Clusters and Competitive Advantage_3 docx

Others for example Rugman,1991; Hodgetts, 1993; Rugman and D’Cruz, 1993; Rugman and Verbeke,1993 share the idea that double and/or multiple-linked diamonds wouldreflect the sources of co

but also by the conditions for resource supply and resource creation in itsproximate environment (Öz, 1999) We shall now look at the most relevantdebates for the purposes of this study

Geographical unit of analysis and applicability to every context

A very interesting debate in the literature is on the most appropriate graphical unit of analysis to apply Porter’s approach In his 1990 study, Porterargues that many of the determinants of advantage are more similar within

geo-a country thgeo-an geo-across countries However, becgeo-ause the geogrgeo-aphic tration of competitive industries is so important he questions whether thecountry is the most appropriate unit of analysis since competitive advantageoften seems to be localized in an area within the country Internationalbusiness scholars, however, tend to take the opposite position Regardingthe EU, for example, Dunning (1993) argues that national diamonds should

concen-be replaced by ‘supranational diamonds’ in order to capture the true petitive advantages of the EU Jacobs and De Jong (1992), on the otherhand, argue that there is a dialectic relationship between divergence andconvergence, and concur with Porter’s (1990) idea that globalization para-doxically leads to more emphasis on local conditions and creates anopportunity for firms to take advantage of them Others (for example Rugman,1991; Hodgetts, 1993; Rugman and D’Cruz, 1993; Rugman and Verbeke,1993) share the idea that double and/or multiple-linked diamonds wouldreflect the sources of competitive advantage better than Porter’s (1990) singlediamond framework does for smaller countries that are highly dependent

com-on com-one or more of the major blocs (Europe, North America and Japan) Atthe micro level the issue is further complicated by the existence of cross-borderclusters (Saner and Yiu, 2000)

Relatedly, some researchers consider that Porter’s approach cannot beused for all countries For instance Rugman (1991) believes that while most

of Porter’s (1990) analysis would work for managers based in the EU, theUnited States or Japan, much of it could not be applied in Canada Themain reason for this, according to Rugman, is that Porter’s study does notincorporate the true significance of multinational activities, an issue thatwill be discussed below Similarly, in Hodgetts’s (1993, p 44) view, ‘sincemost countries of the world do not have the same economic strength oraffluence as those studied by Porter, it is highly unlikely that his model can

be applied to them without modification’ Porter’s emphasis on home marketsand local firms, according to Bellak and Weiss (1993), may be justified inthe case of large countries but is of little relevance for small ones Narula

(1993) and Yetton et al (1992) make a similar point when arguing that since

it is based on and applied to them, the diamond is most relevant for mature,manufacturing-based economies and cannot be used to explain the inter-national competitiveness of developing countries Similarly Davies and Ellis(2000) argue that since Porter generalizes inappropriately from the American

Clusters in the Management Literature 33

experience, developing countries are inadvertently encouraged to pursuepolicies that might be harmful

Sources of advantage: global versus local

Whether sources of advantage are local, as suggested by Porter (1990, 1998),

is another issue that has been subject to severe criticism Porter’s (1990)treatment of multinationals and foreign direct investment in particular hasbeen widely criticized According to Rugman (1991), the narrow understanding

of foreign direct investment is a major conceptual problem with Porter’smodel Relatedly, Davies and Ellis (2000) argue that it is not surprising thatSingapore was not included in Porter’s 1990 book, eventhough it had beenstudied by him: ‘If Singapore’s prosperity were determined by the activities

of firms for whom Singapore is a home base its residents would be poorpeople, but they are not.’ According to Dunning (1993), to suggest that thecompetitiveness of multinationals rests only on their access to the diamond

of competitive advantage in their home countries is ludicrous, regardless ofwhether or not their initial foray overseas was based on such advantages.The geographical dimension of the criticisms of Porter’s attitude towardsFDI is the focus of a work by Lagendijk and Charles (1999), who emphasizethe importance of foreign assets in clustering and suggest that at the regionallevel the issue of multinationality becomes an issue of multiregionality.7Rugman and Verbeke (1993, p 72) challenge ‘Porter’s allegation that thecore competencies of large MNEs and the innovative processes occurringwithin these firms necessarily need to depend upon the characteristics of

a single home base’ They argue that multinationals from small countriesmay rely on a host nation to such an extent that it becomes difficult tomake a distinction between the home base and the host country or countries.8According to Rugman and Verbeke (2001), a major problem with Porter’sapproach is that he concentrates solely on non-location-bound, firm-specificadvantages (FSAs) developed by companies in their home country prior toengaging in FDI, which is only one of many possible combinations that can

be observed empirically in respect of the locational determinants of tive advantage For example one alternative is for non-location-bound FSAs

competi-to be created jointly by subsidiaries located in various countries andexploited throughout the network Here we have an increasingly complexand blurred picture of the relative contribution of FSAs versus CSAs (country-specific advantages) and home CSAs versus host CSAs to overall multinationalcompetitiveness

Another point of disagreement concerns the identification of the homebase of a multinational According to Rugman and Verbeke (ibid.), it isnecessary to define a threshold percentage of core assets, competencies andstrategic decision-making power, below which a firm would be viewed asfunctioning with several home bases In addition, if a firm is able to enhanceits accumulated competencies through interactions with location advantages

in host countries, it will again be viewed as functioning with several homebases This implies that most multinationals will have several home bases,which is in sharp contrast to Porter’s approach It should be rememberedthat Porter (1990) uses the world export shares of industries as a proxy tomeasure international competitiveness at the industry level An industry isalso considered to be competitive when domestic firms in the industry areengaged in substantial outward FDI With regard to inward FDI, a methodo-logical problem arises when a country has an internationally competitivesector (measured by world export share) that is dominated by foreigncompanies What Porter does in such cases is to try to locate the source ofadvantage through in-country research This requires determining whetherthe firms in the industry operate as branches of a multinational company orcan be clearly associated with the host country In the former case theindustry is excluded, and in the latter case it remains on the list of competitiveindustries This is confusing but logical, and the real challenge is to locatethe source of advantage

What is even more confusing is Porter’s (1990) argument that ‘inward FDI

is not entirely healthy’, especially when examples of relatively prosperouscountries such as Singapore, Canada and Ireland, which host considerableinward FDI, are taken into account Dunning (1993) argues that Porter’sinterpretation of the link between FDI and competitiveness rests on the ideathat outward FDI reflects the possession of firm-specific tangible assets thatgive a competitive edge prior to undertaking the FDI While this is a validexplanation of why individual firms are able to engage in FDI, it does notfollow that inward FDI has a negative effect on the competitiveness of therecipient economies (Davies and Ellis, 2000) Recently Lin and Song (1997)have taken up Dunning’s (1995) extension of the diamond framework,which adds ‘multinational business activity’ as a determinant of competitiveadvantage Applying the model to China, Lin and Song conclude that thecountry’s recent success owes much to inward FDI Similar findings areavailable for other countries, including Mexico (Hodgetts, 1993) and Singapore(Chia, 1994) The crucial point here is that foreign investors might and dochoose competitive locations because the environment offered by a particularindustry cluster acts as a magnet for other firms in the industry, so bothnational and foreign firms gravitate to favourable cluster locations even ifcorporate ownership is based elsewhere.9 This being the case, there is no reasonwhy inward FDI should be considered ‘unhealthy’

In summary, many international business scholars (Rugman, 1991, 1992;

Rugman and D’Cruz, 1993; Jacobs and De Jong, 1992; Yetton et al., 1992;

Bellak and Weiss, 1993; Cartwright, 1993; Dunning, 1993; Hodgetts, 1993;Rugman and Verbeke, 1993; Yla-Anttila, 1994) have found fault with Porter’s(1990) insistence that firms’ ability to compete depends on the strength ofthe diamond in their home base As Davies and Ellis (2000) point out, however,this difficulty with the diamond goes deeper than these researchers realize

Clusters in the Management Literature 35

since the argument can be extended to suggest that not only multinationalcompanies but also other companies that are exposed to internationalinfluences in one way or other (for instance via exporting) may sharpen theiradvantages as a result of such interactions If, however, ‘firms in one countryare able to draw upon diamonds in another, the concept of the nationaldiamond is stripped of its content’ (ibid., p 1204), since the whole concept

of the diamond is based on the hypothesis that the sources of competitiveadvantage are local Porter thinks that such criticisms mainly stem from anunnecessary confusion: the geographic scope of competition and the geo-graphic locus of competitive advantage are two different things In his view,competition can be global but the sources of advantage are local (Porter andAmstrong, 1992) It is therefore clear that the two sides of the debate, that

is, Porter versus the international business scholars, are arguing for twocompeting hypotheses: that the sources of advantage are local, or thatadvantages can be sourced globally The point made by international businessscholars, in other words, is in fact a counter-hypothesis rather than a criticism.Needless to say the burden of proof lies on both sides when there are twocompeting hypotheses, and this calls for further empirical research This bookhopes to contribute to this by investigating not only the local circumstances

of but also the global linkages associated with the Turkish clusters

The debate on policy issues

Another noteworthy debate focuses on regional policy issues According toMarkusen (1996b), agglomeration effects are largest for industries that arehigh-tech, knowledge-intensive, innovative and young She implies thatdeveloping countries need these industries because they support a higherstandard of living Porter (1996), however, believes that this perspectivemay be misleading, and that the productivity of an industry matters morethan its being high-tech Markusen also challenges Porter’s argument thatindustrial clusters are the most significant unit of analysis for investigatingregional economic advantage According to her, this is an empirical questionand far from self-evident As an example she cites Seattle, the dynamism ofwhich is explained by the presence of five distinct sectors: shipping, forestry-related activity, aircraft, software and biotechnology (Markusen, 1996b, p 91).Porter agrees with Markusen’s view that the significance of generalized versuscluster-specific agglomeration economies is an empirical question Withregard to the Seattle example, Porter underlines that he is not suggestingthat all clusters in a regional economy have to be connected Another majorpoint of divergence for the two researchers is that Markusen supportsgovernment targeting of particular industries, which in her view is appropriateand effective, whereas for Porter, the whole premise of targeting is flawed.Porter and Markusen also disagree on the types of regional policy that should

be followed Markusen favours a traditional formulation of regional policythat includes broad incentives for firms to locate in less developed regions,

whereas Porter thinks that such measures are doomed to failure According

to him, cluster formation can only be encouraged ‘by locating specializedinfrastructure and institutions in areas where factor endowments, pastindustrial activity, or even historical accidents have resulted in concentrations

of economic activity’ (ibid., p 88) Moreover in Porter’s view there are strongarguments for the greater decentralization of economic policy to subnationalregions, marking yet another area in which he disagrees with Markusen.10Another dimension of policy issues that has been subject to debate is therevitalization of inner-city areas.11 Based on his approach to the locationaldeterminants of competitiveness, Porter (1995a) argues that this task canonly be done through private initiatives based on economic self-interest andcompetitive advantage In the associated debate in the literature, Blakelyand Small (1995) state that Porter’s analysis is incomplete, while Johnson

et al (1995) argue that Porter has devoted too little attention to the role of

the business community in revitalizing such areas In their view, Porter’sassertion that the private sector – in exchange for a more business friendlyenvironment – will step in to fill the gap is not convincing given that thishas rarely happened in the past Businesses need steady customers and reliableemployees, and people who are ‘ill-housed, ill-fed or just plain ill’ cannot beeither (Lowery, 1996, p 64) Overall the critics agree that Porter’s (1995a)approach can serve to supplement other efforts, but it can never be an all-in-one solution or as important as affirmative action In his reply to his critics,Porter (1995b, p 304) insists that many of the criticisms indicate a misun-derstanding of his arguments According to him, as a general principle it isnecessary to view the disadvantages suffered by inner-city areas as an economicproblem and the result of poor strategies and obsolete public policies It istherefore necessary to develop a new strategy for each area, tailored to itsunique characteristics and building on its advantages (ibid., p 333) With regard to the role of government, Porter (1990) believes that clustersoften emerge and grow naturally so there is only an indirect role for thegovernment This is one of the most criticized aspects of his approach Severalscholars (for example Stopford and Strange, 1991; Van den Bosch and deMan, 1994; Öz, 1999) are of the opinion that in developing countries a moreactive part should be played by the government as poor countries cannotafford the luxury of letting market forces determine outcomes In his laterwork Porter (1998) continues to argue that the essential role of government

is to challenge and press industries, and that too much help can underminethe industries’ success A detailed discussion of the ideal level of governmentintervention is beyond the scope of this study However the discussions inthis book on the part played by the government in shaping the sources ofcompetitive advantage of the Turkish clusters examined may provide someinsights into to the role of government in cluster development

3

Industrial Clusters in Turkey

The Turkish business environment, past and present

During the first ten years of the newly established Republic of Turkey(1923–32), state involvement in economic activities was rather limited Thiswas mainly because (1) the basic principles adopted in the Izmir EconomicCongress (1923) committed the government to the establishment of a privateenterprise economy, and (2) some economy-related provisions in the LausanneTreaty (1924) considerably restricted the area in which the governmentcould operate For instance the country was bound to apply the Ottomantariffs for another five years Over this period little was achieved in terms ofindustrialization since the private sector lacked the necessary technologicalcompetence and capital These factors, combined with external ones such asthe Great Depression, were enough to convince the policy makers that theprivate sector could not be entrusted with the task of leading the country’seconomic development This marked the beginning of a new period (1933–45)

in Turkish economic history called ‘etatism’, during which the governmentheavily intervened in the production of goods and services The First FiveYear Industrialization Plan (1934–38) placed strong emphasis on the indus-trialization process, particularly in the case of textiles, iron and steel As

a result of the related policies the pace of industrialization accelerated, withindustry’s share of GNP rising from 14 per cent to 18 per cent during theperiod in question (Kepenek and Yentürk, 1997)

Between the end of World War II and 1960, some attempts were made atliberalization, shaped by a new type of etatism in which the governmentsupported the private sector The transition to a multiparty regime and theprovisions of the Marshall Plan are considered to be the major reasons forthis policy shift Significant investment in energy and motorways as well as

a boom in the housebuilding sector associated with rapid urbanization created

a considerable demand for construction firms, thus promoting the development

of the Turkish construction industry Another feature of the period was thatspecial emphasis was placed on agriculture in accordance with the Marshall

Plan, which brought modern practices to the sector The government wasclearly committed to encouraging the private sector and therefore pursuedpro-business policies However this fostered rent-seeking activities, whichsubsequently became an increasingly deep-rooted problem Interestingly, sincethe pro-business policies did not bring stability, both politicians and businesspeople started to question whether it was possible to achieve stability andliberalization at the same time In this respect it is worth mentioning thateven Prime Minister Menderes, who was very sceptical about planning, had

a change of mind and took certain steps to prepare a development plan inhis last year in office, prior to the military intervention in 1960

The disappointing results of liberalization, together with the tendency where in the world for greater government intervention, caused the militarygovernment of the early 1960s to introduce a 20-year import-substitutiondevelopment strategy for a mixed economy, to be implemented via five-yearplans During this period there were improvements in the growth rate ofoverall output and industrial production Big businessmen were also in favour

else-of a planned approach and stressed the importance else-of having a long-termeconomic strategy to reduce the uncertainty in the economic environment.The need to clarify the boundaries of private sector activity was anotherfactor in this The sense of responsibility felt by the newly emerging bour-geoisie for the economic development process resulted in the establishment

of influential business associations such as TÜSIAD (Bugra, 1994)

The period 1960–80 was a time of unusual political turmoil and there werethree military interventions (in 1960, 1971 and 1980) After these interven-tions, concern about the position of the private sector was soon replaced byconcern about the instability generated by the regimes’ macroeconomicpolicies In the 1970s two additional developments, the oil shock and theCyprus crisis, exacerbated the already bleak scene The coincidence of anunfavourable global economic environment with the political instability inTurkey led the country into a major crisis in the late 1970s, resulting inanother military takeover in 1980 In that year the ‘January 24 Resolutions’introduced a comprehensive stabilization programme under the auspices ofthe IMF and the World Bank The structural adjustment policies adopted inaccordance with the programme were intended to shift the economy from

an inward to an outward orientation, with an emphasis on export-led growth.Reforms were conducted in a number of key areas, one of which was tradepolicy, with the introduction of extensive export promotion measures andthe gradual liberalization of imports The results were impressive in terms ofexports in general and manufactured exports in particular, although theincrease in exports was matched by a boom in imports (Öz, 1999)

In the second half of the 1980s there was a considerable reduction inexport subsidies Tariffs and quotas, and therefore the level of importprotection, were also reduced With the unexpected but comprehensive finan-cial liberalization achieved by making the Turkish lira convertible in 1989,

Industrial Clusters in Turkey 39

the main policies of the liberalization programme were completed Theimmediate result was a worsening of the trade deficit, mainly stemming fromthe increase in imports rather than a decrease in exports, which actuallycontinued to increase gradually (Figure 3.1) and Turkey’s world export shareremained fairly stable

It is argued that the frequent and unexpected changes to key policiescreated a chaotic business environment in Turkey in the 1980s and 1990s(Bugra, 1994) Under the circumstances it was essential for business people

to have good state contacts so that they would at least have a vague ideaabout what was going on In fact, they often complained not about thechanges themselves but about the way they were handled What was worse,however, was that such an environment offered considerable opportunities forabuse Allegations about tax rebates for exports, for instance, caused somescholars to question the export success achieved by Turkey in the post-1980period, and to ask whether the export figures were fictitious (see Arslan andvan Wijnbergen, 1990)

While the 1980s are associated with major reforms, the 1990s are oftenconsidered ‘lost years’ in Turkish economic history (Kumcu and Pamuk, 2001).With regard to the key events that shaped the 1990s, the first was the Gulfcrisis in the beginning of the period, which damaged Turkey’s economicrelations with Iraq In 1994 Turkey faced yet another economic crisis, duemainly to mismanagement of a programme to reduce interest rates Thecustoms union between Turkey and the EU, which had been in effect sinceJanuary 1996, brought challenges as well as opportunities for Turkish industry

Figure 3.1 Exports and imports, Turkey 1982–2000 (US$ 000s)

Sources: SIS (2000); ITC (2002)

Towards the end of the decade the Asian crisis broke out, affecting manyparts of the world The impact of this on the Turkish economy was indirectand occurred after a one-year lag, but the Russian crisis caused considerabledamage to the construction and leather sectors, whose main trading partnerwas Russia In 2000 the government introduced a disinflationary programme,but this collapsed in February 2001 Finally, Turkey implemented yet anotherstabilization programme, under the auspices of the IMF and the World Bankand aimed at ‘empowering the Turkish economy’

Turkey is classified by the World Bank as a middle-income developingcountry It has close ties with the EU, including a customs union agreement

It occupies a very advantageous geographical position, constituting a naturallink between West and East, and recently it has started to take greateradvantage of this, especially in respect of trade and tourism Turkey’s standard

of living, as measured by GDP per capita, has gradually increased (Figure 3.2)but is still rather low at US$ 2200–6080, based on purchasing power parity(PPP) (2001 figures, SPO, 2002) The average annual growth rate of theeconomy, as measured by the rate of growth of real GDP, on the other hand,averaged about 4 per cent in the post-liberalization period This rate, thoughfluctuating widely, was slightly above the average attained by middle-incomecountries (around 2–3 per cent) during the same period (World Bank, 1999).However, although overall domestic production and per capita income havebeen increasing at above average rates compared with other middle-incomedeveloping countries, inequalities in income distribution remain significant.Persistently high inflation rates and external debts, when taken togetherwith Turkey’s ‘grey’ economy, present a bleak outlook for the country’smacroeconomic future This is further complicated by the continuing politicaluncertainty Such an environment is preventing firms from improving their

GDP per capita GDP per capita (PPP)

Figure 3.2 Standard of living, Turkey, 1980–2001 (US$)

Source: SPO (2002)

Industrial Clusters in Turkey 41

competitive advantages Given this picture it is not surprising that Turkeyhas failed to attract much FDI, the annual average being less than US$ 1 billion

in recent years, a figure that compares unfavourably with those achieved byother emerging economies (SPO, 2002)

An examination of the broad characteristics of the Turkish business onment shows that small and medium-sized enterprises account for more than

envir-90 per cent of Turkish firms, but larger firms’ contribution to value-addedand exports are much higher (Taymaz, 1997) Big corporations are a rela-tively new phenomenon in Turkey: of the 405 TÜSIAD member companies,only 22 were incorporated before 1950 (Bugra, 1994, p 55) The 1950s were

an important decade for many of the largest Turkish companies, reflectingthe government’s shift to more liberal policies Many of today’s leadingTurkish construction firms, for example, were either established or made animportant turn in their business during that decade (Öz, 1999)

Family-dominated management of firms of all sizes is a common non in Turkey as there is a lack of confidence in salaried managerial personnel.Educating young members of the family in top universities, integrating aprofessional manager into the family via marriage, and strong relationshipsestablished over the years between family members and professional managers,making the latter ‘part of the family’, appear to be common ways of achieving adelicate balance between professionalization and family control (Bugra, 1994) According to Bugra (ibid., pp 68–9), all Turkish business tycoons havecertain characteristics in common, including family support in commercialactivities at the start of their career, the arbitrary – and rather opportunistic –choice of their initial area of activity, heavy engagement in unrelated diver-sification as the business grows, and good connections especially in statecircles Rent-seeking behaviour is common, and real-estate speculation isparticularly widespread

phenome-The high degree of state involvement in business activity, be it in the form

of subsidized credits, input supply or output demand, has been detrimental

to the Turkish business environment Given the key role of government inthe economy, good connections in government circles have contributedsignificantly to business success The slow bureaucracy and unexpectedchanges in key policies, on the other hand, have caused problems for Turkishbusiness people

Turkey’s position in international competition

This section provides an overview of the evolution of the competitive ture of Turkish industry The analysis is conducted with the help of Porter’s(1990) methodology The basic measure used to determine the internationalcompetitiveness of an industry is its share of world exports, which is defined

struc-as a country’s exports for an industry divided by total world exports for thatindustry in a given year All industries defined in the Standard International

Trade Classification (SITC) are then sorted by world export share at the lowestpossible level of disaggregation (in five-digit detail) Next the cut-off rate iscalculated by dividing the total exports of a country by total world exports.Those industries which have world export shares above the cut-off rate con-stitute the relatively more competitive industries of the country The list ofthese industries is then modified according to additional criteria For example,industries with a world export share that lies between the cut-off rate andtwice its value are checked to exclude ones with a negative trade balance.Also, industries that are among the top fifty in terms of their country’s exportshare (which is defined as the share of an industry in the country’s totalexports) but below the cut-off rate in terms of their world export share areincluded in the list of relatively more competitive industries, provided theyhave a positive trade balance If there is considerable outward foreign directinvestment in an industry, this industry is also included in the list Finally,with the addition of the internationally competitive service sectors the list iscompleted for that particular year and country (Öz, 1999)

The list of competitive industries is used to produce cluster charts Thesereveal the connections between and interrelationships amongst the country’scompetitive industries, and hence the country’s competitive pattern All com-petitive industries are classified into three broad groupings, each of whichincludes different clusters The first group consists of ‘upstream industries’,whose primary products are inputs to the products of other industries Theclusters included in this category are semiconductors/computers, materials/metals, petroleum/chemicals and forest products The second group, ‘industrialand supporting functions’, comprises clusters of multiple businesses, trans-portation, power generation and distribution, office, telecommunications, anddefence The last group is ‘final consumption goods and services’, whichcontains the food/beverage, textiles/apparel, housing/household goods, healthcare, personal, and entertainment/leisure clusters The industries in each clusterare further classified into four groups, revealing the vertical relationshipsamong industries and the depth of national clusters These four groups areprimary goods, the machinery used to produce these goods, the special inputsrequired and the related service industries (Öz, 1999)

We shall now apply the above methodology to recent data on Turkish tries Table 3.1 shows the percentage of exports by cluster and vertical position

indus-in 1992–2000 Turkey’s share of world exports indus-in 2000 was 0.52 per cent,and six clusters of industries had a share above that figure, namely materials/metals (from the upstream industries group), food/beverages, textiles/apparel,housing/household, personal and entertainment/leisure (all from the finalconsumption goods and services group) Of these, textile/apparel had thehighest share with an impressive 2.4 per cent Turkey exports a great variety

of items in this category, mainly primary goods and special inputs Theimportance of the cluster for the Turkish economy is considerable, given that itaccounts for around 37 per cent of the country’s total exports While it has

Table 3.1 Percentage of Turkish exports by cluster and vertical position, 1992–2000

Notes: SC share of country’s total exports (2000); CSC change in share of country’s exports (1992–2000); SW share of world cluster exports (2000); CSW change in share of world cluster exports (1992–2000)

Materials/Metals Forest products Petroleum/Chemicals Semiconductors/Computers

Upstream industries

SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC SW

Primary goods 0.2 0.2 0.0 0.0 3.7 2.5 0.2 0.1 2.3 0.7 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.2 6.3 0.1 Machinery 0.2 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.1 Special inputs 0.0 0.0 0.0 0.0 1.2 0.3 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.2 0.2 Total 0.4 0.4 0.0 0.0 4.9 2.8 0.2 0.1 2.3 0.7 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.2 7.7 0.1

Food/Beverage Textiles/Apparel Housing/Household Health care Personal

Entertainment/

Leisure

Final consumption goods &

services

SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC CSC SW CSW SC SW

Primary goods 9.7 6.9 0.9 0.1 33.1 1.6 2.7 0.7 7.5 2.7 0.9 0.4 0.2 0.2 0.0 0.0 2.0 2.0 0.5 0.5 3.1 1.4 0.6 0.5 56.0 1.2 Machinery 0.3 0.3 0.2 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.2 Special inputs 0.6 2.5 0.2 0.4 3.4 0.2 1.6 1.0 0.8 0.6 0.3 0.4 0.0 0.0 0.4 0.4 1.3 0.7 1.2 1.8 0.0 0.0 0.0 0.0 6.1 0.7 Total 10.6 9.1 0.7 0.1 36.5 1.8 2.4 0.9 8.3 2.1 0.8 0.3 0.2 0.2 0.0 0.0 3.3 1.3 0.6 0.3 3.1 1.4 0.6 0.5 62.4 1.1