The unintended consequences of ethnic pay gap reporting
25 January 2019
SUGGESTED
Markets and Morality
Energy and Environment
Economic Theory
25 June 2024
The Government’s consultation on ethnic pay gap monitoring closed a couple of weeks ago. We await the findings, but it looks pretty certain, given the submissions I have seen, that we will soon see compulsory reporting.
Many of these submissions by businesses and representative bodies support the proposal by Baroness McGregor-Smith, in her 2017 review of Race in the Workplace (an odd title, incidentally, given that the issue is ethnicity, not the widely-discredited concept of race), that all organisations with over 50 employees should be obliged to publish their ethnic pay gap and have a policy in place to show how they are promoting inclusion and equality. Before a company could be considered for a slice of the £100 billion or so of public procurement, its policy and progress towards narrowing the pay gap would have to be approved.
As I have argued before, I think that reporting on pay gaps needs a much more sensitive approach than publishing raw numbers. Pay differentials reflect a host of factors such as experience, qualifications, hours worked, seniority, responsibility and so on. In order to home in on sources of disadvantage (not the same thing as discrimination, incidentally) we need much more information about individuals and their background, of the kind we can get from national datasets such as the Labour Force Survey rather than payroll data from firms.
The problem is that raw differentials at the level of small organisations (which in most cases will only have a handful of ethnic minority employees) conceal more than they reveal. They are all too easily misinterpreted to fit a political narrative of ‘naming and shaming’. This in turn may lead firms to attempt to manipulate the published figures to cast themselves in a better light. In some cases, such manipulations may damage employees – for example when firms outsource poorly-paid work in which ethnic minorities are over-represented.
Some of the problems of ethnic pay reporting are brought home by a recently-published audit of a very large ‘firm’ – the Metropolitan Police.
The audit apparently shows an average difference in pay between White and BAME officers of £1.80 per hour, and for all employees (officers and staff) a mean pay gap of 9.7%. The Met has received a lot of negative publicity as a consequence, particularly as this gap seems to have marginally widened since the last time it was calculated. But does this tell us anything useful at all?
For a start, the categories. ‘BAME’ means ‘Black, Asian and minority ethnic’. But this is easily – and misleadingly – elided to ‘Black’. The Guardian did this in its first report, headlined ‘Black police paid less…’, and was forced to issue a retraction. This focuses on a sub-set of the grouping which is not representative of the category as a whole.
Even when spelt out, the category is so broad as to be meaningless. There are probably approaching 100 distinct minority ethnicities in London. Their experience is very different. Some groups, notably those of Chinese or Indian (Hindu) heritage, are known to do as well as, if not better than, White British. Others – Bangladeshi, Pakistani, Somali – do far worse. There is an important distinction, too, between first generation immigrants and second (and subsequent) generations; the latter fare much better, because of better education and language skills. Then of course there is the distinction between men and women, neglected in this analysis. Black Caribbean women do better in the labour market than White British women on some measures.
Nor is the ‘White’ category as homogenous as it may seem. The Met’s figures lump together disparate groups on the basis simply of skin colour. It is known that those in the ONS category ‘Gypsy or Irish Traveller’ (how has this description survived PC scrutiny?) and Poles and other Eastern Europeans do not do as well as some other ‘White’ groups. There are also big regional differences in average pay. There are substantial numbers of ‘poor whites’ in this country, neglected in the simplistic comparisons favoured by our would-be social engineers.
Another problem with using these crude indicators when dealing with such complexity is that changes over time may produce misleading information. In the Met’s case, they stress that police pay is rigorously determined by roles, and those in the same posts with the same experience, working the same hours and so on will be paid the same. But the overall pay gap has slightly worsened because of their recruitment efforts. The Met has made commendable efforts in trying to recruit a larger proportion of BAME officers and staff. But of course the bulk of new entrants, whatever their ethnicity, start at the bottom of the pay hierarchy, and thus an increased proportion of BAME entrants increases the pay gap.
The Met is used to criticism. But subjecting smallish businesses to the same scrutiny will produce pay gap results which then may generate perverse behaviour such as reducing job openings for lower-skilled workers to improve the indicator.
As I say, we are likely to face compulsory ethnic monitoring of our businesses (some would like this to go down as far as those with as few as ten employees). This may cause some improvements in company recruitment and treatment of staff, but it is unlikely to make fundamental changes in measured pay inequality, which is often rooted in factors beyond the powers of businesses, or government, to eliminate – even if this were desirable.
The danger is that by focusing on some groups rather than others, simplistic policies may exacerbate rather than reduce inequality. It may also, sadly, tend to heighten animosity between these different groups and contribute to the ever-growing disparagement of private enterprise and the wider institutions of a free society.
Many of these submissions by businesses and representative bodies support the proposal by Baroness McGregor-Smith, in her 2017 review of Race in the Workplace (an odd title, incidentally, given that the issue is ethnicity, not the widely-discredited concept of race), that all organisations with over 50 employees should be obliged to publish their ethnic pay gap and have a policy in place to show how they are promoting inclusion and equality. Before a company could be considered for a slice of the £100 billion or so of public procurement, its policy and progress towards narrowing the pay gap would have to be approved.
As I have argued before, I think that reporting on pay gaps needs a much more sensitive approach than publishing raw numbers. Pay differentials reflect a host of factors such as experience, qualifications, hours worked, seniority, responsibility and so on. In order to home in on sources of disadvantage (not the same thing as discrimination, incidentally) we need much more information about individuals and their background, of the kind we can get from national datasets such as the Labour Force Survey rather than payroll data from firms.
The problem is that raw differentials at the level of small organisations (which in most cases will only have a handful of ethnic minority employees) conceal more than they reveal. They are all too easily misinterpreted to fit a political narrative of ‘naming and shaming’. This in turn may lead firms to attempt to manipulate the published figures to cast themselves in a better light. In some cases, such manipulations may damage employees – for example when firms outsource poorly-paid work in which ethnic minorities are over-represented.
Some of the problems of ethnic pay reporting are brought home by a recently-published audit of a very large ‘firm’ – the Metropolitan Police.
The audit apparently shows an average difference in pay between White and BAME officers of £1.80 per hour, and for all employees (officers and staff) a mean pay gap of 9.7%. The Met has received a lot of negative publicity as a consequence, particularly as this gap seems to have marginally widened since the last time it was calculated. But does this tell us anything useful at all?
For a start, the categories. ‘BAME’ means ‘Black, Asian and minority ethnic’. But this is easily – and misleadingly – elided to ‘Black’. The Guardian did this in its first report, headlined ‘Black police paid less…’, and was forced to issue a retraction. This focuses on a sub-set of the grouping which is not representative of the category as a whole.
Even when spelt out, the category is so broad as to be meaningless. There are probably approaching 100 distinct minority ethnicities in London. Their experience is very different. Some groups, notably those of Chinese or Indian (Hindu) heritage, are known to do as well as, if not better than, White British. Others – Bangladeshi, Pakistani, Somali – do far worse. There is an important distinction, too, between first generation immigrants and second (and subsequent) generations; the latter fare much better, because of better education and language skills. Then of course there is the distinction between men and women, neglected in this analysis. Black Caribbean women do better in the labour market than White British women on some measures.
Nor is the ‘White’ category as homogenous as it may seem. The Met’s figures lump together disparate groups on the basis simply of skin colour. It is known that those in the ONS category ‘Gypsy or Irish Traveller’ (how has this description survived PC scrutiny?) and Poles and other Eastern Europeans do not do as well as some other ‘White’ groups. There are also big regional differences in average pay. There are substantial numbers of ‘poor whites’ in this country, neglected in the simplistic comparisons favoured by our would-be social engineers.
Another problem with using these crude indicators when dealing with such complexity is that changes over time may produce misleading information. In the Met’s case, they stress that police pay is rigorously determined by roles, and those in the same posts with the same experience, working the same hours and so on will be paid the same. But the overall pay gap has slightly worsened because of their recruitment efforts. The Met has made commendable efforts in trying to recruit a larger proportion of BAME officers and staff. But of course the bulk of new entrants, whatever their ethnicity, start at the bottom of the pay hierarchy, and thus an increased proportion of BAME entrants increases the pay gap.
The Met is used to criticism. But subjecting smallish businesses to the same scrutiny will produce pay gap results which then may generate perverse behaviour such as reducing job openings for lower-skilled workers to improve the indicator.
As I say, we are likely to face compulsory ethnic monitoring of our businesses (some would like this to go down as far as those with as few as ten employees). This may cause some improvements in company recruitment and treatment of staff, but it is unlikely to make fundamental changes in measured pay inequality, which is often rooted in factors beyond the powers of businesses, or government, to eliminate – even if this were desirable.
The danger is that by focusing on some groups rather than others, simplistic policies may exacerbate rather than reduce inequality. It may also, sadly, tend to heighten animosity between these different groups and contribute to the ever-growing disparagement of private enterprise and the wider institutions of a free society.
SHARE
