The aim of this paper is to provide a logical definition of the international relocalization processes in order to define a more satisfactory way of measuring of such processes. Relocalization, which is a complex phenomenon where market, cooperation and hierarchy elements coexist and interact, takes up an autonomous analytical importance when the processes involve local systems. The reference to local systems brings about a new way of measuring the phenomenon. The results obtained from an empirical enquiry concerning specifically the "made in Italy" sector, indicate the importance of this phenomenon in the '90s and its diffusion in new areas of the country. Furthermore, they provide preliminary and provisional considerations on the effects of these processes on employment in the local systems involved.
1. Introduction
The purpose of this research is to analyse internationalisation of local systems of firms on the basis of the recent experience of the so-called "Made in Italy" sector... This choice is determined by the specific features of this sector, which is characterised by the presence of a large number of small and very small firms interconnected by strong relationships and of competitive advantages that - as the criterion of identification of the sector suggests - cannot be attributed to single firms. In other words, we deal with a productive structure made of local systems organised as industrial districts.
The study of internationalisation processes of this sector, then, is the ideal field where one can try to face the problems arising from the analysis of the phenomenon from the point of view of systems of firms rather than that of single firms. It is well known that from the beginning of the nineties, thanks to the seminal work by Krugman (1991), there is a growing strand of literature that stresses the importance of contextual competitive advantages (both of origin and of destination) as determinant factors in the international growth of industrial systems.
Focusing the attention on systems analyses, however, raises obstacles arising from the fact that this approach to the internationalisation phenomenon needs a revision of the traditional analytical framework and of its tools. The first problem concerns the definition of the phenomenon itself. In strict connection with this the problem of its measurement emerges. The fact that in many studies data on outward processing trade (OPT) are used rather than those on foreign direct investments (FDI) clearly points out the widespread consciousness of this necessity.
The first matter we have to deal with in this paper, then, is that of identifying and defining clearly the phenomenon under investigation. In the next section we discuss these questions and try to explore the theoretical consequences of this new perspective. On the basis of these considerations in the third section we deal with measurement problems of the international relocation of production processes. Inadequacy of standard statistical indicators, then, is the second problem we have to face; to overcome it we propose a specific method of information treatment (section 4). In section 5 we provide an assessment of the dimension of the relocation processes of "Made in Italy" sector. The results obtained contribute to the understanding in particular of the impact of localisation on the employment levels of low-skilled workers and their wages (section 6). Conclusions will be drawn in the last section.
2. Analytical problems
The phenomenon under investigation, then, is international relocation of production. By this expression we indicate the displacement of production, which takes place under the control of the original system, carried out by firms located in the domestic country towards firms located in host countries. These foreign firms can be owned by nationals of the localising country or also by foreigners in case the output (both final and intermediate goods) is not sold on the market but is purchased by the domestic firm presumably to be resold under its own brand name. Local production systems enlarge their geographical scope while retaining their typical complex mix of ties (through markets, hierarchies and interfirm agreements). Thus, something broader than the traditional manufacturing activity related to foreign direct investments (FDI), but also than international trade. In fact, the relationships among the parties in the international localization processes may be regulated by the marketplace – then we are faced with a "normal" specialization process among firms of different countries – or they may be of a hierarchical type – thus manufacturing relocation takes place through an FDI – or they may be governed by some sort of cooperation agreement ... Within the analytical frameworks normally used to study internationalisation processes, then, relocation of production is a complex phenomenon that should be analysed in a framework where market, hierarchical and cooperative elements overlap (see Ietto Gillies, 1991, among others). Speaking of international localization of manufacture rather than dwelling separately on the traditional types of foreign involvement takes on a specific significance only when it is assumed – as in the Italian case - that the important dimension for analytical purposes is that of firm systems, local systems taken as a whole and not that of individual enterprises. That is to say when the rationale underpinning the system is not equivalent to the sum of the rationales of the single enterprises (Cantwell and Iammarino 1998).
The common feature of the three types of transactions lies in that the capacity of controlling them, according to the definition given, is attributed to the system of origin, capacity that is usually considered the key feature of direct investment activities and of international production. The implicit assumption is that there can be a control situation regardless of the existence of legal ties (as in the case of foreign investments). A similar definition is also given by Cowling and Sugden (1986).
In fact, what is the meaning of a legally enforceable control when, by definition, what is important are the benefits and control that cannot be traced to individual enterprises as such? The idea of legal control calls for the identification of the firm that exercises this control; but it is no longer meaningful when control is exercised by a local system. When internationalization involves systems, enterprise networks, the forms of control on foreign manufacture become increasingly complex and do not necessarily entail legal control, which could not be attributed to the system as a whole anyway but to a firm which is part of the system.
Thus, there are reasons to believe that in some cases relocation abroad of manufacture might be likened to international production regardless of the legal form governing the relationships between firms that localizes manufacture and foreign firms. The question is of no small import when analyzing the effects of these processes on the labor markets of advanced countries. The ties that are established among firms, or rather within firm systems, that operate in different countries are more complex than those that arise from a simple international specialization. And if it is true that in the short-run the differences might be significant only because the rationale guiding the transfer abroad of manufacturing activities might vary, in the longer run the effects on employment levels might be much less predictable.
3. Measurement problems
Traditional indicators used for the measurement of FDI are inadequate for the following reasons:
1. only part of the phenomenon may be labelled under the head of FDI;
2. it is difficult to identify legal control situations when systems of very small firms and characterised by systemic advantages are involved
3. technical problems arise at least in the Italian case. Surveys on FDI (and on various forms of non-equity agreements) based on information gathered by the press enable to catch a small part of the phenomenon. As far as international production has mainly involved large firms, or firms belonging to a group, this kind of measurement could be considered reliable. Gathering information on agreements, acquisitions and greenfield investments made by medium and small size firms is a much more difficult task: because the press is seldom informed by firms and, anyhow, the access to these information is denied. Since in recent years the international relocation of production has increasingly involved smaller firms, it should be clear that press-based statistical surveys are not anymore reliable in that they tend to underestimate the phenomenon.
An alternative measure is often proposed using data on outward processing trade (OPT): this is a more comprehensive measure of the phenomenon, but it has some drawbacks which lead again to an underestimation of relocation processes:
1. firms are not always willing to undergo the bureaucratic regulations - which are particularly binding in Italy - for obtaining export license (Forti, 1994).
2. data on OPT cannot take into account processes of horizontal integration among firms of different countries, but only of vertical integration.
3. they register export-import operations carried out by the same firm, ruling out the case of a firm in the home country which exports intermediate products, subsequently re-imported by another firm of the home country. Again, this sort of multilateral agreements is more likely to occur when smaller firms are involved as it is the case for the "Made in Italy" sector largely based on networks of small-sized firms (industrial districts).
4. Methodology
The measurement problem, then, should be solved without separating the different forms of relocation. In this work we use the traditional series of data on international trade flows detailed at a province level and by groups of commodities provided by ISTAT. Information have been gathered both on vertical and horizontal integration. In the former case, international division of labour implies a specialisation of LDC in more labour intensive phases of production. In the latter, the division of labour implies a product specialisation, with less technology intensive products assigned to LDC. The basic idea is to identify flows of intermediate and/or final products which are attributable to processes of vertical and/or horizontal integration among Italian and foreign firms.
First of all we have identified the structure of the production processes of textile & clothing, on one hand, and of leather sector, on the other, choosing an appropriate level of aggregation. The structure used is represented in Table 1.
As to vertical integration, we looked for exports either of raw materials or partly-manufactured products (e.g. yarns and untreated leather) from an Italian province towards a foreign country, which are offset by consistent imports of products of successive phases of production (e.g. knitwear and shoes). In this case, in fact, we can assume that control over foreign firms derives from the control of final product markets. These flows would be attributed to exports linked to vertical relocation and to imports linked to vertical relocation respectively whenever the value of the inward flow of intermediate and final goods is reasonably higher than the outward one, the difference being the value added by foreign production. An example of vertical integration is offered in Table 2.
As to horizontal integration processes we have looked for imports of intermediate or final goods which are not counterbalanced by opposite flows in the preceding phases of production. It should be noticed, however, that it is not easy to distinguish between relocation processes (that imply the control of Italian firms over foreign production) and international specialisation of production (i.e. normal trade integration). In order to limit this ambiguity we have posed some restrictions. First of all, we have considered only imports of intermediate and final goods directed towards the Italian provinces with the largest employment in the sectors considered (ten provinces for textile and clothing, and seven provinces for leather and shoes). The underlying idea is that a firm which displaces part of its production abroad will be more able to control the foreign firm the higher its market power is. Market power which Italian firms hold by definition in the "Made in Italy" sector, but which is likely to be stronger for firms located in those areas where a larger number of firms (or firms of greater size) are concentrated. Secondly, we have attributed to horizontal integration only import flows higher than a predetermined value for each commodity. This choice rests on the assumption that horizontal integration relationships occur when durable relations are set between partners and that the relevance of the flows is a signal of the existence of such a relation. An example of horizontal integration is offered in table 3. above.
In spite of these restrictions, the analysis involves a certain amount of "discretion" by the researcher and may determine quite approximate results, particularly when operating at a very detailed level. However, we hold that our approximation of such a complex phenomenon is far less biased than other ways of representing it (i.e. data on FDI and OPT).
.
5. The extent of international relocation
In this section we present some original estimates of the dimension and the structure of relocation of production of "made in Italy" sectors, as they emerge by using the methodology sketched in section 3. The analysis has been carried out for 1990 and 1998 and has regarded nine countries of the South Mediterranean Area and twelve countries of Eastern Europe.
The relevance of the phenomenon is clearly stated in Table 4 . In 1998 the imports of T-C-L goods which may be attributed to international relocation - then a measure of the value of displaced production - amounted to 5.747 billions lira and to more than 19 percent of total imports of T-C-L goods. More than 80 percent of this flow is linked to vertical integration processes. This means that a productive specialisation by phases prevails among Italian and host countries firms.
A rough assessment of the contribution of international production to the "Made in Italy" sector is given by the ratio of imports due to relocation to the total exports of the sector and to its overall sales. These ratios are 8.3 percent and 3.0 percent respectively. At first sight these values are relatively small and do not allow to explain the international success of this sector in terms of international organisation of production. However, it should be considered that the value of sales per employee referred to foreign sub-contractors is much lower than that of Italian firms. As a consequence, taking into account that average sales/employee of Italian firms producing in these countries (in the made in Italy sector) is likely to be less than 15 percent of Italian one, the substitution of foreign production for domestic production is surely higher (a reasonable estimate could be a value between 15 percent and 20 percent). This is a more relevant figure, and an extrapolation of recent tendencies makes it even more significant.
The importance of our results is even magnified when compared to what emerges using traditional indicators. In table 5 we have calculated the ratios of our estimate to sales of foreign subsidiaries of Italian firms (referred to the same sectors and countries). The first value indicate that our estimate of the phenomenon is 3.9 times larger than what indicators on FDI and other equity agreements would show.
In the last row of table 5 we propose an assessment of the amount of foreign labour force involved in relocation processes. Using the same sales per employee of foreign subsidiaries of Italian firms of the sector we obtain a minimum value of 150 thousands and a maximum value of 210 thousands of employees in 1998.
The data in Table 4 indicate that the phenomenon has reached an important dimension. But the most relevant feature actually is its overwhelming speed of growth. Table 6 try to give a more precise idea of it by showing its medium-term evolution and comparing it to the overall trade of T-C-L within? Italy. The dynamic of trade of T-C-L commodities is clearly connected with the evolution of relocation processes. Imports of T-C-L goods which may be attributed to international relocation have increased in nominal terms by almost seven times. Most of this increase is due to the huge expansion of vertical integration processes: in 1998 the related imports are almost 12 times larger than they used to be in 1990 and the related exports have grown by 5 times in the same period. On the other hand, flows of imports linked to horizontal integration have grown only by three times.
According to the dominant view, the main stimulus to the internationalisation of the sector stemmed from the opening of Eastern Europe countries (EEC), whose geographical proximity and favourable labour conditions made them ideal partners for Italian firms. As Table 7 shows the share of these countries in vertical relocation processes has not substantially changed (from 69.5% to 66.7%) after 1990, indicating that the expected crowding-out on behalf of EEC has not taken place. This result deserves more attention, since it suggests that the renewed political situation in Europe may have determined the acceleration of relocation processes, but that the roots of the phenomenon are of a different nature.
6. International relocalization and effects on the labor market
Thus, in the space of eight years an exceptionally high number of jobs has been created outside the country, especially in light of the fact that the analysis is focusing on one area of the manufacturing sector alone. The size these processes have achieved over the years yields some very interesting information.
The analysis of these changes can be meaningful both in terms of aggregate data and, for some aspects in particular, if a comparison is made especially between the areas where the relocalization processes have been more pronounced and the other areas. In other words, it should not be difficult to spot features attributable to the relocation abroad of manufacture, or stages of manufacture, in the changes in the employment levels and composition within the industries at the beginning of the 1990s.
Let us focus the attention on the employment issue. Is the international development of enterprises a threat to domestic employment? Table 8 was built to try to find an answer to this question; it shows employment patterns by group of regions or provinces by utilizing two sources, Istat’s regional accounting data and employment data provided by Inps. The regional accounting data are in labor units, Inps’s in jobs. The former are available at the regional level until 1995 while the latter are so at the provincial level and until 1996 (with some problems related to the quality of information). Obviously, the area under consideration is that of the traditional industries.
Provinces (regions) are divided on the basis of the ratio of localization share of the country total and employment share of the total for the country. The provinces (regions) where this ratio is higher than 1 have been defined as areas with a high propensity to relocalize manufacturing abroad; the propensity to relocalize of those areas where the ratio is between 0.5 and 1 has been defined as medium, while those lower than 0.5 had a low propensity.
The regional accounting data shown in Table 8 reveal that there is no correlation between international relocalization and employment patterns, both if a comparison is made between the 1980s – with no international localization to speak of – and the 1990s as well as if a comparison is made among the areas that were variously concerned with this type of processes.
To verify the results obtained, a rank correlation index between the imports (per employee) related to the international localization phenomena for each province and the employment changes in the industries under review in the nineties was computed by utilizing Inps’s employment data. Spearman’s rank correlation coefficient was 0.21. In other words, not only do Inps’s data confirm that there is no positive correlation between localization and unemployment, as attested also by the regional accounting data, but it can also be observed that, if anything, this correlation would be negative. Thus, as far as these industries are concerned, there is no evidence that the localization of manufacturing activities abroad entails a cost in terms of higher unemployment.
As underlined above, theoretical analysis suggests that these processes displaced only low-skilled labor. If this is true then it can be inferred that what has taken place is a process whereby low-skilled workers have been replaced by high-skilled ones.
Inps’s data allow us to perform a more in-depth analysis on the composition by skill, but only for blue-collar workers and clerks. No in-depth analysis can be conducted instead within the blue-collars’ ranks. Table 9 shows the clerk share of total workers in 1990 and 1996 for the three groups of provinces already considered in the two preceding tables. The results seem to confirm the assumption that in the areas where the international localization of production have been more marked are those where the clerk share of total workers has been growing the fastest, both in absolute as well as in percentage terms. Thus, this is a further confirmation of the assumption that the main effect of these processes, at least as far as these industries are concerned, has been a specialization by phase which did cause the less skilled components of the labor force to be shed but which saw also a greater concentration in Italy of the more skill-intensive stages of the production processes, thus an increase of high-skilled employment levels.
It might be interesting to understand the trade-off mechanisms between these two components of the labor force. Namely, whether new jobs are the result of internal upgrading (dictated by the market or related to the work of institutions) or are determined by a generation turnover. The information available cannot provide an answer to these questions.
7. Conclusions
The aim of the paper was to address the question of local systems internationalisation in a clear and explicit way. It is our belief that the ideas discussed, the accent on elements of discontinuity with the traditional framework, may contribute to start - rather than to conclude - a debate. Debate which, on the basis of our empirical results, may be usefully carried on as far as are regarded both the identification of phenomena linked to the internationalisation of systems of firms and the relationship between what we have called relocation and international production strictu sensu, especially in the long period.
We are also aware that measurement problems are far from being completely solved. However, the belief that the methodology we have used may be helpful in facing the question has strengthened as we went on with the information processing, even if we also believe that there is space for many improvements and that the combined use of different statistical sources may be fruitful.
As to empirical results, we think that they may be considered suggestive. First of all, the finding that geographical - as well as cultural - proximity has not been an essential factor for relocation choices in the "made in Italy" sector, enhances the idea that these sorts of phenomena will tend to develop at an increasing speed in the future. Secondly, the international restructuring of the sector seems to have reached so relevant proportions that a deeper insight on its impact on both domestic and host countries economic systems is urgently needed. It seems that we are only at the beginning of a process of productive integration among different countries, whose meaning and consequences have still to be explored. Even if data on employment do not confirm the hypothesis that foreign employment crowds out domestic labour, at least in this specific experience, there is little doubt that these processes will affect considerably the volume and the structure of employment (particularly by skill) as well as the productive structure. Similarly we think that equal attention should be paid to the impact of these processes on the industrial systems of the countries mainly involved.
In conclusion, even if we do not give definite answers to the questions posed, we are confident that this work may be a stimulus to new lines of research, debate and further insights.
Andersson U., Fosgren M., 1995, Subsidiary Embededness and its Implication for Integration in MNC, in Schiattarella R. (ed.), New Challenges for European and International Business, Urbino.
Bhagwati J., 1995, Trade and wage: Choosing among alternative explanations, Economic Policy Review 1:42-47.
Bhagwati J., Dehejia V., 1994, International trade theory and wages of the unskilled, in Bhagwati J., Kosters M. H. (eds), Trade and wages: Leveling Wages Down?, Washington: American Enterprise Institute, pp 36 - 75.
Balloni V., Conti G., Cucculelli M., Menghinello S. , 1997, Modelli di impresa e di industria nei contesti di competizione globale, XXI Convegno de "L'industria", Bari.
Becattini G., Rullani E., 1993, Sistema locale e mercato globale, Economia e Politica Industriale n° 80.
Berman E., Bound J., Griliches Z., 1994, Changes in the Demand for Skilled Labor within U. S. Manufacturing: Evidence from the Annual Survey of Manufactures, Quarterly Journal of Economics, 109:367-98.
Brainard S. L., Riker D., 1997, Are U. S. Multinational Exporting Jobs?, NBER Working Paper n° 5958.
Cantwell J., Iammarino S., 1998, MNCs, technological innovation and regional systems in the EU, University of Reading, Discussion papers in international investment and management, series b, vol.10, n.247.
Casson M., 1992, Internalization Theory and Beyond, in P. J. Buckley (a cura di), New Directions in International Business, Elgar.
Cominotti R., Mariotti S., Italia multinazionale, 1986-1998, Angeli.
Conti G., Menghinello S., 1995, Territorio e competitività:l'importanza dei sistemi locali per le esportazioni italiane di manufatti. Un'analisi per province (1985-1994), Rapporto sul Commercio Estero, ICE.
Cowling K., Sugden R., 1987, Market Exchange and the Concept of Transnational Corporation: Analysing the Nature of the Firm, British Review of Economic Issues, 9 (20).
Dunning J. H., 1981,International Production and the Multinational Enterprise, Allen & Unwin.
Dunning J. H., 1993, Multinational Enterprises and the Global Economy, Addison-Wesley.
Faini R., Falzoni A. M., Galeotti M., Helg R., Turrini A., 1998, Importing Jobs and Exporting Firms? A close look at the Labor Market Implications of Italy's Trade and Foreign Direct Investment Flows, mimeo,
Feenstra R. C., Hansen G. H., 1996, Foreign Investment, Outsourcing, and relative wages, in Feenstra R. C., Grossman G. M., Irwin D. A. (eds), The Political Economy of Trade policy, M.I.T. Press.
Feenstra R. C., Hansen G. H., 1996, Globalization, Outsourcing, and Wage Inequality, American Economic Review, papers and proceedings,May.
Forti A., 1994, Il Traffico di Perfezionamento Passivo nelle nuove strategie delle imprese italiane del tessile-abbigliamento, in Istituto Commercio Estero, Rapporto sul Commercio Estero.
Grandinetti R. e Rullani E.,1996, Impresa transnazionale ed economia globale, NIS, Firenze.
Grossman G., 1987, The Employment and Wage Effects of Import Competition in the United States, Journal of International Economic Integration, 2, 1-23, Spring.
Ginsburg A., Simonazzi A. M., 1995, Il settore tessile, abbigliamento, calzature in Europa, Conferenza CNA, Roma.
Krugman P., 1991, Geography and trade, MIT Press.
Krugman P., 1994, Does Third World Hurt First World Prosperity?, Harvard Business Review, 72:13-121.
Krugman P., 1995, Growing World Trade: Causes and Consequences, Brooking Papers on Economic Activity, 1.
Howells J., 1998, Regional systems of innovation, in D.Archibugi et al. (eds), National systems of innovation or the globalisation of technology? CUP
Ietto Gillies G., 1991, International Production, Trends, Theories, Effects, Polity Press.
Lawrence R. Z., Slaughter M. J., 1993, International Trade and American Wages in the 1980s: Giant Sucking Sound or Small Hiccup?, Brooking Papers on Economic Activity, Microeconomics, 2:161-226.
Lawrence R. Z., 1994, Trade, Multinationals and Labor, NBER, Working paper 4836.
Neven D., Wyplosz C., 1996, "Relative Prices, trade and Restructuring in European Industry", CEPR Discussion Paper n. 1451 (august).
Revenga A., 1992, Exporting Jobs? The Impact of Import Competition on Employment and Wages in U. S. Manufacturing, Quarterly Journal of Economics 107:255-84.
Scarso E., 1996, La rilocalizzazione del processo produttivo nei sistemi locali del Veneto: evidenze dal settore moda, Economia e Società Regionale, n° 4.
Slaughter M. J., 1995, Multinational Corporations, Outsourcing and American Wage Divergence, NBER Working Paper, n° 5253.
Storper M., 1995, The resurgence of regional economies ten years later: the region as a nexus of untraded interdependencies, European and regional studies, 2, pp.191-221
Zanfei A., 1985, Cambiamento tecnologico e strategie di internazionalizzazione delle imprese tessili italiane, Economia e Politica Industriale, n° 46.