Ghostwriting and Market Values

In recent years, for example, a troubling pattern has emerged in pharmaceutical research: des­pite strong evidence from clinical trials, as well as enthusiastic support from physicians, prob­lematic outcomes emerge when people begin to use a drug (e.g., high rates of cardiovascular disease, diabetes, or hypertension, or suicidal ideation); we later learn that the drug company was aware of the potential for problems before the drug went to market, that they actively worked to downplay the risks of using the drug, and used ghostwritten publications to get the drug quickly and efficiently to market (Elliot and Landa 2010).

By definition, these forms of ghostwriting must remain hidden from view, so we don’t know how commonplace they are. Moreover, there is little consensus about how to understand the ethical implications of ghostwriting. It is widely assumed that moral problems are likely to arise when professional writers are employed to rush a paper to publication, and when research is guided by the interests of pharmaceutical companies. Philosophers and journal editors have worried that using industry-sponsored ghostwriters can compromise the pursuit of truth by centering market values; and it is sometimes suggested that plagiarism, or a nearby violation of academic integrity, occurs when a scientist puts their name on a paper that they have played little role in writing or revising (Anekwe 2010; McHenry 2010).

When professional writing services are used in pharmaceutical research, the roles of various named authors often remain obscure. In several cases, there is evidence that industry interests have played a role at every possible point during the research process: from the specifica­tion of experimental design, to the analysis and interpretation of data, ghost-managed research aims to make it “easier for pharmaceutical companies to use scientific research to market their products” (Sismondo and Doucet 2010: 275). But achieving these ends requires both increasing the prevalence of research supporting a product in the academic literature, and doing so in ways that make use of high-impact journals, with the assistance of prominent scholars, universities, or research institutions. Put differently, within the existing scientific credit economy, ghost­written articles can only be useful as marketing tools if “they appear to come from a disinter­ested source” (Moffatt and Elliott 2007: 27). And where there is evidence of a robust conflict of interest, scientific norms preclude the publication and dissemination of data. To work around these norms, pharmaceutical companies have taken advantage of the fact that many academics are willing to participate in ghost-managed research; as credit seekers, academics “have a strong interest in publications, particularly in prominent journals” (Sismondo and Doucet 2010: 275). And this is what leads to cases where pursuing scientific truth is disincentivized, while the desire to pursue credit by co-authoring problematic research is enhanced.

It is worth pausing to make it clear why someone would willingly accept credit for a paper, when they played no role in its production.

(1) Industry-sponsored and ghost-managed research tends to be placed in higher-impact journals, and it tends to be cited more frequently than non-industry-supported research (Healy and Cattell 2003; Sismondo and Doucet 2010: 278). In part, this may be because strategic “publication plans” are used to shepherd industry-supported research through the publication process, from the presentation of research at high prestige conferences, to the placement of publications in prominent journals; and professional writers are paid to write compelling and readable articles on the basis of this research. Consequently, medical journals may be more likely to publish ghost-managed research. But in part, this may also be because pharmaceutical companies often purchase large numbers of reprints to dis­tribute to physicians (Sismondo and Doucet 2010: 275).

(2) Existing norms governing scientific authorship allow some people to put their names on nearly completed papers. Cultures of honorary authorship allow the leader of a large lab to be listed as an author, even where they have contributed little to a manuscript. And author­ship is sometimes conferred on the basis of theoretical contributions that were made in conversation. There may be good reasons to worry about the prevalence of these practices. But on their own, they need not compromise the pursuit of scientific truth.

Through the interaction of these factors, a situation emerges where a scientist might come to believe that they can get away with putting their name on a ghostwritten paper. And in this context, they have some incentive to use this low-cost strategy of pursuing academic credit. Of course, many scientists will refuse to take part in this kind of research, as scientific norms play a more significant role in their decision-making than do industry norms. However, the social organization of this kind of research does open up the possibility of engaging in questionable research practices, and potentially fraudulent behavior.

Industry-financed writers can take advantage of this situation to produce articles that pro­mote a company’s interests, while a “key opinion leader” accepts credit for a paper that they have played no substantial role in writing (McHenry 2010). Such practices obscure the role that industry interests have played in determining which data are presented and which data are ignored; and this makes ghost-authored papers look like they are ordinary contributions to scientific knowledge. But perhaps more significantly, ghostwriting and ghost-management can bring about the diffusion of collective moral responsibility in a way that is similar to the diffusion of responsibility that emerges in the context of large-scale collaborative research (see Kukla 2012). To see what this amounts to, we return to Dang’s (submitted) tripartite distinction between different kinds of responsibility.

To begin with, where papers are ghostwritten, it becomes much more difficult—if not com­pletely impossible—to figure out who a scientific product should be attributed to. In the most extreme cases, the named authors have not made any substantial contribution to the research design, and they have not played any significant role in the acquisition, analysis, or interpretation of data; further, they have not drafted the manuscript, or made substantial and intellectually significant revisions to it. They have simply signed off on the final version of the manuscript. Put somewhat differently, it is not just that the causal story is complicated (as it was in the cases we discussed in section 25.3), attributable-responsibility is compromised in these cases because the people who are given credit for the relevant research have played no causal role in its pro­duction. Consequently, papers are attributed to people who are in no real position to vouch for the research, or to offer justifications where they are called upon to do so; and this means that there may be no one who is answerable-responsible for the research product.

The research guiding decisions that are made in this context are “rarely, if ever, made by isolated individuals” (Biddle and Kukla 2017: 229). They are typically made by loosely connected networks of pharmaceutical representatives, researchers, scientists, and professional writing com­panies. This distribution and obfuscation of attributability causes a situation where answerability becomes diffused, as no identifiable individual or group has the capacity to produce the reasons that would justify a scientific outcome; and the diffusion of accountability follows, as there is no identifiable individual or group who is a proper target of praise or blame for the scientific outcome, and no identifiable individual or group who is justly accorded reward or punishment for making claims that are true, erroneous, or fraudulent. While there may be some cases where it is possible to figure out who collected, analyzed, and interpreted the data for a ghostwritten paper, the lack of transparency at every level of the research process can make it impossible to figure out who is answerable, and who should be held accountable for problematic research. And pushing on the named author or “key opinion leader” will at most lead to the acknow­ledgment that they appended their name to a research product they did not really contribute to.

As with the cases we discussed in the previous section, these kinds of ghost-managed and ghost-authored research make it difficult to assign responsibility for fraudulent research to any particular individual. Epistemically, it is often unclear who the pertinent claims should be attributed to, even where there is a fact of the matter. In the case of widely distributed research, it is not clear that anyone is actually answerable or appropriately accountable—no one stands in the right kind of relationship to provide a justification or to be held accountable if the results of a study are inappropriate. In the case of ghostwritten research, by contrast, the intentional obfus­cation of authorship makes it at least impractical to design institutional mechanisms that can be used to hold people accountable for the research that they have actually done. Where fraudulent claims emerge, there will be someone who is causally responsible for entering the problematic claim into the scientific record. But it is not the named author; and the process that leads to the production of the claim is likely to be such a complex and intractable mixture of legitimate science and industry-based interests that it will be impossible to track down where the distor­tion was introduced in the research process. What is more, since the group membership and structure is opaque for the same reasons the individual contributions are, it is just as difficult to hold the group answerable or accountable for any wrongdoing. If there are many cases where no individual nor research group may be held accountable for fraud, this leads naturally to the next question: what ought to be done to ameliorate fraud?

25.5