Authors: Daniel D Kim, Maxwell J Smith
Categories: Original Research, Public Health, ethics, COVID-19
Source: BMJ Public Health
Coercion in public a scoping review of definitions
Authors: Daniel D Kim, Maxwell J Smith
Abstract
Introduction
It is common for certain public health measures to be characterised as ‘coercive’, which often carries negative moral connotations that can fuel opposition to their use. Yet, coercion admits of myriad definitions and justifications. This lack of conceptual clarity not only risks the inconsistent implementation and evaluation of policy but can also undermine public trust and weaken the legitimacy of public health governance.
Methods
A scoping review of the public health literature was conducted to understand (1) how coercion is defined in the public health literature and (2) which justificatory conditions, if any, are identified to support the use of coercion for public health purposes. The search strategy covered five databases. Results were reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guideline.
Results
56 publications were included. Definitions appeared most within the contexts of infectious disease (n=30) and public health ethics (n=34). While definitions lacked consistency and often conceptual precision, several publications (n=16) drew on established philosophical accounts of coercion (eg, Wertheimer, Nozick, Feinberg). We identified five elements central to how coercion is commonly understood and invoked in public limiting alternative options; influencing voluntariness; inducing fear of consequences; using state powers to enforce a rule; leaving someone worse off if they do not comply.
Conclusion
This review reveals an important gap in the public health literature. Despite the ubiquity of the term ‘coercion’ in public health, it remains poorly and inconsistently defined. This ambiguity risks the imprecise and at times inappropriate categorisation and evaluation of public health interventions as coercive, potentially compromising both policy effectiveness and public co-operation. Future research could aim to synthesise a definition of coercion agreeable to most public health stakeholders.
Introduction
A common charge made against some public health measures used in response to the COVID-19 pandemic was that they were ‘coercive’. Concern about the coercive force of public health measures is not new, though this was arguably more pronounced and subject to widespread debate in the context of the pandemic. Examples of public health measures subjected to the charge of coercion included lockdowns, vaccine mandates, travel restrictions, isolation and quarantine, where individuals were expected to comply or face consequences such as termination of employment, fines or denial of services.
It is widely accepted that the use of coercion stands in need of justification. And, for some, coercion carries a negative moral connotation, such that to characterise a public health measure as coercive is to suggest it is bad or wrong. In other words, for some, the claim that a public health measure is coercive counts as a reason to not use it. At the very least, it signals that the use of the measure stands in need of stronger justification relative to non-coercive public health activities. Consequently, it is important to understand and make explicit the grounds on which public health measures could be considered ‘coercive’ and the conditions under which the use of coercion in public health is justified.
A robust philosophical literature exists that has examined and developed multiple accounts of the nature and justification of coercion. There is no agreed upon definition of coercion in this literature, and in fact, accounts of coercion diverge on a number of important points. In general terms, coercion is conventionally understood as a technique to control a person’s actions. Prominent areas of disagreement among theorists include the ways in which physical coercion (ie, use of force) differs from volitional coercion; whether offers and not just threats can be coercive; whether only illegitimate threats count as coercion; whether attempts at making someone do something they would not otherwise do must succeed to count as coercion; and how to assess whether a proposal has made someone ‘worse off’ so as to count as a threat and thus coercion.^1^ Moreover, accounts of coercion distinguish between the kind of coercion that qualifies as wrongful interference, the kind of coercion that reduces another person’s blameworthiness or moral responsibility and the kind of coercion that invalidates another person’s consent.^2^ The sheer variation in how theorists understand coercion raises critical questions about how coercion is being interpreted when it is invoked in public health, whether there is any consistency in this regard, and whether coercion, when invoked in public health, bears resemblance to indepth theoretical treatments of the concept (and if so, which ones).
This article is an attempt to begin answering these questions. It does so via a scoping review that aims to understand (1) how coercion is defined in the public health literature and (2) which justificatory conditions, if any, are identified to support the use of coercion for public health purposes. As a descriptive project, this scoping review aims to understand how coercion is understood when invoked in the public health literature. This can be contrasted with the normative project concerned with how coercion ought to be understood and the ethical appraisal of coercive practices. This normative project is crucial given that coercion is an ethical concept, or at least is used in ways that have ethical significance. While beyond the scope of this paper, we hope mapping the terrain of coercion in the public health literature can help inform such analyses.
Methods
A scoping review was chosen for our purposes because this method is best capable of mapping key concepts across a broad literature and identifying knowledge gaps.^3^ This review is based on Arksey and O’Malley’s framework for conducting scoping reviews,^3^ and our reporting of findings followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guideline.^4^
Eligibility criteria and selection of sources
Articles were eligible to be included in our review if they explicitly pertained to public health (as opposed to acute care, non-health issues, etc) and offered an explicit definition or interpretation of ‘coercion’ or ‘coercive’ in relation to public health activities. Articles merely identifying examples of what were considered by authors to be coercive interventions were not included if they were not also accompanied by an explicit rationale for why the intervention was considered coercive. Publications focusing on coercion at the medical or interpersonal level (eg, involuntary psychiatric hospitalisation, coercive abortion, intimate partner violence, etc), often found in clinical care literature or other fields in health (eg, psychiatry, organ donation, etc), were excluded so as to generate findings more specifically and overtly about public health activities and their coerciveness. Finally, articles were eligible to be included only if they were published in English.
Information sources and search
The search strategy was developed after consultation with an experienced librarian at Western University who specialises in health research. Searches were conducted in five databases—MEDLINE Ovid, Embase, Web of Science, Scopus and CINAHL—on 18 January 2025 (see online supplemental appendix A for each search string). Hand searching and checking reference lists of retrieved publications were also performed iteratively throughout the process. Date of publication was not a sampling criterion as we did not want to preclude an analysis of whether and how definitions of coercion may have evolved over time across different public health challenges.
Specific search keywords branched from two public health (eg, “public health”, “population health”, “community health”, “public health administration”, etc) and coercion (ie, coerc* (coercion, coercive, coercing, coerced, etc)). In applicable databases, terms were exploded and relevant Medical Subject Headings terms were included. As our aim was to identify how coercion is understood, we did not include other possibly related words (eg, compulsory, mandatory, involuntary, etc), since it would not be possible to discern the relationship between those terms and coercion if the term coercion was not also used.
Search results were extracted to the Covidence software (2022) where screening was conducted. All results were screened by the two authors (DDK and MJS) of this review. After duplicates were removed, publications were initially screened with the title and abstract, followed by full text review.
Data items and charting process
The following variables were extracted from each retrieved author, year, type of publication (ie, empirical or non-empirical work), public health theme (eg, vaccination, health promotion, etc) and the definition of coercion provided (see online supplemental appendix B for full table). Data were extracted into a Microsoft Excel spreadsheet, completed independently by DDK and reviewed by MJS. We were also interested in exploring the proportion of literature that focused on (1) COVID-19 (given that we expected the focus on coercion to have increased during the pandemic), (2) infectious diseases versus other public health issues (given that we expected discussion of coercion to be more common in the context of infectious disease control), (3) vaccination versus other public health interventions (given that we expected discussion of coercion to be more common in the context of vaccination vs other public health interventions) and (4) whether articles were published in specialist ethics journals (including public health ethics) or generalist public health journals (given that coercion is an ethical concept and expected explicit definitions of coercion to be more common in ethics-specific journals).
We undertook an inductive, iterative analysis of the included texts to identify recurring elements across the definitions and justificatory conditions of coercion. All extracted definitions were reviewed multiple times in full to build familiarity with the data. Preliminary themes capturing common features of how coercion was described were developed and used to categorise definitions. These themes were iteratively refined through comparison across definitions to maintain consistency and to minimise overlap. This process was conducted by DDK, with themes and categorisations reviewed and discussed with MJS to ensure coherence.
Findings
Figure 1 provides a PRISMA flow diagram of the selection process. A total of 4409 publications were imported for screening, and after removal of duplicates (n=2321), 2088 publications were screened for title and abstract. Based on our inclusion and exclusion criteria, 1761 were excluded and 327 full-text publications were assessed for eligibility. Among the 327 publications, 277 were excluded based on the following definition of coercion was not provided (n=255), publication was not in English (n=15), the word ‘coercion’ was not used in full-text and only located in the abstract, which offered no definition (n=3), coercion was strictly interpersonal (n=3) and full-text publication could not be retrieved (n=1). Hand searching yielded an additional six publications that fit our inclusion criteria. Ultimately, 56 publications were included in this review.
Characteristics of publications
Almost all publications were non-empirical works (n=50) (eg, conceptual or theoretical essays), while only six were empirical studies. Of the 56 publications included in this review, 54% of publications (n=30) focused on infectious diseases, and among these, a little over half (n=16) focused specifically on COVID-19; 34% of publications (n=19) focused on vaccination as opposed to other public health interventions; 61% of publications (n=34) were either ethics focused or published in ethics journals (see table 1). The distinction between ethics and non-ethics works and journals was determined by whether the terms ‘ethics’ or ‘bioethics’ were explicitly used in either the title or text.
Key elements across definitions
Definitions provided in each publication are shown in online supplemental appendix B. The most popular established accounts of coercion cited alongside definitions of coercion in the publications reviewed include those from Alan Wertheimer (n=8), Robert Nozick (n=4) and Joel Feinberg (n=4) (see table 2). This is unsurprising as these are arguably the most prominent accounts of coercion in the philosophical literature. While all three accounts are similar in some respects, each has a unique understanding of coercion. For example, Wertheimer defends a moralised account of coercion where the moral illegitimacy of a proposal is a necessary condition for it to qualify as coercion. Hence, it is difficult, if not impossible, to identify instances of coercion without normative assessment.^5^ This is not so for Feinberg, for example, who defends a non-moralised view of coercion. For him, coercion represents a specific kind of pressure imposed on the coercee’s will but does not require this pressure to be morally illegitimate or otherwise ‘wrong’ to count as coercion.^6^ Feinberg’s account accepts that both threats and offers can coerce, whereas only threats count as coercion for Wertheimer and Nozick. For all three authors, direct compulsion (eg, use of physical force or violence) is distinct from coercion. For Wertheimer and Nozick, coercion is present only when it is successful (ie, coercion takes place only when the coercee acquiesces to it).^7^ As a final example, for coercion to be present, Wertheimer requires that the choice forced on the coercee be such that the coercee has no reasonable choice but to succumb, which must be judged using independent normative (ie, ethical) criterion. This is not so for Nozick, for instance, who merely requires that the consequence threatened by the coercer be undesirable to the coercee and that part of the coercee’s reason for not doing what the coercer demands is to lessen the likelihood they will bring about the consequence that has been threatened. Clearly, then, there are significant implications for how measures and activities in public health will be categorised, evaluated and perceived depending on which of these authors’ accounts is cited to support our understanding of coercion. To complicate matters further, among the extracted definitions that referenced these three philosophers, some referred to more than one philosopher/account. This suggests there may be utility in consolidating important components to form a more comprehensive definition; however, this may prove incoherent where accounts are contradictory.
We identified five distinct elements of coercion that emerged from the 56 extracted definitions (see table 3). These elements are not intended to be understood as necessary or sufficient conditions for coercion (partly because it is not necessarily always clear how authors intended these elements to function within definitions of coercion), but rather reflect recurring features identified within and across definitions in the literature. While some of the elements could be consolidated and may not be mutually exclusive, we have separated them as they were often distinguished by authors in the literature. As such, these elements may overlap analytically and do not necessarily operate at the same level of analysis, with some pertaining to individual experiences (eg, influencing voluntariness or inducing fear of consequences) and others reflecting structural or institutional contexts (eg, use of state power). In several cases, more than one element was included in a single definition. ‘Influencing voluntariness’ and ‘inducing fear of consequences’ emerged as core elements across the various definitions, both individually and in conjunction. By contrast, ‘leaving someone worse off if they do not comply’ was used the least. This was rather unexpected considering that this element is a fundamental component of Wertheimer’s conceptualisation of coercion, which was the most frequently referenced.
We also noted a wide range of detail and precision among the definitions. On one hand, some definitions of coercion were as simple as “A imposing pressure on B so that B complies with A’s will” or that “A enforces different restrictions on B’s rights against their will.” These examples were broad in the sense that authors described a continuum ranging from severe threats and penalties to mild persuasion and incentives.810 On the other hand, some definitions were very precise and noted specific requirements, such as the five elements above, for something to count as coercive. These examples were commonly linked to Wertheimer, Nozick or Feinberg and could generally be found at the severe end of the continuum.
Prominent justifications
Beyond definitions of coercion, we were also interested in understanding whether and how the literature discussed conditions that could justify the use of coercion in public health. Of the 56 articles reviewed, 36 provided explicit justificatory conditions alongside definitions (see online supplemental appendix C for a table of justifications and references). There was a broad consensus that coercive measures are generally more effective in promoting or prohibiting certain behaviours but come at higher costs and place unequal or unfair burdens on individuals.^9 11^ It was generally agreed that the more coercive an intervention is, the more justification it typically requires.^12^ Some suggest a paternalistic justification of coercion may exist in some cases (eg, when used as a means to prevent harm to oneself). Illustrative examples included laws requiring seat belts and helmets, as well as those that ban certain foods and beverages.^13^ Another justification was grounded in the harm principle, which states restrictions on individual liberty may be justifiable when used to prevent harm to others. This principle and justificatory condition were particularly prominent in the context of infectious disease control, in which the actions of individuals (eg, refusing to vaccinate or wear a mask) can pose risk to others.1416
Publications also stressed that the use of coercive measures must be necessary and its application must not be discriminatory.1719 Justificatory conditions for invoking coercive interventions relied on ethical principles such as proportionality, least infringement and solidarity. Proportionality means that the level of coercion must be proportionate to the public health threat.^14^ In some cases, coercive measures are viewed as justified if and only if the threat is serious.^13 20^ Least infringement states that public health actors should seek to minimise the infringement of general moral considerations, like liberty, and/or that less coercive interventions ought to be used first to realise a goal before escalating to more coercive ones.^21 22^ When it is evident less coercive interventions are bound to fail, some articles justify the use of more coercive ones (eg, the most effective way possible).^23 24^ Solidarity is underpinned by collective responsibility and highlights the idea of protecting the common interest, even if this comes with individual costs. In the case that people already have a moral obligation to do what was being enforced by the coercive measure, it was argued as being justified.^25 26^
In addition to these core ethical principles cited in justifications for the use of coercion, cost-benefit ratio was another key consideration for justifying the use of coercive public health measures. This typically takes a consequentialist perspective as it locates the justification of a measure or activity in its having favourable outcomes, that is, a favourable ratio of benefits to costs. Costs were commonly understood in terms of infringements on individual rights and freedoms, reduced quality of life, impacts on equity (eg, who bears the burden) and downstream implications such as trust in public health institutions. Finally, we also identified the duty of ‘collective easy rescue’ as a justificatory condition of coercion, which suggests coercive action may be justified to enforce an individual or group making a small sacrifice when doing so is expected to greatly benefit others.^15 26^ Many of these justificatory conditions and ethical principles work together and are commonly offered in combination.
Discussion
Despite the ubiquity of the term ‘coercion’ in public health, only 56 publications identified in our review of the public health literature offered explicit definitions of the term. While the search strategy was designed to be broad and covered multiple databases, our inclusion criteria required that coercion be explicitly defined or conceptualised when mentioned. Many publications implied coercion or provided examples of what were considered to be coercive measures but did not offer a clear definition or explanation as to why or how such measures should be counted as coercive. This substantially reduced the number of eligible articles. This finding reveals an important gap in the public health despite the widespread usage of the term coercion in public health, there is a lack of (1) clear definitions and (2) consistency among the ones provided.
Identifying the multiplicity of definitions of coercion used in the public health literature permits us to reflect on the advantages and disadvantages between them. Adopting a more restrictive definition—for example, one that requires force or violence in order for something to count as coercion—would perhaps most obviously match intuitions about what coercion entails but at the same time clearly limit its application in public health, since force or violence is rarely used in this arena. This would render the concept less valuable. Moreover, it would mean that many public health measures would not necessarily invite the stronger ethical scrutiny or justification that many may think is warranted when otherwise labelling them as coercive (eg, lockdowns, smoking bans, conditional access to services, etc). On the other end of the spectrum, adopting a broad definition that is inclusive of nudging, persuasion and incentives may also risk the concept being less valuable since it fails to delineate a special class of measures that deserve increased scrutiny. Doing so may leave less room for distinguishing the unique moral dimensions that attach to different methods of affecting behaviour and in turn may afford less nuance to the differential ethical appraisal of such methods.
Others have argued that our understanding of coercion in public health should not discount the public’s experiences with, and perspectives on, coercion.^27^ For instance, it would be problematic should public health proceed with an understanding of coercion that seriously misaligns with how the public understands the term, even if the public is conceptually mistaken about what coercion amounts to. Given the inventory of definitions of coercion now developed as a result of this review, future research may be better positioned to understand and situate public (and other stakeholder) perspectives about what is and is not considered coercive in public health.
Ultimately, variation among definitions of coercion in public health risks inconsistently (and, at times, inappropriately) classifying and appraising public health interventions. But this may be unavoidable. Indeed, some have denied there is a ‘single unified conception of coercion’ across all contexts.^28^ However, this does not mean we cannot aim to be explicit and clear when classifying or appraising an intervention as coercive, including explaining why this is so and whether and when such uses of coercion are ethically justified. Doing so will permit reflection, deliberation and potentially even establish areas of greater conceptual and ethical agreement. In contrast, ambiguity or failing to take seriously the different ways in which coercion can be understood may confuse policy discussions, inappropriately rule in or out certain public health interventions under consideration and even confuse the public, which may undermine trust in public health authorities (eg, if some members of the public believe interventions to be coercive but public health authorities claim they are not).
Finally, it is unsurprising that a diversity of justificatory conditions for coercion exist in the literature, but this is not solely because authors were expected to have different views about what justifies coercion in public health. The diversity of definitions found in the literature is likely partly responsible as well; a definition that restricts coercion to the use of physical force or violence is likely to have more exacting justificatory conditions than a definition that counts nudges and incentives as coercive given that the former will tend to be considered more objectionable than the latter. The upshot is that if we are not clear about which subset of public health activities counts as coercive, we are unlikely to find agreement about when coercion is justified.
Limitations
These findings should be interpreted in light of several limitations. As the first study of its kind, our scoping review sought to understand how coercion is explicitly defined in the public health literature. This required that we restrict the retrieval of literature to that which explicitly uses the terms ‘coercion’ or ‘coercive’ and either provides definitions of the terms or offers a rationale for what counts or does not count as coercion. There is certainly much more to learn about how coercion is understood in public health by analysing proximal and related concepts in the literature, like ‘duress’, ‘mandatory’, ‘compulsory’ or ‘involuntary’, as well as interpreting literature that uses the term coercion but does not define it or interpreting literature that could be said to engage with the phenomenon of ‘coercion’ but does not use the term (ie, uses it implicitly). Accordingly, our review offers an answer to how the public health literature explicitly defines coercion, but does not, and cannot, purport to say anything about the broader and more complex question of how the phenomenon of coercion is understood in public health. And it is unclear whether and the extent to which the former accurately represents the latter. The same limitation applies to the identification of the justificatory conditions for coercion. Another limitation is the descriptive nature of the review. Because this review was intended to map how coercion is understood in the literature, it does not assess the normative adequacy of existing definitions or justificatory conditions. Lastly, only publications in English were included, which may have excluded relevant discussions of the concept.
Conclusion
This scoping review examined how coercion is defined in the public health literature and explored the justifications offered for using coercion for public health purposes. Although there were emerging patterns—for example, references to the prominent philosophical accounts of Wertheimer, Nozick and Feinberg and recurring elements identified across definitions—there remains a noticeable ambiguity and variability in how coercion is interpreted. A lack of specificity can not only lead to inconsistencies in the development and implementation of public health policies but can also stymie evaluation and make it difficult to foster public acceptability for public health activities. While this scoping review offers a synthesis and description of how coercion is used in the public health literature, it does not attempt to analyse nor say anything about how coercion ought to be used in public health. This normative project is crucial, and we hope this mapping of the terrain of how coercion is explicitly defined in the public health literature can serve as a starting point for such analyses. Future research may focus on synthesising a definition or at least essential conditions of coercion that are agreeable to most stakeholders and applicable to the public health context.