International Competitiveness and Comparative Advantage: A Survey ... [PDF]

International Competitiveness and Comparative Advantage: A Survey and a Proposal for Measurement Eckhard Siggel

1

International Competitiveness and Comparative Advantage: A Survey and a Proposal for Measurement by Eckhard Siggel, Department of Economics, Concordia University, Montreal 1455 de Maisonneuve Blvd., Montreal, Quebec, Canada, H3G 1M8 Tel: (514) 848 2424 ext. 3914, e-mail: [email protected]

Abstract The concept of competitiveness, or competitive advantage, has been given numerous interpretations and tends to be ambiguous. Comparative advantage, although rigorously defined in the Ricardian trade model, is also subject to different interpretations when extended beyond the classical trade theory and, particularly, with respect to its measurement. The present paper first reviews the literature that deals with definitions and measurements of these concepts, distinguishing their main characteristics, such as macro vs. micro, static vs. dynamic, positive vs. normative, ex-ante vs. ex-post, as well as the different uses made of the proposed measures. Second, the paper proposes an integrated approach, in which it is demonstrated how competitiveness and comparative advantage are best related to each other and how they differ. The proposed measurement serves the purpose of quantifying the different sources of competitiveness. It is shown how it can be applied to guide policy makers in their evaluation of trade and industrial policies. The latter aspect is illustrated by reference to several applied studies using the method of analysis in the context of policy reforms in India, Mali, Kenya and Uganda. Key words: Competitiveness, Comparative advantage, Price distortions, Policy analysis. JEL classification: F10, F13, F14

1

The author wishes to acknowledge the many fertile discussions of the subject with John Cockburn, as well as the useful comments by two anonymous referees of this journal.

Introduction Although the principle of comparative advantage is one of the oldest and most important concepts in economics, there is some disagreement in the literature about its precise meaning, scope and measurement. The concept of competitiveness or competitive advantage is even more ambiguous, as it has found a wide range of interpretations. The objectives of this paper are to survey the literature, to identify some contradictions and to suggest an approach that avoids such contradictions. This approach is a method of analysis that integrates both concepts and links them to each other. It is an analysis of the sources of competitiveness, which measures competitiveness by using three variants of a unit cost indicator. They are measures of competitiveness in the domestic market, international competitiveness and comparative advantage, respectively. Following a short enunciation of some diverging views of comparative advantage in the first section, the literature survey of section 2 focuses on a limited set of concepts of competitiveness. It attempts to demonstrate that the concept of competitiveness has a legitimate place in the trade literature, although many economists seem to avoid it. The third and fourth sections propose a general measure of comparative and competitive cost advantage that is valid for multiple goods and multiple factors of production. In section 3 the measure is derived from the Ricardian principle of comparative advantage based on one factor and two products. This leads first to the domestic resource cost criterion and finally to a total unit cost criterion. In section 4 the unit cost criterion is further developed as a method of analysis of the sources of competitiveness. One of the merits of this unit cost criterion is that, by replacing the equilibrium prices by observable market prices (including distortions) in costs and revenues, the comparative advantage criterion becomes one of competitive advantage. Section 5 presents a short summary of three applied studies of manufacturing in Mali, India and Kenya using the proposed method.

1

Diverging views of comparative advantage

The principle of comparative advantage is historically tied to the framework of the Ricardian trade model. In that framework the principle is unambiguous as long as only 2

two products are involved. The extension to n products was first demonstrated by Dornbusch, Fischer and Samuelson (1977) and is now standard in most of our textbooks. One crucial aspect of measurement in the context of n goods is not always emphasized, however. It is the necessity to use equilibrium prices in the measurement of costs. If markets are not in equilibrium, wage hikes or currency appreciation could, at least temporarily, eliminate the ability to export any or all goods. This shows that cost comparisons based on market prices cannot be the basis of comparative advantage. It leads to the important distinction between the concepts of comparative and competitive advantage. When costs are measured in terms of (possibly distorted) market prices, we deal with competitive advantage, which is the same as cost competitiveness. The same measurement becomes one of comparative advantage when equilibrium prices are used. It is common in the empirical trade literature to measure comparative advantage by help of the Balassa (1965) index of ‘revealed comparative advantage’ (RCA). This measure reflects the success in exporting of countries relative to a world-wide norm. Exports can result from subsidies or other incentives provided, for instance by exchange rate misalignment. Such incentives can explain competitiveness, but not comparative advantage. Therefore, the RCA index measures competitiveness rather than comparative advantage. The trade literature has also assigned an unduly limited interpretation to comparative advantage. The principle is usually explained in terms of the Ricardian model of two goods, one factor of production and two countries. Some attention has been given to its extension to more than two goods, as mentioned above. The extension to more factors of production, including intermediate inputs has been somewhat neglected. It is true that a large body of literature has dealt with the determinants of trade using different trade theory models. With respect to the generality of the principle of comparative advantage, two positions can be identified in the literature. The first is that comparative advantage is limited to Ricardian and Hecksher-Ohlin-type trade and does not apply to other forms of trade, such as intra-industry trade. This opinion is expressed in most of our current textbooks and seems to be the dominant one. The second is a more general interpretation of the principle. It suggests that a producer has comparative advantage if his/her production costs in terms of equilibrium factor prices are lower than 3

those of an international competitor, irrespective of what the sources of the cost advantage are. The source of that advantage can be the abundance (cheapness) of either primary or intermediate inputs (extended Heckscher-Ohlin), or the use of different technology (Ricardo), or the production at larger scale (Krugman), or any combination of the former sources, such as in the product cycle model (Vernon). The point is that once we move from the two-good to the n-goods world, the measurement of comparative advantage requires the use of monetary costs at equilibrium prices. The principle remains then valid for any number of goods and factors, as well as for any kind of trade. The former and narrower interpretation is based on the assumption that costs are always defined by production functions and factor prices, in other words as potential and not actual costs. This approach reflects an ex-ante view, which is useful for modeling and predicting. It is less useful for the analysis of past performance. In our view, comparative advantage should not be interpreted exclusively as an ex-ante concept. Intra-industry trade can be explained by economies of scale combined with monopolistic competition based on product differentiation, as in the Krugman model. It results from the fact that producers in different countries, while using the same production function, compete against the producers of similar (differentiated) products at a level of costs that is lowered by large scale production. This in turn is the consequence of supplying a larger than the domestic market. Since the attainment of large scale is essential, they maximize profits by concentrating production of certain brands in one location and by exporting. One example of the wider definition of comparative advantage is the product cycle theory developed by Vernon (1966 and 1979). This model attributes comparative advantage in the production of new products to sources that may change over the life cycle of the products. In the early stages of the cycle comparative advantage is based on the first-come advantage of the country in which the product was developed. The cost advantage shifts to lower cost countries, where their advantage is likely to come from Heckscher-Ohlin type factor abundance. In further stages scale economies and learning effects may become the source of comparative advantage. This model has been associated with dynamic comparative advantage (cf. Appleyard and Field, 2001), but it is debatable whether changes of its source and the development of comparative advantage over time deserve the attribute of dynamic. What matter in the present discussion is that 4

economies of scale can be and have been considered a source of comparative advantage. Indeed, that source of comparative advantage is not necessarily different from that of Ricardian comparative advantage. Ricardian comparative advantage leaves it open whether it results from greater skills, more capital or other factors. Large-scale production usually also requires a different technique than small-scale production. It is unrealistic to separate technology from scale. Krugman & Obstfeld in their celebrated textbook (2000) restrict comparative advantage to the factor proportions interpretation, and do not apply it to the general case, in which it is based on costs. They argue, for instance, that “intra-industry trade (manufactures for manufactures) does not reflect comparative advantage” (p.137). In our view, any trade that results in welfare gains needs to be based on comparative advantage, irrespective of the nature of its sources. The sources may be Ricardian productivity differences (or different technologies), or they may be differences in factor endowments that are reflected by factor cost differentials. But they may also include differences in the scale of production, for firms that share the same cost function. Denying this kind of trade comparative advantage is contrary to its very principle and reduces the power of the concept. Admittedly, its measurement is more difficult the wider its application, and it necessitates the consideration of costs. But that is also the case for Ricardian comparative advantage, when moving from the two-good to the multiple-good model. It is true that in classical and neo-classical trade theory comparative advantage is established by cost comparison under autarchy and under trade, whereas in intra-industry trade the cost advantage is captured when markets have been joined and firms have relocated their plants. But this difference does not justify denying comparative advantage to this modern form of trade. Only the sources of comparative advantage are different, just as different as the sources of Ricardian and Heckscher-Ohlin trade.

2

Concepts of competitiveness

A large number of concepts of competitiveness has been proposed in the economic and business literature. This owes to the fact that competitiveness, unlike comparative advantage, has not been defined rigorously in the early economic literature. Thus, over time and after many attempts of definition, it has become a somewhat ambiguous 5

concept.2 Some authors use the term synonymously or in a similar way as comparative advantage, others view it as an economy-wide characteristic. In the present survey we contrast first microeconomic with economy-wide (macroeconomic) interpretations. This distinction is the most fundamental one because the simple extension of micro concepts to the macro level poses problems, as is obvious in the case of comparative advantage. Then we discuss the number and kind of dimensions that the various concepts of competitiveness integrate and measure. A further distinguishing characteristic is the basis for comparison, which ranges from a single firm or industry to groups of countries or the rest of the world. Bilateral and multi-lateral comparisons always require data from one or more foreign countries, whereas unilateral concepts are based on the data of a single country. Further distinguishing attributes discussed are the static or dynamic nature, positive vs. normative, deterministic vs. stochastic, as well as ex-post vs. ex-ante characteristics. It would lie beyond the scope of this survey to analyse the theoretical foundations of each of the concepts, but the perspective taken here is the one of market-oriented economic systems and it emphasizes cost and price competitiveness. It does not cover the whole range of concepts, including those that focus on technological indicators. For such a survey the reader may be referred to the recent survey by Cantwell (2005). The policy perspectives underlying many of the surveyed concepts are addressed in section 2.7 on objectives of competitiveness measurement.

2.1

Macro versus micro-economic concepts

The most controversial kind of competitiveness indicator is the macroeconomic one, although it may be the most popular one. The microeconomic concepts, on the other hand, which apply to single producers or industries3, are less controversial despite the variety of indicators within this group. Although in recent years the public discourse has focused more on the macro concept, it is less well established in economic theory than the micro concept. Countries may compete for market share or for foreign investment, 2

Spence, for instance, recognizes the concept’s complexity and refrains from adopting an exact definition; and so do various other contributors in (Spence & Hazard, 1988). 3 More precisely, the industry level can be referred to as the meso-economic level of analysis, but for

6

but the attributes of stability, good government and profitable investment opportunities, are better summarized as a favorable business climate than competitiveness, in our view. The perhaps best known version of the macro concept is the World Competitiveness Index computed and published yearly by the World Economic Forum and Institute of Management Development (WEF/IMD, annual since 1995). The index is the basis for an international ranking of countries in terms of their business climate. It is a composite of a large number of attributes condensed into a single index. It may serve a useful purpose to international investors, but its theoretical base and, especially, its aggregation method are problematic. A second interpretation of macroeconomic competitiveness that is more in line with the original meaning of the term is an aggregate of the microeconomic concept. In this view an economy is deemed to be competitive if it harbors a large number of internationally competitive enterprises and industries. In other words, it must perform strongly in exports. This idea underlies the concept used by Dollar and Wolff (1993)4, who propose to measure it in terms of productivity, both labour and total factor productivity. Similar approaches are the concepts proposed by Hatsopoulos, Krugman and Summers (1988) and by Markusen (1992). A third approach, equally at the macro level, is that of the real exchange rate (RER), as well as the real effective exchange rate (REER), proposed and used by authors within the IMF (Lipschitz, McDonald, 1991; Marsh, Tokarick, 1994). Since the RER implicitly compares the nominal exchange rate with the purchasing power parity rate, it measures the degree of currency misalignment based on the purchasing power parity assumption. Under-valuation enhances and overvaluation reduces the international competitiveness of domestic producers. This indicator is clearly macroeconomic, but it has also been used as a micro-level concept, by using the price indices of single industries rather than economy-wide price indices (cf. Helleiner, 1991). At the macro level it is essentially a monetary indicator, capturing the distortion of the currency value,

simplicity we consider the meso-economic level as part of the micro level. 4 They define: “A competitive nation is one that can succeed in international trade via high technology and productivity, with accompanying high income and wages”, (Dollar, Wolff, 1993, p.3).

7

rather than factors of real competitiveness, although those are not unrelated to the currency misalignment. It is interesting to note that one of the strongest condemnations of the concept of competitiveness has come from Krugman, who likened it to a “dangerous obsession” (Krugman, 1994) and the debate over it “a matter of time-honored fallacies about international trade being dressed up in new and pretentious rhetoric” (Krugman, 1996). But Krugman has also used the concept, as we saw in the earlier reference, in an economy-wide sense as well as in the industry-specific sense (cf. Krugman, Hatsopoulos, 1987). Microeconomic concepts and indicators of competitiveness have a more solid theoretical base because they focus on the essential characteristics of producers in competition for market share and profits or the ability to export. This ability can be measured by the size or increase of market share (e.g. Mandeng, 1991), by export performance (e.g. Balassa, 1965), by price ratios (e.g. Durand, Giorno,1987), cost competitiveness (e.g. Turner, Gollup, 1997; Siggel, Cockburn, 1995), or by more complex and multi-dimensional indicators (e.g. Porter, 1990; Buckley et al.,1992; Oral, 1993). These indicators differ from each other in terms of various characteristics, especially in terms of the number of dimensions they focus on.

2.2

One- versus multi-dimensional concepts

The number of dimensions included in the measurement of competitiveness is a reflection of the complexity of the concept, but it is also a source of ambiguity. In one of the earlier ground-laying conferences on International Productivity and Competitiveness, Hickman recognized that even among the authors of this conference the concept had a variety of dimensions (Hickman, 1992, p. 6)5. He referred to price competitiveness as the most pervasive concept, but focused on the unit labour cost criterion. Since unit labour cost equals the product of the wage rate, labour productivity and the exchange rate (in international comparisons) one may debate whether this makes it a more-than-onedimensional indicator. Here we classify it as one-dimensional concept because it focuses

5

The conference took place in October 1988 in Palo Alto, California.

8

on the dimension of the real cost of labour, irrespective of how that can be further broken down. Turner and Gollub (1997) and Golub (2000) have also proposed and used this indicator for the measurement of competitiveness at the industry level. An example of a two-dimensional concept is the indicator proposed by Hatsopoulos, Krugman and Summers (1990), which postulates that competitiveness of economies translates into trade balance with rising living standard or real income. The authors argue that export success can always be achieved at the cost of diminished real income, which is then not a reflection of competitiveness. Only if export success occurs with a constant level of welfare can an economy be said to be competitive. A similar concept was proposed by Markusen (1992) and applied by the U.S. Presidential Commission of Competitiveness. The real exchange rate, as well as the real effective exchange rate can be seen as uni-dimensional indicators because they measure the degree of misalignment of the currency. Such misalignment enhances or reduces international competitiveness. RERbased indicators have been proposed by several authors, and with various nuances. The macroeconomic version compares the purchasing power parity (PPP) value with the going exchange rate. If the former exceeds the latter, the currency is undervalued, which contributes to competitiveness. For this conclusion to hold it is necessary, however, that the PPP value be based on price indices of a comparable basket of goods or price indices representing all goods and services, such as the CPI. The RER transforms into a microeconomic indicator of price competitiveness, when the price indices used refer to single products or industries. Various microeconomic studies of single-industry competitiveness use as indicator the relative industry price (relative to one or more foreign competitors), with the exchange rate translating it into the same currency. This kind of indicator has been proposed by Durand and Giorno (1987) and Helleiner (1991), and is used extensively by the OECD-Economics and Statistics Department. Formally it resembles the real exchange rate, except that the prices relate to one industry only, instead of reflecting the general price level. This is an important difference because the RER and REER as macro indicators make a statement about the currency value, whereas as micro indicator they

9

measure the price competitiveness of a country in a particular industry. In that sense this micro-type indicator is a one-dimensional one. The same type of indicator was also used, under the name of purchasing power parity (PPP), by Jorgenson and Kuroda (1995)6 in a study for Industry Canada (Jorgenson, Lee, 2000). The term PPP may not be ideally chosen, because it has a very specific meaning with respect to the currency and requires the use of general price indices, as in international comparisons of GDP (cf. Kravis, Heston, Summers, 1978). Multi-dimensional indicators are the most popular in the business economics literature. Among the micro-economic indicators measuring more than one dimension, the perhaps best known one is the concept of Porter (1990), but Buckley et al. (1992) and Oral (1993) are also interesting attempts to capture more than one dimension of the concept. According to Porter there are four main determinants of competitiveness of enterprises: their strategy, structure and rivalry, the demand conditions they face, the factor supply conditions they encounter, and the conditions of related industries. In fact, it is a multitude of factors that influence the competitiveness of producers, but Porter models them by classifying them under these four facets of a diamond. These facets can be viewed as dimensions along which competitiveness can be measured. Porter’s concept has attracted very wide interest in the business and political communities, perhaps because of its comprehensive nature.

2.3

Bases of comparison

The notion of competitiveness always involves comparisons between producers or industries in different countries. The choice of a foreign competitor is important not only for the numerical outcome, but also for the meaning of the concept. While the Ricardian comparative advantage criterion is most meaningful when applied to a specific foreign country, it is also valid for comparisons with the rest of the world. However, cost and productivity data, as well as transport costs are rarely available for groups of countries. It is possible to replace the foreign cost data by the free-trade prices of imports at the point

6

The authors attribute their method of analysis to Jorgenson and Nishimizu (1978).

10

of entry, i.e. border prices. This procedure assumes that imports will normally come from the lowest bidder, so that the import price is a reflection of best practice. While this approach sets a high standard for competitiveness, the inclusion of transport costs to the point of entry does the opposite, it lowers the standard. A price comparison between the output price of a domestic producer and the free-trade price of a close substitute can then be taken as an indicator of competitiveness vis-à-vis a group of countries in a market close to the home country. In other words, it is a multilateral indicator of price competitiveness. Both, the Domestic Resource Cost (DRC) ratio and the full unit cost ratios proposed later in this paper are using this approach. It permits the analyst to draw conclusions about competitiveness without using data from other than one particular country. However, this does not exclude the possibility of using the same indicator in bilateral comparisons. For that reason, the characteristic of laterality is not included in the table 1, which surveys the concepts according their characteristics. A more precise multilateral indicator takes the different exchange rate valuations of the main trading partners of a country into account. The indicator proposed by Helleiner (1991) serves this purpose as it equals the weighted average of the real exchange rates vis-à-vis several foreign competitor countries, where the weights are the market shares of these foreign competitors in the specific product market. The indicator therefore amounts to a product-specific real effective exchange rate (REER). While the RER is a bilateral indicator, the REER is multilateral as it considers all important trading partners of a country.

2.4

Static vs. dynamic concepts

One of the major limitations of the principle of comparative advantage, both for explaining and predicting trade patterns, is its static nature. Since comparative advantage tends to change over time, the prediction of future trade patterns would require knowledge of how comparative advantage itself changes over time. This is possible by help of knowing the sources or determinants of trade, which is the substance of much of trade theory and policy. Firms or industries that acquire a new and promising technology can be said to enjoy dynamic comparative advantage as they are likely to gain market share. Similar considerations apply to competitiveness. If it is measured by market share 11

for instance, the change in market share can be taken as a dynamic indicator of competitiveness. It is a more pertinent measure of competitive strength than the market share itself. At the micro level Krugman & Hatsopoulos (1987) have used market share as indicator of U.S. competitiveness in manufacturing. Competitive advantage may then be reflected by increased market share. They observed that the international market share of U.S. manufacturing declined in the early 1980s, which they analysed as the reflection of declining competitiveness. The concepts proposed by Porter (1990) and Buckley, Cass and Prescott (1992) include both static and dynamic aspects. The latter integrates three characteristics of firms or industries, their competitive performance, competitive potential and their process of management. It is multi-dimensional and dynamic in that it focuses on the potential of firms to adjust to exogenous changes and to achieve comparative advantage in the future. At the macro level, the indicators proposed by Hatsopoulos, Krugman, Summers (1988) and Markusen (1992) are dynamic in that they measure changes in welfare over time. The real exchange rate (RER) index also conveys a message concerning changes over time. On the other hand, it can be used to measure the degree of currency misalignment at one moment in time. Depending on how, precisely, it is used, it can have a static or dynamic interpretation. Other concepts in the dynamic category are those proposed by Aiginger (1998) and Pitelis (2003). Both refer to entire economies and are based on subjective criteria, such as “factor incomes in line with the aspiration level seen as satisfactory by the people” (Aiginger, 1998, p. 164) or “the improvement of a subjectively defined welfare indicator for a country” (Pitelis, 2003, p. 210). Obviously, these concepts are problematic with regard to measurement and international comparison.

2.5

Deterministic vs. stochastic and ex-post vs. ex-ante concepts

Most concepts proposed in the literature are deterministic in the sense that they measure costs, prices, market shares etc, which are observed and reflect actual performance. A few concepts, however, focus on notions of welfare or potential performance that are not 12

directly observable. They depend on a number of other variables, which are deemed to determine competitiveness according to models of a stochastic nature. These variables are then either chosen as proxies, or they serve as data in statistical analysis of the unobservable indicators. Such concepts add an element of uncertainty about the relevance and statistical significance of the proposed model. Closely related to this distinction is the one between ex-post and ex-ante. Ex-post concepts tend to be deterministic, whereas exante concepts tend to be stochastic in nature. An example of a macro-economic, stochastic and ex-ante type indicator of competitiveness is the one proposed by Fagerberg (1988). This author attempts to explain the market share of a country in world markets by three variables: technical competitiveness reflected by R&D expenditure and patent applications, price competitiveness as reflected by the terms of trade and unit labour cost, and the output capacity. All variables are expressed in terms of growth rates, which makes the indicator also a dynamic one. Among the micro-economic concepts, one that is typically stochastic in nature is the criterion proposed by Swann and Taghavi (1992). The authors infer competitiveness by comparing the expected price of products, based on quality attributes, with the actual price, where the expected price is regressed on the measured quality attributes. If the expected price exceeds the actual one, this is considered as evidence of competitive advantage. While ex-post concepts reflect the outcome of competition, ex-ante concepts measure the readiness for competition or potential competitiveness. For instance large or growing market share is the outcome of successful competition and therefore ex-post competitiveness. An example of such measures focusing on market share is the revealed comparative advantage (RCA) measure proposed by Balassa (1965). According to this indicator, a country has comparative advantage in a particular product if its exports of the product, relative to world exports of the product, are larger than the country’s market share in total exports. Such a larger than proportional market share can, however, result from subsidies or other price distortions instead of high productivity. For that reason the RCA criterion measures competitiveness rather than comparative advantage.

13

The indicators of cost and price competitiveness, as well as various composite indicators based on potential (e.g. Porter, 1990; Buckley et al., 1992; and Oral , 1993), are ex-ante in the sense that they demonstrate a capacity to compete. Ex-post indicators have the advantage that they prove de facto the point of successful competition, but rarely reveal the sources of competitiveness. Ex-ante indicators, on the other hand, tend to show the main sources of the advantage, although the advantage may not yet be realized.

2.6

Positive vs. normative concepts

Closely related to the distinction of deterministic and stochastic concepts is the one between positive and normative concepts. Positive concepts measure what is and normative ones measure what should be. While positive concepts are based on observable reality and do not involve value judgements, normative concepts do. For a firm or industry, strong export performance is de facto evidence of international competitiveness. For a whole economy, strong export performance is not per se evidence of competitiveness because it may be achieved at a cost of income loss. Normative concepts are more frequent in the macro context. A good example is the concept proposed by Markusen (1992) and also attributed to the U.S. Presidential Commission on Industrial Competitiveness. Marcusen defines: "A country is competitive if it maintains a growth rate of real income equal to that of its trading partners in an environment of free and (long run) balanced trade". This criterion is normative in the same sense as most welfare indicators are.

2.7

Different objectives pursued in measuring competitiveness

The wide variety of concepts in the economic and business literature can to some extent be explained by the purpose for which specific indicators have been designed. The measurement of competitiveness differs, depending on whether it is undertaken for the purpose of policy analysis within a specific country, or whether it is used for international comparisons of the business environment. An example of the latter kind is the World Competitiveness Index (WEF/IMD), which is used to rank countries according to a number of conditions that are known to be favorable for business development. Such 14

a ranking can guide international investors in their choice of investment locations, as well as banks in their evaluation of country-specific risks. It can also inform policy makers about the weaknesses and strengths of specific country environments. The main users of competitiveness indicators are government departments designing industrial policies, negotiating trade agreements or writing development plans. Private sector agents like banks and industrial corporations also gain from the analysis, as do the semi-private institutions like chambers of commerce, trade unions and business associations. The methods of economic policy analysis used by government departments and research organizations are of a wide variety. In recent years computable general equilibrium (CGE) models have gained prominence for this purpose, but they are costly, due to their complexity. The literature relevant to this wider purpose is not surveyed here, but Francois and Reinert (1997) provide such a survey, as far as trade policy analysis is concerned. CGE models may not be sufficiently detailed, given their comprehensive nature, and they may not be fully reliable, due to the many simplifying assumptions on which they rely. Competitive and comparative advantage indicators are more manageable and less costly, while permitting the analyst to monitor the impact of policy changes on the competitive environment, the cost structure of industries and the resulting market structure. The concepts and measurements of competitiveness discussed here serve a similar purpose as the CGE modeling approach, but avoid the numerous assumptions about the behavior of economic agents required by CGE models. Industrial policies and strategies can benefit societies if they induce comparative advantage and reduce obstacles to competitiveness. The pursuit of industrial policies in the context of globalization and trade openness is based on the assumptions that comparative advantage changes with the structural transformations of economies, and that it can be developed or enhanced, especially through human resource development. Table 1: Concepts and indicators of competitiveness and their characteristics Proposing author Concept characteristics Measurement criterion or or organization (1) (2) (3) (4) (5) Indicator Macro concepts Lipschitz/McDonald (1991), u s/d det p ea real exchange rate, real Marsh, Tokarick (1994), IMF effective exch.rate 15

Hatsopoulos, Krugman, Summers (1988) Markusen (1992), Dollar/Wolff (1993) Fagerberg (1988) Sharpe (1986) WEF/IMD (annual since 1995) Aiginger (1998), Pitelis (2003) Micro concepts Balassa (1965) Bruno (1965) Buckley et al. (1992) Durand/Giorno (1987), OECD Helleiner (1989) Hickman (1992) Jorgenson, Kuroda (1992) Krugman, Hatsopoulos (1987) Mandeng (1991) Oral (1993) Porter (1990) Siggel/Cockburn (1995) Swann/Taghavi (1992) Turner/Gollub (1997)

two

s/d

det

n

ea

two

s/d

det

n

ea

two m u m m

s d s s/d d

det sto det det det

p n p p n

ea ea ep ea ea

trade balance with rising real income real income growth with free balanced trade productivity market share increase market share world compet. index n.a.

u u m u two u u u u m m u m u

s s d s d s s s/d s/d s s/d s s s

det det det det det det det det det det det det sto det

p p n p p p p p p p p p p p

ep ea ea ea ea ea ea ep ep ea ea ea ea ea

revealed comp.advantage domestic resource cost composite, multi-variable price competitiveness real effective exch. rate unit labour cost price competitiveness market share, change market share, change indust. mastery, unit cost composite, multi-variable full unit cost price/product attribute relative unit labour cost

Concept characteristics: (1) dimensions of concept: u = uni, two, m=multi-dimensional (2) s = static or d = dynamic nature of concept (3) det = deterministic, sto = stochastic nature of concept (4) p = positive, n = normative nature of concept (5) ep = ex-post or ea = ex-ante nature of concept

For instance, an industry may have potential comparative advantage, which is presently not realized because of either skill shortages or infrastructure deficiencies making the industry non-competitive. Industrial policies would then remove such obstacles to competitiveness. An important task in this respect is to identify the sources of comparative advantage and to enhance those activities exhibiting such potential. Clearly, the prediction of future or potential comparative advantage is difficult, because all measurements are based on past performance. Indicators that are of the ex-ante type, although based on measurements of the past, can serve this task. For instance, an indicator that measures the speed of technical change can be used to predict potential comparative advantage. The concepts of competitive and comparative advantage discussed here and their main characteristics are shown in the following table for the purpose of comparison. The table also includes the unit cost concept proposed and derived in the remainder of this 16

paper, which integrates the concepts of comparative advantage and competitiveness. The table provides an impression of the great variety of concepts proposed in the literature and makes an attempt of classifying them according to their main characteristics.

3

The measurement of comparative advantage

In the earlier discussion of conflicting views it was argued that the most popular method of measuring comparative advantage, the one of "revealed" comparative advantage (Balassa, 1965), does not really measure comparative advantage in a rigorous sense. Instead, it is really a measure of competitive advantage, as it reflects export performance, which can result from either real factors or from price distortions and subsidies. The only well-known measure that qualifies as a true measure of comparative advantage is the Domestic Resource Cost (DRC) criterion proposed first by Bruno (1965), discussed by Balassa and Schydlowsky (1968), Bruno (1972), Krueger (72), Srinivasan and Bhagwati (1978) and applied by many others. The next section shows how it is derived from Ricardian comparative advantage, and the section 3.2 extends the DRC criterion to full unit costs.

3.1

From Ricardian comparative advantage to the DRC criterion

In order to extend the Ricardian concept of comparative advantage to its general meaning, there are two obstacles to overcome: first, the extension from two products to n products and, second, the extension to more factors of production from the one-factor Ricardian context. The first of these generalizations was proposed by Dornbusch, Fischer and Samuelson in a seminal paper (Dornbusch, Fischer, Samuelson, 1977). This paper demonstrates that comparative advantage of an industry exists if its labour cost is inferior to the labour cost of the same industry in the foreign country: (1)

a w < a* w*

where a and a* are the domestic and foreign unit labour coefficients, and w and w* are the domestic and foreign equilibrium wages, respectively. It is important to note that this

17

condition is one of absolute advantage as long as the wage rates are not equilibrium rates. The same condition can be written as (2)

w/w* < a*/a,

which means that the relative equilibrium wage must be lower than the domestic labour productivity relative to that of the foreign country. This comparison between the relative wage and the relative productivity can then be used to extend the principle to n products, simply by ranking all industries according to their relative productivities and by using the relative wage as the dividing line between comparative advantage and disadvantage: (3)

ai/a*i > w/w* > aj/a*j

where i represents all products with comparative advantage and j those with disadvantage. The second step of generalization concerns the extension to other factors of production. It follows from the first step by extension of logic. If simple comparison of labour cost is sufficient for comparative advantage, as long as only labour costs matter and equilibrium wages are used, then the same must be true for costs including that of other factors. The DRC criterion makes this assumption by comparing the cost of labour and capital based on their shadow prices with the corresponding cost of the foreign competitor: (4)

a ws + k rs < a* ws* + k* rs*,

where k and k* are the capital coefficients and rs and rs* are the domestic and foreign shadow interest rates, and ws and ws* are the domestic and foreign shadow wage rates. In fact, the DRC criterion goes even one step further since it replaces the foreign factor cost by domestic value added at free-trade prices. The idea of this substitution is that unless one wants to measure comparative advantage vis-à-vis a particular country, the foreign competitor represents the rest of the world and under competitive trade conditions its cost is then equal to the free-trade price of comparable imports. Therefore the DRC criterion of comparative advantage takes on the following form: (5)

(a ws + k rs)/VAw < 1,

where VAw is the free-trade value added, which is assumed to equal its shadow value7.

7

This assumption may not hold if the currency is misaligned. In this case the value must be adjusted.

18

3.2

From DRC to total unit costs

Although the DRC is an acceptable indicator of comparative advantage, it falls short of a perfect measure in two respects. First, intermediate inputs, both tradable and non-tradable ones, may also contribute to comparative advantage. A full-cost-based indicator should be superior to a value-added-based one. One could argue, of course, that tradable inputs are always available at international prices and should not be a source of comparative advantage. However, domestic resource abundance and transport costs may be responsible for import prices to be higher than domestic prices of tradable inputs, thus making them a potential source of comparative advantage. This is even more obvious in case of non-traded inputs. It has also been argued that intermediate inputs can be taken into account in the DRC method by breaking them down into labour and capital contributions using input-output tables. This, however, would make the DRC inappropriate as a cost-benefit indicator at a given stage of transformation (industry). Second, we know that protection leads to inefficient use of resources, including intermediate inputs. Therefore, the value-added computed by eliminating all price distortions is still efficiency-distorted, so that DRC gives a biased message about comparative advantage. An unbiased measure of comparative advantage compares the total domestic cost expressed in terms of shadow prices with the shadow price of output, where the latter is usually taken as the free-trade price of comparable imports. It needs to be adjusted, however, if the exchange rate is misaligned. The resulting criterion of comparative advantage is then that total domestic costs per unit of a product (in shadow prices, TCs) must be inferior to its shadow price: (6)

TCs/Ps < 1,

where Ps is the international or free-trade price corrected for misalignment of the relevant exchange rate. The division of total costs by the shadow price, or the shadow value of output (VOs), to allow for heterogeneous output, is called here the unit cost ratio at shadow prices (UCs). The corresponding criterion of comparative advantage can then be written as: (7)

UCs = TCs/VOs < 1

19

Our use of the term unit cost is a particular one, since it is not understood as costs per physical unit of output. The costs per unit of output are rarely comparable among various competitors, due to the usual differences in product mix and quality. The unit cost ratio (UC) overcomes this problem by assuming that such quality differences are reflected by the prices.

4

Measurement of competitiveness and its sources

The unit cost ratio thus derived from Ricardian comparative advantage, by extending it to multiple goods and factors of production including intermediate inputs, is now easily adjusted to become an indicator of competitiveness. When the equilibrium or shadow prices of both costs and output values are replaced by market prices, we obtain a criterion of domestic competitiveness: (7)

UCd = TCd/VOd < 1

Total costs in domestic prices (TCd) are the ones reported by firms and include various price distortions. The distortions may also include quantity distortions that can be attributed to external factors. For instance, excess costs in energy, transport and communications inputs may be due to failure of public utilities to provide these services at normal costs. Since the analysis of these distortions is somewhat more difficult, we shall get back to this problem at a later stage (section 2.14). The output value (VOd), also includes the distortions of nominal protection and the exchange rate misalignment. UCd is essentially a measure of profitability behind protective barriers to trade. It is, however, not equal to the actual profit margin, because the total cost (TCd) includes the opportunity cost of capital at its going market interest rate. When UCd equals one, this means that the market price covers all costs at market prices, including the opportunity cost of own capital. When UCd is inferior to one, the firm realizes excess profits beyond the market interest rate. These profits can reflect either high productivity of the factors of production or particularly low equilibrium prices of these factors, or price distortions, and all of these are sources of competitive advantage. In addition to domestic competitiveness (profitability) and comparative advantage, there is also the notion of international or export competitiveness. A firm is export competitive if its total cost in market prices is inferior to the international free20

trade price of output. The corresponding unit cost ratio and criterion of export competitiveness is then

(8)

UCx = TCd/VOw < 1,

where international output value differs from its shadow value only in terms of the exchange rate misalignment. The analysis according to the sources of competitiveness proceeds then in the following accounting framework. First, total costs in shadow prices consist of four components, tradable inputs (VITs), non-tradable inputs (VINs), labour cost (LCs) and capital cost (KCs). when each of them is divided by the total value of output, the sum of these components is then equal to UCs. Next, all price and non-price distortions on the cost side, all divided by VOs, are added in order to sum up to UCx. Finally, adding the output distortion to UCx leads to UCd, the indicator of domestic competitiveness.

(9)

VITs/VOs +VINs/VOs +LCs/VOs +KCs/VOs __________ =TCs/VOs=UCs +dpe +dpj +dpje +dw +dpk +dr +ds ___________ =TCd/VOx=UCx +dpp ___________ =TCd/VOd=UCd

(Shadow unit cost of tradable inputs) (Shadow unit cost of non-tradable inputs) (Shadow unit cost of labour inputs) (Shadow unit cost of capital inputs) (Total unit cost at shadow prices) (Exchange rate distortion of output) (Tradable input price distortion) (Exchange rate distortion of tradable inputs) (Wage rate distortion) (Capital goods price distortion) (Interest rate distortion) (Direct subsidy, negative) (Total cost per unit of output at internat. Prices) (Output price distortion) (Total unit cost at domestic prices)

In other words, total unit cost in shadow prices (indicator of comparative advantage) plus all cost distortions adds up to unit cost per output value at free-trade prices (indicator of export competitiveness), plus the output price distortion adds up to unit cost in domestic prices (indicator of domestic competitiveness). This accounting framework can be used to identify, with some limitations, the sources of competitiveness. Since the distortions are individually

21

expressed as proportions of the output value, the highest proportions indicate the strongest influence on unit costs. One of the limitations of this analysis is that for the primary factors of production this accounting procedure cannot be used to measure the relative importance of these factors as sources of comparative advantage. This is because the use of different techniques of production implies the possibility of substitution between the factors. This handicap can be overcome by a stochastic approach discussed later. The method, however, permits to assess the importance of various distortions relative to the shadow value of output. The measurement of distortions deserves, therefore, special attention.

4.1

The measurement of distortions

The distortions that are important in the context of analysing international competitiveness are generally price distortions that are caused by various policies. Subsidies are treated here as price distortions. Price margins stemming from monopoly power are unlikely to be substantial in trade regimes regulated only by the tariff. At the output level we discuss only traded goods, because non-traded outputs are usually assumed to be distortion-free. At the input level, we distinguish traded goods, non-traded goods and services, labour and capital, as seen in the decomposition. Price distortions can be estimated for traded goods and primary factors of production, following the principles of social cost-benefit analysis and shadow pricing. For non-traded intermediate inputs, such as electricity, water and communications, however, it is difficult to distinguish between price distortions and quantity distortions, because shadow prices are not easily available for them. In their cases, other methods of estimation are applied and these may contain quality and quantity aspects.

4.11

Output price distortions

The most important distortion affecting the unit cost ratio UCd is normally the tariff. The difference between the domestic (ex-factory) price and the border (free-trade) price of comparable imports is known as the nominal rate of protection (NRP). In order to capture this distortion adequately one needs to compare the domestic and border prices. In praxis, this is a very difficult exercise and requires the assistance of the producing firms, which

22

know best the product mix and the quality characteristics of the products. The tariff can be used as NRP only in the absence of quantitative restrictions, exemptions and smuggeling.

4.12

Tradable input price distortions Since the number of material inputs is usually large, it is impossible to deflate them,

one by one, for distortions caused by tariffs and quantitative restrictions. One way of dealing with the problem is to identify the main imported materials and to estimate their average NRP by using price comparisons or the collection rate. Applying the average of input NRP to the total value of tradable inputs means assuming that the prices of locally bought tradable inputs include distortions that are similar to those of actually imported inputs. For some inputs, particularly agricultural inputs, this assumption may not hold. If natural resources are purchased at prices that are lower than international ones then they may be regarded as a source of comparative advantage. In other words, their price differential from international prices is not treated as a distortion.

4.13

Exchange rate misalignment Currency misalignment is easy to detect, but difficult to measure, in regimes of fixed

exchange rates. It is even more difficult to argue and measure misalignment when the exchange rate is flexible and market-determined. It is known, nevertheless, that central banks can influence the exchange rate by various kinds of intervention that are sometimes referred to as “leaning towards a higher or lower rate”, depending on the nature of interventions. The effect of over-valuation on unit cost ratios is twofold. On the output side, it raises the shadow value of output above its value at international free-trade prices, and thereby lowers the unit cost ratio UCs. In other words, if the penalty of over-valuation did not exist, comparative advantage would be enhanced. On the input side, the opposite effect occurs for tradable inputs.

4.14

Non-traded inputs

For non-traded inputs (electricity, water, local transport, communications, subcontracts for repair and other services) it is often assumed that these costs are undistorted. In reality, however, these costs can be distorted and impose heavy penalties on producers. It is more 23

difficult, however to separate the distortions from the undistorted costs, because of the nonexistence of international prices. The distortions may also be either at the price level or at the level of quantities consumed. A frequently observed example of the latter kind of distortion are extra costs due to frequent electricity blackouts, or the excessive cost of telephone lines because of their failure to function. The estimation of such distortions may involve ad-hoc approaches and assumptions about “normal” cost levels that are not well established in the literature. It can be, nevertheless, important to include such estimates, especially in order to measure the full extent of policy-induced incentives or disincentives.

4.15

Labour cost distortions

Wages and salaries are in some methods of competitiveness analysis the principal component (cf. the method of relative unit labour cost, e.g. Turner, Gollub, 1997). In our experience with cost structures in low-income countries, their share in total cost of manufacturing industries is often as low as 15 percent. For skilled occupations it is usual to make the simplifying assumption that the paid remuneration reflects the social opportunity cost of these services. For unskilled labour, on the other hand, it is usual to estimate shadow wages and to treat the difference as cost distortion. In addition to, or instead of, the more complex methods of shadow wage determination, one can resort to a simplified estimation based on comparisons with informal sector wages or the marginal product of agriculture, which may be regarded as proxies of the social opportunity cost of unskilled labour.

4.16

Capital cost distortion

The cost of capital includes three components, the financial cost of borrowed capital, the opportunity cost of own capital, and depreciation of the total capital stock. The financial social opportunity cost of capital is taken to equal the shadow interest rate applied to the total value of fixed and non-fixed assets at purchase prices. The shadow interest rate can be computed in various ways. First, assuming that capital is internationally mobile, it is obtained by adding an inflation differential to the international rate (LIBOR). Second, based on the international interest parity condition, one can add to the LIBOR the expected rate of currency depreciation.

24

In contrast, the market opportunity cost of own capital is the market interest rate. The difference between these two interest rates constitutes the main capital cost distortion. The annual depreciation, as reported by the firms, can be treated as undistorted for simplicity. A further source of distortion is due to the payment of import duties on imported capital goods, but since the age structure of the capital stock is not easily available, the estimation of this distortion can be very difficult.

4.2

The sources of competitiveness and comparative advantage

In the framework outlined above it is clear that comparative advantage, as identified by the shadow unit cost criterion, is the main source of competitiveness. We propose to call it real competitiveness as opposed to the “nominal” nature of the components representing price distortions. While the relative importance of price distortions can be measured as their share in total unit cost, this is not possible for the factor components of the shadow unit cost. To analyse the relative importance of tradable inputs, non-tradable inputs, labour and capital as sources of comparative advantage it is necessary to adopt a different approach. The reason for this is the existence of substitution possibilities between these four kinds of costs. A low labour cost component alone cannot signal competitiveness because it may result from high capital intensity or high intermediate input intensity. Therefore, while we can decompose the shadow unit cost into its four components, we cannot rank the four factors in terms of their relative importance for competitiveness, unless we have numerous cost observations of a single industry, which would enable us to compute cost functions. In the absence of such data, i.e. when the costs of an industry or firm are available from a single observation, one can nevertheless pool the data from various industries and examine whether there is a significant tendency for any factor to induce lower or higher unit costs, whenever it is used intensively. For this purpose we plot the unit cost of each factor (for instance unit capital cost, KCs/VOs) over the total unit cost ratio (UCs) on the horizontal axis, as shown for capital in figure 1. If the linear trend line of the unit capital cost has a negative intercept, i.e. if the line KCs/VOs = a + b UCs cuts the average line KCs/VOs = β UCs from below, one

25

can infer that unit costs tend to be high in industries that are particularly capital-intensive. This means that there is a tendency for high unit costs to be related to high capital costs.

Figure 1: Determining the sources of comparative advantage KCs/VOs = a + b UCs KCs/VOs = β UCs

KCs/VOs

average KCs/VOs

average UCs

UCs= TCs/VOs

If the observations, which represent either firms within an industry or different industries within the manufacturing sector, are contained within the marked area of plots in figure 1, and if the trend line is statistically significant, then we conclude in the case shown by figure 1 that capital cost tends to be associated with comparative disadvantage. In this case one or more of the other factors’ trend lines should have a positive intercept, signaling the respective factor to be a source of comparative advantage. If the same kind of statistical analysis is done for unit costs at market prices, it shows the tendency of any one factor to be a source of competitiveness or the lack of it. Such an analysis is less meaningful because it does not reveal whether this tendency is caused by factor price distortions or by the resource allocation decision of the firms. In the case based on shadow unit costs (the case depicted above) we would interpret the observation of capital being associated with high unit costs as evidence of inappropriate technology choice. The industry or sector would then gain comparative advantage by using either more labour-intensive techniques or more intermediate inputs, depending on the significance and slopes of the other trend lines.

26

5

Application of the method in four empirical studies

The method of analysis outlined above has been applied in four case studies of the manufacturing sectors in Mali and Cote d'Ivoire (Cockburn et al. 1999), in India (Siggel, 2001), in Kenya (Siggel, Ikiara, Nganda, 2002) and in Uganda (Siggel, Ssemogerere, 2004). It is also being applied by economists in Vietnam, Iran and South Africa. In this section we provide only a short summary of the conclusions in three of our studies in order to demonstrate what kind of conclusions can be obtained.

5.1

Mali

In this study the enterprises of the (very small) manufacturing sector of Mali were compared with firms of the corresponding industries in Cote d'Ivoire, in order to assess the competitive position of Mali vis-à-vis her neighbour country, which also harbours its access route to the seaport of Abijan. The Malian manufacturers were found to be generally competitive on the protected Malian market, but not on the Ivorian export market. This reflects a lack of comparative advantage coupled with protectionist trade policies in both countries. Thus, Mali's manufacturers face a serious challenge on their local market under further trade liberalization and regional integration. However, the cost reductions required of Malian manufacturers in order to compete locally were only in the order of only 10-20% in most cases. Furthermore, the textile industry shows some potential to become a competitive exporter. Mali's comparative advantage was found to lie in labour-intensive and inputintensive activities, owing mainly to the lower cost of labour and despite lower labour productivity. Capital costs appeared to be higher in Mali, due to lower rates of capacity utilization and smaller scale. Mali’s structure of protection appeared to result in a moderate net cost increase of roughly 8% of production value. This harms potential exporters, and in the context of regional integration also harms the suppliers of the domestic market. Policy recommendations arising from this analysis include first and foremost the lowering of tariffs on intermediate inputs. Financial liberalization and deepening would also help to lower market interest rates and thus allow Malian firms to turn to more stable, long-term sources of financing. Furthermore, investment in national

27

and regional transport infrastructure were found to require special attention to offset the negative effects of Mali' landlocked status.

5.2

India

The study of manufacturing in India was undertaken with the objective to examine whether the beginning trade and industrial reforms of the 1980s had resulted in changes in India's international competitiveness. Therefore, data of 1980/81 were compared with those of 1987/88 and 1991/92. The study was handicapped, however, by the unavailability of protection data in the first and last periods. Therefore, conclusions concerning the whole time period could only be drawn at the level of domestic competitiveness, i.e. behind tariff and non-tariff walls. The full results on export competitiveness and comparative advantage could only be obtained for the year 1987/88. Another difference from the African studies is that for India published data of the annual manufacturing survey were used, which means that the data were more comprehensive in terms of coverage, but less rich in terms of variables and qualitative inputs from interviews. The study revealed that the manufacturing sector remained still very strongly distorted, as India's liberalization was only beginning and the more important policy changes occurred only in the early 1990s. Besides high tariffs and remaining non-tariff barriers, the exchange rate was strongly over-valued by about 60%, which imposed a great hurdle to international competitiveness. The study led to recommendations of further policy reforms, including currency realignment, trade liberalization, fiscal and financial reforms, most of which have been implemented since then. The impact of these reforms are being analysed in a present follow-up study.

5.3

Kenya

In the study of Kenya, recently collected data were used together with data from the mid1980s, in many cases from the same firms. This permitted to examine how Kenya's policy reforms of the late 1980s and early 1990s had affected the country’s manufacturing sector. Interestingly, the sector seems to have gained somewhat in comparative advantage, but lost in terms of competitiveness. This unexpected result is explained by 28

the fact that several major distortions act as obstacles to competitiveness, in particular the very substantial interest rate distortion, caused to a large extent by government policy. Other distortions that affect the manufacturers of Kenya are the high cost of transport, communication and energy, all due to a deteriorating infrastructure of this country. Therefore, the policies of trade and foreign exchange liberalization, although still incomplete, have shown some positive effects in terms of movement towards comparative advantage, but cannot lead to the expected benefits as long as domestic distortions are on the rise. 6

Conclusion

In the survey of the literature it was demonstrated that the most consistent interpretation of the concept of competitiveness is the microeconomic notion of cost competitiveness. It is related to comparative advantage, but differs from it in that it includes in its sources the various price distortions in output value and costs, whereas comparative advantage is based on real factors only. It was demonstrated that the separation of price distortions from the real sources of competitiveness by using shadow prices leads to a framework, in which comparative advantage and competitiveness are clearly defined, closely related but non-identical concepts. The paper has also provided some limited evidence of how this framework can be applied in the analysis of policy reforms in order to show whether such reforms have contributed to an improvement of the resource allocation.

References Aiginger, K. (1998), “A framework for evaluating the dynamic competitiveness of countries”, Structural Change and Economic Dynamics 9, 159-188. Alavi, H., (1990), “International Competitiveness: Determinants and Indicators”, Industry and Energy Department Working Paper No. 29, The World Bank. Appleyard, D.R., A.J. Field jr. (2001), International Economics, 4.Edition, McGraw-Hill Irwin. Balassa, B., (1965), “Trade Liberalization and ‘Revealed’ Comparative Advantage”, Manchester School, No. 33, May.

29

Balassa, B., D. Schydlowsky, (1968), “Effective tariffs, Domestic Costs of Foreign Exchange, and the Equilibrium Exchange Rate”, Journal of Political Economy, 76, No. 3, 348-360. Bruno, M., (1965), “The Optimal Selection of Export-promoting and Import-substituting Projects”, in Planning the External Sector: Techniques, Problems and Policies, New York, United Nations. Bruno, M. (1972), “Domestic Resource Costs and Effective Protection: Clarification and Synthesis”, Journal of Political Economy, vol. 80, 1. Buckley, P.J. et al. (1988), Measures of International Competitiveness: A Critical Survey”, Journal of Marketing Management, 4 (2), 175-200. Buckley, P.J., et al., (1992), “The meaning of competitiveness”, Ch. 2 of Servicing International Markets: Competitive Strategy of Firms, Blackwell, Oxford. Cantwell, J. (2005), “Survey of innovation and competitiveness”, in Oxford Handbook of Innovation, Fagerberg, Mowery and Nelson (eds.), Oxford University Press. Cockburn, J., E. Siggel, M. Coulibaly, S. Vezina, (1999), “Measuring Competitiveness and its Sources: The Case of Mali’s Manufacturing Sector”, Canadian Journal of Development Studies, Vol. XX, No.3. DANE, (1996), Productividad, Competitividad e Internacionalizacion de la Economia, conference volume, Serie de Competitividad, Bogota. Dollar, D., N.E. Wolff (1993), Competitiveness, Convergence and International Specialization, The MIT Press. Dornbusch, R., S. Fisher, P. Samuelson, (1977), “Comparative Advantage, Trade and Payments in a Ricardian Model with a Continuum of Goods”, American Economic Review, 67, No. 5, December, 823-839. Durand, M., C. Giorno, (1987), “Indicators of International Competitiveness: Conceptual Aspects and Evaluation”, OECD Economic Studies, No. 9 Autumn, 147-182. Fagerberg, J. (1988), “International Competitiveness”, The Economic Journal, No. 391, (June), 355-374. Feenstra, R.C.(1989), (ed.), Trade Policies for International Competitiveness, NBER confernce volume, University of Chicago Press. Francis, A. (1989), “The Concept5 of Competitiveness” in A. Francis and P.K. Tharakan (eds.), Competitiveness of European Industry, Routledge, London. Francois, J.F., K.A. Reinert, (1997), Applied Methods for Trade Policy Analysis: A Handbook, Cambridge University Press. Golub, S., (2000), “South Africa’s International Cost Competitiveness”, TIPS working paper No. 14.

30

Hatsopoulos, G., P. Krugman, L. Summers, (1988), “U.S. Competitiveness: Beyond the Trade Deficit”, Science, 241, 299-307. Helleiner, G. K. (1991), “Increasing international competitiveness: A conceptual framework”, in Y. Wen, J. Sengupta, (eds.) Increasing the International Competitiveness of Exports from Caribean Countries, Economic Development Institute , World Bank. Hickman, B.G., (1992), “International Productivity and Competitiveness: An Overview”, in Hickman (ed.), International Productivity and Competitiveness, Oxford University Press (New York). Jorgenson, D.W., M.Kuroda, (1992), “Productivity and International Competitiveness in Japan and the United States, 1960-1985”, in Hickman (ed.), International Productivity and Competitiveness, Oxford University Press (New York). Jorgenson, D.W., F.C.Lee, (2000), Industry-level Productivity and International Competitiveness between Canada and the United States, Industry Canada Research Monograph. Jorgenson, D.W., M. Nishimizu, (1978), “U.S. and Japanese Economic Growth, 1952-1974”, Economic Journal, 88(352), 707-26. Kravis, I., R. E. Lipsey, (1971), Price Competitiveness in World Trade, Columbia University Press. Kravis, I.B., A. Heston, R. Summers, (1978), International Comparisons of Real Product and Purchasing Power, Johns Hopkins University Press, Baltimore (Md.). Krueger, A.O., (1972), “Evaluating Restrictionist Trade Regimes: Theory and Measurement”, Journal of Political Economy, vol. 80, No. 1, January/February. Krugman, P. (1994), “Competitiveness: A Dangerous Obsession”, Foreign Affairs, March/April. Krugman, P., (1996), ”Making Sense of the Competitiveness Debate”, Oxford Review of Economic Policy, vol 12, No. 3. Krugman, P. R., G.N. Hatsopoulos, (1987), “The Problem of U.S. Competitiveness in Manufacturing”, New-England Economic Review, January/February. Krugman, P., M. Obstfeld, (2000), International Economics: Theory and Policy, Addison Wesley, Longman Inc.. Lafay, G., (1987), “Avantage comparatif et compétitivité`, Economie prospective internationale, No. 29, premier trimestre. Leamer, Edward E. (1984), Sources of International Comparative Advantage, MIT Press, Cambridge.

31

Lipschitz L., D. McDonald, (1991), “Real Exchange Rates and Competitiveness: A Clarification of Concepts and Some Measurements for Europe”, IMF Working Paper, March. Lipsey, R.E., I.B.Kravis, (1987), “The Competitiveness and Comparative Advantage of U.S. Multinationals 1957-1984”, Banca Nazionale del Lavoro. Mandeng, O.J. (1991), “International competitiveness and specialization”, CEPAL Review, No. 45, December, 39-52. Markusen, James R., (1992), Productivity, Competitiveness, Trade Performance and Real Income: The Nexus Among Four Concepts, Minister of Supply and Services, Canada. Marsh, L.W., S.P. Tokarick, (1994), “Competitiveness Indicators: A Theoretical and Empirical Assesssment”, IMF Working Paper No. 94/29, March. Oral, M. (1993), “A methodology for competitiveness analysis and strategy formulation in glass industry”, European Journal of Operational Research, No. 68, 9-22. Pitelis, C.N. (2003), “Supply-Side Strategy for Productivity, Competitiveness and Convergence for the EU and the CEECs”, in Foreign Direct Investment in Central and Eastern Europe, (Marinova & Marinow, eds.), Ashgate, Aldershot (U.K.). Porter, M. (1990), “The competitive advantage of Nations”, Harvard Business Review, No.2, March/April. Sharpe, A. (1985), “Can Canada Compete?” Part 1: Measuring Canada’s Competitiveness, Canadian Business Review, Winter. Siggel, E., (2001), “India’s Trade Policy Reforms and Industry Competitiveness in the 1980s”, The World Economy, vol.24, No.2, February, 159-183. Siggel, E. and J. Cockburn, (1995), “International Competitiveness and its Sources: A Method of Development Policy Analysis”, Concordia University Department of Economics, Discussion Paper 9517. Siggel, E., G. Ikiara, B. Nganda, (2002), “Policy Reforms, Competitiveness and Prospects of Kenya’s Manufacturing Industries: 1984-1997”, Journal of African Finance and Economic Development, vol. 5, No. 1, Winter, 72-104. Siggel, E., G. Ssemogerere, (2004), “Uganda’s Policy Reforms, Industry Competitiveness and Regional Integration: A comparison with Kenya”, Journal of International Trade & Economic Development, Vol. 13, No. 3, September, 325-357. Spence, M., H. Hazard, (1988), (eds.) International Competitiveness, Ballinger (Cambridge, Mass.). Srinivasan, T.N., J. Bhagwati, (1978), “Shadow Prices for Project Selection in the Presence of Distortions: Effective Rates of Protection and Domestic Resource Costs”, Journal of Political Economy, vol. 86, no. 1.

32

Summers, L.H., (1988), “Tax Policy and International Competitiveness”, in A.M.Spence & H.A.Hazard, International Competitiveness, Ballinger (Cambridge, Mass.) Swann, P., M. Taghavi, (1992), Measuring Price and Quality Competitiveness, Avebury. ten Kate, A. (1996), “Competitividad Comercial de la Economia Mexicana, 1980-1995”, in DANE, Productividad, Competitividad e Internaciaonalizacion de la Economia, Bogota. Turner, A., S. Golub, (1997), “Towards a System of Unit Labor Cost-Based Competitiveness Indicators for Advanced, Developing and Transition Countries”, Staff Studies for the World Economic Outlook, IMF. ul Haque, I. (1991), (ed.), International Competitiveness, Interaction of the Public and the Private Sectors, Economic Development Institute of the World Bank. Vernon, R. (1966), “International Investment and International Trade in the Product Cycle”, Quarterly Journal of Economics, 80, No. 2, May. Vernon, R. (1979), “The Product Cycle Hypothesis in a New International Environment”, Oxford Bulletin of Economics and Statistics, 41, No. 4, November. WEF/IMD, (1995), The World Competitiveness Report, World Economic Forum and Institute for Management Development, Geneva and Lausanne.

33