EC, September 23, 2009, No 2010-54

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, and in particular the first subparagraph of Article 88(2) thereof, Having regard to the Agreement on the European Economic Area, and in particular Article 62(1)(a) thereof, Having regard to Council Regulation (EC) No 659-1999 of 22 March 1999 laying down detailed rules for the application of Article 93 of the EC Treaty [1], Having called on interested parties to submit their comments pursuant to those provisions [2] and having regard to their comments, Whereas:

1. PROCEDURE

(1) By e-mail of 16 July 2007, Poland pre-notified the Commission of the above-mentioned measure. A meeting with the Polish authorities was held on 2 August 2007.

(2) Pursuant to Article 88(2) of the EC Treaty, by e-mail dated 24 December 2007, Poland notified the Commission of the proposed aid in favour of Dell Products (Poland) Sp. z o.o. The case was registered as N 775-07.

(3) By letters dated 15 February, 2 June and 3 October 2008, the Commission requested additional information, which was provided by the Polish authorities by letters dated 18 April, 1 July, 1 August, 6 August, 17 November and 9 December 2008. Meetings were held with the Polish authorities on 5 March, 26 June and 19 November 2008.

(4) By letter dated 10 December 2008, the Commission informed Poland that it had decided to initiate the procedure laid down in Article 88(2) of the EC Treaty in respect of the aid. By letter dated 18 February 2009, Poland provided its comments on the opening decision.

(5) The Commission decision to initiate the procedure was published in the Official Journal of the European Union [3]. The Commission called on interested parties to submit their comments.

(6) The Commission received comments from Ms Kathy Sinnott, a Member of the European Parliament (letter dated 24 February 2009), from Hewlett-Packard (HP), a competitor of Dell (letter dated 24 February 2009) and from another competitor [4] (letters dated 2 March and 12 March 2009).

(7) By letters dated 24 March and 2 April 2009 the Commission forwarded these comments to Poland, which was given the opportunity to react; its observations were received by letter dated 22 April 2009.

(8) By letter dated 25 May 2009 and by emails dated 17 June and 8 July 2009, the Commission requested further information, which was provided by Poland by letters dated 10 June 2009, and by emails dated 22 June, 13 July and 10 August 2009.

2. DETAILED DESCRIPTION OF THE AID

2.1. The beneficiary

(9) The aid beneficiary is Dell Products (Poland) Sp. z o.o. (Dell Poland) [5], a company wholly owned by Dell Inc. (USA) through a 99,9 % stake held by Dell Global BV (Netherlands) and a 0,1 % stake held by Dell International Holdings VIII BV (Netherlands).

2.2. The investment project

(10) Poland aims to support economic development in Lódz Province by providing regional investment aid to Dell Poland for an investment project in the Lódz Special Economic Zone (SEZ).

(11) Lódz Province is a region with substantial socioeconomic handicaps which is eligible for regional aid under Article 87(3)(a) of the Treaty with a standard regional aid ceiling for large enterprises of 50 % gross grant equivalent (GGE) according to the Polish regional aid map for 2007-2013 [6].

(12) The investment project consists in setting up a new plant for the manufacture of PCs, including desktops and notebooks, and servers [7]. Work on the project started in 2007; the project is expected to be completed by end-2012.

(13) The project involves the creation of approximately 1200 direct jobs, with a possible increase to 3000, which should have a positive impact on the regional and national economy. Further indirect investment in the region by suppliers of Dell Poland is expected to amount to approximately EUR 53 million, which will attract additional investors, producers and subcontractors and therefore further support job creation and economic development.

(14) The investment project comprises total eligible expenditure of PLN 793603000 in nominal value (present value EUR 189578180) [8]. The eligible expenditure is calculated on the basis of investment costs. A breakdown per year of eligible expenditure is presented in the table below.

Eligible expenditure (PLN million, nominal value)

2007 | 2008 | 2009 | 2010 | 2011 | 2012 | Total |

[...] [9] | [...] | [...] | [...] | [...] | [...] | 793,603 |

2.3. The aid

(15) The Polish authorities have confirmed that the application for the aid was submitted by Dell in May 2006. On 19 September 2006 the Polish authorities and Dell signed a Memorandum of Understanding (MoU) which sets outs the conditions of the aid. In a letter dated 13 November 2006 the Polish Ministry of the Economy indicated that the MoU constituted a letter of intent confirming that the granting of the aid was conditional on the Commission's approval.

(16) The total amount of the aid which Poland is planning to grant [10] to Dell Poland amounts to EUR 54,5 million in nominal value, equivalent to PLN 216365000 in nominal value calculated at the exchange rate mentioned in the same document [11], which corresponds to an aid intensity of 27,81 % [12] GGE (present value). The total amount of aid includes (cf. table below):

- the notified aid, covering different forms of ad hoc aid for an amount of PLN 159,92 million in nominal value, which consists of:

(a) a grant for initial investment of PLN 94,64 million in nominal value;

(b) a grant for job creation of PLN 54,08 million in nominal value;

(c) a 55 % reduction in the Lódz SEZ management fees payable by Dell Poland in 2007-2017, equivalent to PLN 6,2 million in nominal value;

(d) costs for the relocation of the energy supply line outside the plot of land belonging to Dell Poland of PLN 5 million in nominal value;

- tax exemptions granted on the basis of existing aid schemes implemented pursuant to Commission Regulation (EC) No 1628-2006 of 24 October 2006 on the application of Articles 87 and 88 of the Treaty to national regional investment aid [13] for an amount of PLN 56,445 million in nominal value, which consists of:

(a) an exemption from corporate income tax for 2008-2018 (granted under aid scheme XR 98/07) [14], equivalent to a maximum of PLN 46,945 million in nominal value;

(b) an exemption from property tax for 2008-2018 (granted under aid scheme XR 164/07) [15], equivalent to PLN 9,5 million in nominal value.

Overall aid (million PLN, nominal value)

NB: No corporate income tax is due on any of the aid.

[]

| 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | Total |

Grant for initial investment | - | - | 58,91 | 7,94 | 5,96 | 21,83 | - | - | - | - | - | 94,64 |

Grant for job creation | - | - | 23,27 | 10,26 | 20,55 | - | - | - | - | - | - | 54,08 |

Reduction of SEZ fees | - | - | 1,4 | 0,6 | 0,6 | 0,6 | 0,6 | 0,6 | 0,6 | 0,6 | 0,6 | 6,2 |

Relocation of energy supply line | - | - | 5 | - | - | - | - | - | - | - | - | 5 |

Corporate income tax exemption [16] | : | : | : | : | : | : | : | : | : | : | : | 46,945 |

Real estate tax exemption | - | - | 1,9 | 0,95 | 0,95 | 0,95 | 0,95 | 0,95 | 0,95 | 0,95 | 0,95 | 9,5 |

Total | | | 90,48 | 19,75 | 28,06 | 23,38 | 1,55 | 1,55 | 1,55 | 1,55 | 1,55 | 216,365 |

(17) The aid is granted on condition that the beneficiary maintains the manufacturing facility at the site for a minimum period of 5 years after the completion of the investment and that the jobs created through the investment are maintained for a minimum period of 5 years from the date the post was first filled.

(18) The Polish authorities have undertaken that financial support for the project will not be cumulated with any other financial support disbursed for the same eligible costs from any other sources.

(19) The Polish authorities have undertaken to submit to the Commission a final detailed report including information on the aid amounts paid and on implementation of the aid contract and any other investment projects at the same establishment or plant within 6 months of payment of the last tranche of the aid.

3. GROUNDS FOR INITIATING THE FORMAL INVESTIGATION PROCEDURE

(20) In its decision to initiate the formal investigation procedure pursuant to Article 88(2) of the Treaty (the opening decision), the Commission expressed doubts as regards the compatibility of the aid with the provisions of point 68 of the Guidelines on national regional aid for 2007-2013 [17] (the RAG).

(21) Point 68 of the RAG reads as follows:

"68. Where the total amount of aid from all sources exceeds 75 % of the maximum amount of aid an investment with eligible expenditure of EUR 100 million could receive, applying the standard aid ceiling in force for large enterprises as per the approved regional aid map on the date the aid is to be granted, and where:

(a) the aid beneficiary accounts for more than 25 % of sales of the product(s) concerned on the market(s) concerned before the investment or will account for more than 25 % after the investment, or

(b) the production capacity created by the project is more than 5 % of the market measured using apparent consumption data for the product concerned, unless the average annual growth rate of its apparent consumption over the last five years is above the average annual growth rate of the European Economic Area's (EEA) GDP,

the Commission will approve regional investment aid only after a detailed verification, following the opening of the procedure provided for in Article 88(2) of the Treaty, that the aid is necessary to provide an incentive effect for the investment and that the benefits of the aid measure outweigh the resulting distortion of competition and effect on trade between Member States."

(22) To conduct the tests referred to in points 68(a) and 68(b) of the RAG, it is necessary to define the relevant product markets and the relevant geographic markets for the products concerned by the project, i.e. desktops, notebooks and servers [18].

(23) In accordance with point 69 of the RAG, it is necessary to determine the relevant product market to which each of the products concerned belongs [19]. The Commission noted in the opening decision (points 56-57 and 63) that desktops and notebooks on the one hand, and servers on the other, form separate product markets, as these two sets of products are not mutually substitutable.

(24) In order to examine whether the project fulfils the condition laid down in point 68(a) of the RAG (the market share test), the Commission analysed Dell's market share at group level of the relevant markets concerned before and after the investment. As the investment project started in 2007 and is expected to be completed by end-2012, the Commission took 2006 and 2013 as reference years [20].

3.1. Relevant product markets for desktops and notebooks

(25) As regards the market definition for desktops and notebooks, the Commission noted in the opening decision (point (61) that a significant price difference exists between desktops and notebooks with similar characteristics, functions and intended use. The Commission also noted that, while substitution occurs primarily in one direction (from desktops to notebooks), a certain customer segment is unlikely to shift from one product to the other. The Commission noted furthermore that switching production and marketing from one product to another could involve significant adjustment costs, depending on the structure of manufacturing and sales operations of manufacturers.

(26) As regards the conditions laid down in point 68(b) of the RAG (the capacity increase test), the Commission noted in the opening decision (point (90) that the production capacity created by the project exceeded 5 % of the EEA market, measured by apparent consumption data (expressed as sales) for each of the three products concerned (cf. table below).

Production capacity created by the project (EEA)

Source: IDC data.

| 2007 | 2012 |

PCs: | [0-1] % | [10-20] % |

Desktops | 0,0 % | [10-20] % |

Notebooks | [0-1] % | [10-20] % |

Servers (whole sector): | 0,0 % | [10-20] % |

(27) As regards the second part of the capacity increase test, the Commission noted in the opening decision (points 95 and 96) that, as regards desktops (considered as a separate product market), the annual growth rate in sales for 2001-2006 was negative in value terms and thus significantly below GDP growth in value terms [21] (cf. table below).

Annual growth rate in sales (2001-2006, EEA)

Source: IDC data.

| Units | Value (factory revenue) |

PCs: | 11,02 % | 4,10 % |

Desktops | 3,00 % | -4,00 % |

Notebooks | 28,60 % | 16,68 % |

Servers (whole sector): | 13,28 % | 7,82 % |

Low-end servers | 14,08 % | 3,47 % |

(28) The Commission pointed out in the opening decision (point (94) that if desktops and notebooks are considered to form a single product market, the aid measure does not exceed the two thresholds set out in point 68(b) of the RAG.

(29) The Commission therefore expressed doubts in the opening decision (points 62 and 105) as to whether desktops and notebooks could be considered to form a single relevant market for the purpose of this Decision.

3.2. Production capacity created by the project

(30) In the opening decision (points 98 and 99) and subsequently as part of the formal investigation, the Commission considered the options of limiting the production capacity created by the project to 5 % of the EEA market in the case of desktops or of taking the view that this limitation is achieved by taking account of the beneficiary's overall capacity in the EEA and not only the effect of the capacity created by the project, i.e. balancing the capacity created by the project with concomitant reductions in capacity in other Member States.

(31) The Commission expressed doubts in the opening decision (points 99 and 105) as to whether it was appropriate to consider the effect the capacity created by the project had on Dell's capacity as a whole in the EEA instead of considering the capacity increase by Dell Poland alone.

3.3. Use of volume figures for calculating growth rates

(32) The Commission pointed out in the opening decision (point (97) that, in general, value figures are better adapted for signalling potentially declining sectors, but that growth rates calculated on the basis of value figures may not adequately reflect the relative dynamism of high-tech sectors, where significant unit price reductions occur principally due to strong competition and to technological change. The Commission also noted in the opening decision (point (100) that the decline in average selling prices of desktops, as expressed by value figures, is an indication of the strength of a market which is subject to continuous quality improvements and where new products are being introduced.

(33) The Commission expressed doubts in the opening decision (points 100 and 105) as to whether it could base its assessment solely on volume figures for the purpose of calculating the growth rates of the product concerned by the investment project established in point 68(b) of the RAG.

3.4. Relevant product market delineation for servers

(34) As regards the market definition for servers, the Commission proposed in the opening decision (points 65-67), in line with previous merger decisions [22], to use segmentation based on product class and price range which distinguishes between: (i) the volume (low-end) server market, consisting of all systems with an average selling price (ASP) below USD 25000; (ii) the mid-range server market, consisting of all systems with an ASP between USD 25000 and USD 499999; (iii) the high-end server market, consisting of all systems with an ASP of USD 500000 and above.

(35) However, the Commission also noted in the opening decision (point (69) that in previous merger decisions [23] it had acknowledged that servers could be differentiated according to x86 and non-x86 instruction set architecture (processor type). Nevertheless, the Commission noted in the opening decision (point (70) that, although segmentation according to product class and price range could not capture all competitive interactions in the server market, alternative market definitions captured market dynamics to a much lesser degree. Given that Dell is only present in the low-end segment, the Commission noted in the opening decision (point (68) that this segment alone should be taken into account.

(36) As regards the market share test, the Commission pointed out in the opening decision (point (83) that Dell's share in the PC market before and after the investment in any of the possible relevant product markets and relevant geographic markets, both in volume and value terms, was below 25 % (cf. table below).

Dell's market share for PCs

Notes: EEA figures exclude Cyprus, Estonia, Iceland, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta.

Source: IDC data, Dell forecasts.

| 2006 | 2012 |

| Units | Value | Units | Value |

PCs - EEA: | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

Desktops - EEA | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

Notebooks - EEA | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

PCs - worldwide: | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

Desktops - worldwide | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

Notebooks - worldwide | [10-20] % | [10-20] % | [10-20] % | [10-20] % |

(37) However, as regards servers, in the event of the x86/non-x86 delineation being used, the Commission pointed out in the opening decision (points 71 and 86) that Dell's share of the server market exceeded 25 % [24] of sales of the product on the market concerned, before or after the investment, only when considering its share at worldwide level of the market segment for x86 servers (cf. table below).

Dell's market share for x86 servers

Source: IDC data, Dell forecasts.

| Units | Value (customer revenue) | Value (factory revenue) |

| 2006 | 2012 | 2006 | 2012 | 2006 | 2012 |

EEA | [20-25] % | : | : | : | : | : |

Worldwide | [25-26] % | [25-28] % | [20-25] % | [20-25] % | [20-25] % | : |

(38) The Commission therefore expressed doubts in the opening decision (points 72 and 105) as to whether the segmentation of servers according to product class/price range constituted the appropriate market delineation.

3.5. Relevant geographic market for servers

(39) According to point 68 of the RAG, the relevant geographic market should normally be the EEA [25].

(40) However, as noted in recital 37, if the x86/non-x86 delineation is used and if the worldwide level is considered as the relevant geographic market for servers, then Dell's share exceeds 25 % of the market segment for x86 servers (cf. opening decision, points 71 and 79-81).

(41) The Commission therefore expressed doubts in the opening decision (points 78-81 and 105) as to whether the relevant geographic market for servers should be considered to be the EEA or the worldwide level.

(42) The Commission pointed out in the opening decision (point (108) that should the information provided in the course of the formal investigation not lead to the conclusion that the aid measure fulfilled the conditions set out in point 68 of the RAG, it would have to conduct an in-depth assessment of the aid measure.

(43) The Commission therefore called on Poland to submit its comments and to provide any information for the purpose of assessing the aid measure, and invited interested parties to submit their comments (cf. opening decision, points 107-109). In addition to comments concerning the doubts expressed as to whether the measure complied with point 68 of the RAG, in order to conduct its in-depth assessment the Commission sought information on the objective of the aid, the appropriateness of the aid instrument, the incentive effect, the proportionality of the aid, the crowding-out of private investment, and the effect on trade (cf. opening decision, point (108).

4. COMMENTS FROM INTERESTED PARTIES

4.1. Comments from Ms Kathy Sinnott

(44) Ms Sinnott expressed concerns about the alleged relocation of manufacturing operations from Ireland to Poland. She also complained that the Polish authorities were able to award aid to Dell whereas it was not possible to award such aid in Ireland.

4.2. Comments from HP

(45) The comments received from HP only concern the definition of the relevant markets for desktops/notebooks and servers.

4.2.1. Comments regarding the relevant product market for desktops/notebooks

(46) HP considers that the relevant product market for PCs includes desktops, notebooks (including netbooks and mini-notebooks), workstations and thin clients. HP notes that these products are based on the same or equivalent technology, have similar or equivalent components and similar capabilities, run on the same or equivalent operating systems and software applications and can be used for the same or similar purposes.

(47) HP considers that there is demand-side substitutability between: desktops and notebooks; high-end desktops or notebooks and workstations (mobile and fixed workstations); desktops or notebooks and thin clients (mobile and fixed thin clients). HP notes that the degree of substitutability is not complete in each case, but it is substantial and sufficient for market definition purposes. HP considers furthermore that any attempt to raise prices for notebooks or netbooks relative to desktops, or vice versa, would cause marginal customers to switch demand to such an extent that reduced demand would outweigh increased revenue from higher prices.

(48) HP takes the view that there is also supply-side substitutability, because a significant proportion of PC manufacturers are active in all four segments and are easily able to switch production and component procurement to meet changing conditions in any particular product segment. HP believes that, different average prices for desktops, notebooks, thin clients and workstations notwithstanding, a single market exists. It takes the view that the prices of all PC products are constrained by other categories of PC due to chains of substitution.

4.2.2. Comments regarding the relevant product market delineation for servers

(49) HP does not take a firm position on the market definition of servers but suggests reconsidering the relevance of the market segmentations used in previous merger decisions and whether there is a unified market for servers.

(50) According to HP, the delineation by product class/price range is based on the availability and segmentation of the data provided by the consultancy firm IDC rather than on sound market analysis considerations.

(51) HP considers that this market delineation is not relevant because: (i) prices are still falling, and customers compare not just the initial purchase price, but the total cost of ownership; (ii) with the emergence of distributed data-processing and scalable servers, customers can replace a higher-end server by clusters or networks of lower-end servers or scale up existing servers; (iii) customers can and do migrate within and between segments; (iv) specific processing tasks can be performed by various types of server; (v) security is available across all ranges of server, and reliability of mid-range or high-end servers can be met e.g. through redundancy in the design of server clusters and networks; (vi) there is supply-side substitutability between servers belonging to different price bands considering that most major manufacturers produce servers across all price bands; (vii) there are cases of new entry by manufacturers in the server market, which does not appear to be limited to a given price band; (viii) the presence of chains of substitution does not suggest a redefinition of the server market; (ix) any cut-off in the continuum of server characteristics and capabilities is arbitrary, as specific price bands appear not to reflect verifiable technical or commercial considerations.

4.2.3. Comments regarding the relevant geographic market for servers

(52) HP considers that the relevant geographic market for servers is at least EEA-wide and possibly at the level of the whole Europe, Middle East and Africa (EMEA) region. In HP's view, servers are often traded across borders and multinational customers may source servers at their convenience, using central rather than distributed procurement. According to HP, there are indications of parallel trade and arbitrage in the server market which suggest that a narrower geographic market definition would not be applicable.

4.3. Comments from a competitor

(53) The comments received from another competitor concern market definition issues and the criteria for the in-depth assessment.

4.3.1. Comment regarding market definition issues

4.3.1.1. Comments regarding the relevant product market for desktops/notebooks

(54) The competitor considers that desktops and notebooks form separate product markets (i.e. the PC sector consists of two separate product markets). The competitor notes that the physical characteristics of desktops and notebooks differ significantly and that mobility is a key distinguishing factor. The competitor indicates that although the performance of desktops and notebooks is not significantly different in the entry-level and mid-range price bands, high-performance desktops remain more powerful than any notebook.

(55) The competitor notes that there remains a certain amount of stability in terms of demand: certain customers are very unlikely to switch from desktops to notebooks or vice-versa. The competitor notes that the shift in sales from desktops to notebooks in recent years has taken place despite remaining price differences between these two products. According to the competitor, if the two products formed a single market, there would be high levels of switching in both directions - and especially, for cost and value reasons, from notebooks to desktops. The competitor notes that although the prices of both products decreased significantly during 2006-2008, the relative decrease for desktops is less marked. According to the competitor, a full price correlation analysis would not show a strong, direct price correlation between desktops and notebooks.

(56) The competitor furthermore indicates that there are certain significant differences in the manufacture, marketing and distribution of the two products. The competitor notes that switching production from one product to another involves a certain investment, takes a relatively long time, and is a strategic decision taken with a view to long-term profitability and market position. According to the competitor, major PC manufacturers face substantial barriers to switching, as evidenced in 2007 with the arrival of netbooks, when incumbent manufacturers were slow to respond by marketing similar products of their own. The competitor points out that if supply-side switching did not involve technical, commercial and financial barriers, this response would have been quicker.

4.3.1.2. Comments regarding the production capacity created by the project

(57) The competitor considers that the notification concerns the investment in Poland and, therefore, that it is the effects of this particular investment which the Commission is required to assess.

4.3.1.3. Comments regarding the use of volume figures for calculating growth rates

(58) The competitor claims that volume figures are not a sufficient basis on which to refrain from conducting a detailed verification of the aid and considers that growth in sales by value should be used as an effective benchmark for the performance of the desktop market [26].

(59) In particular, the competitor notes that value figures provide a truer indication not only of the current performance of the market but also of future prospects for development. Regardless of any increase in the volume of sales, if the value of sales declines, which will entail a fall in revenue obtained, this will deter new market entrants, make commercial expansion unattractive to incumbent manufacturers, and ultimately encourage market consolidation or the exit of certain manufacturers from the market concerned.

(60) The competitor notes that in an efficiently-performing and growing market, a fall in average selling prices will be offset by growth in the volume of sales, preventing any loss of revenue overall. Only in an underperforming market will a fall in average selling prices necessarily also lead to a decline in the value of sales. In the notebook market, average selling prices have also been falling but, given the strong growth in volume terms, the value of sales has not declined over the period.

4.3.1.4. Comments regarding the relevant product market delineation for servers

(61) The competitor considers that the appropriate market delineation for servers is based on instruction set architecture (x86/non-x86 delineation).

(62) The competitor considers that the x86/non-x86 delineation is an industry standard means of classifying servers that is used by manufacturers, customers and market analysts. In the competitor's view, the server sector is clearly divided between two different types of product based on two different technologies: x86 servers, which use industry-standard architecture and non-x86 servers, which use any other architecture. The competitor stresses that these two products are inherently different in terms of their technical characteristics, and this difference informs their performance, functionality, end-use, production, and the customer segments to which they are marketed and distributed.

(63) The competitor agrees with the view expressed by the Commission in the above-mentioned HP/Compaq merger decision that market definitions based on price bands "cannot fully capture all competitive interactions in the technologically highly dynamic server market." The competitor indicates that there are technical and performance differences according to server architecture. In the competitor's view, x86 servers remain the industry-standard for basic computing, but they cannot be used for more demanding functions. Non-x86 servers are capable of performing functions that are beyond x86 servers, which are limited by Intel's original technology. The performance attributes of x86 and non-x86 servers are comparable only at the most basic level. Servers based on x86 architecture are mainly used for common and undemanding tasks, whereas a non-x86 server is likely to be intended for complex processing.

(64) While non-x86 servers range from basic mini computers to very complex mainframes, this is not the case with x86 servers, where examples of high-end processing are rare. Where a company may have a number of x86 servers to deal with day-to-day standard and less demanding processing functions, the same company may invest in a non-x86 server to host a large package (e.g. accounting or computer-aided design). Accordingly, although a customer may be able to use two types of server concurrently, the x86 and the non-x86 servers will operate separately, performing different tasks, and will not interact. The competitor indicates that x86 servers generally run industry-standard Microsoft Windows Server applications, whereas non-x86 servers are usually limited to running proprietary operating systems or Unix. Most software was originally written for the x86 architecture and is not appropriate for a non-x86 server.

(65) As regards demand-side substitution, the competitor claims that given the differences in performance, x86 and non-x86 servers appeal to distinct markets in terms of customer profile and the envisaged practical applications, as demonstrated by the purchasing trends of customers. Typically, a company will purchase multiple x86 servers, but perhaps only a single non-x86 server, reserved for high-end processing tasks rather than daily functionality. Indeed, some businesses will not operate any non-x86 servers and will exclusively use x86 technology. Owing to variations in computer architecture, there are substantial technical barriers to demand-side substitution, with a significant financial impact, as switching from an x86 to a non-x86 server would probably result in a lack of streamlined integration, higher support costs and disruption during installation of the new technology. Aside from the actual cost of the non-x86 server itself, further costs would be associated with the switch in the form of support and training costs, given the additional specialist knowledge required. Further practical difficulties may be encountered by way of restrictions in transferring content, obtaining licenses, and hardware or software incompatibility. The decision to purchase an x86 or non-x86 server is ultimately based upon the end use of the server and the level of performance required, along with compatibility considerations. The price of the server is a secondary concern.

(66) As regards supply-side substitution, the competitor argues that manufacturers regard x86 and non-x86 servers as distinct product markets. Almost every new entrant in the server sector in recent years has entered the x86 market only. If the manufacturing, marketing and distribution processes of x86 and non-x86 servers were similar, a new entrant would enter the server sector by supplying both products. The x86 market is more attractive for manufacturers in view of its profitability. The market is wider and less concentrated - there are numerous x86 manufacturers. The limited customer base for non-x86 servers makes it more difficult for manufacturers to achieve the sales volumes that would make entry cost-effective and profitable in the short to medium term. Incentives are limited to supply-side switching from x86 to non-x86, even without existing technical barriers to new production.

4.3.1.5. Comments regarding the relevant geographic market for servers

(67) The competitor considers that the relevant geographic market for servers is worldwide or covers at least the EEA and the USA, given the correlation between prices and the level of trade between the USA and the EEA.

(68) Furthermore the competitor indicates that there is no significant impediment to the worldwide shipment of servers. In this respect, the competitor points out that the level of trade in servers between the USA, Asia and the EEA suggests that trade in servers is profitable. The competitor indicates that most manufacturers operate at worldwide level and supply both the USA and the EEA because servers are fully compatible (there are no technical barriers, interoperability is complete) and are sold at similar price ranges.

4.3.2. Comments regarding the in-depth assessment criteria

(69) As regards the criteria for the in-depth assessment, the competitor considers that the aid will create a distortion of competition in the common market and will have a negative effect on trade. According to the competitor, the aid benefits a company that enjoys market power. In particular, the competitor clams that the aid will increase capacity in a market in decline and will strengthen Dell's position in a market in which it has a share of supply in excess of 25 %. The competitor claims that the aid serves only to maintain production in a declining market; it can be assumed to promote irrational and uneconomic decisions as to the location of production and the allocation of both public and private resources, and to seriously distort incentives and competition in the market. The competitor considers that the aid lacks any incentive effect because, according to this competitor, Dell would have invested in Poland anyway. According to the competitor, this indicates that the use of ad hoc aid is entirely inappropriate and unnecessary. Moreover, the competitor considers that the aid is disproportionate as similar aid is not available to competitors. Finally, the competitor considers that the positive effects of the aid are limited and highly questionable and do not offset the negative effects of the aid.

(70) The comments from the competitor as regards the criteria for the in-depth assessment are presented in more detail in the relevant headings of subsection 6,7 of this Decision.

5. COMMENTS FROM POLAND

5.1. Comments on the opening decision

5.1.1. General comments

(71) The Polish authorities provided information regarding the background to the decision taken by Dell in 2006 to build a new plant in the light of the need to increase its production capacity to meet forecast demand. They also provided information regarding the timing of Dell's decision to invest in Lódz Province, the conclusion of the agreement with the Polish Government and the notification of the proposed aid. The Polish authorities explained the context of Dell's review of capacity in 2008 and its decision to reduce/cease core manufacturing activities at its plant in Limerick (Ireland).

(72) The Polish authorities provided information to correct certain points of the opening decision.

(73) The Polish authorities notified the Commission of amendments to the Agreement of 15 November 2007, made by way of an annex to the Agreement. These amendments relate to the increase in the amount of the grant for initial investment corresponding to the revised aid amount for the purchase of land and the corresponding modification to the amount of the grant instalment for 2007, the requirements regarding the deposit of an amount corresponding to the grant instalments on an escrow account pending a positive decision of the Commission on the aid, and the qualification of the relevant parts of the aid as a supplement to existing aid schemes.

5.1.2. Comment regarding market definition issues

5.1.2.1. Comments regarding the relevant product market for desktops/notebooks

(74) The Polish authorities consider that the market for PCs as a whole should be considered as the relevant market, without segmentation into sub-markets (desktops or notebooks).

(75) According to submissions made by the Polish authorities prior to the adoption of the opening decision, the comparable characteristics and performance of both products and the declining price differential between the two partly explain the proportionally higher growth in notebook sales. The Polish authorities indicate that Dell Poland is in principle capable of manufacturing desktops and notebooks on the same production line and can rapidly shift production on one line from one product to the other. In view of the significant presence of manufacturers in both segments, the Polish authorities consider that there are no significant barriers to entry for the production of either product (supply-side substitutability).

(76) In their submission of 18 February 2009, the Polish authorities indicate that Dell Poland is capable of manufacturing desktops and notebooks on the same production line, with only minor adjustments, and only incurring marginal costs and productivity losses. The Polish authorities note that, according to point 68 of the RAG, supply-side substitutability is sufficient for two products to form part of the same market.

5.1.2.2. Comments regarding the production capacity created by the project

(77) The Polish authorities indicate that, taking into account the trade-off between the anticipated reduction in production capacity in Ireland and the increase in production capacity in Poland, the overall increase in capacity for desktops would be less than 5 % of the market for desktops at EEA level in 2006.

(78) In this context, the Polish authorities consider that it is appropriate to take into account the overall effect the capacity increase resulting from the project has on the capacity of the Dell group in the EEA as a whole, rather than focusing solely on the capacity increase recorded by Dell in Poland.

(79) The Polish authorities claim that the reduction of capacity in Ireland should be taken into account. According to the Polish authorities, an economic assessment of the effects of the capacity created by the project only makes sense if reductions of capacity taking place within the same geographic market are also taken into account (all the more so since manufacturing operations in Limerick are being closed, which was not planned when the Lódz project was started). The Polish authorities note that these effects can only be determined by looking at the Dell group as a whole (partly because the market share assessment is done at that level). If during the assessment, due to a capacity reduction elsewhere in the EEA, the project no longer exceeds the 5 % limit (and therefore the distortive effects disappear), this must be taken into account. The Polish authorities refer to a previous State aid decision where the Commission was required to assess the production capacity of a company that had relocated a production facility from one Member State to another [27]. According to the Polish authorities, for this case, the Commission had accepted that the project did not create additional capacity and had reasoned that the aid recipient's group should be examined with a view to assessing possible increases of capacity [28].

(80) The Polish authorities also argue against a capacity increase of more than 5 % in the desktop market. According to the Polish authorities, such an increase would only occur if all the capacity achieved by the project were used for producing desktops, leaving no capacity for notebooks. Theoretical capacity cannot be the only measurement; if a realistic assessment of the forecast use of the production facilities is undertaken, the actual combined capacity for desktops in the Limerick plant and the Lódz plant is around half the maximum theoretical capacity for desktops. The Polish authorities note that, in any event, a reduction in the technical capacity of the Lódz plant would not affect the results of the capacity increase test as regards desktops, but would affect production capacity for notebooks (a segment which is growing faster than EEA GDP) and would deprive the investment project of its purpose.

5.1.2.3. Comments regarding the use of volume figures for calculating growth rates

(81) The Polish authorities consider that for high-tech industries, the calculation of the growth rates of the products concerned has to be based on volume figures rather than value figures. They consider that, for these sectors, price reductions often occur as a result of strong competition and rapid technological change and are therefore an indication of the strength of the market. By way of justification, the Polish authorities refer to the cases of State aid to Qimonda and Infineon in which the Commission took volume figures into account [29].

5.1.2.4. Comments regarding the relevant product market delineation for servers

(82) The Polish authorities consider that the most appropriate market segmentation for servers is based on product class and price range. On the supply side, it is not easy for manufacturers of low-end servers to enter the technically demanding high-end market. Most manufacturers concentrate on basic computing servers since there is low demand for advanced servers, resulting in a lack of economies of scale in the remaining two (higher) segments of the market. From the customer's perspective, volume servers respond to different needs to high-end servers and are used in different applications. Customers requiring basic computing will not purchase more costly higher performance servers.

(83) The Polish authorities note that switching from x86 to non-x86 servers implies higher support costs or disruption costs during installation. A switch from basic to more advanced technology will require appropriate adjustments and cause extra costs. This is true not only for a switch from x86 to non-x86 servers, but also from basic servers to more sophisticated servers in general.

(84) As regards supply-side substitution, the Polish authorities consider that a limited volume of the target market in non-x86 servers cannot constitute a basis for defining separate x86 and non-x86 markets. There is only a limited customer base for non-standard functionality and high-performance servers. High-end servers require significant supply-side investments, do not guarantee economies of scale and consequently are much more expensive. By contrast, there is a broad customer base that requires industry standard functionality, regardless of whether it is built on x86 or non-x86 servers.

(85) According to the Polish authorities, marketing and distribution channels are designed according to server performance and computing power rather than operating systems or architecture. Although Dell has the technical capabilities to build high-end x86 servers, to date it has not entered the high-end market in view of this market's specific characteristics and Dell's lack of customer base and distribution channels in this market.

5.1.2.5. Comments regarding the relevant geographic market for servers

(86) The Polish authorities consider that the geographic market for servers should be defined as the EMEA region or even the EEA alone. In their view, compatibility between servers is not the only aspect that needs to be considered. Cost and time of delivery are crucial elements which have a very significant impact on margins. Since prices are decreasing as a result of competition and technology improvements while transport costs are significant, short sales channels are a key issue.

5.1.3. Comments regarding issues for the in-depth assessment

(87) For the purpose of the in-depth assessment, the Polish authorities have provided a full economic analysis of the aid, based on a study produced by consultants LECG Consulting on behalf of Dell Poland which examines the impact of the aid with reference to the requirements laid down by the Commission in the opening decision.

(88) The comments from Poland in respect of the criteria for the in-depth assessment are presented in greater detail under the relevant headings of subsection 6,7 of this Decision.

5.2. Observations on the comments from interested parties

5.2.1. Observations on the comments from Ms Sinnott

(89) The Polish authorities note that Ms Sinnott expresses concerns as regards the legality of the aid granted by Poland "for Dell's move from Ireland to Lódz."

(90) The Polish authorities refer to their submission of 18 February 2009 in which they indicate that the aid granted to Dell is in no way designed to lead to a transfer of activities from Ireland to Lódz. The Polish authorities state that Dell's original plan was to operate both plants in parallel and that both plants have indeed functioned in parallel for some time. The Polish authorities indicate that the decision to build a plant in Lódz was taken in the summer of 2006 (the MoU with the Polish Government was signed on 19 September 2006). The Polish authorities point out that the decision to cease manufacturing activities in Limerick was adopted following an internal review in late 2008 and was influenced, to a large extent, by a deterioration in market conditions as a result of the economic crisis. The Polish authorities indicate that the decision to implement the investment project in Lódz and the decision to cease manufacturing activities in Limerick are not directly related. The Polish authorities point out that when the Polish Government decided on the aid to be granted to Dell Poland, there was no question of closing manufacturing operations in Limerick.

(91) The Polish authorities refer to their submission of 18 February 2009 in which they indicate that Dell intends to maintain significant activities in Limerick. The Polish authorities note that, with the exception of Ms Sinnott's comments, there has been no submission from the Irish authorities or from interested parties from Ireland.

5.2.2. Observations on the comments from HP

(92) The Polish authorities agree with the comments of HP regarding the definition of the product market for PCs and the geographic markets for PCs and servers.

(93) The Polish authorities point out that the comments from HP support the definition of the market for PCs as a single market.

(94) The Polish authorities consider that the comments from HP indicate that relevant markets should be defined at EEA or EMEA level. The Polish authorities consider that the relevant geographic market for servers does not extend beyond the EMEA area because of time constraints and transport costs which constitute major obstacles for worldwide trade in PCs and servers.

(95) The Polish authorities express reservations about HP's comments regarding the definition of the product market for servers. They recognise that while this market is gradually becoming unified as a result of the evolution described by HP, according to the Polish authorities, significant differences between high-end and low-end servers still exist both on the demand side and the supply side. The Polish authorities indicate that low-end servers are mainly used for basic processing, whereas more advanced servers are used for more demanding operations. On the supply side, it is not easy for manufacturers of low-end servers to enter the market for high-end servers. The Polish authorities consider that even though, as HP notes, there are signs of new entry in the server market which do not appear to be limited to one or another of the price bands, most manufacturers concentrate on the low-end segment, reflecting low demand for more advanced servers. The Polish authorities indicate that, from the customer perspective, low-end servers respond to different needs than high-end servers and are used in different applications (e.g. customers that require only basic functionality will not purchase high-performance servers, which are more costly).

5.2.3. Observations on the comments from the non-identified party (the competitor)

5.2.3.1. Claim of confidentiality by the non-identified party

(96) The Polish authorities note that the non-identified party claims confidential treatment not only for its identity, but also for several paragraphs of its submission. The Polish authorities are unable to understand several parts of the submission; this applies in particular to the section on alleged distortions to the market. The Polish authorities point out that the Commission cannot base its decision on any facts of which the Member State has not been informed. They indicate that this includes facts for which interested parties claim confidential treatment [30].

(97) As regards the claim of confidentiality by the non-identified party, the Commission points out that its assessment for the purpose of this Decision is based exclusively on the information provided by the non-identified party in the non-confidential version of its comments, as forwarded to Poland. The Commission has not taken into account the information provided by the non-identified party in the confidential version of its comments [31].

5.2.3.2. Right of the non-identified party to submit comments

(98) The Polish authorities question the right of the non-identified party to submit comments. The Polish authorities note that under Articles 1(h), 6(1), and 20(1) of Regulation (EC) No 659-1999 only interested parties, i.e. parties that might be affected by the granting of the aid in question, are entitled to submit comments. The Polish authorities consider that the comments as transmitted to them do not identify to what extent the non-identified party is affected by the granting of aid to Dell Poland. The Polish authorities urge the Commission to disregard the comments of the non-identified party, which has not demonstrated a sufficient interest pursuant to Article 1(h) of Regulation (EC) No 659-1999.

(99) As regards the right of the non-identified party to submit comments, the Commission points out that the non-identified party referred to in footnote 4 is a competitor of Dell. The Commission considers that, for the purpose of this Decision, this fact is sufficient to demonstrate the legal standing of the non-identified party as an interested party within the meaning of Article 1(h) of Regulation (EC) No 659-1999.

5.2.3.3. Observations on the comments from the competitor regarding market definition issues

(100) The Polish authorities disagree with the comments from the competitor regarding the use of volume figures to calculate growth rates. They refer to their submission of 18 February 2009, which explains why volume figures rather than value figures should be examined to determine whether a market is declining. They point out that the RAG clearly states that the growth rate on the relevant market before the investment should be looked at for the purpose of assessing the measure's compatibility with point 68(b) of the RAG, and that there is no basis on which to derogate from this clearly worded rule and to apply growth forecasts.

(101) In reply to the competitor's comments regarding the market delineation for servers, the Polish authorities argue that they refer to a merger decision taken in 1997 and 1998 to claim that the only appropriate segmentation is x86/non-x86. However, they claim the market has substantially evolved since then. In their view, it is incorrect to assert that the classification into x86/non-x86 is the standard offered by the market researcher IDC, since in its server tracker database IDC offers the possibility of subdividing the market by more than 20 criteria. Servers with higher computing power can also be built based on multiple x86 processors. The required power may be achieved with servers of different architecture and their configurations. Therefore non-x86 servers can fulfil the same functions as x86 servers. It cannot be concluded that high-end processing by x86 servers is not possible. The high-end server segment includes a mix of non-x86 and x86 based servers that run various applications, delivering the same type of end solution. Both x86 and non-x86 servers can run on different operating systems and for specific applications. Customers can choose the type of server to run their applications. Numerous segmentations may be analysed for servers (according to architecture, operating system, database engine compatibility), but computing power and server performance are deciding factors for customers' choices which are reflected in prices; these will be higher if requirements are more complex and specialised.

5.2.3.4. Observations on the comments from the competitor regarding the in-depth assessment

(102) The observations of the Polish authorities on the comments from the competitor in respect of the criteria for the in-depth assessment are presented in greater detail under the relevant headings of subsection 6,7 of this Decision.

6. ASSESSMENT OF THE AID MEASURE

6.1. Classification as State aid

(103) According to Article 87(1) of the Treaty, save as otherwise provided in the Treaty, any aid granted by a Member State or through state resources in any form whatsoever, which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods is incompatible with the common market, insofar as it affects trade between Member States.

(104) The aid is granted by the Polish authorities in the form of direct grants, price reductions and tax exemptions. It can therefore be considered as having been awarded by the Member State and by way of state resources within the meaning of Article 87(1) of the Treaty. The aid is granted to a single company, Dell Poland, and is therefore selective. The aid is granted for an investment relating to the manufacture of personal computers and servers. Since these products are traded between Member States, the measure is therefore likely to affect trade between Member States. The aid granted to Dell Poland will relieve the company of costs which it normally would have had to bear itself and therefore the company benefits from an economic advantage over its competitors. By favouring Dell Poland and its production in this way, the measure therefore distorts or threatens to distort competition.

(105) Consequently, the Commission considers that the notified measure constitutes State aid within the meaning of Article 87(1) of the Treaty.

(106) Having established that the notified measure constitutes State aid within the meaning of Article 87(1) of the Treaty, it is necessary to consider whether the measure can be found to be compatible with the common market.

6.2. Legality of the aid measure

(107) By giving notification of the aid awarded to Dell Poland before putting it into effect, Poland has complied with the individual notification requirement laid down in Article 88(3) of the Treaty (cf. recital 2).

6.3. Legal basis for the assessment

(108) The Commission notes that as the objective of the aid is to promote regional development, the basis for assessing the compatibility of the aid with the common market is the RAG.

(109) The Commission notes that the investment project can be considered as an initial investment within the meaning of point 34 of the RAG, as it involves an investment in material and intangible assets relating to the setting-up of a new establishment.

(110) The Commission notes furthermore that the initial investment project can be considered as a large investment project within the meaning of point 60 of the RAG, as its eligible expenditure is in excess of EUR 50 million, calculated on the basis of prices and exchange rates on the date of notification.

(111) The Commission has therefore assessed the aid measure in accordance with the provisions of the RAG on regional investment aid and, in particular, the provisions of the RAG on aid for large investment projects as well as in accordance with the criteria for the in-depth assessment of regional aid to large investment projects laid down in the Communication of the Commission of 16 September 2009 [32] (the Guidance Communication).

(112) In accordance with the relevant case-law [33], the Commission has to conduct the assessment on the basis of the rules applicable at the time of the decision.

6.4. Compatibility with the general provisions of the RAG

(113) The Commission has verified whether the aid was granted in compliance with the general provisions of the RAG. This assessment led to the following observations:

(114) Pursuant to point 9 of the RAG, Dell Poland is not a firm in difficulty within the meaning of the Community guidelines on State aid for rescuing and restructuring firms in difficulty [34].

(115) Although as a general rule, pursuant to point 10 of the RAG, the Commission does not favour the award of ad hoc regional aid, in view of the information presented in recitals 10 and 13, it can be considered that the project contributes towards a coherent regional development strategy. In particular, the investment project is expected to contribute to the development of economic activity in the region through positive effects in terms of job creation (direct and indirect jobs), training and knowledge transfer and a multiplier effect for additional investment by related services and manufacturers.

(116) The project constitutes an initial investment within the meaning of point 34 of the RAG as it concerns the setting-up of a new establishment (cf. recital 12).

(117) In accordance with point 38 of the RAG, before the start of works on the project, the beneficiary submitted an application for aid and the Polish authorities confirmed in writing that, subject to the Commission's approval, Dell Poland's investment project would be eligible for aid (cf. recital 15).

(118) In accordance with point 39 of the RAG, the beneficiary's own contribution to eligible expenditure is above 25 % (cf. recital 16).

(119) In accordance with point 50 of the RAG, the eligible expenditure of the project was calculated on the basis of the eligible investment costs (cf. recital 14). In accordance with point 58 of the RAG, aid for job creation was calculated on the basis of 2-year employment costs (gross wage costs and other compulsory payments) for newly-hired persons.

(120) In accordance with point 64 of the RAG, by giving notification of the aid before putting it into effect, Poland has complied with the individual notification requirement (cf. recital 2).

(121) In accordance with points 71-75 of the RAG, the rules on cumulation of aid have been complied with (cf. recitals 2 and 15).

(122) The Commission therefore considers that the aid complies with the general provisions of the RAG.

6.5. Compatibility with the provisions of the RAG on large investment projects

6.5.1. Single investment project

(123) The investments conducted by Dell Poland have been followed by investments by two of its suppliers, RR Donnelley Global Turnkey Solutions Poland and Flextronics.

(124) In the opening decision (points 42-51), the Commission considered that sufficient technical, functional and strategic links did not exist between the investment conducted by Dell Poland and the investments conducted by its two suppliers for them to be considered to form part of a single investment project within the meaning of point 60 and footnote 55 of the RAG.

(125) The Commission notes that it has not received any comments as regards the single investment character of the project.

(126) Therefore the Commission confirms that the investments conducted by Dell Poland on the one hand and by the above-mentioned suppliers on the other cannot be considered to form part of a single investment project within the meaning of point 60 and footnote 55 of the RAG.

6.6. Compatibility with the provisions of point 68 of the RAG - verification of the doubts expressed in the opening decision

(127) On the basis of the information provided by the Polish authorities and by the interested parties as a result of the formal investigation procedure, it is necessary to examine whether the doubts expressed in the opening decision have been allayed.

6.6.1. Compatibility with the production capacity test for desktops/notebooks

6.6.1.1. Relevant product markets for desktops and notebooks

(128) In the opening decision (points 62 and 105), the Commission expressed doubts as to whether, for the purpose of the Decision, desktops and notebooks could be considered as forming a single product market or two separate relevant product markets

(129) Having regard to the comments of the Polish authorities and of the interested parties, the Commission notes that there are indications that a certain degree of demand-side substitution does occur in both directions [35] for the two products concerned and that certain manufacturers, notably Dell, seem capable of supply-side substitutability [36]. However, there are also indications that there exists a significant level of non-transient demand for the other product or exclusive demand for either of the products and that, for some manufacturers, significant technical, organisational and cost-related barriers may exist for switching production from one product to the other.

(130) Although the information received during the formal investigation has enabled a better understanding of the PC sector, the Commission takes the view that it is insufficient for it to conduct a meaningful market investigation which would make it possible to define the relevant product market(s) for desktops and notebooks [37]. The Commission notes therefore that the doubts regarding the relevant product market for desktops and notebooks have not been allayed.

(131) In view of the above, the Commission finds that, while it may be possible to leave open the exact definition of the relevant product market for desktops and notebooks, for the purpose of this Decision it is necessary to consider desktops and notebooks separately in order to address the competitive constraints that may exist in the event of desktops being considered as a separate product market [38].

6.6.1.2. Production capacity created by the project

(132) In the opening decision (points 99 and 105), the Commission expressed doubts as to whether it was appropriate to consider the effect of the capacity created by the project on Dell's capacity in the EEA as a whole, as opposed to the capacity created by the project alone.

(133) Having regard to the comments of the Polish authorities presented in recitals 77 and 78, the Commission points out that point 68(b) of the RAG refers to "the production capacity created by the project", which excludes the possibility of taking into account, together with the capacity created by the project, the level of production capacity of other establishments or companies located within the EEA which belong to the recipient. Indeed, if, for the purpose of establishing whether the threshold of point 68(b) of the RAG is met, increases of capacity in one Member State could be offset by reductions in another Member State - as a result of which there would, for instance, be no net increase in capacity at company level - this would mean that at least one important category of potentially harmful State aid cases might escape in-depth scrutiny, namely those aid measures having the effect of increasing production capacity in one Member State at the expense of another Member State, where capacity is reduced because of the aid. As indicated in paragraphs 52-53 of the Guidance Communication, these measures may run counter to the very rationale of regional aid and therefore need to be controlled.

(134) Having regard to the comments of the Polish authorities presented in recital 79, the Commission considers that it is necessary to take into account the effects of the increase of production capacity made possible by the project, as notified, and that reductions in capacity at other production facilities within the EEA cannot be taken into account if such reductions in production capacity do not form part of the investment project for which the aid was notified and are thus not affected by the granting of the aid. The Commission points out that in the decision on State aid case N 158/05 Getrag Ford Transmissions Slovakia s.r.o., cited by the Polish authorities, the anticipated reduction of capacity in one Member State was part of the notified project.

(135) Having regard to the comments of the Polish authorities presented in recital 80, the Commission notes that given the flexibility of production at the plant, the whole of the capacity created by the project could in principle be allocated to the production of any of the three products concerned [39].

(136) In the light of the above, the Commission finds that the reduction in production capacity of a beneficiary in another Member State cannot be taken into account. The Commission concludes that the production capacity created by the project exceeds 5 % of the market in all the product markets concerned.

6.6.1.3. Use of volume figures for calculating growth rates

(137) In the opening decision (points 100 and 105), the Commission expressed doubts as to whether it could calculate the average annual growth rate of the products concerned by the investment project, as per point 68(b) of the RAG, solely on the basis of volume figures.

(138) The Commission notes that the negative growth of desktops in value terms could reflect the strong decrease in the average selling prices of desktops. This negative growth rate in value terms could indicate that the increase in volume is insufficient to compensate for decreasing marginal revenue, which is characteristic of a market in decline. The Commission notes that the same decrease in average selling prices of desktops also applies to notebooks and servers. However, contrary to desktops, the positive growth rate in both volume and value terms for notebooks and servers indicates that the decrease in their average selling prices is more than offset by the increase in the quantities sold.

(139) In view of the above, the Commission finds that, for the purpose of assessing whether desktops constitute a market in decline under the plausible market definition where desktops are considered as a separate product market, it cannot calculate the average annual growth rate of desktops as referred to in point 68(b) of the RAG solely on the basis of volume figures.

6.6.1.4. Conclusion regarding the production capacity increase test for desktops/notebooks

(140) The Commission concludes that in the event of desktops being considered to be a separate product market from notebooks, the aid could raise competition concerns in that it would constitute, for the desktop market, aid to a company in a sector in decline. As regards desktops, it is therefore necessary to conduct an in-depth assessment of the effects of the aid.

6.6.2. Compatibility with the market share test for servers

6.6.2.1. Relevant product market delineation for servers

(141) In the opening decision (points 65-67, 72 and 105), the Commission stated that it was appropriate to follow a market definition based on product class/price range in order to delineate the server market, but expressed doubts as to whether segmentation by product class/price range constituted the appropriate market delineation for servers.

(142) Having regard to the comments received, the Commission considers that its initial view that segmentation of servers by product class/price range constitutes the most appropriate market delineation in this instance could be contradicted by arguments in favour of other plausible market delineations.

(143) In the light of the above, the Commission finds that the doubts as regards the appropriate market delineation which should be applied for the purpose of measuring Dell's share in the server market have not been allayed.

6.6.2.2. Relevant geographic market for servers

(144) In the opening decision (points 78-81 and 105), the Commission expressed doubts as to whether the relevant geographic market for servers should be considered as being EEA-wide, at the level of the EMEA region or worldwide.

(145) As regards the comments of the Polish authorities, the Commission considers that the arguments put forward are specific to Dell's business model and to its operating mode and do not necessarily apply to other server manufacturers or reflect the general functioning of the server market.

(146) The Commission notes that the comments from interested parties indicate that trade in servers occurs at a level broader than the EEA or the EMEA region and that most manufacturers operate at worldwide level.

(147) However, the Commission points out that the geographic market for servers is relevant only when Dell's market share for x86 servers is considered.

(148) In the light of the above, the Commission considers that, while there may be grounds for considering that the geographic market for servers is worldwide, given that doubts remain as regards the appropriate product market delineation for servers, the geographic market for servers can be left open.

6.6.2.3. Conclusion regarding the market share test for servers

(149) The Commission concludes that, depending on the market definition used for servers, the aid could raise competition concerns. In particular, in the event of the x86/non-x86 delineation being used, the measure would constitute, for the x86 segment of the product market, aid to a company with a share exceeding 25 % of the relevant market at worldwide level. As regards the server market, it is therefore necessary to conduct an in-depth assessment of the effects of the aid.

6.6.3. Conclusion regarding the doubts expressed in the opening decision

(150) The Commission considers that the doubts expressed in the opening decision as to whether the aid fulfils the conditions defined in points 68(a) and 68(b) of the RAG have not been allayed.

(151) In accordance with point 68 of the RAG, in the event that either of the two thresholds defined in point 68(a) and 68(b) of the RAG is exceeded for the measure under assessment [40], it is necessary to conduct a detailed verification, following the opening of the procedure provided for in Article 88(2) of the Treaty, that the aid is necessary to provide an incentive effect for the investment and that the benefits of the aid measure outweigh the resulting distortion of competition and effect on trade between Member States.

6.7. In-depth assessment

(152) The in-depth assessment is conducted on the basis of the Guidance Communication, which lays down the assessment criteria for the positive and negative effects of the aid.

6.7.1. Positive effects of the aid

6.7.1.1. Objective of the aid

Comments from Poland

(153) The Polish authorities indicate that Dell's investment in Lódz Province is a long-term commitment (the manufacturing facilities must be maintained for at least 5 years after their completion and the jobs created must be maintained for 5 years) and also constitutes a major precedent which can lay the ground for further investments in the province. It is therefore a credible commitment to Lódz Province which will help to create direct employment and stimulate regional development.

(154) This positive impact is reinforced by the clustering effect the investment project is likely to trigger, by encouraging firms in the same or related industries to locate nearby in order to benefit from external economies of scale or scope.

(155) In particular, Dell's contribution to the creation of jobs in Lódz Province [41] is likely to exceed 2500 [42] direct jobs by January 2010 [43], plus about 1300 indirect jobs from suppliers, subcontractors and services providers located in the region.

(156) In terms of regional development, the Polish authorities also point to the impact of the investment on the following:

- Education: Dell cooperates with two local institutes of higher education [44] (e.g. by granting scholarships, commissioning studies or offering internships) and intends to start working with other similar institutes in Lódz Province [45]. Dell also takes part in the Programme of Human Resources Development in Lódz initiated by the city of Lódz [46]. Similarly to what happened in Ireland [47], this cooperation with local institutes of higher education and local government will contribute to local innovation and to the development of a knowledge-based economy and information society,

- Transfer of knowledge: Dell's plant in Poland is among the most advanced in the world, therefore it is very likely that knowledge will spill over into the local economy. Knowledge spillover effects will be derived from the management and production processes that Dell will introduce in the province, which will result in diffusion of operational, financial, technical, logistic [48] and customer-handling skills, as has already happened in Limerick. Dell will also provide its workforce with a range of training opportunities [49]. The project is expected to contribute to the creation and spread of ICT skills and the establishment of a multi-faceted centre as in Limerick (using operational, financial, technical and customer-handling skills),

- Multiplier effect: Dell's presence will help to raise living standards in Lódz Province by increasing the local economy's output and monetary flows into the province. For the purpose of estimating the potential multiplier effect of Dell's investment the Polish authorities refer to the results of a study by Dell on a similar investment in the USA. The study [50] shows that on average for each job created at Dell, two jobs are created at local level and another 2,5 jobs are created in other regions of the USA. On this basis, Dell's estimated payroll of EUR 17,3 million per annum is expected to create an additional EUR 13 million for new jobs in Lódz Province, and EUR 9,5 million per annum for Poland as a whole.

Comments from the competitor

(157) The competitor expresses doubts about the creation of indirect jobs. It claims that ancillary investors will not expand their production capacities but will instead increase the use of existing facilities in Poland. The competitor also considers that the indirect investments are overestimated and that the benefits resulting from Dell's presence in the province will be limited.

Analysis and conclusion

(158) The Commission notes that the Polish authorities have provided information regarding the objective of the aid and the effects of the investment project on regional development, education, knowledge spillovers, potential multiplier effects, creation of direct and indirect employment.

(159) The Commission considers that the Polish authorities have provided adequate information to demonstrate that the project will contribute to the economic development of Lódz Province.

(160) The Commission considers that the comments from the competitor are unsubstantiated and cannot therefore be taken into account.

(161) In view of the above, the Commission finds that the investment is likely to make an overall positive contribution to economic development in Lódz Province and that as such it will help to fulfil the regional development objective.

6.7.1.2. Appropriateness of the aid instrument

(162) As stated in paragraphs 17 and 18 of the Guidance Communication, State aid in the form of initial investment aid is only one way of overcoming market failures and promoting economic development in disadvantaged regions. Aid constitutes an appropriate instrument if it provides specific advantages compared with other policy measures. According to paragraph 18 of the Guidance Communication, "measures for which the Member State considered other policy options, and for which the advantages of using a selective instrument such as State aid for a specific company are established, are considered to constitute an appropriate instrument."

Comments from Poland

(163) In their submission of 18 February 2009, the Polish authorities explained in detail that the award of State aid to large projects with regional benefits such as the investment by Dell Poland in Lódz was the only appropriate means whereby the economic shortcomings of the province could be addressed (in addition to the general measures that had already been put in place).

(164) On the basis of the findings of the LECG study, the Polish authorities point out that despite general measures to support economic development at national and regional levels through investment in basic infrastructure and productive factors, in particular through the use of the Structural Funds, major regional disparities continue to affect Poland as a whole and Lódz Province in particular. In particular, the data provided in the study indicates that per capita GDP in Lódz Province is below the national average, while unemployment is above the national average and long-term unemployment is also higher than in Poland as a whole. The average wage level in the province is also below the national average. Moreover, the gap between wage levels in Lódz Province and Poland as a whole increased during 2004-2007.

(165) The Polish authorities also point out that general measures implemented in Poland as a whole in order to attract inward investment appear to have had a limited effect on the volume of foreign direct investment (FDI). According to the data provided in the study, Poland has one of the lowest levels of FDI as a percentage of GDP in comparison with other Member States that have joined the Community since 2004.

(166) The Polish authorities consider that the relative underdevelopment of Lódz Province indicates a need for more targeted intervention in the province. The Polish authorities note in this respect that, by benefitting the labour market and promoting infrastructure development, large investments often trigger additional investments through the clustering of economic activity and its associated effects (industrial concentration, indirect growth, technological and knowledge spillovers). State aid granted to support large investment projects can therefore be an effective means of increasing regional development.

Comments from the competitor

(167) The competitor claims that because the aid is provided as ad hoc aid and not under a scheme, there is no obligation for Poland to make it available to all operators in the sector on equal terms. The competitor notes that similar aid could in principle be offered to competitors of Dell, but doubts whether in the current economic climate a grant of over EUR 50 million could be made available to another computer manufacturer.

(168) The competitor points out that, pursuant to point 10 of the RAG, ad hoc aid can be granted only in exceptional circumstances. The competitor considers that the use of ad hoc aid is unjustified and inappropriate as it will increase the market power of a company in a market where it has a significant share of supply, namely servers, and in a market which is in decline, namely desktops.

Observations from Poland on the comments from the competitor

(169) The Polish authorities point out that the competitor's comments do not suggest that the economic disadvantages of Lódz Province could be overcome by other means than State aid to a large investment project. The Polish authorities note that the only point made by the competitor is that similar aid is not available to other operators. According to the Polish authorities, this statement is based on a misunderstanding of the concept of appropriateness as set out in the RAG which requires the Commission to analyse whether the objective of the aid could have been achieved by other means. The Polish authorities consider that the competitor does not make any such suggestions.

Analysis and conclusion

(170) The Commission notes that the Polish authorities base their arguments regarding the appropriateness of the aid instrument on a macroeconomic analysis of the factors affecting economic development at national and regional level in Poland. They indicate that whereas Poland in general and Lódz Province in particular are disadvantaged areas, general measures in Poland as a whole (as opposed to ad hoc aid for large investment projects in given provinces) have had a limited effect on the volume of FDI and that significant regional disparities within Poland remain. The Polish authorities point out that the Lódz Province is one of the regions with the lowest per capita GDP, highest unemployment and lowest average monthly gross wages. The Polish authorities conclude that since market failures prevent normal market functioning, State aid to large investment projects is an effective means of increasing regional development through the concentration of industrial activity and clustering and their associated effects (attraction, spillovers, indirect growth).

(171) As regards the observations by Poland, the Commission notes that the Polish authorities have submitted comprehensive economic evidence that indicates that the granting of State aid to large investment projects with regional benefits such as the project conducted by Dell Poland is an appropriate means to address the economic shortcomings of Lódz Province (in addition to general measures that have already been put in place).

(172) As regards the comments from the competitor, the Commission points out that, with the exception of fiscal aid schemes where aid is granted automatically to qualifying expenditure without any discretion on the part of the authorities, there is no obligation for the Member State to make aid available to all operators in the sector on equal terms, including for aid granted under approved or existing aid schemes. The Commission points out furthermore that, while individual ad hoc has to be justified on the basis of the objective of the aid and its effects, this does not relate to the appropriateness of the aid instrument, which concerns whether the aid is necessary in comparison with general policy measures, regardless of the nature of such aid.

(173) In the light of the above, the Commission finds that the notified aid constitutes an appropriate instrument with which to achieve the objective of the aid.

6.7.1.3. Incentive effect

(174) In the opening decision (point (108), the Commission sought evidence as to whether, and to what extent, the aid had affected the company's economic decisions in order to assess the incentive effect of the aid, as defined in paragraphs 19-28 of the Guidance Communication.

(175) Paragraph 22 of the Guidance Communication mentions two alternative scenarios in which the incentive effect can be proven. In scenario 1 "the aid gives an incentive to adopt a positive investment decision because an investment that would otherwise not be profitable for the company at any location can take place in the assisted region." In scenario 2, "the aid gives an incentive to opt to locate a planned investment in the relevant region rather than elsewhere because it compensates for the net handicaps and costs linked to a location in the assisted region."

(176) Paragraph 23 of the Guidance Communication states that "the Member State should demonstrate to the Commission the existence of an incentive effect of the aid", "specify which scenario applies" and "provide [...] a comprehensive description of the alternative scenario."

(177) The Polish authorities have provided information regarding the incentive effect and proportionality of the aid. In particular, the Polish authorities indicated in their submission of 18 February 2009 that, in 2005-2006, Dell had determined the required characteristics of the project in order to respond to forecasted demand and had decided to carry out the project irrespective of its location. It therefore serves no purpose to consider, as provided for in scenario 1, whether or not the investment would have been carried out.

(178) In respect of paragraph 25 of the Guidance Communication, the Commission notes that the aid falls under scenario 2. It is therefore necessary to consider the effects of the aid in terms of favouring its location in the target region as opposed to another region (for which, in the alternative scenario, it is assumed that no aid is given).

Comments from Poland

(179) In their submission of 18 February 2009, the Polish authorities indicate that, in order to determine the location of the project, Dell analysed and compared the costs and benefits of several potential locations and, as a result, retained two potential locations for the investment project, namely Lódz in Poland and Nitra in Slovakia [51].

(180) In spring 2006 Dell conducted an analysis, presented in a confidential company document submitted by the Polish authorities, to identify and review the cost factors representing the key differences when assessing the comparative advantages of the two locations, which were: [...] [52]. The document also contained a quantitative comparison of the cost difference in net present value (NPV) between the two locations considered for the project [53].

(181) On the basis of this document, the consultant LECG conducted an analysis of the NPV of the cost differential between the two locations considered for the project. Considering the alternative situation, LECG's study suggests that the disadvantages of locating in Lódz Province as opposed to the Nitra region in terms of [...] and [...] are not outweighed by Lódz's advantages in terms of [...] and [...]. The NPV analysis, calculated over 2007-2018, concludes that, overall, in the absence of aid (both in Lódz and Nitra), it would have been more advantageous for Dell to locate the project in Nitra rather than in Lódz (cf. table below).

Present value of cost differential between Lódz and Nitra

NB: The discount factor used for evaluating all cash flows, including those related to the aid, is the weighted average cost of capital (WACC) of [...] % used by Dell in the appraisal of its investment projects in general. Currency conversions use the average exchange rate for September 2006 (date of the MoU between Dell and the Polish Government).

Source: LECG calculations based on data provided by Dell.

(EUR thousand) |

Category/cash flows | NPV years 1-12 |

[...] of Lódz | [...] |

[...] of Lódz | [...] |

[...] of Lódz | [...] |

[...] of Lódz | [...] |

Total NPV disadvantage of Lódz Province (excluding State aid) | -40360 |

(182) The Polish authorities point out that Dell had no additional formal documents comparing the two shortlisted investment locations. Dell had no need to prepare such documents, as its decision-making process was based on the site selection team's recommendations for the Board of Directors. The site selection team thoroughly evaluated all possible sites following a series of visits and negotiations with the authorities on the State aid package in order to put recommendations to the Board of Directors concerning the final decision.

(183) The Polish authorities point out that the MoU, constituting the letter of intent, is the last document confirming Dell's decision to invest in Poland on the agreed terms, in particular encompassing the State aid package.

(184) The Polish authorities note that the investment terms and in particular the aid were subject to lengthy negotiations between Dell and various institutions representing the Polish Government (conducted by the Polish Information and Foreign Investment Agency). Initial meetings took place in April 2005, followed by a series of visits to various locations in Poland. Preliminary offers of aid were made to Dell in late 2005. Negotiations on the wording of the MoU lasted from May 2006 to late July 2006, when the parties reached agreement on the provisional wording. Certain amendments were made during August 2006 and the document was finally signed by all the parties on 19 September 2006, confirming Dell's decision to invest in Poland.

(185) The Polish authorities have provided a copy of a Dell internal document which was presented to the Board of Directors and confirms the decision to invest in Lódz. The documents are dated 6 September 2006, by which time the wording of the MoU had already been adopted by the parties.

Comments from the competitor

(186) The competitor claims that the aid lacks any real incentive effect and believes that Dell has failed to show that the aid provided any real incentive to alter its behaviour at the time that Poland originally agreed to support its investment. The competitor claims that given Poland's central geographic location and the availability of a stable workforce, it has been very successful in attracting foreign direct investment, and that therefore Dell was pursuing a rational economic course that it would have followed without any State aid. The competitor mentions that since the last quarter of 2007 market conditions have changed considerably following the emergence of netbooks, thereby undermining the rationale of Dell's investment in the production of desktops, notebooks and servers.

Observations from Poland on the comments from the competitor

(187) In their reply dated 22 April 2009, the Polish authorities point out that they have submitted detailed economic evidence as to why the aid comprised an incentive effect when Dell decided to locate the plant in Lódz in 2006. They claim that the competitor contradicts itself in several ways: it started by alleging that Dell would have implemented the investment in Poland even without aid but goes on to say that now market conditions have changed Dell would probably no longer decide to build a plant. The Polish authorities also point out that the submission by the competitor fails to mention why present market conditions are relevant for analysing an incentive effect relating to an investment decision that was taken in 2006.

(188) The Polish authorities point out that none of the competitor's comments are relevant when assessing whether the aid comprises an incentive effect. For the Polish authorities, the only relevant issue is whether the investment would have been located elsewhere had the aid not been granted.

Analysis and conclusion

(189) As regards the comments from the competitor, the Commission notes that these are not relevant to an assessment of the incentive effect in scenario 2, according to which the effects of the aid have to be considered in terms of favouring its location in the target region as opposed to another region. Also, the remarks on subsequent changes in market conditions are not relevant to the evaluation of a previous investment decision.

(190) In respect of paragraph 23 of the Guidance Communication, the Commission notes that the Polish authorities have provided clear evidence, as demonstrated in recitals 183-185, that the aid had a real impact on the choice of where to locate the investment, as the decision by Dell to locate in Lódz was taken only after the company had reached agreement with the Polish authorities on aid conditions affecting economic and financial aspects of project implementation in that location.

(191) In respect of paragraph 25 of the Guidance Communication, the Commission notes that the Polish authorities have provided proof of the incentive effect of the aid by providing company documents and a study which shows that a comparison was made between the costs and benefits of locating in the assisted region concerned and those of locating in an alternative region (Nitra, Slovakia). The Commission considers that the comparative scenario presented by the Polish authorities is realistic.

(192) The Polish authorities indicate that the decision to locate the project in Lódz was taken following the decision of the Polish Government to grant aid to Dell to compensate for the economic disadvantages associated with locating in Lódz and the related disadvantages and costs compared with the alternative location considered at that time. In respect of paragraphs 21 and 28 of the Guidance Communication, the Commission considers therefore that the aid helped to change the behaviour of the beneficiary company and that the investment would not have taken place in Lódz Province without the aid.

(193) In view of the above, the Commission finds that the aid provided an incentive for Dell to decide to locate its new manufacturing plant in Lódz by compensating for less favourable investment conditions in comparison with another envisaged location and that, in the absence of the aid, the investment may have taken place somewhere else.

6.7.1.4. Proportionality of the aid

(194) In the opening decision the Commission sought evidence that, given the aid, the return on investment of the project concerned was in line with Dell's normal rate of return on other investment projects, with the cost of capital of the company as a whole or with returns commonly observed in the industry concerned.

(195) In scenario 2 of the Guidance Communication, the aid is deemed proportionate if it is equal to the difference between the net costs to the beneficiary of investing in the assisted region and the net costs of investing in the alternative location.

(196) The Commission notes that the data provided for the analysis of the incentive effect (cf. table below) indicate that the NPV of the aid does not exceed the NPV of the total disadvantages associated with locating in Lódz (total cost differential excluding aid).

Comparison of cost differential between Lódz and Nitra and aid amount

NB: The discount factor used for evaluating all cash flows, including those related to the aid, is the weighted average cost of capital (WACC) of [...] % used by Dell in the appraisal of its investment projects in general. Currency conversions use the average exchange rate for September 2006 (date of the MoU between Dell and the Polish Government).

Source: LECG calculations based on data provided by Dell.

(EUR thousand) |

Total NPV disadvantage of Lódz Province (excluding State aid) | -40360 |

NPV of the aid granted by the Polish authorities | 39432 |

Total NPV advantage of Lódz Province (including State aid) | -927 |

(197) In the light of the above, the Commission finds that the aid is limited to the amount necessary to compensate for the net additional costs of locating in Lódz compared with the alternative location.

6.7.2. Negative effects of the aid

(198) The Commission notes that point 40 of the Guidance Communication indicates that "if, however, the counterfactual analysis suggests that without the aid the investment would have gone ahead in any case, albeit possibly in another location (scenario 2), and if the aid is proportional, possible indications of distortions such as a high market share and an increase in capacity in an underperforming market would in principle be the same regardless of the aid."

6.7.2.1. Crowding out of private investment

(199) Poland and the non-identified interested party have submitted comments with respect to the two sub-criteria set out in the Guidance Communication for the crowding-out of private investment, namely the creation of market power and the creation or maintenance of inefficient market structures.

Comments from Poland

(200) As regards the crowding-out of private investment, the Polish authorities consider it unlikely that competitors will react to the investment project by reducing their investments. They also take the view that the investment project is not conducive to the creation or maintenance of market power or to the maintenance of inefficient market structures.

Comments from the competitor

(201) The non-identified interested party argues that Dell has significant market power which the aid will strengthen. According to the competitor, Dell will be able to use the aid to leverage significant secondary advantages and will gain increased capacity at minimal cost. The competitor also states that aid should not be used to increase capacity in a declining market (desktops).

Analysis and conclusion

(202) As regards market power, the Commission considers that since the aid falls under scenario 2 and given that, as indicated above, the investment project would have been carried out in any event, it is not relevant to consider, as part of the in-depth assessment, the effects that the aid could have in relation to a possible increase in market power, as any such market power would exist irrespective of the location of the investment and would not be affected by the granting of the aid. The Commission therefore takes the view that in this case it cannot be concluded that the market power of the beneficiary has increased as a result of the aid.

(203) Similarly, as regards the creation or maintenance of inefficient market structures, notwithstanding possible indications, noted above, that the investment project could give rise to a significant increase in production capacity in a market which might be considered as being in decline (desktops), the Commission considers that it cannot conclude that the aid is likely to crowd out private investment, since the theory of harm in scenario 2 is not whether competitors are likely to be affected by the investment but rather whether the aid has any effect on the choice of location.

(204) In the light of the above, the Commission finds that the aid does not raise any concerns as regards the crowding-out of private investment.

6.7.2.2. Negative effects on trade

(205) In addition to the potential negative effects resulting from crowding-out of private investment, the Guidance Communication also mentions the potentially negative effects on trade and location which might be associated with aid to large investment projects.

(206) In order to determine the impact of the investment on trade it is first of all necessary to assess whether the aid leads to the displacement of existing activity from one region to another, as opposed to the development of additional economic activity. In the event of displacement, the negative effects for the region where the economic activities are lost (in existing locations or in the alternative location where the investment would have been located in the absence of aid) have to be taken into account in the overall balancing test.

(207) According to paragraph 50 of the Guidance Communication, "when appraising large investment projects subject to this guidance, the Commission should have all necessary information to consider whether State aid would result in a substantial loss of jobs in existing locations within the Community."

Comments from Poland

(208) According to the Polish authorities, Dell's decisions to locate its new manufacturing plant in Lódz and to reduce production at its existing manufacturing plant in Limerick are not interrelated. The Polish authorities indicate that the decision, initially, to reduce manufacturing activities at Limerick and, subsequently, to cease manufacturing activities by closing the Limerick plant, was taken in late 2008, more than 2 years after the decision to proceed with the investment project in Lódz, and reflects a rapid worsening in the global economic outlook.

(209) The Polish authorities indicate that the closure of the manufacturing plant in Limerick, which was announced in January 2009, is expected to lead to the loss of 1900 jobs between April 2009 and January 2010. The Polish authorities indicate, however, that Dell will maintain selected business activities in Limerick. In particular, more than 900 employees will continue to be employed at the plant in Limerick, which will continue to coordinate manufacturing operations and will perform logistics and supply-chain activities across a range of functions including product development, engineering and procurement. The Polish authorities indicate furthermore that Dell will maintain and develop significant activities at other sites in Ireland.

Comments from the competitor

(210) The competitor has not provided any comments on negative effects on trade (in the manner in which this issue is addressed in the Guidance Communication).

Analysis and conclusion

(211) The Commission notes that, on the basis of the information received during the formal investigation, in particular the fact that Dell would have undertaken the new investment even in the absence of the aid (scenario 2), there is no indication of job losses in the Community other than in locations which belong to the Dell group, as mentioned above [54].

(212) The Commission concludes that the recent job losses announced at Dell's Limerick site were not caused by the aid. As set out in recital 193, the investment in new capacity would have been undertaken even in the absence of aid, but at Nitra (Slovakia) as opposed to Lódz (Poland). Given that the aid has changed the location of the investment (from Nitra to Lódz), but not the investment decision itself, it cannot be said that the aid has affected the Limerick site. The Commission also takes note of the Polish authorities' argument that the decision to cease manufacturing activities by closing the manufacturing plant at Limerick is unrelated to Dell's decision to locate the investment project at Lódz.

(213) The Commission finds that the job losses resulting from the Dell group's decision to close down its manufacturing facility in Ireland cannot be regarded as a consequence of the aid granted by the Polish authorities. Accordingly, and notwithstanding other considerations in respect of this matter [55], the Commission considers that such job losses are not a result of the aid.

6.8. Balancing

(214) Having established that the aid provides an incentive to implement the investment in the region concerned, it is necessary to balance the positive effects of the aid against its negative effects.

(215) Pursuant to paragraph 53 of the Guidance Communication, if, without aid, the investment had been located in a poorer region (more regional handicaps - higher maximum regional aid intensity) or a region deemed to have the same regional handicaps as the target region (same maximum regional aid intensity), this would constitute a negative element in the overall balancing test that would be unlikely to be offset by any positive elements because it runs counter to the very rationale of regional aid. However, the Commission takes the view that attracting investment to a poorer region is more beneficial for cohesion within the Community than locating the same investment in a more prosperous region.

(216) The Commission notes that in the absence of the aid awarded by the Polish authorities, the investment would have taken place in another location (in this case, Nitra, Slovakia). However, as noted in paragraph 53 of the Guidance Communication, the Commission considers that "the positive effects of regional aid which merely compensate for the difference in net costs relative to a more developed alternative investment location (and thus fulfils the proportionality test above, in addition to the "positive effect" requirements as to objective, appropriateness and incentive effect), will normally be considered, under the balancing test, to outweigh any negative effects in the alternative location for new investment."

(217) In the present case, the Commission notes that the maximum aid intensity applicable in Západné Slovensko (the area in which Nitra is located) is 40 % GGE [56], whereas in Lódz Province it is 50 %. This difference in maximum aid intensities reflects the fact that, based on data measured in purchasing power standards for 2000-2002, the level of per capita GDP in relation to the EU-25 average was 45,42 % in Západné Slovensko and 41,45 % in Lódz Province.

(218) Lódz Province is therefore considered, for the purpose of the RAG and the Guidance Communication, to be a more disadvantaged region than Západné Slovensko. This means that, a priori, the benefits in terms of Community cohesion of attracting the investment to Lódz Province are to be considered greater than the negative effects associated with the investment not going to the Západné Slovensko region.

(219) Beyond the difference in the level of regional handicaps, there are several other indicators that the aid measure is likely to improve the level of cohesion in the Community.

(220) The Commission notes that in 2006 (when the decision was made to locate the project in Lódz), per capita GDP in relation to the EU-25 average in Západné Slovensko (62,8 %) was higher than in Lódz Province (48 %) [57]. While per capita GDP has increased in both Západné Slovensko and Lódz Province since 2002, growth has been considerably higher in the former. In other words, the difference in wealth levels between the two regions has thus considerably increased over the years.

(221) Furthermore, in 2006, unemployment in Západné Slovensko was 9,19 %, compared with 17,48 % in Lódz Province [58], which reflects higher relative labour demand and the better condition of the labour market in Západné Slovensko [59]. In addition, in 2001-2005 unemployment fell by more than six percentage points in Západné Slovensko, compared with a fall of barely 2,4 % in Lódz Province.

(222) Other indicators also point to a difference. For instance, the Polish authorities refer to the so-called "at-risk-of-poverty rate", which shows a substantial gap between Poland and Slovakia [60]. For 2006, the poverty rate in Slovakia was equal to 11,7 %, which is below the European average. In Poland, on the other hand, the rate was 19,1 %, putting the country above the Community average [61].

(223) Further insight can be obtained from statistics regarding migration rates, which can be viewed as an indicator of regional development as they relate to the opportunities that a region can offer to its inhabitants. Migration rates show a net inflow into Slovakia, whereas Poland has experienced a net outflow in recent years [62]. Net migration rates per 1000 inhabitants in 1995, 2000, 2005 and 2006 were - 0,5, - 0,5, - 0,3, and - 0,9 respectively for Poland and 0,5, 0,3, 0,6, and 0,7 for Slovakia.

(224) In respect of paragraph 54 of the Guidance Communication, the Commission notes, for the reasons described in recital 211, that the aid will not result in substantial job losses in existing locations within the Community, which would otherwise have been likely to be preserved in the medium term.

(225) In view of the above, the Commission finds that, given that the aid is proportional to the difference in net costs for carrying out the investment in the selected location in comparison with an alternative location in a more developed region, the positive effects of the aid in terms of its objective and appropriateness, as demonstrated above, outweigh the negative effects on trade in the alternative location.

(226) In accordance with point 68 of the RAG, and in the light of the in-depth assessment conducted on the basis of the Guidance Communication, the Commission concludes that the aid is necessary to provide an incentive effect for the investment and that the benefits of the aid measure outweigh the resulting distortion of competition and effect on trade between Member States.

7. CONCLUSION

(227) The Commission concludes that the proposed regional investment aid in favour of Dell Poland fulfils all the conditions laid out in the RAG and in the Guidance Communication for it to be considered compatible with the common market within the meaning of Article 87(3)(a) of the Treaty,

Has adopted this decision:

Article 1

1. The State aid which Poland is planning to implement in favour of Dell Poland, amounting to PLN 216365000 in nominal value, and representing a maximum aid intensity of 27,81 % gross grant equivalent, is compatible with the common market within the meaning of Article 87(3)(a) of the Treaty.

2. Implementation of the aid referred to in Article 1(1) is accordingly authorised.

Article 2

The Polish authorities shall submit to the Commission a detailed final report including information on the aid amounts paid, on the execution of the aid contract and on any other investment projects started at the same establishment or plant, within 6 months of payment of the last tranche of the aid according to the notified payment schedule.

Article 3

This Decision is addressed to the Republic of Poland.

Notes:

[1] OJ L 83, 27.3.1999, p. 1.

[2] OJ C 25, 31.1.2009, p. 9.

[3] Cf. footnote 2.

[4] In accordance with Article 6(2) of Regulation (EC) No 659-1999, this interested party has requested that its identity be withheld from the Member State concerned on grounds of potential damage.

[5] Throughout the rest of this Decision, "Dell" is used to refer to the Dell group as a whole or to the company and its management.

[6] Case N 531-06 Poland - National regional aid map 2007-2013 (OJ C 256, 24.10.2006, p. 7).

[7] The Polish authorities have confirmed that no other products besides those concerned by the investment project will be produced at the plant for 5 years after completion of the investment project.

[8] Exchange rate on the date of notification: EUR/PLN 3,7107. Reference rate applicable for Poland on the date of notification: 5,94 %.

[9] Business secret.

[10] As laid out in the MoU of 19 September 2006.

[11] Exchange rate of EUR/PLN 3,97.

[12] Based on discounted amounts.

[13] OJ L 302, 1.11.2006, p. 29.

[14] Information communicated by Member States regarding State aid granted under Commission Regulation (EC) No 1628-2006 on the application of Articles 87 and 88 of the Treaty to national regional investment aid (OJ C 220, 20.9.2007, p. 6).

[15] Information communicated by Member States regarding State aid granted under Commission Regulation (EC) No 1628-2006 on the application of Articles 87 and 88 of the Treaty to national regional investment aid (OJ C 270, 13.11.2007, p. 12).

[16] The investment agreement specifies that the applicable regional aid ceiling must not be exceeded and that the aid will be proportionally reduced if the expected investment amount is not reached.

[17] OJ C 54, 4.3.2006, p. 13.

[18] These are final products which are sold directly on the market. Other products such as handheld computers, peripherals, accessories or computer sub-components are not covered by the project.

[19] Point 69 of the RAG indicates that the relevant product market includes the product concerned and its substitutes considered to be such either by the consumer (by reason of the product's characteristics, prices and intended use) or by the producer (through flexibility of the production installations).

[20] As no data were available on Dell's forecast market share for the relevant markets concerned in 2013, the market share in 2012 was used as a substitute.

[21] The GDP average annual growth rate in the EEA for 2001-2006 was 3,94 % in current prices and 2,01 % in constant prices (source: Eurostat).

[22] Case COMP/M.2609 HP/Compaq, case IV/JV.22 Fujitsu/Siemens, case IV/M.1120 Compaq/Digital.

[23] Case COMP/M.1120 Compaq/Digital, case COMP/M.963 Compaq/Tandem.

[24] According to data submitted by the Polish authorities from an alternative source (Gartner), Dell's market share for x86 servers at worldwide level in volume terms is below 25 %.

[25] As regards desktops and notebooks, since the market shares are in any event below the 25 % threshold indicated in point 68(a) of the RAG, the Commission took the view in the opening decision (point (76) that the corresponding geographic market could be left open.

[26] In addition, the competitor also questions the validity of using growth figures for 2001-2006 in such a rapidly changing sector which is subject to new entrants, technologies and products.

[27] Case N 158/05 Getrag Ford Transmissions Slovakia s.r.o. (OJ C 236, 30.9.2006, p. 33).

[28] Paragraph 82 of the decision mentions the following: "However, in cases where large corporate groups are restructuring their operations, in particular by transferring production from one site to another, and by redeploying the resources freed up at the former site for other purposes, it appears unrealistic to artificially subdivide this activity into a series of different projects."

[29] Case N 872/06 - Individual aid to Qimonda (OJ C 170, 5.7.2008, p. 2), case C 45-03 (ex N 1/03) - Aid in favour of Infineon Technologies-Fabrico de Semiconductores, Portugal, SA (OJ C 235, 1.10.2003, p. 55), case C 86-01 (ex N 334/01) - Aid in favour of Infineon Technologies SC 300 GmbH & Co. KG. (OJ C 368, 22.12.2001, p. 2). In these cases, which concern the memory chip sector, the Commission favoured the use of volume figures essentially because of exceptional price increases in 2000 in comparison with the overall period caused by specific market events in that year. The Commission notes that such similar specific circumstances do not apply to the case at hand.

[30] Case C-234-84 Belgium v Commission [1986] ECR I-2263; Belgium v Commission [1986] ECR I-2321.

[31] The Commission notes in this respect that while it has given due consideration to the matters raised by the non-identified interested party in the confidential version of its comments, it does not consider that these require any further investigation on the grounds of competition concerns.

[32] OJ C 223, 16.9.2009, p. 3.

[33] Judgment in Case C-334-07 P Commission v Freistaat Sachsen, points 53, 56 and 58.

[34] OJ C 244, 1.10.2004, p. 2.

[35] Although demand-side substitution appears to take place primarily in one direction, from desktops to notebooks.

[36] i.e. "suppliers are able to switch production to relevant products and market them in the short term without incurring significant additional costs or risks in response to small and permanent changes in relative prices" (Commission notice on the definition of relevant market for the purposes of Community competition law, point 20, OJ C 372, 9.12.1997, p. 5).

[37] In particular, due to the lack of data, the Commission has not been able to conduct a market analysis based on quantitative methods (e.g. cross-price elasticity, SSNIP test, etc.) which would make it possible to define the relevant market. Calculations nevertheless indicate a 0,92 price correlation between desktops and notebooks, but there is no benchmark against which to consider whether this high correlation is indicative or not of reactivity of demand to small but permanent price variations between the two products. Furthermore, this high coefficient may be attributable to a common factor such as the price of inputs or to a common (downward) price trend rather than to competitive constraints.

[38] Point 27 of the notice on market definition referred to in footnote 34 indicates that "if under the conceivable alternative market definitions the operation in question does not raise competition concerns, the question of market definition will be left open." In opposition therefore, if competition concerns do arise under such alternative market definitions, then these should be considered.

[39] In practice though, the adjusted unit output per product, as provided by the Polish authorities on the basis of Dell's initial projections, is used, not the so-called engineering standard capacity (maximum theoretical capacity).

[40] As these thresholds are not exceeded for notebooks, it is in principle only necessary to assess the distortion of competition and effect on trade as regards the market for desktops and the market for servers. However, given that in the case in point, the plant in Lódz is characterised by supply-side substitutability as regards the production of desktops and notebooks, it is not possible to dissociate the competitive constraints that may affect the notebook market from those in the desktop market. Therefore, whereas for precautionary reasons, desktops and notebooks are considered as separate product markets, the assessment of capacity created by the investment project takes into account the theoretical maximum capacity created, regardless of the actual product breakdown.

[41] Unemployment in Lódz Province is 17,9 %.

[42] The current workforce is in excess of 1700 employees.

[43] This figure is expected to rise to 3000 over the investment period.

[44] Lódz University and Lódz Technical University.

[45] There are five public universities and more than 10 private institutes of higher education in Lódz Province. 20000 students graduate each year (10 % of all Polish graduates).

[46] This programme includes financing of scholarships as well as foreign-language training, internship opportunities and information exchanges on the labour market.

[47] Poland refers to cooperation between Dell and Irish universities (Trinity College, the University of Limerick and the Dublin Institute of Technology).

[48] Dell makes use of sophisticated plant logistics systems and this know-how is likely to be beneficial to other manufacturers in Lódz Province.

[49] e.g. English language classes, soft skills training, SixSigma certification courses.

[50] Dell Inc. Economic impact on national and local markets 2003-2004; Angelou Economics, p. 2.

[51] The Polish authorities indicate that other locations in the Community had been considered (in Ireland, the Czech Republic, Romania and Hungary) before the analysis was restricted to Poland and Slovakia.

[52] The document also examines the conditions surrounding execution of the investment project and implementation of the manufacturing and supply activities forming the objective of the investment project.

[53] The Polish authorities provided the Commission with the company documents and the LECG study during the assessment phase.

[54] The alternative location could be considered as suffering negative effects on trade as a result of the aid (cf. recital 216.).

[55] The Commission notes that Ireland has submitted an application for a contribution from the European Globalisation Adjustment Fund designed to support the reintegration into the labour market of workers affected by the redundancies at Dell and its suppliers in Limerick.

[56] Case N 469-06 Slovakia - National regional State aid map 2007-2013 (OJ C 256, 24.10.2006, p. 6).

[57] Source: Eurostat (GDP at current market prices in purchasing power standards).

[58] Source: Eurostat.

[59] Both regions can be compared as they are of similar size in terms of surface area and population.

[60] The "at-risk-of-poverty rate" shows the share of persons with an equivalised disposable income below 60 % of the national median equivalised disposable income after social transfers.

[61] Lelkes, O., Zolyomi, E., "Poverty Across Europe: the Latest Evidence Using the EUSILC Survey", Policy Brief, October 2008, European Centre for Social Welfare Policy and Research.

[62] OECD, International Migration Outlook 2008. Available at: http://www.oecd.org/document/3/0,3343,en_2649_33931_41241219_1_1_1_37415,00.html