1 Introduction

Publicly funded Western universities have progressively been subject to the governance principles of new public management (Biondi & Russo, 2022; Craig et al., 2014; Hood, 1995), which imply that they should be governed through responsible management, similar to a corporation. This governance paradigm is dominated by the notion of excellence (Nørreklit et al., 2019) and has progressively introduced drastic cuts in university public funding, leading to the establishment of national research authorities and formal research assessments. These exercises rank research quality and reward the excellence of researchers and universities based on the level of their publications (Brinn et al., 2001; Geuna & Martin, 2003; Hicks, 2009; Nicolò et al., 2020; Otley, 2010; Rebora & Turri, 2013; Turri, 2014). Therefore, universities aim to legitimize themselves as research-oriented institutions (Dobija et al., 2019), leading to widespread pressure to publish in the academic accounting community (Carmona, 2006; de Villiers & Dumay, 2014). In turn, the scientific prestige of a university influences private funding, with universities urging their researchers to become “academic performers” (Gendron, 2008). These developments have progressively attenuated the differences between many national university systems and the US context (de Lange et al., 2010; Hopwood, 2008; Paisey & Paisey, 2017; Palea, 2017), where a scholar’s tenure is closely linked to their capacity to produce high-quality publications (Read et al., 1998; Street et al., 1993).

Against a backdrop of “university commercialization” and “performatization,” researchers are increasingly seen as “machines of production” (Guthrie & Parker, 2014; Humphrey & Gendron, 2015), raising concerns about the multiple—often competing—logics (Conrath-Hargreaves & Wüstemann, 2019; Grossi et al., 2020a, 2020b; Pilonato & Monfardini, 2022) introduced by the spread of audit culture and public sector reforms. Governance tools increasingly use IT-based performance measures, such as journal rankings, to capture excellence, leading to broader implications than disciplining scholars and accelerating their scientific output (Nørreklit et al., 2019). A rich academic debate discusses the unintended consequences of the “publishing game” scenario (Broad, 1981) induced by journal rankings and their unbalanced use, which places growing importance on the idea that “you are where you publish” (Moizer, 2009). They include the endangerment of research diversity and innovativeness, threats to local research traditions and relevant niche accounting domains (e.g., public sector, accounting history, accounting education) (van Helden & Argento, 2020), and excessive competition among scholars, which could jeopardize the quality of teaching (Kallio et al., 2017).

Some authors argue that the original purpose of publishing to advance and disseminate knowledge risks being lost (Moizer, 2009), suggesting the danger of goal displacement with possible pernicious effects on researchers’ foci and strategies (Hopwood, 2007; Otley, 2010; Palea, 2017). A rich body of literature reveals critical issues and ambiguities related to university performance measurement (Argento et al., 2020; Vakkuri & Johanson, 2020), showing that linking financial incentives and career progression to these metrics has significant reorientation effects on individual behaviors and epistemic consequences (Guarini et al., 2020). The episode of scholars working in the area of critical studies ‘selected’ for redundancy by a UK business school (Andrew et al., 2021) is a striking case in point, provoking the shocked reaction of the international academic community to this attack on academic freedom (BAFA, 2021; Cho & Vollmer, 2021).

This study contributes to the discourse on publishing pressures by exploring the research topic choice of accounting researchers to ascertain whether and the ways in which these pressures influence this crucial decision. Shedding light on the first phase of the research process can increase our understanding of the dysfunctional effects of the “publishing game”, showing whether they affect what accounting researchers study, in addition to how they conduct research. The research topic choice can be seen as the first evidence of passionate scholarship, defined as “the commitment to a personally meaningful and socially relevant topic, ‘close to the heart’” (Courpasson, 2013, p. 1243). Choosing socially relevant topics reflects passionate scholarship, and the desire to publish is functional when scholars aim to share relevant findings. In contrast, the desire to publish to “score points” as an ultimate goal, regardless of topic relevance, represents a significant departure from the primary purpose of a passionate scholar.

This empirical study complements reflective arguments based on observations from accounting and the broader scholarly landscape (Gendron, 2008; Guthrie & Parker, 2014; Hopwood, 2008), raising concerns that the above-described goal displacement could involve not only the methods but also the topics to cover. The “publishing game” pressures could affect the research topic choice, given that researchers are aware that the probability of getting published in some journals is dependent not only on the quality of their research but also on the topic covered (Czyzewski and Dickinson, 1990). In addition, they can perceive some of the top journals as not as general as their mission statement states, as evidence suggests that they do not publish the most highly cited articles in some topic areas and methodologies (Summers & Wood, 2017).

The focus on the choice of research topic, which has never been investigated in the accounting domain, is also motivated by its considerable influence on the possible societal impact of research. Notable authors stress that addressing a compelling research question can produce socially and/or economically impactful and beneficial insights based on high-quality evidence when the research topic is also relevant outside academia (Broadbent & Unerman, 2011). In contrast, investigating irrelevant or unimportant issues outside academia may represent a significant barrier to research potentially informing practice (Baldvinsdottir et al., 2010; Leisenring & Johnson, 1994; Singleton-Green, 2010; Tucker & Schaltegger, 2016; van Helden, 2019), raising a lost before translation problem of knowledge transfer (Shapiro et al., 2007).

Hence, choosing research topics based solely on publishing opportunity risks increasing the long-standing gap between accounting research and practice (Baxter, 1988; Lee, 1989; Leisenring & Johnson, 1994; Zeff, 1989), as discussed in several studies belonging to different research paradigms (Broadbent, 2016; Dyckman & Zeff, 2015; Evans et al., 2011; Hopwood, 2007; Parker et al., 2011; Sinclair & Cordery, 2016; van Helden, 2019). This dysfunctional consequence would be grave because it could hinder the provision of reliable evidence informing the development of policy and practice for the benefit of society, which is one of the main aims of accounting and broader management research (Unerman, 2020). Indeed, opportunistic research topic choices could lead to a community of scholars becoming increasingly productive in terms of publications but researching less relevant and meaningful topics for society to speed up the publishing process (Humphrey & Gendron, 2015).

Choosing the object of investigation influences the outcomes of the research process and its impact on society. For this reason, we explore the research topic choice, not in isolation, but considering its connections with the broader research process, which we conceptualize building on the diffusion of innovation theory (Rogers, 2003). Following prior literature (Brownson et al., 2006; Gautam, 2008; Tucker & Parker, 2014; Tucker & Schaltegger, 2016), we consider the research process as a particular kind of innovation and apply the diffusion of innovation theory to analyze the translation of research into practice. Considering the translation into practice is relevant for our study, because a researcher’s view of its importance can influence the choice of the research topic. More specifically, we conceptualize two extremes, identifying an impact-oriented process (research process A) and a more opportunistic process driven only by publication objectives (research process B), without placing importance on the translation of research into practice. The risk of prioritizing issues following research process B is particularly relevant given the variety of dysfunctional outcomes of widespread audit culture (Craig et al., 2014) and the threat of goal displacement, with pernicious effects on researchers’ foci and strategies (Hopwood, 2007; Otley, 2010; Palea, 2017).

This study addresses whether and how “publishing game” pressure influences accounting researchers’ choice of research topic through an online questionnaire survey among European accounting researchers. The European setting has relevant common characteristics at a broad level, such as the considerable role of public university funding and the significant impact of austerity policies and research assessment. Hence, European researchers face similar challenges and intensifying pressures for academic publishing (Gebreiter, 2021; Pelger & Grottke, 2015; Salemans & Budding, 2022; Tandilashvili & Tandilashvili, 2022).

The findings indicate that researchers set their research agendaFootnote 1 based on four factors: explicit research requests, short-term publishing opportunities, practice and education needs, and the intellectual needs of the academic community. Using cluster analysis, we categorize respondents according to the perceived relevance of these factors into four clusters, shedding light on different effects of “publishing game” pressures on researchers, who form a heterogeneous community placing different importance on the four factors. One cluster adopts a balanced attitude, while the other three groups prioritize a single specific need: getting published rapidly, supporting practice, or meeting intellectual needs. Hence, the research topic choice can aim at a societal impact (research process A) through different paths taken by balanced, practice-oriented, and curiosity-oriented clusters. By contrast, we document a dangerous effect of publishing pressure on the research topic choice of the publication-oriented cluster, which is mainly driven by the probability of being published in the short term (research process B), thus retreating from the risk of innovation.

This study contributes to the debate on publishing pressures in accounting academia by complementing the contextualized reflections of previous literature with evidence documenting various consequences on the agenda-setting choices of a heterogeneous community. It addresses the unexplored topic of research agenda setting in the accounting domain, showing that the goal displacement induced by “publishing game” pressures can affect what scholars study in addition to how they conduct research. These findings have policy and practical implications that can interest policymakers, university managers, gatekeepers of the publishing process, and the entire academic community.

The remainder of this paper is organized as follows. The next section reviews the prior literature and develops our research question. Section 3 presents the research design and Sect. 4 describes the survey and its respondents. The Sect. 5 illustrates the findings, and the final section discusses the study’s implications and provides the conclusions.

2 Literature review and research question development

In the social science literature, a few articles (Horta & Santos, 2016; Santos & Horta, 2018) address the research agenda of academics, focusing on the higher education domain. Horta and Santos (2016) developed an instrument to measure individuals’ research agenda-setting based on a questionnaire survey targeted at higher education scholars. A subsequent study (Santos & Horta, 2018) used this framework to identify two main groups of higher education researchers: cohesive and trailblazing clusters. Those employing a cohesive research agenda pay particular attention to the development of expertise in the field and aim to converge with existing knowledge, thus conveying a sense of stability and perception of low risk. In contrast, those embracing a trailblazing research agenda seek to expand research into other fields of knowledge by conducting multidisciplinary research and adopting a more risk-taking approach. Researchers in different situations in their careers exhibit remarkably similar clustering in their research agenda, suggesting that some pressures from the academic environment could undermine expected differences and highlight isomorphic pressures to conform to and survive, especially among younger researchers.

Another interesting study (Sandberg & Alvesson, 2011) concentrates on how organizational researchers construct research questions from existing literature. Based on a review of 52 organization articles, Sandberg and Alvesson (2011) suggest that gap-spotting is the most common way of constructing research questions, identifying three versions: confusion, neglect, and application-spotting. They argue that gap-spotting questions are more likely to reinforce or marginally revise existing influential theories than to challenge them and discuss several social norms leading to the prevalence of gap-spotting over problematization. In particular, gap spotting is uncontroversial, less demanding, encouraged by the prevalent crediting norm in academia, and vigorously promoted by research institutions through their measurement systems, all relevant arguments in a “publishing game” setting.

To the best of our knowledge, there is no empirical literature exploring accounting researchers’ research agenda-setting process. The only accounting study on research topic choice (Chow & Harrison, 2002) does not examine how accounting researchers create their research agendas but aims to guide how to generate meaningful or significant topics. Twenty phone interviews with prominent scholars reveal a consensus that research should address topics significantly impacting real-world problems and/or the literature. The contribution to the literature can come from filling a gap in theory, producing salient, novel, or unexpected results, advancing understanding of hard-to-solve research issues, or introducing new procedures.

Accounting studies devote more attention to the topics chosen for the study, their outcomes, and the issues addressed, focusing on journal articles outlining research trends and unanswered questions. These studies typically focus on the scientific production of academics from a specific country (e.g., Beattie, 2005 for the UK), a research movement (Guthrie & Parker, 2017), or published in a set of journals (Unerman & O’Dwyer, 2010; Zengul et al., 2021). These studies provide insights into the topics covered by accounting researchers but do not include research questions addressed by unpublished studies or the drivers of scholars’ choices of the research topic. A similar consideration applies to the extensive literature reviews conducted on widely debated topics, whose goal is to organize and critically discuss the state of the art to propose avenues for future research.

To explore how accounting researchers select the foci of their endeavors, a relevant stream of literature builds upon research on growing publishing pressures in the accounting realm. Indeed, the setting in which accounting scholars set their research agenda is increasingly characterized by a “publishing game” atmosphere, where journal rankings shape the understanding of what constitutes high-quality research, often restricted to what is published in elite US-oriented journals (Willmott, 2011). This narrow definition of research quality (Brinn et al., 2001; Craig et al., 2014) has dysfunctional consequences that are at the center of a lively academic debate involving scholars using different research paradigms and methods (Gendron, 2008; Guthrie & Parker, 2014; Hopwood, 2008; Tuttle & Dillard, 2007; van Helden & Argento, 2020).

The endangerment of research diversity appears to be one of the most damaging effects of the “publishing game” (Annisette et al., 2015; Guthrie & Parker, 2014; Hopwood, 2008; Humphrey & Gendron, 2015; Palea, 2017; Tourish & Willmott, 2015; van Helden & Argento, 2020), which amplifies a hierarchy in terms of research styles and rewards conformity to the US mainstream. Therefore, some research areas, epistemologies, and methodologies are privileged at the expense of others that tend to be ignored in highly-ranked journals, leading to a widespread obsession with theorization (Guthrie & Parker, 2016). Such conformity threatens local research traditions (Komori, 2015; Messner, 2015) and as well as relevant accounting niche domains and related journals, such as public sector accounting (Van Helden & Argento, 2020), accounting history, and accounting education (Sangster, 2015). Indeed, the exclusion of a journal from a high-ranking position translates into a competitive disadvantage in attracting new high-quality submissions and threatens the careers of its authors (Annisette et al., 2015).

Furthermore, this tendency to conformity risks reducing research innovativeness and causing intellectual stagnation (Annisette et al., 2015; Guthrie & Parker, 2014; Humphrey & Gendron, 2015; van Helden & Argento, 2020; Wilkinson & Durone, 2015). The expansion of a paying-off mentality driven by a sense of benefits and rewards expected to materialize in the short run constrains innovativeness in accounting research (Gendron, 2015). The closely connected network of authoring accounting academics in top journals can amplify this effect by being less open to the influx of heterodox or radical ideas, bearing implications for the quality of the production and dissemination of novel ideas in accounting scholarship (Andrikopoulos & Kostaris, 2017). This trend also resonates with the focus of most reviewers on incremental contributions to prior research and the rigor of methods rather than on a paper’s main contribution (Brinn & Jones, 2008; Wood, 2016).

The damaging effects of excessive reliance on journal rankings go beyond research activities, creating extreme competition among scholars (Van Helden & Argento, 2020) and engendering a climate of insecurity, especially for doctoral students and young academics (Annisette et al., 2015; Malsch & Tessier, 2015; Pelger & Grottke, 2015). In this context, academics might experience value conflicts that affect how they perceive their profession and the admiration they have experienced in the past (Tandilashvili & Tandilashvili, 2022). Moreover, a competitive environment can generate harmful unintended consequences, even discouraging academic engagement in other areas. Thus, it might jeopardize the quality of teaching (Kallio et al., 2017) and widen the research-practice gap (Guthrie & Parker, 2014), which is at the center of a long-lasting conversation in the accounting domain (Baxter, 1988; Broadbent, 2016; Dyckman & Zeff, 2015; Evans et al., 2011; Hopwood, 2007; Lee, 1989; Leisenring & Johnson, 1994; Parker et al., 2011; Quagli et al., 2016; Sinclair & Cordery, 2016; Zeff, 1989).

Collectively, research on the growing “publish or perish” atmosphere reveals a severe risk of goal displacement, as such pressure pushes researchers to pursue careerist publishing strategies in which getting into top journals is a more desirable outcome than doing interesting and impactful research (Gendron, 2015; Guthrie & Parker, 2014; Hopwood, 2007; Otley, 2010; Palea, 2017). In other words, the idea that academic quality measured through journal metrics is at the core of academic success risks subverting the true ends of research and relegating its societal impact to a mere cosmetic factor (Van Helden & Argento, 2020).

For this reason, our investigation of research topic selection by accounting researchers requires considering the entire research process and its core objectives, which are at the heart of the fundamental choices made at the beginning of each research journey, including the topic of interest. Researchers who aim to conduct studies producing meaningful and socially relevant results choose topics to investigate in light of the possible research outcomes (i.e., translation of research into practice) since the beginning of the process (i.e., research topic choice). Indeed, they are well aware that the chosen research topic can materially affect several stages of the research process, including its outcomes and impact on society. To explore the research topic choice in light of the broader research process, we mobilize the diffusion of innovations theory (Rogers, 2003), which has already been applied to the research process, conceived of as a particular kind of innovation, to analyze the translation of research into practice (Brownson et al., 2006; Gautam, 2008; Tucker & Parker, 2014; Tucker & Schaltegger, 2016). Prior accounting literature (Tucker & Parker, 2014; Tucker & Schaltegger, 2016) uses the theoretical standpoint of the diffusion of innovation as a lens through which to interpret the perspective of the research process, importing this theory as it has been applied to the research-practice gap in other fields. Thus, it provides one point of departure from which to explicate our understanding of the perceived relationship between academic research and society (i.e., practice) (Tucker & Parker, 2014).

This study draws on the diffusion of innovation as a method theory (Lukka & Vinnari, 2014), which provides a meta-level conceptual system to frame the research process and how the goal displacement induced by publishing pressures can affect its initial essential phase, namely, the research topic choice. More specifically, our investigation builds on the staged approach proposed by Brownson et al. (2006), who conceptualize the diffusion of research findings in four steps: (i) discovery, (ii) translation, (iii) dissemination, and (iv) change. Rigorous research generates knowledge by developing and testing scientific theories and methods (discovery), and generalized findings are then adapted into a form useful to the target population (translation). Translated results are subsequently transmitted to end users (dissemination) to impact practice (change). For our analysis, we incorporate three phases of the research process (Smith, 2003) as components of the discovery stage: (i) selection of the research topic, (ii) choice of the research design, and (iii) conduction of the study. Our study focuses on the first phase of discovery (i.e., selection of the research topic), considering its possible connections with the following steps, which researchers have paid particular attention to since the beginning of the research process.

Within this extended model, we can conceptualize two extremes designed to provide contrast and reflect on different research objectives and topic choices (Fig. 1). One consists of socially and/or economically relevant research, in which the first phase of the discovery stage (i.e., topic choice) is shaped by the desire to have a real-world impact. This illustrates a complete research process (research process A) potentially leading to significant outcomes, addressing issues relevant to policy and/or practice, or generating a long-term impact through theoretical advancement or challenges to current thinking. The other extreme refers to research designed from the beginning to have scientific output (translation stage) as the ultimate goal (research process B), regardless of any possible relevance of the findings outside academic journals.

Fig. 1
figure 1

Research process and topic choice: two extremes

Full size image

These extremes can be conceived of as the endpoints of a continuum, in which research topic selection is somewhat driven by the intent of generating an outcome and not only a scientific output (Fig. 1). In both cases, publishing is a primary objective of a researcher; the difference is that in research process B, it becomes an end in itself, suggesting a goal displacement and possible dysfunctional outcomes of the emphasis on publishing (Gendron, 2008; Guthrie & Parker, 2014; Hopwood, 2008; Tuttle & Dillard, 2007; van Helden & Argento, 2020). These two endpoints of the spectrum also recall the two research career models described by Hermanson (2015); in particular, the above-described research process A shares many features of the career model pursuing problems important to the practice community and interesting to researchers. By contrast, research process B is strictly aimed at achieving publishing objectives and seems in line with the choices of researchers following a career model based on the acceptance of the current rules of the elite academic publishing game. This parallel supports the conceptualization of the research process adopted in this study and highlights an important link between research impact and creativity. Indeed, Hermanson (2015) argues that creativity in accounting research is critical to the impact of our research, the advancement of our discipline, and the personal growth and satisfaction of individual faculty. More generally, an exceedingly conformist research community poses the risk of a disciplinary stagnation, threatening the very sustainability of our profession.

This study explores research topic selection by accounting researchers against a backdrop of intensifying publishing pressures to understand whether their dysfunctional consequences include this fundamental choice, where the strategy of discovery itself lies (Bernal, 1971). Indeed, how scholars choose their objects of investigation is an essential premise for developing meaningful research process A; however, the goal displacement of research process B could bias even this crucial decision, which logically precedes choices about the research paradigm and method. Based on these considerations, our study addresses the following question:

  • How do “publishing game” pressures influence accounting researchers’ choice of research topics?

3 Research design

This study addresses our research question through a two-step analysis of data collected through a questionnaire survey that provides empirical evidence about the factors accounting researchers consider when choosing their objects of investigation. Given the lack of empirical studies on research topic selection in the accounting domain, the questionnaire was designed to consider a variety of elements that are likely to impact this decision (hereafter variables). For each of these variables, respondents were asked to rate their relevance to the research topic choice on a 5-point Likert scale.

The identification of these variables is based on previous literature on accounting research trends and publishing pressures, and aims to consider a broad range of elements due to the exploratory nature of the analysis. More specifically, we consider 14 variables (illustrated in Table 1) that are likely to impact the choice of research topic and frame them by drawing on the classification used by the diffusion of innovation theory for prior conditions (Rogers, 2003). We apply these concepts to the context of the scholars’ choice of the research topic, which can be influenced by (i) their previous research activity; (ii) felt needs/problems posed by other members of the system (e.g., colleagues and practitioners) or by their personal curiosity; (iii) originality of the topic; and (iv) norms of the social system. These four categories are helpful in organizing the variables emerging from the literature on accounting research trends and publishing pressures, whose links with the core objectives of the research process (i.e., research processes A or B) can be more or less clear (Table 1, column 3).

Table 1 Variables potentially affecting the research topic choice
Full size table

In other words, the 14 variables listed in Table 1 aim to explore the topic choice considering a wide range of possible drivers since this decision has not been empirically studied in the accounting domain thus far. These 14 variables consider various possible approaches to the research topic choice and are not all connected to the “publishing game”. Some of them are aligned with research process A (e.g., the perceived relevance of the topic for practice), while others require the joint consideration of more factors to be adequately interpreted (“Research process A or B” in Table 1). For instance, the desire to address debated topics is not necessarily connected to the publishing game and research process B. However, it can be seen as opportunistic when it prevails on the topic’s relevance and is strongly associated with factors that increase the likelihood of a more rapid publishing process (e.g. well-known methods and readily available data). The second step of our analysis (i.e. cluster analysis) allows this joint consideration to adequately interpret the influence of the “publishing game” on scholars’ decisions.

The first category of Table 1 includes variables connected to previous research conducted by respondents, namely, the availability of data and familiarity with the research methods most used to study a topic. These elements are likely to influence the direction of future research projects for various reasons. On one hand, a researcher is expected to remain interested in an area of inquiry over a period of time, gaining the knowledge and in-depth understanding that is undoubtedly required for high-quality and relevant research (research process A). An incremental approach is appropriate for a researcher genuinely interested in conducting rigorous research on relevant topics, for whom a continuous change in research focus could lead to insufficient area knowledge. On the other hand, emphasis on quantitative measures of scientific production can lead to opportunistic choices going beyond a legitimate incremental approach, resulting in opportunistic behaviors (research process B) that can seem rewarding according to the prevailing rules of the publishing game (Hermanson, 2015). Indeed, investigating topics for which a researcher already has available data (DAT) and is familiar with widely applied research methods (MET) can be the result of a paying-off mentality (Gendron, 2015). In this sense, opportunistic behaviors might include duplicate publishing that has been colorfully described as “salami publication” or “least publishable unit,” intended as “one way of squeezing more papers out of a research project” (Broad, 1981).

Another class of variables that could drive research topic choice concerns the needs and problems of accounting researchers. These can range from research problems—expressed by literature, colleagues, practitioners, or deriving from personal curiosity—to needs, such as the need to raise funds to support research activities. All the factors belonging to this category reflect consideration of the research impact, typically related to research process A. The origin of the felt needs can vary, suggesting different types of research impact. For instance, the perceived relevance for practice (PRA) and specific requests by professional associations (PRO), which could also be linked to research grants (GRA), indicate an interest in the practical relevance of research and the intention to engage with practice (Guthrie & Parker, 2014). Intellectual curiosity (CUR) can be a crucial driver of blue-skies research that can later filter across to inform engagement-oriented studies (Unerman, 2020). Similarly, the need to organize knowledge (KN) and deepen the understanding of relevant topics for educational purposes (EDU) reflects the desire to conduct research whose goals are beyond a purely publishing objective. Choosing a research topic based on these two factors reveals positive attention to teaching, which is one of the areas suffering from the increasing urgency of publishing (Tandilashvili & Tandilashvili, 2022). Calls for papers on a particular topic by a journal (JC) or conference (CC) highlight a felt need by the academic community, which could have identified a relevant gap in the literature and used this tool to stimulate relevant research (research process A). However, in the case of goal displacement (research process B), researchers can also see them as a quicker way to increase their score as “academic performers” (Gendron, 2008). These factors could modify their research topic choice in the direction of conformism, which has already been identified in the accounting literature (Tuttle & Dillard, 2007) and poses a severe risk of intellectual stagnation (Gendron, 2008).

The originality of a research topic deserves particular attention when investigating the initial phase of the research process. On the one hand, there is an expectation that research is innovative and original and can foresee phenomena and problems in proposing effective solutions. On the other hand, accounting practice and policymakers might find studies on topics debated in professional journals (DP) useful for their decision-making (research process A). The influence of professional debate on researchers’ topic choices signals engagement with practice, thus addressing the long-standing call to bridge the gap between accounting research and practice (Quagli et al., 2016; Singleton-Green, 2010; Tucker & Schaltegger, 2016). Issues much discussed in academic journals (DJ), even though less innovative, can influence research topic choice by suggesting a greater likelihood of success in the publishing game (research process B) for scholars focusing on those topics. This perception resonates with the risk of intellectual stagnation and low innovativeness in accounting research (Annisette et al., 2015; Guthrie & Parker, 2014; Humphrey & Gendron, 2015; van Helden & Argento, 2020; Wilkinson & Durone, 2015).

The norms of the social system can be expected to play a relevant role as well, especially in contexts where the “publish or perish” pressure is higher. As summarized in the literature review, the desire to “join the club” might lead to the dysfunctional outcome of tailoring research with the mere objective of publishing (research process B), thus overrating the probability of getting published (PUB) (Gendron, 2008; Hopwood, 2008; Palea, 2017). The influence of other scholars’ decisions (COL) is also part of the social system in which scholars make their research topic choices, but can occur in various circumstances, including involvement in a broad research project or the mentorship of a supervisor (Pelger & Grottke, 2015). These considerations highlight the complexity of interpreting factors that potentially affect research topic choices. In most cases, they are not exclusively linked to research process A or B, which suggests a goal displacement induced by the “publishing game”.

Our analysis of the survey responses was divided into two phases to address this complexity. First, we conducted an exploratory principal component analysis to extract a set of underlying dimensions (factors) that affect accounting researchers’ choice of research topic. Employing this analysis, we reduced the 14 variables included in the questionnaire to a smaller number of factors that accounted for a significant part of the total variance in responses. This phase is crucial, as the factors emerging from this analysis do not derive directly from those used in the questionnaire but from the correlations in accounting researchers’ perceptions. Accordingly, these factors have to be interpreted in light of the literature and their composition; this allows us to gain a better understanding of the influence of variables that could refer to different goals associated with topic selection. For instance, the availability of data on a specific topic could be interpreted differently if it is found to be an underlying dimension of a factor together with intellectual curiosity, or only with the probability of being published. The first case suggests that respondents are genuinely interested in a topic and already have data available because of intense research activity (research process A). In contrast, the second could indicate a short-term research strategy that does not exclude “salami publishing” to play the game (research process B).

The second phase of our analysis investigated the researchers’ attitudes towards the factors identified in the first phase to understand how they weighed them in the research topic choice. Indeed, accounting researchers can place different importance on these factors, and it is likely that the respondents do not necessarily share the same view. For example, one group could be more concerned with the potential practical impact of research (research process A), while another is more focused on quickly publishing research (research process B). To better understand researchers’ attitudes, we performed a cluster analysis that categorized respondents according to the perceived relevance they attributed to the factors, thus allowing us to group observations into clusters based on their dissimilarities across the identified factors. This analysis indicates the most relevant factors for different groups, thus addressing our research question on the impact of “publishing game” pressures on the research topic choice.

4 Survey description

We collected data on accounting researchers’ perceptions of the variables affecting their choice of research topic through an online questionnaire survey. The questionnaire included questions regarding the perceived relevance of the 14 variables (described in the previous section) in their choice of research topic, to be rated on a 5-point Likert scale (see “Appendix”). Additionally, the questionnaire included questions on the researchers’ research interests and experience to assess any concentration of responses from a specific research area, and personal information (e.g., country) to have primary data on the academic setting of respondents.

We focus our analysis on the European setting because it has relevant common characteristics at a broad level, even though the detailed context varies from country to country within Europe. University systems rely significantly on public funding and perceive the impact of austerity policies and research assessments. Despite the progressive globalization of research and differences at the national level, European researchers face similar challenges and intensifying pressures on academic publishing (Pelger & Grottke, 2015). Moreover, the European context is characterized by a research-practice gap also rooted in the research topic choice (Singleton-Green, 2010; Tucker & Schaltegger, 2016), suggesting a problem of “lost before translation” in knowledge transfer (Shapiro et al., 2007).

The survey design aimed to minimize some common difficulties identified in survey studies (Young, 1996), including low response rates and response biases. To ensure anonymous answers and encourage participation, the survey was designed to avoid any link between responses and the invitation emails and to limit the number of personal questions that could be used to identify some of the respondents. However, we included an IP address check to exclude the risk of double responses, which could bias the survey findings.

To achieve a relevant number of responses, we disseminated the questionnaire to a vast target population, namely European members of the European Accounting Association (EAA). They represent an international community active in accounting research and encompass a variety of research interests and perspectives on accounting issues, without a specific focus on a single research area. This choice also allowed us to obtain the opinion of a broad population of researchers interested in accounting research, not limited to the authors of leading journal articles (Chow & Harrison, 2002). We also consider that working within European universities is not a requirement to become an EAA member, in line with the association’s objective of providing a platform for the wider dissemination of European accounting research. Therefore, we excluded all responses from professionals and academics working at non-European universities. To this end, we considered the academic affiliations stated by the survey respondents rather than the researchers’ nationality or country indicated in the EAA database.

The lengths of the questionnaire sections were chosen carefully, positioning the easiest and least important questions at the beginning and end of the questionnaire to overcome the most serious effects of possible measurement errors (Andrews, 1984). Additionally, the questionnaire was designed to display possible answers in a randomly selected order to avoid the position of the answers influencing the responses and to allow respondents to add other relevant variables. This option enabled survey participants to indicate additional elements that they rated as important for the research topic choice to ensure that we did not exclude anything that might contribute to the decision under investigation.

The survey was sent to members of the European Accounting Association (EAA) in 2014 using a mailing list available to members on the EAA website. Responses were collected from April 24 to June 12, 2015, with an email invitation sent to EAA members on April 24, 2015, and two reminders sent on May 21 and June 8, 2015, as a follow-up procedure (Dillman, 1978). Data collected in 2015 allow for the assessment of the effects of budgetary restrictions and publishing pressures in the European context, where research assessments intensified in the first decade of 2000 (Rebora & Turri, 2013). Since then, the “publish or perish” culture has flourished, adding to the significance of our research in understanding its impact on accounting academics (Becker & Lukka, 2022). This relevance has been highlighted by the recent special issue on performance management in universities (Manes-Rossi et al., 2022) and the position paper signed by the co-founders of the Responsible Research in Business and Management (RRMN) network (2017). In particular, the sixth principle of that paper stresses the importance of recognizing that the publication itself is not the outcome or the end goal but a step in the journey to scholarly and societal impact.

Participants were recruited by sending 2475 invitation emails that provided an online survey link. We opted to use all email addresses for the invitation emails without excluding some members for their academic status or geographical affiliation, as this information could have been outdated. Instead, we used all email addresses and excluded responses outside of our target population in the second step. 97 emails were returned with delivery problems owing to invalid email addresses and annual or long leaves covering the entire survey period. We considered all other emails as valid invitations, including the ones that received temporary “out of office” automatic answers. As a result, 2378 valid invitations were sent, and 526 responses were collected, with an overall response rate of slightly over 22%. We obtained 448 complete responses, a completion rate of approximately 85%. The high number of survey respondents can be seen as a sign of interest from the academic community in this area of inquiry; moreover, the high response rate suggests a lack of severe issues in questionnaire design or clarity that could bias our findings.

To analyze only data representing European researchers’ viewpoints, we excluded two responses from professionals and 128 from academics affiliated with non-European universities. The final dataset comprised of 318 complete responses from researchers affiliated with European universities. The final response rate was 20.8%, considering all 1526 valid invitations sent to EAA members affiliated with European universities on the official membership list. Table 2 illustrates the geographical composition of the respondents, their academic positions, and their ages.

Table 2 Respondents’ university affiliations, current academic position, and age
Full size table

5 Findings

5.1 Factors driving the research topic choice

Researchers’ responses show that all 14 variables proposed by the questionnaire are relevant to the choice of research topic, with average rates ranging from 2.3 to 4.2 on a 5-point Likert scale (Table 3). In addition, less than 5% of respondents add other elements to the variables included in the questionnaire Their comments do not point to a specific new variable to consider, suggesting that the questionnaire is rather exhaustive.

Table 3 Descriptive statistics of variables influencing scholars’ choice of research topics
Full size table

Intellectual curiosity (CUR) emerges as the most relevant variable, having the highest average value and the lowest variance, with more than three-quarters of respondents giving it at least a 4 rating on a five-point Likert scale. This result may seem unsurprising, but in our view, it is important to confirm that despite many kinds of pressures, the choice of research topic is still, and most of all, an expression of a researcher’s intellectual freedom. Respondents also deem the conditions referred to researchers’ previous research to be relevant, namely the availability of data (DAT) and familiarity with the research methods most used to study a topic (MET). These responses might be interpreted as an indication of a scientific community very focused on specific research objects and methods, as it is reasonable to expect—at least to some extent—from researchers engaged in the international academic arena. It is also important to underline that, ceteris paribus, available data and methodology expertise are likely to lead more rapidly to a publication, especially considering that both these conditions require an in-depth knowledge of specific literature, which is a considerable investment of time. This observation could also be linked to the high relevance attributed to the probability of getting published (PUB), which confirms the pressure of the “publishing game” social norms. Conversely, on average, respondents suggest that research topic choice is scarcely influenced by the decisions of colleagues or tutors (COL).

As for the needs and problems felt by academics or expressed by other constituents of the accounting profession, the need to organize knowledge (KN) and education needs (EDU) are rated as less relevant, with an average score lower than three. Conditions associated with practice needs have mixed results, with general relevance for practice (PRA) perceived as more relevant than a specific request from a professional association (PRO). Furthermore, Table 3 shows the limited relevance of the probability of obtaining research grants (GRA). This finding can be interpreted as evidence that funding constraints do not influence researchers’ key choices; however, it is worth noting that perceptions of this issue vary considerably, as shown by the highest standard deviation in the responses. Call for papers regarding conferences (CC) or journal special issues (JC) appears to be scarcely relevant in the choice of research topic, which might seem to conflict with responses suggesting a considerable influence of the publishing game. A possible interpretation of these findings is that specific calls for papers are associated with intense competition from highly respected scholars, and their deadlines may be incompatible with the time needed to approach a new research topic. Indeed, writing high-quality papers requires a significant investment in time to acquire the necessary in-depth knowledge of the literature before launching a new study. Conditions related to innovativeness (DJ and DP) do not show, on average, perceptions of significant relevance, even though quartiles present very different opinions on this point.

The first step of our two-phase analysis (as described in Sect. 3) is an exploratory principal component analysis aimed at identifying a lower number of factors influencing researchers’ choice of research topic, thus supporting their interpretation. It is a data simplification method that combines many correlated variables (i.e., the 14 variables included in the questionnaire) into fewer underlying dimensions. In other words, it enables the construction of a new set of variables based on the relationships between the original variables, as shown in the correlation matrix (Table 4).

Table 4 Correlation matrix of the fourteen variables influencing scholars’ choice of research topics
Full size table

First, the principal component factors are obtained, and then the analysis focuses on the factors that explain the greatest variance, namely, factors with eigenvalues greater than one. Our analysis shows that the first four factors have eigenvalues greater than one, and these four components explain more than 56% of the 14 variables’ combined variance. Hence, these four factors synthesize 14 variables whose relevance has been rated by the questionnaire respondents. To further simplify the factor structure and interpret the results, a varimax orthogonal rotation was run. After the rotation, the relevance of the 14 variables in the questionnaire was uniformly shared among the four resulting factors (Table 5). All 14 variables included in the questionnaire were used in the analysis, as none of them had factor loadings lower than 0.4 (Field, 2013; Guadagnoli & Velicer, 1988), so the four resulting factors can be considered reliable, especially considering the sample size of our analysis (Hair et al., 1998).

Table 5 Factors and rotated factor loadings (oblique oblimin rotation)
Full size table

Evidence from the exploratory principal component analysis indicates that researchers’ choice of research topic is influenced by four factors, which allows a more unambiguous interpretation of the 14 variables included in the questionnaire. Indeed, many of them could be relevant for research process A and the paying-off mentality (Gendron, 2015) of research process B; hence, interpreting a combination of these variables can clarify the actual drivers of researchers’ choices. We interpret the resulting four factors (Table 6) as follows: (i) explicit research requests, (ii) short-term publishing opportunities, (iii) practice and education needs, and (iv) intellectual needs of the academic community.

Table 6 Factors identified and their underlying variables
Full size table

The first factor (factor 1) includes the probability of obtaining research grants, calls for papers from both journals and conferences, and specific requests by professional associations, which can be collectively interpreted as explicit research requests. This factor reflects engagement and interest in studying topics relevant also for the accounting profession or other funding bodies (Evans et al., 2011), thus suggesting a research topic choice inspired by the goals of research process A.

The second factor (factor 2) comprises the probability of publishing, availability of data, familiarity with the research method most used to study the topic, and considerable debate about an issue in academic journals. The combination of these elements synthesizes short-term publishing opportunities associated with research on a topic. This factor condenses conditions that are likely to lead to a publication in a shorter amount of time and, taken to the extreme, could be interpreted as evidence of the dysfunctional consequences of the “publish or perish” context (research process B).

The third factor (factor 3) includes relevance for practice, financial press debating the topic, and education needs, and simplifies researchers’ perceptions of practice and education needs. The influence of this factor is an important indication of engagement with practice and with an audience outside academia, suggesting a topic choice informed by the objectives of research process A. This differs from the first factor, as researchers’ interpretations filter the needs and problems of the end users of research, which are not necessarily explicitly formulated in specific documents (e.g., call for papers). This result highlights research objectives that are not limited to solving immediate practical problems, but include interpreting reality to figure out issues and opportunities and to provide valuable solutions and insightful viewpoints. This reinforces the idea that the notion that research should solve practical problems is based on a rather simplistic view of knowledge production (van Helden, 2019).

Finally, the last factor (factor 4) includes intellectual curiosity, decisions made by other colleagues or tutors, and the need to organize knowledge, which can be interpreted as the intellectual needs of the academic community. This factor represents a synthesis of conditions related to the intrinsic scholarly desire to learn and investigate and can also be stimulated in dialogue with colleagues. This type of topic selection appears as the premise of research process A and resonates with the features of blue-skies thinkers (Unerman, 2020).

This interpretation of the four factors does not seem significantly biased by cross-loadings, indicating that an item measures more than one dimension. Only two variables (PRO and KN) have cross-loadings on two factors, but we can interpret them without invalidating the analysis in light of their content. It is not surprising to have cross-loadings of specific requests by professional associations and the need to organize knowledge on the factor capturing practice and education needs (factor 3).

Tables 7 and 8 present the descriptive statistics of the four factors and correlation matrix.

Table 7 Descriptive statistics of factors influencing the choice of research topics
Full size table
Table 8 Correlation matrix of factors influencing the choice of research topics
Full size table

Descriptive statistics show that the median relevance is higher for intellectual needs (Factor 4) and short-term publishing opportunities (Factor 2), followed by practical relevance, education needs (Factor 3), and explicit research requests (Factor 1). Evidence confirms that topic choice is influenced by the probability of being published in the short term, suggesting a real risk of goal displacement induced by a “publish or perish” atmosphere (research process B). However, publishing is inherently a relevant objective for researchers, and interpreting this evidence requires a deeper understanding of how short-term opportunities are combined with or prevail over other factors. Further investigation is the focus of the second phase of our analysis.

5.2 The influence of “publishing game” pressures on an heterogenous community

The second phase of our analysis aimed to explore how researchers combine and balance the four previously identified factors in their research topic choice, considering that they might not have common perceptions in this regard. Indeed, some of them could try to balance several factors (e.g., the relevance for practice and the desire to publish rapidly). In contrast, others could perceive a particular factor as more important, showing the emergence of research processes A or B. For this reason, we ran a k-means cluster analysis on the four principal factors identified in the exploratory principal component analysis. This technique allows us to group researchers such that individuals in the same group (cluster) are more similar to each other than those in other clusters, with specific reference to factors driving research topic choice. We used this analysis to explore the extent to which researchers can be grouped into different clusters based on topic choice, as well as the differences between the clusters identified.

The result is a four-cluster solution (Table 9) obtained after considering the cluster size, the Calinski–Harabasz pseudo-F (73.00, untabulated), and whether the cluster solution can be interpreted in light of the theoretical framework and prior literature. First, the analysis suggests the existence of heterogeneous perceptions among accounting researchers who do not seem to share a common view on what counts when deciding the focus of their research efforts. Instead, the four groups of researchers place different importance on the four factors previously identified, and they are balanced in terms of frequencies, without a cluster accounting for a dominant percentage.

Table 9 Statistics for the four factors in researchers’ clusters
Full size table

Most respondents (29.56%) belong to a “balanced” cluster (cluster 2 in Table 9) characterised by significant relevance perceived for all four factors. These researchers choose their research topics mainly based on explicit requests and relevance for practice and education but also consider their intellectual needs and publishing opportunities. This attitude seems in line with research process A previously illustrated, where the choice of a research topic is targeted at many intermediate objectives (including publishing and dissemination of knowledge) with the ultimate aim of producing a societal impact.

When choosing their research topic, the remaining three clusters show a marked prevalence of one factor over another. Two clusters exhibit perceptions aligned to research process A, paying particular attention to intellectual needs (“curiosity-oriented” cluster 4 in Table 9) or practical relevance (“practice-oriented” cluster 3 in Table 9). In contrast, one relevant cluster (cluster 1 in Table 9) can be interpreted as a “publication-oriented” group of researchers who place more emphasis on short-term publication opportunities, seemingly disregarding the other factors when choosing which topic to invest their time in. Its existence and considerable size (24.53%) offer empirical support for the concerns that the pressure to publish leads to the dysfunctional outcome of short-term and careerist academic publishing strategies (Gendron, 2008; Guthrie & Parker, 2014; Hopwood, 2008). We stress that this interpretation is based on evidence indicating that these researchers essentially pursue rapid publishing as the main objective, regardless of or even in contrast to other aims that we purport to serve as members of the academic community. This attitude leads to research process B and goes beyond the conformism induced by a paying-off mentality (Gendron, 2015), threatening not only innovativeness, but also the relevance of accounting research from its very first phase.

Overall, this analysis reveals a complex picture of the European accounting community, which is not composed of researchers who share a common attitude towards what counts when choosing the direction of their studies. Instead, this study suggests that different sub-communities tailor their research topic choices by considering various factors and taking either a balanced approach or placing more emphasis on a specific goal (i.e., rapid publishing, engagement with practice, or intrinsic intellectual needs). The researchers who chose their research topics mainly based on explicit research requests (cluster 3) and practice and education needs (cluster 4) state that they select the foci of their studies with an eye towards making a relevant impact, which is not always their primary goal but is one of the stated purposes of their research. Their topic choice is designed to be the first step of a complete research process, ultimately aiming to impact accounting practice and/or policy makers (research process A).

More specifically, the “practice-oriented” cluster attributes more importance to their interpretation of practice and education needs than to explicit requests to conduct a particular project. Evidence suggests that these researchers are willing to engage with practice, adopting an active approach based on their curiosity and desire to organize knowledge, and opportunities for rapid publishing do not significantly drive their choices. Compared to the “balanced cluster”, these researchers are less affected by the intermediate goal of publishing and are more interested in the impact on accounting practice, also through educational efforts and professional training. The “curiosity-oriented” cluster ranks as least important quick publication opportunities and is only slightly influenced by explicit requests for research from bodies outside academia (e.g. professional association or a Ministry issuing a research grant). This cluster could include blue-skies thinkers who are not deeply engaged with practice or pursue short-term publishing objectives, but can significantly contribute to the advancement of accounting studies by providing insightful interpretations of reality and theoretical support. The low influence of rapid publishing pressure should not necessarily be interpreted as a lack of interest in any scientific output. Rather, this may indicate that these researchers devote their efforts to long-term research projects that can lead to very influential books.

The “publishing game” pressures are far more influential in the “publication-oriented” cluster, which empirically confirms that some accounting researchers choose their research topics, having lost the original purpose of publishing to advance knowledge (Moizer, 2009). The primary goal of this cluster is fast publishing. Short-term publishing opportunities are not an intermediate goal that could be balanced with intellectual or practical interests, thus suggesting a kind of goal displacement from the first phase of the research process. The striking aspect of this kind of response is not limited to the fact that academics react to the growing incentives based on a publishing obsession, which provides empirical support for a growing “fetishism of speed” among academics (Craig et al., 2014). Far more dangerous is the marked contrast that some scholars seem to perceive as meaningful research and publishing objectives, which appears to be a grave threat to the future of our academic community.

6 Discussion and conclusions

This study explored accounting researchers’ choice of research topic, a crucial phase for the societal impact of our research, especially in the current environment of challenges to the legitimacy of academic research and pressures induced by the unbalanced use of journal rankings (Unerman, 2020). More specifically, it has investigated whether and how “publishing game” pressures influence accounting researchers’ choice of research topic, the first phase of the research process, where a researcher is supposed to be free and more creative. Choosing the subject to investigate is closely tied to the outcome of the research process, which we conceptualize building on the diffusion of innovation theory, identifying an impact-oriented process (research process A), and a more opportunistic process driven only by publication objectives (research process B). The risk of a topic choice informed by research process B is particularly relevant given the variety of dysfunctional outcomes of widespread audit culture (Craig et al., 2014) and the threat of goal displacement, with pernicious effects on researchers’ foci and strategies (Hopwood, 2007; Otley, 2010; Palea, 2017).

Findings based on an online survey of European accounting researchers condense the elements influencing research agenda setting into four factors: explicit research requests, short-term publishing opportunities, practice and education needs, and the intellectual needs of the academic community. A cluster analysis of the survey responses reveals that “publishing game” pressures have different effects on researchers, confirming that the reactions to pressures induced by new public management are diverse (Pianezzi et al., 2020). Four clusters hold divergent views on what counts when deciding the direction of a research endeavor: one group adopts a balanced attitude, while the others focus on a single specific need (i.e., getting published rapidly, supporting practice, and meeting intellectual needs).

This diversity of views within the academic community reflects the personal characteristics of researchers and the freedom to research, which is at the heart of our profession and passionate scholarship. Hence, the choice of a research topic can be the starting point of a research process aimed at societal impact (research process A) through different avenues. The relevance to practice and policymakers is significant for “balanced” and “practice-oriented” clusters. “Curiosity-oriented” researchers are mainly interested in addressing the intellectual needs of the academic community without a strong interest in publishing or any immediate impact on practice. This group includes blue skies thinkers who can provide theoretical advancements or insightful reflections that may be impactful in the long term and stimulate future research. In contrast, a considerable group of researchers (“publication-oriented” cluster) claim to be driven in the research initiative only by the probability of being published in the short term (research process B), revealing a dangerous unintended effect of publishing pressure on the research topic choice. Publishing is not seen as a way to diffuse the research results but as the sole purpose of the researcher’s effort, which retreats from the risk of innovation privileging well-mined subject areas.

This study contributes to the discourse on publishing pressures in the academic accounting context by offering empirical insights into the detrimental effect of emphasis on publication metrics on research topic choice. To the best of our knowledge, this is the first study to provide empirical evidence about the unintended effects of the increasing urgency of publishing on research topic choice, which has long been discussed by relying on theoretical reflections or anecdotal evidence. In doing so, it complements the contextualized reflections of previous literature (Argento & van Helden, 2021; Gendron, 2008; Guthrie & Parker, 2014; Hopwood, 2008) with evidence showing that a ‘publish or perish’ environment produces various consequences on the agenda-setting choices of a heterogeneous community. Our analysis empirically confirms that the goal displacement induced by the vicious circle of academic performance objectification (Gendron, 2008; Hopwood, 2007) can affect both how scholars conduct research and what they study. This reveals a lost before translation problem of knowledge transfer (Shapiro et al., 2007) and a paying-off mentality (Gendron, 2015) resonating with auction models, in which scholars are bidders and publications on the journal authority list are the objects to be sold (Nørreklit et al., 2019). These evidence-based reflections emphasize the need expressed by several parties to promote rethinking of university performance measurement systems by identifying the common values around which joint efforts should be made (Manes Rossi et al., 2022).

Our study differs from previous studies in that it sheds light on accounting researchers’ agenda setting, which has not been explored thus far, through a survey with a high number of respondents not limited to the Anglo-Saxon sphere and a specific accounting research area. Furthermore, it responds to the call for research to develop a grounded argument regarding the pitfalls of excessive performance measurement in academia (Gendron, 2008), providing additional evidence that universities are going too far to reshape as enterprises in a competitive setting subject to an audit culture (Argento et al., 2020). Due to its exploratory nature, this investigation could also be of interest to researchers in other social sciences who are increasingly feeling the pressures of a “publish or perish” atmosphere and are facing a growing disconnect between research and practice.

The empirical findings have important policy and practical implications. Policymakers concerned about the efficient use of a country’s tax money spent on higher education should carefully consider the evidence of accounting scholars setting their research agendas, prioritizing publication objectives at the expense of relevance. Our empirical results emphasize the importance of a shift back to content in research assessments (van Helden & Argento, 2020), implying more reading work and a less standardized way to scrutinize academic work. Indeed, the increased use of quantitative metrics affects researchers’ research topic choice, with detrimental outcomes that recall Goodhart’s law, which illustrates how a measure ceases to be a good measure when it becomes a target. Research evaluation policies could usefully consider the heterogeneity of the research community documented in our analysis by designing individual-level incentives that can foster research process A in various forms. Advancements in this area could stimulate research relevant to society and safeguard academic freedom and self-determination, which new public management risks compromise (Argento & van Helden, 2021).

The detrimental effect of “publishing game” pressures on the relevance of research due to a biased topic choice is particularly relevant to university bodies in charge of shaping their university’s performance systems to meet the diverging demands of various stakeholders. The evidence provided in this study can offer a more comprehensive understanding of the unintended consequences of these pressures, which can negatively impact critical aspects such as reputation, trust, and innovation (Grossi et al., 2020a, 2020b). Indeed, excessive emphasis on research productivity might award researchers adopting opportunistic publishing strategies, who will gradually become new decision-makers for the future governance of universities, thus generating a pro-cyclical effect on the orientations of research and publishing strategies of their departments. This tendency reinforces the importance of strengthening the links between strategic planning and performance measurement in universities, taking an overall view rather than the isolated management of these aspects to avoid the unintended consequences of output-oriented systems (Biondi & Russo, 2022). Furthermore, the increased focus on publishing in evaluations risks being combined with relatively little attention paid to teaching and research impacting local contexts. Our findings echo the recommendation that university managers should make more nuanced use of performance measurement systems (Argento & van Helden, 2021) and adopt particular sensitivity in considering the intangible aspects of the research journey, such as the choice of research topics. This attitude can contribute to the development of academics and produce beneficial effects in increasingly important academic domains, such as the third mission (Nicolò et al., 2020), which involves technology transfer, innovation, and social engagement. To this end, managers and superiors could apply a proactive approach to control the career development of researchers, relying on dialogue and avoiding excessive objectification of performance.

The implications of this study are of interest to the gatekeepers of the publishing process and, more broadly, to accounting researchers who seem to have a love-hate relationship with rankings (van Helden & Argento, 2020), which both seduce and constrain (Gendron, 2008). Researchers serving as editors and reviewers can contribute to reducing the trade-off that “publishing-oriented” scholars seem to feel between publishing objectives and the relevance of the topics to be investigated, which results in the dysfunctional research process B. Promoting special issues on relevant topics that risk being neglected for time-consuming data collection could counterbalance the incentives of investigating less relevant topics with readily-available data. The review process, which risks becoming a self-referential echo chamber (Unerman, 2020), is another key area that discourages opportunistic conformity and emphasizes the relevance of research topics and objectives. Recent initiatives such as the Open Accounting manifesto (Alawattage et al., 2021) and the Accounting for Impact network launched in 2022 suggest increased awareness and determination to ward off the risk that our societal contribution is “lost before translation”.

However, our findings should be considered in light of certain limitations. The questionnaire was circulated in 2015 to EAA members, who might attribute more importance to international publishing than to the entire universe of European accounting researchers. In addition, since 2015, the publishing game may have further intensified, and respondents’ behaviors in real-life research choices may differ from the perceptions indicated in survey responses. Additionally, our analysis did not consider the stage of respondents’ careers or their age, which are relevant aspects to understand how the publish or perish game can be related to the need to speed up their career or to answer specific context requirements. A further limitation is that despite a broad degree of consistency within the European university system, this study does not deal with the differences in the detailed context of each country. Researchers’ characteristics and national contexts could be explored in more depth in future studies examining the connections between incentives provided by national research assessment exercises and research topic choice. These aspects are outside the scope of this analysis but can be promising avenues for future research.

Further evidence in this area can help our community guard against the unintended consequences of universities’ performance measurement systems and nurture an environment that promotes academic contribution to the development of policy and practice. We hope that the empirical evidence presented in this study can be a step in this direction, contributing to constructive (hopefully fruitful) dialogue with policymakers and the cohort of university stakeholders. A passionate community of scholars is an essential cornerstone to safeguard the support of various groups in society for academic research (ter Bogt & van Helden, 2012), thus facing challenges to its legitimacy (Unerman, 2020) and reasons for being (Kallio et al., 2020).