Impact of Public Procurement Aggregation on Competition: Risks, Rationale and Justification for the Rules

This paper assesses the risks, rationale and justification for the rules on centralisation and aggregation of public procurement in Directive 2014/24. The paper explores the justifications advanced for the aggregation of purchasing and the countervailing risks it generates. In both cases, it focuses in economic and administrative aspects. It then proceeds to a summary overview of the new rules for the aggregation of public procurement in Directive 2014/24, and emphasised how the Directive is expressly recognising possibilities that clearly exceed the more modest approach in Directive 2004/18. Moving on, it then focusses on the potential justification for certain activities now permitted by the 2014 rules, and engages in a critical assessment of their competitive impact. The paper briefly highlights the far-reaching and not necessarily positive implications that a maximisation of the centralisation and aggregation possibilities under Directive 2014/24 could have, and proposes that strict competition law enforcement will be necessary to avoid undesired consequences. Some suggestions for further research are provided by way of conclusions.

buying power that they allow central purchasing bodies to exercise vis-à-vis economic operators can, however, also thwart the competitive process and constitute breaches of EU competition law as consequence of the unavoidable abuse of a dominant purchasing position (art 102 TFEU), or even of anticompetitive agreements (art 101 TFEU). We also submit that the 'public service' exemption is unavailable because centralised purchasing activities cannot be properly configured as a service of general economic interest (SGEI, art 106(2) TFEU), which triggers State aid risks (art 107(1) TFEU). Thus, adopting aggregated purchasing strategies demands awareness of their implications and the employment of mechanisms to avoid undesired legal, administrative and economic consequences.
Furthermore, we advance that the current regulation of the existing aggregated procurement strategies raises consistency and coherence problems with the rest of public procurement regulation and in particular the principle of competition embedded in article 18(1) of Directive 2014/24. Lastly, we put forward that an increasing tendency towards creation of central purchasing bodies on national level is likely to be replicated in a supranational level with the creation of regional or even European-wide central purchasing bodies. Such centralization tendency, however, poses serious risks that ought to be taken into account and that appear to be overlooked by legislators and procurement practitioners.
The structure of the chapter is as follows. We firstly explore the justifications advanced for the aggregation of purchasing (II) and the countervailing risks it generates (III). In both cases, we focus in economic and administrative aspects. We then proceed to a summary overview of the new rules for the aggregation of public procurement in Directive 2014/24 (IV) and emphasise how the Directive is expressly recognising possibilities that clearly exceed the more modest approach in Directive 2004/18. We then focus on the potential justification of certain activities now permitted by the 2014 rules and engage in a critical assessment of their competitive impact (V). We then briefly focus on the far-reaching and not necessarily positive implications that a maximisation of the centralisation and aggregation possibilities under Directive 2014/24 could have, and propose that strict competition law enforcement will be necessary to avoid undesired consequences (VI). Our conclusions summarise our findings (VII).

II. Justifications for aggregated purchasing
Aggregation strategies may grant benefits to direct and indirect participants of public tenders, the contracting authorities and taxpayers. Aggregation strategies can generate positive effects over the competitive process in the adjudication of public contracts by fostering competition between economic operators, lowering purchasing prices, suppressing duplicated procedures, generating operational economies of scale and reducing public expenditure with the aim of achieving better value for money in public procurement. 6 However, it is our view that the benefits potentially generated by aggregated purchasing will only take place if competition as a process is truly preserved and proper administrative and policy decisions are taken in order to minimize the pernicious side effects generated by aggregated purchasing. Thus, the benefits we discuss infra may only be achieved in aggregation is properly implemented as discussed in detail in sections IV and V, so that the risks spelled out in section III are minimised.

Economic benefits
The economic rationale behind strategies for aggregated purchasing is grounded on the accumulation of public purchasing market power by exerting bargaining power 7 and the creation of bureaucratic economies of scale. 8 Accumulation of public market power is carried out in two different ways. 9 On the one hand, by employing a single contracting authority that will be responsible for making purchases for other contracting authorities, such as is the case with central purchasing bodies (CPBs). 10 On the other hand, by means of public-public collaboration through occasional joint procurement, where two or more contracting authorities either enter into an agreement and collectively coordinate their purchases, or engage in minor forms of collaboration, such common preparation of tender documentation or entrusting one contracting authority with the management of the procedures. 11 Additionally, aggregated purchasing techniques, in particular centralised purchasing framework agreements or dynamic purchasing systems, generate bureaucratic economies of scales as the transaction costs and the total amount of tenders is reduced, thus promoting operational efficiency.

A) Bargaining power
In economic terms, aggregated purchasing aims at pooling public purchasing market power, in a same manner as a purchasing cartel among undertakings seeks to achieve it in order to maximize its profits. 12 In accordance with microeconomic theory, by adopting purchase aggregation strategies contracting authorities acting as a single entity are able to exert buyer power vis-à-vis its suppliers and obtain better terms and conditions when entering into public contracts. 13 Such buyer power, however, can take two different forms: monopsony power and bargaining power. 14 Only aggregated purchasing that fosters the creation of non-abusive bargaining power has sound economic justifications. Aggregated purchasing that generates monopsony effects, on the contrary, will tend to be inefficient. 15 If aggregated purchasing strategies are successful-ie they lead to the increase of purchasing market power, leverage negotiation advantages and increase technical knowledge-bargaining power will be generated. Consequently, the buyer will then be a competitive constrain to the 7 CHARD et al. (2008: NA26). 8 Also supporting this view that aggregated procurement techniques achieves "economies of scale" see CMA (2004: paras. 1.43-1.51);RISVIG HAMER (2014: 201). See also rec (59) dir 2014/24. 9 Somewhat similarly and distinguishing two mechanisms for achieving purchasing aggregation in the healthcare sector, see RACCA (2010: 121). 10 Rec (69) dir 2014/24. 11 Rec (71) dir 2014/24. 12 Paradoxically, anti-competitive agreements between competing undertakings are strictly forbidden under EU/EEA competition law in accordance with art 101 TFEU. For a discussion, based on US law, see CARSTENSEN (2010). 13 This would be the case of bargaining power, see OECD (2009: 22). 14 The term buyer power is used in this contribution as an umbrella term covering both monopsony and bargaining power, despite their very different welfare consequences. For a similar use of the term, see inter alia: OECD (2009); CHEN (2007). Cf with Noll who uses buyer power as a synonym of monopsony power in NOLL (2004NOLL ( -2005. 15 CHEN (2007: 28). suppliers and will be able to obtain more favourable terms and conditions than those offered to other contracting authorities or undertakings acquiring the same goods or services. 16 This, on the other hand, presupposes that aggregated purchase strategies are able to neutralize seller market power, thus acting as countervailing power. 17 This neutralizing force can make competition more vigorous by lowering prices paid and passing on the benefits to tax payers in the form of cheaper and better quality public services. To realize those benefits, however, the supply side must have some previous degree of market power allowing them to charge a price above pure competitive levels (ie the price under "perfect" market conditions). 18 Otherwise, providers will be unable to bring forth better terms and conditions than already offered and will opt for not taking part in the tender procedure. In this perspective, bargaining power generated by aggregation techniques acts as a countervailing force neutralizing the market power of economic operators and forces them to offer the most competitive prices and conditions to contracting authorities.
However, these benefits generated by bargaining power ought to consider exercising public buyer power differs from its exercise by a private entity profit making driven, and therefore, not all assumptions hold true in our case. 19 Firstly, the public buyer is constrained in its acting by EU and national public procurement rules which make exercising buyer power more difficult due to compliance with the principles of equal treatment, non-discrimination and the derived obligation of transparency. Tender procedures themselves impair the full exercise of buyer power because the contracting authority cannot set a purchasing price. Instead, the price is offered by the tenderers. Secondly, buying decision-making of the public sector must take into account other variables than in the private sector, such as being more risk-averse, it cannot (or ought not) compromise availability of goods or services (particularly when they are directly used in the provision of services to the public and where minimum quality standards are applicable) and, by and large, tends to be more conservative. Thirdly, contracting authorities do not aim at maximizing profits and, consequently, will be less oriented towards exercising buyer power. 20 Summing up the idea, by exercising bargaining power through demand aggregation, contracting authorities will be likely to obtain better contractual terms than if they were negotiating only by themselves vis-à-vis economic operators. These buyer power derived benefits will be obtained only if the market is not fully competitive (ie suppliers hold selling power that is resulting in supra-competitive equilibria), the contracting authority is aware of its purchasing power and the purchasing tools are employed in a way that does not preclude its exercise. In welfare terms, extra profits and inefficiencies generated by supply-side market power are then reduced or altogether eliminated. This implies two effects: a welfare distribution between suppliers and buyers and overall increase of efficiency and welfare. 21 16 CMA (2004: para. 1.10). 17 CMA (2004: para. 1.16). Countervailing power as a regulatory mechanism was introduced by GALBRAITH. 18 OECD (2009: 21). 19 Also recognizing this difference with the private buyer power, see CMA (2004: para. 1.13). 20 This latter general idea links back to the concept of x-inefficiency in the public sector. This concept was developed in the seminal work of LEIBENSTEIN. The concept is broadly accepted, although not by all, and still used in the shaping of EU economic policy. See AFONSO et al. (2010). 21 In this case, arguably, the increase in welfare will be positive regardless of the benchmark standard being that of consumer or total welfare. From a pure supplier perspective bargaining power is detrimental because it reduces "monopoly profits".

B) "Bureaucratic economics of scale"
Aggregated purchasing creates bureaucratic economies of scale that foster operational efficiency and reduce transaction costs. 22 In microeconomic terms, and due to the scale of its operational efficiency, aggregated purchasing allows contracting authorities to reduce the overall costs of the procurement activities if compared to the situation where contracting authorities would have carried out individual purchases. These economies of scale may prove to be one of the main reasons behind the rationale of adopting a pro-aggregation policy as they tend to be quite large. 23 Operational efficiency by means of aggregation techniques is achieved in two different ways. Firstly, by recourse to centralisation of purchases that would have been carried out individually by contracting entities, which pooling together eliminates duplicated procurement procedures. Instead of entering into several minor public contracts, aggregation techniques allows the public sector to enter into less frequent but much larger and arguably more complex contracts. Secondly, by entering into framework agreements or dynamic purchasing systems, operational efficiency is obtained as the successive procurement procedures (be they 'mini-competitions' or direct call-offs) carried out after establishing the system are simplified and streamlined (ie setting up the system implies front-loading costs, but the overall operational expenditure is reduced). This may significantly reduce the transaction costs for each successive purchase as there is no need to run a full tender procedure but instead a much simplified and shorter one.

C) Sponsoring of new entrants
Lastly, part of the literature suggests that aggregated purchasing may foster the entrance of new suppliers or encourage participation of local and/or regional SMEs into public procurement markets particularly by overcoming entry barriers or establishing obligations of exclusive purchasing. 24 In our opinion, however, such preferential treatment given to new players will almost always be in breach of the principles of equal treatment, its derived obligation of transparency and the general principle of (undistorted) competition, unless objectively justified and capable of overcoming very strict proportionality tests. 25 Additionally, such measures may constitute granting of unlawful state aid, as the advantages given are of a selective nature. 26

Administrative benefits
Aggregated purchasing strategies may generate other type of efficiencies that by their nature can be qualified as administrative, despite the fact they can also be quantified in economic terms. We distinguish two main administrative benefits: the professionalization of procurement services and the centralization of procedures, which leads to advantages in their control. 22 For some selected works on operational efficiency in the public sector, see: SMITH and MAYSTON (1987); RUGGIERO (1996); ANWANDTER and OZUNA (2002); RANTANEN et al. (2007); and AFONSO et al. (2010). 23 KARJALAINEN (2009

A) Professionalization of procurement services
One of the arguably natural consequences of adopting aggregated procurement strategies is the professionalization of civil servants that carry out public procurement due to the acquired expertise when awarding public contracts in a daily basis, 27 as acknowledged by recital (69) of Directive 2014/24. Procurement aggregation techniques are carried out usually by a group of highly qualified public officers that enter into public contracts on a daily basis. Due to the large amount of operations handled by these officers, it is expected that they will be able to accumulate expertise in carrying out public procurement procedures. 28 Also as clarified by Directive 2014/24, central purchasing bodies may carry out ancillary purchasing activities allowing them to act as "procurement advisors or trainers" for other contracting authorities, further strengthening professionalization of public procurement. 29 The benefits of professionally trained procurement officials are self-evident: skilled procurement officials take less time to plan and design tender procedures, are more efficient at conducting tender procedures and have a better overall knowledge for administering public contracts. Furthermore, professionalization of procurement officers implies that there will be a reduced risk of procedural mistakes because these experts will be law-aware and law-abiding, and consequently less public procurement litigation.

B) Control of procedures
One of the advantages generated by aggregated purchasing is that it grants a larger degree of control over the granting and execution phase of public contracts. 30 Central purchasing bodies or bodies created through ad-hoc public-public collaboration not only carry out purchases, but also administer public contracts on behalf of other contracting authorities-such as framework agreements or dynamic purchasing systems-draft technical specifications, carry out tasks of procurement management, training, contract management, inter alia. 31 These bodies, if entrusted with general managerial tasks in addition to pure procurement practices, are able to have a hands-on control over a large degree of public contracts and their execution as they tend to work as coordination bodies for other contracting authorities.

III. Risks generated by aggregated purchasing
As acknowledged by Directive 2014/24, 32 together with the sorts of benefits we have just discussed (above II), aggregated procurement strategies pose risks to public procurement markets both in the medium and long term, thus demanding that these instruments are carefully monitored to prevent the abuse of public buyer power that would erode the benefits of increased competition. In our view, the importance of these risks should be appropriately recognised. Aggregation techniques and procurement practices, therefore, should be employed by contracting authorities in accordance with the principle of competition embedded in article 18(1) of 27 OECD (2007); ibid (2000). See also STUIJTS et al. (2008). 28 Rec (69)  Directive 2014/24 and with the aim of minimizing these risks whenever entering into public contracts. In the following paragraphs we discuss the economic and administrative risks generated by aggregated procurement strategies.

Economic risks
Aggregation of public buyer power and improper employment of aggregation techniques may lead to restriction of competition and abuse of public purchasing power vis-à-vis economic operators. These situations may take place either by single behaviour or by collusive agreements among contracting authorities when designing tender procedures or imposing contracting conditions, as recognized by recital (59) of the Directive 2014/24. 33 We argue that aggregated purchasing strategies if not properly employed and monitored may cause the following pernicious effects: a) monopsony effect; b) abusive bargaining power; c) increase market concentration; d) waterbed effect; e) lack of incentives to innovate; and f) increase risk of collusion among tenderers. 34 These risks are further incremented if contracting authorities have a short term perspective of focusing on pure "best-value for money" (understood as price cuts) and overlook the competition consequences on market structures in a medium or long term. Due to the complexity of these issues we can only limit ourselves to present them in broad strokes with the intention of raising awareness to public procurement practitioners.

A) Monopsony effect
Monopsonistic buyer power poses the same risks for competition and market players as a monopoly does. 35 Monopsony takes place whenever there is either a sole buyer for a specific product or one very large and dominant buyer that faces competition only from fringe buyers with no market power whatsoever. Thus, it has no competitive constraints from direct competitors. In any case, the monopsonist will exercise its market power to obtain extraordinary contracting conditions by reducing the amount of goods and/or services purchased 36 from the seller to obtain a lower purchasing price. 37 The monopsonist can be at any level of the 33 "(…) the aggregation and centralisation of purchases should be carefully monitored in order to avoid excessive concentration of purchasing power and collusion (…)", rec (59) (2006: 471-474). 36 This conduct is named by SALOP (2004SALOP ( -2005 as "underbuying". "Monopsony conduct involves "underbuying" an input to profitably reduce its price". AREEDA, HOVENKAMP & SOLOW (2006: 442). 37 A similar definition is provided by the OECD when stating: "A firm as monopsony power if its shares of purchases in the upstream input market is sufficiently large that it can cause the market price to fall by purchasing less and cause it to rise by purchasing more". Note, however, how it is stressed the fact that monopsony under this definition can occur even if there is more than a single buyer in the market. OECD (2009: 9). This purchasing price needs to be below competitive levels in order for it to be considered as anticompetitive. See, for example the contribution of Canadian Competition Bureau as well as Germany's contribution in OECD (2009: 142 and 176), respectively. See also GAVIL, KOVACIC & BAKER (2008: 517-518). production chain 38 but it is most likely to be found at an intermediate level, such as it is the case of central purchasing bodies. 39 Monopsony is always negative from a welfare perspective because too few resources are employed, 40 and therefore there are unrealized gains from additional trade, 41 as well as higher prices in the downstream market generating monopsonistic incomes to the powerful buyer. 42 Thus, monopsonists are inefficient buyers.
Despite the fact that monopsony tends to be a rare phenomenon, public procurement markets tend to be scenarios where it might more often take place, 43 Particularly in case of highly centralised procurement activities. This is the consequence of the state usually enjoying a monopoly or quasi monopoly position in the market as the provider of downstream services or goods for which public procurement is used (for example: health provision services; basic educational services; road construction; transportation networks; provision of sewage, etc). Further, by centralizing or aggregating purchases the number of buyers-and "intra"-contracting authority competition-is further restricted and the market power of the central purchasing body is strengthened.

B) Abusive bargaining power
As briefly discussed, bargaining power can generally have a positive welfare effect (above II.1.A). However, we submit that if a dominant buyer is also a dominant provider downstream, the exercise of bargaining power may have negative consequences for welfare. 44 If the buyer is dominant in both markets it will not be constrained to pass on price benefits to end consumers in the form of lower purchasing prices or better quality/accessibility. 45 In public procurement this can take place as contracting authorities usually have a monopoly or quasi monopoly prerogative in meeting needs in the public interest which are not of a commercial or industrial nature, thus facing little competition from private parties. Because of having such a privileged "gateway" position, contracting authorities can impose anticompetitive conditions in the tender documentation-in the form of take it or leave it contracts or very restrictive terms. Some suppliers will be forced to commit to them as otherwise they will have no access to the end customers, hurting competition in the short and the long term. Others, however, will rather exit the public procurement supply market and enter into other economic activity. Also, to pass on 38 JACOBSON and DORMAN (1991: 5). 39 ABA (2012: 52). 40 BLAIR and HARRISON (2010: 44). The same negative consequence is pointed out by the OECD (2009: 9). 41 DOBSON et al. (1998: 12). 42 OECD (2009: 9). Note, however, how higher prices in a downstream market will only occur in the case the buyer uses the acquired input to produce goods and/or services that will be later on sold or offered in a downstream market. In principle, no higher prices in a downstream market will be generated if the input is acquired for inner consumption. 43 Monopsony power requires four main elements to exist, some of which are almost always present in public procurement markets. i) A large dominant buyer with a very high degree of market power; ii) few other buyers with little or no market power; iii) an upward slopping supply curve, which means that each additional purchase the monopsony makes increases the input prices of all the goods and/or services that has purchased before and thus, by cutting purchases it can decrease the price it pays; and iv) not sufficient countervailing supply market power. As a rule of thumb, the existence or lack thereof of monopsony power must be evaluated on a case-by case basis. 44 CHEN (2007: 36 the benefits to consumers, central purchasing bodies must make sure that either: better quality, more quantity and wider accessibility of the services are provided to final consumers or that the cost for society for the provision of these services/goods is also reduced which arguably will imply that less tax-payer's money is spent on public purchases.

C) Market concentration and squeezing of SMEs
Arguably, one the most debated issues concerning aggregated procurement is the tendency of concentrating markets both on the side of the suppliers and the buyers as recognized by the Directive 2014/24. 46 Market concentration and small and medium enterprises ("SMEs") leaving the market may be medium and long term consequences of aggregated procurement strategies, as larger but less frequent public contracts are entered into. By exercising unrestricted public buyer power-which is a conduct that a rational buyer would not engage in 47 -markets will see the number of its actors reduced because certain suppliers are forced to exit due to the lack of profitability and as a natural consequence of the increased competition. 48 This effect of the aggregation of purchases appears to be in clear opposition to the promotion of SME's participation in public procurement, one of the reasons behind the modernization of the rules, as they would be the most affected firms. 49 Three arguments ground this view. Firstly, public buyer power will impose demand side constraints over the competitive process lowering purchasing prices (arguably, a desirable effect) but also reducing the profitability of public contracts for economic operators. Secondly, as the contracts are infrequent but probably larger, participation from large suppliers will be very intense as these contracting operators cannot skip the opportunity to win a large and attractive contract. 50 Thirdly, SMEs will arguably have fewer chances to win the more infrequent but larger contracts 51 as they are unable to compete against larger suppliers due to economies of scale, manpower or simply not being able to satisfy minimum tender requirements. 52 By forcing SMEs out of the market contracting authorities, unknowingly, are not only reducing competition but might also be likely to be infringing article 102 TFEU because of entering into exploitative 46 Rec (69) dir 2014/24. 47 This is one of the underlying assumptions of bargaining models; that a profit-maximizing buyer would not behave irrationally by squeezing its suppliers to the extent that they will be forced to leave the market as it will be left without supply sources. However, microeconomic assumptions do not necessarily reflect the real world. 48 The fact that inefficient suppliers are forced to leave the market should not be a concern and it is not in EU competition law as the ECJ has held in several occasions that competition law should not protect inefficient competitors. 49 "(…) the public procurement rules adopted (…) should be revised and modernised in order to increase the efficiency of public spending, facilitating in particular the participation of small and medium-sized enterprises (SMEs) in public procurement" rec (2) dir 2014/24. For discussion of SMEs in EU public procurement law, see: BURGI (2007). 50 This will, of course, depend on the application of rules on the subdivision of the contract in lots and the possibility to submit bundled offers including rebates in case of award of all or most of the tenderer lots. However, fully discussing these issues would exceed the possibilities of this chapter. practices vis-à-vis its suppliers by, for example, offering unreasonably low purchasing prices or conditions. 53 Increased market concentration in public procurement markets is a serious problem because it reduces market players, competitive pressure and, consequently variety and alternative technical solutions. 54 Additionally, by entering into large and lengthy agreements the contracting authorities become "locked in" with a supplier who, benefiting from the absence of supply-side competition will in the future exercise market power vis-à-vis public buyers and raise prices. It also may increase fierce oligopolistic-competition that can become tacit or express collusion, a topic which we deal with in the following sub-section. 55

D) Waterbed effect
Part of the microeconomic literature suggests that due to techniques that foster the creation of buyer power, the prices paid by other buyers different from the one exercising buyer power will be higher. 56 The intuition behind this waterbed or knock-off effect is simple: suppliers will try to recoup the lost profit by increasing the selling prices that non-powerful buyers will have to pay to acquire their input, and as competition has been reduced, sellers have more market power compared to smaller buyers. 57 In public procurement, this would imply that smaller contracting authorities, or those not having agreements with central purchasing bodies or part of other types of occasional joint procurement, will pay a higher wholesale price than those with public buyer power for the same goods or services. 58 Paradoxically, smaller contracting authorities, for which public procurement tenders are already an economic and administrative heavy burden will see their purchasing prices increased. Consequently, the waterbed-effect is an incentive for this type of contracting authorities to employ aggregation techniques and increasing the risks generated by them-in a potentially vicious circle where contracting authorities can only try to avoid absorbing the negative impacts of contract aggregation and centralisation by furthering precisely those same techniques.
However, despite the waterbed effect, overall purchasing prices could still be lower thanks to aggregation techniques as reductions in price for aggregated procurement are larger than the increase for other smaller buyers. 59 The problem would, in that case, be 'limited' to an issue of cross-subsidisation between contracting authorities participating and not participating in 53 The abusive character of unfairly low purchase prices was raised in one instance (although not specifically in relation to public procurement procedures), but the ECJ dismissed the action on procedural grounds. See Case 298/83 CICCE [1985] ECR 1105. Notably, the possibility that unfairly low purchase prices are considered abusive was not excluded as such by the EU judicature and, consequently, seems to remain as one of the potential conducts covered by art 102(a) TFEU. See below V, and also TREPTE (2001: 275-276). 54 CHARD et al. (2008: NA29). 55 Also of this opinion, ibid, NA32. 56 DOBSON and INDERST (2007: 394). 57 OECD (2009: 23); CMA (2004: para. 1.35). Cf with the view of who argue that despite the intuitive appeal this argument is not entirely satisfying in, DOBSON and INDERST (2008: 342). 58 The same premium over competitive prices will be applicable (or even in a more significant proportion) to fringe non-public buyers that may acquire the same inputs for non-reserved downstream activities. 59 OECD (2009: 23). centralised procurement. Also, if smaller contracting authorities are able to implement "counterstrategies that eliminate or narrow any asymmetries of buying power" they may be able to get similar prices as those paid by contracting authorities employing aggregation techniques. 60 However, the likelihood of these counterstrategies seems limited if small or 'independent' contracting authorities are small and do not have access to highly qualified, professional purchasers.

E) Lack of innovative and investment incentives
The relation between aggregated procurement and innovation incentives is complex. 61 Part of the literature argues that there is a negative side effect on investment and innovation generated by a short-term perspective when employing aggregated procurement . 62 If contracting authorities have an overly strong focus on reduction of purchasing prices, suppliers might be unable to recoup their investment in R&D. 63 However, exercise of aggregated procurement techniques might have the reverse effect and may foster innovative ideas and create incentives for economic operators to invest, as demanded by Directive 2014/24 as "(p)ublic authorities should make the best strategic use of public procurement to spur innovation." 64 Innovation can be spurred for example by employing aggregated purchasing techniques encouraging procedures such as the innovation partnership, 65 allowing variant tenders, 66 and increasing competition among suppliers for innovative products or services. 67

F) Bid riggingcollusion risk
An important effect that aggregated procurement may have on competition in public procurement markets is increasing the incentive and opportunities for economic operators to enter into anticompetitive agreements or tacit collusion. 68 This tactic known as bid rigging allows economic operators to coordinate their behaviour when bidding for the award of public contracts. 69 By doing so, tenderers and not the contracting authority (or, rectius, a competitive process) determine themselves who and under what conditions should win the tender. 70 Unlike the risks mentioned hitherto, the collusive behaviour is not directly carried out by contracting authorities. Instead, aggregated procurement incentivizes tenders to do so.
In our view aggregated purchasing fosters collusionand does not prevent it -due to several reasons. 71 Price signalling and tacit or explicit communication is facilitated as less frequent purchases are carried out by a single entity. This minimizes costs and efforts to try to determine purchasing patterns as by having a "one-stop shop" contracting authority it becomes easier to track purchases. Since the contracts are arguably larger and economic operators cannot afford being "left out" they will have larger incentives to coordinate their behaviour and take turns in a survival strategy. 72 Although we do agree that incentives to cheat in large contracts are higher, 73 so will be the retaliation by other economic operators as the importance of winning such larger tenders is crucial. Also, as aggregated procurement increases market concentration, there will be fewer suppliers that need to coordinate their behaviour, thus creating further incentives to collude. 74

Administrative and legal risks
Aggregation techniques may also cause administrative and legal problems in addition to pure economic efficiency issues. We identify four concerns generated by aggregated procurement: a) risk management; b) decrease of autonomy for contracting authorities; c) principal-agent problem; and d) budget allocation issues.

A) Increase of complex and costly litigation
Employment of aggregation techniques implies the adoption of structures and procedures that are prone to complex and lengthy litigation initiated by economic operators that were not awarded the public contract in dispute. This is in our view derived from three sources. Firstly, as contracts are less frequent and larger in value the incentives to litigate for tenderers will be higher as the stakes are higher. This implies that contracting authorities must have a very high degree of care as the slightest error in the procedure may incentivise economic operators to file claims seeking the annulment of the tender procedure or the contract award.
Secondly, not only litigation risk is increased but also the complexity of that litigation, as not only the public sector professionalizes its procurement activities but the private sector continues to do so. Conflicts may tend to focus on small details or very technical issues. In turn, this overly complex litigation creates a burden for non-specialized judges who will have to deal with difficult issues. Thus, the risk of obtaining wrong decisions at a judicial level is also increased.
Thirdly, due to the larger size and value of public contracts the economic consequences for contracting authorities derived from infringements of public procurement laws increases proportionally. Not only because claims for damages and fines will be larger, but also because remedies such as interim measures suspending procedures may imply that perhaps very large 71 Cf this with the opposite view that centralized purchasing may help prevent collusion: CHARD et al. (2008). 72 Supporting this view on the incentives to collude due to large sizes of public contracts, OECD (2010: 24). 73 CHARD et al. (2008: NA32). 74 ESTEVAN DE QUESADA (2014: 232). amounts of needed goods or services will not be timely and steadily procured until the litigation is finished. 75

B) Decrease of autonomy for contracting authorities
Aggregated procurement techniques also imply de facto lessening contracting authorities' autonomy. This is an issue that will have a much larger impact vis-à-vis smaller local or regional bodies and that can have very significant implications in terms of local and regional autonomy and, more generally, the administrative structure of the State. By entering into aggregated purchasing mechanisms, local contracting authorities transfer their decisional autonomy to a central body with the expectation of obtaining better prices when acquiring goods or services. Now, this central body will make the purchasing decision by either seeking consensus among the different contracting authorities taking part of the aggregated procurement or by simply making an arbitrary decision. Local or regional bodies might feel that their decision making powers have been removed from them or that their view has not been taken into account. This can lead to an increase in political clashes and administrative litigation between the contracting authorities, implicating further expenses. Also, by losing autonomy, local or regional contracting authorities have more difficult the implementation of tailor-made policies as the input they acquire tends to be standardized, as we explain in the section below.

C) Principal-agent problem
Aggregation of purchasing techniques may generate principal-agent effects, a well knoweconomic dilemma. 76 The agent, the entity carrying out the procurement operation, makes procurement decisions that are not directly controlled nor observed by the principal. In our case, the contracting authority using aggregation techniques through an intermediary or in an scenario of public-public occasional joint procurement cannot fully observe the behaviour undertaken by the central purchasing body or the prime contracting authority. If the goals or interest of the agent differ from those of the principal, then conflict of interest between the parties arises. 77 The incentives for deciding which goods to buy may greatly vary between the agent and the principal and, as the decision ultimately lies on the agent, the action taken depends on the risk for the agent when contrasted with the benefits of obtaining the outcome that is in the best interest of the body carrying out the aggregated procurement. 78 Also, another problem raised by the principalagent dilemma is that the latter may be very keen in cutting expenses and achieving operational efficiency but by doing this "threatens lost value (and increased risks, for example, of corruption) in the long run." 79 In public procurement procedures, this implies that aggregated purchasing may lead to the acquisition of standardized or homogenous goods that do not satisfy the specific needs of all contracting authorities part of the aggregation scheme, as the procurement conditions might not be very specific. In other words, securing apparent best value for money (or best valur for money for the agent) may not necessarily achieve customer satisfaction (ie satisfy in full the 75 This scenario appears to be quite unlikely because interim measures in all Member States must be evaluated against a proportionality test, but is not impossible for it to happen. 76 For a thorough analysis of the principal agent problem in public procurement, in general, YUKINS (2010). 77 Ibid, 64. 78 GROSSMAN and HART (1983). 79 YUKINS (2010: 73).
contracting authority for which the procurement is ultimately intended). 80 If such is the case, then the quality of public services will be diminished as one size fits all does not really suit all situations nor it grants proper solutions to end users.
In order to minimize the principal-agent problem, aggregation techniques ought to be under surveillance mechanisms of both economic operators and contracting authorities. Contracting authorities can apply monitoring and supervising strategies to make sure that their needs are as aligned as possible so that customer satisfaction is achieved. Economic operators will act as an indirect control whenever litigating precluded practices by the agents. Nevertheless, despite these supervision measures there will be a residual loss that cannot be completely eliminated but merely mitigated and should also be taken into account. 81 And, in any case, the overall costs of the system may well be significantly increased due to this multiple layers of monitoring.

D) Budget allocation issues and overbuying
Lastly, the economic benefits obtained as a consequence of the purchasing buyer power or operational efficiency might be eroded due to overbuying of goods and budget allocation issues. Contracting authorities might feel the urge to reach the same level of expenditure as the previous year or to match the earmarked budget. 82 They might fear that if they do not do so, then these funds will be taking away from them, particularly in times of economic crisis. Accordingly, contracting authorities might be tempted to overbuy goods or services to reach certain budget targets and, therefore, the efficiencies generated by means of aggregated purchasing are washed away in the purchase of unnecessary goods in the fear of a future reduction of the body's budget. 83 This can be particularly problematic if centralised purchasing bodies adopt independent purchasing decisions when they act as wholesalers for the public sector (below IV). In those cases, if CPBs simply try to achieve the maximum level of economies of scale, they may well end up being left with excessive stocks of unneeded products. This may either result in a pure economic loss and waste in public sector activities, or in a further attempt of the CPB to resell (even at a loss?) the products to non-public markets-which would trigger very significant competition enforcement issues (below V). Overall, the risk of over-procurement is significant and, in our view, gets exacerbated in case of aggregation or centralisation of procurement.

IV. Tools for the aggregation of procurement in Directive 2014/24
As previous sections have shown, it is hard to provide a clear cut answer on whether the net effects of centralisation of procurement are positive or negative, both in economic and in administrative terms. However, it is also clear that centralisation of procurement is a common trend in procurement modernisation in the Member States. In recognition of this general trend, and along the general lines of facilitating public-public cooperation, 84 but in relation to 80 Ibid,[73][74][75][76]66. 82 Also raising this issue, see RACCA (2010: 123). 83 The rationale here appears to be that, at least in EU Member States where the crisis has had a larger impact on the public sector unspent resources are understood by legislators as non-needed funds and not as the result of efficient public procurement practices. 84 Generally, on the functional alternatives, ALBANO and SPARRO (2010). For a general overview of the rules, see ARROWSMITH (2014: 373-377 and 535-540).
'cooperate-to-buy' decisions instrumented through either centralised procurement or occasional joint procurement, 85 Directive 2014/24 regulates certain possibilities that go beyond the primitive rules on centralisation of purchases and the creation of central purchasing bodies contained in article 11 of Directive 2004/18. 86 As the European Commission has stressed, the new Directive makes it easier for contracting authorities to bundle their purchases by using joint procurement procedures or purchasing through a central purchasing body. 87 This can be done on a national or cross-border level. 88 The justification for the increased detail in the regulation of centralised and collaborative procurement, 89 including cross-border cooperation, can be found in recitals (69) to (71) of Directive 2014/24, where the increasing relevance of these procurement techniques is echoed, 90 and an interesting direct reference is made to the potential increase in competition that can derive from the use of these techniques (see above II for further discussion). However, it must be borne in mind that generally (and as pointed out above III), the centralisation of procurement activities creates significant risks of distortions of competition-which, as also already mentioned, is acknowledged in recital (59): 'the aggregation and centralisation of purchases should be carefully monitored in order to avoid excessive concentration of purchasing power and collusion, and to preserve transparency and competition, as well as market access opportunities for SMEs '. 91 Hence, this is an area where particular care should be exercised in trying to avoid distortions of competition.
In that regard, it is important to stress that the rules of Directive 2014/24 deviate in significant ways from what would be desirable from a competition perspective. Central purchasing bodies are now clearly assigned two alternative (or, actually, possibly cumulative or parallel) roles under Directive 2014/24. As explained in recital (69)  Thus, on the one hand, central purchasing bodies can act in support or on behalf of contracting authorities (ie 'act as intermediaries by awarding contracts, operating dynamic purchasing systems or concluding framework agreements to be used by contracting authorities'. Under this 85 On the situations in which recourse to this type of cooperation is desirable, see BAKKER et al. (2008). 86 This provision received very limited attention from the ECJ, given that it was not applicable ratione temporis to the facts in Case C-220/05 Auroux  88 However, for the purposes of this paper, the distinction between internal or cross-border recourse to a central purchasing body will only be taken into consideration in terms of the magnitude of the likely effects on competition. The detailed rules on applicable law established in art 39 dir 2014/24 will not be assessed in detail. 89  instrumental role, central purchasing bodies would be acting as agents of the public sector and their interface with competitive markets will be limited to their recourse to the market for the procurement of the goods, services (or works) ultimately needed by the contracting authorities 'downstream'. Hence, under the currently applicable rules, and particularly under the FENIN-Selex exclusion from the concept of undertaking, 92 'instrumental CPBs' activities would not be controlled by competition law (with the exclusion of State aid rules).
On the other hand, they can act as the actual providers of other contracting authorities (that is, 'act as wholesalers by buying, stocking and reselling'). Under this commercial role, central purchasing bodies would actually engage in an economic activity, as they would be assuming economic risk derived from any mismatch between the volumes they acquire and those they manage to supply to the contracting authorities downstream. This second role of 'commercial CPBs', which would engage in downstream economic activity in competition with 'private suppliers' (or other competing commercial CPBs), would no longer be exempted from the application of competition rules under the current criteria 93 -unless the ECJ extends the FENIN-Selex exemption beyond its current remit, possibly on the basis of a justification more or less tailored to a configuration of CPB activity as a service of general economic interest (SGEI). However, in our view, the (even clearer) economic character of this sort of CPB activities should make them fall completely under the umbrella of competition law (see below V).
Interestingly, there is a third (implicit or derivative) function for central purchasing bodies, which can be directly engaged by contracting authorities for the provision of 'ancillary purchasing activities' 94 and, possibly, remunerated for them. This would create a sort of (potentially free standing) 'advisory CPBs' activity which, in case it is remunerated, would also seem to represent an economic activity that should be subjected to competition rules.
It is now clear that all of these roles are expressly regulated in article 37(1) of Directive 2014/24 as far as instrumental and commercial activities are concerned (which suppresses any legal uncertainty derived from the silence of dir 2004/18), and art 37(4) as regards advisory or ancillary activities. It is also worth stressing that Member States can make the recourse to the central purchasing body mandatory (art 37(1) in fine dir 2014/24). This latter possibility creates very difficult to anticipate competition effects, as it makes the supply of the goods, works or services to the public sector depend on the running of a 'two-sided' platform by the central purchasing body. In that case, depending on the way in which demand is aggregated or bundled, the exclusionary effects on (particularly smaller) suppliers can be very relevant (above III.1.c). Moreover, generally, there seems to be no good reason to impose recourse to the central purchasing body if a given contracting authority can obtain better conditions (ie, better value for money) from an alternative provider. In that case, the principle of competition in article 18(1) of Directive 2014/24 would require carving out an exception from the rule of obligatory recourse to the central body when it is not the one offering the most economically advantageous tender (although, admittedly, this would create practical difficulties if the contracting authority just decides to rely on the central body without carrying out any independent market consultation, under art 40 dir 2014/24 or otherwise). According to the rules in article 37 of Directive 2014/24, recourse to a central purchasing body exempts the contracting authority from complying separately with public procurement rules (on the assumption, obviously and unavoidably, that the central purchasing body is the one bound by them in its market interactions), unless it directly carries out one or more of the phases involved in the procurement process (as indicated in art 37(2) dir 2014/24). 95 Moreover, contracting authorities can award a public service contract for the provision of centralised purchasing activities to a central purchasing body without applying the procedures foreseen in Directive 2014/24. Such public service contracts may also include the provision of ancillary purchasing activities, which as already mentioned implies that there can be an element of remuneration of this service.
Therefore, recourse to central purchasing bodies is fundamentally excluded from the scope of application of Directive 2014/24 in a sort of special case allowing for the use of the negotiated procedure without publication (or by analogy with art 32 dir 2014/24, or even with art 12 dir 2014/24, as the logic and functional rules applicable to in-house provision and public-public cooperation seem rather common), which however has a dubious justification, particularly if the centralised purchasing body is a body governed by public law with private capital participation. 96 Under the rules of Directive 2014/24, centralisation of procurement is seen as a clear device to allow (small) contracting authorities to achieve savings, 97 as well as higher standards of professionalization, 98 and to reduce the administrative burden of running procurement procedures by having recourse to the services of the central purchasing body-in a sort of intermediate solution between a public-public cooperation scheme (for which there would clearly not be a sufficient cooperative element) and an in-house arrangement (for which the control criterion would probably be absent). 99 From the competition perspective, this possibility basically moves the focus of the competition concerns to the market activities of the central purchasing body and increases the likelihood of distortions of competition, and it may as well result in the central purchasing body engaging in a sort of 'market regulation' activity that is difficult to align with the general requirements of the principle of competition. Consequently, it is a development that causes significant source for concern in terms of the development of a procompetitive public procurement system.
Along the same lines, article 38 of Directive 2014/24 sets rules for the carrying out of occasional joint procurement and article 39 of Directive 2014/24 extends the possibility to have recourse to centralised or occasional joint procurement involving contracting authorities from different Member States. These rules aim at settling important legal issues concerning the applicable law and, from a competition perspective, the only provision that is worth mentioning is that article 39(2) establishes that Member States 'shall not prohibit its contracting authorities from using centralised purchasing activities offered by central purchasing bodies located in another Member State'. This could create a certain level of competition between centralised purchasing bodies that could in turn reduce the likelihood of distortions of competition even if their use was made compulsory. However, in a difficult to understand restriction, this same provision allows 95 RISVIG HAMER (2014). 96 Generally, see ØLYKKE and FANØE ANDERSEN (2015). 97 See KARJALAINEN (2011). 98 However, recourse to centralised procurement does not eliminate the need for proper (decentralised) contract management. See ALBANO and ZAMPINO (2012);. 99 For extensive discussion, see SANCHEZ GRAELLS (2015: ch 6).
Member States to limit the possibility of having access to 'foreign' central purchasing bodies that act as wholesalers or as intermediaries-ie from having access to 'commercial CPBs'. That is, a given Member State can decide to allow its contracting authorities to use the services of 'foreign' central purchasing bodies acting as wholesalers, but not of those acting as intermediaries-or viceversa. This is a provision that is difficult to understand if not on the basis of the (likely) constitutional restrictions that some Member States may want to impose on the first possibility (ie, on the possibility of central purchasing bodies acting as wholesalers in competition with private undertakings or, even more, on the basis of a reserve of activity amounting to a monopoly), which they would also be keen on extending beyond their borders. In any case, however, from a competition law perspective, it seems desirable to subject all decisions based on articles 37 to 39 of Directive 2014/24, and 37 especially, to a proportionality test that balances out the expected benefits in terms of reduction in administrative costs and exploitation of economies of scale against the likely distortions of competition in the market where the central purchasing bodies are active (see above II and III). It is further submitted that, in that assessment, the relevance of the latter distortions should be highlighted as a way of providing effectiveness to the principle of competition embedded in article 18(1) of Directive 2014/24. Moreover, depending on the circumstances, the reservation of the activity to the central purchasing body could amount to State aid, which could be difficult to justify on the basis of a potentially nonexisting service of general economic interest (SGEI). Therefore, from a competition perspective, the implementation of article 37 of Directive 2014/24 creates significant risks that will deserve careful consideration by the ECJ in the future.

V. Justification of certain activities permitted by the 2014 rules and critical assessment of their competitive impact
From the cursory analysis carried out in the previous section, there are two issues that stand out and, in our view, are likely to create significant competition problems and litigation in the future. One of them is the temptation of Member State to declare the activities of the central purchasing body a service of general economic interest (SGEI) and, consequently, seek an exemption from the application of competition and State aid rules under article 106(2) TFEU. The second one, which may or may not be linked to such an intended exemption, concerns the possibility for 'commercial CPBs' or 'advisory CPBs' to supply goods or provide services to private buyers in the market. We will assess each of these possibilities in the remainder of this paper.

Can the provision of centralised purchasing activities be defined as an SGEI?
Generally speaking, under article 14 TFEU, Member States retain discretion to define, commission and provide SGEIs, but then they have to comply with EU law rules under article 106(2) TFEU. 100 Thus, in the event that the central purchasing body was considered to be entrusted with the operation of 'services of general economic interest' 101 , articles 14 and 106(2) TFEU will empower the state to relax the regulation of its activities and to allow for certain competition-restricting behaviour (ie, to enact regulation that departs from general EU law and, notably, from the competition rules of the TFEU), as long as it is necessary for the performance, in law or in fact, of the particular tasks assigned to the public contractor. 102 Along these lines, article 106(2) TFEU could arguably be used as the legal basis to exclude the application of EU competition rules in connection with the 'commercial' or 'advisory' activities of central purchasing bodies entrusted with the provision of centralised purchasing activities as an SGEI. If successful, under this strategy article 106 (2)  Overall, though , our opinion is that the 'public service exemption' under article 106(2)TFEU is not available to the exclusion of centralised purchasing activities from competition law scrutiny on the basis that they are (or can be) considered a service of general economic interest (SGEI). Indeed, the proper interpretation of these provisions must rely on the case law of the EU judicature that has set the conditions required to avoid the use of the 'public mission exception' in article 106(2) TFEU to subvert the basic principles that inspire the EU public procurement rules. The basis for such an interpretation is straightforward. The proper performance of the tasks entrusted to the public sector generally does not require a departure from the basic principles of the TFEU and public procurement rules. In the specific case of the exemption as an SGEI of centralised purchasing activities, the limitation that Member States should be unable to overcome is that there is no market failure that requires central purchasing bodies to be exempted from competition rules. It is clear that the private sector is able and willing to supply the public sector with goods and services and, even with the creation of SGEI-covered central purchasing bodies, the ultimate suppliers of the public sector would continue being the same suppliers (or, possibly, some of them, given the exclusionary effects that centralisation can create). Moreover, there are less restrictive ways in which central purchasing bodies can carry out their mission (ie 'intermediary' CPB activity), which would not support an assessment of proportionality in the application of the article 106(2) TFEU exemption to the alternative and market-oriented options (ie 'commercial' and 'advisory' CBP activity). Additionally, in case the CPBs did not have the capacity to meet all existing demand for centralised purchasing services (and this should be seen as a more than theoretical risk, due to funding and other constraints), the justification for the reservation of the activity as an SGEI would also fall foul of the existing case law on the interaction between article 106(2) and 102 TFEU. 103 Finally, and contrary to the purpose of article 106(2) TFEU, the exemption of CPB commercial and advisory activities from competition law scrutiny would be a significant source of potential distortions of competition, would significantly jeopardise the principle of competitive neutrality and could ultimately result in significant losses of social welfare (above III).
It is also worth mentioning that, due to the possibility of entrusting central purchasing bodies with the carrying out of centralised purchasing activities and ancillary purchasing activities (even for remuneration) by direct award under article 37(4) of Directive 2014/24, the impossibility of having recourse to article 106(2) TFEU would also immediately trigger the enforcement of State aid rules under article 107(1) TFEU due to the (practical) impossibility of meeting the fourth Altmark condition in the absence of a procurement exercise. 104

Can 'commercial CPBs' and 'advisory CPBs' engage with private buyers?
Finally, it is worth highlighting that, under the rules in articles 37 and 39 of Directive 2014/24, there seems to be no obstacle for a central purchasing body engaged in commercial and advisory activities to offer its services and supplies to purchasers outside the public sector. This could either be attempted as a general strategy to achieve even larger economies of scale than warranted by the aggregation of public demand, or as a 'minimisation' strategy in cases of overacquisition by the public sector (above III.2.d). In this regard, a systematic interpretation of these provisions together with article 12 of Directive 2014/24-which regulates analogous situations of 'public offer' of goods and services in the context of public-public cooperation and in-house provision-could suggest that offering such supplies and services for an equivalent of up to 20% of the total turnover of the central purchasing body would not be necessarily out of line with the general criteria governing these sort of 'out-of-market' arrangements.
In our view, in case Member States decided to authorise provision of centralised commercial and advisory activities to non-public purchasers, competition rules would be automatically triggered and CPBs' behaviour should be intensely screened to avoid distortions of competition. The main sources of potential distortions would be leveraging of their possibly dominant positions, crosssubsidies that allowed them to engage in predatory behaviour and other unfair commercial practices. Most of them would be covered by Article 102 TFEU and, according to Article 3(2) of Regulation 1/2003, 105 Member States could enforce even stricter standards under their domestic provisions. For all the reasons indicated above, we submit that they should do so. VI. Towards a "Regional or even European Central Purchasing Body"? Or, at least, a "European Network of CPBs"?
Further to the discussion above, and from a more holistic perspective, we are concerned that the combined application of the rules in the new Directive can result in scenarios where competition in public procurement can be significantly affected as a result of excessive (and improperly managed) centralisation of purchases. Indeed, EU public procurement developments and the success of aggregation techniques, in particular central purchasing bodies and framework agreements, appear to be paving the way for the creation of regional central purchasing bodies wherein inter-Member State procurement is carried out a by a regional central purchasing body. If this trend towards aggregation and centralization or purchases continues it is indeed possible to foresee that certain sectors in the EU may advocate for the creation of European (sectoral?) central purchasing bodies that will carry out the procurements for all Member States based on the idea of grouping public buyer power and reducing administrative expenditures.
Already, the Directive 2014/24 appears to be moving towards such direction as one of the new aspects introduced regarding central purchasing bodies is that it precludes Member States from prohibiting its contracting authorities from "using centralised purchasing activities offered by central purchasing bodies located in another Member State." 106 Despite the Directive clarifying that public procurement shall be carried out in accordance to the laws of the Member State of the central purchasing body, it leaves important questions unanswered, such as: which law will govern the contracting authority-central purchasing body relationship? Is this to be left to the application of rules on conflicts of laws? May this be the cause of increased inter-Member State litigation before EU Courts? Will further centralization lead to European-wide one size fits all solutions that, although cheaper, do not properly satisfy the needs of European taxpayers? Furthermore, by setting regional or in a perhaps not so distant future of a hypothetical network of European central purchasing bodies, the benefits resulting from aggregation are possibly increased but also, and to a worryingly degree, the inherent risks to purchasing aggregation techniques. Moreover, at some point, the central purchasing scheme could be 'too big to be regulated' due to political pressures. Thus, in our view, an increasing tendency for Europeanwide centralization is a legitimate source of concern for the well-being of European competition within public procurement markets and market competition in general. The only recommendation that can be made at this early stage, where the creation of such network of CPBs may seem delusional to some, is to apply standard competition rules. Some interesting guiding principles can be found in the more general criteria applicable to the accumulation of buying power (above III), or to the treatment of networks of national entities with legal monopolies and cooperation or mutual recognition schemes. 107 Other than that, guidance from the European Commission on the enforcement and interpretation of articles 37 to 39 of Directive 2014/24 from a competition law perspective would be welcome, particularly if it is preceded by a transparent public consultation.

VII. Conclusion
This paper has adopted an economic and cost-benefit approach to the exploration of the benefits and risks that centralised public procurement techniques can generate, with a particular focus on the activities of central purchasing bodies. It has then briefly explored the new rules included in articles 37 to 39 of Directive 2014/24 and has stressed that most of the relatively new functions that the Directives permit for CPBs should and must be subjected to strict competition law enforcement. More specifically, it has dismantled the idea that most CPB activity can be considered a service of general economic interest (SGEI) and, consequently, rejected the possibility of exempting CPBs from the application of EU competition law on the basis of article 106(2) TFEU. It has also flagged up two sources of particular concern, such as the engagement of CPBs with private purchasers and the potential excessive centralisation if recourse to crossborder centralised public procurement is not restricted and subjected to the appropriate competition controls. We plan to explore some of these issues in further detail in future papers.