Breach of academic values and digital deviant behaviour: The case of Sci-Hub

Rossello, Giulia; Martinelli, Arianna Working Paper Breach of academic values and digital deviant behaviour: The case of Sci-Hub UNU-MERIT Working Papers, No. 2023-009 Provided in Cooperation with: Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), United Nations University (UNU) Suggested Citation: Rossello, Giulia; Martinelli, Arianna (2023) : Breach of academic values and digital deviant behaviour: The case of Sci-Hub, UNU-MERIT Working Papers, No. 2023-009, United Nations University (UNU), Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), Maastricht This Version is available at: https://hdl.handle.net/10419/326859 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-sa/4.0/ #2023-009 Breach of academic values and digital deviant behaviour: The case of Sci-Hub Giulia Rossello and Arianna Martinelli Published 30 March 2023 Maastricht Economic and social Research institute on Innovation and Technology (UNU-MERIT) email: [email protected]u | website: http://www.merit.unu.edu Boschstraat 24, 6211 AX Maastricht, The Netherlands Tel: (31) (43) 388 44 00 UNU-MERIT Working Papers ISSN 1871-9872 Maastricht Economic and social Research Institute on Innovation and Technology UNU-MERIT | Maastricht University UNU-MERIT Working Papers intend to disseminate preliminary results of research carried out at UNU-MERIT to stimulate discussion on the issues raised. Breach of Academic Values and Digital Deviant Behaviour: the Case of Sci-Hub Giulia Rosselloa,b and Arianna Martinelli a aIstituto di Economia, Scuola Superiore Sant’Anna, Martiri della Libert` a 33, 56127 Pisa, Italy; bUNU-MERIT, Maastricht University, Boschstraat 24, 6211 AX Maastricht, the Netherlands. ARTICLE HISTORY Compiled March 29, 2023 ABSTRACT This paper bridges the organisational psychology and the economics of science literature to examine the role of ideology-based psychological contract breach in eliciting mild deviant behaviour in academia. We provide empirical evidence of how the deterioration of academic values related to the diffusion of the “publish or perish” paradigm sparkles copyright violations through Sci-Hub. Based on a representative sample of 2849 academics working in top institutions in 6 European countries, we find that ideology-based psychological contract breach explains Sci-Hub usage, also when controlling for other trivial motivations. The magnitude of the effect depends on contextual and demographic characteristics. Females, foreign and tenured scholars are less likely to respond with digital piracy when experiencing a contract breach of academic values. Our results contribute to prevention policy design, highlighting how policies restoring academic values might also address academic piracy. KEYWORDS Academic Values; Digital Piracy; Deviant Behaviour; Psychological Contract Breach; Sci-Hub. JEL classification: D23, L86, O34 1. Introduction Deviant workplace behaviour (DWB), or more generally counter-productive work behaviour (CWB), such as absenteeism, retaliation, sabotage, aggression, and violence, generates risks and costs for organisations (Spector, Fox, & Domagalski, 2006; Zaghini, Fida, Caruso, Kangasniemi, & Sili, 2016). These hazards also affect academia, and previous literature mostly looked at academic dishonesty and cheating behaviour of college students during exams or evaluations (Farnese, Tramontano, Fida, & Paciello, 2011; S. D. Lee, Kuncel, & Gau, 2020; McCabe, Trevi˜ no, & Butterfield, 2001; Whitley, 1998). However, research-related scandals1, such as the well-known case of Stapel (Bhattacharjee, 2013), exacerbate the growing public interest in scholars’ deviant research-related behaviour or opaque practices. A growing literature documents deviant or misbehaviour in academia, such as data fabrication (Fanelli, 2009), citation manipulation (Fong & Wilhite, 2017), plagiarism (Karabag & Berggren, 2012), questionable research practices (Necker, 2014), and misbehaviour in authorship (Smith et al., CONTACT G. Rossello Email: [email protected] 1See also https://retractionwatch.com for a collection of retraction cases (Last Access November 2022). 2020) to name a few. Nevertheless, not all transgressions are extreme. Some mild ones might be pretty diffused or socially accepted, as in the case of academic software (Rahim, Rahman, & Seyal, 2000; Santillanes & Felder, 2015; Wickham, Plotnicki, & Athey, 1992) or article (Dui´ c, Konjevod, & Grzunov, 2017; Hoy, 2017; Nicholas et al., 2017) piracy. Interestingly, academic piracy mainly harms third parties (e.g. publishers and software companies) and often stems from time or financial constraints. Thus, in most cases, it is undetected, tolerated, or even encouraged by universities. This paper investigates one specific academic deviant behaviour: using Sci-Hub to access scientific literature. The illegal downloading of scientific articles is a form of academic piracy, a mild deviant behaviour that violates publishers’ copyright. We focus on such a mild transgression because it might signal the beginning of negative feedback loops generating future more deviant severe behaviours. The literature interested in preventing DWB describes the diffusion of thereof as a snowball effect. Those studies underline the need to investigate mild transgressions and their determinants, as they might go unnoticed before they escalate (Griep & Vantilborgh, 2018; Spector et al., 2006). Past literature examines escalatory patterns where negative experiences, such as stress, conflicts, changes, and bureaucratic control, generate a perceived sense of injustice and negative emotions that induce CWB feeding subsequent transgressions (J. A.-M. Coyle-Shapiro, Pereira Costa, Doden, & Chang, 2019; Griep & Vantilborgh, 2018). The literature on psychological contract breaches indicates that a trigger of DW is the experience of a contract breach. Every worker has an implicit psychological contract with their organisation, a set of beliefs about jobs’ rights and duties. When the worker perceives a betrayal of those beliefs, the contract has been breached, and this experience elicits deviant behaviours (J. A.-M. Coyle-Shapiro et al., 2019; O’Donohue, Sheehan, Hecker, & Holland, 2007). Besides the extensive empirical evidence that experiencing contract breaches induces deviant behaviour (J. A.-M. Coyle-Shapiro et al., 2019), most studies examine the economic or socioemotional dimensions of psychological contracts, not addressing the characteristics of specific professions. Thompson and Bunderson (2003) highlights that in some cases and professions, the psychological contract might also have an ideological base. Those ideology-based contracts are a credible commitment of workers to pursue a valued cause which goes beyond self-interest and constitutes the nexus of the individual and his/her profession (O’Donohue et al., 2007). Indeed, many professions have a valued cause intrinsic to them called “ideological currency” (Bunderson, 2001; J. A.-M. Coyle-Shapiro et al., 2019). For example, a nurse might feel the responsibility of saving lives (Krause & Moore, 2017) or a researcher might feel the one of advancing the knowledge frontier (O’Donohue et al., 2007). The motivations for the academic profession are the quest for knowledge and confidence in scientific inquiry. These motivations drive academic values: interest in fundamental knowledge and curiosity-driven research, a lack of bureaucratic control and freedom of time allocation, an interest in sharing knowledge, and a reward system based on status and prestige rather than monetary rewards (Sauermann & Stephan, 2013). These values constitute the backbone of the ideology-based psychological contract of the academic profession (Merton, 1973; Sauermann & Stephan, 2013). However, these values are fading under global trends, such as the increase of teaching and administrative load2, of short-term positions and, in par2See the recent article of Kathleen R. Brewer in the Financial Times about the Management Bloat https://www.ft.com/ content/338d7321-bc87-4573-885e-565f34a80b30 2 ticular, the pressure on publishing and funding (Bryson, 2004; Carvalho & Santiago, 2010; Chatelain-Ponroy, Mignot-G´ erard, Musselin, & Sponem, 2018; Siekkinen, Pekkola, & Carvalho, 2020). These trends have increased academic discontent due to the reduction of scholars’ timeallocation freedom, the lower recognition from pursuing curiosity-driven research, and a generalised emphasis on results’ “publishability”. This last change has been popularised with the expression “publish-or-perish”, which emphasises how pivotal publishing is in proving academic lead. Besides the recent acceleration, the “publish-or-perish” potential harm to the fundamental principle of research and academic values has been known since the 1940s (Garfield, 1996). In 1972, the political philosopher Hannah Arendt argued: “This business of ‘publish-or-perish’ has been a catastrophe. People write things which should never have been written and which should never be printed. Nobody’s interested. But for them to keep their jobs and get the proper promotion, they’ve got to do it. It demeans the whole of intellectual life.” Hannah Arendt 13th July 1972 panel discussion titled “Values in Contemporary Society” Today’s centrality of the “publish-or-perish” paradigm makes scientific publishers a crucial third party for the academic profession. Publishers’ concern about bibliometric indicators and rankings is relevant to journals’ marketability. However, it often contrasts with intrinsic research quality. For example, in a provocative commentary article, Kirchherr (2022) refers to 50% of publications in the field of sustainability as academic bullshit where “low-quality articles” are published to enhance journals’ metrics. Besides quality concerns, two facts exacerbate the dystonia between publishers’ interests and academic values. On the one hand, publishers are responsible for appointing journal editors and certifying the “quality” of scientific research through their journals’ prestige and bibliometric indicators. On the other hand, expensive subscriptions to scientific publications can restrict access and harms individuals’ careers. In this paper, we explore how the experience of a breach of academic values (i.e. a loss of faith in knowledge and science), formally an ideology-based psychological contract breach, predicts deviant behaviour toward scientific publishers, as the use of Sci-Hub. We examine it after controlling for common antecedents of digital piracy and trivial explanations for using Sci-Hub (i.e. to save time and for lack of journal access). Additionally, while existing literature on academics’ copyright violation and deviant behaviour focuses on either one scientific discipline (Boudry et al., 2019a; Karabag & Berggren, 2012; Mejia et al., 2017; Necker, 2014), one country (Dui´ c et al., 2017; Meyer & McMahon, 2004), or small samples without systematic survey strategies (Fanelli, 2009; Nicholas et al., 2017, 2019), we overcome these data limitations using new survey data on 2849 academics in 30 top institutions in 6 European countries (i.e. Germany, Hungary, Ireland, Italy, The Netherlands, and Sweden) collected in 2021. Our results indicate that scholars experiencing a contract breach of academic values provoking academic discontent are more likely to use Sci-Hub. Additionally, we find that its relative magnitude changes depending on contextual and demographic moderating factors. Our results underline that demographic characteristics and job security moderate the link between academic discontent and Sci-Hub use. These results imply that any policy aiming to prevent deviant behaviour against publishers must simultaneously address access needs, contextual workplace characteristics, and the consequences of increasing dystonia between copyright 3 and academic values. This paper makes three contributions which provide a fresh and original perspective on mild DWBs. First, we contribute to the theory by combining the organisational psychology literature with the economics of science literature. Relying on both allows underpinning better the peculiarity of academics and the relevance of specific ideological values (i.e. academic values). Second, we provide novel empirical evidence on the diffusion of mild DWBs across disciplines, universities, and countries. We also examine the role of emotions connected to a breach of ideology-based psychological contracts in eliciting mild deviant behaviours. Third, our results inform administrators and policymakers on policies targeting DWBs prevention, specifically academic digital piracy. Identifying DWB triggers and more resilient workers help to design policy interventions at any level better. The paper is structured as follows: Section 2 discusses the theoretical framework and hypothesis development. Section 3 describes the data, the variables and the method, and Sections 4 and 5, respectively, present and discuss the empirical findings. Conclusions will follow. 2. Theoretical Framework and Hypothesis Development From the theoretical perspective, we combine two separate streams of literature. First, we use the organisational psychology literature, which examines the role of psychological contract breaches in eliciting deviant behaviours. Second, we leverage the economics of science literature to account for the academic’s specificities and highlight the ideological dimension of the implicit contract in academic research. The frustration-aggression model (Fox & Spector, 1999), and more recent empirical studies on deviant behaviour (K. Lee & Allen, 2002; Xu, Luo, & Hsu, 2020) emphasise that negative emotions (or feelings) related to the workplace increase the likelihood of deviant behaviours. Such emotions might develop from a breach of the implicit psychological contract that workers form with their organisation (Conway & Briner, 2009). Past research examines how the perception of a contract breach negatively affects employees’ attitudes and behaviours towards the employing organisation, internal parties, and external parties (J. A.-M. CoyleShapiro et al., 2019). Regarding attitudes, the literature indicates that contract breach decreases: employees’ affective organisational commitment (Restubog, Bordia, & Tang, 2006), organisational trust (Robinson & Wolfe Morrison, 2000), public sector commitment (Conway, Kiefer, Hartley, & Briner, 2014), job satisfaction (Conway, Guest, & Trenberth, 2011), perceived organisational support (J. A. Coyle-Shapiro & Conway, 2005), organisational identification (Zagenczyk et al., 2013) and increases turnover intention (Orvis, Dudley, & Cortina, 2008), mistrust (Ann Feldheim, 2007), and organisational cynicism (Johnson & O’LearyKelly, 2003). Similarly, the experience of contract breach negatively affects employees’ behaviour reducing performance (S. P. Costa & Neves, 2017), organisational citizenship behaviour (Restubog, Bordia, Tang, & Krebs, 2010) and increasing deviance (Bordia, Restubog, & Tang, 2008), absenteeism (Deery, Iverson, & Walsh, 2006), turnover (Karagonlar, Eisenberger, & Aselage, 2016) and negative behaviour towards clients or external users (Conway et al., 2014; Deng, Coyle-Shapiro, & Yang, 2018). While the implicit contract between employer and employees certainly relates to economic and professional aspects, some professions entail an ideological element defined as a credible commitment of workers to pursue a valued cause beyond the self-interest and intrinsic to the profession (Bunderson, 2001; O’Donohue et al., 2007; Thompson & Bunderson, 2003). The literature highlights this possibility; however, only a few studies looked at the consequences of such specific psychological contract breaches, failing to account for the specific characteristics of each profession (J. A.-M. Coyle4 Shapiro et al., 2019). The academic profession is characterised by a shared adherence to academic values and beliefs that differ from other “knowledge workers” (Siekkinen et al., 2020). Examining 5000 US life scientists and physical scientists working either in the industry or in academia, Sauermann and Stephan (2013) show that academics have a different value system compared with corporate scientists testifying the existence of academic values. Academic values are implicit, unspoken values, beliefs, and rules about the academic profession. In contrast to commercial values, individuals endowed with academic values “prefer” the quest for fundamental knowledge and curiosity-driven research upon applied incremental research, research freedom over bureaucratic control, peer recognition over monetary rewards, and open disclosure of research results in the form of publication over patent activity (Agarwal & Ohyama, 2013; Merton, 1973; Sauermann & Stephan, 2013). Unlike other types of work, those who choose an academic career are willing to “pay to be a scientist” as they accept lower wages compared to corporate scientists that might experience publishing restrictions (Sauermann & Roach, 2014; Stern, 2004,?).3The diffusion of such values and the willingness to earn less to adhere to them indicates that the implicit contract characterising the academic job has a strong ideological dimension. In the case of academics, beliefs concerning their job relate to terms and conditions (such as duties and rights, job security, career development, and work-life balance) and loyalty to academic value. Any systematic dyscrasia between academics’ experience and expectations about both dimensions breaches the contract between an academic and its institution. Increasing performance-based managerialism, bureaucratisation, university market-like behaviour (Chatelain-Ponroy et al., 2018; Siekkinen et al., 2020; Walsh & Lee, 2015), as well as changes in sizes and the shift to short-term contracts (Bryson, 2004; Cyranoski, Gilbert, Ledford, Nayar, & Yahia, 2011; Hakala, 2009), have enormously changed the academic work. Whether these changes are good or bad, they might have altered the inner perception of academics regarding their job and the profession’s core values, producing more significant responses than general administrative psychological contract breach (Bunderson, 2001). A significant change in the academic profession is the diffusion of the “publish or perish” paradigm. This principle makes scholars more concerned about the publishability of their results, affecting topic choices and reducing scholars’ freedom in pursuing curiosity-driven research. This reduced freedom undermines one of the scientific research’s fundamental principles, harming academic values. A secondary but not less important consequence of the emergence of the “publish or perish” culture is the identification of scientific publishers as a relevant third party, contributing to jeopardising academic values. While scientific publishers restrict access to science and exploit academic free work (i.e. editors and referees often work for free), scholars have little choice about their publishing outlets as scientific publishers act as the certifiers of research quality. Thus, the experience of a contract breach derived from undermining academic values related to the emergence of the “publish or perish” paradigm triggers a reaction towards scientific publishers. Therefore, we hypothesise: H1: Academics perceiving a contract breach because of academic discontent are more likely to violate copyright using Sci-Hub. 3Based on a sample of 164 multiple job offers received by 66 Ph. D.s in Biology, Stern (2004) estimates that scientists accept a wage 14,000$ lower to have the freedom to publish their results. 5 2.1. Moderating Factors While we expect that perceiving an ideology-based contract breach relates positively to deviant behaviour against publishers, some employees’ characteristics might moderate the relation. A large body of literature highlights gender discrimination in the workplace in general (Coffman, Exley, & Niederle, 2021; Heilman & Caleo, 2018; Snizek & Neil, 1992; Trentham & Larwood, 1998) and academia in particular. Female scholars are underrepresented across fields and job ranks (Kahn & Ginther, 2017; Rossello, 2021), less likely to be promoted (De Paola & Scoppa, 2015), are paid less (Barbezat & Hughes, 2005) and obtain lower recognition from co-authorship (Sarsons, 2017). The presence of gender stereotypes and discrimination affects women’s careers and often makes academia a male-dominated environment. However, the women who make it through might have developed strategies to cope with such environment. Recent theoretical work explores the link between female discrimination and resilience (Bridges, Wulff, & Bamberry, 2021). It highlights that females with high individual resilience are more likely to thrive in a male-dominated environment. Thus, females pursuing an academic career might be used to negative experiences in the workplace and resilient in response to academic discontent. This widespread resignation to the current “status quo” makes women decouple their career expectations from their publication performance. In a nutshell, while aware of likely “perish”, women are more indifferent to the “publish or perish” paradigm. This unconcern is supported by the empirical evidence that women publish fewer articles than males colleagues (Rossello, Cowan, & Mairesse, 2020) and are less interested in competition (Dato & Nieken, 2014). Furthermore, a recent meta-analysis confirms that, in general, females are less likely to engage in CWB (Ng, Lam, & Feldman, 2016), and less involved in DWBs such as workplace aggression (Hershcovis et al., 2007), sabotage (Dato & Nieken, 2014), and digital piracy (Dui´ c et al., 2017; Mejia et al., 2017). Thus, we hypothesize: H2 Female academics perceiving a contract breach because of academic discontent are less likely to violate copyright using Sci-Hub than male colleagues. Foreign workers are part of a minority in the workplace and, similarly to females, experience discrimination (Aguirre, 2020; Dupree & Boykin, 2021). The theoretical link between discrimination and resilience applies to workers from a minority too (Bridges et al., 2021). Thus, the foreign workers that continue their careers at universities might have developed high individual resilience to thrive in a context dominated by white males. Additionally, they might be less susceptible to the “publish-or-perish” paradigm for their attitudes towards competition and topic choices. On the one hand, ethnic minorities are less individualistic and responsive to competition than white males (Cox, Lobel, & McLeod, 1991). And on the other hand, they are interested in topics systematically found less likely to be funded (Hoppe et al., 2019) or published (Zeina, Balston, Banerjee, & Woolf, 2020). Past experiences of discrimination can make foreign scholars more resilient to academic discontent, moderating the link between ideology-based psychological contract breaches and deviant behaviour. Thus: H3 Foreign academics perceiving a contract breach because of academic discontent are less likely to violate copyright using Sci-Hub than non-foreign colleagues. Besides gender and nationality, some context-specific features, such as job insecurity, play a role in strengthening the relationship between psychological contract breach and deviant 6 Figure 1 shows the intensity of Sci-Hub usage last year. More than half are non Sci-Hub users, while among those who declared they ever used it, the 97% used it at least once last year and 45% are frequent users (they used it more than 10 times). For this reason we dicotomize our dependent variable and focus the analysis on Sci-Hub usage rather than the intensity of use. Figure 2. Distribution across country (bottom-left) and ERC scientific fields (bottom-right) of the intensity of Sci-Hub use in the last year. Figure 2 displays the distribution of the intensity of Sci-Hub usage last year across countries and academic fields. The figures indicate similar patterns to the previous table, indicating higher country-level variance. These two observations combined suggest that the characteristics of the university system matter more than field differences in explaining the use of Sci-Hub and its intensity. Table 4 reports the estimation results of equations 1 and 2. Column 1 shows regression results including only the controls and the moderators, whereas column 2 including only our main variable IDEOLOGY BASED BREACH. In column 3, we report the results for estimating equation 1 and test H1. The model in column 3 indicates that the experience of an ideology-based contract breach IDEOLOGY BASED BREACH=1) corresponds to a statistically significant increase in the likelihood of using Sci-Hub. Considering the odds ratios, the estimated model indicates a sizable effect. All else being equal, those who experienced a contract breach generating IDEOLOGY BASED BREACH have 75% more odds of using Sci-Hub compared to those who do not experience it (IDEOLOGY BASED BREACH=0). This finding confirms our H1. Looking at the controls and the moderators models 1-2 report results consistent with the existing literature. The likelihood of using Sci-Hub is lower for females (FEMALE) and faculty members (FACULTY), whereas it is higher for foreign scholars (FOREIGN). The likelihood of using Sci-Hub decreases the higher the perceived quality of the genuine product, expressed by the satisfaction individuals report about library services of their institution (LIBRARY SATISFACTION). Several controls have the expected positive sign and drive Sci-Hub use. These are individual moral justification for breaking copyright law (MORAL JUSTIFICATION), scientific publishers’ business model perception as unethical (UNETHICAL PUBLISHERS), excessive teaching load (TEACHING LOAD), and positive social perception surrounding piracy (COLLEAGUES PIRACY PERCEPTION). On the contrary, the likelihood of using Sci-Hub is higher for experienced users (PAST PIRACY), and knowledgeable scholars about copyright (COPYRIGHT KNOWLEDGE). Results in columns 4-6 of table 4 test the moderating effects stated in H2-H4. Since we are estimating a nonlinear model, interpretation of the interaction effects cannot be based only on 13 Table 4. Main Specification and Models with Moderating Effects. The dependent variable is a dummy equal to 1 if the respondent used Sci-Hub in the past and 0 otherwise. Dependent variable: USE SCI-HUB (1) (2) (3) (4) (5) (6) FACULTY -0.835*** -0.802*** -0.802*** -0.808*** -0.753*** (0.0959) (0.0965) (0.0965) (0.0967) (0.100) FEMALE -0.147 -0.165* -0.159 -0.167* -0.163* (0.0960) (0.0965) (0.102) (0.0966) (0.0966) FOREIGN 0.345*** 0.347*** 0.346*** 0.407*** 0.343*** (0.122) (0.123) (0.123) (0.128) (0.123) COPYRIGHT KNOWLEDGE 0.141*** 0.141*** 0.141*** 0.142*** 0.142*** (0.0215) (0.0215) (0.0216) (0.0216) (0.0215) MORAL JUSTIFICATION 0.369*** 0.365*** 0.365*** 0.366*** 0.367*** (0.0384) (0.0384) (0.0384) (0.0385) (0.0385) LIBRARY SATISFACTION -0.0821*** -0.0817*** -0.0816*** -0.0811*** -0.0825*** (0.0292) (0.0291) (0.0291) (0.0292) (0.0292) PAST PIRACY 0.121*** 0.121*** 0.121*** 0.121*** 0.122*** (0.0201) (0.0201) (0.0201) (0.0200) (0.0201) COLLEAGUES PIRACY PERCEPTION 0.148*** 0.148*** 0.148*** 0.148*** 0.149*** (0.0236) (0.0238) (0.0238) (0.0238) (0.0238) TEACHING LOAD 0.342** 0.351** 0.350** 0.348** 0.346** (0.159) (0.160) (0.160) (0.160) (0.160) INSTITUTIONAL TRAINING 0.0414 0.0406 0.0405 0.0395 0.0399 (0.0414) (0.0415) (0.0415) (0.0415) (0.0415) UNETHICAL PUBLISHERS 0.301*** 0.292*** 0.292*** 0.294*** 0.289*** (0.0925) (0.0927) (0.0927) (0.0927) (0.0927) IDEOLOGY BASED BREACH=1 0.746*** 0.560*** 0.591*** 0.682*** 0.792*** (0.133) (0.150) (0.212) (0.167) (0.201) IDEOLOGY BASED BREACH=1 X FEMALE=1 -0.0652 (0.300) IDEOLOGY BASED BREACH=1 X FOREIGN=1 -0.660* (0.358) IDEOLOGY BASED BREACH=1 X FACULTY=1 -0.591* (0.313) COUNTRY Yes No Yes Yes Yes Yes ERC Yes No Yes Yes Yes Yes UNIVERSITY Yes No Yes Yes Yes Yes OBSERVATIONS 2701 2849 2701 2701 2701 2701 Notes: Models are estimated using a logit model and the coefficients are reported. Robust standard errors in parentheses. Legend: ∗p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01 the signs and significance of the coefficients (Ai & Norton, 2003), and we plot the marginal effects. H2 predicts that females experiencing IDEOLOGY BASED BREACH are less likely to use Sci-Hub. Column 4 in table 4 and the marginal effects of the predicted probability in figure 3 indicate we cannot confirm the hypothesis. Overall, we can observe that the interaction term of FEMALE and IDEOLOGY BASED BREACH is negative but not different from zero at 10% significance level. Figure 3 shows that the probability of using Sci-Hub is higher for both males and females experiencing IDEOLOGY BASED BREACH and males appear more responsive to IDEOLOGY BASED BREACH than females. Column 5 in table 4 tests H3 about the moderating role of being foreign. The result supports H3 where the interaction term of IDEOLOGY BASED BREACH and FOREIGN is negative and significant at 10% significance level. Looking at the marginal effects in figure 4, we can observe that the probability that non-foreign academics use Sci-Hub is higher for those experiencing IDEOLOGY BASED BREACH. At the same time, foreigners are equally likely to use Sci-Hub irrespective of their experience of IDEOLOGY BASED BREACH. In other 14 Figure 3. Marginal effects IDEOLOGY BASED BREACH. Model in column 3 table 4 testing hypothesis H2. The x-axis shows FEMALE=1 and MALE=0 while the y-axis is the predicted probability of Using Sci-Hub words, non-foreigners academics are more responsive to IDEOLOGY BASED BREACH than foreigners. Finally, the model in column 6 of table 4 shows support for H4, where we test the moderating effect of having a tenure. The interaction term coefficient is negative and significant at 10% significance level. The marginal effects in figure 5 provide additional details and help interpret the effect. The probability of using Sci-Hub is higher for non-faculty members experiencing IDEOLOGY BASED BREACH, but the probability that faculty members use Sci-Hub is unresponsive to IDEOLOGY BASED BREACH. We can conclude that non-faculty members are more responsive to IDEOLOGY BASED BREACH than faculty members. 15 Figure 4. Marginal effects IDEOLOGY BASED BREACH. Model in column 4 table 4 testing H3. The x-axis shows FOREIGN=1 and NON-FOREIGN=0 while the y-axis is the predicted probability of Using Sci-Hub 16 Figure 5. Marginal effects IDEOLOGY BASED BREACH. Model in column 5 of table 4 test H4. The x-axis shows FACULTY=1 and NON-FACULTY=0 while the y-axis is the predicted probability of Using Sci-Hub 17 4.1. Robustness Checks: Alternative Measures of Ideology-Based Psychological Contract Breach We investigate our results’ robustness using two alternative measures to proxy an ideologybased psychological contract breach related to academic values. We account for the potential drawback of using a dichotomous variable combining the 13 items described in section 3.2. First, we create a composite numeric index aggregating the items specific to the academic profession representing a breach of academic values. We call the variable ACADEMIC BREACH that is the sum of the items (1) teaching responsibilities; (2) administrative responsibilities; (4) being unable to concentrate on my research; (8) the inadequate facilities or funding; (10) the feeling that sometimes my research is a waste of time; (11) the feeling that sometimes my research is a waste of public money; (12) It undermined my confidence in knowledge and science; selected by respondents. The variable ranges from 0 (none of the items is selected) to 7 (all items are selected). Variable PC1 PC2 PC3 PC4 Junior Colleagues Senior Lack of Discontent Behavior Discontent Time TEACHING LOAD -0.0738 0.2008 0.0771 0.4475 ADMINISTRATIVE LOAD -0.3253 0.1915 0.1833 0.3655 COMPETITION LOAD 0.3943 -0.0041 -0.3180 0.2102 LACK RESEARCH TIME 0.0654 0.2182 0.1853 0.6056 CAREER STRESS 0.3974 -0.1324 -0.3459 0.1511 BEHAVIOR JUNIOR COLLEAGUES 0.0815 0.4816 0.1567 -0.2452 BEHAVIOR SENIOR COLLEAGUES 0.1847 0.5556 -0.0257 -0.1489 LACK FUNDING FACILITIES -0.0528 0.3263 -0.2386 0.1428 HOMESICKNESS 0.2303 -0.0501 -0.4243 0.1855 RESEARCH WASTE TIME 0.4745 -0.0676 0.3620 0.0108 RESEARCH WASTE MONEY 0.3667 -0.1098 0.4738 -0.0626 IDEOLOGY BASED BREACH 0.3153 0.0713 0.2264 0.1097 LACK RECOGNITION 0.1246 0.4320 -0.1907 -0.2815 Eigenvalues 1.90 1.31 1.21 1.13 Proportion of variance 0.15 0.10 0.09 0.09 Cumulative prop 0.15 0.25 0.34 0.43 Table 5. Results of PCA on the 13 variables representing the negative aspects of being an academic described in section 3.2 Second, we run a Principal Component Analysis (PCA) on the 13 items, obtaining four main components. We select the four components with eigenvalues above one, and that explain most of the variance (see bottom of Table B and Figure B1 in Appendix ). Table B shows details of the composition of the four components. We consider the first component as the main one representing an ideology-based psychological breach of academic values because it is the one with the highest correlation with the item IDEOLOGY BASED BREACH and related variables (RESEARCH WASTE TIME,RESEARCH WASTE MONEY. We call it JUNIOR DISCONTENT because it is also positively and strongly correlated with variables affecting more young non-tenured researchers (e.g. career stress, competition load) while negatively or mildly related to those more likely to impact senior faculty members (e.g. teaching and administrative load, lack of research time). We call the second component COLLEAGUE BEHAVIOUR because it positively correlates with discontent stemming from colleagues’ behaviour capturing the peer pressure effect. The third component, SENIOR DISCONTENT is the counterpart of the JUNIOR DISCONTENT as it correlates both with the variables representing an ideology-based psychological contract breach and with variables more likely to affect senior tenured scholars negatively. The fourth component, LACK OF TIME, has the highest positive correlation with the variables representing a contract breach related to the lack of time. Table 6 shows the results for equations 1 and 2 using the two broader definitions of the psychological contract breach of academic values defined above. Table 6 confirms our previous results showing the strong effect of an ideology-based psychological contract breach related to academic values in increasing the likelihood of scholars using Sci-Hub. 18 In particular, column 1-4 indicates that one unit increase of ACADEMIC BREACH increased the odds of using Sci-Hub between 17% and 25%. Looking at the interaction terms in columns 2 and 3 of the variable with the moderators FEMALE and FOREIGN, we find that they are both negative and different from zero at 1% significance level, supporting H2 and H3. Also, the variable JUNIOR DISCONTENT shows a sizable effect in affecting the use of Sci-Hub. Columns 5-8 show that we will see between the 25% and the 33% of an increase in the odds of using Sci-Hub for a one-unit increase in JUNIOR DISCONTENT. Additionally, the interaction of the variable with the moderators FEMALE and FACULTY (columns 7-8) provide support for H2 and H4. The coefficients are indeed both negative and statistically significant at 10% significance level. 19 Table 6. Robustness Check. The dependent variable is a dummy equal to 1 if respondent used Sci-Hub in the past and 0 otherwise Dependent variable: USE SCI-HUB (1) (2) (3) (4) (5) (6) (7) (8) FACULTY -0.855*** -0.844*** -0.867*** -0.662*** -0.693*** -0.689*** -0.690*** -0.691*** (0.0959) (0.0961) (0.0963) (0.166) (0.101) (0.101) (0.101) (0.101) FEMALE -0.165* 0.136 -0.172* -0.157 -0.219** -0.210** -0.223** -0.214** (0.0963) (0.170) (0.0967) (0.0968) (0.0978) (0.0975) (0.0980) (0.0982) FOREIGN 0.352*** 0.350*** 0.868*** 0.347*** 0.309** 0.303** 0.322** 0.315** (0.123) (0.123) (0.211) (0.123) (0.125) (0.126) (0.125) (0.126) COPYRIGHT KNOWLEDGE 0.138*** 0.138*** 0.137*** 0.139*** 0.139*** 0.137*** 0.138*** 0.140*** (0.0215) (0.0215) (0.0215) (0.0215) (0.0218) (0.0219) (0.0218) (0.0219) MORAL JUSTIFICATION 0.367*** 0.369*** 0.370*** 0.367*** 0.364*** 0.365*** 0.364*** 0.361*** (0.0384) (0.0384) (0.0385) (0.0384) (0.0388) (0.0388) (0.0388) (0.0388) LIBRARY SATISFATION -0.0770*** -0.0785*** -0.0780*** -0.0777*** -0.0764*** -0.0745** -0.0765*** -0.0763*** (0.0292) (0.0291) (0.0292) (0.0293) (0.0293) (0.0293) (0.0293) (0.0294) PAST PIRACY 0.118*** 0.118*** 0.119*** 0.119*** 0.117*** 0.117*** 0.118*** 0.118*** (0.0201) (0.0201) (0.0201) (0.0201) (0.0202) (0.0202) (0.0202) (0.0203) COLLEAGUES PIRACY PERCEPTION 0.146*** 0.147*** 0.143*** 0.146*** 0.142*** 0.143*** 0.142*** 0.142*** (0.0237) (0.0237) (0.0238) (0.0238) (0.0240) (0.0239) (0.0240) (0.0240) TEACHING LOAD 0.149 0.155 0.155 0.175 0.127 0.123 0.123 0.131 (0.168) (0.169) (0.167) (0.171) (0.197) (0.197) (0.196) (0.196) INSTITUTIONAL TRAINING 0.0376 0.0371 0.0390 0.0378 0.0384 0.0364 0.0378 0.0366 (0.0416) (0.0415) (0.0417) (0.0415) (0.0419) (0.0419) (0.0419) (0.0418) UNETHICAL PUBLISHERS 0.298*** 0.294*** 0.307*** 0.295*** 0.273*** 0.267*** 0.275*** 0.274*** (0.0927) (0.0929) (0.0931) (0.0928) (0.0936) (0.0938) (0.0936) (0.0936) ACADEMIC BREACH 0.154*** 0.224*** 0.214*** 0.209*** (0.0407) (0.0518) (0.0454) (0.0570) JUNIOR DISCONTENT (PC1) 0.222*** 0.272*** 0.241*** 0.282*** (0.0362) (0.0459) (0.0400) (0.0509) COLLEAGUES BEHAVIOUR (PC2) 0.00560 0.00878 0.00626 0.00715 (0.0400) (0.0400) (0.0400) (0.0402) SENIOR DISCONTENT (PC3) -0.0260 -0.0275 -0.0285 -0.0337 (0.0411) (0.0412) (0.0412) (0.0414) LACK OF TIME (PC4) 0.168*** 0.169*** 0.171*** 0.163*** (0.0486) (0.0487) (0.0485) (0.0487) FEMALE=1 X ACADEMIC BREACH -0.173** (0.0790) FOREIGN=1 X ACADEMIC BREACH -0.307*** (0.0996) FACULTY=1 X ACADEMIC BREACH -0.115 (0.0795) FEMALE=1 X JUNIOR DISCONTENT (PC1) -0.117* (0.0700) FOREIGN=1 X JUNIOR DISCONTENT (PC1) -0.0905 (0.0878) FACULTY=1 X JUNIOR DISCONTENT (PC1) -0.128* (0.0727) COUNTRY Yes Yes Yes Yes Yes Yes Yes Yes ERC Yes Yes Yes Yes Yes Yes Yes Yes UNIVERSITY Yes Yes Yes Yes Yes Yes Yes Yes OBSERVATIONS 2701 2701 2701 2701 2701 2701 2701 2701 Notes: Models are estimated using a logit model and the coefficients are reported. Robust standard errors in parentheses. Legend: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01 20 5. Discussion Our results highlight the relationship between ideology-based psychological contract breach and deviant behaviour towards relevant academic third parties. While using Sci-Hub does not directly harm universities, they should prevent it because of the associated reputational and technical costs. Besides the potential security concern that illegal downloads pose to their digital infrastructure13, Sci-Hub usage is pretty unnoticed by universities. Illegal access to scientific literature through Sci-Hub can be considered a mild deviant behaviour. However, we document its pervasiveness as almost half of the scholars in top European universities download papers illegally using Sci-Hub. This broad diffusion suggests acceptance of such deviant behaviour beyond reasons of “simple” journal access. While transgressions towards third parties might be easily overlooked, they diffuse quickly. They might precede the emergence of a slippery slope causing more serious deviant behaviours. Focusing attention on such minor transgressions and designing effective prevention policies could be effective in avoiding subsequent major transgressions. Our results suggest that, while designing policies preventing academic digital piracy, administrators and policymakers should consider the role of academic values and the connected intrinsic ideological currency so crucial for academics. Any university policy attempting to restore academic values (i.e. promoting freedom of research, releasing pressure from bibliometric indicators) might produce a policy spillover into preventing academic piracy. To date, there are no systematic policies to restore academic values or decrease the emphasis on bibliometric indicators. However, the economics of science literature has started investigating at least three intervention areas to lift some “publish or perish” pressure. These are i) moving beyond bibliometric measures, ii) incentives for funding risky research, and iii) the PhD job market. Concerning alternative metrics, a large stream of research highlighted the need for developing metrics based on better data (Mas-Bleda & Thelwall, 2016; Molas-Gallart & R` afols, 2018). Many advocates for using better algorithms to address problematic issues of bibliometric databases such as name disambiguation (Han, Giles, Zha, Li, & Tsioutsiouliklis, 2004; Sanyal, Bhowmick, & Das, 2021), self-citations (Schreiber, 2007; Szomszor, Pendlebury, & Adams, 2020), citations meaning (Budi & Yaniasih, 2022), and authors’ contribution (Shen & Barab´ asi, 2014). Others advocate for incorporating job-market-based measures in university rankings (Cowan & Rossello, 2018; Gonz´ alez-Sauri & Rossello, 2022; Wapman, Zhang, Clauset, & Larremore, 2022). Beyond the scientific effort to correct biases in diffused bibliometric indicators, publishers still retain a central role in developing better metrics. Their improvements and a more critical use accounting for their limitations can restore more trustworthy relations between academic management, publishers, and scholars (Biagioli, 2020). Utrecht University is undergoing an experiment to reduce the emphasis on bibliometric indicators. In June 2021, the University formally abandoned the journal impact factor metric when making faculty hiring and promotion decisions. While it will take some time to evaluate all the effects of this bold decision, some scholars are already sceptical about the decision as the University has yet to identify an alternative measure of scientific performance (Singh Chawla, 2021). Concerning funds for risky research, the economics of science literature (Franzoni, Stephan, & Veugelers, 2022; Wang, Veugelers, & Stephan, 2017) has highlighted funding 13See, for example, the article of Alexander Martin, a Technology reporter for Sky News. Available at https:// news.sky.com/story/police-warn-students-and-universities-against-using-the-pirate-bay-of -science-12250407; last access November 2022 21 agencies’ risk aversion and bias against novelty. Established in 2007, ERC grants should fund curiosity-driven (i.e. characterised by high risk and high gain) research; however, their implementation systematically disadvantages novel and high-risk proposals of young scholars (Franzoni et al., 2022). Incentivising curiosity-driven research through their funding might help to restore academic values. Comparing two funding schemes in life sciences (NIH vs HHMI), Azoulay, Graff Zivin, and Manso (2011) find that funding schemes and incentives matter. The HHMI funding, which emphasises research freedom and creativity, having a longterm focus, promoting intellectual experimentation, and providing feedback, generates more breakthrough innovations than traditional funding (NIH). Moreover, the effect on innovative performance is significant, as predicted by the idea that academic values matter to scholars. Indeed, research freedom and the ability to pursue curiosity-driven research are valuable for scholars and affect their behaviour, performance and eagerness towards CWB. Finally, concerning PhD job market and early career employment, the field of economics provides a test bed for trying to overcome early bibliometric assessment (Coles et al., 2010).14 The economics job market does not consider bibliometric indicators for hiring young scholars. Instead, the Faculty evaluate candidates’ potential by reading their “job market paper”, which is often not published and, therefore, peer-reviewed. Such policy attenuates the emphasis on quantitative bibliometrics indicators and an early quest for publishability. However, this recruitment mechanism is limited at the early career phase and experiments such as the one by Utrecht University are still to come. Even if not strictly related to workplace policies evaluation, our findings help scientific publishers and universities design measures against academic piracy besides relying on copyright law sanctions threats. We highlight the link between a deterioration of academic values and digital piracy, hinting at novel ways to address academic piracy. Beyond improving access to scientific resources, diminishing the emphasis on bibliometric indicators, promoting risky and curiosity-driven research, and developing new career evaluation tools will decrease the likelihood of experiencing an ideology-based psychological contract breach of academic values and break the negative spiral between academic discontent and deviant behaviour. Finally, our investigation of moderating factors also contributes to designing effective prevention policies by identifying worker characteristics that impact the likelihood of engaging in academic piracy. Historically marginalised categories in the job market and academia respond differently to ideology-based contract breaches. Their experience of discrimination and resilience when experiencing academic discontent makes these groups less sensitive and prone to deviant behaviour. These different responses to the academic discontent call for targeting specific prevention policies for these categories. For example, a piracy prevention policy accompanied by an inclusion policy promoting the participation of women and foreigners in academia might generate an environment less prone to such behaviours. Finally, our results indicate specific patterns for non-tenured scholars that should guide targeted prevention policies. Any policy designed to reduce job insecurity (e.g. career counselling, planning of staff turnover) might help prevent such forms of mild deviant behaviours. 6. Conclusion The academic profession has changed dramatically in the last decades. The increase of bureaucratic control, the implementation of performance-based evaluation measures, the pres14The last report about the US PhD job market in Economics is available in (Cawley et al., 2022) 22 Zagenczyk, T. J., Cruz, K. S., Woodard, A. M., Walker, J. C., Few, W. T., Kiazad, K., & Raja, M. (2013). The moderating effect of machiavellianism on the psychological contract breach– organizational identification/disidentification relationships. Journal of Business and Psychology, 28(3), 287–299. Zaghini, F., Fida, R., Caruso, R., Kangasniemi, M., & Sili, A. (2016). What is behind counterproductive work behaviors in the nursing profession? a systematic review. Journal of Clinical Research & Bioethics,7(4). Zeina, M., Balston, A., Banerjee, A., & Woolf, K. (2020). Gender and ethnic differences in publication of bmj letters to the editor: an observational study using machine learning. BMJ open,10(12), e037269. 29 Appendix A. List of universities Name N. emails Country City Founded THE Rank Res Rank Cit Rank N. Students Student/Staff % Int Students Soc. Sci. Hum. Life Sci. Eng. Med. Heidelberg University 9597 Germany Heidelberg 1386 42 71 52 20,020 14.5 18% 1 1 1 1 1 Humboldt University of Berlin 4633 Germany Berlin 1810 80 54 428 33,388 56.3 18% 1 1 1 1 LMU Munich 1996 Germany Munich 1472 32 34 161 34,249 33.6 17% 1 1 1 1 Technical University of Munich 4793 Germany Munich 1868 41 42 209 32,377 39.8 31% 1 1 1 1 University of T¨ ubingen 4503 Germany Tubingen 1477 78 80 242 27,590 36.6 14% 1 1 1 1 Budapest University of Technology and Economics 2635 Hungary Budapest 1782 1001+ 958 1169 19,415 17.7 11% 1 1 E¨ otv¨ os Lor´ and University 2687 Hungary Budapest 1635 601–800 747 768 27,199 14.8 10% 1 1 1 University of Debrecen 1084 Hungary Debrecen 1538 801–1000 1201 1038 26,938 14.1 21% 1 1 1 1 1 Delft University of Technology 7065 Netherlands Delft 1842 78 72 60 19,594 17.1 31% 1 1 1 1 1 University of P´ ecs 2657 Hungary Pecs 1912 601–800 1183 844 16,798 11.7 24% 1 1 1 1 1 University of Szeged 1544 Hungary Szeged 1872 801–1000 976 878 18,859 12.1 20% 1 1 1 1 1 Maynooth University 1489 Ireland Maynooth 1997 401–500 405 456 10,107 28.6 12% 1 1 1 1 National University of Ireland, Galway 1187 Ireland Galway 1845 301–350 336 405 14,453 26.6 18% 1 1 1 1 1 Trinity College Dublin 5437 Ireland Dublin 1592 155 169 296 17,154 21.4 32% 1 1 1 1 1 University College Cork 1423 Ireland Cork 1845 301–350 544 230 17,051 17.6 19% 1 1 1 1 1 University College Dublin 1633 Ireland Dublin 1854 251–300 245 353 23,148 23.2 30% 1 1 1 1 1 Sapienza University of Rome 6566 Italy Roma 1303 201–250 202 424 77,496 22.9 7% 1 1 1 1 1 University of Bologna 11761 Italy Bologna 1088 167 263 221 67,298 24.4 12% 1 1 1 1 1 University of Milan 8602 Italy Milano 1924 351–400 399 362 45,752 20.9 6% 1 1 1 1 1 University of Padua 3252 Italy Padova 1222 251–300 321 289 46,900 21.7 7% 1 1 1 1 1 University of Trento 3793 Italy Trento 1962 301–350 339 333 13,180 20.1 7% 1 1 1 Erasmus University Rotterdam 2224 Netherlands Rotterdam 1913 72 91 41 26,453 26.8 22% 1 1 1 1 Leiden University 4158 Netherlands Leiden 1575 70 60 214 30,178 19 18% 1 1 1 1 University of Amsterdam 7337 Netherlands Amsterdam 1632 66 69 102 24,747 12.8 24% 1 1 1 Utrecht University 6252 Netherlands Utrecht 1636 75 73 141 32,022 14 10% 1 1 1 Karolinska Institute 191 Sweden Stockholm 1810 36 47 66 7,696 8.8 26% 1 1 1 1 Lund University 9214 Sweden Lund 1666 103 100 202 27,443 11.3 19% 1 1 1 Stockholm University 3879 Sweden Stockholm 1878 183 130 235 27,200 19.1 10% 1 1 1 University of Gothenburg 2726 Sweden Gothenburg 1891 191 210 94 19,616 9.8 13% 1 1 1 1 Uppsala University 6115 Sweden Uppsala 1477 111 98 212 25,112 15.8 18% 1 1 1 1 1 Table A1. List of Universities 30 Appendix B. Principal components analysis Figure B1. Screeplot of the Principal Component Analysis on the 13 items, with 95% heteroskedastic bootstrap confidence intervals. 31 The UNU-MERIT WORKING Paper Series 2023-01 Can international mobility shape students' attitudes toward inequality? The Brazilian case by Cintia Denise Granja, Fabiana Visentin and Ana Maria Carneiro 2023-02 Demand-led industrialisation policy in a dual-sector small open economy by Önder Nomaler, Danilo Spinola and Bart Verspagen 2023-03 Reshoring, nearshoring and developing countries: Readiness and implications for Latin America by Carlo Pietrobelli and Cecilia Seri 2023-04 The role of product digitization for productivity: Evidence from web-scraping European high-tech company websites by Torben Schubert, Sajad Ashouri, Matthias Deschryvere, Angela Jäger, Fabiana Visentin, Scott Cunningham, Arash Hajikhani, Lukas Pukelis and Arho Suominen 2023-05 More than a feeling: A global economic valuation of subjective wellbeing damages resulting from rising temperatures by Stephan Dietrich and Stafford Nichols 2023-06 Was Robert Gibrat right? A test based on the graphical model methodology by Marco Guerzoni, Luigi Riso and Marco Vivarelli 2023-07 Predicting social assistance beneficiaries: On the social welfare damage of data biases by Stephan Dietrich, Daniele Malerba and Franziska Gassmann 2023-08 Related or unrelated diversification: What is smart specialization? By Önder Nomaler and Bart Verspagen 2023-09 Breach of academic values and digital deviant behaviour: The case of Sci-Hub by Giulia Rossello and Arianna Martinelli