Cite me! Perspectives on coercive citation in reviewing

Abstract

Purpose

This study builds on previous discussion of an important area for both academics and academic journals – the issue of reviewers inappropriately asking for (or “coercing”) citation of their own work. That situation creates an opportunity for (hopefully a small number of) academics to engage in unethical behaviour, often with the goal of increasing their citation count. This study aims to draw attention to this often-overlooked issue, critically considering potential reviewer motivations and offering possible remedies.

Design/methodology/approach

This study reviews literature and critically discusses this issue, offering a typology for coercive citation suggestions and sharing previously unpublished commentary from Editors of leading journals.

Findings

This study provides a typology of reviewer motivations for coercing citations, suggests potential remedies and considers the positive and negative impacts of these suggestions.

Originality/value

This study identifies an area known from multiple discussions to be important to academics and Editors, where many want changes in journals’ practices. In response, this study provides recommendations for easy changes that would decrease the opportunity for unethical behaviour by reviewers and also, for some journals, improve the quality of reviews.

Citation

Burton, S., Basil, D.Z., Soboleva, A. and Nesbit, P. (2024), "Cite me! Perspectives on coercive citation in reviewing", Journal of Services Marketing, Vol. 38 No. 7, pp. 809-815. https://doi.org/10.1108/JSM-08-2024-0387

Introduction

The inspiration for research can come from unexpected directions. The motivation for our examination emerged from a lunchtime chat between two academics (pseudonyms Sanjay and Linda to preserve anonymity). Sanjay was uncomfortable with a review he had received from a highly-ranked journal. The reviewer requested that he cite several articles that seemed unnecessary and which included a common author. Despite a lack of significant other changes, the reviewer also requested to see the manuscript again after revision – a sign, Sanjay feared, that the manuscript would not be accepted if the suggested citations were not made. Linda commiserated and recounted a similar episode.

Widespread discussion of the issue with other colleagues resulted in near-universal recounting of similar experiences by almost every published academic we talked to. There was general agreement that most authors would include the superfluous citations, out of fear that their manuscript would otherwise be rejected and no one suggested they would raise the issue with the journal’s Editor – because (as questioning revealed), they seemed to be concerned they would be seen as a “difficult” author. That situation triggered us to think about what could be, in some cases, a conflict of interest (COI), by reviewers, who can use their anonymous status to encourage or even coerce, an author to cite them.

In this viewpoint, we address this issue of COI in academic reviewing, focusing in particular on reviewers recommending citation of their own work – a situation that we refer to as “coercive citation”. COI occurs when “a primary professional interest is excessively influenced by an individual’s secondary interest(s)”(Resnik and Elmore, 2018, p. 112). Thus, in the case of the review process, COI occurs when the interests of some of the actors involved in the review process – namely reviewers and/or Editors – are pursued over and above the objective of independent and unbiased evaluation of submitted research. Using an institutional theory framework, we shed light on the perverse incentives in the academic review process, where reviewers can be motivated by self-interest (such as the desire to increase their citation count), rather than by providing an expert, independent review. We dissect the various levels of coercive citation requests, considering the different potential motivations and the damage they could cause, and put forward a reviewer typology reflecting this analysis. We also consider the increased challenges faced by academics and Editors. We discuss the implications and possible actions to decrease these intertwined problems.

Institutional theory

COI and coerced citations may be part of a broader issue – a shift within academe, where professional logics may be giving way to market logics (Leslie and Johnson, 1974). Those within the academic ranks of institutes of higher education (termed “academics” and “IHEs”, respectively) may face competing pressures – to provide objective reviews with high integrity as a service to the profession (professional logics), while demonstrating their own research success through the strong citation metrics desired by their university (market logics). The behaviour of reviewers, of course, is likely to be driven by the mores and reward systems of their institutions, in turn, those institutions represent “constellations of established practices”(Lammers and Barbour, 2006, p. 357), which are not reliant on any particular organization and which tend to remain relatively consistent across time and place. Institutions, and organizations within those institutions, adhere to certain rules of behaviour based on both implicit and explicit views and values, called institutional logics. According to Jackall (1988, p. 112), institutional logics are “the way the world works”. Academics work within various logics, including those of their university, professional bodies, publishing houses and funders. This can position academics between conflicting logics (Thornton and Ocasio, 2008). For example, the institutional logics of the academic profession encourage service to the profession through contributions such as reviewing, yet that service may compete with the organizational logics of an academic university, which may prioritize research over service to the profession. Within the present discussion, professional logics suggest that the sanctity of blind peer review should prevail – when acting as a peer reviewer, academics should have only the best interest of the manuscript under review in mind. However, if the academic is also facing organizational logics that prioritize performance metrics, they may feel pressure to improve their citation metrics by suggesting citations of their own work. Similarly, Editors, who are also academics, are most likely working under the assumption that reviewers follow professional logics by respecting the integrity of the review process but may be witnessing less than professional behaviour from reviewers, such as inappropriate citation recommendations, and/or a lack of willingness to review. Additionally, Editors may be facing coercive efforts themselves to increase citation counts for the journals they lead.

Academic publishing

The academic publishing industry has undergone substantial change over the past several decades. Thornton and Ocasio (1999) identified an industry shift in the 1970s whereby academic publishing went from a professional logics focus, with publishing treated as a profession with emphasis on its purpose, to a market logics focus, centring on the business aspects and financial potential of publishing. Though not the first to highlight this shift (e.g. Tebbel, 1987) that work underscores that academic publishing has largely transitioned to a market-focused logic. This is evident when we consider the $25bn profit margin of the industry, with one major publisher even more profitable than Apple (Schmitt, 2015). Despite that profitability, academic publishing houses are reliant on academics within IHEs for providing the content they publish and reviewing that content (generally for free), as well as a large portion of the readership of that content.

A shift in institutes of higher education institutional logics

The often-quoted admonition to “publish or perish” reflects the pressure for academics to publish, and in recent decades, that pressure has been accentuated by the changing higher education environment. Higher education increasingly operates in a global marketplace, which increases competition. Additionally, lower birth rates have reduced the size of the traditional university-aged demographic. Governments have become more careful with funds allocated to higher education, demanding greater accountability from IHEs and shifting education to a market commodity (Lynch, 2015, p. 190). A key feature of this approach is the “widespread use of performance indicators and league tables, target-setting, benchmarking and performance management” in universities (Deem and Brehony, 2005). Individual academics are pressured to publish in highly ranked journals and to gain high citation counts (de Rond and Miller, 2005; Macdonald and Kam, 2007; Seeber et al., 2019), which serve as a measure of relevance in terms of “knowledge transfer” (Birkinshaw et al., 2016), and influence hiring, promotion and funding decisions (Van Noorden and Singh Chawla, 2019).

While the use of metrics to evaluate the performance of both universities and academics seeks to achieve higher quality research and better outcomes for government funding, Lynch (2015, p. 195) argues that performance measures and rankings can weaken professional values. Focusing on performance metrics can place performance pressure on academics that may manifest in ways that undermine the integrity of the review process (Ioannidis, 2015) and result in undisclosed conflicts of interest (e.g. Wilhite and Fong, 2012; Baas and Fennell, 2019; Seeber et al., 2019). As a result, Lepori (2016) argues for the use of an institutional logics approach to understand the current state of IHEs.

It is perhaps not surprising that these shifting institutional logics would result in shifts for academics as well, consistent with the shift apparent for universities. However, that market approach may at times be in conflict with the professional standards of academe, as academics strive for personal metrics of success. In biomedical research, for example, retractions for ethical misconduct have increased exponentially over the past thirty years (Fang et al., 2012; Economist, 2023). Retractions have also strongly impacted on business and management research (Tourish and Craig, 2020). A recent study by Fire and Guestrin (2019) examined more than 120 million papers across more than 26,000 research fields and found that academics are now treating metrics as an end goal (Fire and Guestrin, 2019). An academic conducting a journal review can increase their citation metrics by encouraging the paper’s authors to cite the reviewer’s work, and we propose that such behaviour poses a potential COI for reviewers – which may explain the behaviour of the (real) academic referred to in our opening anecdote – and many other similar academics.

Coercive citation suggestions

While commentary on COIs for authors is common, much less research has addressed reviewer COIs. In particular, the specific COI of reviewers inappropriately recommending citation of their own work can be difficult to identify but would represent a perversion of the publication process. Without discussion of the problem or specific instructions to reveal it, such behaviour may escape attention, but such discussion is rare. For example, there was no mention of coerced citation efforts in recent publications examining COI in peer review (Resnik and Elmore, 2018), unethical research practices (Schwab and Starbuck, 2017), or journal prestige being driven by citations (Byington and Felps, 2017). In a review of 20 leading management and marketing journals, Clair (2015) found that only 3 (15%) had a reviewer and Editor code of conduct on their website.

We recently conducted an examination of journal practices for discouraging coercive citation requests by reviewers (blinded for peer review). In that work, we reviewed the websites of several randomly selected leading marketing and management journals and surveyed Editors of major marketing and management journals (n = 205). We found that less than 1% provide reviewers with guidance regarding coercive citation. Considering these results, it is perhaps not surprising that coercive citation suggestions occur.

Conceptualizing reviewer motivations for coercive citation suggestions

It is important to understand how COI can arise when crafting recommendations to address it. A reviewer recommending citation of their own work may in fact be following best practice if the reviewer is an expert and recommends appropriate citations. Additionally, recommendations that are not appropriate may stem from well-intentioned, rather than self-serving, motives. To decrease the likelihood that reviewers’ actions are driven, or perceived to be driven, by COI, we propose considering both the potential harm of reviewers recommending citation of their own work, and the intent behind the recommendation. The following thought exercise identifies four categories of reviewer self-citation suggestions. This typology motivates the recommendations we offer in the subsequent section.

The first category of self-citation suggestion would occur in the case of true expertise, where the reviewer is truly an expert on the topic. In that situation, a recommendation to cite the reviewer’s work may be extremely helpful to the author and would be in accordance with professional logics. In that case, the reviewer’s motivation is likely to be constructive rather than self-serving.

A second category of self-citation suggestion may occur when the reviewer has a level of expertise but overestimates the significance or relevance of their own work, genuinely believing that their work should be cited to benefit the paper. Reviewers in this category may have an availability bias (Tversky and Kahneman, 1973), whereby their own work is top of mind and they see it as more central to the literature than it actually is. Such reviewers may not have a self-serving intent; their review may be somewhat helpful, and suggestions of their own work may be moderately appropriate. In these instances, the reviewer may be attempting to follow professional logics but may be misinterpreting the usefulness of their own work.

A third, more problematic, category encompasses reviewers who are consciously seeking a benefit from citation of their work. Reviewers in this group may have research that is legitimately related to the topic (as they have been chosen as reviewers of the paper), but their motivation for suggesting their own work is primarily based on a desire for personal benefit. This type of motivation is likely to be enhanced by a perceived lack of sufficient benefit for reviewing. Reviewing is very time-consuming, and academics generally receive little or no extrinsic reward for that time. Reviewers may therefore justify to themselves the recommendation of their own work as equitable and deserved compensation for their effort, consistent with market logics and in accordance with equity theory. Equity theory, which is a special case of cognitive dissonance, examines the perceptions of inequity felt by employees and predicts that workers who feel they have experienced inequity will seek compensatory balance (Adams, 1963). Equity theory has proven useful for understanding worker responses in situations of pay inequity (Buttner and Lowe, 2017) and idiosyncratic workplace benefits (Marescaux et al., 2019). Reviewers may perceive inequity when reviewing because they are offering their labour without express compensation. Reviewing is rarely if ever a stated part of an academic’s job, so it may feel to a reviewer as if they are volunteering their time when reviewing. Suggesting the citation of their own publications may then seem like a reasonable way to insert equity into the situation. However, this may still be detrimental to a paper under review because it could lead to the inclusion of ill-fitting research that dilutes the paper’s impact and results in future inappropriate citations as others take the presence of those citations as authoritative.

Finally, in what we judge to be the most egregious category of self-citation suggestions (coerced citations), some reviewers may be purely driven by market logics and recommend citation of their own work even if such citation is of little or no benefit to the paper (as appeared to be the case with “Sanjay’s” reviewer). Such behaviour can obviously damage the quality of a paper under review by coercing one or more inappropriate citations and denying the author a potentially well-thought-out review by a reviewer with more ethical motivations. This can also create a negative ripple effect where other authors working on related research see the inappropriate citations and assume they too should cite these works. While authors can never be certain a reviewer is suggesting their own work, marginal value recommendations are the most likely to create suspicion by an author that the reviewer is recommending citation of their own work. That suspicion, in turn, is likely to result in authors distrusting the independence of the review system (as we repeatedly heard in informal conversations with academic colleagues) and, for some, encourage similar behaviour themselves when they review. See Figure 1 for a visual depiction of these reviewer categories.

In summary, a shift from professional logics toward market logics may lead academics toward self-serving behaviours such as coercing citations of their own work. Coerced citations are problematic for several reasons. At the very least, a coerced citation might not add to the paper and could instead dilute its focus and contribution. However, the inclusion of inappropriate citations can have a compounding effect, where others read the published paper and assume the citations are authoritative, so include them in their own work as well, creating a ripple effect. Thus, lower-quality work could become significant through repeated citations. Coerced citations can damage authors’ trust in the objectivity of the review system and introduce inequities in the evaluation of academics’ citations, by rewarding reviewers who are prepared to coerce inappropriate citations. Unless coerced citations are identified and minimized by journals’ practices, they can therefore result in ongoing over-citation of unscrupulous authors and a commensurate benefit to their citation counts.

This thought experiment suggests that reviewers may have different motivations for suggesting citations of their own articles; some recommendations may be helpful and appropriate, and some may be innocent misjudgements by reviewers who are simply not aware of the potential COI involved. Others may be intentionally coercive, motivated to increase personal benefit, in accordance with a market logics approach. Regardless of motivation (which cannot be known), superfluous citations damage the quality of the research and may lead to frustration for authors and inappropriate benefit for unscrupulous reviewers.

In a separate study, we conducted a major survey of Editors of A*, A and B-ranked management and marketing journals to examine whether journals provide reviewers with information regarding COI and coercive citation; largely they do not (blinded for peer review). We also asked Editors at that time for any other comments they had about the instructions their journal gives to reviewers. We now turn to these previously unpublished open-ended comments, specifically focussing on the Editor’s perceptions of reviewers’ need for further guidance. As this is a viewpoint piece, we share these statements as supportive commentary only; in this short piece, we can offer neither the scope nor the depth for a full research study.

Editor comments

We used thematic analysis (Attride-Stirling, 2001; Braun and Clarke, 2006) to analyse open-ended comments by Editors responding to our survey. Several themes emerged: “trust” that reviewers know what they should do and should not do, when they write a review, deferral to publishing house practice and reservations about reviewer responses if they were asked to use a review template. We next briefly share a representative example of each.

Reviewers know what to do

Repeated comments from Editors argued that specific instructions to referees are not necessary, because reviewers know (or should know) what to do:

Since we only invite reviewers who have at least some experience with publishing papers in scientific journals, we assume that they are familiar with what a good review should do and what it looks like (in terms of style, setup, tone). (A journal).

This and similar comments reflect a belief that reviewers know what is expected in a review and provide professional reviews. However, instructions can serve as an important prompt for appropriate behaviour; in contrast, failing to provide guidance may be a missed opportunity to properly orient reviewers’ mindsets.

Deferral to the journal publishing house processes

A second theme in Editors’ explanation for their journal’s guidance (or lack of guidance) for reviewers is that the journal defers to the processes specified by their publishing house, which often provides a standard review template. For example, one Editor responded:

Actually [the review template is] all set by the publishers and I don’t think we have ever tweaked it. you (sic) have made me think that we should. I’m not wholly sure how to [forward a copy of the template] without actually assigning you as a referee for a paper (which is a bit pathetic isn’t it). (A journal).

Problems with templates as a guidance tool

Review templates are a common form of guidance provided to reviewers, but many journals do not use them. Some Editors commented that they personally dislike using templates, that they are not helpful, or that reviewers dislike them:

We do not provide a review template - we’d tried that years ago and had gotten feedback: while some aspects of a template were appreciated, we got a significant volume of responses reporting that a template unnecessarily constrained review comments. (A journal).

Review templates are of course limited, but many provide some form of common rating of the quality of different aspects of a submission, which may provide a more objective way of assessing reviewer reactions, compared to the less comparable rating provided by open-ended qualitative comments by reviewers. By querying authors and designing for purpose, Editors could create templates that suit their journal, rather than relying on boilerplate designs that may be ill-received and not useful.

Interest in providing further reviewer guidance

The open-ended comments revealed that participating in the survey had triggered some Editors to think about whether they should provide further reviewer guidance.

There is [no] template for [reviewers or associate editors]. I am a new editor and am developing them. I would love copies of other journals’ guidelines. (A journal).

Comments such as these suggest that some Editors see the advantages of providing reviewers with additional guidance but are not sure what to do, and given the load and complexity of their position, it may not be seen as a priority.

Examples of better practice

Unsolicited, some Editors gave us examples of what appear to be excellent practice for providing guidance to Associate Editors and for reducing variability in decision-making and instructions to authors. For example, one wrote:

Regarding the information we provide to associate editors, our team has regular teleconferences to discuss both specific manuscripts (e.g., the merit of the submission vis-à-vis the discipline, how to interpret referees’ feedback, etc.) as well as more overarching concerns […]. We’ve found these meetings to be extremely helpful as many submissions are, in their own way, anomalous. And it is through this inductive approach (working from specific manuscripts to generate a more gestalt view of how quality might manifest itself in our journal) that meaningful editorial decisions are made. (These teleconferences supplement our onboarding process where the associate editors work with the managing editor, work with different decision-letter templates, and basically learn the ropes.) (A journal).

Additional concerns regarding conflict of interest and market logics

Two Editors suggested that Editors themselves may be another avenue of concern for coercive citations. One respondent wrote that when Editors serve as reviewers for other journals, some suggest citation of articles published in the journal they edit, with the primary goal of increasing journal citation counts. Another suggested that Editors may in fact be working together to suggest citations to each other’s journals, in a form of citation collusion. These issues are beyond the scope of our work; however, they do suggest that concern over coercive citations should be examined at various levels.

Concluding thoughts

An apparent shift toward market logics in IHEs makes self-serving behaviours such as coercing citation increasingly tempting. Previous evidence suggests that many highly ranked Marketing and Management journals do not appear to have processes in place to counter coercive citations, provide guidance to reviewers and/or standardise reviews in a way that would decrease undesirable practices and encourage a focus on professional logics (blinded for peer review). Building on that evidence, this thematic analysis, albeit brief, suggests that at least some Editors recognise that their review processes can and should be improved. There is no doubt that many journals are doing an excellent job in educating their reviewers about a quality review. As this turn toward market logics continues to grow, steps such as these will become increasingly important.

Our concerns about peer review are of course not new. As one author states, “peer review […] does not work as it is supposed to work” (Macdonald, 2015, p. 272). Another article made extensive recommendations for higher-quality reviews (Allen et al., 2019), providing an excellent resource for journals attempting to improve their review systems. Both of those articles, however, pay only scant attention to the problem of coerced citation, and the problems in peer review identified in Macdonald’s (2015) paper appear to persist.

Reviewers may suggest citations of their own work for a variety of reasons. Often, such suggestions are justified as the reviewer’s work is highly relevant to the manuscript under review. While we believe it is imperative to address the issue of coercive citation, not all self-citation suggestions are in fact coercive, thus making this a complex issue.

We recognise that experiences of coercive citation are mixed. Many with whom we have discussed this topic feel they have experienced coercive citation efforts. This is supported by research; a survey of authors for the journal “Nature” showed two-thirds felt they had received coercive citation suggestions (Chawla, 2019). Coercive citation is relatively more frequent in business and management, compared to other fields, and when Editors coerce citations, newer academics are more likely to be targeted (Wilhite and Fong, 2012). It is also a frequent topic among researchers on social media, even attaining meme status (see for example the “Reviewer Two Must be Stopped” FaceBook group). However, there are also those who feel they have not experienced coercive citation requests and are quite sceptical that such a problem exists. The fact that not everyone has experienced a particular problem does not, however, mean that the problem should not be addressed.

We also recognize that reviewers are offering their time and effort with little benefit for themselves, to serve the wider academy. One insightful viewpoint taking the reviewer’s perspective advocates for a review of the reviewing process, with emphasis on gaining greater equity and recognition for the often-unsung reviewer is by Mogaji (2024). We strongly support that perspective, which suggests there is likely to be a group of reviewers who feel overworked and under-compensated for their review efforts. Attaining high-quality reviews requires a willing reviewer pool, and this will not be available if reviewers are inadequately supported.

We therefore close by urging Editors to take action in this area, with the following recommendations:

• Even if Editors are unpersuaded that a review template would be beneficial, we believe that every review should be accompanied by what is very common for reviews in the medical and public health field: a statement that the reviewer has no COI in writing their review. For reviewers who do not have any COI and who do not recommend citation of their own work (we assume the majority of reviewers), this would require no additional effort beyond something as simple as entering or ticking “no”.

• Reviewers should be asked if they have recommended citation of their own work, and if so, to briefly justify that recommendation. For those who do make such suggestions, only a brief rationale would be required, so the additional effort would not be onerous, and would provide the Editor with a means of quickly assessing the value of the recommendation.

• When a reviewer reveals that they have recommended citation of their own work, the handling Editor (whether the Editor or an Associate Editor) should be able to quickly review the recommendation and decide if it is appropriate. That quick review by the Editor or an Associate Editor should easily identify reviewers whose behaviour merits further review by the relevant journal (or, where appropriate, by the relevant publishing house). So, reviewers who appropriately recommend citation of their work would remain above any suspicion of inappropriate behaviour. In contrast, reviewers who repeatedly recommend citation of their own work, and especially those who recommend citation of multiple publications by themself, would merit closer scrutiny. When the suggestions are deemed inappropriate (coercive citations), counselling, warnings or removal from the review panel would be appropriate.

• Journals should provide reviewer guidance, including clear expectations for reviews, and reviewer training tools. Some standardized questions regarding reviewers’ perceptions of the manuscript would help editors to compare otherwise disparate reviewer responses. This would ideally be by the use of a review template, accompanied by open-ended reviewer comments. Editors and/or journals who are reluctant to add a review template may still benefit from including a few standardized short-answer questions for reviewers, as this would ensure that all reviewers of a manuscript address certain key issues that are most important to that journal.

• Editors should consider how they may add additional benefits for reviewers to the reviewing process. While we recognize that Editors bear a large and often minimally compensated role, maintaining a satisfied reviewer pool is likely to make it easier to secure reviewers, thus reducing the Editor’s workload in the long run. In addition to the excellent suggestions offered in Mogaji (2024), we suggest that Editors consider additional forms of recognition for reviewers, such as public accolades on social media that go beyond the one or two top reviewers for a journal, or possibly sending letters of appreciation that the reviewer can share with their Dean, for those who frequently provide valuable reviews for the journal.

For reviewers, we feel that the changes recommended above would not be onerous. If a reviewer has core work in the area, they would still be able to recommend that work to the author, as they are now. They would simply identify this when submitting their review, and briefly justify the inclusion in a sentence or two for review by the Editor or Associate Editor handling the paper.

These changes should therefore only involve very easy action for reviewers who appropriately ask for citations of their work, but the mere fact of being asked to justify a recommended citation is likely to discourage coerced citations, or at the very least, bring the practice to the notice of Editors and Associate Editors, who would therefore be able to better identify and manage reviewers who inappropriately use reviewing as a way to increase their citation counts.

Just as nearly all large businesses have ethics codes and policies to manage potential or actual violations of ethical principles, we believe it is time for academic publishing – a multi-billion-dollar business – to implement policies and processes to decrease the probability of violations of the ethics codes that every major publishing house details on its website. As Macdonald stated as far back as 2015, “…more depends on this peer review than ever before” (2015, p. 264). Nearly 10 years on, that statement could be repeated. It appears that some journals are moving forward in this area, as more reviewer training programs and guidance are being introduced by leading management and marketing journals (see for example efforts by the Journal of Marketing and the Journal of Advertising, among others) but there is more work to be done.

Figures

Figure 1

Possible motivations for, and results of, coercive citation requests