201 1999, their contribution to total manufacturing output (57%) and GDP (15%) was vastly disproportionate. The productivity differences to EU-15 averages are staggering. In particular pharmaceuticals and the food processing sub-branch of coal concentrates strike out. The latter displays productivity rates in terms of output per employee, which are over 1,000% higher than corresponding EU-15 averages. Similarly, pharmaceuticals record a value added per employee of more than 1,000% of EU-15 averages. The vast differences in labour productivity can be explained by transfer pricing, which resulted in an estimated overvaluation of output and value added by more than 90%. In terms of the macroeconomic relevance of the Entrepôt economy, output and growth figures readjusted to cater for transfer pricing show that annual average GDP growth would have dropped from 8.2% to 6.2% between 1995 and 1999. Furthermore, total labour productivity in manufacturing would have been 5% lower in the same period (Honohan/Walsh 2002: 55, 57). Hence, the performance of the Irish economy was less glittering than officially assumed, as the readjusted aggregate figures were inline with EU averages. Profit repatriations from subsidiaries to the parent concern take place in the form of loyalty payments, loan repayments and fees for hired business services. Consequently, a constant net factor income outflow has appeared and the services position in Ireland’s balance of payments been continuously negative (Paus 2005: 53) As a result, GDP was overvalued by 20% in comparison to GNP in both 1998 and 1999 (Dauderstädt 2001: 7). The large disparity in GDP and GNP growth resembles not a mere statistical peculiarity of the Irish growth process, but it also tarnishes the record of actual wealth creation during the phase of high economic growth. For instance, Irish GDP per head income was 17% higher than the EU average in 2001. In contrast, when measured in GNP per head, the population was actually 20% poorer. Thus, instead of Ireland belonging to the richest member states in the EU, as stated by incomes figures based on GDP, Ireland had actually attained a below EU average income level, if the relevant measurements had been based upon GNP (Dauderstädt 2001: 7). 4.4.2 Linkages The discussion of the direct economic impact of FDI inflows into Ireland has shown that their primary effects have been unquestionably positive. Foreign-owned firms have contributed to the growth in high technology output, exports and employment. However, the picture in terms of the secondary effects of investing TNCs on the productive structure in the form of linkages between foreign-owned and indigenous firms is far less clear cut. The high incidences of intra-industry and related-party trade together with the high import and export intensity of TNC production in Ireland, indicate that the room for substantial linkages between foreign and indigenous firms is low. 202 Pecuniary and Productive Linkages The contribution of TNCs to the Irish exchequer has been considerable, despite the financing of the FDI attraction strategy and TNC transfer pricing. Continuous FDI inflows and the high profitability of TNCs coupled with the relative low levels of taxation greatly increased revenue collection throughout the 1990s. Consequently, the share of corporation tax in total state revenue tripled. Whilst its share stood at below 5% of total revenues in the 1980s (Paus 2005: 58), corporation tax represented 16% of all collected revenues in 2002 (ORC 2003). Following IDA (2002: 5) estimations on the direct taxes paid by its client firms, TNCs under IDA tutelage paid € 2.2 billion in corporate tax, equalling 43% of total corporate tax collected by the government in 2002 (ORC: 2003). Furthermore, FDI inflows also created additional employment and increased direct and indirect spending effects in the domestic economy. In sum, FDI inflows have led to fiscal buoyancy. Increased tax receipts during the 1990s enabled the Irish state to sufficiently reduce the high level of indebtedness, which accumulated throughout the 1970s and the 1980s. Increased revenues driven by TNC profits and investments eventually allowed Ireland’s participation in the European Monetary Union (Murphy 1998: 19).157 Whilst fiscal linkages have been undoubtedly positive, the extent of linkages in the productive structure between foreign-owned and Irish firms is less straightforward. Due to the fact that TNCs mainly use Ireland as a production base for their EU markets, forward linkages in the form of subcontract export-processing is low. Hence, the analysis of TNC-indigenous co-operation is confined to the assessment of backward linkages, whereby Irish firms supply inputs to TNC affiliates. However, relevant figures are marred due to the inability to differentiate between the nationalities of suppliers, the acquisition of Irish-owned firms by TNCs and the distortion of output figures through transfer pricing (Paus 2005: 85). Available statistics, therefore, can indicate only trends. Bearing these measurement constraints in mind, Paus (2005: ibid.) differentiates between intensive and extensive types of cooperation. The latter describes absolute changes in linkages. The former depicts the change in linkages relative to the development of all input purchases. The concept differs to the official methodology, which portrays locally purchased inputs as a proportion of total output. Extensive or absolute linkages rose inline with the increase in FDI inflows. Total expenditure in the Irish economy by all TNCs, which includes payroll costs and purchases of locally produced inputs and services, grew by an annual average of almost 12% between 1990 and 1994 and by 13% between 1995 and 1999 (Forfás 157 Total tax revenue excluding corporate taxes grew by 76%, corporate taxes increased by 258% between 1990 and 1997. As a result, the Exchequer Borrowing Requirement (EBR) was reduced to 1.2% of GDP in 1996. If the collection of corporate taxes had grown by the same extent as other taxes between 1990 and 1996, namely by 55%, then the EBR would have been higher at 2.8% in 1996, slowing the reduction of the overall debt (Murphy 1998: 19). 203 2006: 44). Although the share of locally produced materials recorded higher increases than locally sourced services, the bulk of TNC local spending was on services (Barry et al. 1999b: 63).158 Whilst absolute purchases for local material inputs grew from € 376 million to € 2.7 billion, locally sourced services increased from € 710 million to € 3.8 billion between 1985 and 2001 (Paus 2005: 87). However, services resemble a black box, as they contain low technology manual services as well as sophisticated tasks. Furthermore, although the demand for services by TNCs is associated with economy-wide increases in employment (Barry et al. 1999b: 64), it does not affect the output of Irish-owned manufacturing firms. Hence, in terms of displaying the linkages between foreign and Irish-owned firms it is necessary to concentrate on locally produced material inputs. Table 12 Intensive Backward Linkages, 1985-2001 (%) Manufacturinga Non-foodb Elec. Equip.c Med. Instr.d TNC IOE TNC IOE TNC IOE TNC IOE 1985 23 - 13 33 8 28 18 36 1990 29 - 19 37 14 34 24 27 1995 26 - 20 37 19 30 28 48 1996 25 - 19 38 16 29 26 42 1997 25 - 20 39 19 30 26 54 1998 26 - 23 38 23 26 28 56 1999 25 - 22 36 23 29 21 62 2000 21 69e 19 45 18 25 19 32 2001 22 75e 20 51 20 34 17 44 Intensive linkages are defined as locally purchased material inputs as a proportion of all sourced material inputs. a NACE 15-37; b NACE 17-37; c NACE 30-32; d NACE 33 Sources: Paus (2005: 86, 89) and e Forfás (2005). Extensive linkages measured as absolute purchases of locally produced material inputs by TNCs grew by an annual average of 11% between 1990 and 1994 and a further 16% p.a. between 1995 and 1999 (Forfás 2006: 44). The table displays the development of intensive linkages comparatively for foreign-owned and Irish-owned enterprises (IOE) in the manufacturing sector. The shares are displayed for the total 158 The problem of output related data could be illustrated by comparing absolute TNC expenditure figures with TNC total economy expenditures as a proportion of TNC output. Due to the large increase in net output during the 1990s, economy expenditure growth disappears. Accordingly, the average increase in TNC local expenditure as a proportion of TNC output between 1990 and 1994 falls to 0.3% and is even negative (-9%) between 1995 and 1999 (Forfás 2006: 44). 204 sector in general as well as for individual important foreign-dominated and dynamic sub-sectors. Paus (2005: 86-89) identifies two trends. First, indigenous manufacturing firms source a far higher share of locally produced material inputs than their foreign counterparts. Second, the table shows that TNCs have initially increased their local sourcing of material inputs until 1997/1998. However, from 1998 onwards foreignowned firms reduced local procurement. Contrastingly, non-food processing intensive TNC reverse linkages have nevertheless increased from 13% to 20% between 1985 and 2001, peaking at 23% in 1998. However, these figures fail to display the technological quality of the estimated backward linkages and the actual extent of Irish-owned firms acting as TNC sub-suppliers. Hence, the displayed portrayal of backward linkages fails to assess the developmental contribution of TNCs to indigenous industry. Although data on the amount of Irish-owned TNC sub-suppliers is not available, veritable linkages between foreign-owned and Irish firms are even officially thought to be low as a result of the increased co-location of TNC sub-suppliers to Ireland (DETE 2003: 87). O’Hearn (2001: 180) cites a Forfás survey from the mid-1990s, which estimates that more than two thirds of locally produced inputs purchased by TNC end-producers in the electrical equipment branch originated from sub-supplier TNCs. Paus (2005: 88) quotes a different survey based on information from the national linkage programme, which estimated that only 6.5% of Irish firms with more than ten employees were acting as TNC sub-suppliers. The majority of which were involved in packaging and printing operations. This prominence can explain the displayed large level of Irish-owned output in this manufacturing branch. Aside from packaging and printing service operations, the quality of backward linkages depends to a large degree on the nature of the respective industry and the situation of the respective affiliate in the international production chain of its parent company. In those foreign-dominated activities susceptible to re-export operations or defined by bulk sourcing and production, such as the sub-sector of pharmaceuticals, the extent of high value linkages between foreign and Irish-owned firms is low (Paus 2005: 88). In contrast, the branch of electrical and electronic equipment is seen as the subsector, where backward linkages have developed furthest. Following Enterprise Ireland figures, sub-supplies for TNCs accounted for 42% of indigenous output in this branch. In total, indigenous electronics sub-suppliers were responsible for 57% of employment in Irish-owned firms (Paus 2005: 89). However, the majority of firms specialised mainly in low to medium technology operations. Moreover, following the post-2001 downturn in ICT demand, the industry is experiencing a severe phase of restructuring with a relocation of standardised production to cheaper locations in Eastern Europe. The room for a further deepening of linkages is, therefore, constrained (Paus 2005: ibid). Closely related to the electronics sector is the foreign-dominated area of software localisation, which enabled the development of the indigenous software manual printing branch. However, the rise of CD-Rom and online manuals displaced the majority of Irish software manual 205 printing and publishing firms with only a handful of indigenous companies remaining active (Jacobson/Kirby 2006: 30-32). According to Paus (2005: 90), the most promising signs for backward linkages have developed in the manufacturing branch of medical and optical engineering. A relatively young industry in Ireland, the medical devices sector is at the forefront of new life science innovations. As a result of an active cluster policy by the state, which encouraged the intense co-operation between scientific research and electronics firms, a promising number of Irish-owned firms increasingly established themselves as suppliers to TNCs. The nature of the industry itself promotes increased co-operation. In difference to the electronics sector, medical and optical engineering is not faced with same short duration of the product cycle, comparable levels of standardisation of production as well as competition-induced global sourcing. Furthermore, end-producers are interested in the close proximity of their suppliers. This results from the complexity of product certification and validation processes resulting from high safety standards and requiring high quality and allowing highly specialised and niche production by Irish firms (Paus 2005: 105-106). However, according to official estimations, the number of indigenous firms operating in this sector is low (DETE 2003: 97). Spillover Effects In terms of spillover effects, the evidence is similarly opaque. Although Paus (2005: 125) argues that TNCs served “as the handmaiden of advancements in the technological capabilities of indigenous companies”, econometric evidence is indecisive and fails to produce a clear picture. Again this is linked to the innate measurement difficulties associated with assessing the quantifiable impact of FDI inflows on the Irish productive structure (Ruane 2001: 7-8). Görg and Ruane (2000: 233) show that linkages in the electronics sector have led to increased employment and sales by indigenous firms. Barry (2004a: 26-27) cites econometric surveys by Görg and Strobl (2003; 2002), which find that foreign-firm presence supports the entry of indigenous firms into the same sector or in downstream industries. Görg and Strobl (2002: 1320) maintain that the entry of TNCs into the Irish economy creates additional demand for locally produced intermediary inputs. As a result, the average cost of intermediaries decreases due to the unleashing of scale economies, eventually allowing the entry of indigenous firms as end-producers. Similarly, this process applies to technological spillovers through the transfer of production knowledge and demonstration effects to indigenous firms (Görg/Strobl 2003: 593). As a result, indigenous competitiveness increased in those sectors, which were foreign-dominated, such as office machinery, communications equipment and medical and optical engineering (O’Malley 2004: 74). However, the cited analyses emphasise that these developmental mechanisms are limited to indigenous firms present in the same high technology sectors as TNCs. Hence, the evolution of direct spillovers only applies to a minority of indigenous 206 firms engaged in similar high technology activities. Economy-wide spillover effects from TNCs enhancing indigenous productivity remain unproven (Ruane/U?ur 2006: 102). The low level of substantive linkages and the problematic nature of spillovers is linked to the nature of the TNC. A survey conducted by Hewitt-Dundas (2002: 38 cited in Paus 2005: 109) found that 21% of the 61 high-technology TNC affiliates participating in the questionnaire preferred suppliers within their parent corporations. Ó Riain (2004a: 82-84) emphasises the priority of control over knowledge assets in TNC software affiliates. Centralised control of the affiliate by the parent firm restricts the level of substantial cooperation between Irish and foreign-owned firms as well as the establishment of high value R&D activities in Ireland. Moreover, the much publicly celebrated arrival of US major software producers in the 1990s ignored the actual content of their Irish operations. The fact that between 19% and 29% of workers employed in the foreign-dominated manufacturing sectors in 2000 were classified as highly skilled was taken as proof that a predominant share of TNC employment and therefore operations were not low skilled (DETE 2003: 75-76). However, these figures do not relay any information concerning the quality of work undertaken by highly qualified personnel. Similarly, the share of the manual workforce in the Irish dynamic high technology manufacturing sectors was higher than comparative UK figures, indicating the less technically advanced nature of Ireland’s leading high technology sectors (Ó Gráda 2002: 59). As O’Hearn (2001: 182) shows, the majority of TNC software firms were involved in the localisation of their products for European end-users. The operations entailed the adaptation of software programmes to individual cultural and language formats. These were tasks, which are technically unsophisticated, engage little or no product development and resemble “repetitive assembly-line work” (O’Hearn 2001: ibid), despite these firms employing high levels of skilled staff. The analysis of R&D expenditure by TNCs paints a similar picture of the low strategic function of TNC affiliates located in Ireland. Barry et al. (1999b: 69) maintain that foreign-owned firms contributed to the augmentation of technological competencies in Ireland through increased R&D spending, adding to the evolution of technological spillovers. However, these distinctly upbeat conclusions overlook the fact that R&D expenditures in Ireland have persistently been low in international comparison, lying well behind respective EU and OECD averages (NCC 2003: 43).159 Although absolute business expenditures on R&D by foreign firms rose from € 228 million in 1991 to € 598 million by 2001 (DETE 2003: 78), the increase can be explained by the rise in FDI inflows throughout the period. Relative TNC affiliate BERD growth rates decreased considerably towards the end of the 1990s, as R&D spending increases were almost 10% lower between 1999 and 2001 compared to the 159 According to the NCC (2003: ibid), Ireland was ranked 11th of 16 countries in regard to BERD as a percentage of GDP and 16th of 16 countries in terms of public spending on R&D in 2001. 207 annual growth averages of 15% recorded between 1993 and 1997 (DETE 2003: ibid). Again, BERD intensity figures relative to output are stricken by transfer pricing effects, with TNC BERD as a proportion of output falling from 1% in 1993 to 0.6% in 2001 (DETE 2003: 80). However, TNCs still spent more than 40% on R&D per employee compared to indigenous firms. Nevertheless, these aggregate figures mask that Irish-owned enterprises actually spent more on R&D in individual sub-sectors such as medical engineering, electrical equipment and software development than foreign firms (Paus 2005: 115).160 Furthermore, if expenditure thresholds are applied, the extent of TNC BERD is considerably smaller. Only an estimated five percent of foreign-owned firms spent more than IR£ 1 million (€ 1.3 million) on R&D in 2001 (DETE 2003: 79). Despite the fact that Ireland has one of the highest levels of high technology export specialisation within the OECD (OECD 2001d), the products were predominately designed and developed abroad with Irish affiliates only engaging in process optimisationrelated R&D (Beer 2003: 57). The low levels of R&D therefore illustrate the lack of strategic functions of TNC affiliates located in Ireland (DETE 2003: 81). Thus, the lack of fundamental R&D activities by TNCs emphasises the propensity of TNCs to relocate the production of mature technology abroad, whilst strategic and high value operations such as marketing and R&D activities remain in the home countries of TNCs. Hence, the propensity of TNCs to engage in transfer pricing as well as the low level of high value strategic functions located to the Irish affiliates of TNCs is seen by Ó Riain (2004a: 54) as a sign that, “the revenues generated in Ireland by foreign firms are not being reinvested in developing local capabilities”. 4.4.3 Irish Indigenous Industry Nevertheless, those foreign firms willing to cooperate with Irish suppliers were still unable to find suitable Irish owned firms capable of producing the required inputs (Paus 2005: 109). The lack of indigenous capacity to sufficiently co-operate with TNCs questions the role of the state to augment indigenous industrial capabilities. The following table displays the performance differences in the Irish manufacturing sector. The indicators compare various TNC performance points discussed above with corresponding figures for Irish-owned firms for 2005. Bearing transfer pricing distortions in mind, the differences between foreign-owned and indigenous manufacturing enterprises are starkly visible and serve to illustrate the dichotomous nature of Irish industry. 160 According to Paus (2005: ibid), Irish firms in the electrical and electronics branch spent € 10,000 per employee compared to € 7,500 invested by TNCs. In medical instrumentation, indigenous BERD per employee stood at € 8,500, whereas foreign firms spent €4,700.