Baseline attainment data: legal education, training and post-qualification Supporting paper

Baseline attainment data: legal education, training and post-qualification Supporting paper October 2015

Contents Executive summary... 4 Ethnicity... 4 Gender... 4 Disability status... 4 Social mobility... 5 Intersectionality double disadvantage... 5 Full time, accelerated and part time study... 6 Overall... 6 Introduction... 7 GDL/LPC... 7 Trainees... 8 Other legal education... 8 Qualified solicitors... 8 Data issues... 8 1. Variation in attainment: the GDL... 9 Ethnicity... 10 Gender... 12 Disability status... 14 2. Variation in attainment: the LPC... 15 Ethnicity... 16 Conclusions... 20 Gender... 20 Conclusions... 22 Disability status... 23 Conclusions... 24 3. Variation in attainment: training contract... 25 Ethnicity... 25 Gender... 28 Disability status... 29 4. Variation in career following admission to the roll... 29 03/12/2015 Page 2 of 38 For alternative formats, email contactcentre@sra.org.uk

Ethnicity... 29 Gender... 30 Disability status... 31 Conclusions... 32 5. Conclusions... 33 Annex 1: Variation in attainment, the wider picture... 34 Ethnicity... 34 Trends... 34 Contributing factors... 35 Gender... 36 Trends... 36 Contributing factors... 37 Disability status... 37 Trends... 37 Contributing factors... 38 Page 3 of 38

Executive summary Ethnicity There is an attainment gap between white and Black, Asian and minority ethnic (BAME) students at all levels of legal education white students have a higher pass rate and higher scores and on average are paid more during their training contract. Once qualified, BAME solicitors are less likely to work in higher paid roles in larger firms, and are less likely to be partners. The attainment gap is particularly pronounced for Black students, who as a group have lower scores and lower pass rates than white, Asian and mixed ethnicity students. This gap is also present in other sectors and in education at GCSE, A-level and undergraduate degree level. Across secondary and higher education in general, this gap is not closing over time; as results improve for BAME students, they also improve for white students and so the gap remains. This impacts on efforts to improve diversity in the law through widening access to the profession - if access is not matched by the closing of the gap, inequalities will persist. If a new centralised assessment is introduced, we will need to measure the improvement (or lack of) in the gap, as well as the improvement of BAME students attainment. We should also recognise that there is a large variation in attainment across different ethnic groups, and so we need to monitor these trends, as well as BAME overall. There are also some consistency issues, such as making sure we use the same ethnicity classifications for monitoring across different stages of education. Gender A substantially higher proportion of women than men are currently entering legal education and training. The pass rates for both men and women are very similar for both the GDL and LPC. However, there are proportionately more women obtaining distinctions than men. Once solicitors are qualified, men are conversely more likely than women to be working in higher paid roles in the larger firms and more likely to be partners. A new assessment regime would mean we need be aware that disproportionate outcomes start to appear once women become qualified solicitors, not at the training stage. Therefore, change in the way students train and become qualified is unlikely to have a significant impact on outcomes that only emerge post-qualification. Disability status The LPC and GDL pass rate for students who define themselves as having a disability is less than those who do not. Evidence also suggests that students who do pass the LPC and declare a disability are less likely to have a training contract. However as most people choose not to state their disability status, we cannot know whether this is genuinely disproportionate. Page 4 of 38

When using this data as a baseline for measuring change, we should monitor prefer not to say and non-response rates as well as those who choose to disclose their status. Social mobility Understanding disproportionality in educational attainment related to social background is difficult. This is due to the different proxy measures used by different agencies. No data is available for legal education (GDL/LPC) specifically. Measures used in general secondary and higher education include: whether someone is the first generation of their family to go to university whether they went to a fee-paying school whether they received free school meals There is overall evidence suggesting an educational attainment gap between wealthy and poorer backgrounds beginning early in education. Post-qualification, solicitors who attended fee-paying schools are over-represented in the profession in general, and even more so at partner level, in larger firms and in corporate work. Further research is required to determine the best method of measuring social background. However we may wish to use some of those, particularly if institutions running legal education programmes are already collecting this information about all their students. Intersectionality double disadvantage A common theme emerging is that people who are a member of several different disadvantaged groups have a lower attainment than if they were a part of just one. For example, BAME students and from a low income background are less likely to do well than those who are either BAME from a wealthy background, or white from a low income background. This intersection results in a double disadvantage, meaning that trends on one characteristic can only go so far in telling the full story of the attainment gaps that disadvantage access and achievement in the law. Cross-factor analysis is necessary to fully identify those with the lowest attainment and understand the reasons for the gap. For this reason, it is important that the data we collect on any new assessment allows us to see attainment for groups with more than one characteristic. This may pose some problems with regard to data protection, but can be appropriately managed if built into the plans at an early stage. 1 1 Data protection issues may arise if we publish any analysis of the data which makes individuals identifiable for example, if we published or shared data about trainees at firm level, and any firm had a trainee with a unique combination of characteristics. However, it is Page 5 of 38

Full time, accelerated and part time study The attainment gap becomes more pronounced when we address the mode of study. Full time students are by far the largest group and for the sake of mode of study analysis can be described as the norm. Accelerated full time study has a high pass rate compared with the norm, and a smaller proportion of female, BAME and disabled students are entered on accelerated courses compared to full time courses. Part time courses have a lower pass rate than the norm, and a larger proportion of female, BAME and disabled students undertake part time study. Overall Overall, there is no evidence that disparity in legal educational attainment is due to different factors than is seen in other types of educational attainment. However, the increase post-qualification suggests that firms recruitment has a compounding impact on disparities already established. We can therefore hypothesise that if changes in assessment were to close the attainment gap for any group, this would need to combine with accompanying changes in recruitment to close the achievement gap post-qualification. We can attempt to address the disparity at the point of qualification by introducing a centralised assessment, but there remains the legacy of prior disadvantage in the education system, which may limit the impact that these changes can have. This does not mean that there is no benefit in attempting to address this disparity, but it does mean we should be realistic about the change we can achieve through altering one stage of the educational journey. unlikely that we would wish to publish any data at this level of granularity. In terms of sharing the data, we need to ensure that any assessment provider collecting data make clear how the data will be used when collecting it. Page 6 of 38

Introduction This report has been commissioned as baseline evidence against which to measure the impact of policy changes in education and training, in particular the proposed introduction of changes to assessment. This outlines what we know about how attainment, access, and progression in the profession vary according to: Ethnicity Gender Disability status We are not able to look at social background as we do not currently collect this data. The levels of education covered in this report are: Graduate Diploma in Law (GDL) Legal Practice Certificate (LPC) Training contract (TC) For each of these levels of education, this report looks at attainment data for the cohort undertaking the assessment (or training contract) in the 2013/14 academic year. GDL/LPC We excluded students from the analysis who: Withdrew from the course prior to assessment Were members of a previous cohort and were taking the assessment as a resit or deferred entry Were undertaking part time courses but were only in their first year of study and so had no assessment outcome Did not have valid data about their attainment (e.g. a blank response or not applicable ) Passed the assessment via an exempting law degree. Students were include were those who: And either: Were part of the 2013/14 cohort and were either studying full time or part time and receiving their final assessment outcome in 2014, or studying on an accelerated course Took the assessment and received a pass (distinction, commendation or pass) Took the assessment and received a fail Failed an attempt at an assessment and were referred for resits (described in the report as referred ) Undertook their full period of study but deferred taking the assessment due to extenuating circumstances (described in the report as deferred ). By including and excluding certain students on the GDL and LPC in our analysis, we have attempted to get the most accurate picture of attainment. The same criteria Page 7 of 38

should be used in determining the sample for analysis in future years if we wish to use this report as a baseline. Trainees For trainees (students who have obtained training contracts), we have looked at the diversity of the trainee cohort for 2013/14, and have also looked at the variation in trainee salaries. However, it is worth mentioning that trainees may choose their training contract based on many more factors than salary. We are not using salary to represent quality of life, but instead are using it as a proxy for future earning potential and the status of the firm. Other legal education We do not have a breakdown in attainment for qualifying undergraduate law degrees (often called the qualifying law degree or QLD ). Qualified solicitors The report also looks at the diversity of qualified solicitors, and general secondary and higher education attainment trends, as well as evidence of any factors that are driving variation in attainment. The data about the legal profession post-qualification is for the year 2012/13, as this is the most recent published data available. We have included the more general analysis of attainment trends as an annex (Annex 1), for easy referencing to give the law-specific data some context. By setting a baseline for the difference in attainment for different groups, we will be able to measure any changes when the new assessment is introduced. Data issues There are three important points about the data used in this report. 1. We cannot track groups of students through their education and training. We do not collect data in a way that allows specific groups to be tracked through different stages of education and training. For example, we cannot say how many people from different groups dropped out of each stage if x female solicitors pass the LPC and y female solicitors get training contracts for the subsequent year, this does not mean that x-y solicitors did not get a training contract. Possible solution: In future, we could collect data which allowed us to track different groups. This is called longitudinal data and we should think about whether this is a key type of data we would like to collect. 2. The reliability of the data we have varies. Data quality varies depending on whether collection is compulsory or optional. For example, for the 2013/14 academic year, 65% of trainees who registered with us did not declare their Page 8 of 38

ethnicity. This is different to a prefer not to say response, as trainees have not been directly asked to provide their diversity information as part of the registration process. Instead they are able to update this on mysra should they prefer. We will need to consider balancing the burden on individuals and institutions against the need to collect data to measure progress. Possible solution: We could think about how to get a better response rate to our data collection exercise through things like reminders to students or institutions, or even some incentives to encourage participation. 3. Low statistical power (small numbers) Some of the groups are made up of a comparatively small amount of people. For example, 103 students who took the full time LPC in 2013/14 described their ethnicity as other. This meant that if 11 people failed, the pass rate for students of other ethnicity decreased by over 10%. If 11 white students failed, this constituted a decrease in the pass rate of less than 1% due to there being 2826 white students on the same course. Possible solution: We will measure this effect of small groups and incomplete data sets being more unpredictable through statistical confidence testing. We will look at how confident we are that the result represents the norm and is not a freak occurrence. We can also look at data over a number of years to spot odd years out, and increase understanding of small groups by combining over years to increase statistical power. 1. Variation in attainment: the GDL The pass rate and scores for the main postgraduate law qualifications vary according to protected characteristics. The GDL (often called the law conversion ), is examined through the Common Professional Examination (CPE). This diploma is a one-year course to prepare nonlaw graduates for a career as a solicitor or barrister. The SRA specifies that the pass mark is 40 per cent. This chapter covers trends in results achieved according to ethnicity, gender and disability, and looks at whether students studied full time or part time. The number of students who take the GDL is far lower than the number who take the LPC. The majority of students access the LPC through a qualifying undergraduate law degree. Unfortunately we do not have data about qualifying law degree attainment. This means that the GDL data is not representative of the majority of pre-lpc legal education, and the lower numbers also means that it is less likely to be accurate compared with the LPC data, for the reasons discussed in data issues (3) above. Page 9 of 38

Ethnicity Full time study From figures 1 and 2 below we can see that although the majority of full time students pass the GDL (whether with a pass, a commendation or a distinction) the difference in results varies hugely between students of different ethnicities. In 2014, over 1 in 4 BAME students were referred for resits, compared with less than 1 in 10 white students. When BAME is broken down further, the differences are even more stark; no Black students taking the GDL received a distinction, whereas nearly 1 in 5 white students did, and nearly half (48%) of Black students were referred for resits. Figure 1: Variation in GDL results by ethnicity 2014 (2617 full time students) 2 2 Does not include 245 students who either did not respond or responded with prefer not to say Page 10 of 38

Figure 2: Variation in GDL results by ethnicity 2014 (2617 full time students) 3 Part time study A minority of GDL students study part time over two years, rather than full time over one year. Figure 3 below shows those who studied part time and completed the GDL in 2014 (rather than those who were still in their first year). These results are less reliable as in some of the ethnic groups there are very small numbers of people studying part time. Therefore it is difficult to say whether the 2014 results accurately reflect any disadvantage between ethnicities in part time study. That being said, white students have a higher pass rate and far less white students were referred for resits than other ethnicities, and this reflects the pattern seen in the full time students. It is also worth noting that pass rates for Black students are higher in part time study than full time study. However, again, the numbers are very small and so this result may be less reliable. 3 Does not include 245 students who either did not respond or responded with prefer not to say Page 11 of 38

Figure 3: Variation in GDL results by ethnicity 2014 (705 part time students) 4 Gender Full time study From figure 4 below we can see that pass rates and results are very similar for male and female students the only difference being the marginally higher percentage of women who got a distinction as opposed to a commendation. 4 Does not include 245 students who either did not respond or responded with prefer not to say Page 12 of 38

Figure 4: Variation in GDL results by gender 2014 (2865 full time students) 5 Part time study A minority of GDL students study part time over two years, rather than full time over one year. Figure 5 below shows those who studied part time and completed the GDL in 2014 (rather than those who were still in their first year). As can be seen from Figure 5, the pass rates and results for male students and female students studying the GDL part time are again very similar, much like they are for those full time. 5 Does not include 1 student who did not respond Page 13 of 38

Figure 5: Variation in GDL results by gender 2014 (897 part time students) Disability status In 2014, the majority of students (61%) who took the GDL did not provide their disability. This means that we cannot say with a large amount of certainty whether the data in figure 6 below is an accurate representation of the differences in attainment between students with disabilities and those without. Numbers for part time study are so small (73 people with a disability took the GDL part time) that we have presented the data of full time and part time together. Figure 6: Variation in GDL results by disability status 2014 (3588 students) Page 14 of 38

2. Variation in attainment: the LPC The pass rate and scores for the main postgraduate law qualifications vary according to protected characteristics. The LPC is the one year (full time, can also be taken as two years part time) vocational course for those wishing to become a solicitor, which is undertaken prior to training in a law firm. The SRA specifies that the pass mark is 50%. However, providers can set their own boundaries for commendation and distinction. However the grades above pass are not included on the student's transcript. This chapter covers trends in result achieved in the LPC according to ethnicity, gender and disability, and looks at whether students study full time, part time, or on an accelerated course. Page 15 of 38

Ethnicity Full time study From figures 7 and 8 below we can see that although the majority of full time students pass the LPC (whether with a pass, a commendation or a distinction) the difference in results varies hugely between students of different ethnicities. In 2014, nearly half of all full time white students who took the LPC received a distinction, compared with only around 1 in 5 BAME students. Also around 1 in 4 BAME students were referred for resits, compared with less than 1 in 10 white students. When the BAME group is broken down further (figure 8), there is variation within the group Black students do proportionately worse than Asian students, although both groups have lower results than white students. Figure 7: Variation in LPC results by ethnicity 2014 (4210 full time students) 6 6 Not included are 349 students who either preferred not to say or did not respond Page 16 of 38

Figure 8: Variation in LPC results by ethnicity 2014 (4210 full time students) Accelerated study The analysis above relates to full time students completing the LPC over the normal period of a year. However, some take an accelerated course at either BPP or the University of Law. This is a shorter course (e.g. six months) that is often created as bespoke provision for a large firm or a consortium of large firms. Overall pass rate for this course is higher than for the standard LPC, as can been seen by comparing figure 8 above with figure 11 below. However, the variation in attainment between white and BAME students is very similar to the full time one year LPC students. Figures 9 and 10 show that a higher percentage of the students who take the accelerated course are white: 75% compared with 67% of the full time one year students. This is offset by proportionately less Black and Asian students taking the course. Therefore BAME students could said to be even more disadvantaged on the accelerated course, as they have lower participation compounding their lower results. Figure 9: One year participation Figure 10: Accelerated participation Page 17 of 38

Figure 11: Variation in LPC results by ethnicity 2014 (709 full time accelerated students) 7 Part time study Not all students taking the LPC choose to take the course full time, and instead take the opportunity to undertake it on a part time basis. Reasons for this include affordability and caring responsibilities. Overall pass rate for part time students is lower than for full time, as can been seen by comparing figure 8 above with figure 12 below. Low pass rate is particularly apparent for Black part time students; only around 1 in 3 passed, with the remaining students either failing, being referred for resits due to failing one or more modules, or deferring. 7 Not included are 77 students who either preferred not to say or did not respond Page 18 of 38

Figure 12: Variation in LPC results by ethnicity 2014 (1728 part time students) 8 Figure 13 below shows that Black and Asian students are more likely to undertake part time study than white students. This again raises the issue of a double disadvantage, because the pass rate for the part time course is lower than for the full time. Figure 13: Students taking the LPC (excluding those withdrawn or suspended) by mode of study and ethnicity 8 Not included are 217 students who either preferred not to say or did not respond Page 19 of 38

Conclusions For the LPC there is a significant attainment gap between white and BAME students. More white students pass overall, and more get distinctions, which can be used as a proxy for higher marks. Additionally, proportionately more BAME students study part time, and the results for part time study are lower than those for full time study. Gender Full time study In 2014, 64% of students taking the LPC were women. From figure 14 below we can see that pass rates for the full time LPC (whether with a pass, a commendation or a distinction) are the same for women and for men. However, there are differences in terms of the level of pass achieved; women are more likely to get a distinction than men are, and men are more likely to get a pass than women are. So men are disadvantaged in that fewer men are enrolling on the LPC, and are also as a group achieving lower marks. However, compared to the disparity between BAME and white students, the gap between male and female results is small. Figure 14: Variation in LPC results by gender 2014 (4561 full time students) Accelerated study As with ethnicity above, we have analysed attainment of students undertaking an accelerated version of the LPC, by gender. Page 20 of 38

Interestingly, the participation gap between male and female students on the full time course (62% female, 38% male) is not present on the accelerated course (51% female, 49% male). Figure 15 below shows that the distance between male and female achievement of distinctions is more pronounced among those who take the accelerated LPC, as was the case with ethnicity. However, the percentage of students who got a pass are very similar. As with full time students, the overall pass rate itself is very similar, whether they are male or female. Figure 15: Variation in LPC results by gender 2014 (786 full time accelerated students) Part time study As with ethnicity above, we have analysed attainment of students undertaking a part time version of the LPC, by gender. The participation gap between male and female students on the part time course (68% female, 32% male) is relatively similar to that on the full time course (62% female, 38% male). Figure 16 below shows that unlike on the full time and full time accelerated courses, the pass rate for male and female students is not equal for part time study of the LPC. Instead, male students are less likely to pass than female students are. Page 21 of 38

Figure 16: Variation in LPC results by gender 2014 (part time students) Conclusions Significantly more women than men take the LPC. Overall there is no difference in pass rates for full time students (whether accelerated or not), however women get proportionately more distinctions and commendations, whereas men are more likely to get a pass rather than a commendation or distinction. In terms of part time study, unlike full time students, where there is no difference in pass rate, men have a lower pass rate than women when we look only at those studying part time. Page 22 of 38

Disability status Full time study In 2014, 9% of students taking the LPC had a disability, which reflects the UK student population, where 9.5% have disabilities. 9 From figure 17 below we can see that pass rate for the full time LPC (whether with a pass, a commendation or a distinction) is lower for students with disabilities than for those without, or those who did not provide a response. Figure 17: Variation in LPC results by disability status 2014 (4561 full time students) Accelerated study As with ethnicity and gender above, we have analysed attainment of students undertaking an accelerated version of the LPC, by disability status. Only 5% of students taking the accelerated full time LPC in 2014 had a disability. This constituted 38 people. Additionally, just over 40% of students did not provide a response about whether they had a disability. Therefore, we are not able to draw any meaningful conclusions about the difference in attainment between students with disabilities and those without on the accelerated full time LPC. A more useful line of enquiry may be why students with disabilities are underrepresented on LPC accelerated study programmes. 9 Excludes those who enrolled in a previous year and were undertaking resits. Source of UK student population stats: Equality in higher education: statistical report 2014, Part 2: students, Equality Challenge Unit, 2014 Page 23 of 38

Part time study As with ethnicity and gender above, we have analysed attainment of students undertaking an accelerated version of the LPC, by disability status. 10% of those studying part time LPC had a disability. This is slightly more than for full time students, where 9% had a disability. As can be seen from figure 18 below, the gap in the pass rate between students with disabilities and those without is more pronounced if they are studying part time. Figure 18: Variation in LPC results by disability status 2014 (1924 part time students) Conclusions Students with disabilities have a lower pass rate for the LPC than those without disabilities, whether studying part time or full time. They are also significantly underrepresented in accelerated study programmes, which have a very high pass rate. Page 24 of 38

3. Variation in attainment: training contract In order for a student to become a solicitor, they must currently undertake a period of work-based training with a recognised training provider (law firm). When a student is accepted to train at a firm, they are said to have a training contract and become a trainee. The diversity data we hold on trainees is not as comprehensive as for the other stages. This is due to work we have undertaken to reduce the burden around trainee registration. For example, for the 2013/14 academic year, 65% of trainees did not declare their ethnicity. This is different to a prefer not to say response, as they have not been directly asked to provide their diversity information as part of the registration process. Instead they are able to update this on mysra should they prefer. We do not have information on conversion rates comparable with pass rates for whether people got a training contract. However, we can compare the diversity of training contract holders against that in the general population and those who passed the LPC in the same year. It is important to remember that the training contract figures do not represent a conversion rate between the number of students who passed the LPC in 2012/13 and those who got training contracts for 2013/14. This is because students do not necessarily pass the LPC and obtain a training contract in the same year. To get accurate conversion rate data, we would have to begin collecting longitudinal information about each student. We are measuring attainment of different groups of students at trainee level by looking at the characteristics of trainees, and looking at trainee salaries. Ethnicity Ethnicity of trainees As previously mentioned and shown in Figure 19 below, the majority of trainees registered with us in 2013-14 did not declare their ethnicity. Page 25 of 38

Figure 19: Training contracts registered with the SRA between 1 July 2013 and 30 June 2014, by whether ethnicity of trainee was known or unknown Figure 20 below shows that since we stopped proactively asking trainees to complete a diversity monitoring form (in 2012), the number of trainees with unknown ethnicity has increased. Figure 20: Rise in trainees with unknown ethnicity, 2010-2015 Figure 21 shows the ethnicity of those trainees (2014 cohort) who did provide this information. The distribution is relatively similar to that for students who passed the LPC, suggesting there is little disadvantage related to ethnicity in terms of obtaining a training contract. However this data is unlikely to be as reliable as the other data in this report. Therefore going forward we may need to have more trainees with known ethnicity to understand whether there is any disadvantage in terms of obtaining a training contract. Although Figure 21 below has been included for completeness, we do not believe it represents the full picture. Page 26 of 38

Figure 21: Training contracts registered with the SRA between 1 July 2013 and 30 June 2014, by ethnicity (students of known ethnicity only) Trainee salaries Figure 22 below shows the average salary for trainees of different ethnicities in for the 2014 cohort, and we can see that BAME trainees are on average earning less than white trainees. However, again it is difficult to establish whether this truly reflects all trainees due to the high proportion with unknown ethnicity. Figure 22: Median trainee salary by ethnicity, 2014 10 Figure 23 below shows the percentage of trainees earning at the lower end of the pay scale 11 by ethnicity. Around 1 in 5 BAME trainees are earning a relatively low salary, compared to just over 1 in 20 white counterparts. However, again this comes with the caveats associated with the large amount of trainees of unknown ethnicity. 10 Intake beginning 1 August 2014 11 A salary of below what was formerly the minimum training salary outside of London ( 16,650) has been used as an indicator of the lower end of the salary scale. In the future we may want to look at trainee salaries in terms of percentile distributions or quartiles, rather than above or below former minimum salary, given the minimum salary removal. Page 27 of 38

Figure 23: Trainee salaries above and below benchmark of former minimum salary by ethnicity, 2014 Gender Gender of trainees As can be seen from Figure 24 below, there are significantly more female trainees than male. The percentage of female trainees is very similar to the percentage of female students taking the LPC (64%) and women taking the LPC who had a training contract before they started the course (65%). There is therefore no evidence of disadvantage on the basis of gender in terms of likelihood of becoming a trainee. Figure 24: Training contracts registered with the SRA between 1 July 2013 and 30 June 2014, by gender of trainee Page 28 of 38

Disability status The number of trainees who disclosed a disability was very low, at less than 1%. In 2013/14, 9% of LPC graduates had disclosed a disability, which suggests that either students with disabilities may be less successful in obtaining training contracts than those who do not have one, or they are choosing not to disclose their disability. Trainees either fill out the type of disability they have, or leave it blank, and so we cannot tell whether they preferred not to state their disability, or were a non response. For this reason we may wish to look at ways to improve our data for the future. 4. Variation in career following admission to the roll There is also variation in the careers of solicitors depending on their ethnicity, gender and disability status. Ethnicity 13% of the solicitors we regulate are BAME solicitors. Figures 25 and 26 below are taken from our diversity monitoring report 2012/13, 12 and shows a breakdown of the profession by size of firm and ethnicity, and the percentage of each ethnicity that are at partner level. It shows that BAME solicitors are under-represented in larger firms compared to smaller firms, and are proportionately less likely to be partners than white solicitors. Figure 25: Ethnicity of qualified solicitors by the size of firm they work in, 2012/13 (101,284 individuals) 12 Diversity monitoring statistics 2013, Solicitors Regulation Authority, 2014 Page 29 of 38

Figure 25: Percentage of qualified solicitors of each ethnic background who are partners or equivalent, 2012/13 (101,284 individuals) Gender 43% of the solicitors we regulate are female, and 57% are male. Figures 27 and 28 below are taken from our diversity monitoring report 2012/13, and shows a breakdown of the profession by size of firm and gender, and the percentage of each gender that are at partner level. This shows that female solicitors are better represented in larger firms than smaller firms, but are proportionately less likely to be partners than male solicitors. Figure 27: Gender of qualified solicitors by the size of the firm they work in, 2012/13 (101,284 individuals) Page 30 of 38

Figure 28: Percentage of qualified solicitors of each gender who are partners or equivalent, 2012/13 (101,284 individuals) Disability status Less than 1% of the solicitors we regulate have declared they have a disability. Compared with the general population this is very low, and so it may be that there are more solicitors with disabilities working in the profession who have not declared it. Figures 29 and 30 below are taken from our diversity monitoring report 2012/13, and shows the percentage of solicitors with disabilities at different size firms and those who are at partner level. Figure 29 shows that solicitors with disabilities are better represented in smaller firms, but numbers are low in all types of firm and so it is difficult to draw a conclusion. Figure 30 shows that solicitors with disabilities are proportionately less likely to be partners than average. Overall, only 1% of partners have a disability. This may be due to lack of disclosure as much as disability status. Figure 29: Percentage of solicitors with disabilities by the size of the firm they work in, 2012/13 (101,284 individuals) Page 31 of 38

Figure 30: Percentage of solicitors with disabilities working as partners or equivalent, 2012/13 (101,284 individuals) Conclusions The attainment gap present in legal education is mirrored by an achievement gap post-qualification. We are aware that for some individuals, a position as partner or at a larger firm is not what they choose to do, and that these measures do not represent quality of life. However, as a general proxy measure, they do appear to show a disparity of achievement post-qualification. BAME, women and disabled solicitors are underrepresented at partner level, and BAME and disabled solicitors are proportionately more likely to work at a smaller firm. We can attempt to achieve a more equal level of attainment between different groups of students in legal education. However the data in this chapter shows that there are inequalities that persist post-qualification which may be unrelated to academic achievement. Therefore a new type of centralised assessment is unlikely to remove underlying inequalities post-qualification. Page 32 of 38

5. Conclusions Looking at the disparity of attainment across different stages of legal education, it is clear that there are inequalities depend on the student's background and mode of study. These inequalities persist after qualification. However, we cannot ascertain whether the reasons for these inequalities are directly related to legal education and the legal sector; there are also attainment gaps present in education in general, as can be seen from Annex 1. With this report, we have set out a baseline of the existing attainment gaps, which can be used to measure the impact of a new centralised assessment system. It will be interesting to see whether we are able to narrow some of the attainment gaps through a more consistent method of assessment that is as unbiased as possible. Page 33 of 38

Annex 1: Variation in attainment, the wider picture At university level in the UK, pass rate and degree class vary by protected characteristic. The sections below outline current trends by characteristic, and any evidence of the factors causing these trends. Ethnicity Trends Figure A below shows trends in the number of students who obtained first or 2:1 degrees (as opposed to a 2:2 or lower.) White students consistently achieve higher degree outcomes than students of other ethnicities. Attainment overall has risen, but the attainment gap has remained broadly the same. Figure A: Percentage of UK students of different ethnicities who received a 1 st or a 2:1 in their degree Source: Equality in higher education: statistical report 2014, Part 2: students, Equality Challenge Unit, 2014 However, when the group BAME is broken down, we can see that there is a huge variation in attainment by different ethnic group, and so the catch-all category BAME does not necessarily tell the whole story. 1 st /2:1 attainment ranges from 47% for black graduates to 67% for graduates of mixed ethnicity. Even within broad ethnic groups, attainment varies; for example while 64% of Indian Asian graduates get a 1 st or 2:1, only 54% of Pakistani Asian graduates do. This indicates that the relationship between ethnicity and attainment at degree level is more complex than just a disparity between white and BAME groups. Page 34 of 38

Figure B: Breakdown in percentage of UK students of different ethnicities who received a 1 st or 2:1 in their degree for the academic year 2012/13 (Latest data available) Source: Equality in higher education: statistical report 2014, student data tables, Equality Challenge Unit, 2014 The attainment gap present at undergraduate level and in legal qualifications is also present in other sectors. Similar attainment disparities between white and BAME students have been found, for example, in medicine. 13 Contributing factors Evidence has shown that there is no one key factor that explains the difference in attainment at degree level for different ethnicities. However, some common factors have been identified. Students with better A levels do better in higher education. This is a complex area which is not fully understood, but the university undergraduate degree attainment gap is likely to remain while a similar gap exists at A level. 14 Research has estimated that around half of the disparity between white and BAME students is related to the level of their entry qualifications. 15 13 Woolf, Potts, and McManus Ethnicity and Academic Performance in UK Trained Doctors and Medical Students: Systematic Review and Meta-Analysis in British Medical Journal 342: 901 14 Differences in degree outcomes: Key findings, Higher Education Funding Council for England, 2014. These findings relate to students on full time undergraduate degrees, and look at cohorts starting their degree from 1997-98 through to 2007-08. 15 Richardson, J The under-attainment of ethnic minority students in UK higher education: what we know and what we don t know in Journal of Further and Higher Education 39:2, 278-291, 2015 Page 35 of 38

Students from disadvantaged areas on average do less well in higher education than those with the same prior educational attainment from more advantaged areas. As BAME students are more likely to be from disadvantaged areas, this is a contributing factor. This factor is particularly pronounced for Bangladeshi and Pakistani communities, where 1 in 3 people lives in a deprived neighbourhood. 16 However, even when factors like deprivation and A level attainment are taken into account, there is still an attainment gap. This leads researchers to question whether factors such as unconscious bias in the education system (particularly the role of white middle class norms in the way exams and education in general are framed) and cultural capital (encompassing things like social networks and understanding of higher education behaviours ) are influencing the gap. 17 It is a reasonable assumption that a significant part of the attainment gap in law results from the same factors as the gap in higher education in general; that is factors unrelated to the legal education system or legal sector in particular. However, there have been suggestions that there are some factors specific to law which may explain the wider gap. Cultural family preference for children to pursue careers in law. The head of the Office for Fair Access has discussed that some parents from certain cultural backgrounds put significant pressure on their children to pursue high status careers such as law and medicine, regardless of whether the children themselves want to do this. This could potentially be leading to these students not having the will or the aptitude to complete their course. However, this viewpoint has been challenged as lacking evidence and as an example of unhelpful stereotyping by some commentators. 18 We should be aware of cultural factors in the attainment gap, but be careful not to make assumptions which are largely based on opinions. Gender Trends Interestingly, the gender split in terms of all UK students is far more equal (56.2% female, 43.8% male) than it is in legal vocational education and training. These figures have remained consistent over time, with the proportion of male students increasing at a very gradual rate from 42.7% in 2003/04. 16 Ethnicity and deprivation in England: How likely are ethnic minorities to live in deprived neighbourhoods? ESRC Centre on the dynamics of ethnicity, 2014 17 For example: Broecke & Nicholls Ethnicity and degree attainment, Department for education and skills, 2007; and D Alton-Harrison Awakening the unconscious: Mind the attainment gap in Blended learning in practice, October 2013 18 Are Ethnic Minority Students Being Forced Into Medicine And Law Degrees, As Les Ebdon Claims? The Huffington Post, 2013 Page 36 of 38

However, when we look at taught postgraduates (comparable with the GDL/LPC), the proportion of female students goes up to 61%, similar to the proportion of women GDL/LPC students. This is also the case for students studying any type of law degree they are 60.3% female students. This may go some way to explaining why the percentage of LPC students who are female is so high we can hypothesize that the same disparity is present at qualifying law degree level. 19 In terms of degree attainment, female students getting a 1 st or 2:1, are a few percentage points higher when compared to men, both in general and in all law degrees. Contributing factors Students with better A levels do better in higher education. The higher proportion of female students on the LPC could be partially as a result of the higher female UK degree attainment detailed above, as the qualifying degree class for the GDL/LPC is usually a 2:1, but no causal analysis has been done and so this is currently only a hypothesis. It has additionally been suggested that higher GCSE and A level attainment for female students is a factor in this. 20 Women also make more applications to university than men. 21 Students from disadvantaged areas on average do less well in higher education than those with the same prior educational attainment from more advantaged areas. Evidence from UCAS shows that the difference between higher numbers of female to lower numbers of male university applications is wider in more deprived areas. 22 When compared with UK degree attainment in general, there is less disparity of attainment in the GDL and LPC, where male and female attainment is very similar. Disability status Trends The number of UK students who have a disability has increased steadily over the last ten years, from 5.4% in 2003/04, to 9.5% in 2012/13. 23 However, for taught postgraduate study (comparable with the GDL/LPC), the figure is lower, at 8.3%, and for all those studying some type of law course, the figure is 8%. 24 A figure is also available for taught postgraduate in law, where the figure is 6.1%. 19 The statistics on UK students does not differentiate between qualifying law degrees and other law degrees, and so we have to use this figure as a proxy 20 Gender gap in university admissions rises to record levels, The Guardian, 2015 21 End of cycle report 2014, UCAS, 2014 22 End of cycle report 2014, UCAS, 2014 23 Equality in higher education: statistical report 2014, student data tables, Equality Challenge Unit, 2014 24 Equality in higher education: statistical report 2014, student data tables, Equality Challenge Unit, 2014 Page 37 of 38

This suggests that law is less diverse in terms of disability than average, and we may want to explore reasons for this. Degree attainment for those UK students with a disability (66% 1 st or 2:1) is slightly lower than for those without a disability (68% 1 st or 2:1), but there is nowhere near the type of disparity present when looking at attainment by ethnicity. Contributing factors UK students in general appear to be disclosing their disability status more than students taking the GDL and LPC. It may be that there are lessons we can learn from how universities and UCAS collect disability data to improve our own collection rates. Page 38 of 38