My application for Editor-in-Chief of Tobacco Control

I saw the advert for the Editor-in-Chief of the journal Tobacco Control.

I must apply!

What follows is my application for the position in pre-print form. Following helpful comments and critical reviews, I will finalise and submit by the deadline of 31 January 2023.

Okay, cards on the table: I am not a researcher at all, let alone a respected researcher. Nor do I have experience in journal editing, demonstrable or otherwise. In general, I don’t hold academic publishing in particularly high esteem, and I despair at the quality of much of the published tobacco control research. I think these shortcomings are better seen as strengths, not limitations.

I do, however, have the requisite strong vision that would advance the journal’s reputation, profile and mission to publish engaging and informed content. My application to be Editor-in-Chief is based on the vision set out below.

My vision for the journal

This is my application for the position of Editor-in-Chief of Tobacco Control, which for reasons that will become clear, I will refer to as the journal.

• My vision for the journal
• 1. Mission
• 2. Scientific strategy: curiosity and scepticism
• 3. Embrace Open Science
• 4. Better science through greater scepticism
• 5. Citizen science – a role for nicotine users
• 6. A presumption against policy conclusions
• 7. Peer review, quality control and challenge
• 8. Language and communication
• 9. Conflicts of interest
• 10. The tobacco and nicotine industries
• 11. Editorial Board
• 12. Journal name
• 13. Journal format
• 14. Journal content
• Comments and collaboration

1. Mission

Let us start with clarity about what the mission is not. It is not to help to achieve some desired social or policy objective, such as the end of smoking, tobacco, nicotine, or the tobacco industry. Nor is it to advocate for particular policies or “endgames”. Even if these goals might seem to be desirable to some, they are not part of a legitimate scientific mission. The proper place for normative goals of this nature is through personal advocacy or by joining an organisation engaged in advocacy.

The mission of the journal is to be a platform for insight into and organised scepticism about the phenomenon of nicotine use in society in all its many dimensions.

This exploratory mission aligns with the BMJ job spec, which clearly and rightly avoids specifying any political or policy objectives for the journal.

● Providing the highest possible quality of scientific content to augment the knowledge, impact and practice of researchers and practitioners involved in the field

The BMJ job spec

There is no value, and there could be significant harm in augmenting the knowledge, impact, and practice of professionals or the wider public with poor-quality scientific content that has been distorted to secure a policy or political outcome. Tobacco control science has a poor track record in this regard. The right way to meet the demand for knowledge is through organised scepticism and openness.

2. Scientific strategy: curiosity and scepticism

The new editor will be responsible for the following:

● Developing and maintaining an internationally relevant and coherent scientific strategy and vision for the journal

The BMJ job spec

The scientific strategy will focus on research that deepens insights into the phenomenon of nicotine use in society, including biomedical, technological, attitudinal and behavioural, sociological, economic, commercial, historical, communications, policy and regulatory aspects.

The editorial outlook will favour curiosity at the expense of orthodoxy and insightful heresy at the expense of lazy dogma. Subject to high scientific and analytical standards, nothing will be deemed off-limits. Sacred cows will face ritual slaughter as the evidence and analysis dictate.

Where possible, research that supports informed professional practice and rigorous policymaking will take precedence, provided it is of high quality and not excessively prescriptive. A high premium will be placed on research that evaluates real-world policy and practice impacts and encourages learning from experience and insights from related fields.

The journal will carefully demarcate four types of writing in the same way that some good newspapers do:

1. Reporting (what do we find?): neutral and open reporting of the methods, findings and implications arising from studies, trials, dataset analysis etc.
2. Analysis (how do we interpret scientific findings?): attempts to draw objective insights from data and evidence synthesis, though without excessive policy prescription, unless policy appraisal is the subject of the analysis.
3. Opinion (what should be done?): commentary on the implications arising from the evidence, analysis and broader considerations, including ethics, equity, economics, precaution, and value judgements. The journal will seek balancing and contending views, though without offering a platform to cranks or fanatics.
4. Editorial (how should our knowledge system work?): commentary reflecting the opinions of the journal’s editorial staff. To emphasise its neutrality in relation to nicotine science and policy, editorials will be confined to opinions about science, scientific discourse and academic publishing. All editorials will be signed.

3. Embrace Open Science

The journal will embrace open science in both the letter and spirit, drawing inspiration from the eight ambitions of the European Union for open science, as summarised by University College London below.

The 8 pillars of Open Science are:

1. Findable, Accessible, Interoperable, Reusable [FAIR] data
2. Research Integrity
3. Next Generation Metrics
4. Future of Scholarly Communication
5. Citizen Science
6. Education and Skills
7. Rewards and Initiatives
8. The European Open Science Cloud.

UCL Library: 8 pillars of Open Science

The journal will rapidly transform to become the leading open science journal in the field, benchmarking its progress against Nicotine and Tobacco Research and Addiction. The journal will continue to pursue formal recognition of its Open Science status and Plan S compliance. But above all, these principles will guide editorial practice wherever the opportunities arise. To the extent possible, Open Science will be the organising doctrine for the journal, not merely a compliance exercise.

4. Better science through greater scepticism

The journal should not be fooled into misplaced confidence by its relatively high impact factor of about 7 (here). The academic tobacco control community is a large, well-funded interest group with many researchers vested in a particular narrative. That community can provide a substantial audience for comforting falsehoods and non-sequiturs to bolster a dominant (though flawed) narrative through repetitive citations.

What the journal really needs is a large injection of healthy scepticism. Editors and reviewers will be asked to test, for example, the following (not an exclusive list):

• Toxicology: “the dose makes the poison”, and toxicology is a quantitative discipline. Reviewers should ask, “is that exposure a little or a lot? Far too many studies make a meal out of minuscule exposures without considering materiality. Take metals, for example.
• Comparators: where possible, risks should be placed in context with comparisons to smoking or benchmarks like occupational health exposure standards.
• Counterfactuals: researchers must take care to ask what would have happened in the absence of studied behaviour, technology, or policy. Abstinence should never be assumed. This is especially true of adolescent risk behaviours.
• Confounding: “adjustment” for confounding factors is never complete or sufficiently sophisticated to eliminate confounding. For example, “adjusted for smoking status” almost never reflects the duration and intensity of a smoking history.
• Gateway: claims of gateway effects should always be greeted with scepticism and tested against the rival “common liabilities” hypothesis. This is the idea that associations are explained by the characteristics of the person and their circumstances, not by their product choices.
• Causation: there is a high bar for making causal claims, and causation should not be implied without regard to Bradford Hill’s nine criteria (and subsequent developments). Great care must also be taken to avoid implicit or unintentional causal inference through sloppy language (see below) or policy recommendations that can only be based on a causal claim.
• Associations: the journal should not be a venue for the publication of endless spurious correlations (especially if augmented with false implied claims of causation) that arise from confounding by socioeconomic, mental health, family history, minority status or other obvious stressors.
• External validity: have tests been conducted in realistic conditions that reflect the real-world conditions they are trying to investigate? This problem ranges from the power settings and puff topography of vaping devices (see misleading studies of carbonyl exposures) to guessing the impact of market-based nicotine regulation from limited clinical trials conducted in a completely unrealistic context (see the flawed modelling of New Zealand’s proposed denicotinisation law).
• Animal and cell studies: are the findings over-interpreted, vulnerable to unrealistic exposures that could not be tolerated by humans, and ignoring the human capacity for defence and regeneration? See: Why journalists should stop publishing studies conducted with mice by Arnold Ventures.
• Populations: care should be taken to identify the limitations of study populations (e.g. Twitter users) and selection effects; for example, “dual users” are not the same as people who have quit already or quit easily.
• “The plural of anecdote is not data“: this misguided conceit will not be used to downgrade first-person testimony, user experience, qualitative research and richer accounts of lived experience that are not possible to convey in surveys.

Obviously, a full account of the repeated flaws in tobacco control research is beyond the scope of this humbly-submitted job application. Within six months, the journal editorial office will issue authors, reviewers, and editors with a guide to bad science in the field with a view to raising the quality of submissions and reviews before they even reach the journal. It is regrettable, but this is necessary.

5. Citizen science – a role for nicotine users

One of the eight principles of Open Science is the embrace of “citizen science.” The European Union explains this as follows:

Citizen science can be described as the voluntary participation of non-professional scientists in research and innovation at different stages of the process and at different levels of engagement, from shaping research agendas and policies, to gathering, processing and analysing data, and assessing the outcomes of research. Active engagement with citizens and society has the potential to improve research and its outcomes and reinforce societal trust in science. It can increase – relevance and effectiveness by ensuring that R&I aligns with needs, expectations and values of society – creativity and quality by enlarging the collective capabilities, the scope of research and the quantity and quality of data – transparency, science literacy and confidence of the public in research

European Union. Citizen Science: Elevating research and innovation through societal engagement. 2020

In the case of tobacco and nicotine science, this would mean drawing on the formal and tacit expertise of nicotine users. This would help, for example, in situations where researchers have failed to understand how vaping products are used in practice and why machines can create unrealistic operating conditions. It would help researchers with a better understanding of why people use nicotine and a more textured insight into what tobacco control policies, such as tobacco taxation, mean at the individual or household level. It would stop researchers from asking survey questions that make no sense to users.

The historical antecedents of the modern tobacco control movement are partly grounded in “non-smokers’ rights” movements in which smokers were reviled as a polluting enemy to be thwarted. Smokers were poorly organised and often formed into smokers’ rights front groups by the tobacco industry, thus adding distrust to disgust. That has to stop. Nicotine users are legitimate citizen stakeholders in nicotine science and policy analysis.

Researchers would be encouraged to consult users and ask for feedback on their proposed studies or surveys. Nicotine users would be encouraged to comment on relevant papers and to collaborate with reviewers. A representative would be included on the journal’s editorial board.

6. A presumption against policy conclusions

Policy proposals are too often casually added to the conclusions or discussions of data papers. They should not be. The journal will make a strong presumption against the publication of policy conclusions as an add-on to data-orientated studies. This is because the scope of research papers is almost always too narrow to draw a policy conclusion. According to the OECD:

It is rare that scientific evidence is the only consideration in a policy decision and, particularly for complex issues; many interests may have to be balanced in situations where the science itself may be uncertain.

OECD, Scientific Advice for Policy Making, 2015

Making a policy proposal requires the exhaustive and holistic analysis that goes into policy or regulatory impact analysis. This includes consideration of costs and benefits, counterfactual design, cost-effectiveness, trade-offs, insights into foreseeable unintended consequences, equity considerations, approach to precaution and uncertainty, impacts of errors (false positives, false negatives), isolation of value judgements and considerations of law and policy design. Policy appraisal and evaluation are demanding disciplines in their own right (for example, see UK Treasury Green Book (appraisal), Magenta Book (evaluation) and Orange Book (risk)).

The main useful function of the policy proposals embedded in most research is to provide an implicit declaration of the priors and biases of the authors.

7. Peer review, quality control and challenge

The journal will dispense with the fiction that anonymous pre-publication peer review functions as a guarantee of quality. The “business model” of peer review is broken, with poor professional or economic incentives to do it well, but with the risk that peer review forms an anti-scientific gatekeeping function. Here is Richard Smith, former editor of the BMJ:

Peer review is faith not evidence based, but most scientists believe in it as some people believe in the Loch Ness monster. Research into peer review has mostly failed to show benefit but has shown a substantial downside (slow, expensive, largely a lottery, wasteful of scientific time, fails to detect most errors, rejects the truly original, and doesn’t guard against fraud)

Richard Smith, The optimal peer review system? 8 November 2016

Here are my humbly-submitted proposals for a better approach.

• In summary, the journal must become a trusted source… under my editorship, there would be a strategic shift in focus to the interests of readers and users of the evidence, with less emphasis on protecting the reputations of authors, peer reviewers and editors. For that reason, we would expect more exacting and open peer review and more post-publication criticism, with more corrections, expressions of concern and retractions as appropriate.
• Preprints. The journal will welcome and encourage papers first published in pre-print form and subject to open peer review. The journal will consider the utility of establishing a pre-print facility specific to nicotine science and policy. See an example of a journal, JMIR, that has a process for migration from pre-print to “accepted” status within the journal family: JMIR Formative Research (process). This may be combined with giving authors the final say (below).
• Methodology reviews. The journal would expedite review and publishing for institutions that have conducted methodology reviews. Where efficient or there are doubts about methodology, the journal would conduct an initial methodology review, with full peer review only undertaken if the methodology is sound. See: Daniël Lakens, Is my study useless? Why researchers need methodological review boards, Nature, 2023.
• Open peer review. The journal will publish peer-review comments and responses (including disagreements with peer-review comments). Reviewers will be publicly credited and should consider reviews to be part of the body of their academic work. All versions of papers submitted will be published. This system is similar to that used by the F1000 platform.
• Protection of reviewers. In rare cases, reviewers will be able to request anonymity with a good reason (e.g. to protect against bullying or academic retaliation).
• Ending accept or reject decisions. The journal will offer a system where the editor does not make an accept or reject decision, and the author has the final say over publication. This is the logical extension of a system based on pre-prints and open science. This approach will be adopted by the E-life journal, a Nature publication, to allow the focus instead on public reviews and assessments of preprints.
• Limitations. Critical limitations will no longer be confined beyond a paywall or buried deep in the discussion. The most critical limitations will be included in the abstract to help orientate the reader and media on the significance of the findings.
• Rapid responses. The journal will use the journal’s “Rapid Responses” function to provide a post-publication peer review facility or links to a specialised facility like Qeios. The editor will no longer use moderation to filter or delay criticism, even if it embarrasses the journal’s editorial and review procedures. Moderation rules and functions would be similar to PubPeer (see FAQ). A recent fatal critique of a Tobacco Control paper posted on PubPeer (here) was also submitted as a rapid response to Tobacco Control on 9 November 2022. At the time of writing, more than three months later, it has not been published as a rapid response. How rapid is that? In future, rapid will mean within 24 hours if moderated or instant if not (see below).
• Opening up rapid responses. Rapid Responses would be open to anyone with an ORCID identity without moderation, though breaches of content guidelines would be addressed retrospectively through a complaint-driven review system.
• Right of reply. Any entity, programme, project or person criticised in a published article shall have an unconditional right of reply. On fairness and balance grounds, the right of reply will override any policy limiting publication by tobacco or nicotine companies. Authors making criticisms will be expected to follow basic journalistic practice and put concerns to those criticised prior to publication.
• Corrections. Corrections help to maintain the integrity and scientific value of the published literature. The journal will encourage or require corrections or clarifications where there is evidence of methodological flaws, misinterpretation or ambiguity in reporting findings, or statements falsified by later evidence. The editors will make liberal use of expressions of concern to signal emerging issues with papers.
• Retractions. While following COPE retraction protocols, the journal will deliberately lower the bar for retraction and encourage retractions where the published paper distorts or debases the body of knowledge in the field. Except in cases of research misconduct, author-driven retractions will generally be welcomed without stigma or shame but celebrated as a mark of scientific integrity.

8. Language and communication

The journal will pay close attention to language so that it facilitates scientific understanding rather than obfuscates or confuses readers. The journal will confront poor language practice on multiple fronts:

• Terminology. The terminology and constructs used in a given field should be consistent and convey the same meaning within and between journals. The journal will insist that its authors and reviewers, and editors comply with a formal ontology, such as the AddictO ontology.
• Causal language. The journal will not allow careless, ambiguous, or contrived use of language implying causation where no causal relationship can be justified by the evidence presented. Excessive use of causal language is common in the scientific literature (Haber et al. 2022), though with potentially lethal consequences in tobacco control. There are many examples in the tobacco control literature or implicit or explicit causal claims where none is justified. See the excellent critique of one especially egregious case by Criticher & Siegel. 2021.
• Inappropriate conflations. Some terminology can conflate meanings in a way that obscures rather than enlightens. Three examples follow: (1) “tobacco product” should not include products that do not contain tobacco; (2) the risk characteristics of smoking products should not be attributed to all tobacco products; (3) “EVALI” was a tactical construct introduced by an American federal agency, the CDC, to conflate nicotine and cannabis vaping during a lung injury outbreak that affected the latter but not the former. The guiding principle will be that language should highlight, not obscure, material differences.
• Risk communications. Careful attention will be paid to the use of words and phrases like “safe”, “not safe”, “harm”, “harmless”, and “less harmful”. Very little in life is completely safe, but some behaviours are more or less harmful than others. It is too easy to make statements that are factually true but misleading, incomplete or out of context and easily misinterpreted by an audience. The journal will try to communicate risks accurately and proportionately, taking into account likely reader perceptions.
• Press releases. Given that publication in a journal is a form of validation, the journal will require that its authors do not go beyond their findings in press releases or other communications related to published articles. Authors will be asked to publish links to their press materials or blogs on the journal website either via rapid responses or added as supplementary files or links. These would then be open to post-publication review and criticism along with the substantive paper.
• Loaded language. Certain words can be politicised and used to convey value judgements and stigmatise critics. The journal will not tolerate the use of “loaded” language to convey political messages – examples include “youth vaping epidemic”, “kids” or “children” to refer to adolescents or the use of addict or addiction in some contexts.

9. Conflicts of interest

The journal will no longer indulge the idle misconception that conflicts of interest apply only or primarily to commercial interests. In fact, conflicts of interest are pervasive and inescapable (see Kozlowski, 2016 on the “Conflict-of-Interest Pandemic”).

Conflicts of interest are not even necessarily undesirable – they may simply arise from expertise that commands a high market value.

The journal will be clear that conflicts of interest are disclosed for the purposes of transparency and to aid informed scepticism. COIs are not for use in delegitimising research, which should always be valued on its merits. Nor should COIs be used for making assumptions about the motivation or integrity of researchers. Many researchers work with both high integrity and declared conflicts of interest.

The community of notionally independent tobacco control academics and advocates (which are often indistinguishable) is riven with real conflicts, even though these are never disclosed and rarely understood. Such conflicts include:

• Policy positions. Longstanding personal policy priors or dearly held scientific theories create an interest in not having to recant or admit previous errors or misjudgements. This conflict arises because consistency is overvalued compared to agility in responding to changing fact patterns.
• Motivated funders. Funding by private foundations where the principals have strong policy views creates financial conflicts. Findings discordant with the foundation’s worldview are more likely to result in the loss of future grant opportunities than to change the foundation’s worldview.
• Interventionists. Funding by regulators will encourage findings that support regulation or intervention. Research that concludes “nothing to see here” is unlikely to secure future grants.
• Existential. The field of tobacco control exists mainly because of the harms arising from smoking (cancer etc) and the historic malfeasance of the tobacco industry. Without harm or industry malpractice, there is no more justification for tobacco control than coffee control. Livelihoods, careers, grants, university departments, conferences and personal prestige create a potentially competing interest in finding harm and industry wrong-doing simply to keep going.

10. The tobacco and nicotine industries

The journal will not accept the dogmatic principle set out in the guidance on Article 5.3 of the WHO Framework Convention on Tobacco Control:

Principle 1: There is a fundamental and irreconcilable conflict between the tobacco industry’s interests and public health policy interests.

Guidelines for implementation of Article 5.3 of the WHO Framework Convention on Tobacco Control

Not a scientific principle. This “principle” is a political statement, not an indelible scientific finding immune to challenge. If it was ever valid, it should not be assumed valid indefinitely. If it applies to anyone, it applies to the government officials who foolishly accepted it, not to free-thinking academics. It should not be a barrier to open-minded inquiry or a prevent contact with the industry. Without such contact, most tobacco control experts will remain naive and continue to have little grasp of the ongoing changes in the marketplace and industry.

Alignment of public health and industry goals. The principle, as stated above, is open to easy falsification, for example, by reference to the experience of snus in Scandinavia or the United States FDA’s designation of certain tobacco products to be “appropriate for the protection of public health“. It is possible in some areas for public health and tobacco industry interests to align in key respects: for example, the migration of nicotine products to smoke-free delivery and the diversification of tobacco industry interests to reduce corporate dependence on cigarette revenues.

Improving public health understanding of the industry. The tobacco industry should be regarded with sceptical curiosity as a complex of economic interests worthy of study but not as an enemy (or an ally) to public health. There is no monolithic “Big Tobacco” but a high degree of heterogeneity within and between companies. One important aspect of a renewed mission would be to improve understanding of the strategies, incentives and constraints of these commercial entities in the nicotine field. The current state of knowledge and understanding is pitifully weak, based on cartoon villain stereotypes and a naive understanding of public companies.

Publication of industry research. The case for banning the publication of industry science is vacuous, and its effects are harmful. The industry produces high-quality science for regulatory and product stewardship purposes and is generally cautious in its interpretations. It is harmful to obstruct the publication of such research because it denies readers insights relevant to their field they may not obtain elsewhere. It also denies the opportunity to scrutinise the industry and to mount substantive criticisms, if there are any. It is now rare for tobacco industry science to be subject to credible criticism. However, a system based more on pre-prints, Open Science, and post-publication review would allow substantive criticisms of the industry to be made, debated, and defended. But it would also put an end to the anti-scientific a priori assumption that everything the industry does must be bad or deceitful without even needing to read or understand it.

11. Editorial Board

The journal’s editorial board needs a significant shake-up. It must become more representative of the wide range of views and capabilities in the field. The online listing is out of date, suggesting editorial governance for the journal has been neglected. Those listed on the board need to be actively engaged in maintaining editorial standards, the mission of the journal and its scientific strategy.

There are whole categories of editorial board members that are unnecessary or undesirable, creating barriers to scientific rigour and innovation. For example, “Emeritus” has no meaning on a board, and “Consulting Editors for Policy Advocacy” should have no place in a scientific journal.

A refreshed and streamlined editorial board would:

• reflect a balance of opinions on nicotine and tobacco policies, notably reflecting different views of the direction of policy and practice regarding nicotine;
• include ‘citizen scientist’ representatives of nicotine consumers
• meet every six months online;
• hold the editor accountable for the development and implementation of the mission and strategy;
• consider and deliberate on complaints and disputed requests for correction or retraction with the COPE framework;
• provide an annual report for consideration by readers and the publisher.

12. Journal name

The journal should change its name to something more appropriate for the next twenty years. The name Tobacco Control is not ageing well.

• Tobacco. The subject of interest is the phenomenon of nicotine in society. Tobacco, in its various forms, represents a subset of nicotine products.
• Control. The word “control” suggests a political agenda and preferred Modus Operandi – to impose controls using the legal, informational, and regulatory resources of the state. But what if the optimum approach is to liberalise, respect the autonomy of informed choice, or conclude that no control is justified on risk grounds? A new name should reflect the full range of societal choices available for the management of nicotine as a recreational stimulant.

Though choosing a new name would require a more deliberative process than this humbly-submitted application allows, “Nicotine Science and Policy“, “Nicotine and society“, or just “Nicotine” would be more appropriate in my view.

13. Journal format

There is no point in producing a printed version of this journal. It should move to an online-only format with a focus on the electronic publication of as much related material as possible (datasets, code, more detailed methods descriptions to facilitate replication, press releases etc). There would be no need for type-setting. PDFs, if wanted, would be generated automatically from the online versions (most pre-print servers do this).

Except for specially-themed editions, there is little merit in even publishing discrete editions. A thematic keyword system would be a better way of grouping research than publishing editions containing papers with completely different themes.

14. Journal content

Several sections of Tobacco Control are superfluous (front cover, News Review, The Lighter Side) or betray bias (Industry Watch, Ad-Watch) and should be jettisoned or recast.

The journal content would be dominated by work in progress and subject to peer review and revision. There would be a churn of revision, correction and retraction as authors advanced towards a stabilised and finalised contribution to the published literature. Analysis and commentary would form a valuable forum for debate and criticism. The feel of the journal would capture the dynamism of a scientific mission grounded in curiosity and scepticism in a rapidly evolving field.

Comments and collaboration

I would welcome comments that would help me to strengthen and finalise my application. In a spirit of openness and collaboration, I invite other applicants for the role to draw on this material and take their best possible shot at the job. A rethink and a reset are long overdue.