Quantifying the ethnic penalty

Matt Ford

It's common to hear people talk about how we now live in a ‘post-racial’ society, where merit and effort determine the extent to which people are able to meet their basic needs and achieve their potential. Indeed, this argument is often invoked to counter proponents of positive discrimination. Here I am going to use data to show that it isn't true.

I am going to map out the injustices black people, and where possible young black men, face in a number of areas of their lives. I'm calling these injustices ‘ethnic penalties’ to denote where black people fare less well than similarly placed white people, due directly or indirectly to racism at an individual and institutional level. Using the term ‘ethnic penalties’ is simply because it is a very precise way of identifying the effects of racism across a broad range of measurable outcomes. Some of the data, such as the poverty data I have used, doesn't pinpoint this exactly. But I am making the reasonable assumption, given the other evidence, that it represents the cumulative effects of a range of ethnic penalties, as well as incorporating the effects of other factors such as class.

Excluded from education

I am sure we all agree about the importance of education in self-determination, so excluding certain groups from a good education is hugely detrimental to their well-being. Figure 1 shows the proportion of white British, black African, black Caribbean and Other black pupils achieving five GCSEs grade A* to C including English and maths in 2008/2009 and 2012/2013 (Department for Education, 2014). There has been huge progress in this measure of attainment across the board over a very short period of time. Greater proportions of children from each group are now getting these results, and black African children now outperform white British children. But black Caribbean and Other black children still fall behind.

Steve Strand's 2012 article about his research around ‘tiering’ may be able to throw some light on why this achievement gap exists. His study tried to explain some of the difference in attainment between white British and black Caribbean children in maths and science tests at age 14 by examining patterns in the way different groups were entered into different exam tiers.

At age 14, teachers place children into different levels or ‘tiers’ when preparing for maths and science exams, and the tier you go into affects what mark you can get in the end. You can only achieve the highest mark if you're placed in the highest tier, but if you're placed in the highest tier and you fail to get the lowest expected grade in that exam, you are given an unclassified grade. The teacher uses their professional judgement to decide which tier to place a child in. The decision will be influenced by how well the teacher thinks the child will cope with the content and structure of the tests. Their perception of the child's ability to cope will be based on the child's prior attainment, as well as a range of other background factors, like the child's social class or their level of motivation.

Figure 2 shows the proportions of pupils entered into different tiers of science exams at age 14, and figure 3 shows the same thing for maths exams. You can see that smaller proportions of black Caribbean, Pakistani, Bangladeshi, black African and mixed heritage children are entered into the higher tier exams compared to white British children, and consequently higher proportions of these groups are entered into the lower tier exams.

But these inequalities don't necessarily identify an ethnic penalty or bias on the part of teachers, they may represent real differences between the different ethnic groups. For instance, black Caribbean, Pakistani, Bangladeshi, black African and mixed heritage children being entered into lower tiers at higher rates than white children could just reflect lower levels of attainment or higher levels of poverty among these groups.

Strand's study tried to find out how much of the ethnic inequality identified in the tiering process was down to things like different levels of attainment, and how much was down to bias on the part of teachers. Any of the difference between the proportions of children from different ethnic groups entered into different tiers that couldn't be explained by these background factors would imply that an ethnic penalty exists in the process.

What he found was that after controlling for a wide range of background characteristics, the levels of under-representation of different ethnic groups in the higher tier exams disappeared, and indeed some groups were actually over-represented in the higher tiers. So all other things being equal, black African children, say, will be entered into the higher tiers at exactly the same rate as white British children. However, this wasn't the case for black Caribbean children. They were the only ethnic group who were consistently under-represented in the higher tiers of maths and science exams, even after controlling for a range of background factors. So all other things being equal, for every three white British children entered into higher tier maths and science exams, only two black Caribbean children were. Clearly merit and effort are not enough for black Caribbean children to have the same opportunities as children from other ethnic groups. The conclusion Strand draws is that teachers' perceive black Caribbean children as more behaviourally problematic and this distorts judgements about their academic ability.

This is borne out by government data on school exclusions. You can see in figure 5 that greater proportions of children from most black backgrounds get permanently excluded compared to other ethnic groups. About 0.22 per cent of black Caribbean children get permanently excluded compared to only 0.07 per cent of white children.

There are debates about whether black Caribbean children are actually more behaviourally problematic; whether this is just a pervasive stereotype which leads to discriminatory punishment; or whether it is both of these things reinforcing each other in a vicious cycle. But the report, Getting it, Getting it Right, found that black pupils were punished more severely, more often, and for less serious transgressions than white pupils (Department for Education and Skills, 2006).

Similar ethnic inequalities exist in post-16 education. Figure 6 shows the proportions of different ethnic groups achieving grades AAB or above at A level, of which at least two are in traditional subjects (Department for Education, 2014). Lower proportions of black students achieve this level of attainment than Asian, white, Other, mixed ethnicity and Chinese students. The percentage of black students achieving grades AAB or above is only around half that of most other ethnic groups. 

Black people are also more likely to pursue vocational qualifications rather than A Levels, with 65 per cent of black students pursuing qualifications other than A Levels, compared to 35 per cent across all ethnic groups. Both of these factors are related to the type of further education institution attended, with black people more likely to go to further education colleges than school sixth forms.

Having low A Level results, pursuing vocational qualifications rather than A levels, and attending further education colleges rather than school sixth forms affects people's higher education career in two ways. Firstly, higher tariff universities are biased towards students applying with A Levels directly from school, and so black students tend to get concentrated in newer, post-1992 universities. Secondly, people with A Levels tend to gain higher degree classifications than people with vocational qualifications, so black people end up with lower overall attainment at university as well. As you can see in figure 7, greater proportions of UK-domiciled black Caribbean and black African students get 2.2s and thirds than get 2.1s and firsts (Equality Challenge Unit, 2014).

Figure 8 shows university drop-out rates for UK-domiciled people from different ethnic groups. Black people have the highest drop-out rates of all the ethnic groups, around double that of white people (Equality Challenge Unit, 2014). The point is that the education system as a whole seems to reproduce and exacerbate inequalities between ethnic groups. 

Some of the penalties incurred in education will spill over into employment, but we can identify where the labour market specifically works against black people. I will start by going through some general data where this distinction can't be made, and then move on to some research which pinpoints actual ethnic penalties.

Education isn’t enough

Figure 9 shows the unemployment rate for males aged 16-24 from different ethnic groups from 2002 to 2014 (Department for Work and Pensions, 2015). You can see that young black men consistently have the highest rate of unemployment of any ethnic group, except for in 2013 when there was a massive increase in unemployment for young Pakistani and Bangladeshi men. 

Figure 10 shows unemployment rates for UK-domiciled people from different ethnic groups six months after graduation (Equality Challenge Unit, 2014). New black African graduates have the highest rates, with about 13 per cent of this group being unemployed six months after they complete their degree. Black Caribbean graduates fare relatively well with only around nine per cent being unemployed six months after graduation. But all ethnic minorities are more likely than white people to be unemployed just after qualifying by a significant margin.

But, as with tiering in the education system, some of the differences in levels of unemployment can be attributed to factors like lower overall attainment at university for black people. Research by Rafferty (2012) tried to eliminate the effects of a range of background factors, like degree class, gender and socio-economic background, to see if there was an ethnic penalty in graduate level employment.

Rafferty grouped together people with graduate level qualifications who had similar background characteristics, and then split these groups of similar individuals by their ethnicity. So, for example, he matched all graduates that were men, had a similar degree class in a similar subject, had similar class background and so forth, and then separated these groups by ethnicity. He ended up with a group of white men with similar background characteristics, and a group exactly the same but of black men, and so on for a whole range of different characteristics. He could then compare the employment outcomes of matching black and white groups.

Given the similarities between two matched groups, you would expect that, if no ethnic penalty existed, the black group would have similar employment outcomes to the white group. But instead, what Rafferty found was that groups of black people, who were exactly the same as groups of white people, had higher levels of unemployment, greater proportions of graduates in non-graduate occupations, and lower hourly wage rates.

Figures 11, 12, and 13 show the results of Rafferty's analysis. Figure 11 shows over-representation of UK-born black African, black Caribbean and Pakistani/Bangladeshi graduates in unemployment using the employment levels of white British people as the baseline. The way to read it is that ten per cent of black African men and eight per cent of black African women with graduate level qualifications are unemployed who wouldn't be if they were white, and so on for the other ethnic groups.

Figure 12 shows the over-representation of UK-born black African, black Caribbean and Pakistani/Bangladeshi graduates in non-graduate occupations. The way to interpret it is that 12 per cent of black African men and 18 per cent of black African women with graduate level qualifications are in non-graduate occupations who would be in graduate occupations if they were white (the results for black Caribbean men in this case are not statistically significant).

Finally, in terms of wage penalties, on average UK-born black men with graduate level qualifications earn £3.86 an hour less than similarly placed white men. UK-born black women with graduate level qualifications earn £2.08 an hour less than similarly placed white women.

Kept below the poverty line

Obviously these labour market disadvantages make a significant contribution to the nature and extent of poverty amongst black and Asian people. Income poverty is basically a measure of the proportion of people whose income falls below a certain threshold. This threshold is based on the average income of the whole population. 

Theoretically this means that the people who fall below the threshold don't have enough money to achieve the minimally acceptable standard of living in a given society. For example they won't be able to afford a balanced diet or buy suitable clothing. The recent rise in food-bank use is an example of what happens when lots of people are in income poverty for any length of time.

As you can see from the data in figure 14, black, Asian, other and mixed race people have much higher rates of income poverty than white people; about double the rate (Department for Work and Pensions, 2014). About 40 per cent of all black people in the UK are in income poverty. 

Figure 15 shows income poverty rates for children, working age adults and pensioners from different ethnic groups. Clearly children face the highest levels of poverty, but the ethnic stratification is replicated across age groups. 

The longer someone is in income poverty the worse the effects are likely to be. For example, if a household drops into low income they may be able to weather it out for a while using savings. The longer the household suffers from low income, the more these resources will be depleted, and so household members’ standard of living will begin to decline.

Figure 16 shows rates of income poverty persistence for different ethnic groups (Fisher and Nandi, 2015). It shows the proportion of people in different ethnic groups that were observed as being poor from zero to up to three times over a three year period. Overall you can see that white majority people were least likely to be observed as being poor over the three years. 53 per cent of black African people and 41 per cent of black Caribbean people were observed as being poor at least once over the three year period.

Paying for poverty

It is not unreasonable to believe that many black people will face the extra costs associated with having a low income. The ‘poverty premium’ describes the higher cost per unit poor people pay for certain goods and services due to limited choices.

The premium can arise when accessing credit, cash or insurance, and when paying for fuel, food or utilities. For example, people on low incomes are less likely to be able buy essential white goods outright and so spread the cost over a longer period. These payment plans can mean paying hundreds of per cent more for an item because of the added interest rates. The poverty premium is estimated to add as much as ten per cent more to a minimum household budget (Hirsch, 2013).

Since black people tend to be more likely to be in poverty than white people, they are more likely to have to pay more for essential goods and services. The poverty premium therefore creates a vicious cycle of poverty reproducing and intensifying poverty.

Excluded from financial services

Many black people face restricted access to essential/basic financial products and services (Khan, 2008). They can be excluded from banking, credit, insurance, savings, pensions and financial advice, among other things.

Financial exclusion can limit people's ability to take advantage of certain opportunities and have a negative impact on their quality of life. For instance, it can restrict people's employment options, make it difficult to cope with planned or unplanned changes in circumstances, and result in paying a premium for certain transactions.

Having a bank account is really important and it is the preferred method for paying income of various kinds. Black people are the group least likely to have a regular bank account, and 6.1 percent of black men don't have a bank account at all (Lawton and Platt, 2010).

Savings provide a safety net against planned or unplanned costs and falls in income. Without savings people are therefore more susceptible to declining living standards and debt when these costs arise. Around 55 per cent of black men have no formal savings, compared to around 35 per cent of white men (ibid). For black people who do have savings, the median value, at £2,000, is significantly lower than those of white people, who have average savings of £3,000. In fact, black people have the lowest value savings of any ethnic group.

Pensions are a form of savings which provide income for people when they retire. Out of men of all ethnic backgrounds, black men have the lowest proportion with a private pension (ibid).

Insurance protects people from unexpected costs, enabling them to avoid hardship and debt. Just over 55 per cent of black Caribbean households have home contents insurance, and astonishingly only 35 percent of other black households are covered. In contrast, approximately 80 per cent of white households have contents insurance.

Data on credit and ethnicity is difficult to come by, but we can make some tentative inferences. Given that black people tend to have lower incomes and higher unemployment rates, they are more likely to have limited access to credit. Money they are able to borrow is likely to be charged at significantly higher interest rates than ‘prime’ credit. Given that black people are more likely to need credit to cover the cost of living, this can cause further hardship.

Similarly, data on access to financial advice and ethnicity is sparse. Recent qualitative research found that because black and Asian people are more likely to be unemployed and be on a low income, they are more likely to need advice around debt and benefit issues (Mawhinney, 2010). Given that black people have less access to basic banking services they are more likely to be excluded from formal financial advice. Language barriers and distrust of banks were also identified as factors that may prevent some black and Asian people accessing financial advice.

Post-racial society?

It's clear from the evidence I have briefly sketched out here, that many black people face penalties in many areas of their lives. These penalties also seem to be cumulative, and they intersect with other factors to produce significant inequalities between ethnic groups. The sheer breadth of areas that the ethnic penalty appears to occur in, and I have by no means outlined them all here, suggests that rather than the ‘post-racial’ society that so many have been eager to herald, a person's ethnic background still has a considerable influence on the way people are able to live their lives.


Department for Education (2014), GCSE and equivalent attainment by pupil characteristics: 2013, National and local authority tables: SFR05/2014, London: Department for Education.

Department for Education (2014), Permanent and fixed-period exclusions in England: 2012 to 2013, National tables: SFR28/2014, London: Department for Education.

Department for Education and Skills (2006), Getting it. Getting it right, London: Department for Education and Skills.

Department for Work and Pensions (2014), Households below average income (HBAI): 1994/95 to 2012/13, London: Department for Work and Pensions.

Equality Challenge Unit (2014), Equality in higher education: statistical report 2014, part 2 students, London:Equality Challenge Unit.

Fisher, P., and Nandi, A. (2015), Poverty across ethnic groups through recession and austerity, York: Joseph Rowntree Foundation.

Freedom of Information Response (2014), Response from the Department for Education to Naomi Littlejohn, 24 September 2014. Case Reference: 2014/0033508.

Hirsch, D. (2013), Addressing the poverty premium, Approaches to regulation, London: Consumer Futures.

Khan, O. (2008), Financial Inclusion and ethnicity: An agenda for research and policy action, London: Runnymede Trust.

Lawton, K., and Platt, R. (2010), Review of access to essential services, Financial inclusion and utilities, London: Institute for Public Policy Research.

Mawhinney, P. (2010), Seeking sound advice, Financial inclusion and ethnicity, London: Runnymede Trust.

Rafferty, A. (2012), ‘Ethnic penalties in graduate level over-education, unemployment and wages: evidence from Britain’, Work, employment and society, Vol 26, No. 6 pp. 987-1006.

Strand, S. (2012), ‘The white British-black Caribbean achievement gap: tests, tiers and teacher expectations’, British Educational Research Journal, Vol. 38, No 1, pp. 75-101.