attractions of Scotland to foreign companies include, .... Based on technology and trust. Emphasis on added ...... enter into close 'open-book' relationships. Fig.
Regional Studies, Vol. 27.5, pp. 401417.
Inward Investment and Local Linkages: How Deeply Embedded is ‘Silicon Glen’? I V A N TUROK Centrefor Planning, University of Strathclyde, Richmond Street, Glasgow G1 l X N , U K (Received December 1992; in revised form January 1993) TUROK I. (1993) Inward investment and local linkages: how deeply embedded is ‘Silicon Glen’?, Reg. Studies 27, 401-417. Substantial opportunities were created for local linkages by the sharp growth in output of the Scottish electronics industry in the 1980s. Twelve per cent of material inputs by value are currently sourced locally. Evidence about the nature of these links suggests that many correspond more closely to a simple dependent model than to a self-sustaining development scenario. Most involve capacity- or labour sub-contracting, or supply of relatively bulky, simple components. High value links are limited to branches of foreign suppliers. The reasons for the pattern of sourcing include issues of technology, quality, responsiveness and price. Examples of corporations which stress these attributes are provided. Supply linkages
Scottish electronics industry
Sub-contracting
Dependent scenario
Economic development
TUROK I. (1993) Nach innen gerichtete Investierungen und ortliche Verknupfungen: wie fest verankert ist das ‘Silicon Glen’?, Reg. Studies 27, 401-417. In den achtziger Jahren wurden durch die kraftige Leistungszunahme der schottischen Elektronikindustrie betrachtliche Gelegenheiten zu Verknupfungen am Orte geschaffen. Gegenwartig stammen 12% des materiellen Aufwands dem Werte nach von Auftragen am Orte. Befunde uber die Natur dieser Verknupfungen weisen darauf hin, daB viele eher einem einfachen Dependenzmodell entsprechen als einem sich selbst tragenden Entwicklungsszenario. Die meisten verlangen zusatzliche Kapazitaten oder Arbeitskrafte, oder Zulieferungen verhaltnism8Big umfangreicher, einfacher Bestandteile. Hochwertige Verknupfungen sind auf Zweigunternehmen auslandischer Lieferanten beschrankt. Bei den Grunden f i r das Muster der Herkunftsorte handelt es sich um Fragen der Technologie, der Qualitat, der Aufgeschlossenheit und der Preise. Beispiele von Korperschaften, die diese Attribute betonen, sind beigefugt.
TUROK I. (1993) Investissement direct et liens locaux: dans quelle mesure la ‘Silicon Glen’ est-elle enchlsste?, Reg. Studies 27,401-417. Aux anntes 80 de grandes opportunitts ont CtC crtts en faveur du dtveloppement des liens locaux B cause de la croissance vertigineuse de la production de I’industrie tlectronique ecossaise. En valeur 12% des inputs de mattriaux sont actuellement d’origine locale. Des preuves relatives au caractere de ces liens laissent supposer quc plusieurs se rapportent plus ttroitement B un simple modele de dipendance qu’9 une situation de dtveloppement indtpendant. La plupart comportent la sous-traitance de capacitt ou de main-d’oeuvre, ou l’approvisionnement de pieces relativement grandes et simples. Les liens B forte valeur ne concement que les etablissements des fournisseurs itrangers. La distribution des zones d’origine s’explique par des questions de technologie, de qualid, d’tlasticitt et de prix, entre autres. O n fournit des exemples des socittts qui insistent sur de tels attributs.
Liens d’offre Industrie tlectronique tcossaise Sous-traitance Situation de dtpendance Dtveloppement tconomique
Lieferverbindungen Schottische Elektronikindustrie Vergebung von Auftragen an Subunternehmer Abhangigkeitsszenario Wirtschaftliche Entwicklung
The spin-offs from inward investment for the domestic economy have long been a concern of people interested in economic development. Backward linkages to indigenous suppliers of materials, services and machinery have traditionally been considered one of the main ways in which benefits from inward investment filter through to the host region (e.g. HIRSCHMAN, 1958; KENNEDY,1991). A welldeveloped local supplier base may also attract further investment. It was suggested recently that the main attractions of Scotland to foreign companies include, ‘the educational infrastructure, the supplier base (par-
ticularly f o r electronics), the track record of existing overseas companies and the work of Locate in Scotland in customer care’ (HOOD,1991, p. 20, emphasis added). Historically, emphasis has been placed on the scale and value of local linkage effects and the volume of jobs generated among suppliers, within an essentially static framework. This has been the chief concern of mainstream economic research, reflected in inputoutput and multiplier studies that seek to quantify the short-term impact on employment and income. This has also been the focus of government policy,
401
402
Ivari
demonstrated, for example, by European Commission attempts to impose local content quotas on foreign investment. Local supplier development could also be a way for governments to help create ‘ad&tional’ jobs and improve the balance of payments, since much of the growth should come from import substitution and displacement effects among established firms will be small. The linkage issue has reappeared recently in a different guise, with the burgeoning literature on the geography of new forms of industrial organization, termed ‘flexible specialization’ or ‘post-Fordism’ (see, for example, PIOREand SABEL, 1984; STORPER and WALKER, 1989; AMIN and MALMBERG, 1992; HARRISON, 1992). Greater emphasis is given to the ‘quality’ and dynamic nature of linkages, together with the longer-term implications for local economic development. It is possible to identify two distinctive scenarios emerging from the wide-ranging literature, which I shall term ‘developmental’ and ‘dependent’ (summarized in Table 1). This is a simple and stark dichotomy, admittedly, but it is a useful analytical device to distinguish key features of linkage development discussed in the empirical material presented later. Both scenarios recognize the imperative for major corporations to be more responsive to fluctuating global economic conditions and increasing market diversity. T h s is reflected in the need to accelerate product development and production cycles, and to improve product quality and price competitiveness. It is also associated with certain tendencies towards decentralization and localization. In the developmental scenario, economic pressures are alleged to promote fragmented structures of production for greater flexibhty. The processes of verti-
cal disintegration of large corporations and decentralization of decision-making demand closer, more collaborative relationships between individual plants, suppliers and distributors within the value chain. These encourage stronger geographical clustering to minimize transaction and transport costs, and to facilitate high-level exchange of technical ideas, market awareness and corporate plans. Such ‘clusters’ may become the cores of internally generated growth poles, since the desire of major corporations to increase local sourcing creates market opportunities for new business formation and development. In the process of interaction they transfer valuable technology and expertise to local firms, which become ‘specialist suppliers’. They develop the products and know-how for cumulative growth and the wherewithal to tap new markets at home and abroad. The emphasis in this perspective is clearly on the potential of inward investment to induce all-round development. Corporations get deeply embedded in the local economy through the creation of a network of sophisticated, interdependent linkages, which support the expansion of local firms and generate self-sustaining growth of the cluster as a whole (see, for example, PIOREand SABEL,1984; BEST,1990; PORTER,1990; SCOTT, 3 992). In the dependent scenario flexibility has negative connotations. Local clusters are weak nodes withm a wider network of powerful multinationals. The direct global connections expose local economies to volatile world markets and make them vulnerable to forces of international competition. Moreover, the motives for multinationals extending local linkages are driven more by cost-cutting than by a desire to add value through the exchange of technology and
Table 1. Alternative linkage scenarios: a summary
of the main tendencies
Developmental
Dependent
Nature of local linkages
Collaborative, mutual learning Based on technology and trust Emphasis on added value
Unequal trading relationships Conventional sub-contracting Emphasis on cost-saving
Duration of linkages
hng-term partnerships
Short-term contracts
Meaning of ‘flexibility’
High-level interaction to accelerate product development and increase responsiveness to volatile markets
Price-cutting and short-term convenience for multinationals
Inward investors’ ties to the locality
Deeply embedded High investment in decentralized, multi-functional operations
Weakly embedded Branch plants restricted to final assembly operations
Benefits for local firms
Markets for local firms to develop and produce their own products Transfer of technology and expertise strengthens local firms
Markets for local fmns to make standard, low-tech components Sub-contracting means restricted independent growth capacity
Quality ofjobs
Diverse includmg hgh W e d , high income
Many low skilled, low paid, temporary and casual
Prospects for the local economy
Self-sustaining growth through cumulative expansion of the industrial cluster
Vulnerable to external forces and corporate decisions
403
Inward Investment and Local Linkages
information. Linkages with suppliers are hierarchical and the relationships adversarial rather than cooperative. Considerations of price or short-term convenience are uppermost. Proximate suppliers save corporations labour costs, capital investment and overheads at times of increased competition and falling profits. Suppliers do not participate in the development and technical evolution of the product. Examples include the use of sub-contractors to make standard components previously produced inhouse, or to absorb short-term or cyclical variations in demand- ‘capacity sub-contractors’. Their gross margins are tightly controlled through competitive bidding for contracts. They may be labour intensive to manage large fluctuations in orders without bearing the costs of expensive plant and machinery used discontinuously1 Their work processes are likely to be simple and unskilled staff recruited, often on a temporary, even casual basis. They add much less value to the overall production process than specialist suppliers. Some may be ‘labour-only subcontractors’, where the customer provides whatever components, equipment and expertise are needed on a ‘free-issue’ basis. Such linkages could promote economic and technical dependence. They may undermine the capacity of suppliers to upgrade themselves, by pruning overheads to the bone and depriving them of the resources required to develop the functions of a rounded enterprise, including product development, marketing and strategic management. From this perspective more numerous linkages do not imply improved regional development prospects, stronger local industrial capability or evolution towards a higher skill, higher income economy. Links may prove short-lived or restricted in scope. Inward investments may remain weakly embedded (see, for example, AMINand MALMBERG, 1992; MORRIS and LOWDER, 1992; PERRONS, 1992). This paper contributes to the debate by assembling evidence about the scale and nature of linkages in the Scottish electronics industry. This is an interesting case given its sharp growth and the proposition that a ‘critical mass’ of electronics and supply companies has been established in what is often known as ‘SiLicon Glen’ (HARGRAVE, 1985; BAGGOTT, 1986). Are the developmental and dependent scenarios relevant, and do the linkages conform more closely to one than the other? The paper begins by contrasting the spectacular growth in electronics sales and gross output during the last decade with the modest rise in actual production and employment. Consequently, input purchases have risen steeply and there have been great opportunities for local suppliers of electronic and related components. The third section examines how far this potential has been realized in aggregate terms, followed by a look at the detailed purchasing practices of a sample of major companies in com-
puters and consumer electronics. The fifth section explores some of the reasons for the pattern of local sourcing, focusing on issues concerning technology, quality, responsiveness and price. The conclusion outlines some implications for policy.
ELECTRONICS GROWTH IN SCOTLAND: A POTENTIAL MARKET FOR SUPPLIERS Electronics gross output in Scotland increased fourfold during the 1980s, a very substantial compound growth rate of 14% per annum, while the rest of manufacturing experienced stagnation (SCOTTISH OFFICE, 1992a). By 1990 electronics was a sizeable part of the Scottish economy in some respects, accounting for 20% of gross manufacturing output and 42% of manufactured exports (TUROK, 1993a). Nearly half of all UK exports of computers and peripheral equipment originate in Scotland. However, the direct benefits for the local economy are far less than this suggests, apparent in the relatively slow rise in employment and value added (Figs 1 and 2). Gross output increased at a compound rate of 16% per annum between 1983 and 1989 (the latest date for consistent data available), compared with 7.1% for value added and only 1.8% for employment. Gross output is the preferred indicator of official sources, but is a measure of company sales, not production. Value added is a better guide to the amount of work done by firms in Scotland to develop and manufacture the products sold, and hence of the income generated locally. It is in effect the sales value
7
f
140
2o
-.-.-.-.
1
o!
1
1
1
1
l
1
1
1
Employment Gross output l
1
1
1
Fig. 1. Index of gvoss output and employment for electronics Source: Derived from SCOTTISH OFFICE, 1990, 1991.
404 5000
Ivan Tuvok
-1
0.5 1
A Gross Output
a 0.2
2 Fig. 2. Electronics gross output and value added Note: The original figures expressed in current prices were deflated to constant 1990 prices using the U K PPI for electrical engineering and electronics to take account of the effects of inflation. Source: Census of Production. of firms' output less the cost of materials and services bought from other firms and government bodies.* So it indicates the degree to which basic inputs are transformed and given real value in the process of production. The share of value added in gross output was only 24.2% in electronics in 1989, compared with 34% in the rest of manufacturing (Census of Production). Furthermore, it had fallen steadily from 39.2% in 1983. There was also a big difference between UK- and foreign-owned segments, with the lattcr falling faster and from a lower starting point in 1983 (Fig. 3). The widening gap indicates the rapid growth in electronic products shipped from Scotland, but the slower growth in the amount of actual production here. It seems that the character of the industry has been changing, with less value added as sales increased, and more materials, components and services bought in. This was probably the result of two main processes. First, there was vertical disintegration among existing firms, which were increasingly buying-in more ready-made components, sub-contracting more services and doing less manufacturing in-house. Second, there was a shift towards simple final assembly and test operations among new inward investment plants, linked perhaps to a change in sectoral composition.
B
0.1
-
:
0.0 1983
UK firms Foreign F i s I
I
I
I
I
I
1984
1985
1986
1987
1988
1989
Fig. 3. Amount of value added in gross output for electronics Source: Census of Production. The widening gap between the gross output and value-added trends (Fig. 2 and Table 2) indicates the major opportunities there were for Scottish firms to capture a share of the input purchases, at least in principle. In 1989 the total value of these was considerable at over 23,500m. If this had all been spent in Scotland it could have generated up to 50,000 jobs directly. This estimate is derived by dividing the value of purchases by the average gross output per employee for manufacturing in Scotland in 1989569,000 (SCOTTISH OFFICE, 1992b). Furthermore, no less than 85% of all inputs were purchased by forcign-owned cnterprises (compare 1989 figures in Tables 2 and 3), so their procurement policies were crucial for the prospects of local suppliers. Before examining the proportion of products sourced from Scotland, the diverse pattern of electronics growth should be noted for future reference. Fig. 4 shows that data processing (computers and peripheral equipment) experienced the most dramatic increase in gross output, followed by electronic components (including passive components such as printed circuit boards (PCBs) and connectors, active components such as integrated circuits, and sub-
Table 2. Growth indicatorsfor the Scottish electronics industry ~~
Total sales Gross output Gross value-added 'Input purchases'
1983
1984
1985
1986
1987
1988
1989
1,577 1,573 616 957
2,221 2,299 806 1,493
2,871 2,840 953 1,887
2,692 2,685 882 1,803
3,237 3,278 1,016 2,262
4,105 4,115 1,183 2,932
4,908 4,939 1,197 3,742
Note: Figures are in Am current prices: terms are defined in note 2. Source: Census of Production.
405
Inward Investment and Local Linkages Table 3. Growth indicatorsforforeign-owned electronicsfirms
Total sales Gross output Gross value-added ‘Input purchases’
1983
1984
1985
1986
1987
1988
1989
1,017 1,009 350 659
1,609 1,666 510 1,156
2,144 2,100 620 1,480
1,785 1,771 499 1,272
2,299 2,341 643 1,698
3,111 3,111 769 2,342
3,950 3,979 806 3,173
Note: Figures are in Am current prices. Source: Census of Production.
225 1
Data Processing Components Instrument Eng Other Electronics
1 I T
I
I
I
I
I
I
I
I
I
I
I
I
Fig. 4. Index of gross outputfor electronics sub-sectors Source: SCOTTISH OFFICE, 1990, 1991. assemblies). The next section examines whether the growth of these segments was connected. Other parts of the industry, such as instruments, telecommunications, electronic capital goods and alarms, have declined since 1986. Consumer electronics has risen from a low level recently. This is obscured within the ‘other electronics’ category.
THE ACTUAL LEVEL OF LOCAL SOURCING Comprehensive statistics on the extent of local purchasing by electronics firms in Scotland are not available. However, periodic surveys of the industry have been undertaken by the Scottish Development Agency-now Scottish Enterprise (SE)-in 1986, 1988 and 1990, and by the author in 1991. The samples were selected in different ways so the results are not exactly comparable. However, the response rates were high, particularly for the largest companies, so the results are likely to offer a fairly reliable indication of purchasing patterns and of the trend over time.
Table 4 shows a consistent picture between 1986 and 1991, particularly in the share of products sourced within Scotland. This seems fairly low at 12%, although it would be helpful if there was more comparative data from elsewhere. The surveys also suggest there has been no increase in the degree of integration or ‘clustering’ of the industry and its suppliers over time. Some caution may be warranted at this stage bearing in mind that local suppliers appear to have maintained their share of a sharply increasing total volume of input purchases. It may also be early days in the process of supplier development. If the 12% local sourcing statistic is applied to the overall figure for input purchases in Table 2, the value of purchases from suppliers in Scotland amounts to approximately &450m in 1989. This implies about 6,500 jobs generated directly. This is derived by dividing the value of local purchases by the 269,000 gross output per manufacturing employee. This is not insubstantial, but amounts to only 14% of total employment in electronics in 1989. Table 5 provides a breakdown of product sourcing by company ownership, drawn from the 1991 postal
406
Ivan Tuvok Table 4. Source of material inputs by value (% oftotal) Scotland
Rest of UK
Rest of Europe
Asia
USA
Rest of World
12 15 12 12
23 25 30 30
9 9 16 19
8 30 25 30
38 9 12
5 9
8
1
SDA 1986 survey’ SDA 1988 survey’ SDA 1990 survey3 Author’s 1991 survey4
7
Note: 1. Data available for 101 firms (85% of total jobs in electronics). 2. Data available for 60 (73% of companies employing over 50 people). 3. Data available for 35 (57% of total expenditure by electronics firms). 4. Data available for 106 (46% of all electronics firms,but oriented towards larger ones). Figures in rows may not sum to 100% due to rounding or incomplete reporting in SDA surveys. Service inputs are not included but are generally only a s m a l l fiaction of the value of material purchases. Source: SDA, 1986, 1988, 1990; and author’s 1991 postal survey. Details in the Appendix.
Table 5. Source
of material inputsfor
UK andforeign
j r m s (“/o Oftotal)
UK-owned Foreign firms
Scotland
Rest of UK
Rest of World
25.5 14
55.4 31.1
19.1 54.9
Source: Author’s 1991 postal survey: details in Appendix.
survey. Foreign firms were far more likely to import components and have weak local linkages than U K companies. This is ominous given that most growth has been coming from overseas firms. Incidentally, the proportion of local sourcing is higher in Table 5 than in Table 4 because it is based on averaging the spending pattern of each firm, in effect treating them as equally important. T h s obscures the fact that larger firms tend to buy less locally than smaller firms. Despite the relatively low present levels of local sourcing by foreign firms, there appears to be some interest on their part in raising this. The 1991 postal survey asked whether firms had altered the proportion of material inputs bought in Scotland in the previous three years. Table 6 shows that almost half the foreign firms that responded said the level of local sourcing had increased. Altogether 30% of firms said local sourcing had risen compared with 13% which said it had declined. These statistics are indicative rather than conclusive since evidence of the actual change in the value of local sourcing is not available.
Table 6 . Change in local sourcing in last three years (number ofjirms) More UK-owned Foreign firms Total
Less
19 14 14 0 33 (30%) 14 (13%)
Unchanged
Total
47 17 64 (57%)
80 31 111 (100%)
Source: Author’s 1991 postal survey: details in the Appendix.
THE PATTERN OF LOCAL SOURCING To provide further insights into the impact of foreign firm purchases on the local economy it is important to go beyond the aggregate statistics and examine the kinds of products sourced locally and imported. Fig. 5 shows the value of key component purchases by thirteen of the largest foreign electronics companies in Scotland in 1991. They were selected by SE as the multinationals spending most on material inputs. Eight manufacture data processing equipment (mainly computers) and five make electronic consumer goods. The figures exclude expenditure on basic electronic components such as microprocessors, capacitors and resistors. Very few of these are sourced in the UK, except through distributors stocking catalogued products. The purpose of the SE study was to establish the level of spending on key components and the potential for increasing the amount of local sourcing, either by enhancing the capabilities of local suppliers or by attracting suppliers from elsewhere into Scotland to meet particular component requirements. Since the participating firms recognized the potential benefit of the exercise to themselves, they provided a level of detail about their purchasing patterns that is not normally made available because of the political and market sensitivity of this information. The total value of their product purchases was &1,407m, &319m of which was spent in Scotland. This is a larger proportion than the figures quoted in Table 4, partly because of the above-mentioned omissions. The sample is also slightly untypical in excluding Scotland’s seven semiconductor fabrication plants, which have few local purchases because they form part of integrated global production processes. Nevertheless, the thirteen sampled firms are clearly a very large and important group in absolute terms. Their &319m of local purchases represents a sizeable share of all electronics purchases within Scotland in 1991. Several important points emerge from Fig. 5. First, populated PCBs are by far the most valuable
407
Inward Investment and Local Linkages
Spent elsewhere
H Spent in Scotland
single type of component, as well as the most important sourced locally. They are the essential building blocks of all electronic products, and incorporate active and passive electronic components mounted onto a bare PCB platform. A quarter (26%) of the populated PCBs procured by these firms were made in Scotland. This is surprisingly low considering the relatively straightforward technology involved in pin-in-hole (PIH) assembly and the substantial productive capacity that now exists here (see below). Surface mount technology (SMT) is a more recent and complicated assembly process since everything is miniaturized, but it is still a standard rather than a proprietary technology. There are four large specialist PCB-assembly plants in Scotland: Avex, Phillips, SCI and Timex. They have highly automated PIH machinery and upto-date SMT equipment. All are foreign-owned and they employ about 2,500 people altogether. Two have had long-established plants in Scotland making other things, and diversified into PCB assembly during the 1980s. The other two are recent inward investors which took advantage of the growing market for PCB-assembly work here. So it is foreign-owned rather than indigenous firms that have secured the lion’s share of local PCB-assembly contracts. Second, only 18% of bare PCBs (worth &14m) were sourced locally by the thirteen major customers, with an unquantified additional amount by the PCB assembly firms. Scotland has two of the
three largest PCB manufacturers in the UK in Exacta Holdings and Prestwick Holdings, as well as a host of smaller specialist producers (TUROK,1993b). Exacta and Prestwick are high volume specialists and employ about 1,000 people altogether. They appear to be competitive internationally: one exports over 50% of its output and the other 15-20%. In addition, the technology of P C B production is hybrid and less sophisticated than for many electronic components. Barriers to entry were relatively low in the past, although volume production now demands costly equipment. PCBs are also made to order products with much of the R&D carried out by customers. So proximity to suppliers offers considerable advantages for collaboration. For all these reasons one might have expected a higher level of local purchases. Third, certain kinds of products have distinctly greater local sourcing than the rest. They include keyboards (32%), plastics (45%), cables (61YO), sheet metal (65%), packaging and printed material (79%). Most are apparently simple to manufacture, using standard machinery or labour-intensive methods. They also have relatively low price-to-bulk ratios and high transport costs. They are more straightforward than PCBs to make and bulkier to transport. They are the kinds of products that firms responding to the 1991 postal survey said they had bought more of locally in the last three years (Fig. 6). Many of the local suppliers are Scottish-owned or UK subsidiaries, including Fullarton Fabrication,
408
Ivan Turok
14
.e-a 0
.2
B Y
.d
&
a
..
P
I
i
ln
Fig. 6. Change in products sourced in Scotland in the last three years Note: Thirty-three firms increased and fourteen cut back local sourcing. Source: 1991 postal survey.
Simclar, Lithgow, Kinloch, Mimtec, TFC and Turnkey. These seven employ over 3,000 people altogether. They are simple sub-contractors without their own technology and products. Some are labour-intensive assembly operations with little plant and machinery. Many were started within the last decade and have grown quickly. Some of their customers provide all the materials and equipment they need on free issue, which reduces cash flow difficulties and alleviates the financial constraints small growth firms face. Plastics are slightly different. Much of the recent increase in local sourcing has gone to newly opened branch plants. Companies such as Tenma (from Japan), Tilling (Middlesex), Plastic Engineers (Wales) and Silleck Mouldings (Teesside) have invested in substantial automated plastic moulding equipment. The high cost of investment and inexperience may have deterred Scottish suppliers. Fourth, more sophisticated, higher value components such as disk drives, monitors and power supplies are almost completely imported into Scotland. Until recently there were no significant manufacturers in Scotland. The indigenous disk drive producer Rodime was an exception. It enjoyed great success early on and expanded to 550 local employees in 1987, but then declined and went bankrupt in 1991. A small firm called Domain Power started manufacturing power supplies in the late 1980s, but was taken over by a foreign company after struggling to survive and grow. There have been several inward investments in these sectors since
then, such as Conner Peripherals and YE Data, which make disk drives. The evidence suggests that local sourcing is fairly low overall, with an eighth of electronics inputs acquired in Scotland, but the pattern is also diverse. A sizeable share of relatively bulky components is bought locally, resulting in extensive growth of a fair number of indigenous firms. There is less purchasing of technically sophisticated products, although there are some signs that this may be beginning to change, mainly through foreign suppliers establishing local branches. Some are ‘capacity sub-contractors’ that have invested in high volume equipment, as in plastics and PCB assembly. Others are ‘specialist suppliers’ with their own product technology, as in disk drives. O n the whole, indigenous firms seem to be restricted to making relatively straightforward products, with some exceptions.
TOWARDS A N EXPLANATION: DEMANDING CUSTOMERS A N D SUPPLIER WEAKNESSES The full explanation for the pattern of sourcing is complicated and would require detailed analysis of the characteristics and policies of customers, the performance of suppliers and the relations between the two groups. This would be a major undertaking with several hundred diverse electronics and related enterprises in Scotland (SCOTTISHENTERPRISE, 1991), and given their sensitivity about discussing these issues. In the absence of a comprehensive study the following account is exploratory. The postal survey provides useful preliminary insights. It incorporated questions about the difficulties faced by buyers and suppliers. On the whole, major customers were inclined to blame weaknesses in the supply base for limited local sourcing, whereas suppliers accused customers of stringent purchasing procedures. Seventy-seven firms had tried to increase local procurement. Fig. 7 shows the difficulties encountered. Unavailability of materials such as special metals and chemicals, products such as power supplies and monitors, or basic components such as resistors and capacitors, emerged most often. The scale of t h s response suggests that the supplier infrastructure in Scotland is not as extensive as sometimes alleged. The main reservation about existing suppliers concerned prices. Half the respondents said local firms were more expensive than suppliers elsewhere. Price has become increasingly important as conditions in electronics get more competitive and pressures to cut costs mount. Other drawbacks of local suppliers include inadequate quality, long delivery times and unreliability, which I return to later. There were interesting differences between the attitudes of UK and foreign customers (Fig. 8). Foreign companies appeared more demanding than UK
Inward Investment and Local Linkages
Fig. 7 . D ficulties encountered in increasing local sourcing Note: Seventy-seven firms said they had made an effort to increase local sourcing; sixty-one firms mentioned specific difficulties in doing so. Source: 1991 postal survey.
E Oa70 0.60
i ,
8
i
409
firms, particularly in terms of price and quality. These seem to be key areas for attention if foreign firms are to extend local sourcing voluntarily. Turning to suppliers, the same kinds of difficulties emerged with customers, but from a chfferent perspective. Twenty-three said they encountered problems in meeting the requirements of major customers, including low price expectations, irregular orders, short-term contracts and short lead times (Fig. 9). Price and general short-term behaviour were most important: suppliers felt that unwarranted pressure was exerted on both. Some blamed their customers’ poor organization for the short notice they were often given on orders. Others blamed their customers’ inefficiency for their attempts to secure cost-reductions from suppliers rather than themselves. Some felt customers deliberately kept them on short contracts to maintain pressure and disciplme. To explore these issues in greater depth, detailed interviews were held with the purchasing managers or other senior staff from ten of the largest foreign firms in Scotland (most employing over 500 people), ten of the largest and fastest growing suppliers (all employing over 150 people) and ten smaller, static suppliers. A sample of ten key customers is reasonable since they have a disproportionate share of total purchases. The interviews were supported by an analysis of company accounts to get ‘objective’infor-
UK f m s 0
0 Foreign firms
.H
’c1
0.50
??
0.40 3
“0 0.30
s
0.20
8
a 0.10 0.00
Fig. 8. Difficulties in increasing local sourcing: U K and foreign firms Note: Thirty-four U K firms and twenty-seven foreign firms mentioned specific difficulties. Source: 1991 postal survey.
410
Ivan Turok
18
f
44
:
”0 12
jj 10
S 8
2
6
4 2
n
Fig. 9. Dgjculties suppliersface with major customers Note: Twenty-three firms said they faced difficulties meeting the requirements of local customers. Source: 1991 postal survey.
mation on growth rates, investment, income distribution and other performance measures (details in TUROK, 1993a). Surprising diversity emerged between the purchasing practices of different firms, reflecting the nature of production processes, the speed of technical change in market segments, the nationality of ownership and global corporate strategies. Yet, there were common patterns too, promoted by growing international competition. Leading companies were introducing more advanced procurement procedures and more elaborate supplier evaluation criteria, in order to help reduce the time-to-market of their products, to raise quality standards and to control prices. It is helpful to structure the analysis around four broad factors: technology, quality, responsiveness and price. Their order of importance differs between firms and examples are provided to elaborate and illustrate each. Evidence is presented in this form partly because of the considerable difficulty in securing equivalent information on a wider basis.
Technology Technology is an important barrier to increased sourcing from indigenous firms. Relatively few are specialist suppliers of technology-based components. This is why many firms in the postal survey stated that certain products were unavailable. They import these components or encourage their suppliers at
home to establish branches in Scotland to service them more effectively. The undeveloped state of indigenous suppliers is linked to difficulties securing financial support, lack of management experience and less risky opportunities in other segments. Firms seeking to develop their own products claim to be hampered by the lack of long-term ‘patient’ resources from financial institutions or government agencies. Experience shows that technology takes time and is costly to develop, obvious barriers-to-entry for new small firms. The managers of some suppliers that succeeded in raising funds and enjoyed early success-such as Rodimemade serious strategic errors which jeopardized their existence. Established U K electronics firms in Scotland-such as Ferranti, Marconi and Racal-have been preoccupied with small-batch, high-priced products for protected military markets. This has left them relatively uncompetitive and impeded their diversification into expanding markets for high volume civil electronics goods (see, for example, MCKINSEY & Co., 1988). It is unclear whether technology is becoming more or less important in relation to other factors such as price and quality. On the one hand, the growth of ‘open systems’ in data processing means a decline in proprietary technology and computers becoming more standard commodities. This should shift the emphasis in component purchasing towards price and away from specialist technology. On the other hand, technology is becoming important in more subtle ways. It is helpful to distinguish between ‘hard’ technology -associated with investment in tangible products and processes-and ‘soft’ technology which is linked with skills and knowledge. Some customers are looking for suppliers to offer extra added value beyond simple subcontracting. They want suppliers to have equipment they do not possess themselves, to provide expertise with their own development and production engineers, or to provide knowledge about the market and competitors that stems from operating internationally. These save customers’ costs, provide an intelligent resource to assist in problem-solving, and help identify ways of increasing quality and cutting waste. It is a moot point how far customers are willing to pay for these resources through higher product prices. Some differentiate between ‘preferred suppliers’, who are specialists and tend to operate world-wide, and other suppliers of lower value components and simple sub-contractors, whose overheads they are inclined to curtail. Sun Microsystems is one of a new breed of electronics companies that attach particular importance to suppliers with technology. Its corporate strategy is highly focused and has proved very successful in the decade since it was established. It concentrates on designing and manufacturing computers for a speci-
41 1
Inward Investment and Local Linkages fic, expanding market -workstations. The key factor for success in this competitive sector is the ability to introduce new products quickly and in large volumes. To ensure flexibility and rapid response the company tries to contain its physical expansion and avoids vertical integration, a traditional corporate growth strategy. Instead, it limits in-house functions to the highest value-added activities and relies on external suppliers for the bulk of components. Its close ‘partnership’ relationshps with leading global suppliers start with component design and extend through to volume production. Suppliers are expected to invest heavily in new production techniques and product development. There is a high degree of joint working and risk sharing. Sun opened its third manufacturing and supply facility-its first outside the US-in Scotland in 1990. Europe was the fastest-growing regional market and Sun’s intention was to become a global company, so it had to establish a manufacturing presence here. The plant is restricted to final assembly of the central processing unit of workstations, automated testing, burn-in and packaging. This is a simple operation, although there are also plans to start assembling its PCBs locally. Few inputs are sourced in Scotland, partly because of the demanding nature of its requirements for most components. First, suppliers need to operate on a global basis, to provide engineering support for Sun’s manufacturing plants in the US. Ultimate approval of suppliers takes place at US headquarters, which militates against European suppliers unfamiliar with American standards. Other US companies with European plants, such as Hewlett Packard and IBM, purchase on a regional rather than a global basis, which makes things easier for local suppliers. Second, suppliers have to be able to expand to high volume production rapidly to match Sun’s strategy of being quick to the market with new products. This demands considerable resources and experience. Small indigenous firms cannot cope with t h s requirement. Third, Sun expects suppliers to be open about their cost structures and seeks to secure progressive price reductions across its components of about 15% on average each year. There is pressure on suppliers because 7580% of the cost of production is in the material inputs. This is a key feature of its cost-competitiveness. Sun states that European suppliers are often unable to match its specificationsor reluctant to enter into close ‘open-book’ relationships. Fig. 10 shows the immediate source of inputs for its Scottish plant. The European share exaggerates the ultimate spend here because over 90% comprises PCB sub-assemblies from an English supplier and much of their value consists of basic electronic components from Asia. The other components sourced
Asia nonJapan - 15
Ask Japan - 49%
Fig. 10. Source of Sun’s material inputs, by value: 1991-92
Source: Company. within Britain have relatively low value: metalwork, plastics and packaging cost less than &lm. Sun also buys country kit assemblies (keyboard, mouse and manuals) from a firm that packages them in Scotland, but buys them elsewhere. Sun sub-contracts this function because the skills involved are low and it is space intensive. Many of Sun’s supply linkages seem to have much in common with the value-added, technology-based links described in the developmental scenario. But few of the beneficiaries have been Scottish, let alone European. Interestingly, the company had been in the process of searching for European partners when the plant was opened, because of E C local content rules. However, these were relaxed soon afterwards, which was fortuitous for Sun since it removed the obligation to increase local sourcing. Compaq is another firm with similar procurement patterns. Quality Most corporations expect rising standards of suppliers in terms of the proportion of defective components and the sophistication of their manufacturing control systems. They expect progressive elimination of ‘waste’, defined as the amount of equipment, materials, space and workers’ time not directly adding value to the product. Quality was the second major drawback mentioned in the postal survey (Fig. 8). ‘Total Quality Management’ implies a companywide approach focusing on prevention rather than detection and remedial action. Suppliers are expected to introduce techniques to assist monitoring of production, such as Statistical Process Control. They verify the reliability of processes, eliminating the need for incoming inspection of components by customers and so saving them time and money. They also make it easier for suppliers to engineer product and process improvements by providing a complete record of each stage of production and data measuring the effectiveness of efficiency efforts.
412
Ivan Turok
It is common for major electronics firms in Scotland to require their suppliers to operate independently certified quality assurance systems such as BS5750 or ISO9000, which specify procedures to ensure consistent standards. They are intended to build in reliability at every stage to reduce defects and cut time spent on modifying designs, correcting faults and re-working products. Many sub-contractors are critical of these formal systems. They appear simple and straightforward, and are therefore popular for customers to insist upon as a matter of routine. However, they are costly to introduce because of the extensive documentation. They can also be largely irrelevant because of their comprehensive character. For some suppliers, they are little more than marketing tools to ensure they get short-listed for contracts. Motorola attaches importance to quality beyond standard approval systems. It has two major plants in Scotland employing over 3,000 people altogether. One makes semiconductors and is long-established. The other assembles cellular telephones. The former concentrates on wafer fabrication and testing, with some design and assembly work. Low defect rates are crucial, since process yields determine profitability and chips control the reliability of the products they are used in. This is also important in cellular telephones, because the customer base includes businesses and governments, who pay a premium for quality considerations. The emphasis on quality is pervasive: For many years Motorola has built its business on a fundamental objective of total customer satisfaction. In its quest to attain this, the area of customer quality expectations was a key consideration. In 1981, the company set a goal of 10 times improvement in quality in the ensuing 5 years. This strategy proved so successful that in 1986, a further goal of Six Sigma quality-i.e. near perfection-was set for 1992. The key words to achieving near-perfect quality are cooperation and partnership, with suppliers and with customers (MOTOROLA, 1990).
The semiconductor plant introduced a complex supplier appraisal system in 1989, based on a range of criteria including physical and financial capabilities, lead times, use of SPC, scheduling, inventory management and supporting documentation. These are condensed into measures of quality, service and delivery, and weighted so%,25% and 25%, respectively, to produce an overall index of performance. Once suppliers are selected, appraisal is an ongoing process linked with training to secure improvements over time. The plant also compares its purchasing achievements against others in the company. Motorola has also used its appraisal procedure to reduce its supplier base and build long-term partnerships with key firms. It has been moving away from
old-style arm’s-length purchasing, based on multiple sourcing and competitive bidding, towards more interactive and selective relationships. It has traditionally sourced materials and components worldwide, with much from Japan. Recent procedures may have prompted several overseas suppliers to establish local branches here to service Motorola and other semiconductor plants more effectively. Examples are Shin-Etsu, Heraeus, Countdown and OCG. Compugraphics makes Motorola’s photomasks and is the only notable Scottish-owned supplier. Overall, Motorola’s links with suppliers are strongly influenced by quality considerations. The shrinking supplier base is one effect, as it strengthens ties with ‘world-class’ vendors. Few indigenous firms have benefited, although several foreign firms have branches which contribute several hundred jobs to the local economy between them. Its links with suppliers are of the durable kind outlined in the developmental scenario, except that-like Sun-few of them are based in Scotland and even fewer are under local ownership and control. Responsiveness
Corporations expect suppliers to demonstrate a high level of commitment to them, including a positive and accommodating attitude, willingness to react at short notice, flexibility in the service provided and on-time delivery. Some also want suppliers to have procedures to utilize their sales forecasts as planning tools and to reducc their lead times once firm orders are placed. The introduction of systems designed to cut stocks and shorten production cycles-such as just-in-time- have increased pressure on suppliers. Geographical proximity can reinforce dependability by enabling firms to exert more control over suppliers. The use of electronic data interchange among a few major companies to speed up communication and cut paper-work reduces some but not all of the requirements for physical closeness. The critical issue for several of the companies interviewed was the attitude of the supplier’s ownermanager. This is fundamental to all aspects of responsiveness and is the ultimate safeguard for customers. IBM attaches particular importance to responsiveness. It is the largest single foreign-owned electronics establishment in Scotland with around 2,300 employees. It is also by far the biggest customer of local suppliers and is responsible for supporting several thousand jobs among them. In the 1980s it introduced a more rigorous supplier evaluation scheme, which assesses the attitudes of key staff as well as their technical and commercial capabilities. According to a senior IBM manager, standard approval procedures such as BS 5750: don’t look at the one thing that makes a supplier tickand that’s the people. We sub-contract a lot of our prob-
Inward Investment and Local Linkages
413
lems to our supplier base-that’s the way we do business. They are our manufacturers and risk sharers in partnership, so we have to know what makes them tick. We start by evaluating the managing director: what he likes as a person, his abilities, the level of his commitment and his business goals. We then assess the extent to which these aims and philosophies are reflected throughout the company (quoted in REES,1990, p. 62).
The Greenock plant manufactures mostly personal computers (PCs) for Europe. Until the early 1980s it made small batches of larger computers and was highly integrated vertically. The introduction of PCs transformed the process into mass production. The high volumes (currently 1.5 million units a year) and volatile sales pattern prompted much greater use of sub-contractors. In addition, the parent company did not want the plant to employ many more people, for a variety of reasons related to the full employment commitment to its permanent staff and the damage that would be inflicted on the local economy and the company’s reputation from a downturn. It also wanted to limit the scale of investment in machinery in view of the rapid shifts in technology and uncertainty about its lifespan. Consequently, ‘capacity sub-contractors’ have been used to manage uneven demand and off-load production volumes. Suppliers engaged in high volume PCB assembly include the foreign-owned plants mentioned above. Smaller indigenous subcontractors are used to absorb unexpected peaks in demand and to make mature products no longer needed in large batches. They have to show flexibility and dexterity, able to respond quickly to urgent orders of different kinds. IBM provides the materials and equipment on free issue for about 15% of purchases. Successful sub-contractors have reserves of people willing to undertake assembly work at short notice, often during evenings and weekends. Firms acknowledge that their operations are largely reactive: their planning horizons are short, their fixed costs and overheads are minimized, and there is little scope or inducement to invest. The large volumes put out to sub-contractors and the short turn-around often expected mean that proximity to the plant is quite important. This is reflected in the unusually large share of inputs sourced within Scotland and the rest of Britain (Fig. 11). Other factors include the simple nature of many basic components, the ability of IBM in the past to afford the higher costs of sourcing these in Britain rather than Asia because of its market position, and the company’s long-standing policy of contributing something to the localities in which it operates. There is an element of self-interest in this, since local purchases help to ensure a supplier infrastructure for the future. Local sourcing also enhances image and reputation. The standards of responsiveness expected differ between companies. Several recent inward investors
Rest of UK - 31%
Fig. 11. Source of IRM’s material inputs, by value: 1990 Source: Company. are critical of indigenous sub-contractors. Considering their turnover levels, management structures are said to be underdeveloped, machinery obsolete, operational procedures inadequate, training limited and staff wages and commitment low. Some prosperous owner-managers are thought to be complacent and have little incentive to modernize. Since they tend to rely heavily on a few key companies, others are wary of using them for fear of an inferior service. To conclude, some of IBM’s purchases are governed by the need for responsiveness. It seeks simple sub-contractors willing and able to handle discrete ordcrs at short notice. Proximity is an advantage, resulting in numerous links with local firms and substantial resource flows. Many of these relationships appear to be based on short-term convenience, consistent with the dependent scenario. Price
Price has always been important in sourcing decisions. The trend towards increased outsourcing and rising productivity (e.g. Fig. 2) helps to explain why the cost of external inputs is even more significant. For individual companies input purchases can represent 85% or more of the gross cost of production. Sourcing decisions traditionally focused on the ‘piece price’, i.e. the price quoted by suppliers per component. This put UK firms at a disadvantage compared with those in Asia, in particular, where labour costs were lower and suppliers’ time horizons longer. Most corporations now use a measure of the total cost, including freight, import duties, inspection, defects, engineering support to suppliers, delays in delivery and managing stock. One effect is to redress the balance between local and foreign suppliers, particularly for bulky items such as plastics and metalwork with high costs of transport and stock-holding. Forward-looking companies seek suppliers with the capability to improve productivity continuously
414
Ivan Turok
in order to secure steady price cuts. Some offer their own engineers and accountants to help suppliers identify ways of improving efficiency. A figure of 15% year-on-year price reductions is often mentioned as a target. This is difficult for labourintensive sub-contractors lacking machines that can be speeded up or processes that can be upgraded incrementally. There are limits to the productivity increases that can be secured through work intensification (TUROK, 1989). One way customers achieve price cuts is to compare suppliers against each other and foster rivalry among them. In some cases this is explicit and firms are ranked in ‘benchmarking’ exercises on detailed productivity and cost criteria. This is also used to identify areas of potential cost-saving. Not surprisingly, some suppliers acknowledge frustration with the demands for price reductions. They say they are expected to submit their lowest possible bid to win the initial contract and then to find scope for further efficiencies subsequently. Consumer electronics companies appear to attach most importance to price considerations, presumably because price competition has traditionally been fiercer than in data processing or military electronics. In addition, Japanese plants have a more disciplined, ‘lean’ approach to production, apparent even on casual inspection. They may also be better placed than US firms to compare prices in Asia. Many are part of giant, vertically integrated corporations that manufacture many components in-house, so there is a natural inclination to source them internally. Some managers acknowledge that sourcing from parent firms may also be a means of repatriating profits through transfer pricing. Mitsubishi Electric has two major consumer electronics plants in Scotland employing 1,300 people altogether. One makes video cassette recorders (VCRs) and the other colour televisions. Both are assembly plants although the VCR facility also has a drum-cylinder polishing and finishing line, a higher value-added activity. The markets for VCRs and TVs are highly competitive with little to choose between the quality of the main producers. Price is a key determinant of success. The VCR plant was opened in 1983 and has expanded steadily in output and employment. In 1991 it produced 450,000 units and employed 800 people. It is a replica of Mitsubishi plants in Japan, without the higher level functions. The VCRs are Japanese designs with minor adaptation for European technical and cosmetic requirements. The assembly process is tightly organized. Material inputs represent 75% of the cost of production, and most are sourced from Japan (Fig. 12). Foreign suppliers in Britain are used for specific items such as tuners, remote controls and power supplies. Local sub-contractors are used for plastics, metal-
Fig. 12. Source of Mitsubishi’s material inputs, by value: 1991
Source: Company.
work, packaging, printing and basic PCB and mechanical assemblies. The sub-contracted assembly work duplicates internal capacity but helps to deal with the seasonality of VCR sales by absorbing production peaks. Sub-contractors also assemble the oldest products whose low volumes do not justify Mitsubishi devoting its own high capacity production line to them. Mitsubishi is obliged by E C regulations to procure 40% of inputs in Europe. Otherwise, it would source less here since local suppliers are found to be more costly than in Japan. Some are also more wary of open-book negotiations about pricing structures for fear of losing control. In addition, the Scottish plant is obliged to purchase many inputs through the parent corporation in Japan. It tends to select Japanese suppliers for historical reasons and convenience. There is also greater market power in purchasing components for the Scottish and Japanese plants together, enabling cheaper deals to be negotiated. To conclude, Mitsubishi and other Japanese plants such as JVC and OK1 attach great importance to supplier prices. Other corporations may not do so to quite the same extent since they operate in different markets, are less vertically integrated and have different needs. The emphasis on price militates against local suppliers, except for bulky items and urgent requirements. E C policy encourages more purchases withm Europe than would otherwise occur. These features do not imply strong, two-way developmental linkages within Scotland.
C 0N C L U S I 0 N The evidence assembled suggests that inward investment in the Scottish electronics industry is not very deeply embedded. Overall, about 12% of material inputs are sourced locally, which seems fairly low. This proportion has at least stayed constant over a period of sharply increased total purchases. Com-
Inward Investment and Local Linkages parative analysis for other regions and industries would help to put this into perspective. More complete evidence about the nature of local linkages is difficult to collect and further research is needed in Scotland before definitive conclusions can be reached. The evidence gathered to date indicates that many of the links with indigenous suppliers correspond more closely to the simple dependent model than to the developmental scenario. There are few high value-added, technology-based linkages, except with the branches of foreign suppliers in Scotland. Their high-level R&D, strategic management and marketing functions are based elsewhere, so local learning and transfer of technology may be limited. Many links with local firms involve capacity or labour sub-contracting, or the supply of bulky nonelectronic components on a reactive basis. They serve useful purposes for major corporations, including cost-saving, convenience and less risk. There are clearly benefits for the local economy in terms ofjob numbers, and some sub-contractors have grown rapidly to employ hundreds of people. However, the extent to which firms have developed into more rounded enterprises is unclear. It requires further investigation, probably on a sectoral basis. None of the electronic sub-contractors interviewed had developed their own products or technology, although some had begun to tap wider markets. Bare PCB manufacturers are good though fairly isolated examples (TUROK,1993b). Overall, the simple conclusion to be drawn from the pattern and nature of linkage formation is that the prospects for self-sustaining, internally generated growth of the Scottish electronics industry cannot yet be described as promising. Whether it is possible to progress from an essentially dependent situation to a developmental one remains unanswered. From the point of view of inward investors, the reasons for the limitations of local sourcing include technological weaknesses of indigenous firms, doubts about their quality and reliability, and their relatively high prices for some products. Other factors include the branch-plant character of many foreign firms, their absence of strong product design and procurement functions locally, and their protected supply relationships with their parent corporations or global suppliers. There appears to be scope for some change in the situation. There is certainly a need considering the level of import penetration. It may require policy actions at local and European levels to make a significant difference, and should be directed at the supplier and customer sides of the linkage relation. There is a large potential market in Scotland for components of various kinds. Some inward investors are looking for more local suppliers because of the advantages of proximity. Measures to build up the capacity of the supply base, through targeted investment, training
415
and enterprise development, might prove effective at capturing a larger share of their purchases. Competitive suppliers on price and quality might progress to exploit wider markets in Britain and Europe. Cooperation between focused suppliers could provide the integrated package of products and services that seems to be expected by some corporations. Scottish Enterprise have a small team of engineers working to improve the management skills and quality controls in a group of local sub-contractors. At first sight, their Supplier Development Programme appears to be achieving useful results with individual firms. In view of the considerable potential, it ought to be expanded, given a wider remit and a higher profile. Many sub-contractors require substantial investment in advanced equipment to compete for high volume contracts, as in PCB assembly work. There are still considerable amounts produced elsewhere. There is even scope for local firms to secure a larger share of orders for cables, metalwork, keyboards and plastics. Given the tradition of the Scottish mechanical engineering industry with similar basic processes, this should not be infeasible. Some local suppliers need persuasion to reinvest more of their surpluses to raise productivity and shift their activities in the high-income, hgh-skill direction. Some customers are looking for sub-contractors offering additional value-added services. The challenge to establish firms with distinctive products and technology is bound to be difficult. This probably requires long-term commitment to an intensive kind of support that has not been acceptable in Britain for years, although it is pursued with apparent effectiveness in some other countries. Established firms such as Ferranti and Marconi with advanced technology experience in military electronics could be good starting points for efforts to diversify into expanding civil sectors. With a ready market for high value components on the doorstep, the potential could be considerable. Joint ventures with inward investors is another option that has not been seriously considered. They might aim to build the collaborative, technologybased partnerships that many multinationals are seeking to develop. In the past, the struggle to attract inward investment has placed regions in a weak position, reluctant to convey the impression they arc making any impositions. Co-ordination by the EC might provide the framework within which localities could begin to negotiate closer, mutually beneficial relationships with foreign investors. The E C could also insist that conditions are attached to regional assistance to support efforts to develop local suppliers. Selective assistance in the U K already has employment obligations attached and could be extended to include linkage conditions. There are few technical reasons why some corporations source so little within Europe.
416
Ivan Turok
Acknowledgements- The research was supported by the Economic and Social Research Council (award number 232783). The author acknowledges thc assistance of staff from the Scottish Office Industry Department and Scottish Enterprise in providing unpublished data, and the cooperation of many Scottish electronics companies in surveys and interviews. Helpful comments on an earlier draft were reccived from Peter Booth, Keith Hayton and two anonymous referees.
NOTES 1. Electronics is defined in the paper to include data processing equipment, electronic components, electronic instruments, telecommunications, electronic capital goods and consumer goods, i.e. the following ten activity headings from the 1980 SIC: 3302, 3444,3453, 3710, 3732, 3433, 3441, 3442, 3443 and 3454. The evidence is drawn from an eighteen-month study of the electronics industry, including a comprehensive postal survey of firms in 1991 (see Appendix), follow-up interviews with ten major customers, ten large local suppliers and ten smaller suppliers in 1992, secondary survey data from government sources, consultations with industry experts and a review of local and national newspapers for the last decade. 2. A summary of the official definition of terms is as follows: Sales of goods produced = sales during the year, whether or not they were produced in the year of the return. It also includcs sales of goods made from materials given out to sub-contractors or outworkers. The value of sales is the ‘net selling value’, i.e. the amount charged to customers less VAT, trade discounts, etc. Transfers of goods produced by a business to departments not covered by the return are treated as sales, valued as if sold to an independent purchaser. Gross output = the value of total sales and work done (including industrial services such as repairs and maintenance, installation work and technical research for other organizations) adjusted by the change during the year of work in progress and goods on hand for sale. Value added (gross value added at factor cost) = gross output minus the cost of purchases of materials for use
in production, purchases of goods for merchanting, output minus the cost of purchases of materials for use in production, purchases of goods for merchanting, services received and rates. This approaches the definition of net output or value added in national accounts statistics.
APPENDIX A postal questionnaire survey of all electronics manufacturing firms in Scotland was undertaken between June and September 1991. The definition of electronics included the ten SIC activity headings listed above. The basic source of information on company names and addresses was a range of business directories produced by local authorities and government agencies responsible for promoting business and economic development in Scotland. These were checked and sifted to exclude firms whose main activity was not electronics manufacturing. After follow-up calls and postal reminders, a total of 148 usable responses were received from 232 relevant establishments, representing a very satisfactory response rate of 64%. The four sectors with most firms were passive components (SIC heading 3444) with 20% of the firms, followed by electrical instruments (3443- 19%), active components and sub-assemblies (3453- MY0), and data processing (3302- 13%). The responding firms employed 32,387 people altogether-about 70% of the official figure for all electronics jobs in Scotland. Almost half (49%) of all the jobs in responding firms were in just eight plants employing over 1,000 people: 92% of the jobs were in 53 firms employing over 100 people. At the other end of the spectrum, firms with between 1 and 24 staff (38% of all firms) had only 2% of the jobs. Half the jobs were lowskilled manual occupations such as assembly operatives, followed by skilled technical jobs (22%), clerical and administrative (12%), managerial and executive (9%), and skilled manual (8%). Nearly a third (32%) of the plants had opened bctween 1986 and 1991, 31% between 1980 and 1985, 21% between 1970 and 1979, and 17% before 1970. Six plants had an annual turnover of more than &100mand 32 between 210 and &100m,compared with 34 firms with sales of less than &0.5m.A longer summary of the results is available on request from the author.
REFERENCES AMINA . and MALMBERG A . (1992) Competing structural and institutional influences on the geography of production in Europe, Environ. Plann. A 24, 401-16. BAGGOTT M. (1986) The vital support industries, Scot. Econ. Develop. Rev. No. 4, pp. 14-17, Scottish Development Agency, Glasgow. BESTM . H. (1990) The New Competition: Institutions of Industrial Restructuring. Polity Press, Cambridge. O F PRODUCTION (various years) Obtained from Scottish Office Industry Department, Glasgow. CENSUS HAKGRAVE A. (1985) Silicon Glen: Reality or Illusion?. Mainstream, Edinburgh. HARRISON B. (1992) Industrial districts: old wine in new bottles?, Reg. Studies 26, 469-83. A . (1958) The Strategy of Economic Development. Yale University Press, New Haven, CT. HIRSCHMAN HOODN . (1991) Inward investment and the Scottish economy, Roy. Bank Scot. Rev. 169, 17-32. K. (1991) Linkages and overseas industry, in FOLEYA . and MCALEESE D. (Eds) Overseas Industry in Ireland. Gill KENNEDY & Macmillan, Dublin. MCKINSEY COMPANY INC. (1988) Pevformance and Competitive Success: Strengthening Competitiveness in U K Electronics. N E D O , London.
Inward Investment and Local Linkages
417
MORRISA. S. and LOWDEK S. (1992) Flexible specialisation: the application of theory in a poor-country context: Leon, Mexico, Int. /. Urban Reg. Res. 16, 190-201. MOTOROLA (1990) European Corporate Fact Sheet 1990/92. Motorola, East Kilbride. PERRONS D. (1992) Undefended cities and regions facing the new European order, report on an international seminar, Int. J . Urban Reg. Res. 16, 318-20. PIOREM. and SABELC . F. (1984) The Second Industrial Divide. Basic Books, New York. PORTER M. (1990) The Cornpetitive Advantage ofNations. The Free Press, New York. REESR. (1990) The best of working partnerships, Boardroom Report, November, pp. 6044, DTI, London. SCOTTA . J. (1992) The role of large producers in industnal districts: a case study of high technology systems houses in Southern California, Reg. Studies, 26, 265-75. SCOTTISH DEVELOPMENT AGENCY (1986) Scottish electronics companies survey, SDA, Glasgow. SCOTTISH DEVELOPMENT AGENCY (1988) Scottish electronics companies survey, S D A , Glasgow. SCOTTISH DEVELOPMENT AGENCY (1990) Scottish electronics companies survey, SDA, Glasgow. SCOTTISH ENTERPRISE (1991) Electronics and support compaiiies in Scotland, SE, Glasgow. SCOTTISH OFFICE(1990) The electronics industry in Scotland, Statist. Bull. No. C1.3, Edinburgh. SCOTTISH OFFICE (1991) The electronics industry in Scotland, Statist. Bull. No. C1.4, Edinburgh. SCOTTISH OFFICE (1992a) Index of industry production and construction for Scotland, Statist. Bull. No. D2.4, Edinburgh. SCOTTISH OFFICE(1992b) Scottirh Economic Bulletin, No. 45. HMSO, Edinburgh. STORPER M. and WALKER R. (1989) The Capitalist Imperative: Tewitory, Technology and Industrial Growth. Basil Blackwell, Oxford. TUROKI. (1989) Evaluation and understanding in local economic policy, Urban Studies 26, 587-606. TUROK 1. (1993a) Contrasts in ownership and development: local versus global in Silicon Glen, Urban Studies 30, 365-86. TUROK I. (1993b) Making connections: the growth of a Scottish cluster in electronics. Strathclyde Pap. Plann. No. 24, Centre for Planning, University of Strathclyde.
