Why are Ethnic Minority Teacher Paid Differently in England: A Preliminary Analysis of the School Workforce Census

Why are Ethnic Minority Teacher Paid Differently in England: A Preliminary Analysis of the School Workforce Census

Stephen Gorard, Yiyang Gao, Nadia Siddiqui, and Beng Huat See

Durham University Evidence Centre for Education

DOI: https://dx.doi.org/10.47772/IJRISS.2024.803318S

Received: 20 September 2024; Accepted: 25 September 2024; Published: 23 October 2024

ABSTRACT

This paper uses data from the School Workforce Census in England, 2015-2022, to examine differential pay rates for teachers of different ethnic groups. Part of the reason for this is to help understand why ethnic minority teachers are under-represented in the teacher workforce. There are surface differences between the pay of different groups. Black African and Black Caribbean teachers tend to be paid more than average, and Pakistani teachers tend to be paid less. But there are also differences in age, qualifications, region, and phase of education. Taking all of the background and context variables into account, there is no evidence that ethnicity is a clear factor in determining pay. Of course, the reasons why teachers of different ethnicity teach in particular phases and regions may be part of that explanation. However, we will have to look elsewhere to explain the disproportion between ethnic minority teachers and pupils in England, and for the most effective policy levers to increase the supply and retention of minority ethnic teachers. Policymakers and schools could still consider salary incentives to try and attract more minority ethnic undergraduate students to a teaching career, where recruitment and retention of minority ethnic teachers is a challenge.

INTRODUCTION

A well-paid teaching workforce can be an important factor in teacher supply and long-term retention. Existing evidence has shown that raising teacher salaries can attract more skilled individuals in the profession (Chelwa et al. 2019). However, disproportionate salary differences on the basis of teacher characteristics such as their gender, ethnicity, language, country of origin, and route to teaching qualifications can make teaching profession less attractive and reduce retention (D’Amico et al. 2017, Akiba et al. 2012).

According to teacher workforce data in England 21% teachers identified themselves as from minority ethnic backgrounds (DfE 2024), and this has been increasing over time (Kelly 2024). Teachers from minority ethnic groups are underrepresented in leadership positions such as heads, deputy heads, and assistant heads. In 2023/24, 15.9% of White British teachers held leadership roles, compared to 11.1% of White minorities, 10.0% of Black or Black British, and 8.3% of Asian or Asian British teachers (DfE 2023). The disparities in teachers’ salaries may be attributed to bottlenecks in the workforce pipeline, where minority ethnic teachers encounter obstacles in obtaining promotions or deployments to higher-salary leadership roles (McNamara et al. 2009). What about the pay of classroom teachers?

The analysis presented here is part of a larger project looking at the recruitment and retention of minority ethnic teachers in England, involving three systematic reviews of prior evidence, a national survey of teachers, interviews, case studies, and secondary analyses of the School Workforce Census, National Pupil Database, and UCAS and TALIS figures on the recruitment of teachers.

We have shown that minority ethnic teachers in England are far less common as a proportion than minority ethnic pupils are (Gorard et al. 2023a). And that this disproportion is related to the recruitment and retention of minority ethnic teachers (See et al. 2024), and to the treatment, behaviour and outcomes of minority ethnic pupils (Gorard et al. 2003b). This disproportion matters. Therefore, we are exploring why the disproportion occurs, and how the recruitment and retention of ethnic minority teachers could be improved. One relevant issue could be differential pay – if ethnic minority teachers are paid less than their peers, ceteris paribus, this could be a factor in their under-representation.

This paper looks at patterns of differential pay for minority ethnic teachers. Following this introduction, the paper includes a summary of the methods used, comparative findings of pay by ethnicity and other factors, and a logistic regression analysis explaining high and low pay in terms of teacher background characteristics and qualifications. The paper ends with a brief discussion of the implications and next steps.

METHODS USED

The dataset used in this paper is the Schools Workforce Census for England, 2015-2022, containing individual records of teachers and other educational staff. It contain information on training and qualifications, contracts, role, phase, promotion, pay, and background characteristics such as region, sex and ethnicity. This paper focuses on teacher pay by ethnic group, and potential explanatory factors for any differences.

This is achieved through comparisons of means, and cross-tabulations of frequencies, and a logistic regression model using all available variables to explain whether teachers have relatively high or low pay. In this way, we can begin to assess whether any ethnic groups are high or low paid once other factors are accounted for. For this regression analysis only, missing cases for age and months since achieving qualified teacher status (QTS) are replaced with the overall mean for that variable.

The first degree qualification, the first QTS qualification, and Masters or PhD qualification, are retained for each candidate (some had 33 or more qualifications listed). The qualification codes are converted to broad subject areas for each qualification level. NQF unspecified qualifications are listed as first degree equivalent where no other degree is listed. The eight subject areas are:

• Art including Media Studies, crafts, joining
• English including Welsh literature, creative writing
• Services includingYouth Work, Careers
• Social sciences including Politics, Policy, Social Work
• Engineering including Design, Technology, Ceramics, Textiles
• Agriculture including Land management, Forestry, Food Production
• Health including Veterinary
• MFL and culture including Asian studies

COMPARATIVE RESULTS

This analysis is based on full-time classroom teachers in England, with a particular focus on the pay and ethnicity of teachers. There are no suitable figures for 2019, but we have retained the empty column for clarity. There are just over 300,000 cases in the dataset in each year (Table 1). There are too few teachers from a Traveller/Roma background to report analyses on (this is a stipulation of the data owner). Travellers have consistently low average pay.

Table 1 – Number of full-time classroom teachers in England, 2015-2022

Table 2 shows the actual average salary for each minority ethnic group, in pounds Sterling (£). This gives a sense of salary scale, growth over time, and differences between groups. As the clear majority, the pay of White British teachers is roughly equivalent to the overall national figure, but slightly below average in every year except 2020. There is no sign of any particular trend over time. The lowest pay is consistently for Pakistani teachers, below the national average in every year but with no particular sign of a trend. Black African and Black Caribbean teachers clearly have clearly had the highest pay in every year.

Table 2 – Base pay by ethnicity of teachers, 2015-2022

Overall, the standard deviation of base pay is £8,626

Future tables will mostly not present these actual salary figures. Instead, they will show the differences between the pay for each sub-group and year, compared to the overall pay for that year, and divided by the overall standard deviation of pay. This is a standard “effect” size. Table 3 is an example, and it shows the same data as Table 2. It is clearer to see the main differences and trends. Teachers of Pakistani origin, for example, always have considerably lower than average pay, with perhaps a slight improvement over time. Teachers of Black Caribbean origin have pay which is substantially higher than average each year (“effect” size of around +0.3).

Table 3 – Effect sizes for differences in base pay by ethnicity of teachers, 2015-2022

Compared to overall pay each year

There will be several reasons for these differences in teacher pay by ethnicity, including regional geography (Table 4). Teachers in London, for example, will tend to have higher pay and/or a London allowance, although it is not entirely clear that this “incentive” to teach in London is needed or justified given that this is not the region with greatest teacher shortages (Gorard 2018). This higher pay is represented in base pay, gross pay, and additional payments.

Table 4 – Effect sizes for differences in base pay by economic region, 2015-2022

Only two areas in England have above average pay for teachers (Table 5). Teachers in Inner London have pay which is a massive +0.57 of a standard deviation above average, and this difference has been growing over time. Teachers in Outer London have pay 0.36 of a standard deviation above average, again increasing over time. If teachers in other regions had the same level of pay as in London then there would likely be no shortages in the schools there. The idea of using financial incentives appears reasonable to encourage teachers to work in hard-to-staff or disadvantaged schools (See et al. 2020), but it is nowhere near as clear that extra payments should be given to teachers to teach in what are already the most popular areas to live in. Weighting teacher pay towards London in this way, and the damage it causes to teacher supply elsewhere, is a national policy, although it is not clear that policy-makers are aware of it, or that it has been properly debated.

The lowest pay is in Yorkshire and the Humber, and this has been relatively static over time, along with the East of England, and the South East, where pay has decreased relative to other areas over time. These regional differences, coupled with residential figures for ethnic minority teachers in England (Gorard et al. 2023), can help explain at least part of the differences in pay by ethnic group. Black African and Black Caribbean teachers live disproportionately in London, compared to White British, Chinese, Indian and Pakistani. Bangladeshi teachers are also disproportionately in Inner London.

Low pay in the South East may somehow be linked to the much higher and growing pay in London, but it clearly shows that regional pay is not directly linked to local house prices. The South East and the East of England have the highest house prices in England outside of London, and so if pay were related to house prices then we would expect to see these two regions have somewhat higher than average pay. They do not. In fact they have substantially lower than average pay for England, and so London remains a highly-paid anomaly (Table 5).

Table 5 – House prices by region of England, January 2023

Male teachers are paid slightly more than female teachers, and the gap has increased slightly over time, rather than reducing. This will be examined further is later analyses. Additionally, there are more male teachers who are Black African (40.4%) than in other ethnic groups (overall 28.2%). This could also explain part of the difference in pay by ethnicity, but does not apply to Black Caribbean teachers (24.1% male); nor can it explain the low pay of Pakistani teachers (26.6% male).

Understandably, whether a teacher has qualified teacher status (QTS) is related to their pay (Table 6). The gap between teachers with QTS and those without has remained quite similar over time. As with all of the characteristics considered so far, if QTS is linked to ethnicity then this could help explain at least part of the pay gap.

Table 6 – Effect sizes for differences in base pay by QTS status, 2015-2022

All figures are similar, but the percentage of teachers with QTS is somewhat lower for Black Caribbean teachers, which does not help to explain their higher average pay (Table 7).

Table 7 – Percentages for QTS by ethnicity, all years combined

In England, there is a range of ways in which teachers are trained initially, including traditional university routes, school-based development, and recognition of overseas training. The lowest paid teachers, in terms of how they got their qualified teacher status, are those who started on Teach First, or who trained via a School Direct or School Direct salaried route (Table 8). Teachers from all three of these “on the job” training routes have been catching up with average pay over time, but are still paid substantially below average. Teachers trained overseas, or in Northern Ireland, are paid substantially more than average. And this higher payment has increased hugely over time. Those qualifying via a PGCE (postgraduate) have been consistently paid slightly above average in every year.

Table 8 – Base salary by QTS route, 2015-2022

Again, this difference in pay by qualifying routes may be a factor in the pay differences by ethnic group. The route for most teachers is not recorded. Of the rest, ethnic minority groups are more likely than White British teachers to have been trained overseas.

The phase of education may also matter. Teachers in secondary, and middle-deemed secondary, schools consistently earn more than primary and middle-deemed primary school teachers. Teachers in sixth-from, college and other 16 plus sites earn more again, although the long-term trend is downwards (Table 9). Ethnic minority teachers are more common in secondary than in primary schools, whereas White British teachers are evenly balanced between these phases. Again, this can partly explain differences in pay, as secondary teachers tend to earn more.

Table 9 – Base pay by school phase, 2015-2022

Teachers in special schools but not PRUs (Pupil Referral Units) tend to be paid less than average. Many of the annual figures are volatile where there are few schools, or a new school type is introduced (Table 10). It is clear that secondary teachers are paid more than primary teachers, with teachers in Free schools 16-19 and CTCs paid the most. Teachers in voluntary-aided schools are paid more than average, and those in voluntary-controlled schools receive less than average. There are no clear differences in employment of ethnic groups by different school types, and so that is unlikely to help explain their differential pay.

Table 10 – Base salary by school type, 2015-2022

There is a small positive correlation between the base pay and age of teachers, how long ago they attained QTS (if they did) and how recent the year is for that pay (Table 11). This makes sense, because of pay for experience and incremental pay scales.

Table 11 – Correlation between base pay, and age, months since QTS, and census year – all years combined

Black African and Black Caribbean teachers are slightly older on average, and this may help explain their higher pay (Table 12). Pakistani teachers are among the youngest and this may help explain their lower pay.

Table 12 – Mean age and months since QTS by ethnicity, all years combined

Note: the figure for Travellers is combined over all years of data, and so is greater than the minimum threshold of 10 cases.

Logistic regression model

A lot of factors have been considered so far, as possible explanations for differential pay by ethnic groups. To collate and simplify the picture. logistic regression is used to look collectively at all of the factors presented above, that may influence teacher pay. Linear regression would require a large number of dummy variables, because most of the variables used above are categorical. A binary outcome was created representing whether base pay was below £34,550 or not. This yielded 50% of cases below, and 50% at or above, that pay.

The possible predictors were entered in order of biographical appearance (e.g. from age and sex to current school type), mixed with their strength as predictors. Within each group, each predictor is entered separately at first to assess their quality as predictors. In the first group, age of teacher was the best single predictor raising the percentage of pay predicted correctly from 50% to 72.3%, followed by months since QTS (7.17%) which would be correlated with age, then Census year (57%) to account for pay inflation, and the sex of the teacher (52.7%). Together they raise the percentage predicted correctly to 82%. The QTS status of the teacher, and their QTS route make a small further difference. Once the background characteristics and QTS status of teachers are taken into account, there is no role for school type or individual teacher ethnicity (Table 13).

Table 13 – Predictors of teacher pay, in logistic regression

This suggests that ethnicity is not in itself a factor in differential pay by ethnicity, including the higher average pay for Black African and Black Caribbean teachers, and the lower average pay for Pakistani teachers. Age and prior qualification can explain all of the difference that can be explained by these variables (Table 14).

Table 14 – Predictors of teacher pay, coefficients from logistic regression

If ethnicity is considered alone, then it raises the percentage predicted correctly from 50% to 51.1% (i.e. it is less relevant than the sex of teacher). If ethnicity is considered as part of the first group of background predictors, but last in the group because it is the weakest predictor, it raises the percentage predicted correctly from 82.0% to 82.1%%. Ethnicity itself does not appear to be a major factor in the differential pay of teachers. Rather it is the differential age and experience of different ethnic groups that explains their pay.

DISCUSSION

There is a clear disproportion between the number of minority ethnic teachers and pupils in England. This shortage of minority teachers can have serious implications for recruitment and retention of teachers, and for the treatment of, and outcomes for, minority ethnic pupils. This paper has used a large-scale dataset to look at whether minority ethnic teachers are paid less than their peers, once other factors are taken into account. In general, they are not. In fact, Black classroom teachers are the most highly paid. The results suggest that any differences in pay can be explained in terms of differences in age, qualifications, phase of schooling and economic region. Differences in age, qualifications, phase of schooling and region could still be linked to ethnicity. Pay incentives could still be used to try and attract more ethnic minority students to a teaching career, especially outside London and the West Midlands. Our project will examine both of these ideas further. However, we will also examine other possible explanations, via structured reviews of evidence, a national survey, interviews and case studies with teachers.

REFERENCES

1. Akiba, M., Chiu, Y. L., Shimizu, K., & Liang, G. (2012). Teacher salary and national achievement: A cross-national analysis of 30 countries. International Journal of Educational Research, 53, 171-181.
2. Chelwa, G., Pellicer, M., & Maboshe, M. (2019). Teacher salary and educational outcomes: Evidence from the rural hardship allowance in Zambia. South African Journal of Economics, 87(3), 255-282.
3. D’Amico, D., Pawlewicz, R. J., Earley, P. M., & McGeehan, A. P. (2017). Where Are All the Black Teachers? Discrimination in the Teacher Labor Market. Harvard Educational Review, 87(1), 26-49,155-156. http://ezphost.dur.ac.uk/login?url=https://www.proquest.com/scholarly-journals/where-are-all-black-teachers-discrimination/docview/1883114075/se-2
4. Department for Education DfE (2023). School leadership in England 2010 to 2020: characteristics and trends. School leadership in England 2010 to 2020: characteristics and trends – GOV.UK (www.gov.uk)
5. Department for Education DfE (2024). School Workforce in England: Reporting year 2023. School workforce in England, Reporting year 2023 – Explore education statistics – GOV.UK (explore-education-statistics.service.gov.uk)
6. Gorard, S. (2018) Education policy: Evidence of equity and effectiveness, Bristol: Policy Press
7. Gorard, S., Chen, W., Tan, Y., Gazmuri, C., See, BH, Tereshchenko, A, Demie, F. and Siddiqui, N. (2023a) The disproportionality of ethnic minority teachers in England: trends, patterns, and problems, Routledge Open Research, The disproportionality of ethnic minority … | Routledge Open Research
8. Gorard, S., Siddiqui, N., See, B., Tereschenko, A. and Demie, F. (2023b) Ethnic proportionality of teachers and students, and the link to school-level outcomes, Education Sciences, 13, 8, 838, Education Sciences | Free Full-Text | The Ethnic Proportionality of Teachers and Students and the Link to School-Level Outcomes (mdpi.com)
9. Kelly, A. (2024). Representativeness and diversity within the teaching profession in England, 2010–2020. International Journal of Educational Development, 108, 103067. https://doi.org/10.1016/j.ijedudev.2024.103067
10. McNamara, O., Howson, J., Gunter, H. and Fryers, A. (2009). The leadership aspirations and careers of black and minority ethnic teachers. Birmingham: NASUWT. Leadership aspirations and careers of black and minority ethnic teachers (equitableeducation.co.uk)
11. See, BH, Gorard, S., Morris, R. and el-Soufi, N. (2020) What works in attracting and retaining teachers in challenging schools and areas?, Oxford Review of Education,  https://doi.org/10.1080/03054985.2020.1775566