What works to reduce prejudice and discrimination? - A review of the evidence

Executive Summary

Objectives

• This is a review of the international literature and evidence on what measures are considered to be most effective in tackling prejudice. It provides an overview of the key theoretical debates on prejudice reduction, outlines some of the most common interventions, and suggests some lessons that may influence the design of policy interventions in the future.
• The report focuses on high quality empirical studies, a mixture of interventions that were evaluated in 'real world' settings also lab-based psychological experiments.
• The report also reflects on the appropriateness and applicability of such interventions for tackling different types of prejudice in Scotland. This includes exploring some of the key lessons in relation to anti-sectarianism work.

What is prejudice?

• The spectrum of prejudice ranges from instances of overt discrimination or hate crime, to much more subtle 'everyday' expressions, such as so-called 'banter', or ignoring or excluding certain people or groups, even unwittingly. This report sought to explore measures which tackle these more subtle forms.
• Prejudice should be viewed as a process in a set of relationships between people. The report therefore focuses on intergroup relations rather than the apparently prejudiced characteristics of individuals, moving away from an individual pathological approach towards seeing prejudice as a social problem which requires social change.
• Most people do not consider themselves to be 'bigoted', and may be unaware that their attitudes or behaviours could be deemed prejudiced or discriminatory. As such, it may be important to begin from an assumption that people will recognise themselves as prejudiced. Even people who consider themselves tolerant and are consciously non-prejudiced can have 'implicit bias' which is instinctive and often activated without being recognised.
• Given this uncertainty, interventions aimed at tackling prejudice should be based on an approach which is flexible, allows different perspectives to be debated, and does not blame or castigate. There should be sensitivity about the terminology used. Perhaps drawing people's attention to the concept of implicit bias - in a sensitive and blameless way, as opposed to making a moral judgment - might be a useful way to raise awareness and provoke honest discussion.

Theories on prejudice reduction

• Theories of prejudice reduction can be roughly divided into two camps: theories of intergroup contact, whereby association with other groups reduces negative attitudes and promote inclusivity; and theories which focus on how exposure to information about other groups can challenge the way people think about them - (sometimes referred to as 'education' interventions).
• Lab-based and field studies have continually confirmed the effectiveness of contact, highlighting its ability to challenge prejudice by reducing intergroup anxiety and increasing empathy for other groups (the two underlying mechanisms).
• However, highly prejudiced people are more likely to deliberately avoid intergroup contact, so thinking about how to promote opportunities for contact and remove barriers for those less likely to seek it is vital. This is why interventions are often deployed, usually based on some form of focused education.

Types of prejudice-reduction interventions

• There are a mixture of lab-based interventions and evaluations of prejudice-reduction initiatives 'in the field', however the majority of studies are controlled and experimental, take place in psychology laboratories, and often with psychology students as participants. Fewer studies take place in 'real-life', in schools or communities for example.
• The most frequently studied interventions tend to fall into one of three categories: mid-long term educational strategies (including but not limited to school-based interventions); short-term diversity training courses; and media campaigns.
• It is important to emphasise that the evidence on 'what works' is very limited, but there is still value in drawing upon some of the most promising messages and themes that might help to inform and improve future initiatives.

Key lessons - General

Aim for a broad commitment to reducing prejudice, not one-shot interventions

• One objective of this review was to determine whether specific anti-prejudice initiatives (e.g. anti-sectarian, anti-racist, anti-homophobic etc.) would work better than a 'catch-all' focus on prejudice. Abrams (2010) states that this has not yet been subject to sufficient testing, but since prejudices towards different groups appear to have different developmental trajectories, there is value in treating these as different problems, with potentially different causes and solutions that require reflection and sensitive intervention designs.
• However, to avoid 'prioritising' certain types of prejudice over others (perhaps as a result of political/media pressure), there could be a broad prejudice-reduction framework with flexibility which allows for a focus on specific forms where necessary and regional and local sensitivity.
• Evidence suggests that one-off activities make less impact; better results come from sustained activities over a period of time. Some short-term projects may still be effective, however these should be part of a wider framework that emphasises long-term education and opportunities for long-term contact with the potential for cross-group friendships.
• Furthermore, interventions should take place within a broader context of commitment to diversity in terms of institutional and cultural change. For example, there is perhaps less of a priority for organisations holding diversity training courses when groups such as women, ethnic minorities, or people with disabilities, are significantly underrepresented in senior positions within their workforce.

Certain techniques more effective than others

• Interventions should be based on social-psychological theories and key lessons from the literature: those that are not rooted in these tend to be less effective. Techniques based on an overly simplistic or idealised notion of what 'should' work can be counterproductive, especially if not applied with care.
• For example, despite good intentions, direct attempts at persuading people to recognise and change their attitudes have been known to be ineffective and often have unintentional negative effects. Diversity training in particular risks 'backfiring' by reinforcing minority ethnic stereotypes, essentialising group categories, and drawing attention to difference and inequality.
• In contrast, certain techniques are known to be much more successful with less risk of negative impacts. Interventions which facilitate positive intergroup contact, or are based on principles of perspective-taking or empathy-induction are considered to be effective. In education, cooperative learning and the use of curriculum which embeds positive messages of intergroup contact are also promising.

Handle issues sensitively

• Rather than 'instructing' what types of behaviours, language, or attitudes are 'wrong' - something that is often subjective and contested - teaching skills and disposition, such as critical thinking and empathy, is likely to be more effective.
• Acknowledging and discussing historical events may be helpful in terms of breaking down existing barriers and challenging the residual prejudice apparently stemming from historical conflict and poor relations.
• The 'backlash' effects discussed above emphasise the importance of knowing the area and of carefully designing programmes and initiatives. The literature strongly supports the principle of peer engagement, suggesting that change is best affected from within peer groups where possible. This could involve participants who previously took part in programmes helping to shape and facilitate future initiatives. The 'credibility of the messenger' is highly important.

Monitoring and evaluation central to success

• Interventions should have a clear strategy, and should be carefully monitored throughout. Evaluation should be considered at design stage, not as an afterthought. Need to go beyond self-reporting as it is vital not to confuse 'feedback' with evidence of impact.
• A 'What works' approach should be broken down into: for whom, in what context, for how long, and in what way. 'What doesn't work' is equally important - we should not be afraid to document what failed, and why.

Key lessons - Scotland

• A clear caveat is that we need to be wary about what conclusions can be drawn from 'what works' in more problematic settings that will be relevant or appropriate to Scotland. Many of the key prejudice-reduction interventions have taken place in areas in which ethnic or other prejudice results in or is exacerbated by overt conflict, or at least has done in recent times. This has not been the case in Scotland, so although these may involve useful strategies that could help to influence prejudice-reduction initiatives more broadly, direct application may not be appropriate.
• High levels of contact does not necessarily remove the existence of stereotyping or discrimination: the example of gender relations highlights this. In terms of Catholic/Protestant relations, people are very integrated in terms of families and other relationships. However, the research shows that perceptions of continued existence of sectarianism in Scotland are extremely high.
• Sectarianism therefore is either a problem of perception, or more complex than simply tensions between two separate groups. It may be the case that 'contact works' to an extent, but there 'residual' problems which are said to still exist - perhaps a result of historical 'grudges' and myths about our own histories and the histories of others. If the latter is the case, these might be tackled better through specific education or re-education.
• People have to be willing to confront and challenge their prejudices: it is very difficult to compel someone to change against their will. Yet we have established that most people do not recognise themselves as prejudiced, and this is particularly complex given that the attitudinal research on sectarianism shows that people tended to think of sectarianism as happening 'elsewhere' - not in their local area/community and not themselves personally. There may therefore be a lack of willingness among people to engage.
• Moreover, this complexity is compounded by the lack of consensus among people regarding what sectarianism actually is, again highlighted in the qualitative research. Sectarianism is a contested term, and some groups feel that what is sometimes deemed sectarian behaviour - for example, certain songs or language used - are legitimate expressions of culture or identity. This tension may be addressed in part by rooting initiatives in people's own experiences and understandings, rather than 'instructional' approaches which impose value judgements.
• Finally, it is important to be mindful of not facilitating the reproduction of particular assumptions or stereotypes. Dramatic interpretations of the issue - for instance, 'hard-hitting' messages or media clips - could risk alienating sections of the audience who would not recognise overt violence as a feature of their lives or communities.