Online mandatory “antisemitism and racism training” for the NHS: serious proposal or performative “sheep dip”?

Racism and antisemitism are incompatible with the founding values of the NHS. They damage patient treatment and access to care, undermine organisational effectiveness and cause serious harm to staff. https://www.rogerkline.co.uk/post/not-an-optional-extra-the-price-of-not-tackling-race-discrimination-in-the-nhs

The Government has proposed that a cornerstone of responding to both is to be mandatory online training on “antisemitism and racism” for all 1.5 million NHS staff in England.

The NHS rests on a bedrock of research. Research strongly suggests this proposal is likely to be ineffective, potentially counter-productive and is likely to become a performative alternative to the sustained evidence-based work that is needed to address racism and antisemitism in the NHS.

Context

There is a wealth of evidence that overt racism (abuse, harassment and violence) is increasing whilst covert racism remains embedded in recruitment, career progression, discipline, bullying, the response to speaking up and the everyday “microaggressions” that NHS Black and Minority Ethnic (BME) staff report. There has almost no improvement in the measures of NHS race discrimination since data began being systematically collected in 2015 or since a Labour Government was elected in July 2024. Moreover, there is evidence of an increase in antisemitism being reported in the NHS especially since the start of the war in Gaza.

Racism and antisemitism have deep roots in society. Modern day racism has its roots in the slave trade of 500 years ago, developed as a means of justifying the murder and enslavement of peoples who were described as morally and intellectually inferior - and subhuman.

Antisemitism has roots several centuries further and can be traced back to an early Christian belief in Jewish culpability in the death of Christ. The Nazi Holocaust may be more than eight decades ago but its memory still scars the experience of Jewish communities in this country.

Both experiences are deeply embedded in the culture of our country and inevitably impacts the lives of the 28% of NHS staff of BME heritage and the Jewish members of the NHS workforce. The focus of racism may vary (thus the recent steep rise in Islamophobia) but its influence and harm remains. It was in that context that the government announced (16th October 2025) that:

“All 1.5 million NHS staff will be required to complete updated mandatory antisemitism and antiracism training, with existing equality, diversity and human rights programmes” https://www.gov.uk/government/news/government-to-tackle-antisemitism-and-other-racism-in-the-nhs

The purpose of training on diversity and discrimination

This initiative raises numerous questions. The Government statement refers to tackling “antisemitism and other forms of racism” whereas the NHS England response to it references a “request for action on racism including antisemitism”. The former is focused on antisemitism whilst the latter is focused on racism including antisemitism. Neither makes more than a passing reference to Islamophobia (14 words in 1594 words) which is arguably a key element of current expressions of racism by and towards NHS staff.

A second question is whether the focus of the planned training is on overt racism and antisemitism (explicit abuse, violence, harassment, “racist banter”) or is it also on tackling covert racism and antisemitism, the more hidden and subtle but no less damaging discrimination which is ubiquitous in everyday NHS working life?  This is important since the approaches needed might be quite different depending on whether the focus is on the former (tackling overt “hate” is prominent in the Government statement) or on the latter which may nor manifest itself as “hate” at all and can exist in unintended ways unchallenged by staff who believe themselves to not be racist at all

The NHS has relied upon a set of policies, procedures and training to mitigate or prevent workplace discrimination with a singular lack of success. The approach is dominated by methodological individualism – an ineffective reliance on individual raising concerns and the NHS effectively responding to them as Kline (2023) argues  https://bmjleader.bmj.com/content/7/4/314 

This approach is not proactive and preventative and it is frequently not even safe for individuals to raise concerns which might be addressed as Warmington and Kline (2024) show https://www.brap.org.uk/post/toohottohandl This approach is simply not going to work without embedding evidence based intervention, accountability and leaders who can demonstrate thy care = and has resulted in glacial progress across most metrics. https://www.england.nhs.uk/publication/nhs-workforce-race-equality-standard-2024-data-analysis-report-for-nhs-trusts/

One further tricky question is how antisemitism is defined. The Government has adopted one definition but its implications are contested since, for example, there is a fierce debate (including amongst Jewish people of whom I am one) about whether, and to what extent, criticism of the Israeli Government actions on Gaza might reasonably be regarded as antisemitic.

The effectiveness of training

There are many hundreds of academic studies which have sought understand whether (and if so what sort of) training can mitigate racism and discrimination. They are part of a wider debate, within healthcare, for example.

In a recent systematic review of evidence regarding the efficacy of statutory and mandatory training in health and care: Ashley et al (2024) concluded that

Mandatory training is considered fundamental to establishing and maintaining high standards of professional practice. There is little evidence however, of the training either achieving its required learning outcomes, or delivering improvement in outcomes for patients. Whist organisations may be hitting their compliance target for mandatory training, is the purpose missing the point? https://www.tandfonline.com/doi/full/10.1080/0142159X.2024.2331048

They found:

"A lack of consensus regarding mandatory training rationale, core topics, duration, and optimum refresher training period. Currently, mandatory training does not consistently translate to widescale improvements in safe practice or improved patient outcomes.

"The rationale for mandating training programmes remains undefined. The assumption that mandatory training is delivering safe practice outcomes is not supported by studies included in this review.".

An earlier US study by Peterkin (2014) came to similar conclusions

"The VA (Veterans Affairs) biennially requires every employee to complete a combined 5.5 hours of computer-based training on the topics of information security, workplace harassment, and government ethics. The annual cost for the three core trainings has been estimated at $40 million. Yet, the VA has neither generated nor found any evidence to establish that these mandatory trainings improve workplace performance." https://www.ncbi.nlm.nih.gov/books/NBK384612/pdf/Bookshelf_NBK384612.pdf 

Training or development or both?

Training that seeks to influence behaviours and personal development may be quite different to training which focuses on additional knowledge and skills. To influence behaviours and understanding (including of racism and antisemitism) a focus on personal development rather than just the accumulation of knowledge is necessary, focused on exploring hidden assumptions, emotions and behaviours. To do that effectively requires much more than sharing information and skills. Such work would seek to undermine assumptions and ways of thinking. It can be stressful and upsetting in ways that the imparting of skill and information is not. That in turn requires a safe space and the opportunity to share both the challenges and develop their understanding in ways that may be uncomfortable. Such work is necessary of participants are to be effective allies in tackling racism.

Two central findings emerge from research on tackling racism. Tackling racism requires individuals to change themselves not just absorb information but will also need to give the confidence and understanding that much discrimination is subtle, covert, and difficult to prove but nonetheless devastating in its impact. Research is clear moreover that a primary reliance on policies, procedures and training will not in isolation change the culture and behaviours of an organisation. That underpins covert discrimination. Tackling antisemitism similarly requires, amongst other elements, an understanding of its origins, why it is so resilient to challenge and the deep harm it causes.

Methodological concerns and impact.

In a systematic review of diversity training outcomes Alhejji, H et al (2015) found:

(a) a literature that is fragmented and diverse in terms of publication outlets; (b) researchers conduct diversity-training outcomes research in a diverse range or organizations, sectors, cultural and training contexts; (c) studies primarily reflect the business case or learning perspectives; and (b) existing studies have significant methodological limitations.

They specifically found that “there was limited evidence of social justice outcomes”. https://researchonline.gcu.ac.uk/en/publications/diversity-training-programme-outcomes-a-systematic-review/

Yates et al (2022) suggest that how training of any kind is evaluated is problematic.  In a scoping review of the use of self-evaluation, they concluded:

"whilst self-Assessment is often assumed to be essential for learning, this assumption has been extensively rebutted with a lack of correlation between self-assessment and competence." https://www.tandfonline.com/doi/10.1080/0142159X.2022.2093704

Effectiveness of diversity training.

“Anti racism training” Is not new. There was a wave of such training in the UK and elsewhere some forty years or so ago. The most authoritative review *Wrench J et al (1987) raised serious questions about its effectiveness: https://search-library.ucsd.edu/discovery/fulldisplay?vid=01UCS_SDI:UCSD&docid=alma991016409329706535&context=L

Specific programmes for tackling antisemitism exist but there is little academic evaluation or consensus on what is effective – as one report put it “Everyone has a plan to fight antisemitism. Few have studied what actually works”  Asaf Elia-Shalev Jewish Telegraphic Agency January 6 2025 https://www.jta.org/2025/01/06/united-states/everyone-has-a-plan-to-fight-antisemitism-few-have-studied-what-actually-works

The evidence on diversity training generally is not encouraging.

Two-thirds of US human resources specialists three decades ago reported they believed diversity training has no positive effects on women’s or ethnic minorities career progression and on managerial diversity. Sara Rynes and Benson Rosen, “A Field Survey of Factors Affecting the Adoption and Perceived Success of Diversity Training,” Personnel Psychology 48, no. 2 (1995);

There is ample evidence that training alone does not change attitudes or behavior, or at least not by much and not for long. A review of 985 studies of anti-bias interventions (Paluck and Green 2009) found little evidence that training reduces bias. https://www.researchgate.net/publication/23316190_Prejudice_Reduction_What_Works_A_Review_and_Assessment_of_Research_and_Practice

Kulik C et al (2008) found 27 studies that documented improved knowledge of, or attitudes toward, diversity, but most found small, short-term improvements on just one or two of the items measured. https://www.researchgate.net/publication/285177024_Common_Goals_and_Golden_Opportunities_Evaluations_of_Diversity_Education_in_Academic_and_Organizational_Settings 

Some recent studies used the implicit association test (IAT) to assess whether unconscious bias can be affected by training. Forsher P (2018) in a meta-analysis of 426 studies found weak immediate effects on unconscious bias and interestingly weaker effects on explicit bias. Effects dissipated very quickly. within a few days. Most of these studies look at interventions that mirror corporate and university training in intensity and duration. Forscher suggests, moreover, that change in unconscious bias training does not lead to change in discrimination. The key to improving the effects of training is to make it part of a wider program of change. https://pubmed.ncbi.nlm.nih.gov/31192631/

Diversity training may be counter-productive.

Dobbin F and Kalev A (2016) argue that diversity and harassment training programs frequently focus on blame, legal consequences, and unconscious bias. Employees are often told they are biased, and managers are informed that they will be held accountable if employees are accused of discrimination. This is counterproductive because employees tend to resist these messages, inadvertently increasing discriminatory behaviour. Employees may leave these courses either feeling they have been accused of being bad people or deciding how questionable their behaviour can become before crossing legal thresholds. https://hbr.org/2016/07/why-diversity-programs-fail

Training intended to tackle bias may also activate stereotypes i.e. do the opposite of what is intended. Trying to get people to asking people to suppress stereotypes may well tend to reinforce them Macrae C (1994)  https://www.researchgate.net/publication/232482010_Out_of_Mind_but_Back_in_Sight_Stereotypes_on_the_Rebound

Plaut, V et al (2011) found diversity training typically encourages people to recognize and fight the stereotypes they hold, and suggest this may simply be counterproductive. https://psycnet.apa.org/record/2011-08638-001

Extensive research show that efforts to control what staff think and how they behave meets resistance whereas when a goal is seen as coming from within the organisation, with their involvement and agreement, motivation for achieving that goal increases. This may be specifically true for diversity training. Legault L et al (2011) found that when pursuing a goal as originating internally, commitment rises, but when they frame motivation as originating externally, rebellion increases and produced more explicit and implicit prejudice than did not intervening at all..  https://pubmed.ncbi.nlm.nih.gov/22123778/ 

See also on the risks of backfiring: Legault L,  Chung J, et al   (2007). https://journals.sagepub.com/doi/abs/10.1177/0146167206298564?download=true

Kulik C (2008) op cit. found employers mandate training in the belief that people hostile to the message will not attend voluntarily, but research suggests that forcing them to come may do more harm than good.

The UK Government Equalities Office and The Behavioural Insights Team (2018) reviewed which interventions were more likely to reduce gender equality in UK workplaces The concluded

“Diversity training Diversity training can help raise awareness but is unlikely to change behaviour. Some research in the US has found that mandatory diversity training either does not change the number of women in management positions, or actually reduces it.

"This backfiring may be for a number of reasons. It may be because people resent being made to do something and so do not take the training seriously. The training might also bring to mind unhelpful stereotypes which people then act upon, or the training might make people think that the organisation has now solved its diversity problems. They did not  diversity training as an effective means of doing so”. https://www.bi.team/wp-content/uploads/2018/06/GEO_BIT_INSIGHT_A4_WEB.pdf

A more recent review from the same office (2025) makes no mention of diversity training altogether.  https://www.bi.team/wp-content/uploads/2018/06/GEO_BIT_INSIGHT_A4_WEB.pdf

What might have an impact?

One relatively effective intervention did appear to demonstrate sustained improvement through diversity training. Devine P et al (2012) found that a more extensive curriculum, based in strategies proven effective in the lab, can reduce measured bias. Their 12-week intervention, took the form of a college course and included a control group, with the most effective participants being those who were concerned about discrimination and who engaged fully. https://www.sciencedirect.com/science/article/abs/pii/S0022103112001369

Bezrukova et al (2008) suggest diversity training is more effective if it not only builds awareness about biases, but also develops people’s interpersonal skills in a way that reduces bias. It also suggests that training interventions must not be isolated activities, but instead integrated with other diversity-related initiatives into broad cohesive programmes. Katerina Bezrukova, Aparna Joshi and Karen A. Jehn, “Can We Teach Diversity? A Review of Diversity Trainings in Educational and Organizational Settings,” (2008), Working Paper, Psychology Department, Santa Clara University.

In a subsequent meta-analysis of 260 independent samples Bezrukova et al (2016) found that training effects on cognitive learning did remain stable but the effects of diversity training on reactions and attitudinal/affective learning decayed over time and that positive effects of diversity training were greater when training was complemented by other diversity initiatives, targeted to both awareness and skills development, and conducted over a significant period of time  https://pubmed.ncbi.nlm.nih.gov/27618543/

A UK Equality and Human Rights Commission review of Unconscious Bias Training (UBT) (Atewologun et al (2918) found that:

• UBT is effective for awareness raising when using an Implicit Association Test2 (IAT) followed by a debrief, or more advanced training designs such as interactive workshops or longer-term programmes to reflectively reduce biases. UBT can be effective for reducing implicit bias, but there is no evidence that it can eliminate it;

• UBT interventions are not generally designed to reduce explicit bias and those that do aim to do so have yielded mixed results or very small effects;

• Using the IAT and educating participants on unconscious bias theory is likely to increase awareness of and reduce implicit bias;

• The evidence for UBT’s ability effectively to change behaviour is limited. Most of the evidence reviewed did not use valid measures of behaviour change;

• There is potential for back-firing effects when UBT participants are exposed to information that suggests stereotypes and biases are unchangeable. https://www.equalityhumanrights.com/sites/default/files/research-report-113-unconcious-bais-training-an-assessment-of-the-evidence-for-effectiveness-pdf.pdf 

In a Chartered Institute of Personnel Development (CIPD) review, Gifford J et al (2019) summarised the research evidence and noted that, while unconscious bias training can increase people’s awareness and knowledge of diversity issues, this evidence is generally based on self-reported measures, which may not be reliable especially when done as an isolated activity.

They concluded that there is no conclusive evidence that diversity training changes attitudes. The CIPD review noted that there is typically no sustained impact on behaviour and emotional prejudice following UBT, which is not enough in itself to create diverse and inclusive organisations.  They suggested that emerging evidence suggests that a perspective-taking approach, which promises to raise awareness and empathy, has potential to create buy-in for EDI strategy.  https://westminsterresearch.westminster.ac.uk/item/vqzy3/diversity-management-that-works-an-evidence-based-view

Goransson (2025) found that different instructional methods (a mix of experiential exercises with lectures) allowed individuals to engage with the message through multiple channels, which might strengthen the effectiveness, given that people experience things differently, thus increasing the likelihood of prejudice reduction in heterogeneous workplace samples. a duration at least three times longer, delivered across two separate sessions, and conducted in-person rather than online. https://www.tandfonline.com/doi/full/10.1080/01419870.2025.2478268#abstract

That is similar to the findings of Bezrukova (2018) op cit that the length of the training program was related to effect sizes, suggesting that multiple training sessions are beneficial for lasting effects.

Noon (2017), in a critical review, suggests the risk of “a quick fix rather than the start of an ongoing and possibly lengthy process of reflection, discussion and awareness raising in keeping with cooperative learning approaches”. He suggests that

“even if ‘unconscious bias training’ has the theoretical potential to change behaviour, it will depend on the type of racism: symbolic/modern/colour-blind, aversive or blatant. In addition, even if those deemed racist are motivated to change behaviour, structural constraints can militate against pro-diversity actions. Agency is overstated by psychology-inspired ‘unconscious bias training’ proponents”.  https://journals.sagepub.com/doi/abs/10.1177/0950017017719841 

Summary

In summary, diversity training, may have a positive impact on cognitive awareness under certain circumstances. However:

Many interventions were shown to have little impact if at all

• Those that did have some impact found it was generally short-lived

• Diversity training was even less effective for overt bias as opposed to implicit bias

• Diversity training has a serious risk of being counter-productive especially if it is imposed and made mandatory

• The most effective training involves sustained training over a period of time

• In person training was stressed as essential to embed learning from diversity training – preferably within existing teams

• Diversity training is very unlikely to have any sustained impact unless combined with a range of other, evidence-based interventions which emphasise accountability and sustained engagement with managers and staff.

• Forms of training that are not one one-off, are not primarily online and are not mandatory, may have more likelihood of success but crucially there must be an opportunity and willingness for participants to raise difficult issues, reflect deeply on their own prejudices and there must be skilled tutors providing challenge to embedded ideas and to signpost ways forward.

 However, the learning and personal development such training can enable is only likely to be effective if integrated with changes to employment relations that debias processes (and insert transparency and accountability into decision making.  Such interventions do have a solid evidence base as long as they are proposed and implemented by leaders who are prepared to challenge themselves, and be challenged, about their prejudices and the privilege they have previously taken for granted. See, for example, Kahneman 2021 https://www.scirp.org/reference/referencespapers?referenceid=3432662; Bohnet 2018 https://www.hup.harvard.edu/file/feeds/PDF/9780674986565_sample.pdf

These are not abstract issues. The NHS has been told it must engage in precisely the online mandatory “sheep dip” approach towards diversity training that research s suggests is ineffective and potentially counter-productive.

Moreover, this approach risks being seen by some employers as an alternative to precisely the evidence based sustained work that is needed and which has been so conspicuously absent from much NHS work to tackle racism or antisemitism (or Islamophobia).