Printer Friendly

A development club and groupings in Europe, and growth strategies.


In 2000 the European Council of the European Union launched an agenda for reforms in order to strengthen employment, economic reforms and social cohesion. In 2005 the European Council issued conclusions on 'Re-launching the Lisbon strategy: A partnership for growth and employment'. Slovenia is a transition country that joined the EU in 2004. In 2005 it launched the 'Framework of economic and social reforms in order to increase the welfare in Slovenia'. These are just two instances of recent attempts to formulate a growth policy. The formulation of growth policy which would accelerate and sustain a high growth rate is of great concern to policy-makers.

A subject of growth policy and the question of finding a framework for formulating a growth policy for specific nation-state conditions has recently come onto the research agenda. Rodrik (2005), Hausmann et al. (2005), Aghion and Howit (2005), and Sapir et al. (2004) have made recent contributions here.

What should an analytical framework discuss or advise about the growth policy of a particular country? Does economic theory give straightforward instructions for how to form a growth policy? Rodrik (2005, pp. 993, 1009) reports that no country has experienced rapid growth without minimal adherence to higher-order principles of sound economic governance--property rights, market-oriented incentives, sound money, and fiscal solvency. He points out that the key is to realise that general theoretical principles do not translate directly into specific economic growth policy recommendations. A similar statement could be made with regard to theories and models of economic growth. Pritchett (1997, p. 15) expressed the opinion that it is conceivable there is an all-purpose theory of economic growth, but it is more plausible that the appropriate growth policy will differ according to the particular situation. I conclude from these two statements that general growth theory does not suffice to formulate growth policy. Growth theory deals with big ideas, general factors and principles of economic growth. At this level, it is of general validity and it is institution-free. Growth policy, on the other hand, cannot be formulated a priori; it is contingent on the local economic and political context. Unlike economic principles, growth policies do not hold general validity. Rodrik (2005) suggests that the successful acceleration of growth is frequently related to non-standard growth policies (eg China, East Asian countries, Ireland).

Many growth policies are possible. The problem is to know which particular growth policy to chose for a particular country. I pose the following question: can a grouping of countries into a club or groupings according to some common characteristics reduce the range of possible growth policies? I shall elaborate on the question by extending the growth strategy concept to a group of countries. Is it possible to apply one growth strategy to a group of countries? For example, is it possible to apply a single growth strategy to the 25 (or more) countries of the European Union? I maintain it is possible to extend the concept of a growth strategy to a group of countries with certain similar features. The thesis of the article is that it is useful, in terms of growth policy, to arrange European countries into several groupings and that there is a mainstream growth pattern in the grouping.

The article is structured as follows. In the first part I will present convergence theories and theories of the stages of growth and competitive development in order to identify bigger ideas concerning the theories of growth. By combining the ideas of three theoretical sources I define the concept of development clubs. Based on data for the 1955-2000 period I identify the development groupings of countries in Europe. Further on I discuss the question of whether certain groups of countries form development clubs in Europe. I then examine the dynamics of development clubs and groups. The concept of development groups and clubs is finally applied to discuss some implications for the growth policies of the EU, groups of European countries and some individual European countries.


Stages of growth and stages of competitive development In the 1950s Rostow's famous theory of stages of growth emerged (Rostow, 1960). W.W. Rostow used the notion of stages as a tool for analysing economic growth over a longer historical time, spanning the last hundred years. His thesis was that countries go through stages of growth. The decisive moment in the growth process is the 'take-off'. Different countries find themselves at different stages at a certain moment of time. The idea was that there is a linear movement through the stages--from the lowest through the take-off up to the highest. Looking from today's perspective, we can say there is an implicit thesis of a linear convergence towards the highest stage, a stage of mass consumption. Thirty years later, Michael Porter (1990) employed the notion of stages to explain the theory of the competitive advantages of nations. National economies exhibit a number of stages of competitive development reflecting the characteristic sources of advantages of the nation's firms in international competition. Porter's theory suggests three distinct stages of national competitive development: factor-driven, investment-driven, and innovation-driven. It is not inevitable that nations progress linearly through the stages, from lowest to highest. It is possible that a national economy can become stuck in one stage; it is even possible that it slides down to a lower stage of competitive development. Such patterns are possible because a competitive advantage is defined relative to other economies. In such a setting, the absolute convergence of all countries is impossible. There are always differences in competitive advantages among nations.

Absolute and conditional convergence and convergence clubs

Absolute convergence is based on neoclassical growth models for closed economies as presented by Solow (1956). Absolute convergence implies that the per capita growth rate tends to be inversely related to the starting level of output per person. The neoclassical model predicts that each economy converges to its own steady state. The steady state is determined by the saving rate (s), population growth rate (n), and rate of technological progress (g). According to the model (Mankiw, 1995), income per capita (y) converges to its steady-state ([y.sup.*]) as follows: dy/dt = -[lambda](y-[y.sup.*]), where [lambda] is the rate of convergence.

In a particular application to a cross-country case, the absolute convergence thesis would predict that poor economies grow faster than rich ones if the economies were similar in respect of their preferences and technology. The assumption underlying cross-country growth studies is that all countries obey a common linear specification. Barro and Sala-i-Martin (1992) evaluated arguments on convergence and found that the per capita growth rate is negatively related to the log of the initial per capita GDP and that there is a convergence over the whole sample of countries only if different conditions are taken into account (compare Mankiw et al., 1992).

The next stage in this line of argument is the idea of a 'convergence club'. The thesis is that there is convergence within a certain group of countries, forming a convergence club. Baumol (1986) observed (for the 1950-1980 period) the remarkable convergence of output per labour hour among freemarket industrialised nations. He called these countries a convergence 'club'. He suggested there are three convergence clubs. Centrally planned economies are members of a convergence club of their own (the second club). This club displays inferior performance to that of free-market industrialised countries, which form the first club. The third club consists of poor, less-developed countries. These countries are still largely excluded from the homogenisation processes (Baumol, 1986, p. 1080). Baumol noted that in the 1950-1980 period there was little convergence among the groups (convergence clubs). The idea of convergence clubs has sparked an enormous number of articles on the subject (Islam, 2004; Desdoigts, 1999; Quah, 1996). Durlauf and Johnson (1995) summarised the results showing that cross-country growth behaviour in the convergence club model is nonlinear, exhibiting multiple regimes. Only countries associated with the same steady state obey a common linear regression.


According to Chatterji (1992), a convergence club implies the existence of a set of countries which are driven to steady state with equalised real per capita incomes. For my purposes of studying growth strategy such a definition is too narrow. I refer to the statement of Baumol and Wolff (1988, p. 1159) that the crucial issue is how countries achieve membership in the convergence club and on what basis they are sometimes ejected. In order to include this issue relevant to growth policy I have defined the concept of a 'development club'. A development club is formed by a group of countries that are at the same (income) stage of growth, which are converging to a common steady state (eg convergence club), and which are at the same stage of competitive development as revealed by their common patterns of growth.

Buchanan (1965) wrote that his theory of clubs applies to the organisation of membership or sharing arrangements where 'exclusion' is possible, and that the theory of clubs is a theory of optimal exclusion, as well as one of inclusion. The excludability question is therefore important in the discussion of clubs. What is the excludability factor in the case of development clubs? According to Schumpeter (1934) and Porter (1990), the excludability factors are innovation, technology and entrepreneurship. All of these are reflected in frontier and high-tech technologies. These technologies are not public goods and are protected by different schemes of property rights, and are not a subject of technology transfer. A club can only be formed by those countries which are in a position to protect their advanced position through the non-transferability of high technology and non-mobility of highly educated and research-based labour. Only the most advanced countries may form a development club.

I make a distinction between a development club and a development grouping. I define a development grouping as a group of countries that show some common similarities but not those that are excludable as in the case of a development club. If technology is a public good there cannot be a club based on exclusion by technology. We shall work on the assumption that high technology is exclusive and protected, that medium technology is on the market, and that low technology is a public good which is generally available. It is not possible to form a club on the basis of medium and low technology. Countries which are in such a stage of development that they can only use medium and low technology may be grouped in groupings but not in a club. (1)

There is a growth pattern of a development club which may not be taken over by imitation. Growth patterns of development groupings are transferable, for example, they are not excludable, and as such they cannot provide a basis for forming a club.

Growth patterns of the different development clubs

For the analysis we have in mind an aggregate-level analysis will not suffice. It is helpful to divide a whole economy into sectors. The traditional division into agriculture, manufacturing, and services will do for the lower stages. Yet for higher growth stages a different structure is relevant, for example a structure by the level of technology. Let us assume that every economy consists of three sectors. Each sector differs from the others by its level of technology. A level of technology is described by indicators of educational attainments and by research and development efforts) Sector 1 is at a high-technology level. Part of this sector is characterised by frontier or top technology. Top technologies are associated with highly educated labour, high R&D efforts, and creativity. Such technology is protected by intellectual property rights. It is not on the market. It is in no way a public good. The other part of the high-technology sector is one step behind the frontier. This technology is internationally transferable only through foreign direct investment (FDI) to countries at a similar development level. Sector 2 is at the middle level of technology. This technology is available in the market and easily transferable through FDI. This sector is two steps behind the frontier and one step behind the high-technology level. Sector 3 is at a low-technology level. It is one step behind the middle-level technology. This technology is, in fact, a public good and easily available. To introduce this technology into production, investment and middle-and low-skill labour is needed.

Economies at different stages of growth have a different structure of production in terms of the level of their technology. Most developed economies have a greater share of top- and high-level technology sectors. Economies of the second grouping have a larger share of middle- and low-technology sectors. The economies of the third grouping do not have top- and high-level technology sectors.

Innovation- (knowledge-) based growth

According to our definition, the first development club consists of economies which are high-income economies, innovation based and which have a large share of top- and high-level technology sectors. Economies are in an innovation-driven stage of competitive development.

Economies at the top of the first development club develop frontier technologies and high-tech technologies. [A.sup.*] is a productivity parameter attached to the frontier technology. The rate of growth of frontier technologies ([g.sup.*]) is based on higher education (u) and R&D activities (d).

[A.sup.*] (t) = (1 + [g.sup.*])[A.sup.*](t - 1) (1)

[A.sup.*](t) = f(u, d)[A.sup.*](t - 1) (2)

Other economies of the first development club try to follow the frontier technology through imitation and catching-up (the first term in (3)), and they innovate at the top of the high-technology level (the second term in (3)):

[DELTA]A = f([h.sub.1], k)([A.sup.*] - [A.sub.1]) + f(u, d)[A.sub.1](t - 1) (3)

[A.sub.1] is a productivity parameter attached to the high technology. ([A.sup.*] - [A.sub.1]) represents the technology gap vis-a-vis the frontier. The first term in the above equation represents the case where an innovation involves implementing technologies that have been developed elsewhere in the peer-level countries. To do this, human capital (h1) and investments in capital (k) are necessary. The second term represents activity whereby the country is making a leading-edge innovation that builds on and improves the leading-edge technology in its industry. To do this, it has to engage highly skilled labour (u) and R&D efforts (d).

A model of growth for this stage of growth is a model of innovation and growth of technological change. Aghion and Howitt (2005, p. 5) elaborated a similar model in more detail. It has to be noted that economies at this stage of development grow through innovations and they cannot grow basically on the basis of the imitation and implementation of technologies from another development club. Their growth is endogenously driven. Frontier technology is not on the market. If economies in the group want to eliminate the distance vis-a-vis the frontier technology of peer countries they have to do this solely through their own growth efforts.

Efficiency-based growth

The second development grouping is trying to catch up with the first grouping's technology level and to innovate at the top of middle-level technology. Economies are at the stage where they can basically grow through the imitation and implementation of technologies from other countries. Their innovation is mostly of the implementation type. Factors of their development are: capital deepening, k; human capital, [h.sub.2], based on learning type higher education; and R&D efforts, d.

[DELTA]q = f(k, [h.sup.2])([A.sub.1] - [A.sub.2]) + f([h.sup.2], d)[A.sub.2](t - 1) (4)

q is the productivity of labour. [A.sub.2] is a productivity parameter attached to medium technology. ([A.sub.1] - [A.sub.2]) is the productivity gap. Countries of the second grouping are at an intermediate stage between innovation and investment-factor-driven stages of development. I call this stage efficiency-based growth because it has the potential to be better exploited. In reference to older literature on socialist countries we may call this type of growth an intensive type of growth. A string from this development club to the first development club is realised through [A.sub.1], high technology, to which some of the economic sectors strive to catch up. If this succeeds, a transition to the higher development club is possible.

Factor- and investment-based growth

The third development grouping is in the early stages of investment-driven development. The representative technology level is low technology which is a public good. A model of growth is founded on the extensive use of factors based on domestic and foreign investment. Economies try to speed up the growth of their output by catching up and keeping up, and implementing and transferring technologies from other countries. We may also call this pattern of growth an extensive type of growth.

[DELTA]y = f(k)([y.sup.*] - y) + f(k, [h.sub.3])[A.sub.3](t - 1) (5)

y is product per capita. Benchmark income per capita ([y.sup.*]) is some average of countries in Europe or in the EU. ([y.sup.*] - y) is the income gap. [A.sub.3] is productivity parameter attached to low technology. Human capital, [h.sub.3], is of a secondary level. The growth of countries in this grouping is exogenously driven because it is driven by FDI and directed towards foreign markets.

Growth strategies

From this review of the three distinct growth patterns it follows that there are three distinct broad growth strategies. The mainstream growth strategy of the first development club is innovation based. It is a growth pattern that cannot be imitated by and transferred to other group of countries. This is a unique development club strategy. On the other hand, there are growth patterns which can be transferred. These are the groupings' strategies. This means that there are growth strategies which are transferable and others which are not. The club formation is based on growth patterns which are not easily transferable.


Groupings by income stage: convergence clubs I shall define and identify development clubs in Europe. I restrict the empirical analysis of an illustrative nature to the sample of European countries. (3) In principle, they form a homogenous group to such a degree that it is possible to talk about a comparison of growth strategies. Desdoigts (1999) suggested that economic structures are quite homogenous within each continent. I assume that groupings of countries on the world scale might require a different analytical framework than is provided in this article. Pryor (2005, 2006), for example, made an entirely different characterisation of economic systems for OECD countries and for developing countries.

I will make an initial grouping of countries by income per capita. Countries may be grouped, like the World Bank does, according to high-income, middle-income and low-income levels. I group all European countries into income groups according to their gross domestic product (GDP) per capita in current prices expressed in purchasing power parity (PPP) US dollars or in international dollars at the initial date (eg 1955). The data used here stem from United Nations and Word Bank publications, and Maddison (2003).

The grouping of countries according to GDP per capita discontinues after 1990. Three countries, Czechoslovakia, Yugoslavia and the USSR disintegrated into many countries. Instead of having one group called Eastern Europe it is sensible to form two new groups: Central Eastern Europe (CEE) and South East Europe (SEE). (4) I identify four groups of countries in Europe in the time period around 2000.

These four income groups of countries are potential development groupings for which growth strategies can be formulated. (5) Next we shall try to find out whether these groupings converge within the groupings in order to form convergence clubs.

The convergence of countries within D-clubs

We can expect that there is convergence within a convergence club. We assume that each convergence club has its own steady state. My supposition is that the steady state of the convergence club is determined by the steady state of the most developed countries in the convergence club. This means that countries converge toward the upper frontier of the development level of the convergence club.

Based on data from Maddison (2003) I checked data for three groups of European countries for the period 1955-2000 (Table 1). Maddison called these groups of countries: 12 Core-Western European Countries, four South-Western European Countries, and Eastern Europe.

I made a provisional calculation of the convergence of the G1-group, namely the countries of Core Western Europe. Sigma convergence, measured with a standard deviation, is not present. (6) The coefficient of standard deviation is pretty much the same in 1955, 1985 and 2000. Beta convergence is, however, present in the 1955-2000 period. (7) I applied one version of convergence equations and obtained the following result:

GYP = B - b(RGDP55)


where GYP is the growth rate of real per capita GDP (in international dollars) during the 1955-2000 period and RGDP55 is the 1955 level of per capita real GDP. Original data are transformed into natural logarithms (ln). Using the method of linear regression the estimated equation of the regression line is shown in (6). Coefficient 'b' is negative and the correlation coefficient is high which suggests that the per capita GDP of the 12 Core Western European countries, based on Maddison's sample data, converged between 1955 and 2000. One should be cautious about these results. There may be fluctuations in sub-periods between convergence and divergence. There was a sigma divergence in the 1955-1970 period and sigma convergence afterwards. The calculations suggest that the results of convergence could be highly sensitive to the initial and final time periods. The results of the regression analysis show there was a strong convergence within the G1 group or CWE countries.

Group G2, the SWE countries, is too small to consider convergence. The situation in the G3 group--the counties of Eastern Europe (EE) and the USSR--may have formed in the past period 1955-1985 some kind of convergence club as Baumol (1986) indicated. Yet history intervened and the group disintegrated so it no longer exists.

In Table 2 I present new groupings for 2003. Groups G1 and G2 are slightly different. Group 3 from Table 1 (the period 1955-1990) has split into two new groups G3/1 and G3/2, Central Eastern Europe and South East Europe, which are relevant for the period after 2000. For the countries of Central and South East Europe (groups G3/1 plus G3/2 in Table 2) it cannot be empirically established whether their per capita incomes are converging or not. The time period is too short and the very existence of many new individual countries is too short to reveal any pattern of growth and process of convergence. It is out of question to consider groupings of CEE and SEE countries as a possible convergence club.

The distance between groupings: convergence or divergence?

When comparing the average per capita income (in international dollars) of different groups of countries in different time periods we see a diverse situation. Convergence among the groups was not present in the past five decades of the post-war period in Europe. The economic order which prevailed in 1955-1985 collapsed in the 1990s and the gap between the upper and lower groups expanded. Divergence, not convergence, was taking place (Table 3).

In the 1986-2000 period some major political and economic changes took place in Europe. The development left groups G1 (12CWE) and G2 (4SWE) intact in the 1990s. Group G1 was even trying to incorporate G2 into its own club. Such was the case with Ireland. The countries of Eastern Europe and the former USSR (G3-group) were not converging in the 1955-1985 period. They began to diverge in the 1980s and in the 1990s. As a group they have disintegrated and split apart, both politically and economically. There was a major repositioning of almost all member countries of this group in 2000. The group of G3 countries (Table 1) split into the groups G3/1 and G3/2 (Table 2). At the beginning of this century, we had two groups (Table 2): Central Eastern Europe and South-East Europe. The G3/1 group was trying to catch up. The situation of the G3/2 group in 2000 resembles the development situation of LDCs on other continents like Asia. Such a situation was not seen in Europe in the 1955-1985 period.

Convergence club in Europe

To sum up: we have established one convergence club in Europe and two additional groupings according to income per capita. In order to establish whether the convergence club is also a development club, and whether the groupings are also development groupings, we will have to investigate whether the growth patterns are based on innovation, efficiency or on the extensive use of growth factors. Since only a convergence club may become a development club we shall pay more attention to the question of the presence of innovation-based growth patterns. If a European convergence club (eg G1 or CWE) reveals common characteristics of innovation-based growth patterns then we can identify the case of a development club in Europe.

Grouping of countries by the growth pattern index

In order to identify the growth pattern of different income groups of European countries I collected some generally available data to serve as indicators of growth factors. On the basis of the growth pattern described in the third section, I selected a few indicators which should reflect the mainstream growth pattern of the first grouping (innovation- and knowledge-based growth) as distinct from other groupings. Indicators were selected on the basis of theory and not on the basis of best fit. The other criterion was that indicators should be easily accessible in official publications of national or international organisations. The third criterion was that it is possible to assume that the higher value of an indicator goes with a higher level of a growth pattern. On the basis of indicators I constructed a growth pattern index.

Indicators of the growth pattern reflect research efforts [(1) expenditures on R&D, as a % of GDP (period 1996-2002), (2) researchers in R&D per million people (period 1990-2001)], educational attainment [(3) public expenditure on education, as a % in GDP (year 2002), (4) tertiary gross enrolment ratio, % of relevant age group (year 2001/02)], economic structure [(5) share of services in GDP, in % (year 2000), (6) employment in high-tech services], and factor intensity [(7) total employment rate (year 2004), and (8) the ratio of outward and inward FDI (year 2003)]. (8)

On the basis of these four groups of eight indicators I calculated the 'growth pattern index' (GPI). The GPI is calculated in a similar way as the Human Development Index (HDI) is calculated (UNDP, 2003, pp. 340-341). Performance in each indicator is expressed as a value between 0 and 1 by applying a dimension index. (9) Once dimension indices have been calculated, the GPI is a simple average of the eight dimension indices. For each country a GPI has been calculated. This enabled me to rank countries according to the value of their GPI and to group countries according to the range values of their GPI. Three groups of countries were thereby formed.

It is interesting to compare the groupings of countries according to their income per capita (Table 2) and groupings according to their growth pattern index (Table 4). Ireland and Italy are in the first income group but they are in the second group according to their GPI. Greece is in third GPI group, although it is in second income group. Estonia and Latvia are ranked in the second group according to the GPI, but by income they are in a third group. In this way I identified a group of countries for which the grouping by income per capita and by growth pattern does not coincide.

Ireland, Italy and Greece are ranked by income per capita in a higher group than by the growth pattern. The opposite is the case with Estonia and Latvia. It would be interesting to speculate what forces will eventually prevail: the income ranking or the growth pattern ranking. Is it possible to speculate that growth patterns will facilitate Estonia and Latvia making a speedy catch up to join the second income group? Is Ireland's income position sustainable with its current growth pattern?

Even such a simple classification of growth patterns shows there is no uniform growth pattern to reach a certain income group. It is, however, of interest that 11 countries of CWE form a single group by income per capita as well as by GPI. This may indicate that there is after all a mainstream growth pattern for most of the group of developed countries. Such a statement would substantiate the specification of growth patterns of the different development clubs that we provided at the beginning of the article (section Development Clubs and Groupings).

Migration of countries between development groupings

In a comparison of income per capita groups in the post-2000 period (Table 2), the period before 2000, for example the period 1955-1990 (Table 1), we see that the membership of the development groupings is very stable. Only small countries occasionally alter their group membership. None of the large countries has shifted to an upper grouping or descended to a lower stage in the last 50 years.

In the 1980s the G2 (SWE) group consisted of Ireland, Greece, Portugal and Spain. In comparison to the core group of countries these countries were less developed. Due to the principle of solidarity, the European Economic Community (EEC) formulated structural and cohesion policies in order to enable these countries to catch up. This policy was successful. Ireland has left the G2 group and joined the G1 group. Malta, Cyprus and Slovenia joined the G2 group in 2004. It is likely that some countries of today's G3/1 countries will join the G2 group in the future. Membership of the G2 group is partially transitory.

When examining the membership of development groupings by income in the 1970-1990-2000 periods (Tables 1 and 2), we see that only two countries migrated from a lower to a higher development grouping: Ireland in the 1990s and Slovenia after 2000. It would be of interest to elaborate on the experience of these two countries' accelerated growth. Which particular conditions and factors are needed for a country to leapfrog developmentally? Is there a lesson to be learned? Such questions have been matter of recent research into growth strategies (Rodrik, 2005). I will present one possible and tentative explanation.

I will apply the growth policy framework presented in this article to two different countries of interest. In the focus of my interest are Slovenia and Ireland, while Austria is referred to as a benchmark country. These three countries exhibit three different growth paths. Austria is a high-income country using high-income niches in manufacturing and services. Ireland has achieved a high-income position through a bypass growth path. Slovenia reveals some endogenous catching up.

The cases of Ireland, Austria and Slovenia indicate that small countries have an alternative way leading towards high-income countries. Of course, a country must have considerable institutional, human and technical proficiency to be able to excel in high-income niches or be the gateway of high-tech products to the markets of developed countries and to join the group of income-leading countries.

Ireland: exogenously-driven high-tech bypass growth path

A country from the lower grouping can narrow its gap vis-a-vis the higher grouping by imitation and transfers, but it cannot eliminate the gap altogether in this way. A follower remains a follower unless it is able to add in something that originates endogenously. A second possibility is that it achieves a bypass. Ireland was able to move from G2 to G1 not only because it was able to imitate G1 but because it was also able to bypass the border of G1 by acquiring the experience and resources from the most developed area in the world, for example the USA. (10) Slovenia was able to join the G2 club from G3/1 positions not just by imitating the G2 club but also by transferring some growth factors from G1 (eg the biggest trading partners of Slovenia were members of the G1-club, eg Germany, Italy, France, Austria).

Such bypass growth by imitation and transfer is only available to those countries that are at least two development stages behind the highest stage (eg the USA). The possibility of bypass growth allows some optimism for those countries in the lower groupings to catch up. The lower the position of the grouping, the greater is the pool of opportunities for fishing for imitation and transfers not just from the next grouping on the ladder, but also from a higher up development grouping.

In the 1991-2004 period Ireland increased its index of GDP per head of population (at current market prices, EUR-15 = 100) by 73 percentage points in euro terms (from 69 in 1991 to 142 in 2004), and by 50 percentage points in purchasing power standards (PPS) (from 77 in 1991 to 127 in 2004) (PC, 2004). This is an extraordinary performance. I am looking here for strategy-related lessons that are also applicable to other countries.

Why was Ireland able to move from stage 2 to the highest income ladder in Europe? With the transfer of American growth factors to Ireland, Ireland was able to pass from stage 2 to 1 within Europe. American firms at a high-tech level were investing in Ireland in order to gain access to the markets of developed European countries (eg G1 group countries). This has helped the Irish economy to leapfrog. (11)

Ireland could not have been upgraded to the G1 grouping solely through the imitation of G1 practices. It also did not have its own innovation capacity for this step upwards. Ireland's high-tech growth seems to have been exogenously driven in a given time period. The US high-tech firms chose Ireland as their springboard to enter European markets.

The structural indicators in Table 5 show that Ireland is a high-income country. It is, however, open to discussion whether it reveals the innovation-based growth patterns we would expect of a country that is a member of the first development club. Its ranking by GPI is much lower than its ranking by income per capita.

Slovenia: endogenous catching-up

Slovenia, as the second case study, in the 1991-2004 period increased the index of its GDP per capita (at current market prices, EUR-15=100) by 21 percentage points in current euro terms (from 30 in 1991 to 52 in 2004) and by 12 percentage points in PPS terms (from 60 in 1991 to 72 in 2004) (EC, 2004). Slovenia has narrowed the gap with the average of the EU-15 and, as such, it represents a case of successful catching up. Slovenia achieved this endogenous catching up through high, but not too high, investment, by a high level of educational attainment, by its considerable R&D efforts, and by the radical restructuring of industries where global competition from developing countries was the strongest (eg textile and shoe industry). Slovenia reveals (Table 5) good structural indicators for its development stage; many of them are comparable to those of Austria and Ireland. Slovenia does not create new technology. It implements and upgrades the technologies from other medium-level-technology countries and achieves high export growth. Slovenia's growth rate is endogenously driven. It is important to note that Slovenia's ranking by income per capita coincides with its ranking by growth pattern index. Slovenia is following the mainstream growth pattern of its income grouping.

Dynamics of development groupings due to the global competition

The dynamics of European development groupings cannot be understood within the European framework alone. The impact of countries from other continents has to be taken into account. There are two sources of outside competition. One comes from the world's most developed area, for example America. The other comes from developing areas, for example East Asia (China, India and others), which are on a similar or even lower level of development to the European countries of the G3 grouping (eg factor- and investment-based growth).

Competition from the world's most developed area (eg the USA)

For the developed part of EU member states the USA is a benchmark of economic performance. An overview of the EU's and the US' performances in GDP per capita over the last five decades is a useful starting point (Table 3). At the outset, the EU enjoyed a period of strong convergence towards the US level of GDP per capita, with its GDP per capita level rising from less than 50% of the US level in the 1950s to close to 70% by the 1970s. Over the subsequent period up to 1995 the convergence process stalled. The convergence process went into reverse (Commission 2003, p. 101). The stalling of the convergence process at a relatively low level of convergence was an obvious concern for EU policy-makers during this period (European Council, 2000; Kok, 2004; Sapir et al., 2004).

As the developed core Western European countries are competing and benchmarking themselves with the USA, the medium-developed countries of the SWE, and less-developed countries of CEE and SEE are looking to the US for sources of growth in order to catch up (in the case of G2) and keep up (in the case of G3) in the European environment. They do not face competition from America. They are simply too far behind. It may happen that they would attract some FDI from America in order to serve markets in Europe. The existence of the lead development club outside Europe may contribute to the convergence of development groupings in Europe. This thesis was illustrated with the case of Ireland.

Global competition from developing countries (eg China, India)

The second type of competition confronting European countries comes from East Asia's developing countries (South Korea, China, India, etc). This competition affects the development groupings in Europe differently. It provides global competition in sectors where CWE used to be good but is losing its competitiveness, for example low-technology and labour-intensive sectors. Due to global competition the countries of the first development club have to cut their employment in these sectors and reallocate production facilities to countries of other European countries and to East Asia. On the other hand, they are strengthening their high-technology sectors in order to avoid direct competition. Since the growth of the development club is mostly endogenous, foreign and global competition is boosting the competitiveness of the countries in this club.

The European countries of the G3 grouping face global competition in a different way. The competition from East Asia is direct and frontal. East Asia produces similar products, with similar technology and even cheaper labour. They all target the same markets, the markets of America and Europe. The European periphery cannot escape this competition by switching over to higher technological products. It has to face the global competition right away. If it loses, it will be driven away from the markets and from the sources of FDI. Such global competition may drive a wedge between the core and the periphery in Europe. It may slow down the catching-up process in Europe and may even contribute to the divergence of some countries of a third development grouping in Europe.


For a development club to be recognised as such three conditions should be fulfilled: similar levels of income per capita, convergence to the common steady state, and a similar pattern of growth. Based on 1955-2000 data, the group of countries in G1, for example Core Western European countries, could be described as a development club in Europe. They form the first development club with their high incomes per capita that converge, and with high technology, knowledge- and innovation-based growth patterns. They form a club because there is no transfer of high technology toward lower-placed groupings. There is no movement of high-skilled labour towards lower-placed groupings. The first development club is exclusive in terms of high technology, R&D activity and highly educated people. These features are not transferable to lower-placed groupings. Individual countries from the second development grouping may shift towards the first grouping or club, but only with some original undertakings.

Other countries do not form any development club. They may be brought together into groupings. The group of countries of G3 are diversified and reveal some common patterns. I would describe this group as a third development grouping with a growth pattern of factor- and investment-based growth stages. (12) Countries of South Western Europe (group G2 in Table 1 or G2 in Table 2) form an intermediate grouping. They do not form a stable grouping. Both of these groupings do not converge in income per capita within groupings and nor do they reveal exclusive growth patterns. They, therefore, do not form a development club.

We were able to identify three development groupings in Europe, and one of them can be identified as a development club. Membership of a convergence club (where incomes per capita converge) is not identical to membership of a development club (where the incomes per capita and growth patterns of countries converge as well). Italy and Ireland have high income per capita, while indicators of their growth patterns indicate that their growth patterns differ from the mainstream growth pattern of the development club. This is a tentative conclusion which should be further researched. The experience implies that there are different paths to the upper income group.

Frontier and high-technology, R&D activities, and highly educated and trained labour are located in countries of the first development club. They do not move freely outside the confines of the club. They do not move downward towards lower-placed development groupings. This feature institutes the club characteristics of a development grouping. Besides this, there is a brain drain from the lower to highest levels of development groupings which reinforces the club position of the first development grouping. The free movement of labour in Europe aggravates the divergence between the first and other development groupings in Europe because it is only in one direction, from the lower to higher level.


Chatterji (1992) showed (on sample of 109 countries) that there are two mutually exclusive convergence clubs, one for the 'rich' and one for the 'poor'. He distinguished between high and low convergence clubs. Similarly, Quah (1996) noticed twin-peak dynamics: there is a group of rich countries collecting together, and a group of poor countries collecting together, and a middle income class that is vanishing. He explains the twin-peak dynamics with varying degrees of capital market imperfections. I would rather attribute development club dynamics to high-technology and high-skilled labour market imperfections. Howit and Mayer-Foulkes (2005) distinguish convergence clubs by growth patterns, which is similar to my position. They predict that countries will fall into three convergence clubs: group 1--R&D; group 2--implementation, and group 3--stagnation. At the same time, they expect that due to technology transfers the R&D and implementation club would converge, yet the stagnation club would diverge because technology transfer does not reach this group of countries. My framework does not predict convergence between the first and second groups, although individual countries may switch between groupings, and it does leave open the divergence of a third grouping. Pryor's (2005) institutionally defined economic systems of the OECD countries could be complementary to my development club/groupings by indicating toward institutional patterns. He found four clusters of economic systems of the OECD nations in 1990, which are relevant to my analysis: South European (Greece, Italy, Portugal, Spain), West European, Nordic nations, and the Anglo-Saxon group (Switzerland, Ireland, UK). What is important for my analysis is that the South European group also appears in my analysis, and that three economic system clusters appear in the first development club.


In the post-2000 period differences in levels of economic development among European countries have been large: larger, perhaps, than ever before in recent history. There are two distinct groupings of countries according to their level of development: developed and less developed countries. These two groupings of countries face different development challenges. Western Europe (or the developed part of Europe) is lagging behind the most developed and competitive country in the world--the USA. Western Europe would like to re-establish the economic position of Europe as one of the most competitive and dynamic economic areas in the world. This aim is reflected in the Lisbon agenda. The benchmark for growth in this part of Europe is the most developed part of the world, for example the USA (Kok, 2004; Sapir et al., 2004). The integration of Europe is an instrument for enhancing the economic power of Europe in the global context.

Central and South East Europe (or the less-developed part of Europe) face the challenge of economic development. These European countries have in the past five decades experienced the processes of catching up, convergence and divergence in economic development. The benchmark for this part of Europe is some average of Europe. The frontier development grouping is simply too distant to be a benchmark for these countries. Most countries in this grouping see joining the EU a form of development aid and as an opportunity for the faster catching up with the rest of Europe.

We have argued that the growth patterns of the developed economies of core Western Europe form a development club and differ fundamentally from the growth patterns of the less-developed economies of South East Europe as a developing grouping. We imply, therefore, that the growth policy of economies in the development club would fundamentally differ from the growth policy that is appropriate for the economies of the third European grouping. Accordingly, benchmarking, cherry-picking and best practices must be put into the context of limited transferability.

The existence of three distinct groupings of countries in terms of growth characteristics imply that more than one set of growth policies is needed in Europe. So far, the EU formally has one growth policy: the structural policy and Lisbon strategy are meant for all countries. Implications of our framework would be that the Lisbon agenda is more suitable for the first grouping of countries. A third grouping of countries needs a separate growth policy, eg structural and cohesion policy with elements of development policy.


Aghion, P and Howitt, P. 2005: Appropriate growth policy: a unifying framework. Mimeo, Harvard University-Brown University, 9 August 2005.

Barro, R and Sala-i-Martin, X. 1992: Convergence. Journal of Political Economy 100(2): 223-251.

Baumol, W. 1986: Productivity growth, convergence, and welfare: What the long-run data show. American Economic Review 76(5): 1072-1085.

Baumol, W and Wolff, E. 1988: Productivity growth, convergence, and welfare: Reply. The American Economic Review 78(5): 1155-1159.

Buchanan, J. 1965: An economic theory of clubs. Economica 32: 1-14.

Chatterji, M. 1992: Convergence clubs and endogenous growth. Oxford Review of Economic Policy 4: 57-69.

Desdoigts, A. 1999: Patterns of economic development and the formation of clubs. Journal of Economic Growth 4: 305-330.

Durlauf, S and Johnson, P. 1995: Multiple regimes and cross-country growth behaviour. Journal of Applied Econometrics 10: 365-384.

European Commission. 2003: The EU economy: 2003 review. European Economy, No. 6, DG Economic and Financial Affairs.

European Commission. 2004: The EU economy: 2004 review. European Economy, No. 6, DG Economic and Financial Affairs.

European Council. 2000: Presidency Conclusions. Lisbon European Council, 23 and 24 March.

Hausmann, R, Pritchett, L and Rodrik, D. 2005: Growth accelerations. Journal of Economic Growth 10: 303-329.

Howit, P and Mayer-Foulkes, D. 2005: R&D, implementation, and stagnation: A Schumpeterian theory of convergence clubs. Journal of Money, Credit, and Banking 37(1): 147-177.

Islam, N. 2004: What have we learnt from the convergence debate?. In: George, D, Oxley, L and Carlaw, K (eds). Surveys in Economic Growth. Theory and Empirics. Blackwell Publishing: Oxford pp. 83-135.

Kok, W. 2004: Facing the challenge. The Lisbon strategy for growth and employment. Report from the High Level Group chaired by Wim Kok. European Communities: Luxembourg.

Maddison, A. 2003: The World Economy: Historical Statistics. Development Centre of the Organisation for Economic Co-operation and Development: Paris.

Mankiw, G. 1995: The growth of nations. Brookings Papers on Economic Activity 1: 275-326.

Mankiw, G, Romer, D and Weil, D. 1992: A contribution to the empirics of economic growth. Quarterly Journal of Economics 107 (2): 407-437.

OECD. 2005: OECD Science, Technology and Industry Scoreboard, 2005 Edition, OECD: Paris.

Porter, M. 1990: The Competitive Advantage of Nations. The Free Press: New York.

Pritchett, L. 1997: Divergence, big time. Journal of Economic Perspectives 11(3): 3-17.

Pryor, E 2005: Market economic systems. Journal of Comparative Economics 33(1): 25-46.

Pryor, F. 2006: Economic systems of developing nations. Comparative Economic Studies 48: 77-99.

Quah, D. 1996: Twin peaks: Growth and convergence in model of distribution dynamics. Economic Journal 106: 1045-1055.

Rodrik, D. 2005: Growth strategies. In: Aghion, P and Durlauf, S (eds). Handbook of Economic Growth, Volume 1A. Elsevier, North-Holland: Amsterdam pp. 967-1014.

Rostow, WW. 1960: The Stages of Economic Growth. Cambridge University Press: Cambridge.

Sapir, A, Philippe, A, Giuseppe, B, Martin, H, Jean, P-F, Dariusz, R, Jose, V, Helen, W, with Marco, B, Marion, N and Peter, MS. 2004: An Agenda for Growing Europe: The Sapir Report. Oxford University Press: New York.

Schumpeter, J. 1934: The Theory of Economic Development. Oxford University Press: New York.

Solow, R. 1956: A contribution to the theory of economic growth. Quarterly Journal of Economics 70: 65-94.

TrendChart. 2005: European Innovation Scoreboard 2005. Comparative Analysis of Innovation Performance. European Commission: Brussels.

UNCTAD. 2002: World Investment Report 2002. United Nations: New York.

UNDP. 2003: Human Development Report 2003. United Nations Development Programme: New York.

(1) I use the term 'development club' in an analogy to 'convergence clubs', which is frequently used in the growth literature. However, it has a broader meaning. Howit and Mayer-Foulkes (2005) use the term 'growth club', yet they do not define it. My terms 'club' and 'grouping" are distinct from the term 'economic system' used by Pryor (2005, 2006) who forms clusters based on 'indicators of complementary institutions'.

(2) 0ECD (OECD Science, Technology and Industry Scoreboard, 2003) divided manufacturing industries into high-technology, medium-high-technology, medium-low-technology and low-technology groups after ranking industries according to their R&D intensities (eg R&D expenditures divided by value added, and R&D expenditures divided by production (OECD, 2005)).

(3) Not all European countries are included. Countries with a population of less than 1 million are excluded due to a lack of consistent data collection by some international institutions (eg the World Bank).

(4) For the period after 2000, the countries Serbia and Montenegro, Bosnia and Herzegovina, and the Republic of Moldova are not included for practical statistical reasons. I have not included Russia in SEE because it could be a separate group, which I am not concerned about in this article.

(5) As mentioned, some countries were left out. The grouping of the left out countries would be as follows. Luxembourg would be included in CWE, Cyprus and Malta would be grouped in SWE, while Bosnia and Herzegovina, Serbia and Montenegro and Moldova would be included in the SEE group.

(6) Sigma convergence is when the dispersion of cross-section levels of GDP per capita diminishes over time.

(7) Beta convergence is when in a cross-section country data regression of growth rates on initial levels of GDP per capita, the coefficient on initial levels is negative.

(8) Sources for these indicators are World Development Indicators 2004 (World Bank), TrendChart, OECD (2005), Eurostat, World Investment Report 2004 (United Nations).

(9) Dimension index = (actual value minimum value)/(maximum value-minimum value).

(10) The USA forms a development club of its own. This is the top development club. This aspect of analysis is not elaborated in this article. I refer to this fact in continuation of the article.

(11) Let me illustrate this thesis. UNCTAD (2002, p. 20) developed a transnationality index to compare the transnationality of countries in which TNCs operate. The transnationality index of a country is based on two FDI variables and two variables related to foreign firms' operations

in a host country: (1) FDI inflows as a percentage of gross fixed capital formation; (2) FDI inward stock as a percentage of GDP; (3) value added by foreign affiliates as a percentage of GDP; and (4) employment by foreign affiliates as a percentage of total employment. The simple average of these four shares results in the Transnationality Index of a host country. The transnationality index may be used as a measure of a country's openness and its ability to imitate, learn from, and transfer development factors from other countries. Europe's most developed host economy in 1999 was Belgium and Luxembourg (a transnationality index of 66.0), followed by Ireland (35.7), Sweden (33.1), the Netherlands (25.2) and Denmark (17.9) (UNCTAD 2002, pp. 20-21). Total exports of Ireland increased almost eightfold between 1985 and 2000, from $10 billion in 1985 to $ 76 billion in 2000. FDI inflows rose even faster, from $ 164 million in 1985 to $ 24 billion in 2000 (UNCTAD 2002, p. 172). Foreign affiliates accounted for a 90% share of Irish exports in 1999. Ireland was a country that was open to transnational operations.

(12) This supposition is not verified in the present paper.


Faculty of Economics, University of Ljubljana, Kardeljeva ploscad 17, 1000 Ljubljana SI-Slovenia. E-mail:
Table 1: Groups of countries according to per capita
GDP, 1955-1985 (1990 international dollars)

Country 1955 1970 1985

G1: Core Western
 Europe (CWE)
 Austria, Belgium,
 Denmark, Finland, France,
 Germany (a),
 Italy, Netherlands, Norway,
 Sweden, Switzerland,
 United Kingdom
 Average (b) 6,658 10,924 15,479

G2: South Western
 Europe (SWE)
 Ireland, Greece, Portugal, 9,160
 Average 2,922 6,051

G3: Eastern Europe (EE)
 Albania, Bulgaria,
 Czechoslovakia, Hungary,
 Poland, Romania, Yugoslavia
 Average 7 EE 2,495 4,315 5,892
USSR 3,313 5,575 6,701

Source: Maddison, 2003.

(a) Germany with its 1991 borders.

(b) Unweighted arithmetic mean.

Table 2: Groups of countries according to GDP per capita in 2003

Countries PPP USD Index

G1: Core Western Europe (CWE)
 Austria, Belgium, Denmark,
 Finland, France, Germany,
 Ireland, Italy, Netherlands,
 Sweden, UK, Norway, Switzerland
 Average GDP per capita (a) 29,946 100
G2: South Western Europe (SWE)
 Greece, Portugal, Spain,
 Average GDP per capita 19,905 66
G3/1: Central Eastern
 Europe (CEE)
 Czech Rep., Estonia, Hungary,
 Latvia, Lithuania,
 Poland, Slovakia, Croatia
 Average GDP per capita 12,801 43
G3/2: South East Europe (SEE)
 Bulgaria, Romania, Macedonia,
 Albania, Ukraine, Belarus
 Average GDP per capita 6,322 21

Source: Human Development Report 2003, UNDP, 2003.

(a) Unweighted arithmetic averages.

Table 3: Indexes of per capita GDP 1955-2000 (international
dollars) by G-groups

Groups 1955 1970 1980 1990 2000

G0 (USA) 173 150 132 138 142
G1 (12 CWE) (a) 100 100 100 100 100
G2 (4 SWE) (a) 46 60 62 66 80
G3 (HE and USSR) (a) 47 50 44 37 --

Source: based on Maddison, 2003.

(a) From Table 1.

Table 4: Groupings of countries according to the growth
patterns index (GPI), around 2000

 Range of
Countries values of GPI

G1: Core Western Europe (CWE) 0.46-1.0
 Austria, Belgium, Denmark,
 Finland, France, Germany,
 Netherlands, Sweden, UK,
 Norway, Switzerland
G2: South-Western and Baltic 0.31-0.45
 countries (SWB)
 Italy, Ireland, Spain, Portugal,
 Slovenia, Estonia, Latvia
G3: Central and Eastern 0.0-0.30
 Europe (CEE)
 Greece, Czech Republic, Hungary,
 Poland, Lithuania, Slovakia,
 Bulgaria, Romania

Table 5: Structural indicators of the relative growth
performances of Austria, Ireland and Slovenia

Indicator Year AT IE SI

1. GDP per capita (PPS, 2003 111 122 71
2. Labour productivity 2003 96 120 70
 (PPP, EU-15=100) (a)
3. Educational attainment 2003 84 86 91
 (20-24) (%)
4. R&D expenditure 2003 2.2 1.2 1.5
 (% of GDP)
5. Business investment 2003 20.3 19.7 21.1
 (% GDP)
6. Gross enrolment ratio, 2001/02 57 47 61
 tertiary (b)
7. Researchers in R&D per 1990-2001 2313 2190 2258
 million people
8. Information and 2002 5.3 4.0 4.9
 expenditures, %
 of GDP
9. Ratio of inward/outward 2003 1.0 5.8 2.4
 stock of FDI (1)

Sources: Kok, 2004; World Development Indicators 2004
(World Bank); (1) World Investment Report 2004 (UNCTAD).

(a) Per person employed.

(b) % of relevant age group.
COPYRIGHT 2007 Association for Comparative Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Senjur, Marjan
Publication:Comparative Economic Studies
Geographic Code:4E
Date:Dec 1, 2007
Previous Article:Compounded effects of external crises on GDP growth.
Next Article:The rise of the transaction sector in the Bulgarian economy.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters