Bio-based and Applied Economics 9(1): 25-52, 2020 ISSN 2280-6180 (print) © Firenze University Press ISSN 2280-6172 (online) www.fupress.com/bae Full Research Article DOI: 10.13128/bae-9936 Drivers and barriers of process innovation in the EU manufacturing food processing industry: exploring the role of energy policies Federica deMaria*, annalisa ZeZZa National Council for Agricultural Research and Economics – Research Centre for Agricultural Policies and Bio- economy, via Po, 14 - 00198 Roma, Italy Abstract. This paper investigates the driving forces that can promote or impede pro- cess innovation adoption in the European food manufacturing industry. The study uses a logit model applied to Community Innovation Survey (CIS) data containing information on innovation at the industry level for 15 EU Member States. Results suggest the relevance of many factors, internal and external to the enterprise, such as size and organization of business practices on one hand, and networking activities and cooperation agreements within the supply chain on the other hand. We also focus on energy policy variables as process innovation determinants. Energy policies imple- mentation, energy price and the availability of public funds, show a significant impact on process innovation adoption in the European food processing industry. Keywords. Environmental Regulation, Energy policy, Innovation, Cooperation activities, Food manufacturing industry, EU. JEL codes. L66, O3, Q4, Q16. 1. Introduction In October 2014 the European Council agreed on a new 2030 Framework for climate and energy, including EU-wide targets and policy objectives. This strategy aims to help the EU achieve a more competitive, secure and sustainable energy system and to meet its long-term 2050 greenhouse gas reductions target. The figures for renewables and energy efficiency have subsequently been increased in the Targets for 2030 context including, among others, a 40% cut in greenhouse gas emissions compared to 1990 levels and an indicative target for an improvement of at least 32.5% energy efficiency at EU level. The food processing industry makes a significant and increasing contribution to the overall energy consumption and GHGs emitted in the food chain (OECD, 2017). Eurostat considers the total energy consumed by the food processing industry (including bever- *Corresponding author. E-mail: federica.demaria@crea.gov.it Editor: Valentina Raimondi. 26 Federica DeMaria, Annalisa Zezza ages and tobacco) to be 29178.5 Mtoe in 2015. This amount represents 28% of the energy embedded in the food chain (Monforti- Ferrario et al. 2015), 2.7% of the average final energy consumption and the 10% of the average energy consumption in the manufactur- ing industry, with Denmark and Croatia having the larger shares. Although the European food processing and beverage sector has gradually improved its energy efficiency – meas- ured as the ratio value added/energy consumption - gaining competitiveness and reducing GHGs emissions in the last decades, still it has a strong potential for decreasing energy consumption both as a result of process optimization and plant system improvements (OECD, 2017; Ladha-Sabur et al, 2019). Technical potential energy savings have been esti- mated to be 22% (compared to 2004) by 2030 (Altmann et al, 2010; ICF, 2015). Energy use reduction and energy recovery from waste are two important methods to reduce pro- duction costs in the food processing industry (OECD, 2017; Altmann et al. 2010; Mon- forti- Ferrario et al. 2015). Kaminski and Leduc (2010) have identified the most important systems and processes where significant energy-efficiency improvements can be achieved in the EU’s food industry: steam, motor and pump, compressed air systems, process cool- ing and refrigeration, and buildings heating and lighting. The EU Lisbon Strategy considers innovation as one of the most important factors to enhance productivity, competitiveness, and sustainability in the economy. Literature, starting from Schumpeter’s studies (1934, 1942) has tried to understand what the internal and external factors influencing process innovation are (Cohen, 1996; Galende and de la Fuente, 2003). Many factors determine the firm’s capacity to innovate, ranging from tech- nical capabilities, financial structure, market needs, network relationships, regulations, and subsidies. Enterprises, however, can also be deterred from engaging in innovation or fail to bring in new processes or products because existing barriers. Recognizing the nature of these barriers is important both from a policy and a management point of view (D’Este et al. 2012). Literature on innovation has focused mainly been on high tech industries and only a few studies have considered low-tech traditional sectors such as the food processing industry. Some specificity characterizes innovation in the food processing industry. Firstly, food firms are mainly process innovation oriented (Minarelli et al., 2014). Secondly, the production of new technologies is usually developed by upstream industries and R&D expenditure is low compared to other sectors (Garcia-Martinez and Briz, 2000; Triguero et al., 2013). Thirdly, most innovations are incremental rather than radical (Garcia-Martinez and Briz, 2000). Fourthly, product typologies and process phases are extremely differenti- ated and difficult to uniform (Capitanio et al., 2010). Finally, SMEs are characteristical- ly prevalent in the agri-food sector. Literature shows that SMEs behave differently with respect to large enterprises relative to innovation adoption, with larger firms being more innovative than smaller ones (Galizzi and Venturini, 1996). This distinction has impor- tant policy implications for policy design. Recent research (Minarelli et al., 2014) indicates that, in the EU, SMEs are a very heterogeneous group regarding their innovation profile, particularly in the food sector. The relation between size and innovation is not straight- forward, as other factors influence firms’ behaviour as workforce in-house capabilities and the engagement in R&D activities (Avermaete et al. 2004). The small average size is con- sidered one of the main barriers to innovation in Southern European countries (Garcia- 27Drivers and barriers of process innovation Martinez and Briz, 2000; Capitanio et al. 2010). Nevertheless, the food processing industry is becoming more technologically inten- sive to maintain better process control, exploit economies of scale and guarantee food safety and quality, facing market competition through new products and processes devel- opment. Capitanio et al. (2009) analysing the driving factors of innovation in the Italian food processing industry conclude that the determinants can be found both in internal and external factors. Among the first, there is the human capital qualification, a clearer ori- entation to quality products, organizational changes, and relation capacity development. On the external side, factors such as the increasing competition and demand have a rel- evant role in shaping the innovation process. Different determinants emerge when ana- lysing process and products innovation separately. In the first case, the of human capital and market relationships qualities are the most important drivers, while in the of pro- cess innovation case, the firm’s financial structure and capital intensity are the most rel- evant. The firm’s location, meaning socioeconomic context relevance, has a positive and significant impact on process or product innovation adoption. With regard to Spain, research has underlined the machinery and equipment suppliers contribution to innova- tion diffusion (Garcia Martinez and Briz, 2000), together with significant path depend- ence (Triguero et al., 2013). Environmental and market factors role seems higher in the food processing industry compared to other manufacturing industries. Cleaner technolo- gies1 can improve the production process efficiency by reducing materials and energy con- sumption, improving firm’s productivity and competitiveness (Del Rio, 2005; Frondel et al., 2007; Arimura et al., 2007). Regarding the adoption of cleaner processes in the Span- ish food industry, such as material recycling and water management processes, Cuerva et al. (2014) find that firm organizational capabilities - i.e. the implementation of an envi- ronmental management system - are an important driver, while no significant relationship with public support is found. The same result is confirmed in other studies on France and Germany (Belin et. al. 2011) and for the EU-27 (Triguero et al. 2013). Small food firms contribute substantially to the food processing industry economic performance and are considered to play a key role in achieving sustainable economic growth in local economies. In the EU, the food and drink sector turnover, in 2017 was 1,192 billion with 294,000 firms. The food processing industry is dealing with several challenges related to the sus- tainability productive processes, being a large surface water and energy user. However, as previously mentioned, it has a strong potential of decreasing energy consumption both a result of process optimization and plant system improvements (OECD, 2017; Ladha-Sabur et al, 2019). On the same vein, we focus on the food processing industry investigating for factors that can statistically be associated with the firm’s innovation process. We study the adop- tion of process innovation in the food processing industry in the EU, considering the role of energy prices and policies2 role. According to the CIS 9 survey, 30% of firms in the 1 Process innovations are usually grouped in clean technologies and end-of-pipe technologies (Kemp and Volpi 2008; Rave et al, 2011). 2 The main limitation of this study concerns the fact that it does not exploit the endogeneity issues; the panel is too short and the use a lagged variable may overfit the endogenous lagged dependent variable- 28 Federica DeMaria, Annalisa Zezza F&B industry that have innovated have reduced their energy consumption while 20% have contributed to reduce energy use or the CO2 “footprint” during the consumption or use of a good or service by the end-users. Financial measures count for half of the policies addressing energy efficiency in industry. The Odyssee- MURE project shows that in many countries the policies in place include a broad mix such as co-operative measures (e.g. agreements among enterprises on energy efficiency), cross-cutting measures with sector-specific characteristics (e.g. indus- try eco-tax with reduced rates ); fiscal/tariff measures (e.g. tax deduction for energy sav- ing investments in businesses); information/education/training (e.g. advice programs for industry, energy management systems); legislative/informative (e.g. Mandatory execution of energy audits in large enterprises); legislative/normative (for e.g. CO2 emission fee for large emitters; new market-based instruments. About 10% of overall measures have been introduced under the Energy Efficiency Directive (2012/27/EU), especially measures introduced under Article 7 (energy efficiency obligations and/or alternative measures), mandatory audits (Article 8) and new certification/qualification schemes. Nevertheless, most energy measures are not EU-related but national measures, particularly those rated with a high impact (Odyssee-Mure, 2015). The main questions addressed in this paper are: i) Are networking and cooperation activities between research institutions and food companies relevant in generating and promoting process innovation? ii) Is the high energy price a driver for boosting process innovation? iii) Does environmental regulation together with policy stringency stimulate the process innovation introduction? iv) If yes, what is the most effective driver? The paper is structured as follows. Section 2 presents the hypothesis that will be tested in the model. Section 3 introduces the dataset mainly based on the Eurostat Community Innovation Survey and the empirical model. Section 4 reports the econometric estimates. Section 5 shows the main conclusions. 2. Conceptual background and hypotheses The empirical literature on the determinants of innovations focuses on four groups of factors impacting production and innovation adoption: technological capabilities, organizational capabilities, market pull and external influences related to the regula- tory push/pull and potential existence of networks (Cuerva et al., 2014, Horbach, 2008, Wagner, 2009). Technological capabilities refer to knowledge resources, human skills and access to internal or external funds and are common drivers for all kinds of innovation. Organizational capabilities have a strong impact on green innovations; for example, the quality management systems adoption is often linked to the implementation of environ- mental management systems (Mazzanti et al. 2008). Market pull factors relate to con- sumers’ preferences or customers’ requirements for new products as well as the search for new niche markets. Among external innovation sources, literature considers the regula- tory push/pull effect as very relevant (Rennings, 2000). Regulations have been found to be significantly more important for environmental innovation compared to other innova- tion (Horbach et al. 2012). At the same time, government policies play a role in inducing the creation of new cleaner technologies and also in the adoption of already existing technologies by firms 29Drivers and barriers of process innovation (Veugelers, 2012). The EU growth strategy Europe-2020 seeks to booster innovation and collaboration across actors in the supply and innovation chains and private companies, and to strengthen cooperation among research institutions and firms, in addition to pro- moting more effective and efficient public financial support for innovation activities. Fur- thermore, the EC Green Deal Communication notes the role of new technologies in pro- viding additional benefits in the transition to a sustainable economy. A recent OECD (2017) study on energy efficiency in the agri-food sector identifies four broad groups of barriers: structural, behavioral, availability and policy. Structural barriers encompass issues such as limited know-how on implementing energy-efficiency measures, or fragmented and under-developed supply chains. Such barriers prevent an end-user from adopting an energy-efficient technology or practice: for example, low edu- cational attainment and ageing farmers impede the adoption of new potentially energy- efficient technologies. Behavioral barriers include situations in which limited awareness or end-user inertia inhibit an opportunity pursuit. Inertia represents the resistance to change and risk, and the more radical the change, the higher the barrier will be (Sorrel et al. 2000). It can lead to preferring interventions with quick and low investments and returns, thus slightly modifying the production system with short pay-back criteria may be explained by risk aversion (Jaffe and Stavins, 1994). An unfavorable perception or treatment of risks can inflate energy-efficiency projects costs or lead to the underestima- tion of risks associated with changes in energy prices. Uncertainty about energy prices can also limit energy efficiency measures because of higher perceived risks. Risks management associated with energy costs and availability in agri-food businesses are largely determined by business size, with larger businesses being more likely to be proactive in managing risks from volatility in energy and commodity prices (OECD, 2017). Availability barriers include situations in which the decision-maker is interested in and willing to innovate, but barriers, for example, a lack of capital access might prevent an upgrade to a new heating system or the availability and diffusion of technology and inno- vations (OECD, 2017). Policy barriers are policy-induced market distortions which result in market condi- tions hindering energy efficiency. For example, energy subsidies can crowd-out public spending and private investment, encourage excessive energy consumption, reduce incen- tives for investment in renewable energy, and accelerate the depletion of natural resources (McKinsey and Company, 2010) i.e. encouraging more fossil fuels or energy usage inten- sive production . Cagno et al. (2013) find that the major perceived barrier for Italian manufacturing SMEs in the food processing industry, regarding the adoption of energy efficiency tech- nologies, is represented by high investment costs. Same or also insufficient profitability and low capital availability. The identification of barriers to innovation is crucial for effective policy design. According to the Eurobarometer survey of SMEs in the EU (Eurobarometer, 2016), most common barriers are represented by uncertain market demand and returns. Other causes are the lack of funds or qualified personnel and in general, low technological capabilities. This barriers typology is expected to be more pervasive for SMEs, particularly in sectors with non–energy-intensive production processes (Fleiter et al., 2012, Trianni et al., 2013) such as the food processing industry. 30 Federica DeMaria, Annalisa Zezza Hence, we explore the following research hypotheses: H1a) Networking and cooperation activities between research institutions and food compa- nies are positively associated with process innovation. Research shows that networking and cooperation effects are unclear, indeed some SMEs benefit from positive effects from cooperation to achieve innovations (De Jong and Vermeulen 2006; Van Gils and Zwart, 2004; Batterink et al., 2010; Omta, 2002); while others experience problems (Hoffmann and Schlosser, 2001; Caputo et al. 2002; Kauf- mann and Todtling, 2002). The importance of cooperation has risen steadily alongside the complexity, risk and cost of innovation activities. Innovation cooperation influences innovation activities through the pattern of collaborative relationships and partner type involved (Vinding 2002). This relationship is mutually reinforcing - external linkages facil- itate innovation, and at the same time innovative outputs attract further collaborative ties (Powell and Grodal 2005). Companies that continuously cooperate with different exter- nal subjects such as suppliers, customers, competitors, and research organization improve both knowledge sharing and market knowledge acquisition by the firms, resulting in expansion of the firm’s existing knowledge base, which in turn advances a firm’s innova- tion capability (Zhou & Li, 2012). Such collaboration has been identified in literature as one of the most important external predictors of innovations (Alexiev et al., 2016; Clauss and Kesting, 2017; Heirati et al., 2016). In addition, a company may establish collabora- tion with other business partners, such as technology providers and researchers (Bigliardi and Galati, 2013). In fact, through networking, a company can extend its range of skills by an effective contractual arrangement (Martino and Polinori, 2011). Vertical cooperation offers more possibilities for innovation SMEs because cooperation is often used to acquire external know-how, in particular where firms have neither R&D employees nor the special technical requirements necessary to engage in R&D activities (Gellynck et al., 2007; Gel- lynck and Khüne, 2010; Laperche and Liu, 2013). Collaborative innovation networks are defined when members participate in new product development and innovation processes (Alexiev et al., 2016; Möller & Halinen, 2017). The role of firm network relationships and internationalization has been investigated by Cainelli et al. (2011) for Italian firms, finding to have a strong effect on the environmental innovations adoption by internationalized firms while being less clear for locally oriented firms. Yet, scholars show that when firms cooperate with universities or research institutes, the overall effects on innovation capacity is positive (Hájek and Stejskal, 2018). Research also demonstrates that participating in cooperation networks makes companies more prone to undertaking sustainable oriented innovation (Melano-Levado, 2020; Klewitz and Hansen, 2013). Resorting to cooperation agreements (e.g. Cainelli et al. 2012; DeMarchi 2012; Del Río et al. 2013) and external knowledge sourcing (e.g. Del Río et al. 2013; Ghisetti et al. 2015) are thus particularly important and “complement” investments made in organizational and technological capabilities (e.g. Horbach 2008; Demirel and Kesidou 2011; Horbach et al. 2012). H1b) Information is positively associated with a process innovation. This hypothesis follows the previous one. Even in this case, several studies reveal that access to information facilitates the use of scientific knowledge, enhancing innovation and 31Drivers and barriers of process innovation increasing the food processing industry competitiveness (Ciliberti et al. 2016). However, firm size matters in this regard, where small companies rely on universities or research institutes while larger enterprises might have the capabilities needed to put new ideas into practice (King et al, 2003; Ciliberti et al. 2016). Lasagni (2012) suggested that innovation performance in SMEs can be higher when they strongly collaborated with users, custom- ers and suppliers. His results also showed that SMEs can be more successful in product development when they closely work with research institutes. This suggests that there can be specific types of partners preferred by SMEs. Gomez et al (2016) examine a panel of manufacturing firms in Spain to verify the extent to which the use of internal and exter- nal sources of information generate product and process innovation. Their results show that, although internal sources are influential, external sources of information are key to achieve innovation performance. Furthermore, the importance of external sources of information varies depending on the type of innovation (product or process) considered. To generate process innovation, firms mainly rely on suppliers while, to generate product innovation, the main contribution is from customers. H2a) The higher energy price stimulates process innovation. A relatively higher energy price in a country, as a result of country’s energetic struc- ture and energy taxation, will induce a technological change heading to higher energy efficiency. Ghisetti and Rennings (2014) found that innovation leading toward reductions in the use of energy or materials per unit output positively affect the firm’s competitive- ness, while externality-reducing innovations hamper the firm’s competitiveness. Cainelli & Mazzanti (2013) find that policies targeting the manufacturing sector are likely to induce innovation adoption in the services sector, especially when considering innovation prac- tices aimed at abating CO2 emissions and improving energy efficiency. Yet, the study of Popp (2002), on the standard inducement mechanism, confirms that both energy prices and the quality of knowledge exert a significant and positive effect on patenting. H2b) Environmental Policy Measures stimulate firms’ process innovation adoption. The relevant contribution of Porter and Van der Linde (1995b) has paid attention to “Porter hypothesis”, according to which a good environmental innovation can lead to an increase in firms’ performance, for instance through a reduction in energy or materials use. However, since firms are not always aware of the opportunities from eco-innovation, a strict and effective environmental regulation is required in increasing this awareness. Therefore, environmental policy seems to be an important eco-innovation driver and deserves specific attention. According to Porter and van der Linde (1995a), environmental standards can foster innovation but under three well established conditions. Firstly, they must create maxi- mum opportunities for eco-innovation, letting the industry choose its own approach to innovation. Secondly, regulation should foster continuous improvement in any technology. Thirdly, the regulatory process should not leave uncertainty at every stage of implementa- tion. The type of regulation or policy and the way in which it is implemented is impor- tant. It could lead firms to effectively address environmental problems. The stringency of the policy and the terms in which it is defined are equally important, since uncertainty depends on these factors. Several empirical studies (ZEW, 2001; Rehfeld at al. 2006; Reid 32 Federica DeMaria, Annalisa Zezza et al. 2008; Belin at al. 2009) find a positive correlation between innovation and regula- tions. Porter and van der Linde, (1995a/b), Kemp et al. (2001) and Jänicke et al. (2002) show that strict environmental regulations stimulate innovation in several ways, such as advantages created by the development of “green” technologies. However, firms are not able to recognize the environmental innovation cost saving potentials as in the case of energy or materials savings (Horbach and Rennings, 2007). This leads many of them to believe that an environmentally virtuous behavior is a burden rather than an asset (Kemp and Andersen, 2004). Therefore, regulations and policies can be a catalyst and help them to understand the potential benefits of environmental innovations. Popp (2009) argues that in general, market-based policies are thought to provide greater incentives for inno- vation, as they provide rewards for continuous improvement in environmental quality. Conversely, command-and-control policies penalize polluters who do not meet the stand- ard, but do not reward those who do better than mandated as the command-and-control regulations direct a specific level of performance. 3. Dataset, Variable and estimation methods 3.1 Data The dataset used in the analysis is based on the biennial CIS surveys carried out from 2010 to 2014 (CIS 8 and CIS 9). The CIS questionnaire addresses several elements of firms such as size, i.e. firm’s size, turnover, employees, cooperation activities, source of informa- tion, public financial supports, innovation expenditures, and innovation activities. The CIS survey provides information on sectors innovativeness, different types of innovations and various aspects of innovation development. The survey allows to distin- guish firms that can easily be categorized into innovating and non-innovating. Table 1 reports data on process innovation implemented by enterprises across the EU-28 between 2012 and 2014. The highest proportion of enterprises that have devel- oped process innovation is observed for Belgium (46,8%), Netherlands (33,4%), Portugal (37,2%) and Lithuania (42,3%) in 2012/2014, while rates are lower for Bulgaria, Hungary, Romania and Slovakia , ranging from 5% to 9%. Our panel (CIS 8 and CIS 9) includes all the in the firms belonging to the food manu- facturing industry (NACE Code C10-12). For data availability reasons, we have restrict- ed the analysis to the following EU countries: Bulgaria, Cyprus, Croatia, Czech Republic, Germany, Estonia, Hungary, Italy, Lithuania, Portugal, Spain, Romania, Slovenia, Slovakia and Norway. After removing missing value, the sample contains 4618 observations which are used in the analysis. Data on energy policy come from the MURE project (Mesures d’Utilisation Ration- nelle de l’Energie)3 which provides information on energy efficiency policies and their impact assessment in EU countries. The score classifies the EU member states based on scoring energy efficiency poli- cies and trends. It aims to provide comparison indicators and comparable characteristics helping countries to understand whether their policies are comparable or better than in 3 https://www.odyssee-mure.eu 33Drivers and barriers of process innovation other countries or whether they can learn from other countries to improve their policies. It ranges between 0 and 5, with 0 meaning “worst” and 5 “best”. Countries with a lower score are Cyprus, Hungary and Croatia; conversely, Spain, Norway and Slovenia reported the highest score. The energy price yearly data come from Eurostat database available at the following link: “https://ec.europa.eu/eurostat/web/energy/data/database”. Prices are provided without taxes, with VAT and with all taxes included. A detailed explanation of the variables definition and measurement is reported in appendix (Table A.1). 3.2 Model and estimation The choice to adopt a process innovation is represented by a binary logit model where the dependent variable (process innovation adoption hereafter proc_inno_adop) is a bina- ry variable (yes=1/no=0) based on the response – at the firm level - on the introduction of innovations in the previous three years. Let y the dependent variable observed and the latent variable satisfying the single index model (1) Even if is not observed, we do observe Table 1. Enterprises in the food processing industry that have introduced process innovation. 2010/2012 2012/2014 2010/2012 2012/2014 Belgium 35.5 46.8 Lithuania* 18.4 42.3 Bulgaria* 9.6 9.2 Luxembourg 43.8 22.6 Czechia* 27.2 25.3 Hungary* 6.1 7.4 Denmark 26.8 19.0 Malta : 29.8 Germany* 22.3 17.4 Netherlands 27.0 33.4 Estonia* 37.3 21.9 Austria 20.0 26.0 Ireland : 51.7 Poland 8.1 8.5 Greece 29.0 30.3 Portugal* 35.8 37.2 Spain* 20.0 18.5 Romania* 5.1 5.1 France 25.0 26.3 Slovenia* 23.1 28.7 Croatia* 15.2 21.8 Slovakia* 9.1 17.3 Italy* 32.6 31.7 Finland 38.4 30.1 Cyprus* 26.9 : United Kingdom 17.0 23.6 Latvia 21.2 14.3 The * symbol is indicating the countries used in the analysis. Source: Eurostat, Community Innovation Survey 34 Federica DeMaria, Annalisa Zezza (2) (0 for non-innovative firms and 1 for the otherwise). From 1 and 2 we have: (3) Where F(x’β) is the cumulative distribution function of -εj. The logit model specifies that cumulative standard logistic is: (4) and the marginal effect is: (5) Thus, we estimate the following equation: proc_inn_adopit = α + β1 ln(turnoverit) + β2fundsit + β3lmarketit + β4rmacit + β5gnewmktit + β6orgbupit + β7coit + β8int_info_sourcesit + β9other_info_sourcesit + β10policy_mit + β11epit + fei + fet + εit (6) where i denotes countries, t = 2010, 2012 and fei and fet represent country and time fixed effects respectively. The dependent variable (proc_inn_adop) is a dummy that has been built using three indicators of the CIS: a) INPSD which considers the introduction onto the market of a new or significantly improved production method ; b) INPSLG which considers a new or significantly improved logistic, delivery or distribution system; c) INPSSU that considers the introduction onto the market of a new or significantly improved supporting activities. It takes a value of 1 if a new or significantly improved method of process or distribution has been introduced, 0 otherwise. We have included the following control variables chosen on the basis of their rele- vance for firm characteristics and strategies: - lnturnover, measured as the natural logarithm of the turnover. Literature has found that size affects the propensity to innovate, emphasizing the difficulties for small and medium enterprises. Indeed, the small average size is considered one of the main bar- riers to innovation in Southern European countries (Garcia-Martinez and Briz, 2000; Capitanio et al. 2010). Yet, scholars find that farm size has a positive, albeit small, 35Drivers and barriers of process innovation effect on innovation, which is in line with the general innovation adoption litera- ture (Lapple et al., 2015; Feder et al., 1985; Sauer and Zilberman, 2012). Moreover, as highlighted in the Diederen et al. (2003) study, agricultural farm size explains differ- ences in adoption. Similarly, the work of Hashi and Stojcic (2012) show that larger firms are more likely to embark, to invest on innovation activities but with decreasing innovation output depending on the firm’s size; - funds reflecting the availability of public support to innovation. It takes a value of 1 if enterprises have benefited from public (Regional, National, European) support to innovation and 0 otherwise. Marzucchi and Montresor (2017) note that public fund- ing for innovation generally increases the innovation adoption and environmental innovation particularly. Hyytinen and Toivanen (2003) analyze the effects of public policy, measured by government funding, on the behavior of privately owned, small and medium sized enterprises (SMEs) in Finland. Their findings pointed out that gov- ernment funding disproportionately helps innovation and growth of firms in indus- tries that are dependent on external finance; - lmarket representing firm’s prevalent market. It is a dummy variable equal to 1 if firms operate in the EU/international market and 0 for national/regional market. Regarding the access to foreign markets, literature has pointed out that international has been associated with successful innovation development (Oliveira and Carvalho, 2010; Salavisa et al., 2012; de Faria et al., 2010; Romijn and Albaladejo, 2002); - orgbup representing organizational practices. It includes new business practices or new method of organizing work responsibilities and external relationships. A part of the literature supports the view that having a structured organization is important in achieving innovation. Laursen and Foss (2003), find that interdisciplinary teams, quality circles4, employees’ proposals collection system, planned job rotation, delega- tion of responsibility, integration of functions and performance related significantly lead to innovation. O’Connor and DeMartino (2006) agree that organizational struc- ture and incentive systems are key elements in the innovation success. Prester and Bozac (2012) in analyzing companies over 20 employees on the European Manufac- turing Innovation Survey (EMIS) in Croatia report findings similar to Laursen and Foss (2003). Therefore, organizational practices have impact in achieving innovation. We test our hypothesis by considering the following variables: - co is the variable cooperation agreements which includes active participation among companies or institutions on innovation activities. The aim of any cooperation agreement is that of introducing external knowledge to the firms. Studies show an uncleared networking and cooperation effect. Some SMEs benefit from positive coop- eration effects to achieve innovations (De Jong and Vermeulen 2006; Van Gils and Zwart, 2004; Batterink et al., 2010; Omta, 2002); while others experience problems (Hoffmann and Schlosser, 2001; Caputo et al. 2002; Kaufmann and Todtling, 2002); - int_info_sources considers the internal sources of information; it is equal to 1 if its CIS score is more than 2 and 0 otherwise. Scholars find that innovations are developed by using knowledge from a diverse set of internal and external sources of information 4 Jones et al. (2008) define quality circles in a following manner: “The company uses quality circles, defined as regular meetings between employees where they discuss issues related to immediate job tasks and make sugges- tions to improve production processes”. 36 Federica DeMaria, Annalisa Zezza (Gomez et al., 2016; Amara and Landry, 2005). Furthermore, the influence of each source is different depending on the innovation type. Internal sources and suppliers are the main contributors in the case of process innovation (Gomez et al. 2016); - other_info_sources takes into account external sources of information. It considers information from suppliers, competitors, consultant or from other sources as scien- tific journal, which allows firms to generate new ideas and developing innovations by merging this kind of information with their internal ones (Lefebvre et al., 2015; Lee et al., 2010). This variable considers all the information sources other than internal one. It takes a value of 1 if all the information sources report a CIS score more than 2 and 0 otherwise; - policy_m represents the score attributed to policies and measures at national level in terms of success in achieving energy efficiency in the industry end-use sectors (see data description). Regulations and policies can be a catalyst and help firms to under- stand the potential benefits of environmental innovations. - ep refers to yearly energy price data. Energy prices are considered together with the policy measures. We include prices with taxes in the model. A relatively higher energy price in a country will induce a technological change in favor of higher energy effi- ciency. (Ghisetti and Rennings, 2014; Cainelli & Mazzanti, 2013; Popp, 2002). Descriptive statistics on selected variables used in the estimated model are shown in appendix (Table A.2). About 64% of the considered firms have introduced process innovations. Data show the importance of internal sources of information (68%) and the machinery and equip- ment acquisition (52%). At the same time, most of the enterprises in the sample operate in the EU markets (89%). Only 26% of the firms in the sample are engaged in cooperation activities or received public funds for innovation (26%). Energy price (with tax) has high variability in the EU countries ranging from a minimum value of 0.07 and a maximum value of 0.25. Finally, the successful policies in energy saving show high variability rang- ing between 0 and 4.38. The correlation matrix is reported in table 2. Correlations are moderate implying that there is a low collinearity risk issues and redundancies in this set of variables. All the con- trol variables are positively correlated with the dependent variable except for “other infor- mation sources” and “energy price”. 4. Results and Discussion We have run three different models with the scope of investigating the energy policy variables effect on process innovation, (Table 3). The first one considers the main innova- tion process drivers and barriers and the successful policy measures (policy_m); the sec- ond one takes into account the effect of energy policy variables (ep); while the third one considers the two policy variables effects (ep and policy_m). Most of the hypotheses are confirmed by the results with models 1, 2 and 3, satisfying all the tests. The discussion below concerns model 3 in table 3. The variable lnturnover displays a positive and significant coefficient (+0.1288). This result is different from the study of Garcia-Martinez and Briz (2000) or Capitanio et al., (2010), where the small average size hamper innovation. As regards to the other control variables we find a positive and sig- 37Drivers and barriers of process innovation Ta bl e 2. C or re la tio n M at rix . Pr oc es s in no va tio n ad op tio n (in no _p ro c_ ad op ) In te rn al in fo rm at io n so ur ce s ( in t_ in fo _s s) O th er in fo rm at io n so ur ce s (o th er _i nf o_ ss ) A cq ui sit io n m ac hi na ry (r m ac ) Fu nd s C oo pe ra tio n (c o) Bu sin es s or ga ni za tio n (o rg bu p) Tu rn ov er (ln tu rn ov er ) Po lic y m ea su re s (p m ) En er gy p ric e (e p) Pr oc es s in no va tio n ad op tio n (in no _ pr oc _a do p) 1 In te rn al in fo rm at io n so ur ce s ( in t_ in fo _s s) 0. 13 75 * 1 O th er in fo rm at io n so ur ce s ( ot he r_ in fo _s s) -0 .1 36 3* 0. 24 17 * 1 A cq ui sit io n m ac hi na ry (r m ac ) 0. 31 02 * 0. 09 28 * -0 .0 48 4* 1 Fu nd s 0. 16 77 * 0. 07 89 * -0 .0 15 8 0. 12 84 * 1 C oo pe ra tio n (c o) 0. 18 41 * 0. 19 31 * 0. 04 24 * 0. 10 05 * 0. 24 66 * 1 Bu sin es s or ga ni za tio n (o rg bu p) 0. 25 85 * 0. 15 57 * -0 .0 26 5* 0. 13 69 * 0. 10 21 * 0. 19 96 * 1 Tu rn ov er (ln tu rn ov er ) 0. 20 36 * 0. 07 46 * -0 .2 04 0* 0. 03 57 * 0. 12 69 * 0. 26 78 * 0. 21 34 * 1 Po lic y m ea su re s (p m ) 0. 11 19 * 0. 01 51 -0 .0 18 5 -0 .2 21 5* 0. 00 85 -0 .0 12 4 0. 05 83 * 0. 06 39 * 1 En er gy p ric e (e p) -0 .1 53 4* -0 .0 47 9* 0. 20 90 * -0 .0 90 1* 0. 01 88 -0 .0 63 9* -0 .0 35 4* 0. 03 80 * -0 .0 54 1* 1 38 Federica DeMaria, Annalisa Zezza nificant influence in fostering process innovation adoption for machinery acquisition (rmac), business practices organization (orgbup) and public funds (funds). As regards to the machinery acquisition (rmac +1.8995), our result is in line with the main literature which shows that machinery acquisition foster process innovation adoption. (Ciliberti et al 2016). Business practices organization reports a positive and statistically significant coefficient (orgbup +0.8891). This result is supported by the following studies Laursen and Foss (2003); O’Connor and DeMartino (2006); Prester and Bozac (2012); while Silva et al. (2008) find a negative relationship between the propensity to innovate and the organi- zational rigidities. Public financial support for innovation seems to be a process innova- tion driver (+0.2613) with a positive and statistically significant coefficient. The opening of new markets (gomkt) and market localization (lmarket) report insignificant coefficient. Focusing the discussion on the first two hypothesis, results reveal that the presence of cooperation agreements (co +0.3929) and networking activities are positively associated with innovation process. As literature points out, these activities facilitate learning about new opportunities and can improve market access and economies of scale and scope (de Faria et al. 2010; Cassiman and Veugelers, 2002; Lopez, 2008). Quantitative empirical studies on external knowledge sourcing provide evidence that involving a large number of external sources of knowledge in innovation is a promising choice for large firms (Lakha- ni et al. 2006; Laursen and Salter 2006). Information sources reveal a different pathway depending on their nature. Our analy- sis suggests a relevant role for the internal sources (int_info_sources +0.2622); while the others (other_info_sources -0.5494) are negatively correlated with the process innova- tion. The impact of various information sources is not straightforward as their use can be public and private – universities, journals, conferences and suppliers among many oth- ers - which may generate costs that must be considered. In some cases, the over-search of external sources may take too much time and slow down the innovation process. Addi- tionally, excessive reliance on external information sources can increase coordination and monitoring costs and could affect the creation of knowledge stocks within the firm. Com- paring results with existing literature, we find that the information sources affect the gen- eration process innovation as in Ciliberti et al 2016. However, this finding is dissimilar to other empirical evidence which shows that firms should always look for external informa- tion which can then be embodied into innovation (Köhler et al. 2012; Costa et al. 2015). In some cases in acquiring external information, companies demonstrate openness and ability to scan the market and identify opportunities which allow them to be more effi- cient in implementing innovation and decrease the risk of product failure (Stewart-Knox and Mitchell 2003; Avermaete et al. 2004; Wei and Wang 2011). As underlines by Tether and Tajar 2008; Lee et al. 2010, diverse information sources (from suppliers, competitors, consultants) are complementary and, if merged with the existing knowledge, allow to cre- ate new knowledge useful for innovation. The third and fourth hypothesis have tested through the impact of the implementa- tion at the country level of energy saving policy measures (policy_m). The positive and statistically significant coefficient of policy_m (+3.3934) indicates that energy policies adopted by EU countries boost innovation. Firms innovated, i.e. adopting new or mak- ing changes to the organization of the productive process. Very similar is the energy price effect (ep +7.8284), which clearly shows that a high energy price is an incentive to modify 39Drivers and barriers of process innovation the productive process. Our results are also consistent with what is found in Rennings and Rammer (2009) revealing that process innovations are more strongly aimed at cost reduc- tion, since increasing energy and/or material efficiency are associated with lower costs per unit. On the same vein, Popp (2002) exposes the strong and positive impact energy prices on new innovations. Similarly, Rennings et al. (2008), Del Rio Gonzalez (2005) find that regulation pressure and corporate image are the main drivers for adopting green technolo- gies in the Spanish pulp and paper industry. Table 4 merges the results of the model with the underlying hypothesis, showing that the model performs reasonably well, with energy policy, cooperation activities, financial Table 3. Logit regression result from panel data. (1) Proc_inn_adoption (2) Proc_inn_adoption (3) Proc_inn_adoption Proc_inn_adoption lnturnover 0.1274*** 0.1288*** 0.1288*** (0.0243) (0.0244) (0.0244) funds 0.2605** 0.2613** 0.2613** (0.0994) (0.0992) (0.0992) lmarket -0.0088 -0.0087 -0.0087 (0.0865) (0.0865) (0.0865) rmac 1.8986*** 1.8995*** 1.8995*** (0.0903) (0.0903) (0.0903) gnewmkt 0.0599 0.0517 0.0517 (0.0376) (0.0381) (0.0381) co 0.3934*** 0.3929*** 0.3929*** (0.1073) (0.1073) (0.1073) orgbup 0.8888*** 0.8891*** 0.8891*** (0.0919) (0.0918) (0.0918) Inte_info_sources 0.2620** 0.2622** 0.2622** (0.0896) (0.0896) (0.0896) Other_info_sources -0.5122*** -0.5494*** -0.5494*** (0.1208) (0.1237) (0.1237) policy_m 1.9569** 3.2951** (0.8876) (1.133) ep 7.8284* 7.8284* (4.5457) (4.5457) N 4645 4645 4645 adj. R2 0.41 0.40 0.40 chi2 1191.8519 1196.1113 1196.1113 BIC 4559.1268 4564.8505 4564.8505 VIF 1.20 1.21 1.21 Time dummies and country dummies are included in the model. Standard errors in parentheses. Vari- able statistically significant at * p < 0.05, ** p < 0.01, *** p < 0.001 respectively. 40 Federica DeMaria, Annalisa Zezza public supports, and technical capabilities play an important role in creating a favourable environment for processes innovation. Table 4. Merging Hypothesis and results. Hypothesis Results H1a) Networks and cooperation activities promote process innovation YES ⇒ + and significant H1b) Sources of information stimulate process innovation Internal ⇒ YES ⇒ + and significant Other ⇒ YES ⇒ - and significant H2a) Successful policies measures stimulates firm to introduce process innovation YES ⇒ + and significant H2b) Energy price is positively associated with process innovation YES ⇒ + and significant 4.1 Robustness Check In order to test the robustness of our results, further elaborations are provided (Table 5). Results were confirmed when we ran the logit model considering the original variables of the CIS survey instead of the transformed variables used in section 3. Again, coopera- tion, information sources, policy variables and energy price, are positively associated with process innovation. Results were also confirmed when we used variables acting as barriers instead of drivers. The estimated models are consistent; indeed, the impact of most explanatory variables is statistically significant and different from zero. Results of the confusion matrix5, which describes how many actual and predicted values exist for different classes predicted by the model, indicate that the model fit quite well with both estimation techniques having a per- centage of corrected classified value about 78% (Table A.3 in appendix). Results from VIF test suggest that variables are uncorrelated with each other. Toler- ance is different from zero and the variance inflation is low. Evidence of good fit is reflected in a ROC curve (figure 1 in appendix), the area under the ROC curve is equal to 0.83 meaning that 83% of the observations are correctly classified. 5. Conclusions The food processing industry is a sector mainly constituted by SMEs, with a low pro- pensity to adopt process innovation. It represents one of the four sectors that consumes more energy in Europe although, at a large extent, it is not considered energy intensive and therefore covered by the EU Emissions Trading System (ETS); it also is considered to have a high energy saving potential. A major barrier that literature finds is the low impor- tance attributed to energy consumption in non-energy intensive industries as in the case 5 A confusion matrix is a table used to describe the performance of a classification model on a set of test data for which the true values are known. It tells us how many actual values and predicted values exist for different classes predicted by the model. 41Drivers and barriers of process innovation of food processing industry. Another barrier is the lack of high expected returns ad short payback times. Furthermore, SMEs have limited access to information, low energy share on their expenditures, too high transaction costs for fund searching, cost disadvantages in Table 5. Robustness Logit estimation. (1) Proc_inn_adoption (2) Proc_inn_adoption (3) Proc_inn_adoption Proc_inn_adoption       lnturnover 0.1163*** 0.1100*** 0.1052***   (0.0248) (0.0255) (0.0259) funds 0.0878 0.0881 0.0902   (0.0992) (0.0996) (0.0996) lmarket -0.0453 -0.0343 -0.0449   (0.0885) (0.0889) (0.0893) rmac 1.5594*** 1.5537*** 1.5553***   (0.0914) (0.092) (0.092) gnewmkt -0.0017 0.0011 -0.0001   (0.039) (0.0393) (0.0392) orgbup 0.7801*** 0.7812*** 0.7771***   (0.0929) (0.0936) (0.0937) ssup 0.3931*** 0.3991*** 0.3996***   (0.0431) (0.0433) (0.0433) scom -0.0871 -0.0888 -0.089   (0.0452) (0.0456) (0.0456) sins 0.1445** 0.1416** 0.1395**   (0.0478) (0.0479) (0.048) co 0.2694* 0.2749* 0.2677*   (0.1071) (0.1078) (0.1079) policy_m 4.8399*** 4.5146** 4.4198**   (1.2369) (1.3724) (1.3801) ep 15.4363** 15.3776** 15.0575**   (4.8897) (4.8956) (4.9124) obsprs   0.0219 0.0241     (0.0455) (0.0454) obsfin   -0.0441 -0.0429     (0.0384) (0.0384) empedu     0.1113       (0.1003) N 4425 4398 4398 adj. R-sq  0.346 0.34  0.34  chi2 998.4058 994.3702 999.5503 BIC 4358.8634 4345.35 4352.528 Time dummies and country dummies are included in the model. Standard errors in parentheses. Vari- able statistically significant at * p < 0.05, ** p < 0.01, *** p < 0.001 respectively. 42 Federica DeMaria, Annalisa Zezza obtaining or developing innovation. All these problems need to be specifically addressed by policy measures. In particular, small food firms contribute substantially to the food processing industry economic performance and are considered to play a role in achieving sustainable economic growth in local economies. Nevertheless, the small size of the indus- tries and their less energy intensive use clarifies why it is difficult to apply energy policy measures and, why the optimization process is often not a priority for company managers. Being among the largest users of surface water and energy among the manufacturing sectors, the food processing industry needs to reduce its energy consumption and improve its energy efficiency. However, changes in energy efficiency are often difficult because they depend on several factors such as the energy used technical performance, the importance of energy trans-formations, climate conditions, the structure of each economic sector that uses energy (MEDENER, 2013). In this study, we have addressed the following issues: a) the role of cooperation agree- ments between food processing industries and research institutes in promoting process innovation adoption; b) whether or not the high price of energy encourages process inno- vation adoption; c) whether or not the environmental regulation and the severity of the policies stimulate process innovation adoption; d) which of these two factors, energy price and successful policy measures, are the most effective tool in promoting process innova- tion adoption. Through our models, we have addressed the issues related to process inno- vation adoption in the EU food processing industry. As regards to the first two hypoth- eses, namely the role of cooperation and network activities, on the one hand (H1a) and the role of information (H1b), our study confirms that cooperation agreements encour- age SMEs to adopt new process innovations. Networking activities in this regard are rel- evant because they allow SMEs to acquire all the knowledges that is unavailable within the firm. The role of different information sources is uncertain. Indeed, the internal sources availability seem to encourage the adoption of new processes innovation. External sources information availability shows a negative sign that could be due to the higher costs related to the acquisition of this kind of knowledge and the manager’s or owners’ attitudes. Concerning the role energy policy, results confirm the key role of energy prices and energy policies, with the energy price coefficient having a higher weight. Furthermore, results confirm the role and the relevance of drivers like financial resources availability at the enterprise level, the presence of new organizational methods, the positive role of R&D firms’ engagement and cooperation activities. These are impor- tant findings, in particular, if we consider that SMEs have limited access to information, low energy share on their expenditures, too high transaction costs for fund searching, cost disadvantages in obtaining or developing innovation. Our results support previous research in identifying the main areas for policy action. Process innovation adoption in the food processing industry could be enhanced by meas- ures addressed to: - information cost reduction in order to support informed choices. One example is the support energy auditing in SMEs as a tool to track energy consumption and costs throughout a facility and identify opportunities to reduce energy use, increasing entrepreneurs’ awareness. - Contrasting the low private investment in R&D in SMEs for process and energy sav- ing innovations through public-funded R&D or promoting enterprises aggregation 43Drivers and barriers of process innovation in networks. This action can take several forms research public of financing activities that require partnership with the private sector including technology providers and or facilitating partnership agreements between sectoral entrepreneurs and technology providers. - Promoting policy coherence in EU policies impacting on energy use and innovation i.e. CAP, EU Cohesion policies, Energy and Climate policies as well the Bioeconomy and the Circular economy Strategies. - Reinforcing the role of partnership tools along the whole supply chain in the agri- food sector enhancing cooperation towards sustainable production, thus creating the necessary conditions by which green labels could deliver and increase the demand for sustainable products. - Considering that policy instruments are located at different government levels (EU, MS, regional or local) when dealing with the appropriate policy mix, and increase policy coherence through the whole governance system in which policy tools operate (Borràs and Edquist 2013). Acknowledgments This article has been co-funded by the EU Project EnreMilk. Grant Agreement No.: 613968 (FP7-KBBE-2013-7-single-stage). The views expressed in this article are exclusive- ly those of the authors. We thank the two anonymous referees for their outstanding comments and suggestions. References Alexiev, A.S., Volberda, H.W., Van den Bosch, F.A. (2016). Interorganizational collabora- tion and firm innovativeness: Unpacking the role of the organizational environment. Journal of Business Research, 69 (2) (2016), pp. 974-984. Altmann, M., Michalski, J., Brenninkmeijer, A., & Tisserand, P. (2010). Overview of Ener- gy Efficiency measures of European industry. Brussels, Belgium: European Parlia- ment’s Committee on Industry, Research and Energy (ITRE), 51. Amara N, Landry R (2005) Sources of information as determinants of novelty of innova- tion in manufacturing firms: evidence from the 1999 statistics Canada innovation survey. Technovation 25(3): 245–259. Arimura, T., Hibiki, and A., Johnstone, N. (2007) “An empirical study of environmental R&D: what encourages facilities to be environmentally-innovative ?” in Johnstone, N. (ed.) Corporate Behaviour and Environmental Policy Cheltenham UK: Edward Elgar in association with OECD. Avermaete, T., Viaene, J., Morgan, E.J., Pitts, E., Crawfor, N., Mahon, D. (2004). Determi- nants of products and process innovation in small food manufacturing firms. Trends in food science and technology, 15, 474-483. Batterink, M.H., Wubben, F.M., Klerkx, L., Omta, S.W.F. 2010. Orchestrating innovation networks: the case of innovation brokers in the agri-food sector. Entrepreneurship and Regional Development, 22(1), January, 47-76. 44 Federica DeMaria, Annalisa Zezza Battisti, G., Stoneman, P. 2010. How Innovative are UK Firms? Evidence from the Fourth UK Community Innovation Survey on Synergies between Technological and Organ- izational Innovations. British Journal of Management, Vol. 21, 187–206 (2010) Belin, J., J. Horbach, Oltra. V. 2011. Determinants and specificities of eco-innovations – an econometric analysis for France and Germany based on Community Innovation Survey. Cahiers du GREThA n° 2011-17. Bigliardi, B., and Galati, F. (2013). Innovation trends in the food industry: the case of functional foods. Trends in Food Science & Technology 31 (2), 118-129, 2013. 514, 2013. Borrás, S., & Edquist, C. (2013). The choice of innovation policy instruments. Technologi- cal forecasting and social change, 80(8), 1513-1522. Bröring, S., Cloutier, L.M. (2008). Value-creation in new product development within con- verging value chains. An analysis in the functional foods and nutraceutical industry. British food journal, 110 (1), 76-97. Cagno, E., Worrell, E., Trianni, A., & Pugliese, G. (2013). A novel approach for barriers to industrial energy efficiency. Renewable and Sustainable Energy Reviews, 19, 290-308. Cainelli G., Mazzanti M., Montresor S. (2012) Environmental Innovations, Local Net- works and Internationalization, Industry and Innovation, 19(8):697-734. Cainelli G., Mazzanti M., Zoboli R. (2011) Environmental innovations, complementarity and local/global cooperation: evidence from North-East Italian industry, Interna- tional Journal of Technology, Policy and Management, 11(3/4):328-368. Capitanio, F., Coppola, A. and Pascucci, S. (2009), “Indications for drivers of innovation in the food sector”, British Food Journal, Vol. 111 No. 8, pp. 820-38. Capitanio, F., Coppola, A. and Pascucci, S. (2010), Product and process innovation in the Italian food processing industry. Agribusiness (26): 503-518. Caputo, A., Cucchiella, F., Fratocchi, L., Scacchia, F., Pelagagge, P.M. (2002). A method- ological framework for innovation transfer to SMEs July 2002. Industrial Manage- ment & Data Systems 102(5):271-283. Cassiman, B., Veugelers, R., 2002. R&D cooperation and spillovers: some empirical evi- dence from Belgium. American Economic Review 92 (4), 1169–1184. Ciliberti, S., Bröring, S., Carrisi., L.2016. External Knowledge Sources as Drivers for Cross-Industry Innovation in the Italian Food Sector: Does Company Size Matter? September 2016. The International Food and Agribusiness Management Review 19(3):77-98 Clauss, T., and Kesting, T. (2017). How businesses should govern knowledge-intensive col- laborations with universities: An empirical investigation of university professors. Industrial Marketing Management 62 (April 2017), 185-198. Cohen, W.M. and Klepper, S. (1996), Firm Size and the Nature of Innovation within Industries: The Case of Process and Product R&D. The Review of Economics & Sta- tistics 78: 232-243. Cuerva, M.C., Triguero-Cano, Á. and Córcoles, D., 2014. Drivers of green and non-green innovation: empirical evidence in Low-Tech SMEs. Journal of Cleaner Production, 68, pp.104-113. D’Este P., Iammarino S., Savona M, von Tunzelmann N. (2012) What hampers innova- tion? Revealed barriers versus deterring barriers, Research Policy, 41(2):482-488. 45Drivers and barriers of process innovation Dahlander, L., and Gann M.V. 2010 How open is innovation? Research Policy, 2010, vol. 39, issue 6, 699-709 De Faria, P., Lima, F., & Santos, R. (2010). Cooperation in innovation activities: The importance ofpartners.Research Policy, 39,1082–1092. de Jong, Jeroen P.J., Vermeulen Patrick A.M., (2006). Determinants of Product Innovation in Small Firms: A Comparison Across Industries. International Small Business Jour- nal Volume: 24 issue: 6, page(s): 587-609 De Marchi V. (2012) Environmental innovation and R&D cooperation: Empirical evi- dence from Spanish manufacturing firms, Research Policy, 41(3):614-623. Del Rio Gonzalez P (2005): Analysing the Factors Influencing Clean Technology Adop- tion: A Study of the Spanish Pulp and Paper Industry. Business Strategy and the Environment, 14:20–37. Demirel, P., and Kesidou, E. (2011), Stimulating different types of eco-innovation in the UK: Government policies and firm motivations, Ecological Economics, 70, (8), 1546-1557. Diederen, P., Van Meijl, H., Wolters, A., Bijak, K. (2003). Innovation adoption in agricul- ture: innovators, early adopters and laggards. Cahiers d’Economie et de Sociologie Rurales, INRA Editions,2003, 67, pp.29-50. hal-0120104 Eurobarometer (2016), Flash Eurobarometer 433, Innobarometer 2016 – EU business innovation trends, http://www.horizon2020news.it/wpcontent/uploads/2016/07/ fl_433_sum_en.pdf Evangelista, R., Iammarino, S., Mastrostefano, V., Silvani, A. 2002. Looking for regional system of innovation: evidence from the Italian innovation survey. Regional Studies, 36(2), pp. 173-186. Feder, G., Just, R.E., Zilberman, D. (1985). Adoption of agricultural innovations in devel- oping countries: a survey. Econ. Dev. Cult. Change 33, 255–298. Fleiter, T., Schleich, J. and Ravivanpong, P., 2012. Adoption of energy-efficiency measures in SMEs—An empirical analysis based on energy audit data from Germany. Energy Policy, 51, pp.863-875. Frondel, M, Horbach, J, Rennings, K. (2007). End-of-pipe or cleaner production? An empirical comparison of environmental innovation decisions across OECD coun- tries. Business Strategy and the Environment, 16 (8) (2007), pp. 571-584. Galende, J. and de la Fuente, J. M. (2003) Internal factors determining a firm’s innovative behavior. Research Policy 32 (2003) 715–736. Galizzi, G. and Venturini, L. (1996). Product innovation in the food processing industry: Nature, characteristics and determinants. In G. Galizzi & L. Venturini (Eds.), Eco- nomics of innovation: The case of food processing industry (pp. 133–153). Heidel- berg: Physica-Verlag. Garcıa-Martınez, M. and Briz, J. (2000). Innovation in the Spanish food & drink industry. International Food and Agribusiness Management Review 3: 155-176. Gellynck X., Vermeire, B., Viaene, J., (2007). Innovation in food firms: contribution of regional networks within the international business context. Entrepreneurship & Regional Development, 19 (3), 209-226. Gellynck, X., Kühne, B. (2010). Horizontal and vertical networks for innovation in the tra- ditional food sector. International Journal on Food System Dynamics, 1 (2), 123-132. 46 Federica DeMaria, Annalisa Zezza Ghisetti, C., Marzucchi, A. Montresor, S. (2015). The open eco-innovation mode: an empirical investigation of eleven European countries. Res. Policy, 44 (5) (2015), pp. 1080-1093, 10.1016/j.respol.2014.12.001. Ghisetti, C., Marzucchi, A., Montresor, S., 2013. Does external knowledge affect envi- ronmental innovations? An empirical investigation of eleven European countries, INGENIO (CSIC-UPV) Working Paper Series No.2013-01, Valencia. Ghisetti, C., Rennings, K., 2014. Environmental innovations and profitability: How does it pay to be green? An empirical analysis on the German innovation survey, Journal of Cleaner Production, 75, 106-117 Gómez J, Salazar I, Vargas P. Sources of information as determinants of product and pro- cess innovation. PLoS One. 2016;11: e0152743. pmid:27035456- Greene, W. H. (2008). Econometric analysis, 6th ed. Upper Saddle River N.J., Prentice Hall. Gulati R 1998. Alliances and networks. Strategic Management Journal, Volume 19, Issue 4, April 1998 Pages 293–317 Hájek, P., and Stejskal, J. (2018). R&D Cooperation and Knowledge Spillover Effects for Sustainable Business Innovation in the Chemical Industry. Sustainability 2018, 10, 1064; doi:10.3390/su10041064 Hashi, I., and Stojcic, N. (2013), The impact of innovation activities on firm performance using a multi-stage model: Evidence from the Community Innovation Survey 4, Research Policy, 42, (2), 353-366. Heirati, N., O’Cass, A., Schoefer, K., & Siahtiri, V. (2016). Do professional service firms benefit from customer and supplier collaborations in competitive, turbulent envi- ronments? Industrial Marketing Management, 55, 50–58. Hoffman, W.H., Schlosser, R., 2001. Success factors of strategic alliancesin small and medium-sized enterprises: an empirical survey. LongRange Planning 34, 357–381. Horbach J. (2008) Determinants of environmental innovation—new evidence from Ger- man panel data sources, Research Policy, 37(1):163-173. Horbach, J., Rammer, C., & Rennings, K. (2012). Determinants of eco-innovations by type of environmental impact—The role of regulatory push/pull, technology push and market pull. Ecological Economics, 78, 112-122. Hyytinena, A., and Toivanenab, O. (2005). Do financial constraints hold back innovation and growth? Evidence on the role of public policy. Research Policy Volume 34, Issue 9, November 2005, Pages 1385-1403. ICF (2015), Study on Energy Efficiency and Energy Saving Potential in Industry from pos- sible Policy Mechanisms, Report for EC - https://ec.europa.eu/energy/sites/ener/ files/documents/151201%20DG%20ENER%20Industrial%20EE%20study%20-%20 final%20report_clean_stc.pdf. Jaffe, A., Stavins, R., 1994. The energy-efficiency gap. What does it mean? Energy Policy 22 (10), 804-810. Jänicke M. and Jacob K. (2002): “Ecological modernisation and the creation of lead mar- kets”, FFU-report 03-2002, Forschungstelle fuer Umweltpolitik, Freie Universitaet Berlin. Jones, DC, P Kalmi, T Kato and M Mäkinen (2008). The incidence and determinants of employee involvement–evidence from the finnish Manufacturing Sector. Available at http://klaara.etla.fi/files/1974 Dp1131.pdf. 47Drivers and barriers of process innovation Kaminski, J. and Leduc, G., ‘Energy efficiency improvement options for the EU food industry’, Polityka Energetyczna, 13(1), 2010, pp. 81-97. Kaufmann, A., Todtling, F., 2000. Science–industry interaction in the process of innova- tion: the importance of boundary-crossing between systems. Research Policy 1241 (2000), 1–14. Kemp, K. and Volpi, M. (2008) “The diffusion of clean technologies: a review with sug- gestions for future diffusion analysis”, Journal of Cleaner Production, 16S1 (2008) S14eS21. Kemp, R. and Andersen, M. M. (2004): “Strategy for Eco-Efficiency Innovation”, IMR Strategielijnen project voor VROM, Final version, 1 p. King, D.R., J.G. Covin, and W.H. Hegarty. 2003. Complementary resources and the exploi- tation of technological innovation. Journal of Management 29(4): 589–606. Klewitz, J., and Hansen, E.G. (2013). Sustainability-oriented innovation of SMEs: a sys- tematic review. Journal of Cleaner Production xxx (2013) 1e19. Köhler C, Sofka W, Grimpe C. 2012. Selective search, sectoral patterns, and the impact on product innovation performance. Research Policy 41 (8): 1344-1356 Ladha‐Sabur, A., Bakalis, S., Fryer, P. J., & Lopez‐Quiroga, E. (2019). Mapping energy con- sumption in food manufacturing. Trends in Food Science and Technology, 86, 270– 280. Laperche, B., Liu Z. (2013). SMEs and knowledge-capital formation in innovation net- works: a review of literature. Journal of innovation and entrepreneurship, 2-21. Läpple, D., Thorne, F., 2018. The Role of Innovation in Farm Economic Performance: Generalised Propensity Score Evidence from Irish Dairy Farms. J. Agric. Econ. 1–24. https://doi.org/10.1111/1477-9552.12282. Lasagni, A. (2012). How can external relationships enhance innovation in SMEs? New evi- dence for Europe. Journal of Small Business Management, 50(2), 310-339. Laursen, K., and Foss, N. 2003. New human resource management practices, complemen- tarities and the impact on innovation performance. Cambridge Journal of Econom- ics, 27(2): 243-263. Lee, A. H., Wang, W. M. and Lin, T. Y. 2010. An evaluation framework for technology transfer of new equipment in high technology industry. Technological Forecasting and Social Change 77(1): 135-150. Lefebvre, V.M., De Steur, H. and Gellynck, X. (2015), “External sources for innova- tion in food SMEs”, British Food Journal, Vol. 117 No. 1, pp. 412-430. https://doi. org/10.1108/BFJ-09-2013-0276. López, A., 2008. Determinants of R&D cooperation: evidence from Spanish manufactur- ing Martino G., Polinori P., (2011): Networks and organizational learning: Evidences from broilers production. British Food Journal113, 871-885. Mazzanti M., Montini A., Zoboli R. (2008) Complementarities, Firm strategy and Envi- ronmental Innovations: Empirical evidence for a local manufacturing system, Envi- ronmental Sciences, 5(1):17-40. McKinsey and Company (2010), Energy efficiency: a compelling global resource. MEDENER (2013). Energy efficiency trends in Mediterranean countries, Tunis, MEDEN- ER, July (http://medener-indicateurs.net/uk/documents-fourth-reunion.html). 48 Federica DeMaria, Annalisa Zezza Melane-Lavado, A., & Álvarez-Herranz, A. (2020). Cooperation Networks as a Driver of Sustainability-Oriented Innovation. Sustainability, 12(7), 2820. Minarelli, F., Raggi, M. and Viaggi, D. (2014). Distinguishing the innovation behaviour of micro, small and medium food enterprises. Journal on Chain and Network Science 14(2): 95-102. Möllera, M., Halinenb, A. (2017). Managing business and innovation networks. From strategic nets to business fields and ecosystems. Industrial Market Journal 67(2017) 5-22. Monforti-Ferrario, F., & Pascua, I. P. (Eds.). (2015). Energy use in the EU food sector: State of play and opportunities for improvement. Publications Office. O’Connor, G., and DeMartino, R. (2006). Organizing for Radical Innovation: An Explora- tory Study of the Structural Aspects of RI Management Systems in Large Established Firms. October 2006. Journal of Product Innovation Management 23(6):475 – 497. Odyssee Mure (2015). Energy Efficiency Trends and Policies In Industry, https://www. odyssee-mure.eu/publications/archives/energy-efficiency-trends-policies-industry. pdf OECD (2011), Towards a green growth: a summary for policy makers. OECD Publishing. http://dx.doi.org/10.1787/9789264111318-en. OECD (2017), Improving Energy Efficiency in the Agro-food Chain, OECD Green Growth Studies, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264278530- en OECD-EUROSTAT (1997), Oslo Manual. Proposed Guidelines for Collecting and Inter- preting Technological Innovation Data, OECD, Paris Oliveira L and Carvalho H. (2010) Why Firms Do Not Enrol in Socio-Technical Networks -Empirical Evidence from Portugal. Sociology of Science and Technology, Institute for the History of Science and Technology/Russian Academy of Sciences1: 73-84. Omta, S.W.F., 2002. Innovation in chains and networks. Journal on Chain and Network Science, 2(2): 73-80. Popp, D. (2002). Induced Innovation and Energy Prices. American Economic Review vol. 92, no. 1, March 2002. Popp, D. (2009): “Policies for The Development and Transfer of Eco-Innovations: Lessons from The Literature”, OECD Global Forum on Environment on Eco-Innovation, Paris. Porter, M. E., and van der Linde, C. (1995b), “Toward a New Conception of the Environ- ment-Competitiveness Relationship”, Journal of Economic Perspectives 9(4): 97-118. Porter, M. E., and van der Linde, C. (1995a), “Green and Competitive - Ending the Stale- mate”, Harvard business review 73: 120-134. Powell, W. and Grodal, S. (2005). “Network of Innovators,” In: J. Fagerberg, D. Mowery and R. R. Nelson, Eds., The Oxford Handbook of Innovation, Oxford University Press, Oxford, 2005, pp. 56-85. Prester, J., Bozac, MG. (2012). Are innovative organizational concepts enough for foster- ing innovation? International Journal of Innovation Management · February 2012. Rave, T.; Goetzke, F.; Larch, M. (2011). “The determinants of environmental innovations and patenting: Germany reconsidered”, IFOWorking Paper Series No. 97, IFO Insti- tute for Economic Research at the University of Munich. 49Drivers and barriers of process innovation Rehfeld K.M., Rennings K. and Ziegler A. (2006): “Integrated Product Policy and Envi- ronmental Product Innovations - An Empirical Analysis”, Ecological Economics. Reichstein, T., & Salter, A. (2006). Investigating the sources of process innovation among UK manufacturing firms. Industrial and Corporate Change, 15(4), 653-682. https:// doi.org/10.1093/icc/dtl014. Rennings, K., 2000. Redefining innovation—eco-innovation research and the contribution from ecological economics. Ecological Economics, 32(2), pp.319-332. Romijn H and Albaladejo M. (2002) Determinants of innovation capability in small elec- tronics and software firms in southeast England. Research Policy31: 1053-1067. Salavisa I, Sousa C and Fontes M. (2012) Topologies of innovation networks in knowl- edge-intensive sectors: Sectoral differences in the access to knowledge and comple- mentary assets through formal and informal ties. Technovation32: 380-399. Sauer, J., Zilberman, D., 2012. Sequential technology implementation, network externali- ties, and risk: the case of automatic milking systems. Agric. Econ. 43, 233–252. htt- ps://doi.org/10.1111/j.1574-0862.2012.00579.x. Schumpeter, J.A., 1934. The Theory of Economic Development. Harvard University Press, Cambridge, MA. Schumpeter, J.A., 1942. Capitalism, Socialism and Democracy. Harper & Row, New York. Silva, M., Leitao, J., Raposo, M. (2008). Barriers to innovation faced by manufacturing firms in Portugal: How to overcome it for fostering business excellent? International Journal of Business Excellence, 1(1–2), 92–105. Sorrell S, Schleich J, Scott, S, O’Malley E, Trace F, Boede U, Ostertag K, and Radgen P, Reducing barriers to energy efficiency in public and private organizations Energy research centre science and technology policy research (SPRU), University of Sus- sex, Brighton, 2000. Stewart-Knox, B. and P. Mitchell, 2003. What separates the winners from the losers in new food product development? Trends in Food Science & Technology 14(1): 58-64. Tether, B., and Tajar, A. (2008). Beyond industry-university links: Sourcing knowledge for innovation from consultants, private research organisations and the public science- base. Research Policy, 2008, vol. 37, issue 6-7, 1079-1095. Trianni, A., Cagno, E., Worrell, E., & Pugliese, G. (2013). Empirical investigation of ener- gy efficiency barriers in Italian manufacturing SMEs. Energy, 49, 444–458. Triguero, A., Moreno-Mondéjar, L., Davia, María A., 2013. Drivers of different types of eco-innovation in European SMEs. Ecological Economics. Elsevier, vol. 92(C), pages 25-33. Van Gils, A., and Zwart, P.S. (2004). Knowledge Acquisition and Learning in Dutch and Belgian SMEs: December 2004European Management Journal 22(6):685-692 DOI: 10.1016/j.emj.2004.09.031. Veugelers, R., 2012. Which policy instruments to induce clean innovating? Research Poli- cy, 41(10), pp.1770-1778. Vinding, A. (2002), International Diffusion and Transformation of Knowledge in the Pro- cess of Product Innovation, excerpt of Ph.D. dissertation, Aalborg University. Wagner, M. 2009. Innovation and Competitive Advantages from the Integration of Strate- gic Aspects with Social and Environmental Management in European Firms. Busi- ness Strategy and the Environment 18, 291–306 (2009). 50 Federica DeMaria, Annalisa Zezza Wei, Y. and Wang, Q. (2011). Making sense of a market information system for superior performance: The roles of organizational responsiveness and innovation strategy. Industrial Marketing Management 40(2): 267–277. ZEW (coord) (2001). “The Impact of Clean Production on Employment in Europe – An Analysis using Surveys and Case Studies (IMPRESS)”, Draft Final Report of Con- tract, No: SOE-1-CT-98-1106. Zhou, K.Z., and Bingxin Li, C. (2012). How knowledge affects radical innovation: Knowl- edge base, market knowledge acquisition, and internal knowledge sharing. Startegic Management Journal, Volume 33, issue 9 p.1090-1102. APPENDIX Table A.1. Variables’ description. Variable CIS Description Variable in the model Re-coded Variable Expected sign INPSPD Introduced onto the market a new or significantly improved method of production Dependent Variable: process innovation adoption dummy variable: 0=No; 1=Yes   INPSLG Introduced onto the market a new or significantly improved logistic, delivery or distribution system INPSSU Introduced onto the market a new or significantly improved supporting activities TURN Total turnover Size total turnover in euros + FUNDS Public funding from local or regional authorities External Financial capacity dummy variable: 0=No; 1=Yes + MOTHER Local/regional market (within country) lmarket dummy variable: 0=No; 1= Yes + RMAC Acquisition of machinery RMAC   + GOMKT Increase market share GOMKT   + ORGBUP New business practices for organizing work or procedures  ORGBUP dummy 0= No , 1=Yes + CO Cooperation arrangements on innovation activities Cooperation activities dummy variable: 0=No; 1=Yes + SENTG Sources from within the enterprise or enterprise group Int_info_sources dummy variable: 0=not important; 1=important + SSUP Suorces from suppliers of equipment, materials etc other_info_sources dummy variable: 0=not important; 1=important + SCOM Sources from Competitors and other enterprises of the same industry SINS Sources from consultants, commercial labs or private R&D institutes 51Drivers and barriers of process innovation Variable CIS Description Variable in the model Re-coded Variable Expected sign SUNI Sources from Universities or other higher education institutes SCLUP Clients or customers from the public sector SCON Sources from professional conferences, trade fairs, meetings SJOU Sources from Scientific journals, trade/ scientific publications SPRO Sources from Professional and industry associations Policy_m Energy saving successful policy measures Policies 0= .5 True D defined as dipend_1 != 0   Sensitivity Pr(+| D) 91.48% Specifity Pr(-|~D) 56.16% Positive predictive value Pr(D| +) 78.88% Negative predictive value Pr(~D|-) 78.64%   False + rate for true ~D Pr(+|~D) 43.84% False - rate for true D Pr (-| D) 8.52% False + rate for classified + Pr(~D|+) 21.12% False - rate for classified - Pr(D| -) 21.36% Correctly classified 78.82% Figure 1. ROC Curve. 0. 00 0. 25 0. 50 0. 75 1. 00 S en si tiv ity 0.00 0.25 0.50 0.75 1.00 1 - Specificity Area under ROC curve = 0.8288 The role of trust and perceived barriers on farmer’s intention to adopt risk management tools Elisa Giampietri1, Xiaohua Yu2, Samuele Trestini3,* Drivers and barriers of process innovation in the EU manufacturing food processing industry: exploring the role of energy policies Federica DeMaria, Annalisa Zezza Step-by-step development of a model simulating returns on farm from investments: the example of hazelnut plantation in Italy Alisa Spiegel1,*, Simone Severini2, Wolfgang Britz3, Attilio Coletta2 The role of group-time treatment effect heterogeneity in long standing European agricultural policies. An application to the European geographical indication policy Leonardo Cei1, Gianluca Stefani2, Edi Defrancesco1 The impact of food price shocks on poverty and vulnerability of urban households in Iran Ghasem Layani1, Mohammad Bakhshoodeh1, Mona Aghabeygi2*, Yaprak Kurstal3, Davide Viaggi3