Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching1

The article summarises and elaborates a norm-critical and intersectional study of the legal education in Gothenburg, and reflects on the broader pedagogical project of discussing the study among the faculty. A norm-critical perspective seeks to critically examine what is taken for granted and perceived as normal, and subsequently what is perceived as exceptions. An intersectional perspective acknowledges power structures as complex social structures with diverse origins. Together they provide a useful tool for analysing how different social positions are represented in the assignments of the legal education. The study showed, among other things, that men, Nordic people, adults and people from middle and upper class were overrepresented compared to the Swedish society as a whole. The study resulted in important discussions at the Department of Law in Gothenburg on whether representation matters, and if so what should be done. The article seeks to present the study itself and the theoretical thinking behind the study, and to discuss possible ways of dealing with its results.


Situating the study Background and aims of the study and this article
In 2015, members of the student association Feministiska Juriststudenter (FEMJUR) [Feminist Law Students] knocked on the door of the Vice Head of the Department of Law at the University of Gothenburg (hereinafter the Department). They showed him an investigation they had done concerning the representation of different categories of people in the assignments used in three courses in the LL.M. Programme. The investigation showed a strong overrepresentation of men compared to women. It also showed that almost all people had names that sounded Nordic and that all families represented appeared to be heterosexual (FEMJUR Göteborg 2015). The management of the department became concerned about this narrow representation and decided to conduct a study to figure out if the situation was similar in the other courses in the LL.M. Programme. The study (image 1) was conducted by the author of this article in cooperation with project assistants Jessica Thornström and Ylva Lindh, both at that time law students. Image 1. Cover of the report of the study (Mägi 2018).
The LL.M. Programme at the University of Gothenburg, as in many other places, has a pedagogical tradition of letting the students work with a large number of assignments consisting of a brief story connected to a range of legal questions. The aim of the study was to get a comprehensive picture of

A brief introduction to intersectionality and norm-critical theory
The norm-critical study of the LL.M. Programme used intersectionality and norm-critical theory to analyse the results in relation to social norms and power structures. The American lawyer and professor Kimberlé Crenshaw introduced the concept of intersectionality in 1989 while trying to explain how discrimination based on race and sex often happens simultaneously and that they are interconnected (Crenshaw 1989). The intersectional theory says that it is not possible to fully understand, for example, gender structures alone, but only in relation to other structures (De los Reyes and Mulinari 2005; Lundberg and Werner 2016). There has been critique of how intersectionality has been expressed. Some models have stacked identities on top of each other in trying to get the full picture -gender, class, ethnicity, sexuality, age -but have still needed to end with an "etcetera clause", which has been called "Oppression Olympics" (Wennberg and Pylkkänen 2012). Intersectionality theory later developed into an understanding of power structures as complex social structures (De los Reyes and Mulinari 2005).
The concept normkritik [norm-critical] was introduced in Sweden in 2010 in relation to pedagogy. Norm-critical pedagogy has emerged, together with other critical pedagogical perspectives, as a reaction to pedagogy that attempts to include what is considered different but only by tolerating it. This kind of pedagogy might include a broader representation of people to avoid discrediting certain social groups. At the same time, tolerating marginalised social positions as "the other" still maintains the norm as a position that is taken for granted (Bromseth and Darj 2010;Kalonaityté 2014).
Norm-critical studies have their roots in intersectional analysis and other critical perspectives, but are not limited to any particular power structures.
Instead, such studies focuses on social norms and power structures in general. Norm-critical studies also mark a transition from attention on only marginalised groups, such as homosexual and transgender people, to also examine privileged groups, such as heterosexual and cisgender people. 4 Moreover, norm-critical studies take a further step by studying the very mechanisms that (re)produce social norms, such as divisions between people as well as hierarchies between aspects perceived as different. This marks a transition from attention on social groups to attention on social positions shaped by social norms. Fundamentally, this is done by critically examining and exposing what is taken for granted and perceived as normal or neutral, and subsequently what is perceived as exceptional to such norms (Bromseth and Darj 2010). Image 2. "Let's Talk About Race" Part 1. Photo: Chris Buck (chrisbuck.com) The cover picture of the report illustrates the norm-critical perspective (image 2). A light-skinned girl is standing in front of a wall of dark-skinned dolls. By showing a contrast to a common social reality, the picture highlights which children more often find toys that represent themselves, their families and their people. Toy stores often have more light-skinned dolls and have to make an active choice to display dark-skinned dolls. This is an example of how light skin is taken for granted, while dark skin is perceived as an exception. The mechanisms behind social norms that reproduce differences between people based on skin colour are complex and have a long history in colonialism, slavery and discrimination. This will be elaborated on in the discussions below. Another, simpler, example of norms and exceptions is how to distinguish between vegetables with or without pesticides. To find vegetables cultivated without pesticides in a grocery store, one has to look for certain organic markings. If there is no marking, we can assume that pesticides have been used. Thus, cultivation with pesticides appears as the norm and organic cultivation appears as an exception to this norm. The norm of cultivation with pesticides is only apparent in light of the exception of organic cultivation. As norms are taken for granted, they are often invisible until someone challenges the norm with an exception.
These examples also shows how norms are not necessarily good, and not necessarily bad. Moreover, social norms are neither necessarily intentional nor do they even have to relate to what is most common. Social norms are cognitive structures needed to understand the world. However, some of them are related to power structures of inequality and exclusion.
A norm-critical perspective can analyse several power structures at the same time. By studying what appears as normal and what appears as an exception to such norms, we can analyse how a phenomenon relates to many aspects. Thus, it is not necessary to decide in advance what structures are to be studied. This means that the results may be unexpected, unveiling norms that the researcher did not have in mind in beforehand. While it is also possible to use norm-critical methods based on a specific theoretical understanding, there has been criticism that some norm-critical studies projects are too closely linked to political ideologies. So far, explicit normcritical methods have mostly been used in practical change work and only recently in academic work. Most critical perspectives problematizing power structures are criticised for not being neutral or objective, meaning that they do not describe a reality independent of human impact. From another perspective of theory of knowledge, studying a problem from different viewpoints, being transparent and describing one's premises, assumptions and methods lead to neutrality and objectivity (Svensson 2018).
As mentioned above, the study analysed the material on the basis of eight perspectives: gender, gender identity or gender expression, ethnicity, religion or other belief, ability, sexual orientation, age and class/socioeconomic status. Deliberately, the study has not engaged in developing complex theories on all of these aspects. Neither has it engaged in discussions about the relationship between social norms and law, but rather the relationship between social norms and pedagogy in legal education. The purpose of the study was to communicate to faculty colleagues how our assignments relate to diversity and equality. Therefore, the study has applied only an elementary theoretical framework that there are social norms and power structures related to these eight aspects. This means that, on a structural level, some positions related to these aspects have, or have had, the function of being norms in society and holders of privileges. At the same time, other positions have, or have had, the function of being exceptions to such norms and been subject to disadvantages. The study did not aim to describe or determine what the social norms and power structures are in Swedish society as a whole. Instead, the study analysed the assignments to see if there were any patterns in relation to our theoretical framework. According to our framework, for example, men are structurally more normative and have superior positions in society compared to women, acknowledging that this is a simplification and that there are individual deviations from this structure. As elaborated on below, this basic theoretical framework has been sufficient to obtain a meaningful analysis of the results of the study.

Summary of the study: method, findings and conclusion An overview of previous gender and norm-critical studies of and projects in legal education in Sweden
The Department of Law at the University of Gothenburg was not the first in Sweden to undertake a study of its LL.M. Programme from a gender and/or norm-critical perspective. As early as 1993, Anita Dahlberg (1993) wrote an article illustrating how assignments and exam questions used in the LL.M. Programme at Stockholm University presupposed male experiences and relationships, including some explicitly sexist content. Since then, several studies at different universities have examined their legal education, mostly focusing on literature but also assignments, from either a gender perspective or a more general norm-critical perspective. Studies of representation among authors of the literature showed that law students read textbooks mainly authored by men that also refer to other men as legal authorities (Stridsberg and Westerstrand 1999;Bredberg et al. 2011). 5 Other studies have concerned the content of the literature and the assignments. One study at Uppsala University showed that women are referred to by their gender while men are referred to as human beings with no gender, moreover that power relationships between men and women are never discussed, and finally that legal areas with high relevance for women are hardly discussed at all (Stridsberg and Westerstrand 1999). A later study at Uppsala University showed that the literature is based on gender stereotypes that present society as more unequal than reality, and at the same time overlook the inequalities that exist, making the literature legally deficient (Eriksson and Burman 2007). A study at Lund University showed that the literature is inadequate in terms of critical discussions of law in general, that the gender perspective is almost invisible, and that the language and examples in the literature lead to outdated stereotypes (Bredberg et al. 2011). One further study at Uppsala University showed that the literature demonstrates a binary view on gender, refers only to countries in the Western world, and a number of grounds of discrimination are not mentioned at all (Ekström et al. 2015). A recent study at Stockholm University (conducted after the study described in this article) showed that representation in the assignments differed greatly from the actual population, with the socioeconomic group workers being less represented for example. There were also some gender stereotypical assignments (Welsapar, Johansson and Klamberg 2017).
Further projects reflected on the processes of introducing the gender perspective in legal education. One project at the University of Gothenburg reported on the frustration among teachers, difficulties in disentangling the issues of gender in education and gender equality in the workplace, the need for explicit support from management, and the long-term and ongoing work to achieve change in legal education (Eriksson-Zetterqvist and Svensson 2001). Another project at Umeå University reported on fear of the gender perspective as a threat to academic freedom and of the feminising (with negative connotations) of legal education. A number of teachers did not participate, but several teachers also showed greater interest in 5 See also an inventory of literature and material with a gender perspective on six law programmes in Sweden (Gutman and Gustafsson 2011).

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching participating in discussions in a field of knowledge that was new to them (Nordborg, Ågren and Burman 2002). An article reflecting on both projects mentioned above concluded that Sweden´s Higher Education Act (Chapter 1 Section 8) demands utility-related attitudes as well as critical and reflective attitudes (Burman, Svensson and Ågren 2003). Another project at Lund University reported problems with the implementation of the gender perspective due to the teacher's lack of knowledge, with the gender perspective being played down, a lack of time and no formal coordination between teachers (Persson et al. 2014).
In summary, all the studies and projects conducted have concluded that the legal education that students receive at Swedish universities does not represent women and men equally, nor other social positions, and that it is not necessarily easy to raise these issues at the departments and faculties. At the same time, all the aforementioned studies and projects have focused either solely on gender or mainly on the literature used in legal education. The study conducted at the University of Gothenburg was the first to include a broad norm-critical study of assignments.

Method
The norm-critical study of the LL.M. Programme discussed in this article used quantitative methods to examine representation and description of different social positions. The material chosen was a selection of the assignments used in legal education, as they are products of the Department and within the hands of the Department to change more easily than literature for example. The study was limited to the assignments given in the first three mandatory years of the LL.M. Programme at the University of Gothenburg, as all students encounter these assignments. 6 From this material, the study made a selection of about one third of all assignments organized in an archive. The study coded all the characters who feature in the selected assignments in relation to the eight variables based on their connections to different social positions (gender, gender identity or gender expression, ethnicity, religion or other belief, ability, sexual orientation, age and class/socioeconomic status). The study also coded each character in the assignments in relation to five variables according to how they were described (activity, dependency, social sphere, role, and negativity/positivity). The coding of this material was than compiled to results for every course and aggregated for the whole LL.M. Programme. Finally, the results were analysed in comparison to the population in Sweden and in relation to our theoretical framework of normative and privileged positions as well as marginalised and disadvantaged positions. 7 The first part of the study concerned the extent to which the assignments represented a diversity of social positions. The project received some comments in this part, such as "Isn't it a contradiction that a norm-critical study is so excessively focused on counting representation?" and "Isn't there a risk of confirmation bias in assuming that certain representations are Nordic, heterosexual and so forth when this is not explicitly stated in the assignment?" The study admits that there are challenges in categorising people in different social positions, as this process simplifies reality, and subsequently risks assigning and reinforcing identities not necessarily shared by the people involved. However, in this study the assignments analysed were only fictional stories, and not about real people. The study also only analysed how a character appears to the reader. The study has neither tried to determine anything about what the person "really" is, nor tried to capture the intention of the author. Furthermore, the authors of the study continuously reflected on our methods in an ambition to conduct the coding as objectively as possible.
The first part of the study focused on identifying and coding whether a character depicted explicitly belonged to any of the eight social positions studied, such as a "citizen of the Democratic Republic of Kongo", meaning a character of non-Nordic origin. Secondly, the study analysed how a character depicted appears to the reader by whether the character's name sounds male or female, Nordic or non-Nordic, etc. What the study looked at is how the actual text signals different social positions in the specific context of a Swedish LL.M. Programme. As this context is normative, such norms inevitably affect the coding. For example, the name "Maria" has been coded as Nordic, although the name is also common in many other regions of the world. The author may have thought of "Maria" as non-Nordic, and there are many ways to interpret "Maria" as not necessarily Nordic by problematizing the underlying meaning of the text. At the same time, a character named "Maria" appears to be Nordic to most Swedish readers if there is no other information in the text that makes her appear to be non-Nordic, such as a family name or other information about her heritage. To deal with this normative effect on the coding of the material, the study has explicitly made the coding generous, in favour of non-normative positions. This means that for example the name "Tanja" was coded as non-Nordic, as it sounds Russian to many Swedish readers, even though it is also a fairly common name in Sweden. In this case, the name "Tanja" in itself bears information that makes the character appear to be at least potentially non-Nordic. Through these dual measures, the study is assessed as having achieved a nuanced coding of what students encounter in reading and working on the assignments. In many cases, a character's gender, ethnicity, sexual orientation and so on were not apparent at all.
The coding of class/socioeconomic status was conducted in two steps. Firstly, the study noted occupation and/or education as working class, middle class, or upper class according to a simplification of Statistic Sweden´s socioeconomic classification (SCB 1982). Secondly, it noted the value of the assets owned by the character in the story, their wealth, as less than average (up to SEK 200,000), average (SEK 200,000 -5,000,000) or higher than average (SEK 5,000,000). Thirdly, it weighted these two aspects of occupation/education and wealth together as lower, middle or higher class/socioeconomic status, deliberately using own terms specifically for this study.
The second part of the study examined how the assignments described people of different social positions with regard to how active they appear to be, how independent, what social sphere they appear to act in (public or private), what role they play in the story and how negative or positive they appear to be. In the first step, the study coded all adjectives, verbs and adverbs connected to every character. The second step, of the study analysed activity, independence, etc., based on these words together with an overall assessment of the context as a whole. Again, the study has not tried to determine how a person "really" is, nor tried to capture any implicit intentions of the author, but merely coded how the character is described as they appear to the reader.
To give an example of the coding process, below is one assignment in Family Law.
Hulda and Malte have begun to disagree about just about everything and mostly about financial issues. The reason is that Malte has started gambling in Åby [a racecourse for horses] and has not been particularly successful. Currently, their combined wealth is as follows: Hulda has approximately SEK 500,000 net in marital property, while Malte has approximately SEK 300,000 net, also in marital property. In just the past six months, Malte has gambled away about SEK 100,000, and if he continues at this rate, Hulda is afraid of losing a large part of her own marital property in a possible future property settlement -which, under the circumstances, does not seem entirely unrealistic or distant. Is Hulda able to request property division during marriage, request that the spouses enter into a prenuptial agreement or do anything else to take care of herself? (Archive 2.5,author's translation.) Regarding representation, the study coded Malte as a man and Hulda as a woman. Their names are also a word game in Swedish, Malte referring to man [husband] and Hulda to hustru [wife]. Both were coded as appearing to be Nordic (Swedish-sounding names), heterosexual (married, different sex couple), adult or elderly (18-64 or 65+) and with average wealth (assets worth SEK 200,000-5,000,000). It was also noted that their gender identity/expression, religion/belief and ability do not appear. Regarding the description, Malte is described by the verbs "have begun to disagree", "has started gambling", "has gambled away" and the adverbs "not particularly successful". Hulda is described by the verbs "have begun to disagree" and the adjective "afraid of losing marital property". There are also questions about her legal options directed to the reader as a law student. Based on these descriptions, the study has coded Malte as active and independent (gambling -not in relation to Hulda) while Hulda is coded as passive and reactive (being afraid -in relation to Malte's gambling). One could argue that Hulda is active and independent -actually being angry and planning a divorce to avoid losing her property. However, in the text there is no explicit basis for this interpretation. A future divorce is indeed described as realistic and there are questions on how to plan for a divorce under such circumstances. Nevertheless, such plans are not explicitly connected to

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching Hulda as an active agent, but instead as an object. Therefore, the text itself describes her as passive. Furthermore, the study also coded Malte as active in both the private sphere (disagreement with his wife) and public sphere (the racecourse) while Hulda is active in only the private sphere (disagreement with her husband). Additionally, it coded both to be in leading roles and not supporting roles, and finally Malte is described in negative terms (risking his own and his wife's economy on unsuccessful gambling) and Hulda in neutral terms (being afraid for her economy). This is one example of how the study has coded the assignments. In this way, the study coded representation of gender, ethnicity, sexual orientation, age and class/socioeconomic status, as well as coding the descriptions of activity, independence, social sphere, roles and negativity/positivity in connection with these social positions. A similar coding process was applied to all the studied assignments. The part on representation examined 388 assignments, coded 1,189 characters and analysed them on the basis of eight variables. The part on description examined 225 assignments, coded 784 characters and analysed them on the basis of five variables. This resulted in a total of 13,432 coded data points. This data was analysed at an aggregated level in order to study any patterns related to social norms and power structures in the assignments as a whole.

Over-representation of normative positions
The first set of results of the norm-critical study of the LL.M. Programme concerned the representation of people in different social positions. This differed a lot between the different courses of the LL.M. Programme. However, the most interesting results are results with statistical significance at an aggregated level that includes all of the courses. Initially, the study compared representation of gender, ethnicity, age and class/socioeconomic status in the assignments with the population in Sweden according to Statistics Sweden (SCB).
Starting with gender, the assignments had 63% men and 37% women, while the population in Sweden has 50.06% men and 49.94% women (SCB 2017a). Thus, men are clearly overrepresented in the assignments compared to the population (Figure 1). Moreover, gender was not apparent for 25% of the characters in the assignment stories. According to FEMJUR, teachers as well as students often refer to non-gendered characters as "he" in pedagogical practice. In that case, men would appear to be even more overrepresented. Concerning ethnicity, 9% appear to be non-Nordic people and 91% Nordic people in the assignments. In comparison, 22% of the population in Sweden have a foreign background, meaning that they are born abroad or both parents were born abroad, and 78% do not have a foreign background (SCB 2002(SCB , 2017b. Thus, Nordic people are clearly overrepresented in the assignments ( Figure 2). Moreover, the definition of non-Nordic people in the study is also narrower than Statistics Sweden's definition of foreign background, making the difference actually even greater. It was also obvious that non-Nordic people who figured in the assignments, more often than other social positions, included references to known people (fictional or real) such as George R. R. Martin (American writer), Gustav Mahler (Austrian composer), Zlatan (Swedish soccer player of Balkans origins) and Professor Balthazar (figure in Yugoslavian cartoon TV series from the 1970s).

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching 15 Figure 1. Representation concerning ethnicity.
Looking at age, the assignments had 5% explicitly young people (0-17 year olds), 92% explicitly adult people (18-64 year olds) and 3% explicitly elderly people (65+ year olds). This can be compared to the population in Sweden where 21% are young, 59% are adults and 20% are elderly (SCB 2017c). Adults are overrepresented in the assignments, while young and elderly people are underrepresented ( Figure 3). Our study coded the young and elderly who were explicitly under 18 or over 65 years. The ambiguous cases were coded as adults, which makes the differences compared to the population smaller, but still significant. When it comes to class/socioeconomic status, the assignments had 20% people of lower, 67% of middle and 13% of higher class/socioeconomic status. At the time, the study was not able to find statistics about the population in Sweden that were compatible enough with our coding. A new study however shows, with slightly different definitions than our study, that the working class in Sweden is 38-46% of the population, the middle layer 46-54% and the self-employed and bourgeoisie 9% of the population (Theborn 2018). As an approximate comparison, there were fewer characters in the lower category in the assignments than working class people in the population, with more middle category characters in the assignments than people in the middle layer. This means characters from the lower class/socioeconomic status are probably underrepresented in the assignments, while the middle category is overrepresented (Figure 4). In summary, the results showed, with statistical significance, that men, Nordic people and adults are overrepresented in the assignments compared to the population in Sweden. Similarly, women, non-Nordic people, the young and the elderly, are underrepresented. Class/socioeconomic status seems to follow the same pattern. Interestingly, all of the overrepresented social positions have structurally normative and/or superior positions in relation to our theoretical framework, while the underrepresented social positions are structurally non-normative and/or subordinate. The study argues that the representation in the assignments thus contributes to maintaining and reproducing such social norms.

Non-normative positions marginalized
Another interesting set of results concerning representation was that a number of social positions occurred only very marginally. These were Erik Mägi Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching transgender people, religious people and people with disabilities. This also applied to a number of ethnic groups. By transgender people, the study means people who have a gender identity or gender expression that transcends the traditional divisions of man and woman. By cisgender people, the study means people whose gender identity and expression correspond to their sex assigned at birth. In the material, there was only one character with gender-crossing expression, making representation of this social position 0.1%. That assignment concerned Labour Law and discrimination of a woman performing in a drag show parodying famous boy bands. 8 There were no characters with gendercrossing identity, nor with explicit cisgender identity. For the vast majority of 99.9%, gender identity and expression was not apparent ( Figure 5). There was also only one assignment including characters with explicit religious beliefs, which also means 0.1% representation. The assignment concerned Real Property Law and the board of a co-operative housing association being "conservative with religious values" and thereby sceptical about a same-sex couple becoming members of the association. 9 There were no explicitly non-religious or secular characters. Also in this case, religion or other belief was not apparent for the vast majority of 99.9% of the characters in the assignments ( Figure 6). 8 Drag king / drag queen is a person who uses social attributes and symbols to push the boundaries of male and female. Often for the purpose of entertaining or playful expression of a part of their identity (RFSL 2017). 9 There was also an assignment concerning a religious legal entity, a Christian community "New Fundamentalists" having their own translation of the Bible. The study, however, only analyzed human beings and not legal entities. There were fourteen characters in the material with a physical or mental disability, making the representation 1.5%. Our interpretation of disability was extensive including for example disc herniation, "sperm of bad quality" and unspecified special needs and illnesses. There was no character with a cognitive disability, and also no character explicitly fully able-bodied. In most cases, 98.5%, ability was not apparent (Figure 7). Moving on to ethnicity, the non-Nordic characters in the material were a very heterogeneous group. To get a more nuanced picture, the study also analysed the representation of some ethnic sub-groups. Half of the non-Nordic characters had names signalling a background in the Western World. The study compared this to the representation of people from outside the Western world. As the material did not show a background from all parts of the world, outside the Western world in this material primarily meant Eastern Europe, Africa and Asia while the Western world primarily meant North America and Western Europe (including the Nordic countries). There were 27 characters appearing to have a background in countries outside the Western world. This means 4% compared to 96% of the characters appearing to have a background in the Western world, including Nordic people (Figure 8). The study also analysed the representation of characters from regions with people typically targeted by racism. This category of people might also be referred to as racialized people, non-white people or people from the global majority. In this study, people targeted by racism refers to people appearing to originate from Africa and Asia, which are regions with people typically targeted by ethnic stereotypes (there were no characters in the material appearing to have origins from non-European descendants in other regions of the world). In this study, "white" people refers to people able to pass norms of whiteness by appearing to originate from Europe or European descendants. The study writes "white" in quotation marks to indicate that those people are not actually white and that whiteness is a social construct related to several social and biological markers. 10 There were 15 characters appearing to have a background in regions with people typically targeted by racism and ethnic stereotypes. This means 2% of the characters appearing in the stories were people typically targeted by racism and 98% appeared to be "white" people ( Figure 9). The study also analysed the representation of national minorities among the group of Nordic people in the assignments. The officially recognized national minorities in Sweden are Sami, Roma, Jews, Sweden Finns and Tornedalers. 11 In the material, there were three characters with Finnishsounding names and no other indications of characters with connections to any of the national minorities. This means 0.5% of the characters appearing to have a background in one of the national minorities compared to 99.5% appearing as Majority Swedes (Figure 10). Finally, looking at sexual orientation, there were 14 characters appearing to be homosexual or bisexual (because of a same-sex relationship). This can be compared to 139 characters appearing to be heterosexual (because of a different-sex relationship) and 791 characters with no indication of sexual 11 Sweden´s Act on National Minorities and Minority Languages (2009:724).

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching orientation. This means 9% were represented as explicitly homo/bisexual compared to 91% as explicitly heterosexual (Figure 11). The representation of homo/bisexual characters in the assignments is thereby approximately the same as in the population in Sweden. This is the only social position with a balanced representation in the material. This is probably due to intentional inclusion by the authors of the assignments. Figure 11. Representation concerning sexual orientation.
To summarize, in the assignments there were a number of social positions that were highly marginalised with some of these hardly appearing at all (despite the generous interpretation applied to non-normative social positions). This means that the Law students very seldom encounter assignments that include characters from these positions during their compulsory courses in their legal education. All of the marginalised positions are structurally non-normative or subordinate according to our theoretical framework. In several cases, the structurally normative positions such as cisgender people, non-religious or secular people and able-bodied people were not apparent either. In this way, the very norms of gender identity, religion/belief and ability are invisible, giving the students hardly any opening to problematize them. The study thus argues that the assignments contribute to rendering a number of non-normative social positions invisible. This can be compared to the representation of sexual orientation, where a comparable proportion of homosexual and bisexual people were represented as in the general population in Sweden, thereby making the norm of heterosexuality visible too.

Intersectional enforcement of unbalanced representation
The study also found some results with statistical significance on representation concerning intersections of different social positions. Looking at ethnicity, in the group of non-Nordic people there were 32% women and 68% men, compared the group of Nordic people that had 37% women and 63% men ( Figure 12). This means the underrepresentation of women is higher in the non-Nordic group than in the Nordic group. This shows how gender and ethnicity intersect with each other. In other words, there are fewer non-Nordic women, which mean the underrepresentation is intersectionally reinforced by the combination of gender and ethnicity. There are also more Nordic men, which means an intersectionally reinforced overrepresentation. Similar patterns could be seen concerning the intersectional underrepresentation of women in the higher class/socioeconomic status. Other intersectional underrepresentation were not ensured with statistical significance but are nonetheless interesting because they show how some particular intersections of social positions are invisible in the assignments: • No examples of homo-or bisexual people of non-Nordic origin.
• No examples of disability in middle or higher class/socioeconomic status. • Few young and no elderly people of non-Nordic origin.

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching

Diverse individual descriptions
The second part of the norm-critical study of the LL.M. Programme concerned how the assignments described different social positions. Our main aim was to gather information about the descriptions at an aggregated level. At the same time, the study's authors also kept their eyes open for inappropriate individual assignments. Actually, the study could not find any clear examples of open sexism, racism or other offensive expressions in the analysed material. However, a number of individual assignments did reproduce some stereotypes. For example, there are several examples of older and richer men being married to younger women. More than a few assignments also describe characters from lower class acting in an unreasonable way, such as one example of an electrician being described with the adjective "raging" and the verbs "trespasses", "had taken pills", "had been drinking beer", "drives the truck straight through the gates" and "kicks the policeman" (Archive 2.7, author's translation). There were also a number of examples of young characters being discredited. One example is Max, a boy of 11 years old, being described with the phrases "a real rascal", "in a short period of time… had broken two windows, started a number of fights and lit a fire in a trash can" and "puts a rolled newspaper soaked in gasoline under Märta's car and then lights it". And finally, "afterwards Max explains that he and his dad do not like Mercedes, but prefer Volvos, and that he burned up the car to avoid seeing it" (Archive 2.9, author's translation). Such assignments might have some entertainment value, often well needed in legal education. At the same time, they feed prejudices against working class and young people as having negative attitudes and behaviour. Some assignments also clearly go against stereotypes, for example a male preschool teacher or a single dad taking full responsibility for his child. There are also several female bosses or entrepreneurs described as active and independent, for example with phrases such as "runs a business", "starts", "moves on", "teaches [a male college]" and "confronts" (Archive 2.14, author's translation). Sometimes women are in relationships with a man who is younger or from a lower class. In several of the assignments, there are also non-stereotypical representations such as same-sex relationships, without that aspect being significant for the legal issues.
In total, the students need to work with assignments that both reproduce and counteract stereotypes. We argue that this combination may help the students to reflect on social norms. This of course is dependent on how such assignments are used in pedagogical practice, which this project did not study. Furthermore, we focused most of our attention on the descriptions at an aggregated level.

Unequal descriptions at the aggregated level
The study shows interesting results for the descriptions at the aggregated level of all courses together. Above in the section on method, the article referred to an assignment where a man is gambling, described as active and independent, while a woman is afraid of losing her property, described as passive and dependent. Our study shows that this is not an isolated case but an example of a broader structure. At the aggregated level, the results show, with statistical significance, that the assignments describe men as more active than women (men 86.3% and women 74.4%), more independent (men 76.8% and women 63.4%), less dependent (men 17.9% and women 27.6%) and to a higher degree in the public sphere (men 55.3% and women 45.9%). This means men are described as more active, independent and in the public sphere while women to a higher degree are described as passive, dependent and in the private sphere ( Figure 13). The study argues that these descriptions of men and women reproduce traditional gender roles where men appear to have higher status and importance than women. The study also found results with statistical significance for similar structures concerning class/socioeconomic status. For example, characters from the higher class/socioeconomic status were almost all described as

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching 25 active (96.6%), especially in connection with business and as employers.
One example is an owner of a large corporation described with the verbs "agree", "establish", "appoint", "manage" and "push through" (Archive 2.13,author's translation). No people of the higher category are described as passive (0%). This may be compared to people of the middle category being described to a lesser degree as active (83.6%) and more as passive (13.6%), and people of the lower category as even less active (65.2%) and even more passive (23.9%) ( Figure 14). Furthermore, the study found that people of the higher class/socioeconomic status are never described in the private sphere (0%). The study also found that some social positions are more often than others described in negative terms. This includes men (often in relation to criminality and violence), non-Nordic people (often criminality), young people (often criminality and drugs) and people from lower class/socioeconomic status (often social problems). At the same time, Nordic people, adults, and people from middle class/socioeconomic status are more often described in neutral terms (in widely varying ways). The only social position described as more positive, with statistical significance, is people from higher class/socioeconomic status. One example of this is how two upper class characters are described with the adjectives "hardworking", "genius", "successful" and "highly competent" (Archive 2.7, author's translation). Interestingly enough, it is again primarily structurally normative/superior social positions according to our theoretical framework that are to a higher degree described in positions of higher status and greater importance. In this way, the study argues that the assignments reproduce stereotypes of a number of different groups as having less capacity to do good, and other groups as being the moral yardstick others are measured by, and finally the upper class being better and superior.
In the results, there were also two exceptions to this pattern. The first exception is that men are more often described in negative terms, although our theoretical framework regards them as having a structurally normative position. Masculinity is an interesting phenomenon and more complex than the study was able to elaborate on. A brief reflection, however, is that traditional male gender roles may well include doing negative things to others. The other exception is that non-Nordic people, people with disabilities, and people from lower class/socioeconomic status are more often described as main characters and less in supporting roles. This may be because the authors of the assignments have made active choices to include characters from these social positions. If these underrepresented positions were not included by active choice, they would probably be equally distributed as main and supporting roles. As mentioned above, among the non-Nordic characters there were also more references to known people, supporting the same argument. If this assumption of active choices is correct, it shows that the authors are willing to include people from non-normative positions, and at the same time, that such inclusion may contribute to upholding stereotypes.

The LL.M. Programme does not reflect the diversity of society, and not equally
As a conclusion on representation, the norm-critical study of the LL.M. Programme showed statistically significant imbalances regarding gender, ethnicity, age and class/socioeconomic status, when comparing the assignments at an aggregated level with the population in Sweden. Furthermore, the study also showed that social positions concerning gender identity and expression, religion or other belief, ability and different aspects of ethnicity appear only marginally. The study also showed that these imbalances and invisibilities correlate with structures of normativity and superiority. In other words, the structurally normative and/or superior positions of "white", Nordic, secular and adult (cis)men without disabilities, from the middle and higher class/socioeconomic status, are

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching overrepresented in the assignments, compared to the corresponding nonnormative and/or subordinate positions. With one exception however: the representation of homo-and bisexual people in the assignments was largely equivalent to their proportions in the population in Sweden.
The study on descriptions showed individual examples of assignments both reproducing and contradicting stereotypes. At the aggregated level however, the study showed statistically significant results that men, Nordic people, adults as well as people from the middle and higher class/socioeconomic status are generally described as being more important or more significant than the other social positions examined. Again, there are similar correlations with normative structures, with two exceptions. Men are described in a negative manner to a higher degree than women are. In addition, non-Nordic people, people with disabilities, and people from the lower class/socioeconomic status are more often in main roles and less as supporting roles. However, those exceptions might still fit into normative structures, as discussed above.
In summary, the study showed that the assignments of the LL.M. Programme at the University of Gothenburg do not reflect the diversity of society and do not describe people in different social positions equally. The study argues that the representation in the assignments thereby affirms and upholds ideas about which people are taken for granted as "normal" and which are considered to be "exceptions" to such norms, and also reproduces notions of which people have higher status and are more important than others. The gender and norm-critical studies that have been made of other LL.M. programmes in Sweden show that the University of Gothenburg is far from alone in this. 12 In addition to studying the present situation, the project also proposed and problematized measures to improve the degree of diversity and equality in legal education. As the structures that reproduce non-diversity and inequality are complex, there are certainly pedagogical challenges to accomplishing this. Many people have become engaged in developing pedagogical practice in relation to these issues. The next section of this article will present and discuss one such set of ideas aiming to address these issues: norm-critical pedagogy. The section then goes on to situate this approach in relation to the question of what degree of representation might be desirable in legal education. The discussion then ends with the more elementary question of whether diversity and equality really matter when teaching law.

Making use of norm-critical pedagogy
Norm-critical pedagogy, along with other critical pedagogies, have emerged as a response to what has been called "tolerance pedagogy", which is about tolerating what is considered different. But in merely tolerating the "others", social norms are not questioned, only broadened a little. Normcritical pedagogy instead aims to questioning the norms of what is taken for granted (Bromseth and Darj 2010;Kalonaityté 2014). Teachers may assume that people from all kind of social positions are represented in the room, which enables an approach of "we/you" rather than "we/them". Such perspectives do not presuppose a consensus on the existence or origins of social norms and power structures. How would such a consensus be possible to achieve on such complex matters? Instead, aware reflections on these issues may open the way for discussions were different standpoints are able to meet.
The authors of the first anthology on norm-critical pedagogy in Sweden have described their journey towards a norm-critical pedagogy in relation to social norms and power structures in roughly four general phases (Bromseth and Darj 2010). A teacher who finds their educational material to have imbalanced representation may want to make one or more social groups more visible. We can take people targeted by racism and ethnic stereotypes as an example. To start mentioning such people, for example, people of African heritage, indigenous peoples, or Swedes adopted from Asia is an important step forward. Such references make these social positions more visible and broaden the representation. However, there is always the risk that their inclusion will be stereotypical in nature, for example African people behaving strangely or mostly being angry. If people who are often targeted by ethnic stereotypes are not portrayed with forethought, there is a risk of merely reproducing the existing norms in society. Some ethnic stereotypes are rooted so deeply in our history that

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching they find their way into our minds unintentionally, and even without our awareness. When people of African heritage are portrayed, for example, there is a significant risk that they will be perceived against a background of deeply rooted racist prejudice, which has historical roots in the slave trade and afterwards and still impacts dark-skinned people even in modern times. Becoming conscious of the risk of reproducing stereotypes often leads to another phase of aiming for equal treatment. For teachers in law this often appear as appealing. This phase may have a focus on writing and talking about "white" people and people targeted by racism in an equal way, for example by describing people from both groups in positions with high status as well as low status. This kind of effort often does better justice to different social groups. At the same time, it risks maintaining "white" people's privilege invisible as well as the disadvantages that affect people targeted by racism.
Queer theory has contributed the perspective of switching attention from people appearing as "others" to focusing on people appearing as the norm. It does this by exposing heteronormativity as a structure that takes heterosexuality for granted if nothing else is clearly indicated. An additional pedagogical phase may therefore be to make privilege visible and explicitly talking about "white" people and the conditions that apply to them as a social group. This may include examples such as having easier access to the labour market and high-ranking positions in society, segregation in the housing market, and better treatment by the police and other authorities. Such social structures may in fact have legal relevance and affect legal processes, which might be important for law students to study. This phase requires the teacher to have knowledge of the actual circumstances or at least convey a critical approach to the students. Otherwise, this phase also risks upholding different stereotypes.
A fourth phase may be to focus less on people and social groups and instead focus attention on the very social mechanisms that create divisions and hierarchies between different social positions. For example, a teacher may show how norms of whiteness may make our spontaneous picture of a judge to be a "white" person. In addition, problematizing how dark-skinned judges are portrayed risks unintentionally reproducing ethnic stereotypes. The teacher may also encourage students to learn about how different ethnic groups have been historically disadvantaged so that they know what prejudices may need to be addressed to be able to provide legal counsel in a respectful manner, with for example Muslims, Roma people or people of African heritage.
All of these pedagogical phases are useful in themselves and may lead to developing legal education to improve its representation of diversity and its equality. Since individual teachers and the universities that teach law in Sweden have different starting points, different kinds of measures may be relevant.

No simple solutions to complex problems
A recurring question from the faculty concerned what level of representation they should strive to achieve, and how to implement such representation without the assignments becoming artificial or seeming to be forced. As a response to these discussions, the project team stressed that there are no obvious answers to such questions, as there are no simple solutions to complex problems. If we stick to the issue of diversity in assignments, we need to relate to several parameters. One point of reference might be to try to reflect the population. Another could be to try to make people in marginalized social positions visible and therefore make sure to include more transgender people or Roma people or people with cognitive disabilities than in the population, to name some examples. Another could be that a teacher might want the assignments to reflect the social reality of the specific legal questions at hand, for example, men committing more crimes than women do. The teacher might argue that working with such assignments helps students to prepare for their profession. Alternatively, a teacher might want to focus only on the aspects that are legally relevant and have neutral characters in their assignments, if for no other reason than not to make legal education more difficult than it already is. Yet another teacher might want to stress the aim of developing the student's ability to think critically. To achieve this aim, the teacher might use examples that go against stereotypes, forcing students to think about them, or stereotypical examples that the teacher helps the students to problematize.
All of the mentioned aspects or points of reference may be legitimate. Different choices will produce different results, and will need different pedagogical approaches and skills to be successful. Therefore, there is no point in producing a common standard for representation. It is also not

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching necessary, or even ideal, that all teachers or all law courses use the same approach. In other words, the project team argued that we need diverse approaches concerning diversity. Taking into account the thoughtfulness required in applying flexible measures described above, the project nevertheless proposed measures for the Department of Law at the University of Gothenburg to implement on these issues -some on the concrete level and some on a structural level. These may be of interest to other LL.M. Programmes at other universities. The teachers at the Department were encouraged to be creative in finding ways to get more diversity into the assignments. One concrete proposal was to use pictures in addition to texts, to use multifaceted material from reality such as verdicts, or to study legal questions relevant to different social positions such as discrimination or hate crimes. The report also mentioned some social positions that could receive special attention when designing assignments: • More woman in general and more often described as active, independent and in the public sphere. Men more often described in a positive sense not connected to traditional male gender roles, for example showing care for other people. • More non-Nordic people, especially people from other regions than the Western world and people other than those with "white" European descendants. In addition, to let non-Nordic people be represented with fewer references to known people. • Transgender people and religious people or organizations could be visible at least occasionally. In addition, more people with disabilities: physical, mental, and cognitive. • More explicitly young people, especially teenagers and young adults portrayed as active, independent and described in positive ways. Also more elderly people, in particular still alive and in contexts other than the law of succession. • More people from the working class and lower socioeconomic status, especially described as active and in positive ways.
However, more important measures should be taken on a structural level. Diversity and equality are not primary about changing names in these assignments. They need ongoing work on many levels. The project therefore proposed five structural measures to the Department: 1) To adopt clear objectives on diversity and equality in formal policy; 2) To task the teachers with discussing and reporting on how they relate to diversity and equality; 3) To develop financial incentives for teachers to work on such issues; 4) To develop structures for inviting and handling criticisms about such issues from students in a professional way; 5) To evaluate and follow up the work.

Do diversity and equality matter when teaching law?
The aim of the broader project in the study was not only to show how the assignments relate to diversity and equality, but also to open the way for discussions about what this means for the Department. This proved to be a far from straightforward task. In this debate, several positions have been taken. A fair number of the teachers were positive. They had already reflected on the issues in their own teaching and welcomed the discussions about the LL.M. Programme as a whole. Some of the teaching staff agreed in principle with the findings of the study, but were concerned that a normcritical approach would distract them from more important tasks at hand. That is, the aim of legal education is not to teach students to count women and men, but to teach students the law. There were also teachers who felt exhausted by the study and its recommendations. They felt that teachers already need to deal with growing numbers of students, ever-increasing administrative demands, and now they are also expected to rethink their pedagogical approach. Yet others were concerned that a norm-critical approach to legal education would lead to nothing beyond providing a veneer of political correctness on an unequal society. That is, what would really change -except that legal assignments would become cumbersome and appear forced and artificial -by for example, letting a judge be a black lesbian in a wheelchair? Finally, a small group of teachers were outright negative to the study and came up with ways of mocking the issue as a whole.
The discussions among the faculty tended to be most polarised concerning the methods used by the study, and especially the label "normcritical". Reports from other projects show that similar problems seem to have occurred in relation to the label "gender". 13 Although all studies need to choose a specific method to come to its findings, most researchers do not want to be remembered for the label used for their analysis tool. A lesson learned from a project which aimed to intervene in existing behaviour is that the presentation of such a study should try to connect to goals that are common to the involved people. At the Department of Law at the University of Gothenburg, "norm-critical" or "gender" are certainly not goals shared by all faculty. However, all teachers do value the quality of the education. Therefore, discussions on diversity and equality as aspects of the quality of education seemed to be more fruitful. Still, there was one issue arising in those discussions that the project had not been prepared for. Do diversity and equality matter when teaching law?
The author of this article, as well as many colleagues at the Department, are convinced that diversity and equality in legal education do matter enough to focus attention on them. To conclude the discussion in this article, I will discuss three significant aspects of why a lack of diversity and equality may decrease the quality of legal education: the risk of keeping injustices invisible; the risk of law students from various backgrounds finding themselves excluded; and the risk of future lawyers not being fully prepared for objective assessments. The discussion in this article concerns the pedagogy of legal education, and not how the law itself relates to diversity and equality.
Our societies rest on injustices in many respects. As an example, the economic prosperity of Europe and North America is partly based on the transatlantic slave trade in the past. Even Sweden was involved in the slave trade, into the 19 th century. Although Sweden was responsible for a rather small share of the total slave trade, it was large compared to the size of its population (Rönnbäck 2018). The slave trade, and its associated exploitation of and violence against African people, has been justified by ideas of black and white people being different and white people being superior to black people. Although this situation has changed dramatically over time, there is still discrimination against African heritage people as well many other ethnic groups.
American professor Iris Marion Young have described oppression as having a number of different aspects: economic exploitation, marginalization, powerlessness, cultural dominance, and violence. Cultural dominance is the process by which dominant groups represent themselves, their experiences and their culture as universal. These groups are well represented and described in nuanced ways, while other groups are rendered invisible or described in stereotypes. Through this, the dominant groups distinguish people as either "us" (themselves) or "them" (the others). The dominant groups measure the others in relation to themselves, in relation to what they perceive as reasonable, good and normal. This mechanism keeps all oppression invisible, and provides justifying reasons for discrimination (Young 1990). Using this lens to look at the assignments of the LL.M. Programme at the University of Gothenburg, the study did not find any clear racial stereotypes. However, there were only very few examples of assignments that included people from, or with a heritage from Africa or Asia. In Young's terms, the assignments thereby reproduce cultural dominance by invisibilization.
What concrete effects might that have? If African and Asian people are invisible in legal education, will the students get any chance to reflect on historical and current racial discrimination and other injustices? Will the students even get a starting-point for such discussions? If students nevertheless initiate such discussion, will the issue be viewed as something central to their legal education or something more peripheral? The ambition of working towards diversity and equality in legal education have been described by some teachers as aiming to politicise legal education. However, I argue that the conscious or unconscious maintenance of an imbalanced representation, resulting in invisibilization of injustices, could be equally described as a politicized phenomenon.
To continue with the second aspect of the significance of representation, we will take a quick look at what the study has found about the students' backgrounds. The majority of the law students at the University of Gothenburg are women, 66% women compared to 34% men (Benson 2016) ( Figure 15). Moreover, about 31% of the law students have a foreign background, meaning born abroad or having at least one parent born abroad. A total of 69% do not have a foreign background (SCB 2018) ( Figure  16). 14 The imbalance of representation among the law students compared to the assignments is even higher than the imbalance compared to the general population in Sweden (compare with Figures 1 and 2 above).
14 The project did not have access to this statistic when writing the report.

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching 35 Figure 15. Representation concerning gender, comparison with students Figure 16. Representation concerning ethnicity, comparison with students Do the female students find enough examples that mirror them to find themselves included in the material used in their legal education? Do the students with a foreign background? 15 What might be the results if such groups of students did find themselves excluded? The study did not have relevant statistics about other kind of backgrounds for the law students but there are probably students who are religious, have a disability or a transgender identity more frequently than the few examples of these social positions found in the assignments used in their education. A similar situation probably also applies to students from national minorities and students having origins outside the Western world or who are from regions with people typically targeted by racism. If people in such social positions already have an internalised self-identity of exclusion and invisibility, resulting from the cultural domination in society, is there a risk that the marginalisation in the assignments helps to maintain such a self-identity? And, on the contrary, would a more balanced representation, or even an overrepresentation, help to contradict such an internalisation? Such is the situation for homo-and bisexual students who can find their own social positions represented in a roughly equivalent degree to that of the population. The third aspect I will discuss concerns the preparation of future lawyers. In many positions, lawyers are obliged to conduct objective assessments. The LL.M. Programme is not only an institution aimed at gathering knowledge. Higher education is also obliged to develop the student's ability to make independent and critical assessments according to Sweden´s Higher Education Act (Chapter 1 Section 8). Furthermore, legal education is a major factor in socialising law students into the community of lawyers (Burman, Svensson and Ågren 2003). Ideally, this should be a community able to identify and react to stereotypical thinking. Lawyers become better in their profession if their thinking is based more on factual reasoning than on prejudices. Stereotypical thinking impedes critical thinking and, at worst, threatens the rule of law. Therefore, it is important that those who teach law reflect on what ideas, values and norms are communicated via how we choose to represent different social positions. How does imbalanced representation in assignments affect the student's ability to develop awareness and critical perspectives? Clearly, it is not only the assignments in themselves that matter, but also the way the assignments are presented and discussed in pedagogical practice. Some teachers provide stereotypical assignments for a pedagogical reason and make sure that the prejudices that students project onto the assignments are problematized. However, how do we manage to make visible and discuss not only individual examples of stereotyping, but also the imbalance at the aggregated level? An unconscious, uncritical or passive approach to diversity and equality from the teachers makes it harder for the students. Moreover, if the students take the initiative to problematize such assignments, are the teachers prepared to handle this situation?

Erik Mägi
Lessons from a Study of Diversity and Equality in Legal Education: Methodology, Findings and Relevance for Teaching

Conclusion: Fulfilment of the purpose?
Diversity and equality in legal education seem to be issues that engage the teachers, at least in terms of talking. So far, the Department of Law at the University of Gothenburg has not made any formal decisions on any measures more than that these issues will be dealt with in the future reform of the LL.M. Programme. However, there have been extensive discussions on the subject at different levels in the organization. A press agency article about the study and its findings was also spread widely in Sweden's national newspapers, presenting the Department of Law as taking these issues seriously (TT 2018). There is far from any consensus on how to relate to diversity and equality in education, but very few teachers at the Department are any longer unaware of the issues. Furthermore, two of the courses studied by FEMJUR that had particularly imbalanced representation have recently been updated because of other reasons. It is clear, without any formal study being made, that the updated courses have assignments with broader representation than their predecessors did. However, the number of people from different social groups are not the most important part of how legal education relates to diversity and equality. In my view, the most important thing is to take a step from an unconscious and passive approach towards a more conscious and critical perspective. Thus, the project seem to have at least partly fulfilled its purpose by rocking the boat.