Pre-commercial Procurement and the Marketing of Innovation: A New Innovation Policy Instrument or “Old Wine in New Bottles”?

We analyse the peculiarities of pre-commercial procurement (PCP) within the EU innovation policy kit, interpreting it as a risk-shifting mechanism. While most studies assume the point of view of the public procurer, we take the suppliers’ one, with a focus on SMEs. We conclude that PCP, operating in the phase of the specification of technical requirements, has the peculiar role of favouring a better match between supply and demand as for innovative goods and can allow exploiting the competitive advantages that SMEs display in the early phases of the technological process, if interpreted as an innovation-leaking process.


Introduction
A new consensus has emerged at EU level on the importance of demand-side policies as an engine for innovation (European Commission 2005, 2006, 2014a).This has stemmed from the empirical results on their effectiveness (Nelson, 1982), and in comparison with supply-side instruments (Guerzoni & Raiteri, 2015;Aschoff & Sofka, 2009).
The new approach places an emphasis on the slowdowns and interruptions of the innovation process that can happen after investment in experimental research has occurred and before the commercialization phase has been reached.
The research and innovation life cycle whereby an idea is transformed into a viable marketable product (or service) is traditionally divided into four phases: phase 1 consists of the exploration of solutions and feasibility studies; phase 2 covers R&D activity up to the creation of prototypes; in phase 3, through continued R&D the first pre-series batch of pre-commercial volumes of pre-products is produced in order to carry out a field test that will then validate the product (or not); phase 4 sees the production and commercialization of large scale volumes of the product take place.The R&D risk level usually decreases progressively through the four phases assuming value "4" at the beginning of the first phase and reaching "0" at the end of phase 4 (European Commission, 2006).
Because of the new policy approach, instruments that address phases after experimental research (basic research subsidies) and before commercialization (public procurement for innovation, PPI, previously called public technology procurement) have entered the European Commission innovation policy kit, and these include pre-commercial procurement (PCP).
PCP is a kind of procurement of R&D services aimed at stimulating R&D activity in the period before the production of prototypes or first-test series and without entailing the commitment to buy innovative commercial products.Because of its intermediate position between supply-(research subsidies) and demand-side (commercial public procurement) measures, it has been defined as the 'missing link' in the EU innovation cycle (European Commission, 2006).Part of the literature has, however, been reluctant to accept this interpretation, considering it either among demand-side or supply-side policy tools (Edquist & Zabala-Iturriagagoitia, 2015).
The new policy paradigm especially targets small and medium sized enterprises (SMEs).Though PCP is not directly addressed to SMEs, its antecedents are the US SBIR and the UK SBRI programmes, both of which were specifically tailored for small firms.Moreover, EU documents often refer to the opportunities that PCP can offer to SMEs (European Commission, 2006, 2011, 2014a).The connection between innovation and SMEs is not new within EU policies: the 1986 SME Action Programme contained measures aimed at improving the competitiveness of SMEs by enhancing their ability to innovate; the issue was revisited in the Lisbon Strategy and in the Small Business Act (European Commission, 2008, 2009a).Nevertheless, the effectiveness of PCP in supporting innovation in SMEs is questionable (Cepilovs, 2012).
Against this background, we review the existing literature and policy documents on innovation marketing to address the question whether PCP is an effective tool to stimulate marketable innovation in a SMEs perspective.We first focus on the specific role that SMEs can play in the innovation process and at the market failures that can hinder it (section 2.1).We then review the literature on the limitations that traditional public commercial procurement faces (section 2.2).Based on this, we analyse whether PCP represents a more effective alternative for stimulating innovation and growth skills in SMEs (section 3), given its structure as deriving from the EU legal framework (section 4).The final section comments on the results and their policy implications.

Innovation and SMEs
In the nineties, the Swedish economy proved unable to deliver innovative products despite its remarkably high commitment to R&D.Edquist and McKelvey (1998) coined the term the 'Swedish paradox' to describe the phenomenon which a few years later was also detected in the German economy (Kuhlmann, 2003).In fact, the incapacity to translate top scientific performance into marketable innovation represented a widespread European problem.It could be argued that one of the causes of the EU paradox is the delay of SMEs compared to their US counterparts, as one comes to productivity and innovation (European Commission, 2009a, 2014a).European SMEs reveal lower innovative capacity than their US counterparts since most of them do not survive very long (Santarelli & Vivarelli, 2007).The relevance of the performance of SMEs to economic policy appears obvious, if one considers that they are the prevalent form of business organisation in the EU (Cepilovs, 2012).Moreover, the economic literature has recognised the substantial role that SMEs can play in the economy as a source of innovative activity (Thurik, 1996;Noteboom, 1994;Burrone & Jaiya, 2005;Audretsch, 2002) and a stimulus for economic development (Biggs, 2002;Newberry, 2006), industrial growth and job creation (Acs & Audretsch, 1987), entrepreneur development (Johnson & Cathcart, 1979) and increased competitiveness (Song & Perry, 1997).It follows that enhancing the possibility for innovation to occur at the SME level could result in a widespread positive effect on the economy as a whole.Innovation deadlocks specific to SMEs have thus been addressed (European Commission, 2011).
The 'European paradox' marked the shift from the 'science-technology paradigm' to the 'innovation paradigm' in EU policy making.The traditional approach to innovation policy was based on the idea that firms' ability to innovate rested on the amount of investment in R&D.This is typically less than socially optimal, because of market failure instances connected to the characteristics of scientific knowledge and technological development.This approach dates to Arrow (1962) and Schumpeter (1975) and provides the theoretical rationale for the granting of intellectual property rights (IPRs) and the public funding of R&D.
The European paradox, however, showed that R&D investment is not a sufficient condition for innovation.At the beginning of the new millennium, European policy making turned to a new theoretical rationale, provided by the literature on national innovation systems (Lundvall, 1992;Nelson, 1993;Lundvall & Borràs, 1997).The new perspective focusses on systemic and institutional aspects of innovation and on the firms' learning process.It identifies typologies of market failure other than those related to the properties of scientific knowledge.These include failure in the provision of infrastructures, transition failure, and lock-in failure (Smith, 2000) and learning failure, exploitation-exploration and variety-selection trade-off, appropriability traps and complementarity failure (Malerba, 2002).
The innovation paradigm literature also emphasises the role of the users of innovation in its generation, development and diffusion, with reference to lead (von Hippel, 1986), sophisticated (Malerba et al., 2007) and experimental (Rogers, 1962) users and to the demand driven effects stemming from the interactions between innovative firms and innovative users (Lundvall, 1985;Malerba, 2007).Thus, innovation policy shifted from the emphasis on supply-side measures to a more comprehensive approach, based on a mix of supply and demand instruments.Among the latter, PPI became a systemic policy, aiming at pulling innovation through the demand for new products and solutions (Porter, 1990;Edler & Georghiou, 2007;European Commission, 2003, 2005, 2006).
The new paradigm also focuses on the place of SMEs within innovation policy, since they are considered strong players in knowledge intensive sectors, with cost effective R&D activity and the capability to transform research results into marketable products.This is because SMEs appear to have a comparative advantage over large enterprises (LEs) due to their flexibility and customisation of products in the early, fluid phases of the technological cycle (Utterbach & Suarez, 1993;Rothwell & Zegveld, 1982).It is also recognised that they are more capable to 'catch' individual consumers and to increase the range of products and are able to enter markets in which economies of scale are absent and that are, therefore, left aside by larger enterprises.On the other hand, SMEs are at a disadvantage in the later, crystalized phases of the technological cycle, when standardisation is associated to gain from scale economies in production.Size related constraints are not associated with the production phase only.SMEs find it more difficult to attract funds than other innovative firms (European Commission, 2012, 2013), since financing SMEs is generally riskier than financing LEs (Berger & Udell, 2005;Ayyagari et al., 2006;Beck et al., 2008).This is because size forces SMEs to concentrate on a small number of projects and this prevents them from differentiating and spreading risk.Also, scarcity of funds can hinder investment in R&D facilities and in the employment of high-skilled workers (Acs & Audretsch, 1990) and this often forces firms to restrict their search for new markets to local opportunities or at most to the national context, thus giving up the search for business opportunities in foreign markets.
Venture capital mitigates these problems (Engel, 2002), given its efficiency in operating under conditions of uncertainty and information asymmetries which characterise innovative projects in small and young firms (Caselli et al., 2008).However, venture capitalists concentrate on sectors with high perspective economic returns in the short-and medium-term.(Gans & Stern, 2000), which leaves out the smallest, youngest and most experimental firms, especially if development costs are large and returns on investment are expected only in the long-term (Apostol, 2017).
Another mitigating factor could be getting access to knowledge and production networks (Zanfei, 1994); however, there is a lack of these between SMEs, as the latter usually limit their interactions to customers and local partners on the value chain (Kaufmann & Toedtling, 2002).
When markets fail to deliver the right incentives, there is scope for public intervention to overcome innovation deadlocks.

PPI and SMEs
Size generates a trade-off between flexibility and availability of resources that reduces the likelihood that SMEs will engage in innovation activities.Public intervention can represent an option to not lose the advantages deriving from the peculiarities of SMEs, if it is tailored to their innovation potential.The literature shows that this lies in their ability to adapt to users' needs and this is important in the fluid phase of the innovation cycle.Policy should target the early stages of the technological process (Dalpé et al., 1992;Mazzucato, 2011).For instance, it could support the creation of knowledge and production networks to encourage SMEs to pool their R&D facilities and work forces; or support the creation of venture capital funds aimed at financing innovation in SMEs.Also, in the presence of hindrances on the demand side, the government could take the role of a lead or experimental user (Georghiou, 2006).This could be done through public procurement for innovation (Edquist & Hommen, 1998).A public agency could play the role of 'quality leader', acting, for instance, as an experimental user in the early stages of product cycles, when technical uncertainty is high (Malerba et al., 2007;Uyarra & Flanagan, 2009) and commercial volumes do not reach the critical mass necessary to reach the break-even point.A classic example of successful public procurement for innovation is the development of the semiconductor industry in the US in reaction to stimulus from the Department of Defence (Rothwell, 1994).In markets with important societal needs (Edquist & Hommen, 1998;Jaffe et al., 2004;OECD, 2014), the government might intervene to organise demand, through the specification of requirements based on socially desirable objectives.Successful examples include Japan's Rental Agency (JECC) with regard to the development of the domestic computer market and the NUTEK experience in Sweden (Nilson, 1994).In these cases, public procurement served as a catalyst, the motivation for intervention being to stimulate demand rather than to improve the provision of a public service (Note 1).
Public agencies also provide credibility to small firms (Edler & Georghiou, 2007), thus helping them in entering markets for innovation.
The conditions under which PPI can lead to innovation (Geroski, 1990;Rothwell & Zegveld, 1981;Uyarra et al., 2014) can, however, be at odds with those facilitating the access of SMEs to public tenders.
Successful PPI requires an 'intelligent' cooperation between customer and supplier (European Commission, 2005).SMEs are at a disadvantage from this point of view, since they lack an extensive network (Woldesenbet & Worthington, 2018) and experience in dealing with the public sector.The structure of the procurement process is perceived as complex (Glover, 2008;Woldesenbet & Worthington, 2018), beginning with the possibility of getting to hear of the call, which is reported as being difficult to obtain for SMEs which lack previous contracts with the public agency (European Commission, 2005).Easing SMEs' access can be obtained by the streamlining of bureaucratic procedures and the provision of information (GHK, 2010), for instance using national portals that allow public agencies to advertise tender opportunities and of streamlined pre-qualifications for use by SMEs (European Commission, 2005).However, recent experiences of using electronic solutions to carry out the procurement process have resulted in a large market share of contracts for SMEs, but contracts of low monetary value, reflecting what happens in physical markets (Albano et al., 2015).Also, while the bundling of demand and use of selection criteria privileging size or experience are procurer's risk management techniques, they might make it difficult for SMEs to participate and win (Cabral et al., 2006;European Commission, 2008).This reduces the chance of successful market entry and overall innovation, when the latter comes through young, small firms, since large bundles give an advantage to large and mature incumbents (European Commission, 2005).Possible solutions are (European Commission, 2005): i) co-sourcing, that is, awarding the contract to more than one firm, thus also maintaining technological competition (Cabral et al., 2006); ii) multiple slots, in case the presence of scale economies is dubious, so that, if they are not, SMEs have an opportunity to bid and win; iii) availability of subcontracting, which entails, however, a minor role for SMEs within the public procurement market.

The Case for PCP
Overcoming the barriers that SMEs experience in accessing public procurement is among the reasons for introducing PCP (European Commission, 2006, 2011, 2014a).Though open also to LEs, the design of PCP is similar to that of the US SBIR programme (one should also note that in the US companies with less than 500 employees are considered 'small', whereas in the EU the limit is 50 for 'small' and 250 for 'medium' enterprises-European Commission, 2009b).
PCP is a kind of procurement of R&D services by which the public procurer shares and limits the risk and the effort to pull R&D activities until the production of a prototype or first test series is achieved, without entailing the commercial procurement of products that are ready for the market.Since information improves and uncertainty declines over the course of the innovation cycle, separating the initial activities from the rest of the procurement process considerably reduces risk, allowing the public procurer to test the innovative products before ordering them on a large scale.
There are also specific reasons for the adoption of the instrument in the context of EU innovation policies.First, deregulation and the subsequent loss of the monopsonistic power of old state monopolies, especially in the telecommunications, transport and energy sectors, has weakened the demand pull.This has meant that the role that public procurers used to play as first buyers in their interactions with private suppliers of intermediate goods no longer exists (European Commission, 2006).Second, the pre-commercial nature of PCP exempts it from national treatment and non-discrimination obligations under art.XV GPA WTO and from the general prohibition of state aid under art.87 (1) of the EC Treaty, provided that the procedure guarantees competition and transparency and that risks and benefits be shared at market conditions between the public procurer and the supplier of R&D services (Note 2).
Abiding by the EU and international legal norms in force creates a model of PCP with a specific structure.Its main characteristics are: a) the separation between the R&D phase and the phase of the production of commercial products, b) the sharing of risks and benefits at market conditions, c) the competitive development in phases (Ramboll, 2008).Like the US SBIR and the UK SBRI programmes, PCP is a model of competitive funnelling.It is a single procedure, with specific contracts for each phase (phase 1, exploration of solutions; phase 2, construction of prototypes and phase 3, production of first test series) of the innovation process and two points of intermediate evaluation.At the end of phase 1 and of phase 2, the best projects are selected and go on to the next step.It is thus a phased process with multiple suppliers, who participate in a selection process at the end of which the (eventual) commercial tender might be won by firms that have not participated in the pre-commercial phase (Figure 1). on in elated rs.In a risk ening of competition in the commercial phase ensures the best value for money.These two effects are to the advantage of the public procurer, in that they limit the degree of risk that is shifted from the seller to the buyer.By the same token, however, leaving a significant part of the risk on the supply-side, this can discourage participation, especially by firms that are small, young and have no experience in public tenders.
Another problem is that the separation from the commercial phase, especially in the case of catalytic procurement, might leave a gap between the characteristics of the product obtained in the R&D phase and the quality/price ratio threshold required by the market for mass production.This is a case of innovation deadlock: suppliers having reached the level of technological 'know how' to produce innovative goods, but in a context where demand is lacking or not sufficiently articulated, so that the innovative solutions do not receive the investment necessary to be materialized into successful products.Another factor which can contribute to the deadlock is that obtaining funds is particularly difficult at this stage: public support is no longer available, as the product is beyond the research phase, and private support is not yet available as the product is still in the pre-commercial phase.A typical example of this is the above-mentioned case of energy saving technologies that encountered problems.
SMEs, on the one hand, can benefit from the reduced contract scale, but, on the other hand, their participation might be discouraged by the difficulties that remain at the point when the innovation should reach the market.In fact, they might foresee their weaker position vis-à-vis large firms in the commercial tender.
Striking a balance between the pros and cons for SMEs of procuring R&D only up to the phase of prototyping is possible if the public procurer considers the pre-commercial phase and the commercial one as distinct, but not separate.That is, the procurement process should envisage financial and non-financial support tools, following PCP and helping SMEs to reach the commercial phase, or, at least, allowing them to recover the costs incurred in the pre-commercial phase if they do not win the commercial tender.
Size-related difficulties can be summarised deriving from scarcity of funds and or human resources.Therefore, SMEs that have already participated in PCP should be given the opportunity of access to dedicated venture capital funds (Note 3) (as in the US SBIR) and to the contact network of the public procurer.Prior to this, a form of support could be to assist SMEs to create a solid business plan, a necessary prerequisite for finding financial support for the marketing of their products.

Competitive Development in Phases
There are, however, other aspects of the PCP procedure that can promote the participation of SMEs (European Commission, 2006), counterbalancing the problems and enhancing the advantages of the separation from the commercial phase.PCP is, in fact, also characterised by a competitive development in phases.
PCP covers the innovation cycle until the point when the new products are tested to ensure that they comply with the functional requirements requested by the procurers.Limiting the amount of money awarded at each phase can deter LEs from applying to PCP tenders (Apostol, 2012).The size of the tasks and the value of the contract typically increase as the project develops so enabling SMEs to prepare for large-scale production.In the perspective offered by Malerba et al. (2007), they have time to develop the product to meet the market quality threshold level and, in this way, PCP can be used to break the lock-in deriving from the presence of sophisticated buyers.Moreover, PCP allows SMEs to gradually increase the financial requirements necessary to perform the task, so that they grow as entities, alongside the innovation project that they are driving.This is a chance for SMEs to step out of their traditional subcontractor role.
The fact that development occurs in phases also means it is a competitive process.Competition and the presence of more than one firm should guarantee supply, prevent monopoly and ensure better value for money and more creativity.These are advantages in the interests of both the public procurer and society as a whole.However, competition can reduce the suppliers' incentives to participate in the procedure.There are several obstacles that can hinder the participation of a plurality of competing firms, which might explain the empirical finding that PCP has not been effective in decreasing prices with respect to other forms of procurement (Bedin et al., 2015).Some of these obstacles are, however, more relevant for the LEs context than that of SMEs or can be alleviated by SMEs' participation.First, there may be few suppliers in the relevant sectors: securing SMEs participation is thus particularly important (Yukins & Piga, 2012).Second, the funding initially available for each participant might be lower than in projects without multiple suppliers (whereas from the procurer's perspective, multiple suppliers can mean higher initial costs, but these will be recovered since the price will be lower under competitive conditions): this is more a disincentive for LEs than for SMEs, as argued above.Another disincentive may derive from the possibility that rivals gain disclosure to industrial secrets.Finally, due to the separation from the commercial phase, the winners of the pre-commercial phase do not necessarily obtain the contract for large scale deployment.This might imply that the commercial tender is won by a non-innovating firm, which reduces the incentive to participate in the pre-commercial tender.
The obstacles that may hinder the presence of a multiplicity of suppliers must be tackled with ad hoc measures, since competitive development is a crucial feature of PCP.First, each supplier should typically maintain IPRs, so that even those who do not pass to the successive phases can benefit from their investment in R&D.Second, the business plan should be reconsidered at the end of each phase, to ensure that all potential benefits are reaped.Finally, the degree of risk sharing should be different in the different phases, with the procurer endorsing most of the costs in the initial phase.It should also be noted that SMEs that have undergone competitive development and had a first customer testing validating their solutions have increased credibility (Edler & Georghiou, 2007).This in turn increases their ability to compete globally and to attract venture capitalists, as for instance, in the NHS UK case (European Commission, 2014b).

The Sharing of Risks and Benefits
Risks deriving from the separation from the commercial phase and the competitive nature of PCP must be counterbalanced by awarding producers a larger share of the benefits of the innovation process.This economic reason, besides the legal restrictions on state aid explains the shared risks -shared benefits approach, based on agreements on the assignment of IPRs.This element is an important step in the innovation cycle.In one of the most famous examples of private technology procurement, the cooperation between Toyota and Nippon Steel (Hellman, 1993;Edquist & Hommen, 1998), which led to an innovation in the steel industry in 1983, an agreement on the exploitation of research results preceded the beginning of large-scale production.PCP does not involve an exclusive use of research results by the public procurer, but these are typically shared with suppliers and other stakeholders (for instance, another public procurer).This feature is directly linked to the fact that research results are not commercial products yet.
IPRs are a key benefit of R&D projects and therefore of risk-benefit sharing.Moreover, they are strictly connected to the presence of multiple suppliers and to the other key factors that make PCP of interest to SMEs.
The problem of the assignment of IPRs can be identified with respect to its two extreme solutions: a) assigning IPRs to the supplier involves the risk of a vendor lock-in situation for the procurer, who has borne the product development costs; b) assigning the IPRs to the procurer reduces the incentive to innovate and to share risk, and so does not allow the supplier to exploit the innovation in the market.
IPRs sharing can ensure that both the public procurer's and the supplier's fundamental needs are met.The procurer should be able to freely apply the innovation, directly using it or licensing it to other producers to guarantee competition.The supplier on the other hand should be able to commercially exploit it with other customers.In PCP, IPRs are attributed either to the procurer or to the supplier in a non-exclusive way.In the first case, the public procurer allows the supplier to commercialise products, thus obtaining risk sharing, through a price reduction or a participation in the costs of managing IPRs.In the second case, as in the US SBIR, the procurer keeps the right to freely use the innovation or to license it to third parties (or to require the supplier to license it to third parties at market conditions).
The degree of sharing should depend on the contribution that each party has given to the development of the innovation and on the ability to exploit IPRs.In general, the more innovative the solution and the larger the investment required, or the higher the probability of commercial exploitation and future improvements, the more IPRs should be left with the market.Instead, the higher the risk of vendor lock-in and uncertainty, the more IPRs should be left with the public procurer (European Commission, 2007).
Of course, the participation in the tender is less attractive if fewer rights are granted to the supplier.From the public procurer's point of view, the right to use the innovation, to let other parties use it and to obtain a return for letting others use it, is granted both by entitling the public procurer with the IPRs and by leaving them with the supplier but granting the public procurer a licence.The procurer's intent to limit the use of the innovation by third parties is obtained by assigning the IPRs to the public procurer or by assigning them to the supplier but granting the public procurer an exclusive licence.
Leaving the IPRs with the supplier is essential for SMEs and can change the configuration of the competitive funnelling process to be to their advantage, on the one hand, reducing the share of risk borne by them, and, on the other hand, giving them a chance to grow globally by reselling solutions to other markets (European Commission, 2014b).
One thing needs to be stressed within this contextual framework: suppliers should be able to license the IPRs resulting from the R&D activities undertaken within the project to both other participating firms and external