Although formally now there was a democratic system with elections, the reality was that the power remained in the hands of the elite who largely ran the country in the interests of the landowners.

11

External shocks like wars and economic crises had leaded to new ideas and ideologies and thereby to collective insubordination among the middle class, which lead to transformations in the distribution of de facto political power. Moreover, The Industrial Revolution formed the ground root to the emergence of political movements. This had also caused gradual changes in de jure political power, in the form of movements towards a formally social democracy by Juan Péron. Péron, an absolutist dictator, had gained a lot of support among the people during World War ΙΙ. Even though he was arrested in 1945, the protesting Argentinean people managed to get him back in power. This way he became president and regained de jure political power. With this power he managed to restructure the economic institutions in Argentina; he improved the social structure of the economy, encouraged investments and nationalized all big companies and actually the entire economy. Economic welfare increased and there was increasing support for Péron. Nonetheless, Péron was a very authoritarian leader, who was not afraid to use violence. The established unions also didn’t have much power at that time, because of the direct guardianship of the government. When the economy had to suffer decay in the fifties, Péron decided to resign under the political pressure of the army, the church and the elite. The years between 1955 en 1973 can be characterized by political chaos, with many coups d´état and takeovers. Though a new president had prohibited all political parties with the idea to create political rest, the support from the Argentinean population to Péron had been very extensive. So extensive that the new president was forced to recognize his party again. Thus, because of the massive support of Péron, his implicit de facto political power, the newly created political institutions were not strong enough to prevent Péron from having political influence and regaining a certain amount of de jure political power. However, his come back had not been much of a success, due to the problems of political chaos. Between 1976 and 1983 a military authority under the leadership of general Videla took control. His leadership was one of military suppression. This type of extorting bad political institutions, like bureaucratic politics, lead to an at least equally bad economic situation. This severe economic situation was also for a large part caused by the restructuring of economic institutions under the Videla administrations towards more liberalization, which caused a huge increase in public debt and in inflation. Due to miserable economic circumstances, bad political institutions and military pressure from remainders of the Videla era, promising president Raúl Alfonsín could not improve the situation during his leadership. Carlos Menem, winner of the following elections improved economic institutions; he limited inflation, reallocated public savings and stimulated free (international) trade, partly through approaches to the IMF and the World Bank for help. His economic liberalization was actually more successful. However, he gained de facto political power by favoring the powerful elites and abused this power by appointing friends for political positions and by reforming the political institutions in his favor by introducing the possibility of obtaining a second term of office. Using the theory of Acemoglu (Source : Acemoglu 2004): although he improved certain economic institutions, Menem abused his political power to obtain more power and to be able to reform the political institutions of Argentina in his favor. A combination of formal political democracy and stagnating social democracy caused existing institutions of political democracy to lose credibility. (Source : Buve 2004) Although the intentions of Menem might not always had been the right ones, he managed to successfully reform economic institutions after many years of economic decay. After other presidents have had a difficult time under economic decay, Nestor Kircher has become the current president of Argentina. In 2002, Argentina faced an enormous financial crisis; to high a public debt and an unsustainable one-to-one exchange rate of the peso with the American dollar, resulting in an overvaluation of the peso.

12

These problems lead to very serious credibility problems in the Argentinean economy. (Source : Van Reemst, van der Doef, Frank, Hotwijk & Thielen, 2001/2002) Partly because of a serious political crisis, in 2002, the Argentinean government has finally dropped the socalled currency board, an important cause of these current economic problems. The reason why Argentina had not already dropped this currency board has something to do with political and institutional impediments. A fundamental cause is again a struggle of power between the central and decentralized governments, already referred to in the first part of this chapter. Two main problems have arisen in Argentina. In the first place the system of a combination of taxation by the central government and spending of the provincial governments is inefficient, because the system causes free rider and moral hazard behavior of the provincial governments due to weak budget constraints. This has created such a large public debt. Secondly both central and provincial governments do not want to lose their formal de jure political power and that is why a complex and unclear system of regulation is implemented; there are no incentives for both central and provincial governments to create an institution to supervise their implementation of established agreements. This is why the system lacks decisiveness to make certain important decisions, like the decision to drop the currency board. 3. Conclusions The colonization of the country of Argentina by the Spanish was one of extraction of natural resources and without the establishment of clear political institutions. This is why the initial political institutions (at the time of the colonization) were not very elaborate and didn’t secure property rights very well. These initial institutions caused much of political instability and shifts and de facto and instability between de jure and de facto political power. De facto political power has generally been in the hands of elite groups like landowners, while de jure political power was executed by different, mostly authoritarian, political leaders. This asymmetry of powers has led to multiple coups d’états and other takeovers of de jure political power. Authoritarian political leaders have, several times, changed political institutions in their favor. They just did not want to loose power. They have also influenced economic institutions, like Some leaders, like Urquiza, Péron and Menem have succeeded in improving the institutions. They managed to secure property rights, to modernize the economy and to improve labour marktet conditions. Others like Vidéla have only worsened them. Hid librtation policy has lead to increasing public debt and inflation. The Industrial Revolution caused the need of social economic and social political developments in the western countries, like the emergence of unions and free elections. However, the, often self-enriching, political regimes in Argentina have never really succeeded in encouraging these developments. A severe lack of social democracy and political struggles between central governments and powerful elites, have led to a serious lack of decisiveness within the Argentinean government. This lack of decisiveness has led to many economic difficulties during the second half of the 20th century. The conclusions are summarised in Figure 2.1, derived from the model of Acemoglu. (Acemoglu, 2004)

13

Figure 2.1

United States of America 1. History In 1492 Columbus, sponsored by the Spanish King to find a direct western all-sea route to Asia, accidentally discovered the America’s. Shortly after this all major European powers (France, England, the Netherlands, but most notably Portugal and Spain) founded settlements in the New World. At the end of the 16th century, with its overwhelming military and maritime power, Spain had conquered most of South and Central America and large parts of North America (Mexico). These countries contained an enormous wealth in natural resources and were rather densely populated. In its conquest for gold and silver the Spanish rulers established a set of institutions aimed at extracting resources rather than respecting the property rights of the majority. During the first half of the 16th century the other European powers had only played a role in the margin. England and the Netherlands, however, were rapidly gaining economic and military strength and began to challenge the Spanish supremacy. Especially England began to claim territory in the New World and in response to this, but also to punish England from wandering from the Catholic path, did Philip II send forth his naval “Armada” in the year of 1588. As noted by William Elson (1904, Ch.5) few events in history have been more far reaching than the destruction of the Spanish Armada. It marked the end of the Spanish dominion of the sea, and, where Spain had conquered Mexico and most of South and Central America, the British became the colonizers of the present-day United States of America. 2. Early Institutions Because the native population of North America was small in number, relatively lowdeveloped, and not endowed with large quantities of silver and gold the British had little to gain from a set of institutions aimed at extracting resources and plundering the natives. Instead the British founded (in total 13) self-supporting colonies along the Atlantic Coast and, other than the operation of the Navigation Acts to regulate trade, allowed the colonies to go their own way. These colonies were:

14

Massachusetts New Hampshire Connecticut Rhode Island New York, Pennsylvania New Jersey Delaware Maryland Virginia North Carolina South Carolina Georgia

THE NEW ENGLAND COLONIES

THE MIDDLE COLONIES

THE SOUTHERN COLONIES

Source: Bloy (2004)

In these colonies only partial British rule applied. “Each of the thirteen separate colonies had a Governor appointed by the Crown. He was responsible only for trade and defense. The daily administration -by-laws, internal taxation and so on - was in the hands of local, very democratically elected colonial assemblies. Each of these had two Houses: the Assembly (the colonial equivalent to the House of Commons) and Council (the equivalent to the House of Lords). The Council was made up of important persons in each colony.” (Source: Bloy, 2004) A direct consequence of the political institutions as set forward by the British was that a broad cross-section of colonists (with significant investment opportunities) was in charge of establishing the economic institutions in the colonies. The resulting economic institutions, as a consequence, ensured broad-based property rights with rather effective enforcement, relatively good access to economic resources, and strong entrepreneurial incentives. Basically, the foundation of the economic success of the British colonies was already present in its earliest political and economic institutions. 3. War of Independence In 1763 England had won the Seven Year’s War against France and gained full control over Canada. Suddenly the size of British territory in North America had doubled which put a severe strain on the Royal treasury. To increase revenues Britain started to tighten its control over the 13 colonies and impose various regulations and taxes. During the 17th and 18th century, however, the 13 British colonies had grown vastly in economic strength and become highly independent from Britain. Not only did the colonists denounce the heavy taxation, they passionately resisted any reduction in self-governance. “The conflict escalated and King George III issued a proclamation on August 23, 1775, declaring the colonies to be in a state of rebellion. On July 4, 1776, the Continental Congress adopted a Declaration of Independence. Armed conflict between America and England lasted until 1783. Known as the Treaty of Paris, the peace settlement acknowledged the independence, freedom and sovereignty of the 13 former colonies, now states, to which Great Britain granted the territory west to the Mississippi River, north to Canada and south to Florida, which was returned to Spain. The 13 colonies were now "free and united independent states" - but not yet one united nation. The success of the Revolution gave Americans the opportunity to give legal form to their ideals as expressed in the Declaration of Independence, and to remedy some of their grievances through state constitutions. As early as May 10, 1776, Congress had passed a 15

resolution advising the colonies to form new governments. On a national level, the "Articles of Confederation and Perpetual Union" produced by John Dickinson in 1776, were adopted by the Continental Congress in November 1777, and they went into effect in 1781. The governmental framework established by the Articles had many weaknesses, for example the national government lacked the authority to set up tariffs, to regulate commerce and to levy taxes. It lacked sole control of international relations: a number of states had begun their own negotiations with foreign countries. Nine states had organized their own armies, and several had their own navies. In May 1787, a convention met in Philadelphia to draft a new Constitution which established a stronger federal government empowered to collect taxes, conduct diplomacy, maintain armed forces and regulate foreign trade and commerce among the states. The Constitution divides the government into three branches, each separate and distinct from one another. The powers given to each are delicately balanced by the powers of the other two; and each branch serves as a check on potential excesses of the others.” (Source : History of the United States, 2004) So, without the interference of Britain, the economic framework of the United States became a little bit more favorable towards long-term economic growth. But as the colonies already had quite “good” economic institutions, and the involvement of Britain in its North American colonies had never been large, the main improvements were indeed in the political sphere. The new political system placed strong checks on the legislative, executive, and judicial branch of government1, preventing those in power from creating a set of economic institutions that are beneficial for themselves yet detrimental for the rest of society. This not only ensured the future survival of the “good” economic institutions but also led to a stable and trustworthy government, both of course in favor of economic growth. Figure 3.1 captures the events of the War of Independence in the framework as proposed by Acemoglu, Johnson, and Robinson (2004) The political institutions before the War of Independence had been introduced by the British and divided de jure political power between the governor (trade and defense) and the colonists themselves (responsible for everything else). The distribution of resources, and consequently the de facto political power in the colonies, was absolutely in favor of the colonies. Yet the British had one very important trump: they provided the much needed protection against the French. In return for this protection the colonies paid taxes to the Crown and there was a balance in the system. As mentioned before, this delicate balance was disturbed when the British defeated the French and the colonies became substantially less dependent on the U.K. for their military protection. At the same time, the U.K. started to use its de jure political power to a) further increase its de jure political power at the expense of the power of the colonies and b) substantially raise the tax burden on the colonies. These “external shocks,” as Acemoglu, Johnson, and Robinson (2004, p.6) would refer to them, modified the balance of power -and gave rise to such a discontent with the British among the colonists- that they eventually resulted in the War of Independence.

1

Noteworthy is that even today the political system of the United States is still essentially the same as it was set forth in the constitution of 1787!

16

Figure 3.1

After the War of Independence, as can be seen in figure 3.1, the distribution of resources was of course not entirely the same as before the war. The important changes in de facto political power however were mostly the result of the diminution of British military power in the former colonies. With the British gone the confederate colonies were able to introduce a completely new set of political institutions which, as mentioned earlier, placed strong checks on the legislative, executive, and judicial branch of government. This resulted in a stable and trustworthy government and, as was also mentioned before, in an economic framework that became even more favorable towards long-term economic growth. 4. Civil War When slavery was introduced in North America in the 17th century it soon died in the New England and Middle colonies where it did not work for a society based on small farms and shops. But it flourished in the Southern colonies where the climate facilitated large plantations which required large amounts of workers. Before (and during) the American War of Independence this imposed no problem, as the differences between the North and South were of minor importance compared to their common interest in opposing Britain. After the War of Independence, differences between the North and South were worked out through compromises. During the course of the 19th century, however, when the industrial revolution caught momentum in the North and the balance of economic and political power began to shift, slavery eventually led to the largest conflict in the history of the United States of America. Of major importance was that the southern states remained predominantly agricultural with an economic and social system highly dependent on plantations and slavery. While slavery for them had become an economic institution distanced from theoretical and ideological issues, public opinion in the North began to turn against the concept of human slavery. “But even as the need to protect [the institution of slavery] grew, the ability, or at least the perceived ability of the South to do so was waning. Southern leaders grew progressively more sensitive to this condition. In 1800 half of the population of the United States had lived in the South. But by 1850 only a third lived there and the disparity continued to widen. While northern industrial opportunity attracted scores of immigrants from Europe in search of freedom the South's population stagnated. Even as slave states were added to the Union to balance the number of free ones, the South found that its representatives in the House had been overwhelmed by the 17

North’s explosive growth. More and more emphasis was now placed on maintaining parity in the Senate. Failing this, the paranoid theory went, the South would find itself at the mercy of a government in which it no longer had an effective voice.” (Source : Harrison, 2004) Clearly the North had a distinct advantage in its ability to produce soldiers and supplies and thus to win the war, yet the South’s fear to lose their way of life was big enough to secede from the Union and form their own confederacy and risk going to war. As President Lincoln was determined to keep the Union together and did not allow the Confederate States to secede, the conflict eventually led to the American Civil War of 1861-1865. Figure 4.1 again captures the events leading to the U.S. Civil war in the framework as proposed by Acemoglu, Johnson, and Robinson (2004). The basic shock that can be identified is the industrial revolution, which -over time- shifted the balance of both de facto and de jure political power towards the Northern states. At the same time public opinion in the North starts to turn against the single most important economic institution of the Southern states: human slavery. The South feels more and more threatened by the North, culminating in the seceding of eleven southern states from the union. As mentioned before, President Lincoln did not allow them to form an independent nation which is why proceedings eventually culminated in the U.S. Civil War. Figure 4.1

In retrospect it is clear that the war was won by the North and that the political situation returned to the pre-war situation. The Southern states re-joined the U.S.A. and were forced to abolish the (economic) institution of slavery. Although a major event in America history with important social implications, the Civil War only led to minor changes in the economic institutions. In conclusion, the Civil War is a good example to demonstrate the relevance of the framework of Acemoglu, Johnson, and Robinson (2004) but the event itself did not significantly affect the long-term economic growth perspectives of the U.S.A.

18

3. Conclusions The colonization of the America was not one of extraction and plundering natives, because the country did not have many natural resources of interest. Instead the English settled in the country and decided to create economic institutions that ensured broad-based property rights with rather effective enforcement, relatively good access to economic resources, and strong entrepreneurial incentives. Because of fairly democratic political institutions, de facto political power was, very soon, in the hands of the population of the thirteen separate colonies. The de jure power was for a small part in the hands of Governors appointed by the Crown. However, most de jure power was given to the very democratically elected colonial assemblies. The established political and economic institutions had lead to economic welfare and strong autonomy, especially in the northern states. These favorable economic developments in the northern states, facilitated by the Industrial Revolution, caused a large immigration flow into the northern states. This immigration inflow leaded to a shift property rights in favor of the northern states. The de jure power of these states in the House increased and these states finally overpowered the southern states. As already mentioned, the Civil War is a good example to show the relevance of the framework of Acemoglu, Johnson, and Robinson (2004) but in itself, the event did not significantly affect the long-term economic growth perspectives of the U.S.A

Comparison Argentina had been developing very well during the second part of the nineteenth century, but it has been by far outpaced by the United States since World War II. A fundamental cause for this development is the difference in colonization between the two countries. Generally, in Argentina natural resources were extracted and not much attention was paid by the colonists to establishing stable economic and political institutions. In contrary, in the United States, the English decided to found good institutions, because they intended to establish themselves in this country. The main consequence for the United States was real and stable democracy. Instead, in Argentina there was formal but no real democracy. Instead there have been authoritarian regimes and very powerful self-enriching elites. Moreover there has been instability in political and economic institutions and in de jure political power. Before World War II, Argentina had been able to develop pretty well, because of large resources of natural resources and intensive relationships with other western countries. However, after World War II, developments in Argentina completely turned around and years of political chaos followed. Mainly because of the lack of good institutions, Argentina faced troubles with the arrival of the Industrial Revolution and the trend of liberalization. On contrary, in the United States, especially the northern states, prospered very well. The de facto and de jure increase of power of the northern states created the possibility for them to further improve institutions. For example, slavery was completely abolished. Nowadays, Argentina still faces the problem of vague political institutions and the fear of political leaders to loose political power. The distribution of power between central and decentralized governments is unclear and this causes a serious lack of decisiveness. On contrary, the United States have only gradually improved their stable political and economic institutions. Moreover, nowadays, the political system of the United States is still essentially the same as it was set forth in the constitution of 1787.

19

References Acemoglu, D., Johson, S. and Robinson, J. "Institutions as the Fundamental Cause of LongRun Growth." NBER working paper 10481 (2004) Arda, Atilla. and Edwin Lambrechts. “Institutionele problemen achter de crisis in Argentinië”, Tijdschrift voor Politieke Ekonomie 2003 derde jaargang 24(4) 3-15 Buve, Raymond. “Democratie in Latijns- Amerika: voor vrienden de regering, voor vijanden de wet”. Internationale Spectator januari 2004 Jonker, Marcel. The economic crisis of Argentina; needed: a government and a plan, final paper Véganzonèz, Marie-Ange and Carlos Winograd. Argentina in the 20th century: an account of long-awaited growth. OECD, 1997 Argentinië. Bronnen: Van Reemst, 2001, Doef, P. van der 2001, Frank, N. 2002, Holtwijk, I. 2001, Thielen, J. 2002 The Origins of the American War of Independence. Bloy, M. Oct. 4th 2004 History of the United States - Revolutionary Period and New Nation. American Information Resource Center (AIRC), U.S. Embassy in Warsaw & U.S. Consulate General in Krakow, Poland. Oct. 4th 2004 The American Civil War – The Causes. Harrison, T. Oct. 4th 2004

20

II. 1 (G62) Investment policies

Jeffrey Oudeman (analyst) Marcel Mulken (statistician) Jasper van Hekken (manager) Paul Gille (analyst)

EDUCATION POLICIES Human capital is generally seen as a key factor that contributes to high levels of income and economic growth. Human capital is the result of education and on-the-job training, but also results from experience and learning externalities. The OECD provides the following definition of human capital: The knowledge, skills, competencies and attributes embodied in individuals that facilitate the creation of personal, social and economic well-being (OECD,2001). This paper is about what should be done by policy-makers to make sure education is arranged in the best way. So how can education contribute best to economic growth? First we are going to see why people chose to go to school and we will check whether they go to school enough; are individual returns to education high enough to make sure people take enough education or should people be stimulated to go to school more because positive externalities exist? In the second part of this paper we will compare the USA and Europe. In the US education is more general and concept-based, while in most European countries (Germany) education is more skillspecific. The question is: What is better for economic growth? Should we invest in general education or skill-specific training?

1. Education and its effect on economic growth The European Union (E.U.) wants to be by 2010 the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion (European Council, Lisbon 2000). Also in the Netherlands politicians are talking a lot about our knowledge-based economy and about how to be more innovative and increase our human capital (which encompasses characteristics like education, work experience and health) etc. One of the political parties, which is now one of the political parties in office, D66, even made education their main topic during the election campaign last year. At the presidential election in the US last year, education policies played a very important role. 1.1 Growth – education links Education has many direct and indirect effects (& externalities) both on micro and macro level. The figure below gives an overview of these effects.

Source: Dahlin, B.G. , “The Impact of Education on Economic Growth,” Duke University We like to focus on the macro level and the effect of two kinds of education on economic growth. Based on the above diagram (of course with underlying assumptions and the scientific proof for the effects as presented) we can say that education causes higher economic growth in four ways: 1. Increased earnings (higher productivity) 2. Increased earnings of neighbours (learning through observation) 3. Higher participation in the labour force which results in a increased labour force 4. Lower population growth and better health of population (and labour force) The first one of the above list is the most important, so education causing higher productivity and thus higher economic growth, that’s what we will review and discuss. While it might seem obvious that there is a link between education or human capital on the one hand and a nation’s growth rate of GDP on the other hand, it is quite difficult to find statistical evidence for such a link; some empirical studies even find the link to be insignificant. Imperfect techniques to measure the education of individuals and the aggregate human capital of an economy might be among the reasons why this link turns out to be insignificant in some empirical studies. Disagreement in the results of those empirical studies sometimes does arise because of different measures of education and different definitions of human capital. For example the number of years spent on school, the quality of the schooling, the nature of the curriculum and the student’s effort are all components, which should be included in an ideal measure of an individual’s education. Still we think economic literature as well as empirical studies generally supports the view in which education and human capital are positively related to the growth rate of GDP. 1.2 Education and its returns The idea of positive educational externalities is that the benefits of individually acquired education may not be restricted to the individual but might spill over to others as well - that is, to other individuals in the same industry, city, region or economy. Indeed if social returns at the macro level exceed the private returns on the micro level, increased public support for education is justified. There is compelling evidence that human capital increases productivity, suggesting that education really is productivity-enhancing rather than just a device that individuals use to signal their level of ability to

the employer.1 Whether there is insufficient investment in human capital, however, is subject to debate. We can distinguish between the return of education for individuals and for the society as a whole. When we look at individuals we mostly look at the earnings. People prefer a high wage and therefore are willing to invest in their skills and capabilities. The question is how much they benefit from more schooling and at what cost. It is also the case that education stimulates not only individuals but also shift outward to the public, so education also has a positive effect on others. It is obvious that when we stimulate education, people get more “human capital” and therefore become more skilled. When people become more skilled they can stimulate a countries economy and also earn more individually. But when we look at the earnings only, 40% of the variation is explained by measures such as educational qualifications, literacy and work experience. When we look at how human capital can affect earnings we find that earnings depend on: • years of education • skills that do not always come from more education, but from the other sources (like background), and • factors that are not quantified (like race) The standard approach to explain the variations in wages, where the explanatory variables include years of schooling, either age or a simple proxy for experience, and other characteristics, is based on the work of Mincer. The formula is: Ln w = α + β0S+ β1E+ β2E2 This equation gives a relation between the logarithm of wages (w) to the years of schooling (S) and labour experience (E). It has to be said that the evidence that earnings are positively associated with schooling is robust and uncontroversial. A problem is that, because of a lack of suitable data, important data that probably are correlated with both schooling and earnings has to be included. It also seems probable that the costs and benefits of education vary across individuals. But when we look at the figure below we can see that there is a relation between schooling and wage.

1

“The returns to education: A review of the empirical macro-economic literature” Sianesi and van Reenen

We can say that not all earnings variation is due to human capital. People vary in the extent to which they trade of earnings against other job factors, like job-satisfaction and working hours. It’s also that in real life we haven’t got a perfectly working market, so individual earnings is not determined entirely by each person’s productive capacity and the level of individual investment. Factors like discrimination on grounds of race and gender or shocks in technology are mostly beyond control of the individual wage earner. 1.3 Government intervention We can ask ourselves if education should be stimulated and, if so, in what way? What are for example the roles of schools, family and social environment? What kind of intervention could help to improve education and thereby the competitiveness of a country. Parent have an important influence on education. In many well-development countries, young people whose parents have completed some tertiary education, are about twice as likely to participate in tertiary education as those of parents that haven’t got such a high schooling level. Well-educated parents can stimulate their children in different ways. They have skills which can be thought to them and also have a better ability to pay for their children compared to less educated (wealthy) people. It has to be said that this connection between parental education and skills is a modest one. ‘Rich children’ can have better opportunities but this won’t say that therefore they become more skilled. Another aspect worth noting when we look at intervention is the social environment. Not only parents play a role in decisions but also other family members, colleagues and friends. Beside that you have people who are trusted and well informed about career and educational opportunities and there are organizations who provided trainings to make people more skilled. Access to these ‘instruments’ will depend on individual skills as well as the existence of networks. The benefits of guidance can be seen individually through better learning strategies, career decisions and greater satisfaction. The collective benefits could include better-targeted human capital investment and better job and skill matching which can lead to a higher output. Hence, for children with uneducated parents or those in an education-adverse social environment, the private return to schooling is probably lower than the social return. Intervention is justified, for example in the form of enhancing access to schools targeted at special groups.

2. Two types of education and expenditures on it in US and EU Skill-specific or general education: which should we favour? We distinguish two types of education, namely a specialized, skill-specific, vocational education and a concept-based, general education. Europe (we will focus on the EU) has a focus on specialized, skillspecific, vocational education and fostering it at the upper-secondary (secondary schooling from the age of 14) and tertiary level. While the opposite is true for the United States, as we can see in Table 1. Table 1: Education Indicators % Upper % Upper University NonUniversity NonCountry Secondary Secondary Net Entry University Tertiary University in General in Vocational Rate Tertiary Attainment Tertiary Education Education Attainment Return France 47 53 33 8 11 18 Germany 23 77 27 10 13 17 Italy 28 72 8 Netherlands 30 70 34 22 EU 42.4 57.6 US 52 8 25 9 Source: Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER

University Tertiary Return 14 11 10 11 13

The educational system in Europe is known as rigid and inflexible relative to the system in the United States. Another difference is that the channelling of students into either vocational or general

education starts earlier in Europe. While in the United States there isn’t even a separate stream for vocational education at the upper secondary level, in the EU 57.6% of the students were enrolled in the vocational stream. General education is primarily imparted at the universities and at that level we see the United States have the highest net entry rate and the highest tertiary attainment from the countries included in table 1. Of course this is reflected in the differences we see in education expenditures, as described in table 2, between the U.S. and EU-member states. The percentage of GDP spend on primary and secondary education (in 1998) are comparable in the U.S. and the European countries. But we see (relative) expenditures on university tertiary education are much higher in the United States, for example 2.3% of their GDP goes to university tertiary education, while in Germany this is only 1%. On the other hand (relative) expenditures on vocational tertiary education are larger in Europe. Per student figures reflect this as well, for example in PPP dollars, in the US $19,802 was spend per student on the university tertiary education, while in Germany it was only $10,139 in 1997. Thus we may say education expenditures in the US and in Europe do explain the differences as described in table 1. Table 2: Education Expenditures Expend. Expend. Country /GDP /GDP Primary+ Vocational Secondary Tertiary

Expend. /GDP University Tertiary

Expend. per student Vocational Tertiary

Expend. per student as % of per capita GDP Vocational Tertiary 36 48 36 31

Expend. per student University Tertiary

Expend. per student as % of per capita GDP University Tertiary France 4.4 0.3 0.9 7,636 7,113 34 Germany 3.7 0.4 1 10,924 10,139 44 Italy 3.5 0.1 0.8 6,283 6,295 28 Netherlands 3.1 1.2 7,592 10,796 44 US 3.7 2.3 19,802 61 Source: Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER 2.1 Growth differences between the US and the EU We have just seen the differences in education between the US and the EU. Our choice is to see whether we can determine if those differences in education are one of the reasons for the growth differences between the US and the EU. Table 3 gives us information about the growth rates of Real GDP per capita. Table 3: Growth Rates of Real GDP per Capita Country 1970-1980 1980-1990 1990-1998 1999 US 2.1% 2.3% 2.0% 3.2% Germany 2.6% 2.0% 1.0% 1.4% Italy 3.1% 2.2% 1.2% 1.3% France 2.7% 1.8% 0.9% 2.5% UK 1.8% 2.5% 1.7% 1.7% Source: Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER We see that during the 1970s many European countries grew at a faster rate then the United States. But the general picture shows us the growth in European countries slowed down, resulting in quite a growth gap between the United States and the European countries during the 1990s and for sure during the ICT-revolution (Information and Communication Technology). We know that even in 2004 there is a huge growth gap between the US and EU, even after the bursting of the ‘ICT-bubble’ during

2000/2001. Of course it’s clear that this growth gap has many reasons and causes, but the differences in education could indeed be one of them. While European focus on vocational education was appropriate during the technologically tranquil times of the 1950s an 1960s, this focus may have also resulted in European technology adoption and economic growth being impeded from the 1980s onwards. To see if the above view is correct we take a closer look at the growth rates in the manufacturing sector. The key hypothesis we want to investigate is the following: Vocational education enables workers to very productively operate established technologies, while general education enables workers to adapt more easily to new technologies. So technology adoption is faster in a country with more generally educated workers, which should result in higher productivity (and growth rates). When we review growth rates in output per hour (productivity) we expect the US to have higher growth rates compared to the European countries. The manufacturing sector is most likely the sector where we could see best if our expectations are right. Table 4: Growth Rates in Output per Hour, Manufacturing 1978-1984 1985-1991 1992-2002 Country US 2.9% 2.4% 4.3% Germany 2.4% 2.2% 2.7% Italy 3.8% 1.7% 1.6% France 4.8% 3.6% 4.2% UK 3.5% 4.6% 2.9% Source: Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER Table 4 shows us that the growth rates in the US are indeed (much) higher during the 1990s (except for France), while in the early 1980s the EU still has higher growth rates. So we see a shift during those two decades from lower growth rates in the manufacturing sector in the US, to much higher growth rates compared to the ones of the EU. A gap occurs from the 1980s on and it seems to get bigger over the years. When we take into account that the gap between the US and the EU is even larger when we examine technology-driven industries (8.3% average annual productivity increase in the 1990s in the US compared to 3.5% in the same industries in the EU), we have a strong indication that our assumption and theory is right. Pharmaceuticals, office machinery and computers, motor vehicles, aircraft and spacecraft are a few of the industries classified as technology-driven industries. Generally we can say about these industries that (new) technology and (product) innovation play a very important, crucial role. Firms in these industries are very much dependent on technology and innovation and they can’t afford it to lack (too much) behind of competing firms. A good example is the manufacturing industry of computers, with several very big companies (like IBM, Dell, Packard Bell and Compaq), who every day need to improve their computers to keep being competitive. Because indeed during the last 25 years we have been experiencing an increased growth rate of technological progress with several new technologies. Which according to our assumptions and theory should result in higher growth rates (of productivity) for countries like the US, who have relatively more general educated workers compared to countries like Germany, who relatively have less general educated workers. Several economists say Europe has suffered from a “technology deficit” relative to the US. And there is lots of direct evidence that this is indeed the case. Just one example where Europe lags behind the US in the usage of new technology is the ICT capital. Table 5 shows us the ICT contribution to output growth. Table 5: ICT Contribution to Output Growth (%Points) 1980-1985 1985-1990 Country US 0.28 0.34 Germany 0.12 0.17 Italy 0.13 0.18

1990-1996 0.42 0.19 0.21

Source: Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER The US, Germany as well as Italy have been experiencing an increasing contribution of ICT to output growth, but we see that in the US the ICT contribution rose much faster. So again we see the gap between US and the European countries is only increasing. Once more we of course have to realize and emphasize that the different types of education is not the only explanation for the growth differences between EU and the U.S. over the years. It is also not our aim to review those other (possible) explanations, like the introduction of the computer, the oil crisis and the economic policy from the governments (the current big deficits from the U.S. and tax cuts from the recent years for example). But we do want to briefly mention some more of the evidence and theoretical and quantitative findings found by for example D. Krueger about our subject. In economic literature and economic scientific research is found much evidence to state that education indeed helps to cope with technical change, one of the main and important assumptions when we review the role of the two different types of education. The same we can say about the evidence for increase in growth rate of embodied technological change. This is important because D. Krueger argues that ‘the emergence of a gap between the US and Europe since the 1980s is related to the almost concurrent increase in the rate of technological change’. He constructed a model to test the effect of the different types of education and education policies between the US and EU. His model suggests that ‘while European education policies that favor specialized, vocational education may have worked well during the 1960s and 1970s when technologies were more stable, they may have contributed to slow growth and increased the European growth gap relative to the US during the information age of the 1080s and 1990s when new technologies emerged at a more rapid pace.’ And his quantitative analysis with a calibrated version of his model shows that the role of education may be significant. One of the examples shows 0.6 percentage point of the 1 percentage point growth gap between the US and EU in the 1990s is explained (in his model of course) by differences in education policy, far more than the product and labor market regulations do explain this growth gap. 2.3 Economic theory, scientific research and evidence Of course over the years there is a lot of scientific research been done on education from an economic point of view. Education (human capital) has a prominent place in economic theories. Many models have been developed within the last decades, also especially about education and its effects on economic growth. Examples are the “macro-Mincer”, and new growth theory. There are two main approaches in how human capital with or without externalities is introduced in new growth theory: • The incorporation of human capital as a factor input, for example by adapting the Solow model. • Explaining the process of knowledge accumulation by relating it directly to human capital accumulation, or indirectly via research and development (R&D) activity. It’s beyond our goal to discuss those kinds of models into detail, but it is interested to see if the conclusions they draw are similar with what we found out so far. Dirk Krueger indeed argues that ‘the recent growth gap between the US and Europe may be partially explained by Europe’s stronger focus on vocational education, compared to the US.’ And he says ‘If our hypothesis is valid, education reform towards more general education in Europe may have beneficial consequences for technology adoption and economic growth. There are signs that such reforms are underway’. More generally Brian G. Dahlin says: ‘Although the empirical evidence is difficult to interpret, macroeconomists in the last decade have stressed the importance of human capital’s contribution to growth: The main engine of growth is the accumulation of human capital – or knowledge- and the main source of differences in living standards among nations is a difference in human capital.’ One thing is sure education is a topic of ongoing research. 3. Conclusion We have seen that education has direct and indirect effects (externalities) on both micro and macro level. We focussed on higher productivity on macro level and the effect caused by two different types

of education, namely specialized, skill-specific, vocational education and a concept-based, general education. Education indicators showed us that Europe has a far more vocational trained labour force, while the US labour force is more generally educated. Education expenditures (government policy) do reflect this picture and of course are mainly causing it as well. Since the end of the 1980s the United States had higher economic growth compared to key member states of the EU, like Germany. The growth gap increased during the 1990s and we see a quite similar picture when we look at the growth rates in output per hour in the manufacturing sector. This leads us to the conclusion that one of the causes of these growth gaps between the US and the EU is indeed the differences in education. The technological adoption in the US is (much) higher than in the EU, so during the past two and a half decade with an increased growth rate of technological progress with several new technologies. Economists talk about the “technology deficit” from the EU relative to the US, which we endorsed by giving the example of ICT contribution to output growth. Also with respect to ICT contribution we see the gap between the US and Europe has been increasing. So education reform to more general education in Europe may have beneficial consequences for technology adoption and economic growth. Theoretical and quantitative findings from D. Krueger do support the above. Despite disagreement about definitions and measurements of education (and human capital) and imperfect techniques to measure it, economists have done a lot of research on education and developed several theories. Empirical evidence is quite hard to obtain and interpret and some empirical studies even find the link between education & human capital and a nation’s growth rate of GDP to be insignificant, still everybody seems to agree on the importance of human capital and thus education. References • • • • • • • • • •

Krueger, D. , “US-Europe Growth Differences: The Role of Education,” University of Pennsylvania, CEPR and NBER Dahlin, B.G. , “The Impact of Education on Economic Growth,” Duke University Krueger, D. and Kumar, K.B. (2003), “Skill-specific rather than General Education: A Reason for US-Europe Growth Differences?,” ‘education policy analysis 2002’ OECD Mamuneas, Savvides and Stengos, april 2002, ‘economic development and the return to human capital: A smooth coefficient semiparametric approach’ Johannes Hers CPB rapport 1998,‘Human capital and economic growth: a survey of the literature’ Jonathan Temple, ‘Growth effects of education and social capital in the OECD countries’ Barbara Sianesi and John Van Reenen, march 2002‘The returns to education: a review of the empirical macroeconomic literature’ http://www.house.gov/jec/educ.htm http://www.oecd.org

II. 3 (G21) Structural Policies Ben van Gils Marc Rooijackers Peter van Oudheusden Bart van de Gevel

DOES LIBERALIZING TRADE LEAD TO FASTER GROWTH? Introduction Leading economic institutions like the OECD, World Bank and IMF have been strongly advising countries to open up to the world economy. Increased openness could lead to higher growth and could cause convergence of less developed countries to developed countries. Countries with liberalizing trade policies can therefore get easier access to loans. More and more developed countries even tie their development aid policies of developing counties to those, who have open trade policies. In the last decennium empirical evidence has been presented that these policies do actually improve economic performance in growth. However a recent paper of Rodriguez and Rodrik (2000) criticize the used methods of these researches and doubt some of its results. Winters (2004) comes to following conclusion in his overview paper on the role of trade on growth:1 “While there are serious methodological challenges and disagreements about the strength of the evidence, the most plausible conclusion is that liberalization generally induces a temporary increase in growth” In this paragraph we will give an overview on some of the more influential studies that conclude that openness of a country does lead to more growth. First we will look at the much-cited study by Dollar (1992). Secondly we will discuss the study done by Warner and Sachs (1995) and by Frankel and Romer (1999). Also we discuss the critics made by Rodriguez and Rodrik (2000) on these three studies. We will end with a short conclusion. Dollar The first study that we discuss is the study by Dollar (1992). This was one of the first empirical studies that dealt with the empirical relationship between growth and trade. Dollar argues that the reason for this is that outward orientation is hard to measure across countries. According to him it combines two factors. The first factor is the level of protection and the second the amount of variability in the exchange rate. Because the latter doesn’t reflect a country’s long run trade policy Dollar focuses on the first factor. But how can we measure this factor? For this he introduces an index of country i’s relative price level (RPL). This index can be calculated by the following formula: RPLi = 100 × ePi / PUS , where e is the exchange rate in dollars per unit domestic currency and Pi is the consumption price for a country i. This formula implies that if al goods are tradable and there are no trade barriers then RPL would be 100 for every country. But the problem with the formula is that there are also non-tradables and that these will cause the value not be 100 for each country even with open trade. Dollar argues that differences in factor endowments between countries are the cause of different prices in non-tradables. He solves this problem by regressing the 1

Pp F4

1

price levels on factor endowments.2 He uses this relationship to calculate the measure by which a countries real exchange rate is distorted away from the free trade level. Dollar sorts the countries on the measure of distortion. On average he finds that countries in Africa are most distorted followed by countries in Latin America. Asia is less distorted than those two regions but more distorted than Europe and the Middle East. Besides this expected result he also finds some anomalies. Peru has a very low price level and Japan, Korea and Taiwan a surprisingly high level. To eliminate the anomalies he includes a measure of variability of the real exchange rate. The outward orientation index is being created by a weighted average of the two variables above. Dollar regresses his measure for outward orientation on growth. Important is that he adds the investment rate in his regression equation. This because both should produce more rapid growth and exclusion of the investment rate could cause a too great effect of outward orientation on growth. The regression shows there is a positive relationship between growth and outward orientation. But this result should be dealt with caution because outward orientation could also pick up effect of other relevant omitted variables and there could be backward causation. Dollar also includes a dummy variable for Africa in the regression equation. Even with this dummy included there is still a significant positive relationship. He divides his dataset in four quartiles bases on outward orientation and then estimates the growth for each of the quartiles. For the highest quartile he estimates a per capita growth rate of 2.9%, for the second quartile a growth of 0.9%, the third quartile -0.2% and the closest quartile a growth of -1.3%.3 Dollar comes on basis of this to the following conclusion:4 “The results strongly imply that trade liberalization, devaluation of the real exchange rate, maintenance of the real exchange rate could dramatically improve growth performance in many poor countries.” Sachs and Warner A second influential paper is titled “Economic convergence and economic policies” by Sachs and Warner (1995). In this paper they ask the question why poor countries do not grow rapidly to close the gap with rich counties. They think policy failures are the main cause why they can’t narrow the gap. This in contrast with other views which give human capital or technology the central role. In their paper they use two subsets of appropriate policies. The first set is related to property rights and the second is related to integration of the economy in international trade. For this paper we will look particularly at the second subset. This subset contains four criteria:5 (1) a very high proportion of imports covered by quota restrictions; (2) for Sub-Saharan Africa, high proportions of exports are covered by state export monopolies and state-set prices; (3) a socialist economic structure; (4) a black-market premium over the official exchange rate of 20 percent or more. Failure on one of these criteria does indicate that a country is not qualifying for the openness subset. When a country passes the criteria above it qualifies for the subset related to integration in the world economy.6 The subset related to property rights is constructed in a 2

Dollar explains the intuition behind this method as follows: “ The residuals of this regression indicates whether prices are high or low, given its factor endowments, and from these residuals can be constructed a cross-country index of real exchange rate distortion.” (Pp 526) 3 We include the sensitivity analyses of these estimates in table 1 in the appendix. 4 Pp 540 5 These descriptions are based on the ones by Sachs and Warner page 10 and 11. For how the find data for the criteria I refer to there article. 6 In their study they use a dummy 0 for qualifying and 1 for disqualified.

2

similar way. They now use regression analysis to confirm their hypothesis that qualifying countries display a strong tendency to convergence. In the appendix table 3 we include the results of there most important regression. In the second column of the table the regression contains only the two the subset variables: the political variable (PNQ) and the openness variable (ONQ). They come to the conclusion that both of these variables are important for growth and convergence. This result still applies when other variables are included to regression analysis like human capital and technology. This means that qualifying countries do better then non-qualifying countries. Important for our discussion is that they indicate that the influence of open trade policies is even larger and more significant than politics. Frankel and Romer The third study, which claims a positive relationship of trade liberalization and the last that we’ll discuss here, is the study of Frankel and Romer (1999). In there study they begin by pointing on the problems concerning previous studies on the topic. The biggest problem is that the effects of trade on growth, which previous studies show, could be caused by backward causation. It could be that a country changes its trade policies to more outward oriented when it grows. This implies that the trade is not the cause but the result of growth. To sort out which of these explanations is the most plausible Frankel and Romer use geographical variables as instrumental variables for international trade. The reason for this is, they argue, that knowing how far a country is from other countries provides considerable information on the amounts it trades. They base this claim on literature on the gravity model of trade.7 They even argue that geography does not influence income other than through trade. The size of a country is the second variable that could influence the within-country trade. By using these instruments they could determine the real influence of trade on growth. This influence is independent on a countries income. In the appendix we have enclosed table 4 with the basic results of the paper. In the first column we see that an increase of the share of trade with 1 percent point increases income per person 0.9 percent. The standard error estimate of 0.25 shows that the relationship is significant. Also there is positive relationship between size and income per person. But this relationship is only marginal significant. Frankel and Romer argue that the OLS estimates understate the influence of trade on income. Besides the OLS estimates also include instrument variables estimates (IV). This is because there are reasons why there should be a positive correlation between trade and the error term in OLS.8 There is now even a larger positive relation between trade and growth. A one percentage point increase in trade, in these estimates, will increase income per person with 2 percent. However, the relationship is not as significant as estimated by OLS. Lastly they test the results for robustness. They do this by removing outliers, correcting for differences in continents, taking oil producing countries into consideration9 and testing the instrument variables on endogenous components. None of these changes has a major effect on the results. Their conclusion is in summary that trade does raise income. The critics by Rodriguez and Rodrik The above studies conclude that trade does indeed foster growth. In their paper Rodriguez and Rodrik (2000) give some skeptics on the cross-national evidence. They don’t argue that there 7

See for evidence Linneman 1966, Frankel et al.1995 and Frankel 1997. They give in the paper the following four reasons why the IV estimates are larger than the OLS estimates: 1 Countries that adopt free trade policies are likely to adopt other policies that raise income. 2 Countries that are wealthy for reasons other than trade are likely to have better infrastructure and transportation. 3 Countries that with low trade may have to rely on tariffs to finance government spending. 4 Increases in income of sources other than trade may increase the variety of goods that households demand away and shift the composition of their demand away from basic commodities toward more processed lighter weight goods. Pp 391 t/m 392 9 In the 98-country sample which the table in the appendix contains these countries where already excluded. The relationship between trade and growth stays however significant and even increases somehow. 8

3

is no such relationship but that the evidence should be taken with some caution. We will look at the criticism they raise about the studies we have discussed earlier in our paper.10 They begin by criticizing the paper of Dollar (1992). First they argue that the distortion variable constructed by Dollar has some serious conceptual flaws.11 The problem with this concept is that economies that combine import barriers with export taxes will be less protected according to relative price level (RPL). Also countries, which will soften import restrictions by export subsidies, will look more protected than those that do not. They conclude that the measure of Dollar is sensitive to the form of trade restrictions. A second point of critique on the study is that the RPL method works best if the law of one price (LOP) holds. But as already noted above Dollar himself knew that there were problems with this law. According to Rodriguez and Rodrik the problems are more serious. First a significant part of the distortions are caused by monetary and trade policies. This is a problem because for example an appreciation of the exchange rate would raise the relative price of importcompeting and export-competing goods. This would cause a high distortion rate without trade policy having anything to do with it. Dollar’s measure of distortion fails also when transport costs or geographical factors influence the variable. The third point of critique is the index variability chosen by Dollar. This index has nothing to do with trade restrictions but measures instability at large. Dollar included this variable to eliminate anomalies in his dataset. The last point of skepticism that is being raised is that the results aren’t robust with respect to the distortion variable. Rodriguez en Rodrik show that with improved data and included variables like continental dummy’s, initial income and initial schooling the distortion index is not robust.12 This was the most important index used by Dollar. The variability index remains robust but this index doesn’t represents outward orientation. Rodriguez and Rodrik also criticize Sachs and Warner’s study, mainly with respect to the criteria used for openness. They argue that the individual components, which are mainly responsible for the strength of Sachs-Warner dummy, are not indicators for openness. The two dummies that are main distributors to its statistical power are the black market premium (BMP) and the state monopoly of exports (MON). To show this they construct a new dummy only consisting the above two variables BMP and MON. When this new dummy is used in a regression together with other variables, the latter have little predictive power. But what do the variables BMP and MON represent? Rodriguez and Rodrik argue that the BMP is in fact a dummy for Sub-Saharan Africa and the only information we can extract from it is that African economies have grown more slowly. This is because non-African countries, with restrictive policies towards export, escape disqualifying for this dummy and Africa countries with restrictive export policies but reformations in the late eighties are also overlooked. The second dummy BMP, which in theory could give inside on trade barriers shows also something different according to Rodriguez and Rodrik. They argue that the premium rather reflects a wide range of policy failures not only concerning trade barriers. Because Sachs and Warner only disqualify a country with a BMP of more than 20% they say the dummy measures the effect of widespread macroeconomic and political crises. They conclude that the Sachs and Warner measure is correlated with alternative explanatory variables. Therefore, it is too risky to draw any conclusions of it. The third pro-trade evidence on growth was the paper by Frankel and Romer (1999). Also this paper gets its share of criticism by Rodriguez and Rodrik. First they claim that the paper concerns the volume of trade instead of trade policies. Secondly they argue that the geographical instrument variable used in the paper is not a valid instrument. The reason for this is that geography influences growth in more ways than only trade. They give as an example the quality of institutions through colonialism, wars and migration. Also geography could influence the quantity and quality of natural endowments. Because of this they re-run

10 They also discuss papers of Edwards(1998), Ben-David(1993) and in short three others (Lee 1993, Harrison 1996 and Wacziarg 1998). 11 For the proof we refer to pages 15,16 and 17 of the paper of Rodriguez and Rodrik (2000) 12 They used a more recent version of the Summer-Heston database (Mark 5.6 instead of Mark 4.0)

4

the Frankel and Romer regression with additional geography variables and come to the conclusion that trade becomes insignificant.

Conclusion We have seen three studies, which claim to have found empirical evidence that liberalizing trade policies could improve growth. Important is that most of the studies only predict a moderate effect of trade policies on growth. These three studies are only a couple of numerous studies of the last decade, which have made the same claims. Although these results have caused a wide support on free trade policies by eminent economic institutions there is also criticism. Rodriguez and Rodrik have shown that the studies have some major shortcomings. They have the concern that trade policies crowded out other institutional reforms with potentially greater payoffs. We join Winters (2004) in his conclusion that the empirical work shows that it is most plausible that trade induces a temporally moderate positive effect on growth. But we also agree with Rodriguez and Rodrik that institutional reforms are probably a better bet for developing countries, to pursue growth. References Dollar, D. (1992), Outwarted-oriented Developing Economies Really Do Grow More Rapidly: Evidence from 95 LCD’s, 1976-1985, Economic development and Cultural Change, Vol. 40, pp523-544 Edwards, S.(1998), Openness, Productivity and Growth: What do We Really Know?, The Economic Journal, Vol. 108, No. 447 pp 383-398 Frankel, J.A., and Romer D.(1999), Does Trade Cause Growth?, The American Economic Review, Vol. 89, No. 3, pp 379-399 Rodriguez, F and Rodrik, D. (2001), Trade policy and economic growth: a skeptic’s guide to the cross-national evidence, Macroeconomics Annual 2000, pp 261-324 Romer, P. (1994). New goods, old theory and the welfare cost of trade restrictions. NBER working paper series nr. 4452 Sachs, J.D. and Warner, A.M.,(1995), Economic convergence and economic policies, NBER working paper series, Working Paper No. 5039 Winters, A.L. (2004), Trade Liberalization and economic performance: An overview, The Economic Journal, 114 February, F4-F21

5

Appendix Table 1 Sensitivity Analysis Regression Number Constant N = 95 DC 1 1,84 2

2,87

N = 48 PC 3

1,23

4

2,17

N = 24 PC 5

1,86

Real Exchange Rate Distortion Variability

Investment

-0,033 [3,55] -0,045 [4,91]

-0,8 [2,9] -0,07 [2,71]

0,18 [5,24] 0,16 [5,2]

-0,029 [4,02] -0,033 [5,45]

0,02 [0,64] 0,03 [0,96]

0,2 [3,53] 0,15 [3,05]

-0,027 [2,6]

0,04 [0,86]

0,08 [0,96]

Dummy

R2 0,4

4,66 [4,01]

0,49

0,38 5,24 [4,37]

0,57

0,26

Source: Economic Development and Cultural Change (Dollar 1992) Table 2 Outward orientation and other measures

1 2 3 4

GDP per capita Distortion 2606,13 88,22 1603,04 110,79 1883,83 140,54 1003,96 174,00

Variability 0,09 0,11 0,15 0,28

(% Of GDP) 2,86 0,85 -0,24 -1,30

(% of GDP) 20,67 16,87 18,03 14,60

Deze tabel is een samenvatting van de appendix uit het desbetreffende artikel waarbij 1 = erg open, 2 = open, 3 = gesloten en 4 = erg gesloten

Source: Economic Development and Cultural Change (Dollar 1992)

6

Table 3 Growth effects of the political and openness measures

Constant (t-ratio) SOC EDU RIGHT

1 9,912 [5,344] -1,267 [-1,820] -0,624 [-1,391] -0,670 [-1,227]

PNQ BMP OWQID EXM

-0,764 [-1,852] -2,246 [-6,026] -0,848 [-1,958] -1,885 [-3,584]

ONQ LGDP70 R bar 2 Mean dv. Standard Error Sample Size

2 8,282 [4,408]

-0,839 [-3,8] 0,446 1,583 1,587 95

-2,770 [-5,851] -0,568 [-2,595] 0,311 1,512 1,824 114

Source: Economic convergence and economic policies (Sachs and Warner 1995) Table 4 Trade and Income Estimation Constant Trade Share Ln population Ln Area Sample size R2 SE of regression First stage F on excluded instrument

OLS 7,4 [0,66] 0,85 [0,25] 0,12 [0,06] -0,01 [0,06] 150 0,09 1

IV 4,96 [2,20] 1,97 [0,99] 0,19 [0,09] 0,09 [0,10] 150 0,09 1,06 13,13

Source: Does trade cause growth? (Frankel and Romer 1999)

7

OLS 6,95 [1,12] 0,82 [0,32] 0,21 [0,1] -0,05 [0,08] 98 0,11 1,04

IV 1,62 [3,85] 2,96 [1,49] 0,35 [0,15] 0,2 [0,19] 98 0,09 1,27 8,45

II.2 (G22) Structural policies Pieter van Erp Dirk Hendrix Marcel Jonker Luc Verschuren

PRODUCT MARKET COMPETITION AND INNOVATION The best of all monopoly profits is a quite life.

John Hicks (1935)

1. INTRODUCTION In March 2000, leaders of European countries established an agenda to make the European economy by 2010 the most competitive and dynamic in the world. They expressed the ambition to become, within a decade, “the most competitive and dynamic knowledge based economy in the world, capable of sustainable growth with more and better jobs and greater social cohesion” In the ‘Lisbon strategy’ detailed policy recommendations were agreed upon, covering matters such as research, education, training, internet access and on-line business. (Source: Towards a knowledge based society) A cornerstone of the Lisbon strategy is its focus on the completion of the internal European market. This emphasis is derived from the notion that (further) reform of economic, social, and administrative regulations can produce significant and long-lasting benefits by means of inducing stronger product market competition within the European Union. Increased competition, subsequently, is thought to result in lower prices, improved product quality, increased consumer choice, and, most importantly, in increased innovation. In assessing the link between competition and innovation an important distinction should be made between static and dynamic efficiency. This distinction will be introduced in the first part of this paper, which covers the theoretical background of the relation between competition and innovation. The subsequent section then discusses the deregulation process and its results in the European telecommunications sector. Chapter four, then, will provide a similar case study of the European air transport sector. The fifth and final chapter of this paper will address the conclusions based on the theoretical evidence, empirical evidence, and case studies as presented in this paper.

2. THEORETICAL BACKGROUND

2.1 Competition and static efficiency

In (product) markets characterized by imperfect competition firms often seek to lower their production in order to create scarcity rents. As a result resources are moved to less productive sectors in the economy. Moreover, in product markets characterized by oligopolistic competition companies also tend to invest heavily in competing for market shares and entry deterrence. (See e.g. Scherer 1980 and Wilson 1990). Imperfect competition thus leads to inefficiencies resulting from resource misallocation. Besides from resource misallocation inefficiencies also result from opportunity for slack that is introduced by monopoly power. Several reasons for so-called ‘X-inefficiencies’1 have been introduced: - Regarding managerial effort, it is often impossible -or at least costly- to directly observe input and effort of managers. To overcome this principal-agent problem reward structures are often based on relative performance. Yet, in the case of imperfect competition there is less opportunity for comparison of performance. This creates opportunities for slack, and leads to inefficiencies. (See Nalebuff and Stiglitz 1983) - An alternative explanation of inefficiencies resulting from imperfect competition is based on the observation that, in the presence of market power, profits are less sensitive to the actions of managers. Consequently owners have less incentive to ensure that managerial effort is kept high; again resulting in inefficiencies. (See Nickell 1996:727) - Imperfect competition may also influence the effort of workers when monopoly rents are shared with workers. This might occur because it makes the life of managers more comfortable (see e.g. Smirlock and Marshall 1983) or because workers demand a fair wage/effort ratio compared to managers (see e.g. Akerlof and Yellen 1988). If higher rents derived from monopoly paper are indeed shared with workers,2 these rents are either captured in the form of higher wages 1

X-inefficiency is the term introduced by Liebenstein (1966) to capture inefficiencies generated from noncompetition, i.e. low effort of managers and workers to reduce costs, improve quality, introduce new ways of doing things, etc. 2 Both Krueger and Summer (1986a,b) and Dickens and Katz (1986a,b) provide impressive evidence that rent sharing is indeed quite common across different types of industries.

2

or reduced worker effort. Consequently there is a direct connection between higher rents derived from monopoly power and 1) lower worker effort and 2) stronger incentives to engage in unproductive activities to protect vested interests. Both lead to inefficiencies resulting from imperfect competition. In conclusion, increasing competition can lead to efficiency gains by reducing both allocative inefficiencies and X-inefficiencies. These are so-called ‘one-off efficiency’ improvements, or ‘static efficiency’ improvements, because it improves efficiency at a specific moment in time only. As will be addressed in chapters four and five, increased competition can markedly improve static efficiency in markets that were previously characterized by imperfect competition, resulting in higher quality products and most noticeably lower prices for consumers.

2.2 Competition and dynamic efficiency

While an efficient use and allocation of resources at any moment in time is obviously important, in the medium and long run, it is dynamic efficiency that matters most for growth in living standards. (OECD 2002:3) Here ‘dynamic efficiency’ or ‘ongoing efficiency’ improvements can be broadly defined in terms of productivity growth through innovations. Several theoretical channels through which competition leads to innovation and thus increases dynamic efficiency have been proposed: - “Darwinian effect: Intensified product market competition could force managers to speed up the adoption of new technologies in order to avoid loss of control rights due to bankruptcy (Aghion et al., 1999). More generally, firms should innovate to survive under competitive pressure (cf. Porter, 1990). - Neck-and-neck competition: In a simple model of “creative destruction”, the incumbent firms unlike new entrants have no incentives to innovate. Under a more gradualist technological progress assumption with incumbent firms engaged in step-by-step innovative activities, competition could increase innovation. It is because more intensive product market competition between firms with “neck-and-neck” technologies will increase each firm’s incentive to acquire or increase its technological lead over its rivals. - Mobility effect: In the learning-by-doing model of endogenous growth, the steady-state rate of growth may be increased if skilled workers become more adaptable in switching to newer production lines (namely, Lucas effect). In this case, more competition between new and old production lines (parameterised by increased substitutability between them) will induce skilled workers to switch from old to newer lines more rapidly (Aghion and Howitt, 1996).” (Source: Ahn 2002:7)

2.3 The Possible Trade-off between Static and Dynamic Efficiency

3

In the line of reasoning of Schumpeter (1943) the organization of firms and markets most conducive to solving the static problem of resource allocation (static efficiency) is not necessarily most conducive to innovation and technological progress (dynamic efficiency). While there is general consensus that increased competition may bring static efficiency gains, Schumpeter argued that market power, and not competition, is the driving force behind innovation: - “Firstly, some form of temporary ex post market power is required for firms to have the incentive to invest in R& D. - Secondly, the rents from ex ante market power provide firms with the internal financial resources for innovative activities. - Finally ex ante market power also helps reduce uncertainty associated with excessive rivalry which tends to undermine the incentive to invest.” (Kim and Phillips 2003:2) Note that if this trade-off between competition and static efficiency versus dynamic efficiency indeed exists, the Lisbon strategy would be fallacious. Stimulating competition in European markets would indeed bring allocative efficiency gains and result in better products and lower prices; yet in the long run it would hurt innovation and economic growth instead of helping Europe to become “the most competitive and dynamic knowledge based economy in the world.”

2.4 Is Competition conducive to innovation?

The basic Schumpeterian model indicates that innovation and growth are declining with competition because the monopoly rents from innovation tend to dissipate more quickly when there is stronger competition. However, as mentioned by Scarpetta and Tressel (2002:5), extensions of this model yield a more complicated picture and often lead to opposite conclusions (see e.g. Aghion and Howitt, 1998; Boone, 2000). Empirical evidence should resolve the matter, but the amount of empirical evidence is unfortunately limited due to data availability. Nevertheless the general conclusion of the available empirical literature is a positive relationship between competition and innovative activity. (See e.g. Nickell (1996), Blundell et al. (1995) and Bassanini and Ernst (2002), OECD 2002, etc.) An example is for instance recent OECD research (2002) which used the extent of anticompetitive product market regulation (PMR) as a proxy for the strength of product market competition and R&D intensity as a proxy for innovative activity. Both proxies are of course not perfect; R&D spending for example measures inputs rather than outputs from the innovative process, and disregards several other important aspects of innovation; yet the cross-country pattern of R&D intensity and the extent of product market deregulation does suggest a positive relationship between product market competition and innovation:

Figure 2.4.1 R&D intensity& product market deregulation (1998)

4

(Source: OECD 2002)

Important empirical evidence as presented in Aghion et al. (2002), which is in support of an inverted U relationship between competition and innovation, also indicates that competition has a positive effect on innovation. Yet under fierce competition firms lack the capacity, (and in highly concentrated industries a monopolist lacks the incentive), to invest in innovation. Therefore, if a market moves from monopoly to perfect competition, innovative activity generally increases, but comes to a halt when competition gets too intense. This would not only be a logical conclusion, it also combines strong arguments for and against a positive relationship between competition and innovation.

4. CASE STUDY: THE EUROPEAN TELECOM SECTOR

Historically, the European telecommunication sector was characterized by strong national monopolies. This environment created a strong national orientation for the sector with the consequent loss of the potential opportunities of a European-wide market. In 1999, for example, it could happen that a short-distance call from two neighboring European states was more than 3 times as expensive as a call from the US to Europe. (Source: Ten years without frontiers 2002:29) To improve efficiency, reduce costs, and stimulate innovation the telecommunications markets have been fully liberalised in most of the EU as of 1 January 1998. The program to liberalize the

5

telecommunications market was part of the wider process of the economic integration of Europe. The Treaty of Rome (March 25th 1957) initiated this process and it was accelerated through the European Community's internal (single) market programme in the mid-eighties. For the telecommunications sector especially the Maastricht Treaty (EU treaty, 1993) was very important for the integration of the telecommunication sector and its internal network. A couple of common policies were initiated from 1984 on to establish common development lines. The most important aspect of this first phase policies were: • • • •

Standards development to solve the problems of national fragmentation created by different national specifications. Common research; shared programmes between the operators and industry at European level (R&D framework programmes). Special development programmes for the least developed regions of the EU. Programmes in the context of structural funds (adopted in 1986). Initial tentative steps towards a common European position in the international telecommunication arena. (Source: http://europa.eu.int/ISPO/infosoc/telecompolicy/en/tcstatus.htm#I.)

By 1987, the second phase of the community policy was initiated with the publication by the “Commission of the Green Paper on de development of the common market for telecommunication services and equipment”3. This Green Paper opened the debate on the telecommunication regulatory environment in Europe. From this moment on great steps towards the European single market were made, eventually resulting in full market liberalisation, harmonisation of conditions for a common regulatory framework, and promotion of European players in the telecommunications market. All changes in policy and the effects this had on the markets of course brought some major benefits for everybody living in these markets. Liberalisation in the telecommunication sector resulted in benefits such as: •

Greatly increased choice for consumers. Over 95 % of the population in 12 Member States can now choose between more than 5 operators for long distance and international calls. A choice of more than 5 operators for local calls is also possible in 8 Member States

•

Competition between these operators is spurring innovation. There are now more advanced services than ever before. The average level of internet penetration in EU households, for example, was around 40 % in June 2002 – up from 18 % in March 2000. Furthermore, high-speed internet access is progressively gaining ground, and in October 2002 there are 10.8 million retail broadband customers in the EU.

•

Competition, combined with technological progress, is also bringing down prices. For example, tariffs charged by the old national monopolies for national calls have been

3

Towards a dynamic European economy: Green Paper on the development of the common market for telecommunications services and equipment, COM(87) 290 final, 30.07.19

6

reduced by around 50 % on average since liberalisation, and those for international calls by around 40 %. New operators in many Member States offer even lower prices, even for local calls: new entrant tariffs for national calls are up to 56 % lower and for international calls up to 65 % lower in some countries. •

As a result, the cost of a basket of national calls - including fixed charges and subscriptions - has fallen for both business and residential users since 1996. Business users pay, on average, 30 % less for the same service, while residential users pay 16 % less.

(Source: Ten years without frontiers 2002:29)

In conclusion, with the arrival of the full liberalization of telecommunication sector in the majority of the Member States of the European Union on 1 January 1998, a 10-year process of harmonization and liberalization was completed. Old-style monopolies have been broken up, which led to increasing competition. Because of this process the transparency and predictability of the market increased and economic performance, development, and innovations in the market were stimulated.

5. CASE STUDY: AIR TRANSPORT

Similar results have been accomplished in air transport. Historically, the aviation sector was characterized by tight regulation on the basis of bilateral agreements. Flying between two major European cities usually implied, as recent as five years ago, that one had to choose between one of the national airlines. “Three successive packages of liberalizing measures – adopted during the 1990s – have effectively turned all European-owned and controlled airlines, regardless of the Member State in which they are legally established, into “Community carriers” with equal rights of access to all the Community’s markets and equal responsibilities under the law. The result is that any EU airline can now operate on any route within the Community. This has triggered a number of developments in the industry: • The number of promotional fares available has increased dramatically. This is due to the presence of low-cost carriers in the market and the way in which established carriers have responded to them (see box 10). According to a recent study prices of promotional fares decreased by 41 % between 1992 and 2000. While some of the very lowest prices may not prove sustainable, the importance of promotional fares in the industry is unlikely to change.

7

• The number of “Community Carriers” offering scheduled passenger and freight services has risen from 119 in 1992 to 133 today, having peaked at 140 in 2000. Since 1993 an average of more than 20 new carriers per year have been set up in the EU. Over 40 carriers formed in 1993 have shut down or been taken over. All this points to dynamism and lively competition in the sector. • The number of routes linking Member States has risen by 46 % since 1992 – giving passengers a wider choice of destinations and carriers. The number of routes where more than two carriers are competing rose from 61 in 1992 to 100 in 2001. On such routes, business, economy and promotional fares were around 10 %, 17 % and 24 % lower.” (Source: Ten years without frontiers 2002:27) Increased competition in the air transport sector thus led to more intra-EU routes, more competition, and significantly lower prices. Moreover, especially the low-cost carriers implemented a completely new and innovative business model to compete with the established national airlines. Essential elements of this business model are a single type fleet of planes, fast turn-rounds, use of cheap secondary airports; no frills or free drinks during the flight; and lowcost booking almost entirely over the internet. In Europe, indeed 99% of cheap intra-EU flights are now booked through the internet. (Source: The economist) Additionally, increased competition in the air transport sector also forces established airlines to adopt the innovative internet-selling and yield-management techniques of their low-cost rivals. In conclusion, the liberalization of the EU air transport market is another example of increased competition leading to lower prices, while simultaneously greatly stimulating innovative behavior across the sector. 6. CONCLUSION

Increasing competition by means of regulatory reform can lead to large static efficiency gains in markets characterized by imperfect competition. Several channels have been discussed through which increasing competition leads to a reduction in allocative inefficiencies and so-called Xinefficiencies. There is general consensus in the literature, and also abundant empirical evidence, that increasing competition in these markets leads to higher quality products and services, increased consumer choice, and lower prices. Besides the two case-studies performed in this paper, table 1 shows many more examples of increased competition leading to strong price reductions in real terms:

8

(Source: OECD 1997:11) While there is general consensus in the literature and empirical evidence that increasing competition can lead to static efficiency gains, there is still some controversy about the link between competition and dynamic efficiency gains. Our two case studies, in line with the general conclusion of the empirical literature, indicated a positive relationship between competition and innovation. However, following Schumpeter (1943) and empirical evidence of Aghion et al. (2002), some degree of market power is needed to provide firms with the internal financial resources for innovative activity and some degree of market power also reduces uncertainty associated with excessive rivalry. Therefore, in order for the Lisbon Agenda to succeed and Europe to become “the most competitive and dynamic knowledge based economy in the world, capable of sustainable growth with more and better jobs and greater social cohesion,” we conclude that product market deregulation and the completion of the internal market has been sound economic

9

policy to increase static and dynamic efficiency in the European Union. Two caveats, however, should be taken into account. Firstly, increasing competition on product markets characterized by imperfect competition will indeed increase static efficiency, but creating too much competition might actually hurt innovation and thereby economic growth. Secondly, increasing competition by means of product market deregulation should not become a one-size-fits-all approach based on past success. Each market is different and programs to further deregulate and increase competition in for example the electricity, gas, and financial markets have to take this into consideration. As long as the European Union takes this into account we believe that similar results can be achieved as those of the programs that deregulated the telecommunications and air transport sector.

References Ahn, S. (2002) “Competition, Innovation and Productivity Growth: A review of Theory and Evidence,” OECD Economics Working Paper No. 317. Aghion, P; Howitt, P. (1996) “Research and development in the growth process”, Journal of Economic Growth, Vol. 1, Nr. 1: pp. 49-73. Aghion, P; Howitt, P. (1998), Endogenous Growth Theory, MIT Press, Cambridge and London. Aghion, P; Dewatripont, M; Rey, P. (1999) “Competition, financial discipline and growth”, Review of Economic Studies, 66, pp.825-52. Aghion, P; Bloom, N; Blundell, R, Howitt, P. “Competition and innovation: An Inverted U Relationship,” The Institute for Fiscal Studies, Working Paper, 02/04 Akerlof, G; Yellen, J. (1988) “Fairness and Unemployment,” The American Economic Review, Vol. 78, No.2, Papers and Proceedings of the One-Hundredth Annual Meeting of the American Economic Association, May: 44-49 Bassanini, A; Ernst, E. (2001) “Labour market institutions, product market regulation and innovation: Cross-country evidence”, OECD Economics Department Working Papers Blundell, R; Griffith, R; Van Reenen, J. (1995) “Dynamic count data models of technological innovation”, The Economic Journal, 105, March, pp.333-344. Boone, J. (2000), “Competition”, Center for Economic Research Discussion Paper No.2000-104, Tilburg University, Netherlands, October. 10

Dickens, W; Katz, L. (1986a) “Interindustry Wage Differences and Industry Characteristics,” NBER Working Paper No. 2014, September 1986 Dickens, W; Katz, L. (1986b) “Industry and Occupational Wage Patterns and Theories of Wage Determination,” mimeo., March 1986. Kim, J; Phillips, B. (2003) “Competition and Economic Development” Seoul Competition Workshop Krueger, A; Summers, L. (1986a) “Efficiency Wages and the Inter-Industry Wage Structure,” mimeo., 1986 Krueger, A; Summers, L. (1986b) “Reflections on the Inter-Industry Wage Structure,” Harvard Institute of Economic Research Discussion Paper No. 1522, July 1986 Liebenstein, H. (1966) "Allocative Efficiency and X-Efficiency". American Economic Review, June: 392-415. Nalebuff, B; Stiglitz, J. (1983) “Prizes and Incentives: Towards a General Theory of Compensation and competition,” The Bell Journal of Economics, Vol. 14, No. 1, Spring: 21-43 Nickell, S. (1996) “Competition and Corporate Performance,” The Journal of Political Economy, Vol. 104, No. 4, August: 724-746 OECD (1997), The OECD Report on Regulatory Reform I-II, Paris OECD (1999), Implementing the OECD Jobs Strategy: Assessing Performance and Policy, Paris OECD (2000), OECD Economic Outlook, No. 67, Paris. OECD (2001), OECD Economic Studies: Special Issue on Regulatory Reform, No. 32, Paris. OECD (2002), OECD Economic Outlook, December, Paris Porter, M. (1990) The Competitive Advantage of Nations, London: Macmillan Press. Scarpetta, S; Tressel, T (2002) “Productivity and Convergence in a Panel of OECD Industries: Do Regulations and Institutions Matter?” OECD Working Paper 342, September: 1-40 Scherer, Frederic (1980) Industrial Market Structure and Economic Performance, Second Edition. Chicago: Rand McNally.

11

Schumpeter, J. (1943) Capitalism, Socialism and Democracy, Harper and Row, New York Smirlock, M; Marshall, W. (1983) “Monopoly power and expense-preference behavior: theory and evidence to the contrary,” Bell Journal of Economics, Vol.14, 166–178. “Ten Years without Frontiers,” (2002) EU Working Paper, http://europa.eu.int/comm/internal_market/10years/docs/workingdoc/workingdoc_en.pdf “Towards a knowledge based society,” Europe in 12 lessons, http://europa.eu.int/abc/12lessons/index8_en.htm “Turbulent skies,” (2004) The Economist, July 8th www.economist.com/business/displaystory.cfm?story_id=2897525 Vickers, J. (1995) “Concepts of Competition,” Oxford Economic Papers, New Series, Vol. 47, No. 1, January: 1-23 Wilson (1991) Strategic Models of Entry Deterrence; The Handbook of Game Theory, R. Aumann and S. Hart (eds.), Amsterdam: North-Holland/Elsevier Science Publishers

12