single market: inferences from intra-industry trade patterns

of 20

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
20 pages
0 downs
: inferences from intra-industry trade patterns Marius Brülhart Department of Economics, University of Manchester, Manchester, UK and Robert J.R. Elliott Department of Economics, Manchester Metropolitan
: inferences from intra-industry trade patterns Marius Brülhart Department of Economics, University of Manchester, Manchester, UK and Robert J.R. Elliott Department of Economics, Manchester Metropolitan University, Manchester, UK Introduction During the negotiation and implementation stages of the 1992 programme, prominent economists anticipated that the pressures for industrial re-structuring among EU[1] countries would be considerably stronger than during previous episodes of European integration. Krugman (1987, p. 364) put it as follows: The question now is whether the further expansion of trade in progress will be equally easy to cope with. The unfortunate answer is, probably not. Greenaway and Hine (1991, p. 620) suggested that specialisation in Europe may have entered a new phase, and that this could pose greater problems for adjustment. These expectations were based on two main developments, both of which were connected to the phenomenon of intra-industry trade (IIT), the two-way exchange of goods with similar production requirements. First, the discovery of high and growing IIT levels in the 1970s had produced a wave of new thinking by trade theorists, which shifted the emphasis of the models away from country-specific trade determinants, generically termed comparative advantage, towards industry-specific factors such as increasing returns and external economies. Models of the new trade theory, although explaining the existence of IIT, generally predicted that a fall in trade barriers would promote concentration and re-location of industries near their largest markets[2]. Second, a popular, albeit loosely defined, hypothesis has emerged in the literature, according to which high levels of IIT were indicative of relatively low tradeinduced adjustment costs[3]. Some studies in the late 1980s found evidence of stagnating IIT growth, and therefore concluded that adjustment pressures were becoming more severe. The rapid growth of trade flows among EU countries (Table I) was therefore expected to result in growing factor-market friction The authors are grateful to the participants in the SPES project on Trade specialisation and market structure in Community for their co-operation, and to David Greenaway, Dermot McAleese, Chris Milner and two anonymous referees for useful comments. Journal of Economic Studies, Vol. 25 No. 3, 1998, pp , MCB University Press, Journal of Economic Studies 25,3 226 Table I. The post-war boom in intra-eu trade, , imports and exports Country ImportsExports Percentage EU ImportsExports Percentage EU Imports Exports Percentage EU Imports Exports Percentage EU Belgium-Luxembourg Denmark France Germany Greece Ireland Italy The Netherlands Portugal Spain United Kingdom European Union Country ImportsExports Percentage EU ImportsExports Percentage EU Imports Exports Percentage EU Imports Exports Percentage EU Belgium-Luxembourg Denmark France Germany Greece Ireland Italy The Netherlands Portugal Spain United Kingdom European Union Notes: All goods, current market prices, in percentage of GDP; percentage EU: proportion of intra-eu trade in total trade Source: EU Commission which in turn could fuel protectionist sentiment and undermine the integration project. This paper explores the validity of the second reason for the anticipation of growing adjustment pains in Europe[4]. Greenaway (1987) was first to find that IIT may have declined in the EU countries during the late 1970s. These suspicions were confirmed by Globerman and Dean (1990), Greenaway and Hine (1991) and Mucchielli (1988), which all observed IIT trend reversals for several OECD countries based on relatively small data sets. No comprehensive and recent evidence, however, has as yet been produced to verify these suggestions. Even if we did find confirmation of a generalised reversal in post-second World War IIT trends, it might be misleading to draw direct inferences from such a discovery on adjustment costs. The traditional interpretation of IIT has been challenged recently by work on the measurement of changes in IIT, now commonly referred to as marginal IIT (MIIT). The traditional Grubel-Lloyd (GL) index of IIT is a static measure, in the sense that it describes trade patterns for one time period. Hamilton and Kniest (1991) have shown that the observation of a high proportion of IIT does not justify a priori any prediction of the likely pattern of change in trade flows. Industrial adjustment, however, is a dynamic concept, relating to the re-allocation of resources over time. Even an observed increase in static IIT levels between two periods could hide a very uneven change in trade flows, attendant with inter- rather than intra-industry adjustment, and with asymmetric rather than symmetric changes. Hence, it might be misleading to infer from rises/falls in IIT that trade expansion entails relatively low/high adjustment costs. We therefore complement the analysis of traditional IIT measures by a survey of MIIT patterns in the EU. The paper is organised as follows. In section two, we explore the theoretical link between IIT and adjustment costs. Section three presents the relevant measurements. A comprehensive and statistically disaggregated analysis of IIT and MIIT patterns among the countries of the EU is provided in section four. This survey covers the period from , the latter being the deadline for the implementation of the as well as the last year for which customs data on intra-eu trade flows are available. In section five, we report results of some exercises which examine explicitly the relationship between MIIT and structural change. Section six concludes by summarising and commenting on our main findings Intra-industry trade and adjustment Some theoretical aspects Definition of adjustment costs. Adjustment costs arise from temporary inefficiencies when markets fail to clear instantaneously in response to changes in demand or supply conditions. The type of adjustment this study examines is the welfare loss arising from trade-induced factor switching costs or unemployment. Journal of Economic Studies 25,3 228 Two important clarifications need to complement this definition. First, with respect to trade-induced unemployment, economic theory suggests that this is strictly temporary in nature. While the duration of the adjustment process cannot be determined a priori, the underlying expectation is that market forces tend towards full employment. The discussion of adjustment costs, therefore, does not contribute to the theory of permanent unemployment or the noninflationary rate of unemployment. In practice, market rigidities combined with hysteresis effects may extend the duration of trade-induced unemployment. Hence, trade-induced unemployment could conceivably be long-term. The theory of adjustment, however, is firmly rooted in neo-classical thinking, where the market-based price mechanism is relentlessly driving the economy towards full employment[5]. Second, we need to explain how trade can be termed a cause for adjustment pressures. The size and composition of trade flows are not exogenous. Rather, they result from underlying factor endowments, consumer preferences, technologies, income levels and policy regimes of the trading countries. When we talk of trade-induced changes, we therefore implicitly allude to ulterior causes which are manifested in the structure of trade flows. This conception is easiest to grasp in a setting of trade liberalisation. In this case, any change that can be tracked to the change in the trade-policy regime is defined as tradeinduced. Adjustment affects all production factors. Economists have devoted their attention mainly to adjustment in the labour market. The most accessible theoretical framework for a discussion of adjustment issues is the specificfactors model, which has been expounded concisely by Neary (1985). This model assumes a small open economy which produces and consumes an exportable and an importable good facing perfect competition in all markets and given world prices. Labour can move between the two sectors (but not between countries), all other factors are fixed (the specific factors), and there are diminishing returns to factor inputs. Imagine an export boom, which is equivalent to a fall in the relative demand for importables, triggered by some measure of trade liberalisation. If adjustment were perfectly smooth, the economy would instantly attain a new equilibrium where the unique economywide wage in terms of the exportable has fallen, and some workers have switched employment from the contracting import sector to the expanding export sector. However, Neary (1985) has shown two potential sources of adjustment costs: (1) imperfect substitutability of labour between the two sectors; or (2) nominal wage rigidity. In the first scenario, workers are temporarily tied to their sectors and cannot switch jobs without costs, but wages are perfectly flexible. Labour immobility may arise for various reasons, such as sector-specific skills, geographical immobility or firm loyalty. If, following an export boom, workers in the import sector are not able or willing to switch to the export sector, relative wages in the import sector have to fall. Over time, as more and more workers are lured into producing exportables by higher wages, the economy is likely to move to the long-run equilibrium, but in the meantime, adjustment costs will become manifest through intersectoral wage differentials. Wage differentials per se are not an adjustment cost. They are, however, indicative of the need for resources to be used up in the transfer of labour from contracting to expanding activities. Such adjustment services, comprising job searching, re-training and relocation, represent a cost to workers from switching occupations, and they make temporary wage differentials a necessary condition for factor-market adjustment (Baldwin et al., 1980)[6]. In the second scenario, the wage rate is sticky in a downward direction and driven by the expanding sector. In this situation, the initial reaction to the export boom is a rise of the overall wage level, dictated by the higher demand for labour in the exportables sector. Since the wage level is above the marketclearing level, total demand for labour falls short of total supply, and a number of workers in the importables sector are left without jobs. Again, if market forces are allowed to operate over time, the unemployed can bargain down the wage rate, but in the meantime, adjustment costs will take the shape of involuntary unemployment. The hypothesis of IIT and smooth adjustment. Numerous authors postulate that high or growing IIT implies low adjustment costs[7]. Yet, this hypothesis has not been defined rigorously. It is, however, possible to formulate the IITadjustment hypothesis in terms of the specific-factors model outlined above. According to the IIT literature, adjustment is smoother in terms of temporary wage disparities and unemployment if the expanding and contracting activities are contained within the same industry, than if they represent two different industries (Greenaway and Milner, 1986). This hypothesis implicitly makes at least one of the following two assumptions: (1) the mobility of labour is greater within industries than between industries, ceteris paribus; or (2) relative wages are more flexible within industries than between industries, ceteris paribus. The first justification for the expectation of smooth intra-industry adjustment has great intuitive appeal. If we define IIT as the exchange of goods with similar production requirements, it is implied that labour requirements are more similar within than between industries. If the skills acquired by the workers and managers of a contracting firm can be applied without much re-training in an expanding firm of the same industry, then labour mobility may well be higher within industries than between them. Where IIT reflects intra-firm trade, workers can simply be transferred from one department to another. The problem is that we cannot assume a priori that the statistical product categories underlying empirical calculations of IIT actually correspond to this definition of industries. Some recent studies, for instance, suggest that a considerable proportion of reported IIT consists of exchanges of low quality products for 229 Journal of Economic Studies 25,3 230 high-quality varieties of the same industries[8]. This phenomenon, termed vertical IIT (VIIT), casts doubt over the homogeneity of statistical product groups. The second hypothesis seems less plausible. The main impediments to wage flexibility are minimum-wage legislation and contractual wage agreements between labour market institutions. Since such constraints generally apply at the level of the entire economy or of individual industries, they might actually be expected to allow greater wage flexibility between industries than within them. If temporary wage inflexibility through industry-wide centralised bargaining is the dominant cause of adjustment problems, then adjustment costs would be greater when trade shocks are intra-industry than when trade alters the relative positions between industries. In conclusion, the homogeneity and adaptability of industries, as defined in trade statistics, can only be determined through empirical investigation. Particular attention should be paid to the conceptual difference between: (1) the intra- and inter-industry similarity of products in terms of their production requirements; and (2) the intra- and inter-industry flexibility of factor prices. Most attention has so far been paid to the former aspect. However, it seems quite plausible that adjustment costs caused by inflexible relative wages which have more serious implications in the context of traditional welfare analysis[9] are more pronounced within industries than between them, since wage agreements are mostly made at the level of individual industries. Hence, the adjustment-smoothing effect of relative physical homogeneity of industries could be offset to some extent by relatively greater intra-industry wage rigidity. The traditional analysis of intra- and inter-industry differences in factor ratios should therefore be complemented by investigations of price adjustments. Some indicative evidence on the link between IIT and adjustment is given in section five, but this brief discussion of the relevant issues highlights the importance of a more refined empirical examination, which would necessitate richer statistical information, ideally a firm-level panel of labour flows and wages. 3. Measuring intra-industry trade with relevance for adjustment Traditional IIT analysis The empirical literature of IIT uses almost exclusively an index proposed by Grubel and Lloyd (1975), which is given by: (1) where X t and M t exports and imports of a particular industry in year t. The statistical properties of this index have been analysed thoroughly (see Greenaway and Milner, 1986). Application of this measure involves some degree of arbitrariness in the definition of an industry (the categorical aggregation problem ). One way of minimising the categorical aggregation problem is to apply a very narrow industry definition, by resorting to highly disaggregated trade data. This maximises the likelihood that factor substitutability is greater within than between industries, as stipulated by the adjustment hypothesis. The results reported in section three are therefore calculated at the four and five-digit levels of the SITC classification, where up to 3,118 industries are distinguished. A second much-debated issue in IIT measurement lies in the choice of whether or not to adjust industry-level indices for aggregate trade imbalance. We have chosen to report unadjusted GL indices, based on the view that the case for adjusting these measures is weak both on economic and on statistical grounds (see Kol and Mennes, 1986). 231 Marginal intra-industry trade Even though it is computed from trade flows, the GL index is a static measure in the sense that it captures the degree of sectoral trade overlap in one time period (usually a year). In so far as exports correlate with production, the GL index therefore contains information on the degree of intra-industry specialisation, which is equivalent to the degree of symmetry in industry structures across countries, in one particular period of time. Many writers have taken an intuitive leap from cross-country (a)symmetry of industry structures at a particular time to cross-country (a)symmetry of changes in these industry structures[10]. No formal case has yet been made to underpin this line of reasoning. Furthermore, empirical evidence suggests that country-specific determinants of cyclical fluctuations and structural change often dominate industry-specific factors, even among the highly integrated countries of the European Union[11]. Hence, the degree of cross-country symmetry of industry composition is a weak predictor of cross-country symmetry in changes of this composition. The a priori case for a direct inference from IIT to balanced change in trade, and from static (a)symmetry to dynamic (a)symmetry, is therefore weak. Changes over time in the GL index were shown by Hamilton and Kniest (1991) not to be an adequate measure of the structure of trade changes. The juxtaposition of corresponding GL indices for different time periods conveys some information on the structure of trade in each of these time periods, but it does not allow conclusions on the structure of the change in trade flows. An analysis of trade changes can be conducted by reference to the concept of marginal IIT (MIIT), of which Hamilton and Kniest (1991) proposed a first index. This index was shown to be flawed by Greenaway et al. (1994b)[12]. Building on the suggestions of the latter analysis, some simple measures of MIIT were developed by Brülhart (1994). The Brülhart (1994) A index is a transposition of the GL index to trade changes: Journal of Economic Studies 25,3 232 where n stands for the number of years constituting the relevant adjustment period[13]. This is also written as: (2) (3) The A index, like the GL index, varies between 0 and 1, where 0 indicates marginal trade in the particular industry to be completely of the inter-industry type, and 1 represents marginal trade to be entirely of the intra-industry type. Where a country s exports and imports in a particular industry grow or shrink at a similar absolute rate (high A), trade-induced adjustment is likely to occur at the intra-industry level, while the overall performance of the industry is determined by factors which tend to affect all countries symmetrically, such as global demand or technology changes. The A index therefore captures the degree of cross-country symmetry in trade changes. Conversely, where a country s exports and imports in a particular industry show diverging trends (low A), both the trade-induced asymmetrical forces for geographical interindustry adjustment and the exogenous factors determining the fate of the industry across all countries are likely to be relevant. The A index (like the GL index) can provide results which are relevant for multilateral studies by relating to overall adjustment pressures. Yet, it does not contain any information as to the relative trade performance of industries in particular countries. In terms of net exports, inter-industry adjustment can reflect trade specialisation into or out of particular industries. Hence, Brülhart (1994) suggested the following index: (4) where This coefficient can take values ranging between 1 and 1. It is twodimensional, containing information about both the proportion of MIIT and country-specific sectoral performance. First, the closer B is to zero, the higher is MIIT, whereas at both 1 and 1 it represents marginal trade to be entirely of the inter-industry type. Second, sectoral performance is defined as the
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks