photo of abstract painting

Literature Review into ‘harmful’, ‘inadequate’ and ‘good’ supervision.

I have reviewed the literature into ‘harmful’, ‘inadequate’ and ‘good’ supervision to have a better understanding of the extent of ‘harmful’ Supervision and what might be needed to address this at an individual practitioner and professional community level.

Ellis et al (2014) composed twenty-one criteria to unequivocally identify incidences where a Supervisor has violated ethical standards on more than one occasion, these include physical threats, sexual relationships, harmful-exploitative dual relationships, cruel and victimising evaluations, using drugs together, public humiliation and acts of racial or cultural discriminating aggression. Ellis et al (2017) presents eleven anonymous accounts of supervises who experienced these criteria and three reactions from self-appointed ‘supervision experts’, drawn from locations across four continents, to identify make recommendations. He defines as Supervision as ‘harmful’ when it results in emotional, psychological, and/or physical trauma, (p.7), significantly he includes harm identified by the supervisee and harm when a supervisee reports one of the criteria identified above. Therefore, he provides evidence that ‘harmful supervision’ is happening internationally, at an alarming rate across multiple mental health disciplines, but also that only 5%-12% of US and 7% of Irish supervises were able to self-identify that Supervision they are currently receiving is harmful and only 25-36% of US and 23% of Irish supervises were able to self-identify it in hindsight (Ellis et al, 2017, p.5). Ellis and his team draw out four problem areas, firstly, the literature focuses on how supervisors should behave, rather than how supervises can recognise ‘harmful supervision’, secondly supervises are uninformed about their rights and responsibilities, thirdly the literature does not point to systemic interventions for dealing with the widespread, systemic nature of ‘harmful supervision’ leaving the supervisee who identifies harmful experiences unsupported and finally asserts traumatised supervises are reluctant to participate in research meaning the problem is, understandably, under-reported.

The responses from ‘supervisory experts’ included opinions that incidents of ‘harmful supervision’ should be left to legal processes, not researchers (therefore not to be shared outside of successful prosecutions) and that it is the responsibility of the supervisee and/or their employer rather than the professional community, I think this points to a systemic stance which protects the supervisor’s power over others’ interests. However, I agree with Ellis when he argues that as the standard of care for supervisors is higher within Ethical Frameworks than the law, professional organisations need to hold harmful supervisors to account, they have a duty of care to supervisees, who are instructed to be in Supervisory relationships and an obligation to clients that ‘adequate supervision’, after all, ‘by definition, all harmful supervision is inadequate supervision’ (Ellis, 2017:5) 

Notably, supervises who reported ‘harmful supervision’ to legal and/or professional bodies did so after competing training or professional recognition, illustrating how entrenched power dynamics at the community level inhibit supervises who recognise harm from acting. However, despite even in these cases employers allowed ‘harmful supervisors’ to continue, demonstrating the power of their leverage and the naivety of expecting supervisees to prevent harmful behaviour (2014:9). 

Ellis et al (2014) then classified reports of  ‘harmful supervision’ by both supervisor-protective and supervisee-protective criteria, when the criteria used protected the supervisor the number of incidents were lower (27%-40%) than the supervise-protective critieria (66-73%) and significantly higher again when there was only a single incidence of harm with 77% reporting a single incidence of harm when a supervisee-protective stance was taken (2014:10-11).

Ellis is not alone, Madison (2009) references Unger’s finding that 50% of supervises have inadequate supervision such as insufficient attention, while 15% supervises report being traumatised by supervision. Nelson and Friedlander (2001) interviewed thirteen Supervision trainers who reported supervisors uninterested in exploring the Supervisory relationship, insufficient Supervision, having to support their supervisor, power struggles with angry supervisors and needing to rely on informal peer support and professionals other than their supervisor. Gray et al (2001) interviewed thirteen psychotherapy trainees and found they all experienced a negative counterproductive psychotherapeutic event which they believed weakened the Supervisory relationship which they felt had a negative impact on their client work, moreover they did not disclose this to the supervisor and concluded the supervisor was unaware of the event. As a professional community we currently do not enable those with less power to speak, so harmed supervisees silence themselves and disengage from Supervisory relationships and professional community, at least until they are in a more powerful position. This is incompatible with the current professional position that Supervision is necessary and ineffective Supervision poses a risk to clients and the supervises themselves.

Ellis (2001) reviews the studies above and problematises their conflation of harmful and inadequate supervision, the criteria he proposed (2014) describes Supervision as ‘harmful’ only when the consequence is “psychological trauma, …functional impairment [and] debilitating general mental or physical health” which, while the behaviour may not be illegal it violates basic professional standards and is in opposition to the purpose of psychotherapy, perpetuating individual and community level violence (p.402). Ellis calls for an epidemiological study of graduate and recent graduates, which, if it validates his estimation that ‘one sixth or more of our trainees are traumatised in supervision’ (p.403). To address this, he proposes that training courses include training psychotherapy students about of their rights, how to use Ethical Frameworks protectively and helping them prepare proactively to the threat of ‘harmful supervision’ (2001; p404). However, Frazier et al (2009) demonstrates traumatised psychotherapy students most frequent response to harm is that “it’s all my fault”, ‘harmful supervision’ is part of a systemic context of harmful training and education generally and needs a systemic response, such as that envisaged in Madison’s (2023) ‘Democratic Focusing’. 

The prevalence of harm, according to the statistics above, speaks of a culture in which supervises entering Supervision are likely to have been personally harmed, to have witnessed colleagues being harmed, or know of Supervision as potentially harmful. Supervisory harm can be normalised, unseen and habitual, happening alongside harmful familial, professional, educational, and organisational experiences. Supervisees may be habituated to functionally defending themselves in relationships. ‘Good supervision’ creates space for supervises to acknowledging past experiences of harmful relationships and experience themselves in a professional relationship guided by PCA values wherein the Supervisor demonstrates and holds space for becoming emotionally safe for ourselves and others. 

Somewhat obviously, Ellis (2001) concludes that supervisors who are emotionally cold, directive and judgmental are ineffective, while ’good supervision’ is attentive to the Supervisory relationship. Jensen (2007) adds that ‘good supervision’ is not a quantifiable technique or limited to any model based on tasks and the sharing of client material but is recognised in a relationship which welcomes individual competencies and experiences through a felt knowing. 


Ellis, M. (2001). ‘Harmful Supervision, a Cause for Alarm: Comment on Gray et al’. Journal of Counseling Psychology 48(4) 401–406. doi:10.1037/0022-0167.48.4.401

Ellis, M. V., Berger, L., Hanus, A. E., Ayala, E. E., Swords, B. A., & Siembor, M. (2014). ‘Inadequate and harmful clinical supervision: Testing a revised framework and assessing occurrence’. The Counseling Psychologist, 42, 434–472. doi:10.1177/0011000013508656

Ellis, M.V. Englann J. Taylor, Dylan A. Corp, Heidi Hutman & Kelsey A.

Kangos (2017). ‘Narratives of harmful clinical supervision: Introduction to the Special Issue’, The Clinical Supervisor, 36:1, 4–19, DOI: 10.1080/07325223.2017.1297753

Frazier, P., Anders, S., Perera, S., Tomich, P., Tennen, H., Park, C., & Tashiro, T. (2009). ‘Traumatic events among undergraduate students: Prevalence and associated symptoms’. Journal of Counseling Psychology, 56, 450–460. doi:10.1037/a0016412


Jensen, P. (2007). ‘Special Section: On Learning from Experience: Personal and Private Experiences as the Context for Psychotherapeutic Practice’. Clinical Child Psychology and Psychiatry.;12(3):375-384. doi:10.1177/1359104507078468

Madison, G. (30 0ct 2023) Embodied Democracy SEA Conference Presentation. 2015, London [Accessed 15 November 2023]

Leave a Reply